* [RFC v9 00/22] i386 cleanup
@ 2020-12-08 19:48 Claudio Fontana
2020-12-08 19:48 ` [RFC v9 01/32] accel/tcg: split CpusAccel into three TCG variants Claudio Fontana
` (34 more replies)
0 siblings, 35 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Peter Maydell, Eduardo Habkost, Paul Durrant,
Jason Wang, Marcelo Tosatti, qemu-devel, Peter Xu,
Dario Faggioli, Cameron Esfahani, haxm-team, Claudio Fontana,
Anthony Perard, Bruce Rogers, Olaf Hering, Emilio G . Cota,
Colin Xu
v8 -> v9: move additional methods to CPUClass->tcg_ops
do_unaligned_access, transaction_failed and do_interrupt.
do_interrupt is a bit tricky, as the same code is reused
(albeit not usually directly) for KVM under certain odd conditions.
Change arm, as the only user of do_interrupt callback for KVM,
to instead call the target function directly arm_do_interrupt.
v7 -> v8: add missing CONFIG_TCGs, fix bugs
* add the prerequisite patches for "3 tcg" at the beginning of the
series for convenience (already reviewed, queued by RH).
* add CONFIG_TCG to TCGCpuOperations and tcg_ops variable use
* reduce the scope of the realizefn refactoring, do not
introduce a separate cpu_accel_realize, and instead use the
existing cpu_exec_realizefn, there is not enough benefit
to introduce a new function.
* fix bugs in user mode due to attempt to move the tcg_region_init()
early, so it could be done just once in tcg_init() for both
softmmu and user mode. Unfortunately it needs to remain deferred
for user mode, as it needs to be done after prologue init and
after the GUEST_BASE has been set.
v6 -> v7: integrate TCGCpuOperations, refactored cpu_exec_realizefn
* integrate TCGCpuOperations (Eduardo)
Taken some refactoring from Eduardo for Tcg-only operations on
CPUClass.
* refactored cpu_exec_realizefn
The other main change is a refactoring of cpu_exec_realizefn,
directly linked to the effort of making many cpu_exec operations
TCG-only (Eduardo series above):
cpu_exec_realizefn is actually a TCG-only thing, with the
exception of a couple things that can be done in base cpu code.
This changes all targets realizefn, so I guess I have to Cc:
the Multiverse? (Universe was already CCed for all accelerators).
v5 -> v6: remove MODULE_INIT_ACCEL_CPU
instead, use a call to accel_init_interfaces().
* The class lookups are now general and performed in accel/
new AccelCPUClass for new archs are supported as new
ones appear in the class hierarchy, no need for stubs.
* Split the code a bit better
v4 -> v5: centralized and simplified initializations
I put in Cc: Emilio G. Cota, specifically because in patch 8
I (re)moved for user-mode the call to tcg_regions_init().
The call happens now inside the tcg AccelClass machine_init,
(so earlier). This seems to work fine, but thought to get the
author opinion on this.
Rebased on "tcg-cpus: split into 3 tcg variants" series
(queued by Richard), to avoid some code churn:
https://lists.gnu.org/archive/html/qemu-devel/2020-10/msg04356.html
* Extended AccelClass to user-mode.
user-mode now does not call tcg_exec_init directly,
instead it uses the tcg accel class, and its init_machine method.
Since user-mode does not define or use a machine state,
the machine is just passed as NULL.
The immediate advantage is that now we can call current_accel()
from both user mode and softmmu, so we can work out the correct
class to use for accelerator initializations.
* QOMification of CpusAccelOps
simple QOMification of CpusAccelOps abstract class.
* Centralized all accel_cpu_init, so only one per cpu-arch,
plus one for all accels will remain.
So we can expect accel_cpu_init() to be limited to:
softmmu/cpus.c - initializes the chosen softmmu accel ops for the cpus module.
target/ARCH/cpu.c - initializes the chosen arch-specific cpu accelerator.
These changes are meant to address concerns/issues (Paolo):
1) the use of if (tcg_enabled()) and similar in the module_init call path
2) the excessive number of accel_cpu_init() to hunt down in the codebase.
* Fixed wrong use of host_cpu_class_init (Eduardo)
v3 -> v4: QOMification of X86CPUAccelClass
In this version I basically QOMified X86CPUAccel, taking the
suggestions from Eduardo as the starting point,
but stopping just short of making it an actual QOM interface,
using a plain abstract class, and then subclasses for the
actual objects.
Initialization is still using the existing qemu initialization
framework (module_call_init), which is I still think is better
than the alternatives proposed, in the current state.
Possibly some improvements could be developed in the future here.
In this case, effort should be put in keeping things extendible,
in order not to be blocked once accelerators also become modules.
Motivation and higher level steps:
https://lists.gnu.org/archive/html/qemu-devel/2020-05/msg04628.html
Looking forward to your comments on this proposal,
Ciao,
Claudio
Claudio Fontana (23):
accel/tcg: split CpusAccel into three TCG variants
accel/tcg: split tcg_start_vcpu_thread
accel/tcg: rename tcg-cpus functions to match module name
i386: move kvm accel files into kvm/
i386: move whpx accel files into whpx/
i386: move hax accel files into hax/
i386: hvf: remove stale MAINTAINERS entry for old hvf stubs
i386: move TCG accel files into tcg/
i386: move cpu dump out of helper.c into cpu-dump.c
i386: move TCG cpu class initialization out of helper.c
target/riscv: remove CONFIG_TCG, as it is always TCG
accel/tcg: split TCG-only code from cpu_exec_realizefn
target/arm: do not use cc->do_interrupt for KVM directly
cpu: move cc->do_interrupt to tcg_ops
cpu: move cc->transaction_failed to tcg_ops
cpu: move do_unaligned_access to tcg_ops
accel: extend AccelState and AccelClass to user-mode
accel: replace struct CpusAccel with AccelOpsClass
accel: introduce AccelCPUClass extending CPUClass
i386: split cpu accelerators from cpu.c, using AccelCPUClass
cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn
hw/core/cpu: call qemu_init_vcpu in cpu_common_realizefn
cpu: introduce cpu_accel_instance_init
Eduardo Habkost (9):
tcg: cpu_exec_{enter,exit} helpers
tcg: make CPUClass.cpu_exec_* optional
tcg: Make CPUClass.debug_excp_handler optional
cpu: Remove unnecessary noop methods
cpu: Introduce TCGCpuOperations struct
cpu: Move synchronize_from_tb() to tcg_ops
cpu: Move cpu_exec_* to tcg_ops
cpu: Move tlb_fill to tcg_ops
cpu: Move debug_excp_handler to tcg_ops
MAINTAINERS | 19 +-
accel/accel-common.c | 105 +++++
accel/{accel.c => accel-softmmu.c} | 60 +--
accel/accel-softmmu.h | 15 +
accel/accel-user.c | 24 ++
accel/kvm/kvm-all.c | 2 -
accel/kvm/kvm-cpus.c | 26 +-
accel/kvm/kvm-cpus.h | 2 -
accel/meson.build | 4 +-
accel/qtest/qtest.c | 25 +-
accel/tcg/cpu-exec.c | 70 +++-
accel/tcg/cputlb.c | 6 +-
accel/tcg/meson.build | 9 +-
accel/tcg/tcg-all.c | 14 +-
accel/tcg/tcg-cpus-icount.c | 138 +++++++
accel/tcg/tcg-cpus-icount.h | 19 +
accel/tcg/tcg-cpus-mttcg.c | 134 +++++++
accel/tcg/tcg-cpus-mttcg.h | 19 +
accel/tcg/tcg-cpus-rr.c | 298 ++++++++++++++
accel/tcg/tcg-cpus-rr.h | 21 +
accel/tcg/tcg-cpus.c | 539 +++-----------------------
accel/tcg/tcg-cpus.h | 8 +-
accel/tcg/user-exec.c | 6 +-
accel/xen/xen-all.c | 24 +-
bsd-user/main.c | 11 +-
cpu.c | 71 ++--
hw/core/cpu.c | 30 +-
hw/i386/fw_cfg.c | 2 +-
hw/i386/intel_iommu.c | 2 +-
hw/i386/kvm/apic.c | 2 +-
hw/i386/kvm/clock.c | 2 +-
hw/i386/microvm.c | 2 +-
hw/i386/pc.c | 2 +-
hw/i386/pc_piix.c | 1 +
hw/i386/x86.c | 2 +-
hw/mips/jazz.c | 4 +-
include/hw/boards.h | 2 +-
include/hw/core/accel-cpu.h | 25 ++
include/hw/core/cpu.h | 92 ++---
include/hw/core/tcg-cpu-ops.h | 75 ++++
include/{sysemu => qemu}/accel.h | 16 +-
include/sysemu/accel-ops.h | 45 +++
include/sysemu/cpus.h | 26 +-
include/sysemu/hvf.h | 2 +-
include/sysemu/kvm.h | 2 +-
include/sysemu/kvm_int.h | 2 +-
linux-user/main.c | 7 +-
meson.build | 1 +
softmmu/cpus.c | 12 +-
softmmu/icount.c | 2 +-
softmmu/memory.c | 2 +-
softmmu/qtest.c | 2 +-
softmmu/vl.c | 8 +-
target/alpha/cpu.c | 18 +-
target/arm/cpu.c | 26 +-
target/arm/cpu64.c | 5 +-
target/arm/cpu_tcg.c | 8 +-
target/arm/helper.c | 4 +
target/arm/kvm64.c | 4 +-
target/avr/cpu.c | 13 +-
target/avr/helper.c | 4 +-
target/cris/cpu.c | 30 +-
target/cris/helper.c | 4 +-
target/hppa/cpu.c | 15 +-
target/i386/cpu-dump.c | 537 +++++++++++++++++++++++++
target/i386/cpu.c | 418 ++------------------
target/i386/cpu.h | 120 +-----
target/i386/{ => hax}/hax-all.c | 5 +-
target/i386/{ => hax}/hax-cpus.c | 29 +-
target/i386/{ => hax}/hax-cpus.h | 2 -
target/i386/{ => hax}/hax-i386.h | 6 +-
target/i386/{ => hax}/hax-interface.h | 0
target/i386/{ => hax}/hax-mem.c | 0
target/i386/{ => hax}/hax-posix.c | 0
target/i386/{ => hax}/hax-posix.h | 0
target/i386/{ => hax}/hax-windows.c | 0
target/i386/{ => hax}/hax-windows.h | 0
target/i386/hax/meson.build | 7 +
target/i386/helper-tcg.h | 112 ++++++
target/i386/helper.c | 539 +-------------------------
target/i386/host-cpu.c | 198 ++++++++++
target/i386/host-cpu.h | 19 +
target/i386/hvf/cpu.c | 65 ++++
target/i386/hvf/hvf-cpus.c | 27 +-
target/i386/hvf/hvf-cpus.h | 2 -
target/i386/hvf/hvf-i386.h | 2 +-
target/i386/hvf/hvf.c | 3 +-
target/i386/hvf/meson.build | 1 +
target/i386/hvf/x86_task.c | 2 +-
target/i386/kvm/cpu.c | 148 +++++++
target/i386/{ => kvm}/hyperv-proto.h | 0
target/i386/{ => kvm}/hyperv-stub.c | 0
target/i386/{ => kvm}/hyperv.c | 0
target/i386/{ => kvm}/hyperv.h | 0
target/i386/kvm/kvm-cpu.h | 41 ++
target/i386/{ => kvm}/kvm-stub.c | 0
target/i386/{ => kvm}/kvm.c | 3 +-
target/i386/{ => kvm}/kvm_i386.h | 0
target/i386/kvm/meson.build | 8 +
target/i386/kvm/trace-events | 7 +
target/i386/kvm/trace.h | 1 +
target/i386/machine.c | 4 +-
target/i386/meson.build | 39 +-
target/i386/{ => tcg}/bpt_helper.c | 1 +
target/i386/{ => tcg}/cc_helper.c | 1 +
target/i386/tcg/cpu.c | 173 +++++++++
target/i386/{ => tcg}/excp_helper.c | 1 +
target/i386/{ => tcg}/fpu_helper.c | 33 +-
target/i386/{ => tcg}/int_helper.c | 1 +
target/i386/{ => tcg}/mem_helper.c | 1 +
target/i386/tcg/meson.build | 14 +
target/i386/{ => tcg}/misc_helper.c | 1 +
target/i386/{ => tcg}/mpx_helper.c | 1 +
target/i386/{ => tcg}/seg_helper.c | 1 +
target/i386/{ => tcg}/smm_helper.c | 2 +
target/i386/{ => tcg}/svm_helper.c | 1 +
target/i386/{ => tcg}/tcg-stub.c | 0
target/i386/{ => tcg}/translate.c | 1 +
target/i386/trace-events | 6 -
target/i386/whpx/meson.build | 4 +
target/i386/{ => whpx}/whp-dispatch.h | 0
target/i386/{ => whpx}/whpx-all.c | 4 +-
target/i386/{ => whpx}/whpx-cpus.c | 29 +-
target/i386/{ => whpx}/whpx-cpus.h | 2 -
target/lm32/cpu.c | 13 +-
target/m68k/cpu.c | 12 +-
target/microblaze/cpu.c | 23 +-
target/mips/cpu.c | 21 +-
target/moxie/cpu.c | 10 +-
target/nios2/cpu.c | 14 +-
target/openrisc/cpu.c | 12 +-
target/ppc/translate_init.c.inc | 23 +-
target/riscv/cpu.c | 25 +-
target/rx/cpu.c | 18 +-
target/s390x/cpu.c | 15 +-
target/sh4/cpu.c | 14 +-
target/sparc/cpu.c | 18 +-
target/tilegx/cpu.c | 10 +-
target/tricore/cpu.c | 8 +-
target/unicore32/cpu.c | 14 +-
target/xtensa/cpu.c | 16 +-
141 files changed, 3007 insertions(+), 2051 deletions(-)
create mode 100644 accel/accel-common.c
rename accel/{accel.c => accel-softmmu.c} (64%)
create mode 100644 accel/accel-softmmu.h
create mode 100644 accel/accel-user.c
create mode 100644 accel/tcg/tcg-cpus-icount.c
create mode 100644 accel/tcg/tcg-cpus-icount.h
create mode 100644 accel/tcg/tcg-cpus-mttcg.c
create mode 100644 accel/tcg/tcg-cpus-mttcg.h
create mode 100644 accel/tcg/tcg-cpus-rr.c
create mode 100644 accel/tcg/tcg-cpus-rr.h
create mode 100644 include/hw/core/accel-cpu.h
create mode 100644 include/hw/core/tcg-cpu-ops.h
rename include/{sysemu => qemu}/accel.h (94%)
create mode 100644 include/sysemu/accel-ops.h
create mode 100644 target/i386/cpu-dump.c
rename target/i386/{ => hax}/hax-all.c (99%)
rename target/i386/{ => hax}/hax-cpus.c (71%)
rename target/i386/{ => hax}/hax-cpus.h (95%)
rename target/i386/{ => hax}/hax-i386.h (95%)
rename target/i386/{ => hax}/hax-interface.h (100%)
rename target/i386/{ => hax}/hax-mem.c (100%)
rename target/i386/{ => hax}/hax-posix.c (100%)
rename target/i386/{ => hax}/hax-posix.h (100%)
rename target/i386/{ => hax}/hax-windows.c (100%)
rename target/i386/{ => hax}/hax-windows.h (100%)
create mode 100644 target/i386/hax/meson.build
create mode 100644 target/i386/helper-tcg.h
create mode 100644 target/i386/host-cpu.c
create mode 100644 target/i386/host-cpu.h
create mode 100644 target/i386/hvf/cpu.c
create mode 100644 target/i386/kvm/cpu.c
rename target/i386/{ => kvm}/hyperv-proto.h (100%)
rename target/i386/{ => kvm}/hyperv-stub.c (100%)
rename target/i386/{ => kvm}/hyperv.c (100%)
rename target/i386/{ => kvm}/hyperv.h (100%)
create mode 100644 target/i386/kvm/kvm-cpu.h
rename target/i386/{ => kvm}/kvm-stub.c (100%)
rename target/i386/{ => kvm}/kvm.c (99%)
rename target/i386/{ => kvm}/kvm_i386.h (100%)
create mode 100644 target/i386/kvm/meson.build
create mode 100644 target/i386/kvm/trace-events
create mode 100644 target/i386/kvm/trace.h
rename target/i386/{ => tcg}/bpt_helper.c (99%)
rename target/i386/{ => tcg}/cc_helper.c (99%)
create mode 100644 target/i386/tcg/cpu.c
rename target/i386/{ => tcg}/excp_helper.c (99%)
rename target/i386/{ => tcg}/fpu_helper.c (99%)
rename target/i386/{ => tcg}/int_helper.c (99%)
rename target/i386/{ => tcg}/mem_helper.c (99%)
create mode 100644 target/i386/tcg/meson.build
rename target/i386/{ => tcg}/misc_helper.c (99%)
rename target/i386/{ => tcg}/mpx_helper.c (99%)
rename target/i386/{ => tcg}/seg_helper.c (99%)
rename target/i386/{ => tcg}/smm_helper.c (99%)
rename target/i386/{ => tcg}/svm_helper.c (99%)
rename target/i386/{ => tcg}/tcg-stub.c (100%)
rename target/i386/{ => tcg}/translate.c (99%)
create mode 100644 target/i386/whpx/meson.build
rename target/i386/{ => whpx}/whp-dispatch.h (100%)
rename target/i386/{ => whpx}/whpx-all.c (99%)
rename target/i386/{ => whpx}/whpx-cpus.c (73%)
rename target/i386/{ => whpx}/whpx-cpus.h (96%)
--
2.26.2
^ permalink raw reply [flat|nested] 87+ messages in thread
* [RFC v9 01/32] accel/tcg: split CpusAccel into three TCG variants
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 8:34 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 02/32] accel/tcg: split tcg_start_vcpu_thread Claudio Fontana
` (33 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Philippe Mathieu-Daudé,
Eduardo Habkost, Paul Durrant, Jason Wang, Marcelo Tosatti,
qemu-devel, Peter Xu, Dario Faggioli, Cameron Esfahani,
haxm-team, Claudio Fontana, Anthony Perard, Bruce Rogers,
Olaf Hering, Emilio G . Cota, Colin Xu
split up the CpusAccel tcg_cpus into three TCG variants:
tcg_cpus_rr (single threaded, round robin cpus)
tcg_cpus_icount (same as rr, but with instruction counting enabled)
tcg_cpus_mttcg (multi-threaded cpus)
Suggested-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
---
accel/tcg/meson.build | 9 +-
accel/tcg/tcg-all.c | 8 +-
accel/tcg/tcg-cpus-icount.c | 147 +++++++++++
accel/tcg/tcg-cpus-icount.h | 17 ++
accel/tcg/tcg-cpus-mttcg.c | 117 +++++++++
accel/tcg/tcg-cpus-mttcg.h | 21 ++
accel/tcg/tcg-cpus-rr.c | 270 ++++++++++++++++++++
accel/tcg/tcg-cpus-rr.h | 20 ++
accel/tcg/tcg-cpus.c | 484 ++----------------------------------
accel/tcg/tcg-cpus.h | 13 +-
softmmu/icount.c | 2 +-
11 files changed, 646 insertions(+), 462 deletions(-)
create mode 100644 accel/tcg/tcg-cpus-icount.c
create mode 100644 accel/tcg/tcg-cpus-icount.h
create mode 100644 accel/tcg/tcg-cpus-mttcg.c
create mode 100644 accel/tcg/tcg-cpus-mttcg.h
create mode 100644 accel/tcg/tcg-cpus-rr.c
create mode 100644 accel/tcg/tcg-cpus-rr.h
diff --git a/accel/tcg/meson.build b/accel/tcg/meson.build
index 19b9343d5b..f39aab0a0c 100644
--- a/accel/tcg/meson.build
+++ b/accel/tcg/meson.build
@@ -12,4 +12,11 @@ tcg_ss.add(when: 'CONFIG_SOFTMMU', if_false: files('user-exec-stub.c'))
tcg_ss.add(when: 'CONFIG_PLUGIN', if_true: [files('plugin-gen.c'), libdl])
specific_ss.add_all(when: 'CONFIG_TCG', if_true: tcg_ss)
-specific_ss.add(when: ['CONFIG_SOFTMMU', 'CONFIG_TCG'], if_true: files('tcg-all.c', 'cputlb.c', 'tcg-cpus.c'))
+specific_ss.add(when: ['CONFIG_SOFTMMU', 'CONFIG_TCG'], if_true: files(
+ 'tcg-all.c',
+ 'cputlb.c',
+ 'tcg-cpus.c',
+ 'tcg-cpus-mttcg.c',
+ 'tcg-cpus-icount.c',
+ 'tcg-cpus-rr.c'
+))
diff --git a/accel/tcg/tcg-all.c b/accel/tcg/tcg-all.c
index fa1208158f..e42a028043 100644
--- a/accel/tcg/tcg-all.c
+++ b/accel/tcg/tcg-all.c
@@ -104,8 +104,14 @@ static int tcg_init(MachineState *ms)
tcg_exec_init(s->tb_size * 1024 * 1024);
mttcg_enabled = s->mttcg_enabled;
- cpus_register_accel(&tcg_cpus);
+ if (mttcg_enabled) {
+ cpus_register_accel(&tcg_cpus_mttcg);
+ } else if (icount_enabled()) {
+ cpus_register_accel(&tcg_cpus_icount);
+ } else {
+ cpus_register_accel(&tcg_cpus_rr);
+ }
return 0;
}
diff --git a/accel/tcg/tcg-cpus-icount.c b/accel/tcg/tcg-cpus-icount.c
new file mode 100644
index 0000000000..d3af3afb6d
--- /dev/null
+++ b/accel/tcg/tcg-cpus-icount.c
@@ -0,0 +1,147 @@
+/*
+ * QEMU TCG Single Threaded vCPUs implementation using instruction counting
+ *
+ * Copyright (c) 2003-2008 Fabrice Bellard
+ * Copyright (c) 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+#include "sysemu/tcg.h"
+#include "sysemu/replay.h"
+#include "qemu/main-loop.h"
+#include "qemu/guest-random.h"
+#include "exec/exec-all.h"
+#include "hw/boards.h"
+
+#include "tcg-cpus.h"
+#include "tcg-cpus-icount.h"
+#include "tcg-cpus-rr.h"
+
+static int64_t tcg_get_icount_limit(void)
+{
+ int64_t deadline;
+
+ if (replay_mode != REPLAY_MODE_PLAY) {
+ /*
+ * Include all the timers, because they may need an attention.
+ * Too long CPU execution may create unnecessary delay in UI.
+ */
+ deadline = qemu_clock_deadline_ns_all(QEMU_CLOCK_VIRTUAL,
+ QEMU_TIMER_ATTR_ALL);
+ /* Check realtime timers, because they help with input processing */
+ deadline = qemu_soonest_timeout(deadline,
+ qemu_clock_deadline_ns_all(QEMU_CLOCK_REALTIME,
+ QEMU_TIMER_ATTR_ALL));
+
+ /*
+ * Maintain prior (possibly buggy) behaviour where if no deadline
+ * was set (as there is no QEMU_CLOCK_VIRTUAL timer) or it is more than
+ * INT32_MAX nanoseconds ahead, we still use INT32_MAX
+ * nanoseconds.
+ */
+ if ((deadline < 0) || (deadline > INT32_MAX)) {
+ deadline = INT32_MAX;
+ }
+
+ return icount_round(deadline);
+ } else {
+ return replay_get_instructions();
+ }
+}
+
+static void notify_aio_contexts(void)
+{
+ /* Wake up other AioContexts. */
+ qemu_clock_notify(QEMU_CLOCK_VIRTUAL);
+ qemu_clock_run_timers(QEMU_CLOCK_VIRTUAL);
+}
+
+void handle_icount_deadline(void)
+{
+ assert(qemu_in_vcpu_thread());
+ int64_t deadline = qemu_clock_deadline_ns_all(QEMU_CLOCK_VIRTUAL,
+ QEMU_TIMER_ATTR_ALL);
+
+ if (deadline == 0) {
+ notify_aio_contexts();
+ }
+}
+
+void prepare_icount_for_run(CPUState *cpu)
+{
+ int insns_left;
+
+ /*
+ * These should always be cleared by process_icount_data after
+ * each vCPU execution. However u16.high can be raised
+ * asynchronously by cpu_exit/cpu_interrupt/tcg_handle_interrupt
+ */
+ g_assert(cpu_neg(cpu)->icount_decr.u16.low == 0);
+ g_assert(cpu->icount_extra == 0);
+
+ cpu->icount_budget = tcg_get_icount_limit();
+ insns_left = MIN(0xffff, cpu->icount_budget);
+ cpu_neg(cpu)->icount_decr.u16.low = insns_left;
+ cpu->icount_extra = cpu->icount_budget - insns_left;
+
+ replay_mutex_lock();
+
+ if (cpu->icount_budget == 0 && replay_has_checkpoint()) {
+ notify_aio_contexts();
+ }
+}
+
+void process_icount_data(CPUState *cpu)
+{
+ /* Account for executed instructions */
+ icount_update(cpu);
+
+ /* Reset the counters */
+ cpu_neg(cpu)->icount_decr.u16.low = 0;
+ cpu->icount_extra = 0;
+ cpu->icount_budget = 0;
+
+ replay_account_executed_instructions();
+
+ replay_mutex_unlock();
+}
+
+static void icount_handle_interrupt(CPUState *cpu, int mask)
+{
+ int old_mask = cpu->interrupt_request;
+
+ tcg_handle_interrupt(cpu, mask);
+ if (qemu_cpu_is_self(cpu) &&
+ !cpu->can_do_io
+ && (mask & ~old_mask) != 0) {
+ cpu_abort(cpu, "Raised interrupt while not in I/O function");
+ }
+}
+
+const CpusAccel tcg_cpus_icount = {
+ .create_vcpu_thread = tcg_start_vcpu_thread,
+ .kick_vcpu_thread = qemu_cpu_kick_rr_cpus,
+
+ .handle_interrupt = icount_handle_interrupt,
+ .get_virtual_clock = icount_get,
+ .get_elapsed_ticks = icount_get,
+};
diff --git a/accel/tcg/tcg-cpus-icount.h b/accel/tcg/tcg-cpus-icount.h
new file mode 100644
index 0000000000..cbcf76b413
--- /dev/null
+++ b/accel/tcg/tcg-cpus-icount.h
@@ -0,0 +1,17 @@
+/*
+ * QEMU TCG Single Threaded vCPUs implementation using instruction counting
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#ifndef TCG_CPUS_ICOUNT_H
+#define TCG_CPUS_ICOUNT_H
+
+void handle_icount_deadline(void);
+void prepare_icount_for_run(CPUState *cpu);
+void process_icount_data(CPUState *cpu);
+
+#endif /* TCG_CPUS_ICOUNT_H */
diff --git a/accel/tcg/tcg-cpus-mttcg.c b/accel/tcg/tcg-cpus-mttcg.c
new file mode 100644
index 0000000000..dac724fc85
--- /dev/null
+++ b/accel/tcg/tcg-cpus-mttcg.c
@@ -0,0 +1,117 @@
+/*
+ * QEMU TCG Multi Threaded vCPUs implementation
+ *
+ * Copyright (c) 2003-2008 Fabrice Bellard
+ * Copyright (c) 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+#include "sysemu/tcg.h"
+#include "sysemu/replay.h"
+#include "qemu/main-loop.h"
+#include "qemu/guest-random.h"
+#include "exec/exec-all.h"
+#include "hw/boards.h"
+
+#include "tcg-cpus.h"
+#include "tcg-cpus-mttcg.h"
+
+/*
+ * In the multi-threaded case each vCPU has its own thread. The TLS
+ * variable current_cpu can be used deep in the code to find the
+ * current CPUState for a given thread.
+ */
+
+void *tcg_cpu_thread_fn(void *arg)
+{
+ CPUState *cpu = arg;
+
+ assert(tcg_enabled());
+ g_assert(!icount_enabled());
+
+ rcu_register_thread();
+ tcg_register_thread();
+
+ qemu_mutex_lock_iothread();
+ qemu_thread_get_self(cpu->thread);
+
+ cpu->thread_id = qemu_get_thread_id();
+ cpu->can_do_io = 1;
+ current_cpu = cpu;
+ cpu_thread_signal_created(cpu);
+ qemu_guest_random_seed_thread_part2(cpu->random_seed);
+
+ /* process any pending work */
+ cpu->exit_request = 1;
+
+ do {
+ if (cpu_can_run(cpu)) {
+ int r;
+ qemu_mutex_unlock_iothread();
+ r = tcg_cpu_exec(cpu);
+ qemu_mutex_lock_iothread();
+ switch (r) {
+ case EXCP_DEBUG:
+ cpu_handle_guest_debug(cpu);
+ break;
+ case EXCP_HALTED:
+ /*
+ * during start-up the vCPU is reset and the thread is
+ * kicked several times. If we don't ensure we go back
+ * to sleep in the halted state we won't cleanly
+ * start-up when the vCPU is enabled.
+ *
+ * cpu->halted should ensure we sleep in wait_io_event
+ */
+ g_assert(cpu->halted);
+ break;
+ case EXCP_ATOMIC:
+ qemu_mutex_unlock_iothread();
+ cpu_exec_step_atomic(cpu);
+ qemu_mutex_lock_iothread();
+ default:
+ /* Ignore everything else? */
+ break;
+ }
+ }
+
+ qatomic_mb_set(&cpu->exit_request, 0);
+ qemu_wait_io_event(cpu);
+ } while (!cpu->unplug || cpu_can_run(cpu));
+
+ qemu_tcg_destroy_vcpu(cpu);
+ qemu_mutex_unlock_iothread();
+ rcu_unregister_thread();
+ return NULL;
+}
+
+static void mttcg_kick_vcpu_thread(CPUState *cpu)
+{
+ cpu_exit(cpu);
+}
+
+const CpusAccel tcg_cpus_mttcg = {
+ .create_vcpu_thread = tcg_start_vcpu_thread,
+ .kick_vcpu_thread = mttcg_kick_vcpu_thread,
+
+ .handle_interrupt = tcg_handle_interrupt,
+};
diff --git a/accel/tcg/tcg-cpus-mttcg.h b/accel/tcg/tcg-cpus-mttcg.h
new file mode 100644
index 0000000000..d1bd771f49
--- /dev/null
+++ b/accel/tcg/tcg-cpus-mttcg.h
@@ -0,0 +1,21 @@
+/*
+ * QEMU TCG Multi Threaded vCPUs implementation
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#ifndef TCG_CPUS_MTTCG_H
+#define TCG_CPUS_MTTCG_H
+
+/*
+ * In the multi-threaded case each vCPU has its own thread. The TLS
+ * variable current_cpu can be used deep in the code to find the
+ * current CPUState for a given thread.
+ */
+
+void *tcg_cpu_thread_fn(void *arg);
+
+#endif /* TCG_CPUS_MTTCG_H */
diff --git a/accel/tcg/tcg-cpus-rr.c b/accel/tcg/tcg-cpus-rr.c
new file mode 100644
index 0000000000..ad50a3765f
--- /dev/null
+++ b/accel/tcg/tcg-cpus-rr.c
@@ -0,0 +1,270 @@
+/*
+ * QEMU TCG Single Threaded vCPUs implementation
+ *
+ * Copyright (c) 2003-2008 Fabrice Bellard
+ * Copyright (c) 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+#include "sysemu/tcg.h"
+#include "sysemu/replay.h"
+#include "qemu/main-loop.h"
+#include "qemu/guest-random.h"
+#include "exec/exec-all.h"
+#include "hw/boards.h"
+
+#include "tcg-cpus.h"
+#include "tcg-cpus-rr.h"
+#include "tcg-cpus-icount.h"
+
+/* Kick all RR vCPUs */
+void qemu_cpu_kick_rr_cpus(CPUState *unused)
+{
+ CPUState *cpu;
+
+ CPU_FOREACH(cpu) {
+ cpu_exit(cpu);
+ };
+}
+
+/*
+ * TCG vCPU kick timer
+ *
+ * The kick timer is responsible for moving single threaded vCPU
+ * emulation on to the next vCPU. If more than one vCPU is running a
+ * timer event with force a cpu->exit so the next vCPU can get
+ * scheduled.
+ *
+ * The timer is removed if all vCPUs are idle and restarted again once
+ * idleness is complete.
+ */
+
+static QEMUTimer *tcg_kick_vcpu_timer;
+static CPUState *tcg_current_rr_cpu;
+
+#define TCG_KICK_PERIOD (NANOSECONDS_PER_SECOND / 10)
+
+static inline int64_t qemu_tcg_next_kick(void)
+{
+ return qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL) + TCG_KICK_PERIOD;
+}
+
+/* Kick the currently round-robin scheduled vCPU to next */
+static void qemu_cpu_kick_rr_next_cpu(void)
+{
+ CPUState *cpu;
+ do {
+ cpu = qatomic_mb_read(&tcg_current_rr_cpu);
+ if (cpu) {
+ cpu_exit(cpu);
+ }
+ } while (cpu != qatomic_mb_read(&tcg_current_rr_cpu));
+}
+
+static void kick_tcg_thread(void *opaque)
+{
+ timer_mod(tcg_kick_vcpu_timer, qemu_tcg_next_kick());
+ qemu_cpu_kick_rr_next_cpu();
+}
+
+static void start_tcg_kick_timer(void)
+{
+ if (!tcg_kick_vcpu_timer && CPU_NEXT(first_cpu)) {
+ tcg_kick_vcpu_timer = timer_new_ns(QEMU_CLOCK_VIRTUAL,
+ kick_tcg_thread, NULL);
+ }
+ if (tcg_kick_vcpu_timer && !timer_pending(tcg_kick_vcpu_timer)) {
+ timer_mod(tcg_kick_vcpu_timer, qemu_tcg_next_kick());
+ }
+}
+
+static void stop_tcg_kick_timer(void)
+{
+ if (tcg_kick_vcpu_timer && timer_pending(tcg_kick_vcpu_timer)) {
+ timer_del(tcg_kick_vcpu_timer);
+ }
+}
+
+static void qemu_tcg_rr_wait_io_event(void)
+{
+ CPUState *cpu;
+
+ while (all_cpu_threads_idle()) {
+ stop_tcg_kick_timer();
+ qemu_cond_wait_iothread(first_cpu->halt_cond);
+ }
+
+ start_tcg_kick_timer();
+
+ CPU_FOREACH(cpu) {
+ qemu_wait_io_event_common(cpu);
+ }
+}
+
+/*
+ * Destroy any remaining vCPUs which have been unplugged and have
+ * finished running
+ */
+static void deal_with_unplugged_cpus(void)
+{
+ CPUState *cpu;
+
+ CPU_FOREACH(cpu) {
+ if (cpu->unplug && !cpu_can_run(cpu)) {
+ qemu_tcg_destroy_vcpu(cpu);
+ break;
+ }
+ }
+}
+
+/*
+ * In the single-threaded case each vCPU is simulated in turn. If
+ * there is more than a single vCPU we create a simple timer to kick
+ * the vCPU and ensure we don't get stuck in a tight loop in one vCPU.
+ * This is done explicitly rather than relying on side-effects
+ * elsewhere.
+ */
+
+void *tcg_rr_cpu_thread_fn(void *arg)
+{
+ CPUState *cpu = arg;
+
+ assert(tcg_enabled());
+ rcu_register_thread();
+ tcg_register_thread();
+
+ qemu_mutex_lock_iothread();
+ qemu_thread_get_self(cpu->thread);
+
+ cpu->thread_id = qemu_get_thread_id();
+ cpu->can_do_io = 1;
+ cpu_thread_signal_created(cpu);
+ qemu_guest_random_seed_thread_part2(cpu->random_seed);
+
+ /* wait for initial kick-off after machine start */
+ while (first_cpu->stopped) {
+ qemu_cond_wait_iothread(first_cpu->halt_cond);
+
+ /* process any pending work */
+ CPU_FOREACH(cpu) {
+ current_cpu = cpu;
+ qemu_wait_io_event_common(cpu);
+ }
+ }
+
+ start_tcg_kick_timer();
+
+ cpu = first_cpu;
+
+ /* process any pending work */
+ cpu->exit_request = 1;
+
+ while (1) {
+ qemu_mutex_unlock_iothread();
+ replay_mutex_lock();
+ qemu_mutex_lock_iothread();
+
+ if (icount_enabled()) {
+ /* Account partial waits to QEMU_CLOCK_VIRTUAL. */
+ icount_account_warp_timer();
+ /*
+ * Run the timers here. This is much more efficient than
+ * waking up the I/O thread and waiting for completion.
+ */
+ handle_icount_deadline();
+ }
+
+ replay_mutex_unlock();
+
+ if (!cpu) {
+ cpu = first_cpu;
+ }
+
+ while (cpu && cpu_work_list_empty(cpu) && !cpu->exit_request) {
+
+ qatomic_mb_set(&tcg_current_rr_cpu, cpu);
+ current_cpu = cpu;
+
+ qemu_clock_enable(QEMU_CLOCK_VIRTUAL,
+ (cpu->singlestep_enabled & SSTEP_NOTIMER) == 0);
+
+ if (cpu_can_run(cpu)) {
+ int r;
+
+ qemu_mutex_unlock_iothread();
+ if (icount_enabled()) {
+ prepare_icount_for_run(cpu);
+ }
+ r = tcg_cpu_exec(cpu);
+ if (icount_enabled()) {
+ process_icount_data(cpu);
+ }
+ qemu_mutex_lock_iothread();
+
+ if (r == EXCP_DEBUG) {
+ cpu_handle_guest_debug(cpu);
+ break;
+ } else if (r == EXCP_ATOMIC) {
+ qemu_mutex_unlock_iothread();
+ cpu_exec_step_atomic(cpu);
+ qemu_mutex_lock_iothread();
+ break;
+ }
+ } else if (cpu->stop) {
+ if (cpu->unplug) {
+ cpu = CPU_NEXT(cpu);
+ }
+ break;
+ }
+
+ cpu = CPU_NEXT(cpu);
+ } /* while (cpu && !cpu->exit_request).. */
+
+ /* Does not need qatomic_mb_set because a spurious wakeup is okay. */
+ qatomic_set(&tcg_current_rr_cpu, NULL);
+
+ if (cpu && cpu->exit_request) {
+ qatomic_mb_set(&cpu->exit_request, 0);
+ }
+
+ if (icount_enabled() && all_cpu_threads_idle()) {
+ /*
+ * When all cpus are sleeping (e.g in WFI), to avoid a deadlock
+ * in the main_loop, wake it up in order to start the warp timer.
+ */
+ qemu_notify_event();
+ }
+
+ qemu_tcg_rr_wait_io_event();
+ deal_with_unplugged_cpus();
+ }
+
+ rcu_unregister_thread();
+ return NULL;
+}
+
+const CpusAccel tcg_cpus_rr = {
+ .create_vcpu_thread = tcg_start_vcpu_thread,
+ .kick_vcpu_thread = qemu_cpu_kick_rr_cpus,
+
+ .handle_interrupt = tcg_handle_interrupt,
+};
diff --git a/accel/tcg/tcg-cpus-rr.h b/accel/tcg/tcg-cpus-rr.h
new file mode 100644
index 0000000000..1936fd16ab
--- /dev/null
+++ b/accel/tcg/tcg-cpus-rr.h
@@ -0,0 +1,20 @@
+/*
+ * QEMU TCG Single Threaded vCPUs implementation
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#ifndef TCG_CPUS_RR_H
+#define TCG_CPUS_RR_H
+
+#define TCG_KICK_PERIOD (NANOSECONDS_PER_SECOND / 10)
+
+/* Kick all RR vCPUs. */
+void qemu_cpu_kick_rr_cpus(CPUState *unused);
+
+void *tcg_rr_cpu_thread_fn(void *arg);
+
+#endif /* TCG_CPUS_RR_H */
diff --git a/accel/tcg/tcg-cpus.c b/accel/tcg/tcg-cpus.c
index da1c63d8f6..f2b9bbf99e 100644
--- a/accel/tcg/tcg-cpus.c
+++ b/accel/tcg/tcg-cpus.c
@@ -1,5 +1,7 @@
/*
- * QEMU System Emulator
+ * QEMU TCG vCPU common functionality
+ *
+ * Functionality common to all TCG vCPU variants: mttcg, rr and icount.
*
* Copyright (c) 2003-2008 Fabrice Bellard
* Copyright (c) 2014 Red Hat Inc.
@@ -33,436 +35,12 @@
#include "hw/boards.h"
#include "tcg-cpus.h"
+#include "tcg-cpus-mttcg.h"
+#include "tcg-cpus-rr.h"
-/* Kick all RR vCPUs */
-static void qemu_cpu_kick_rr_cpus(void)
-{
- CPUState *cpu;
+/* common functionality among all TCG variants */
- CPU_FOREACH(cpu) {
- cpu_exit(cpu);
- };
-}
-
-static void tcg_kick_vcpu_thread(CPUState *cpu)
-{
- if (qemu_tcg_mttcg_enabled()) {
- cpu_exit(cpu);
- } else {
- qemu_cpu_kick_rr_cpus();
- }
-}
-
-/*
- * TCG vCPU kick timer
- *
- * The kick timer is responsible for moving single threaded vCPU
- * emulation on to the next vCPU. If more than one vCPU is running a
- * timer event with force a cpu->exit so the next vCPU can get
- * scheduled.
- *
- * The timer is removed if all vCPUs are idle and restarted again once
- * idleness is complete.
- */
-
-static QEMUTimer *tcg_kick_vcpu_timer;
-static CPUState *tcg_current_rr_cpu;
-
-#define TCG_KICK_PERIOD (NANOSECONDS_PER_SECOND / 10)
-
-static inline int64_t qemu_tcg_next_kick(void)
-{
- return qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL) + TCG_KICK_PERIOD;
-}
-
-/* Kick the currently round-robin scheduled vCPU to next */
-static void qemu_cpu_kick_rr_next_cpu(void)
-{
- CPUState *cpu;
- do {
- cpu = qatomic_mb_read(&tcg_current_rr_cpu);
- if (cpu) {
- cpu_exit(cpu);
- }
- } while (cpu != qatomic_mb_read(&tcg_current_rr_cpu));
-}
-
-static void kick_tcg_thread(void *opaque)
-{
- timer_mod(tcg_kick_vcpu_timer, qemu_tcg_next_kick());
- qemu_cpu_kick_rr_next_cpu();
-}
-
-static void start_tcg_kick_timer(void)
-{
- assert(!mttcg_enabled);
- if (!tcg_kick_vcpu_timer && CPU_NEXT(first_cpu)) {
- tcg_kick_vcpu_timer = timer_new_ns(QEMU_CLOCK_VIRTUAL,
- kick_tcg_thread, NULL);
- }
- if (tcg_kick_vcpu_timer && !timer_pending(tcg_kick_vcpu_timer)) {
- timer_mod(tcg_kick_vcpu_timer, qemu_tcg_next_kick());
- }
-}
-
-static void stop_tcg_kick_timer(void)
-{
- assert(!mttcg_enabled);
- if (tcg_kick_vcpu_timer && timer_pending(tcg_kick_vcpu_timer)) {
- timer_del(tcg_kick_vcpu_timer);
- }
-}
-
-static void qemu_tcg_destroy_vcpu(CPUState *cpu)
-{
-}
-
-static void qemu_tcg_rr_wait_io_event(void)
-{
- CPUState *cpu;
-
- while (all_cpu_threads_idle()) {
- stop_tcg_kick_timer();
- qemu_cond_wait_iothread(first_cpu->halt_cond);
- }
-
- start_tcg_kick_timer();
-
- CPU_FOREACH(cpu) {
- qemu_wait_io_event_common(cpu);
- }
-}
-
-static int64_t tcg_get_icount_limit(void)
-{
- int64_t deadline;
-
- if (replay_mode != REPLAY_MODE_PLAY) {
- /*
- * Include all the timers, because they may need an attention.
- * Too long CPU execution may create unnecessary delay in UI.
- */
- deadline = qemu_clock_deadline_ns_all(QEMU_CLOCK_VIRTUAL,
- QEMU_TIMER_ATTR_ALL);
- /* Check realtime timers, because they help with input processing */
- deadline = qemu_soonest_timeout(deadline,
- qemu_clock_deadline_ns_all(QEMU_CLOCK_REALTIME,
- QEMU_TIMER_ATTR_ALL));
-
- /*
- * Maintain prior (possibly buggy) behaviour where if no deadline
- * was set (as there is no QEMU_CLOCK_VIRTUAL timer) or it is more than
- * INT32_MAX nanoseconds ahead, we still use INT32_MAX
- * nanoseconds.
- */
- if ((deadline < 0) || (deadline > INT32_MAX)) {
- deadline = INT32_MAX;
- }
-
- return icount_round(deadline);
- } else {
- return replay_get_instructions();
- }
-}
-
-static void notify_aio_contexts(void)
-{
- /* Wake up other AioContexts. */
- qemu_clock_notify(QEMU_CLOCK_VIRTUAL);
- qemu_clock_run_timers(QEMU_CLOCK_VIRTUAL);
-}
-
-static void handle_icount_deadline(void)
-{
- assert(qemu_in_vcpu_thread());
- if (icount_enabled()) {
- int64_t deadline = qemu_clock_deadline_ns_all(QEMU_CLOCK_VIRTUAL,
- QEMU_TIMER_ATTR_ALL);
-
- if (deadline == 0) {
- notify_aio_contexts();
- }
- }
-}
-
-static void prepare_icount_for_run(CPUState *cpu)
-{
- if (icount_enabled()) {
- int insns_left;
-
- /*
- * These should always be cleared by process_icount_data after
- * each vCPU execution. However u16.high can be raised
- * asynchronously by cpu_exit/cpu_interrupt/tcg_handle_interrupt
- */
- g_assert(cpu_neg(cpu)->icount_decr.u16.low == 0);
- g_assert(cpu->icount_extra == 0);
-
- cpu->icount_budget = tcg_get_icount_limit();
- insns_left = MIN(0xffff, cpu->icount_budget);
- cpu_neg(cpu)->icount_decr.u16.low = insns_left;
- cpu->icount_extra = cpu->icount_budget - insns_left;
-
- replay_mutex_lock();
-
- if (cpu->icount_budget == 0 && replay_has_checkpoint()) {
- notify_aio_contexts();
- }
- }
-}
-
-static void process_icount_data(CPUState *cpu)
-{
- if (icount_enabled()) {
- /* Account for executed instructions */
- icount_update(cpu);
-
- /* Reset the counters */
- cpu_neg(cpu)->icount_decr.u16.low = 0;
- cpu->icount_extra = 0;
- cpu->icount_budget = 0;
-
- replay_account_executed_instructions();
-
- replay_mutex_unlock();
- }
-}
-
-static int tcg_cpu_exec(CPUState *cpu)
-{
- int ret;
-#ifdef CONFIG_PROFILER
- int64_t ti;
-#endif
-
- assert(tcg_enabled());
-#ifdef CONFIG_PROFILER
- ti = profile_getclock();
-#endif
- cpu_exec_start(cpu);
- ret = cpu_exec(cpu);
- cpu_exec_end(cpu);
-#ifdef CONFIG_PROFILER
- qatomic_set(&tcg_ctx->prof.cpu_exec_time,
- tcg_ctx->prof.cpu_exec_time + profile_getclock() - ti);
-#endif
- return ret;
-}
-
-/*
- * Destroy any remaining vCPUs which have been unplugged and have
- * finished running
- */
-static void deal_with_unplugged_cpus(void)
-{
- CPUState *cpu;
-
- CPU_FOREACH(cpu) {
- if (cpu->unplug && !cpu_can_run(cpu)) {
- qemu_tcg_destroy_vcpu(cpu);
- cpu_thread_signal_destroyed(cpu);
- break;
- }
- }
-}
-
-/*
- * Single-threaded TCG
- *
- * In the single-threaded case each vCPU is simulated in turn. If
- * there is more than a single vCPU we create a simple timer to kick
- * the vCPU and ensure we don't get stuck in a tight loop in one vCPU.
- * This is done explicitly rather than relying on side-effects
- * elsewhere.
- */
-
-static void *tcg_rr_cpu_thread_fn(void *arg)
-{
- CPUState *cpu = arg;
-
- assert(tcg_enabled());
- rcu_register_thread();
- tcg_register_thread();
-
- qemu_mutex_lock_iothread();
- qemu_thread_get_self(cpu->thread);
-
- cpu->thread_id = qemu_get_thread_id();
- cpu->can_do_io = 1;
- cpu_thread_signal_created(cpu);
- qemu_guest_random_seed_thread_part2(cpu->random_seed);
-
- /* wait for initial kick-off after machine start */
- while (first_cpu->stopped) {
- qemu_cond_wait_iothread(first_cpu->halt_cond);
-
- /* process any pending work */
- CPU_FOREACH(cpu) {
- current_cpu = cpu;
- qemu_wait_io_event_common(cpu);
- }
- }
-
- start_tcg_kick_timer();
-
- cpu = first_cpu;
-
- /* process any pending work */
- cpu->exit_request = 1;
-
- while (1) {
- qemu_mutex_unlock_iothread();
- replay_mutex_lock();
- qemu_mutex_lock_iothread();
- /* Account partial waits to QEMU_CLOCK_VIRTUAL. */
- icount_account_warp_timer();
-
- /*
- * Run the timers here. This is much more efficient than
- * waking up the I/O thread and waiting for completion.
- */
- handle_icount_deadline();
-
- replay_mutex_unlock();
-
- if (!cpu) {
- cpu = first_cpu;
- }
-
- while (cpu && cpu_work_list_empty(cpu) && !cpu->exit_request) {
-
- qatomic_mb_set(&tcg_current_rr_cpu, cpu);
- current_cpu = cpu;
-
- qemu_clock_enable(QEMU_CLOCK_VIRTUAL,
- (cpu->singlestep_enabled & SSTEP_NOTIMER) == 0);
-
- if (cpu_can_run(cpu)) {
- int r;
-
- qemu_mutex_unlock_iothread();
- prepare_icount_for_run(cpu);
-
- r = tcg_cpu_exec(cpu);
-
- process_icount_data(cpu);
- qemu_mutex_lock_iothread();
-
- if (r == EXCP_DEBUG) {
- cpu_handle_guest_debug(cpu);
- break;
- } else if (r == EXCP_ATOMIC) {
- qemu_mutex_unlock_iothread();
- cpu_exec_step_atomic(cpu);
- qemu_mutex_lock_iothread();
- break;
- }
- } else if (cpu->stop) {
- if (cpu->unplug) {
- cpu = CPU_NEXT(cpu);
- }
- break;
- }
-
- cpu = CPU_NEXT(cpu);
- } /* while (cpu && !cpu->exit_request).. */
-
- /* Does not need qatomic_mb_set because a spurious wakeup is okay. */
- qatomic_set(&tcg_current_rr_cpu, NULL);
-
- if (cpu && cpu->exit_request) {
- qatomic_mb_set(&cpu->exit_request, 0);
- }
-
- if (icount_enabled() && all_cpu_threads_idle()) {
- /*
- * When all cpus are sleeping (e.g in WFI), to avoid a deadlock
- * in the main_loop, wake it up in order to start the warp timer.
- */
- qemu_notify_event();
- }
-
- qemu_tcg_rr_wait_io_event();
- deal_with_unplugged_cpus();
- }
-
- rcu_unregister_thread();
- return NULL;
-}
-
-/*
- * Multi-threaded TCG
- *
- * In the multi-threaded case each vCPU has its own thread. The TLS
- * variable current_cpu can be used deep in the code to find the
- * current CPUState for a given thread.
- */
-
-static void *tcg_cpu_thread_fn(void *arg)
-{
- CPUState *cpu = arg;
-
- assert(tcg_enabled());
- g_assert(!icount_enabled());
-
- rcu_register_thread();
- tcg_register_thread();
-
- qemu_mutex_lock_iothread();
- qemu_thread_get_self(cpu->thread);
-
- cpu->thread_id = qemu_get_thread_id();
- cpu->can_do_io = 1;
- current_cpu = cpu;
- cpu_thread_signal_created(cpu);
- qemu_guest_random_seed_thread_part2(cpu->random_seed);
-
- /* process any pending work */
- cpu->exit_request = 1;
-
- do {
- if (cpu_can_run(cpu)) {
- int r;
- qemu_mutex_unlock_iothread();
- r = tcg_cpu_exec(cpu);
- qemu_mutex_lock_iothread();
- switch (r) {
- case EXCP_DEBUG:
- cpu_handle_guest_debug(cpu);
- break;
- case EXCP_HALTED:
- /*
- * during start-up the vCPU is reset and the thread is
- * kicked several times. If we don't ensure we go back
- * to sleep in the halted state we won't cleanly
- * start-up when the vCPU is enabled.
- *
- * cpu->halted should ensure we sleep in wait_io_event
- */
- g_assert(cpu->halted);
- break;
- case EXCP_ATOMIC:
- qemu_mutex_unlock_iothread();
- cpu_exec_step_atomic(cpu);
- qemu_mutex_lock_iothread();
- default:
- /* Ignore everything else? */
- break;
- }
- }
-
- qatomic_mb_set(&cpu->exit_request, 0);
- qemu_wait_io_event(cpu);
- } while (!cpu->unplug || cpu_can_run(cpu));
-
- qemu_tcg_destroy_vcpu(cpu);
- cpu_thread_signal_destroyed(cpu);
- qemu_mutex_unlock_iothread();
- rcu_unregister_thread();
- return NULL;
-}
-
-static void tcg_start_vcpu_thread(CPUState *cpu)
+void tcg_start_vcpu_thread(CPUState *cpu)
{
char thread_name[VCPU_THREAD_NAME_SIZE];
static QemuCond *single_tcg_halt_cond;
@@ -518,29 +96,36 @@ static void tcg_start_vcpu_thread(CPUState *cpu)
}
}
-static int64_t tcg_get_virtual_clock(void)
+void qemu_tcg_destroy_vcpu(CPUState *cpu)
{
- if (icount_enabled()) {
- return icount_get();
- }
- return cpu_get_clock();
+ cpu_thread_signal_destroyed(cpu);
}
-static int64_t tcg_get_elapsed_ticks(void)
+int tcg_cpu_exec(CPUState *cpu)
{
- if (icount_enabled()) {
- return icount_get();
- }
- return cpu_get_ticks();
+ int ret;
+#ifdef CONFIG_PROFILER
+ int64_t ti;
+#endif
+ assert(tcg_enabled());
+#ifdef CONFIG_PROFILER
+ ti = profile_getclock();
+#endif
+ cpu_exec_start(cpu);
+ ret = cpu_exec(cpu);
+ cpu_exec_end(cpu);
+#ifdef CONFIG_PROFILER
+ qatomic_set(&tcg_ctx->prof.cpu_exec_time,
+ tcg_ctx->prof.cpu_exec_time + profile_getclock() - ti);
+#endif
+ return ret;
}
/* mask must never be zero, except for A20 change call */
-static void tcg_handle_interrupt(CPUState *cpu, int mask)
+void tcg_handle_interrupt(CPUState *cpu, int mask)
{
- int old_mask;
g_assert(qemu_mutex_iothread_locked());
- old_mask = cpu->interrupt_request;
cpu->interrupt_request |= mask;
/*
@@ -551,20 +136,5 @@ static void tcg_handle_interrupt(CPUState *cpu, int mask)
qemu_cpu_kick(cpu);
} else {
qatomic_set(&cpu_neg(cpu)->icount_decr.u16.high, -1);
- if (icount_enabled() &&
- !cpu->can_do_io
- && (mask & ~old_mask) != 0) {
- cpu_abort(cpu, "Raised interrupt while not in I/O function");
- }
}
}
-
-const CpusAccel tcg_cpus = {
- .create_vcpu_thread = tcg_start_vcpu_thread,
- .kick_vcpu_thread = tcg_kick_vcpu_thread,
-
- .handle_interrupt = tcg_handle_interrupt,
-
- .get_virtual_clock = tcg_get_virtual_clock,
- .get_elapsed_ticks = tcg_get_elapsed_ticks,
-};
diff --git a/accel/tcg/tcg-cpus.h b/accel/tcg/tcg-cpus.h
index 8b1d9d2abc..279ba72e1f 100644
--- a/accel/tcg/tcg-cpus.h
+++ b/accel/tcg/tcg-cpus.h
@@ -1,5 +1,7 @@
/*
- * Accelerator CPUS Interface
+ * QEMU TCG vCPU common functionality
+ *
+ * Functionality common to all TCG vcpu variants: mttcg, rr and icount.
*
* Copyright 2020 SUSE LLC
*
@@ -12,6 +14,13 @@
#include "sysemu/cpus.h"
-extern const CpusAccel tcg_cpus;
+extern const CpusAccel tcg_cpus_mttcg;
+extern const CpusAccel tcg_cpus_icount;
+extern const CpusAccel tcg_cpus_rr;
+
+void tcg_start_vcpu_thread(CPUState *cpu);
+void qemu_tcg_destroy_vcpu(CPUState *cpu);
+int tcg_cpu_exec(CPUState *cpu);
+void tcg_handle_interrupt(CPUState *cpu, int mask);
#endif /* TCG_CPUS_H */
diff --git a/softmmu/icount.c b/softmmu/icount.c
index 020a201a01..dbcd8c3594 100644
--- a/softmmu/icount.c
+++ b/softmmu/icount.c
@@ -396,7 +396,7 @@ void icount_start_warp_timer(void)
void icount_account_warp_timer(void)
{
- if (!icount_enabled() || !icount_sleep) {
+ if (!icount_sleep) {
return;
}
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 02/32] accel/tcg: split tcg_start_vcpu_thread
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
2020-12-08 19:48 ` [RFC v9 01/32] accel/tcg: split CpusAccel into three TCG variants Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 9:03 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 03/32] accel/tcg: rename tcg-cpus functions to match module name Claudio Fontana
` (32 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
after the initial split into 3 tcg variants, we proceed to also
split tcg_start_vcpu_thread.
We actually split it in 2 this time, since the icount variant
just uses the round robin function.
Suggested-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
accel/tcg/tcg-all.c | 5 ++++
accel/tcg/tcg-cpus-icount.c | 2 +-
accel/tcg/tcg-cpus-mttcg.c | 29 +++++++++++++++++--
accel/tcg/tcg-cpus-mttcg.h | 21 --------------
accel/tcg/tcg-cpus-rr.c | 39 +++++++++++++++++++++++--
accel/tcg/tcg-cpus-rr.h | 3 +-
accel/tcg/tcg-cpus.c | 58 -------------------------------------
accel/tcg/tcg-cpus.h | 1 -
8 files changed, 71 insertions(+), 87 deletions(-)
delete mode 100644 accel/tcg/tcg-cpus-mttcg.h
diff --git a/accel/tcg/tcg-all.c b/accel/tcg/tcg-all.c
index e42a028043..1ac0b76515 100644
--- a/accel/tcg/tcg-all.c
+++ b/accel/tcg/tcg-all.c
@@ -105,6 +105,11 @@ static int tcg_init(MachineState *ms)
tcg_exec_init(s->tb_size * 1024 * 1024);
mttcg_enabled = s->mttcg_enabled;
+ /*
+ * Initialize TCG regions
+ */
+ tcg_region_init();
+
if (mttcg_enabled) {
cpus_register_accel(&tcg_cpus_mttcg);
} else if (icount_enabled()) {
diff --git a/accel/tcg/tcg-cpus-icount.c b/accel/tcg/tcg-cpus-icount.c
index d3af3afb6d..82dbe2cacf 100644
--- a/accel/tcg/tcg-cpus-icount.c
+++ b/accel/tcg/tcg-cpus-icount.c
@@ -138,7 +138,7 @@ static void icount_handle_interrupt(CPUState *cpu, int mask)
}
const CpusAccel tcg_cpus_icount = {
- .create_vcpu_thread = tcg_start_vcpu_thread,
+ .create_vcpu_thread = rr_start_vcpu_thread,
.kick_vcpu_thread = qemu_cpu_kick_rr_cpus,
.handle_interrupt = icount_handle_interrupt,
diff --git a/accel/tcg/tcg-cpus-mttcg.c b/accel/tcg/tcg-cpus-mttcg.c
index dac724fc85..f2b892a380 100644
--- a/accel/tcg/tcg-cpus-mttcg.c
+++ b/accel/tcg/tcg-cpus-mttcg.c
@@ -33,7 +33,6 @@
#include "hw/boards.h"
#include "tcg-cpus.h"
-#include "tcg-cpus-mttcg.h"
/*
* In the multi-threaded case each vCPU has its own thread. The TLS
@@ -41,7 +40,7 @@
* current CPUState for a given thread.
*/
-void *tcg_cpu_thread_fn(void *arg)
+static void *tcg_cpu_thread_fn(void *arg)
{
CPUState *cpu = arg;
@@ -109,8 +108,32 @@ static void mttcg_kick_vcpu_thread(CPUState *cpu)
cpu_exit(cpu);
}
+static void mttcg_start_vcpu_thread(CPUState *cpu)
+{
+ char thread_name[VCPU_THREAD_NAME_SIZE];
+
+ g_assert(tcg_enabled());
+
+ parallel_cpus = (current_machine->smp.max_cpus > 1);
+
+ cpu->thread = g_malloc0(sizeof(QemuThread));
+ cpu->halt_cond = g_malloc0(sizeof(QemuCond));
+ qemu_cond_init(cpu->halt_cond);
+
+ /* create a thread per vCPU with TCG (MTTCG) */
+ snprintf(thread_name, VCPU_THREAD_NAME_SIZE, "CPU %d/TCG",
+ cpu->cpu_index);
+
+ qemu_thread_create(cpu->thread, thread_name, tcg_cpu_thread_fn,
+ cpu, QEMU_THREAD_JOINABLE);
+
+#ifdef _WIN32
+ cpu->hThread = qemu_thread_get_handle(cpu->thread);
+#endif
+}
+
const CpusAccel tcg_cpus_mttcg = {
- .create_vcpu_thread = tcg_start_vcpu_thread,
+ .create_vcpu_thread = mttcg_start_vcpu_thread,
.kick_vcpu_thread = mttcg_kick_vcpu_thread,
.handle_interrupt = tcg_handle_interrupt,
diff --git a/accel/tcg/tcg-cpus-mttcg.h b/accel/tcg/tcg-cpus-mttcg.h
deleted file mode 100644
index d1bd771f49..0000000000
--- a/accel/tcg/tcg-cpus-mttcg.h
+++ /dev/null
@@ -1,21 +0,0 @@
-/*
- * QEMU TCG Multi Threaded vCPUs implementation
- *
- * Copyright 2020 SUSE LLC
- *
- * This work is licensed under the terms of the GNU GPL, version 2 or later.
- * See the COPYING file in the top-level directory.
- */
-
-#ifndef TCG_CPUS_MTTCG_H
-#define TCG_CPUS_MTTCG_H
-
-/*
- * In the multi-threaded case each vCPU has its own thread. The TLS
- * variable current_cpu can be used deep in the code to find the
- * current CPUState for a given thread.
- */
-
-void *tcg_cpu_thread_fn(void *arg);
-
-#endif /* TCG_CPUS_MTTCG_H */
diff --git a/accel/tcg/tcg-cpus-rr.c b/accel/tcg/tcg-cpus-rr.c
index ad50a3765f..f3b262bec7 100644
--- a/accel/tcg/tcg-cpus-rr.c
+++ b/accel/tcg/tcg-cpus-rr.c
@@ -144,7 +144,7 @@ static void deal_with_unplugged_cpus(void)
* elsewhere.
*/
-void *tcg_rr_cpu_thread_fn(void *arg)
+static void *tcg_rr_cpu_thread_fn(void *arg)
{
CPUState *cpu = arg;
@@ -262,8 +262,43 @@ void *tcg_rr_cpu_thread_fn(void *arg)
return NULL;
}
+void rr_start_vcpu_thread(CPUState *cpu)
+{
+ char thread_name[VCPU_THREAD_NAME_SIZE];
+ static QemuCond *single_tcg_halt_cond;
+ static QemuThread *single_tcg_cpu_thread;
+
+ g_assert(tcg_enabled());
+ parallel_cpus = false;
+
+ if (!single_tcg_cpu_thread) {
+ cpu->thread = g_malloc0(sizeof(QemuThread));
+ cpu->halt_cond = g_malloc0(sizeof(QemuCond));
+ qemu_cond_init(cpu->halt_cond);
+
+ /* share a single thread for all cpus with TCG */
+ snprintf(thread_name, VCPU_THREAD_NAME_SIZE, "ALL CPUs/TCG");
+ qemu_thread_create(cpu->thread, thread_name,
+ tcg_rr_cpu_thread_fn,
+ cpu, QEMU_THREAD_JOINABLE);
+
+ single_tcg_halt_cond = cpu->halt_cond;
+ single_tcg_cpu_thread = cpu->thread;
+#ifdef _WIN32
+ cpu->hThread = qemu_thread_get_handle(cpu->thread);
+#endif
+ } else {
+ /* we share the thread */
+ cpu->thread = single_tcg_cpu_thread;
+ cpu->halt_cond = single_tcg_halt_cond;
+ cpu->thread_id = first_cpu->thread_id;
+ cpu->can_do_io = 1;
+ cpu->created = true;
+ }
+}
+
const CpusAccel tcg_cpus_rr = {
- .create_vcpu_thread = tcg_start_vcpu_thread,
+ .create_vcpu_thread = rr_start_vcpu_thread,
.kick_vcpu_thread = qemu_cpu_kick_rr_cpus,
.handle_interrupt = tcg_handle_interrupt,
diff --git a/accel/tcg/tcg-cpus-rr.h b/accel/tcg/tcg-cpus-rr.h
index 1936fd16ab..2e5943eda9 100644
--- a/accel/tcg/tcg-cpus-rr.h
+++ b/accel/tcg/tcg-cpus-rr.h
@@ -15,6 +15,7 @@
/* Kick all RR vCPUs. */
void qemu_cpu_kick_rr_cpus(CPUState *unused);
-void *tcg_rr_cpu_thread_fn(void *arg);
+/* start the round robin vcpu thread */
+void rr_start_vcpu_thread(CPUState *cpu);
#endif /* TCG_CPUS_RR_H */
diff --git a/accel/tcg/tcg-cpus.c b/accel/tcg/tcg-cpus.c
index f2b9bbf99e..86fd09545a 100644
--- a/accel/tcg/tcg-cpus.c
+++ b/accel/tcg/tcg-cpus.c
@@ -35,67 +35,9 @@
#include "hw/boards.h"
#include "tcg-cpus.h"
-#include "tcg-cpus-mttcg.h"
-#include "tcg-cpus-rr.h"
/* common functionality among all TCG variants */
-void tcg_start_vcpu_thread(CPUState *cpu)
-{
- char thread_name[VCPU_THREAD_NAME_SIZE];
- static QemuCond *single_tcg_halt_cond;
- static QemuThread *single_tcg_cpu_thread;
- static int tcg_region_inited;
-
- assert(tcg_enabled());
- /*
- * Initialize TCG regions--once. Now is a good time, because:
- * (1) TCG's init context, prologue and target globals have been set up.
- * (2) qemu_tcg_mttcg_enabled() works now (TCG init code runs before the
- * -accel flag is processed, so the check doesn't work then).
- */
- if (!tcg_region_inited) {
- tcg_region_inited = 1;
- tcg_region_init();
- parallel_cpus = qemu_tcg_mttcg_enabled() && current_machine->smp.max_cpus > 1;
- }
-
- if (qemu_tcg_mttcg_enabled() || !single_tcg_cpu_thread) {
- cpu->thread = g_malloc0(sizeof(QemuThread));
- cpu->halt_cond = g_malloc0(sizeof(QemuCond));
- qemu_cond_init(cpu->halt_cond);
-
- if (qemu_tcg_mttcg_enabled()) {
- /* create a thread per vCPU with TCG (MTTCG) */
- snprintf(thread_name, VCPU_THREAD_NAME_SIZE, "CPU %d/TCG",
- cpu->cpu_index);
-
- qemu_thread_create(cpu->thread, thread_name, tcg_cpu_thread_fn,
- cpu, QEMU_THREAD_JOINABLE);
-
- } else {
- /* share a single thread for all cpus with TCG */
- snprintf(thread_name, VCPU_THREAD_NAME_SIZE, "ALL CPUs/TCG");
- qemu_thread_create(cpu->thread, thread_name,
- tcg_rr_cpu_thread_fn,
- cpu, QEMU_THREAD_JOINABLE);
-
- single_tcg_halt_cond = cpu->halt_cond;
- single_tcg_cpu_thread = cpu->thread;
- }
-#ifdef _WIN32
- cpu->hThread = qemu_thread_get_handle(cpu->thread);
-#endif
- } else {
- /* For non-MTTCG cases we share the thread */
- cpu->thread = single_tcg_cpu_thread;
- cpu->halt_cond = single_tcg_halt_cond;
- cpu->thread_id = first_cpu->thread_id;
- cpu->can_do_io = 1;
- cpu->created = true;
- }
-}
-
void qemu_tcg_destroy_vcpu(CPUState *cpu)
{
cpu_thread_signal_destroyed(cpu);
diff --git a/accel/tcg/tcg-cpus.h b/accel/tcg/tcg-cpus.h
index 279ba72e1f..b7ca954e13 100644
--- a/accel/tcg/tcg-cpus.h
+++ b/accel/tcg/tcg-cpus.h
@@ -18,7 +18,6 @@ extern const CpusAccel tcg_cpus_mttcg;
extern const CpusAccel tcg_cpus_icount;
extern const CpusAccel tcg_cpus_rr;
-void tcg_start_vcpu_thread(CPUState *cpu);
void qemu_tcg_destroy_vcpu(CPUState *cpu);
int tcg_cpu_exec(CPUState *cpu);
void tcg_handle_interrupt(CPUState *cpu, int mask);
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 03/32] accel/tcg: rename tcg-cpus functions to match module name
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
2020-12-08 19:48 ` [RFC v9 01/32] accel/tcg: split CpusAccel into three TCG variants Claudio Fontana
2020-12-08 19:48 ` [RFC v9 02/32] accel/tcg: split tcg_start_vcpu_thread Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 9:10 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 04/32] i386: move kvm accel files into kvm/ Claudio Fontana
` (31 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Philippe Mathieu-Daudé,
Eduardo Habkost, Paul Durrant, Jason Wang, Marcelo Tosatti,
qemu-devel, Peter Xu, Dario Faggioli, Cameron Esfahani,
haxm-team, Claudio Fontana, Anthony Perard, Bruce Rogers,
Olaf Hering, Emilio G . Cota, Colin Xu
Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
accel/tcg/tcg-cpus-icount.c | 24 ++++++------
accel/tcg/tcg-cpus-icount.h | 6 +--
accel/tcg/tcg-cpus-mttcg.c | 10 ++---
accel/tcg/tcg-cpus-rr.c | 74 ++++++++++++++++++-------------------
accel/tcg/tcg-cpus-rr.h | 2 +-
accel/tcg/tcg-cpus.c | 6 +--
accel/tcg/tcg-cpus.h | 6 +--
7 files changed, 64 insertions(+), 64 deletions(-)
diff --git a/accel/tcg/tcg-cpus-icount.c b/accel/tcg/tcg-cpus-icount.c
index 82dbe2cacf..9f45432275 100644
--- a/accel/tcg/tcg-cpus-icount.c
+++ b/accel/tcg/tcg-cpus-icount.c
@@ -36,7 +36,7 @@
#include "tcg-cpus-icount.h"
#include "tcg-cpus-rr.h"
-static int64_t tcg_get_icount_limit(void)
+static int64_t icount_get_limit(void)
{
int64_t deadline;
@@ -68,37 +68,37 @@ static int64_t tcg_get_icount_limit(void)
}
}
-static void notify_aio_contexts(void)
+static void icount_notify_aio_contexts(void)
{
/* Wake up other AioContexts. */
qemu_clock_notify(QEMU_CLOCK_VIRTUAL);
qemu_clock_run_timers(QEMU_CLOCK_VIRTUAL);
}
-void handle_icount_deadline(void)
+void icount_handle_deadline(void)
{
assert(qemu_in_vcpu_thread());
int64_t deadline = qemu_clock_deadline_ns_all(QEMU_CLOCK_VIRTUAL,
QEMU_TIMER_ATTR_ALL);
if (deadline == 0) {
- notify_aio_contexts();
+ icount_notify_aio_contexts();
}
}
-void prepare_icount_for_run(CPUState *cpu)
+void icount_prepare_for_run(CPUState *cpu)
{
int insns_left;
/*
- * These should always be cleared by process_icount_data after
+ * These should always be cleared by icount_process_data after
* each vCPU execution. However u16.high can be raised
- * asynchronously by cpu_exit/cpu_interrupt/tcg_handle_interrupt
+ * asynchronously by cpu_exit/cpu_interrupt/tcg_cpus_handle_interrupt
*/
g_assert(cpu_neg(cpu)->icount_decr.u16.low == 0);
g_assert(cpu->icount_extra == 0);
- cpu->icount_budget = tcg_get_icount_limit();
+ cpu->icount_budget = icount_get_limit();
insns_left = MIN(0xffff, cpu->icount_budget);
cpu_neg(cpu)->icount_decr.u16.low = insns_left;
cpu->icount_extra = cpu->icount_budget - insns_left;
@@ -106,11 +106,11 @@ void prepare_icount_for_run(CPUState *cpu)
replay_mutex_lock();
if (cpu->icount_budget == 0 && replay_has_checkpoint()) {
- notify_aio_contexts();
+ icount_notify_aio_contexts();
}
}
-void process_icount_data(CPUState *cpu)
+void icount_process_data(CPUState *cpu)
{
/* Account for executed instructions */
icount_update(cpu);
@@ -129,7 +129,7 @@ static void icount_handle_interrupt(CPUState *cpu, int mask)
{
int old_mask = cpu->interrupt_request;
- tcg_handle_interrupt(cpu, mask);
+ tcg_cpus_handle_interrupt(cpu, mask);
if (qemu_cpu_is_self(cpu) &&
!cpu->can_do_io
&& (mask & ~old_mask) != 0) {
@@ -139,7 +139,7 @@ static void icount_handle_interrupt(CPUState *cpu, int mask)
const CpusAccel tcg_cpus_icount = {
.create_vcpu_thread = rr_start_vcpu_thread,
- .kick_vcpu_thread = qemu_cpu_kick_rr_cpus,
+ .kick_vcpu_thread = rr_kick_vcpu_thread,
.handle_interrupt = icount_handle_interrupt,
.get_virtual_clock = icount_get,
diff --git a/accel/tcg/tcg-cpus-icount.h b/accel/tcg/tcg-cpus-icount.h
index cbcf76b413..b695939dfa 100644
--- a/accel/tcg/tcg-cpus-icount.h
+++ b/accel/tcg/tcg-cpus-icount.h
@@ -10,8 +10,8 @@
#ifndef TCG_CPUS_ICOUNT_H
#define TCG_CPUS_ICOUNT_H
-void handle_icount_deadline(void);
-void prepare_icount_for_run(CPUState *cpu);
-void process_icount_data(CPUState *cpu);
+void icount_handle_deadline(void);
+void icount_prepare_for_run(CPUState *cpu);
+void icount_process_data(CPUState *cpu);
#endif /* TCG_CPUS_ICOUNT_H */
diff --git a/accel/tcg/tcg-cpus-mttcg.c b/accel/tcg/tcg-cpus-mttcg.c
index f2b892a380..9c3767d260 100644
--- a/accel/tcg/tcg-cpus-mttcg.c
+++ b/accel/tcg/tcg-cpus-mttcg.c
@@ -40,7 +40,7 @@
* current CPUState for a given thread.
*/
-static void *tcg_cpu_thread_fn(void *arg)
+static void *mttcg_cpu_thread_fn(void *arg)
{
CPUState *cpu = arg;
@@ -66,7 +66,7 @@ static void *tcg_cpu_thread_fn(void *arg)
if (cpu_can_run(cpu)) {
int r;
qemu_mutex_unlock_iothread();
- r = tcg_cpu_exec(cpu);
+ r = tcg_cpus_exec(cpu);
qemu_mutex_lock_iothread();
switch (r) {
case EXCP_DEBUG:
@@ -97,7 +97,7 @@ static void *tcg_cpu_thread_fn(void *arg)
qemu_wait_io_event(cpu);
} while (!cpu->unplug || cpu_can_run(cpu));
- qemu_tcg_destroy_vcpu(cpu);
+ tcg_cpus_destroy(cpu);
qemu_mutex_unlock_iothread();
rcu_unregister_thread();
return NULL;
@@ -124,7 +124,7 @@ static void mttcg_start_vcpu_thread(CPUState *cpu)
snprintf(thread_name, VCPU_THREAD_NAME_SIZE, "CPU %d/TCG",
cpu->cpu_index);
- qemu_thread_create(cpu->thread, thread_name, tcg_cpu_thread_fn,
+ qemu_thread_create(cpu->thread, thread_name, mttcg_cpu_thread_fn,
cpu, QEMU_THREAD_JOINABLE);
#ifdef _WIN32
@@ -136,5 +136,5 @@ const CpusAccel tcg_cpus_mttcg = {
.create_vcpu_thread = mttcg_start_vcpu_thread,
.kick_vcpu_thread = mttcg_kick_vcpu_thread,
- .handle_interrupt = tcg_handle_interrupt,
+ .handle_interrupt = tcg_cpus_handle_interrupt,
};
diff --git a/accel/tcg/tcg-cpus-rr.c b/accel/tcg/tcg-cpus-rr.c
index f3b262bec7..0181d2e4eb 100644
--- a/accel/tcg/tcg-cpus-rr.c
+++ b/accel/tcg/tcg-cpus-rr.c
@@ -37,7 +37,7 @@
#include "tcg-cpus-icount.h"
/* Kick all RR vCPUs */
-void qemu_cpu_kick_rr_cpus(CPUState *unused)
+void rr_kick_vcpu_thread(CPUState *unused)
{
CPUState *cpu;
@@ -58,62 +58,62 @@ void qemu_cpu_kick_rr_cpus(CPUState *unused)
* idleness is complete.
*/
-static QEMUTimer *tcg_kick_vcpu_timer;
-static CPUState *tcg_current_rr_cpu;
+static QEMUTimer *rr_kick_vcpu_timer;
+static CPUState *rr_current_cpu;
#define TCG_KICK_PERIOD (NANOSECONDS_PER_SECOND / 10)
-static inline int64_t qemu_tcg_next_kick(void)
+static inline int64_t rr_next_kick_time(void)
{
return qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL) + TCG_KICK_PERIOD;
}
/* Kick the currently round-robin scheduled vCPU to next */
-static void qemu_cpu_kick_rr_next_cpu(void)
+static void rr_kick_next_cpu(void)
{
CPUState *cpu;
do {
- cpu = qatomic_mb_read(&tcg_current_rr_cpu);
+ cpu = qatomic_mb_read(&rr_current_cpu);
if (cpu) {
cpu_exit(cpu);
}
- } while (cpu != qatomic_mb_read(&tcg_current_rr_cpu));
+ } while (cpu != qatomic_mb_read(&rr_current_cpu));
}
-static void kick_tcg_thread(void *opaque)
+static void rr_kick_thread(void *opaque)
{
- timer_mod(tcg_kick_vcpu_timer, qemu_tcg_next_kick());
- qemu_cpu_kick_rr_next_cpu();
+ timer_mod(rr_kick_vcpu_timer, rr_next_kick_time());
+ rr_kick_next_cpu();
}
-static void start_tcg_kick_timer(void)
+static void rr_start_kick_timer(void)
{
- if (!tcg_kick_vcpu_timer && CPU_NEXT(first_cpu)) {
- tcg_kick_vcpu_timer = timer_new_ns(QEMU_CLOCK_VIRTUAL,
- kick_tcg_thread, NULL);
+ if (!rr_kick_vcpu_timer && CPU_NEXT(first_cpu)) {
+ rr_kick_vcpu_timer = timer_new_ns(QEMU_CLOCK_VIRTUAL,
+ rr_kick_thread, NULL);
}
- if (tcg_kick_vcpu_timer && !timer_pending(tcg_kick_vcpu_timer)) {
- timer_mod(tcg_kick_vcpu_timer, qemu_tcg_next_kick());
+ if (rr_kick_vcpu_timer && !timer_pending(rr_kick_vcpu_timer)) {
+ timer_mod(rr_kick_vcpu_timer, rr_next_kick_time());
}
}
-static void stop_tcg_kick_timer(void)
+static void rr_stop_kick_timer(void)
{
- if (tcg_kick_vcpu_timer && timer_pending(tcg_kick_vcpu_timer)) {
- timer_del(tcg_kick_vcpu_timer);
+ if (rr_kick_vcpu_timer && timer_pending(rr_kick_vcpu_timer)) {
+ timer_del(rr_kick_vcpu_timer);
}
}
-static void qemu_tcg_rr_wait_io_event(void)
+static void rr_wait_io_event(void)
{
CPUState *cpu;
while (all_cpu_threads_idle()) {
- stop_tcg_kick_timer();
+ rr_stop_kick_timer();
qemu_cond_wait_iothread(first_cpu->halt_cond);
}
- start_tcg_kick_timer();
+ rr_start_kick_timer();
CPU_FOREACH(cpu) {
qemu_wait_io_event_common(cpu);
@@ -124,13 +124,13 @@ static void qemu_tcg_rr_wait_io_event(void)
* Destroy any remaining vCPUs which have been unplugged and have
* finished running
*/
-static void deal_with_unplugged_cpus(void)
+static void rr_deal_with_unplugged_cpus(void)
{
CPUState *cpu;
CPU_FOREACH(cpu) {
if (cpu->unplug && !cpu_can_run(cpu)) {
- qemu_tcg_destroy_vcpu(cpu);
+ tcg_cpus_destroy(cpu);
break;
}
}
@@ -144,7 +144,7 @@ static void deal_with_unplugged_cpus(void)
* elsewhere.
*/
-static void *tcg_rr_cpu_thread_fn(void *arg)
+static void *rr_cpu_thread_fn(void *arg)
{
CPUState *cpu = arg;
@@ -171,7 +171,7 @@ static void *tcg_rr_cpu_thread_fn(void *arg)
}
}
- start_tcg_kick_timer();
+ rr_start_kick_timer();
cpu = first_cpu;
@@ -190,7 +190,7 @@ static void *tcg_rr_cpu_thread_fn(void *arg)
* Run the timers here. This is much more efficient than
* waking up the I/O thread and waiting for completion.
*/
- handle_icount_deadline();
+ icount_handle_deadline();
}
replay_mutex_unlock();
@@ -201,7 +201,7 @@ static void *tcg_rr_cpu_thread_fn(void *arg)
while (cpu && cpu_work_list_empty(cpu) && !cpu->exit_request) {
- qatomic_mb_set(&tcg_current_rr_cpu, cpu);
+ qatomic_mb_set(&rr_current_cpu, cpu);
current_cpu = cpu;
qemu_clock_enable(QEMU_CLOCK_VIRTUAL,
@@ -212,11 +212,11 @@ static void *tcg_rr_cpu_thread_fn(void *arg)
qemu_mutex_unlock_iothread();
if (icount_enabled()) {
- prepare_icount_for_run(cpu);
+ icount_prepare_for_run(cpu);
}
- r = tcg_cpu_exec(cpu);
+ r = tcg_cpus_exec(cpu);
if (icount_enabled()) {
- process_icount_data(cpu);
+ icount_process_data(cpu);
}
qemu_mutex_lock_iothread();
@@ -240,7 +240,7 @@ static void *tcg_rr_cpu_thread_fn(void *arg)
} /* while (cpu && !cpu->exit_request).. */
/* Does not need qatomic_mb_set because a spurious wakeup is okay. */
- qatomic_set(&tcg_current_rr_cpu, NULL);
+ qatomic_set(&rr_current_cpu, NULL);
if (cpu && cpu->exit_request) {
qatomic_mb_set(&cpu->exit_request, 0);
@@ -254,8 +254,8 @@ static void *tcg_rr_cpu_thread_fn(void *arg)
qemu_notify_event();
}
- qemu_tcg_rr_wait_io_event();
- deal_with_unplugged_cpus();
+ rr_wait_io_event();
+ rr_deal_with_unplugged_cpus();
}
rcu_unregister_thread();
@@ -279,7 +279,7 @@ void rr_start_vcpu_thread(CPUState *cpu)
/* share a single thread for all cpus with TCG */
snprintf(thread_name, VCPU_THREAD_NAME_SIZE, "ALL CPUs/TCG");
qemu_thread_create(cpu->thread, thread_name,
- tcg_rr_cpu_thread_fn,
+ rr_cpu_thread_fn,
cpu, QEMU_THREAD_JOINABLE);
single_tcg_halt_cond = cpu->halt_cond;
@@ -299,7 +299,7 @@ void rr_start_vcpu_thread(CPUState *cpu)
const CpusAccel tcg_cpus_rr = {
.create_vcpu_thread = rr_start_vcpu_thread,
- .kick_vcpu_thread = qemu_cpu_kick_rr_cpus,
+ .kick_vcpu_thread = rr_kick_vcpu_thread,
- .handle_interrupt = tcg_handle_interrupt,
+ .handle_interrupt = tcg_cpus_handle_interrupt,
};
diff --git a/accel/tcg/tcg-cpus-rr.h b/accel/tcg/tcg-cpus-rr.h
index 2e5943eda9..54f6ae6e86 100644
--- a/accel/tcg/tcg-cpus-rr.h
+++ b/accel/tcg/tcg-cpus-rr.h
@@ -13,7 +13,7 @@
#define TCG_KICK_PERIOD (NANOSECONDS_PER_SECOND / 10)
/* Kick all RR vCPUs. */
-void qemu_cpu_kick_rr_cpus(CPUState *unused);
+void rr_kick_vcpu_thread(CPUState *unused);
/* start the round robin vcpu thread */
void rr_start_vcpu_thread(CPUState *cpu);
diff --git a/accel/tcg/tcg-cpus.c b/accel/tcg/tcg-cpus.c
index 86fd09545a..e335f9f155 100644
--- a/accel/tcg/tcg-cpus.c
+++ b/accel/tcg/tcg-cpus.c
@@ -38,12 +38,12 @@
/* common functionality among all TCG variants */
-void qemu_tcg_destroy_vcpu(CPUState *cpu)
+void tcg_cpus_destroy(CPUState *cpu)
{
cpu_thread_signal_destroyed(cpu);
}
-int tcg_cpu_exec(CPUState *cpu)
+int tcg_cpus_exec(CPUState *cpu)
{
int ret;
#ifdef CONFIG_PROFILER
@@ -64,7 +64,7 @@ int tcg_cpu_exec(CPUState *cpu)
}
/* mask must never be zero, except for A20 change call */
-void tcg_handle_interrupt(CPUState *cpu, int mask)
+void tcg_cpus_handle_interrupt(CPUState *cpu, int mask)
{
g_assert(qemu_mutex_iothread_locked());
diff --git a/accel/tcg/tcg-cpus.h b/accel/tcg/tcg-cpus.h
index b7ca954e13..d6893a32f8 100644
--- a/accel/tcg/tcg-cpus.h
+++ b/accel/tcg/tcg-cpus.h
@@ -18,8 +18,8 @@ extern const CpusAccel tcg_cpus_mttcg;
extern const CpusAccel tcg_cpus_icount;
extern const CpusAccel tcg_cpus_rr;
-void qemu_tcg_destroy_vcpu(CPUState *cpu);
-int tcg_cpu_exec(CPUState *cpu);
-void tcg_handle_interrupt(CPUState *cpu, int mask);
+void tcg_cpus_destroy(CPUState *cpu);
+int tcg_cpus_exec(CPUState *cpu);
+void tcg_cpus_handle_interrupt(CPUState *cpu, int mask);
#endif /* TCG_CPUS_H */
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 04/32] i386: move kvm accel files into kvm/
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (2 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 03/32] accel/tcg: rename tcg-cpus functions to match module name Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 9:17 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 05/32] i386: move whpx accel files into whpx/ Claudio Fontana
` (30 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
MAINTAINERS | 2 +-
hw/i386/fw_cfg.c | 2 +-
hw/i386/intel_iommu.c | 2 +-
hw/i386/kvm/apic.c | 2 +-
hw/i386/kvm/clock.c | 2 +-
hw/i386/microvm.c | 2 +-
hw/i386/pc.c | 2 +-
hw/i386/x86.c | 2 +-
meson.build | 1 +
target/i386/cpu.c | 2 +-
target/i386/cpu.h | 2 +-
target/i386/helper.c | 2 +-
target/i386/{ => kvm}/hyperv-proto.h | 0
target/i386/{ => kvm}/hyperv-stub.c | 0
target/i386/{ => kvm}/hyperv.c | 0
target/i386/{ => kvm}/hyperv.h | 0
target/i386/{ => kvm}/kvm-stub.c | 0
target/i386/{ => kvm}/kvm.c | 0
target/i386/{ => kvm}/kvm_i386.h | 0
target/i386/kvm/meson.build | 3 +++
target/i386/kvm/trace-events | 7 +++++++
target/i386/kvm/trace.h | 1 +
target/i386/machine.c | 4 ++--
target/i386/meson.build | 4 +---
target/i386/trace-events | 6 ------
25 files changed, 26 insertions(+), 22 deletions(-)
rename target/i386/{ => kvm}/hyperv-proto.h (100%)
rename target/i386/{ => kvm}/hyperv-stub.c (100%)
rename target/i386/{ => kvm}/hyperv.c (100%)
rename target/i386/{ => kvm}/hyperv.h (100%)
rename target/i386/{ => kvm}/kvm-stub.c (100%)
rename target/i386/{ => kvm}/kvm.c (100%)
rename target/i386/{ => kvm}/kvm_i386.h (100%)
create mode 100644 target/i386/kvm/meson.build
create mode 100644 target/i386/kvm/trace-events
create mode 100644 target/i386/kvm/trace.h
diff --git a/MAINTAINERS b/MAINTAINERS
index 68bc160f41..5b3eced829 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -425,7 +425,7 @@ M: Paolo Bonzini <pbonzini@redhat.com>
M: Marcelo Tosatti <mtosatti@redhat.com>
L: kvm@vger.kernel.org
S: Supported
-F: target/i386/kvm.c
+F: target/i386/kvm/
F: scripts/kvm/vmxcap
Guest CPU Cores (other accelerators)
diff --git a/hw/i386/fw_cfg.c b/hw/i386/fw_cfg.c
index e06579490c..fae1bb380f 100644
--- a/hw/i386/fw_cfg.c
+++ b/hw/i386/fw_cfg.c
@@ -21,7 +21,7 @@
#include "hw/timer/hpet.h"
#include "hw/nvram/fw_cfg.h"
#include "e820_memory_layout.h"
-#include "kvm_i386.h"
+#include "kvm/kvm_i386.h"
#include CONFIG_DEVICES
struct hpet_fw_config hpet_cfg = {.count = UINT8_MAX};
diff --git a/hw/i386/intel_iommu.c b/hw/i386/intel_iommu.c
index 70ac837733..361b6cd238 100644
--- a/hw/i386/intel_iommu.c
+++ b/hw/i386/intel_iommu.c
@@ -37,7 +37,7 @@
#include "sysemu/kvm.h"
#include "sysemu/sysemu.h"
#include "hw/i386/apic_internal.h"
-#include "kvm_i386.h"
+#include "kvm/kvm_i386.h"
#include "migration/vmstate.h"
#include "trace.h"
diff --git a/hw/i386/kvm/apic.c b/hw/i386/kvm/apic.c
index dd29906061..07bebc1282 100644
--- a/hw/i386/kvm/apic.c
+++ b/hw/i386/kvm/apic.c
@@ -17,7 +17,7 @@
#include "hw/pci/msi.h"
#include "sysemu/hw_accel.h"
#include "sysemu/kvm.h"
-#include "target/i386/kvm_i386.h"
+#include "kvm/kvm_i386.h"
static inline void kvm_apic_set_reg(struct kvm_lapic_state *kapic,
int reg_id, uint32_t val)
diff --git a/hw/i386/kvm/clock.c b/hw/i386/kvm/clock.c
index 24fe5091b6..2d8a366369 100644
--- a/hw/i386/kvm/clock.c
+++ b/hw/i386/kvm/clock.c
@@ -20,7 +20,7 @@
#include "sysemu/kvm.h"
#include "sysemu/runstate.h"
#include "sysemu/hw_accel.h"
-#include "kvm_i386.h"
+#include "kvm/kvm_i386.h"
#include "migration/vmstate.h"
#include "hw/sysbus.h"
#include "hw/kvm/clock.h"
diff --git a/hw/i386/microvm.c b/hw/i386/microvm.c
index 5428448b70..3ed6adff83 100644
--- a/hw/i386/microvm.c
+++ b/hw/i386/microvm.c
@@ -51,7 +51,7 @@
#include "cpu.h"
#include "elf.h"
-#include "kvm_i386.h"
+#include "kvm/kvm_i386.h"
#include "hw/xen/start_info.h"
#define MICROVM_QBOOT_FILENAME "qboot.rom"
diff --git a/hw/i386/pc.c b/hw/i386/pc.c
index 17b514d1da..299aaba8e9 100644
--- a/hw/i386/pc.c
+++ b/hw/i386/pc.c
@@ -61,7 +61,7 @@
#include "sysemu/qtest.h"
#include "sysemu/reset.h"
#include "sysemu/runstate.h"
-#include "kvm_i386.h"
+#include "kvm/kvm_i386.h"
#include "hw/xen/xen.h"
#include "hw/xen/start_info.h"
#include "ui/qemu-spice.h"
diff --git a/hw/i386/x86.c b/hw/i386/x86.c
index 5944fc44ed..88d0c70e12 100644
--- a/hw/i386/x86.c
+++ b/hw/i386/x86.c
@@ -52,7 +52,7 @@
#include "elf.h"
#include "standard-headers/asm-x86/bootparam.h"
#include CONFIG_DEVICES
-#include "kvm_i386.h"
+#include "kvm/kvm_i386.h"
#define BIOS_FILENAME "bios.bin"
diff --git a/meson.build b/meson.build
index e3386196ba..198298e9d8 100644
--- a/meson.build
+++ b/meson.build
@@ -1467,6 +1467,7 @@ trace_events_subdirs += [
'target/arm',
'target/hppa',
'target/i386',
+ 'target/i386/kvm',
'target/mips',
'target/ppc',
'target/riscv',
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 5a8c96072e..b9bd249c8f 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -30,7 +30,7 @@
#include "sysemu/hvf.h"
#include "sysemu/cpus.h"
#include "sysemu/xen.h"
-#include "kvm_i386.h"
+#include "kvm/kvm_i386.h"
#include "sev_i386.h"
#include "qemu/error-report.h"
diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index 88e8586f8f..9ecda75aec 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -22,7 +22,7 @@
#include "sysemu/tcg.h"
#include "cpu-qom.h"
-#include "hyperv-proto.h"
+#include "kvm/hyperv-proto.h"
#include "exec/cpu-defs.h"
#include "qapi/qapi-types-common.h"
diff --git a/target/i386/helper.c b/target/i386/helper.c
index 034f46bcc2..a1b3367ab2 100644
--- a/target/i386/helper.c
+++ b/target/i386/helper.c
@@ -24,7 +24,7 @@
#include "qemu/qemu-print.h"
#include "sysemu/kvm.h"
#include "sysemu/runstate.h"
-#include "kvm_i386.h"
+#include "kvm/kvm_i386.h"
#ifndef CONFIG_USER_ONLY
#include "sysemu/tcg.h"
#include "sysemu/hw_accel.h"
diff --git a/target/i386/hyperv-proto.h b/target/i386/kvm/hyperv-proto.h
similarity index 100%
rename from target/i386/hyperv-proto.h
rename to target/i386/kvm/hyperv-proto.h
diff --git a/target/i386/hyperv-stub.c b/target/i386/kvm/hyperv-stub.c
similarity index 100%
rename from target/i386/hyperv-stub.c
rename to target/i386/kvm/hyperv-stub.c
diff --git a/target/i386/hyperv.c b/target/i386/kvm/hyperv.c
similarity index 100%
rename from target/i386/hyperv.c
rename to target/i386/kvm/hyperv.c
diff --git a/target/i386/hyperv.h b/target/i386/kvm/hyperv.h
similarity index 100%
rename from target/i386/hyperv.h
rename to target/i386/kvm/hyperv.h
diff --git a/target/i386/kvm-stub.c b/target/i386/kvm/kvm-stub.c
similarity index 100%
rename from target/i386/kvm-stub.c
rename to target/i386/kvm/kvm-stub.c
diff --git a/target/i386/kvm.c b/target/i386/kvm/kvm.c
similarity index 100%
rename from target/i386/kvm.c
rename to target/i386/kvm/kvm.c
diff --git a/target/i386/kvm_i386.h b/target/i386/kvm/kvm_i386.h
similarity index 100%
rename from target/i386/kvm_i386.h
rename to target/i386/kvm/kvm_i386.h
diff --git a/target/i386/kvm/meson.build b/target/i386/kvm/meson.build
new file mode 100644
index 0000000000..1d66559187
--- /dev/null
+++ b/target/i386/kvm/meson.build
@@ -0,0 +1,3 @@
+i386_ss.add(when: 'CONFIG_KVM', if_false: files('kvm-stub.c'))
+i386_softmmu_ss.add(when: 'CONFIG_KVM', if_true: files('kvm.c'))
+i386_softmmu_ss.add(when: 'CONFIG_HYPERV', if_true: files('hyperv.c'), if_false: files('hyperv-stub.c'))
diff --git a/target/i386/kvm/trace-events b/target/i386/kvm/trace-events
new file mode 100644
index 0000000000..b4e2d9e4ea
--- /dev/null
+++ b/target/i386/kvm/trace-events
@@ -0,0 +1,7 @@
+# See docs/devel/tracing.txt for syntax documentation.
+
+# kvm.c
+kvm_x86_fixup_msi_error(uint32_t gsi) "VT-d failed to remap interrupt for GSI %" PRIu32
+kvm_x86_add_msi_route(int virq) "Adding route entry for virq %d"
+kvm_x86_remove_msi_route(int virq) "Removing route entry for virq %d"
+kvm_x86_update_msi_routes(int num) "Updated %d MSI routes"
diff --git a/target/i386/kvm/trace.h b/target/i386/kvm/trace.h
new file mode 100644
index 0000000000..46b75c6942
--- /dev/null
+++ b/target/i386/kvm/trace.h
@@ -0,0 +1 @@
+#include "trace/trace-target_i386_kvm.h"
diff --git a/target/i386/machine.c b/target/i386/machine.c
index 233e46bb70..1614e8c2f8 100644
--- a/target/i386/machine.c
+++ b/target/i386/machine.c
@@ -3,9 +3,9 @@
#include "exec/exec-all.h"
#include "hw/isa/isa.h"
#include "migration/cpu.h"
-#include "hyperv.h"
+#include "kvm/hyperv.h"
#include "hw/i386/x86.h"
-#include "kvm_i386.h"
+#include "kvm/kvm_i386.h"
#include "sysemu/kvm.h"
#include "sysemu/tcg.h"
diff --git a/target/i386/meson.build b/target/i386/meson.build
index a1a02f3e99..0209542a8a 100644
--- a/target/i386/meson.build
+++ b/target/i386/meson.build
@@ -18,7 +18,6 @@ i386_ss.add(when: 'CONFIG_TCG', if_true: files(
'smm_helper.c',
'svm_helper.c',
'translate.c'), if_false: files('tcg-stub.c'))
-i386_ss.add(when: 'CONFIG_KVM', if_false: files('kvm-stub.c'))
i386_ss.add(when: 'CONFIG_SEV', if_true: files('sev.c'), if_false: files('sev-stub.c'))
i386_softmmu_ss = ss.source_set()
@@ -28,8 +27,6 @@ i386_softmmu_ss.add(files(
'machine.c',
'monitor.c',
))
-i386_softmmu_ss.add(when: 'CONFIG_HYPERV', if_true: files('hyperv.c'), if_false: files('hyperv-stub.c'))
-i386_softmmu_ss.add(when: 'CONFIG_KVM', if_true: files('kvm.c'))
i386_softmmu_ss.add(when: 'CONFIG_WHPX', if_true: files(
'whpx-all.c',
'whpx-cpus.c',
@@ -42,6 +39,7 @@ i386_softmmu_ss.add(when: 'CONFIG_HAX', if_true: files(
i386_softmmu_ss.add(when: ['CONFIG_HAX', 'CONFIG_POSIX'], if_true: files('hax-posix.c'))
i386_softmmu_ss.add(when: ['CONFIG_HAX', 'CONFIG_WIN32'], if_true: files('hax-windows.c'))
+subdir('kvm')
subdir('hvf')
target_arch += {'i386': i386_ss}
diff --git a/target/i386/trace-events b/target/i386/trace-events
index 789c700d4a..d166f9d5e0 100644
--- a/target/i386/trace-events
+++ b/target/i386/trace-events
@@ -1,11 +1,5 @@
# See docs/devel/tracing.txt for syntax documentation.
-# kvm.c
-kvm_x86_fixup_msi_error(uint32_t gsi) "VT-d failed to remap interrupt for GSI %" PRIu32
-kvm_x86_add_msi_route(int virq) "Adding route entry for virq %d"
-kvm_x86_remove_msi_route(int virq) "Removing route entry for virq %d"
-kvm_x86_update_msi_routes(int num) "Updated %d MSI routes"
-
# sev.c
kvm_sev_init(void) ""
kvm_memcrypt_register_region(void *addr, size_t len) "addr %p len 0x%zu"
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 05/32] i386: move whpx accel files into whpx/
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (3 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 04/32] i386: move kvm accel files into kvm/ Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 9:21 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 06/32] i386: move hax accel files into hax/ Claudio Fontana
` (29 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
MAINTAINERS | 5 +----
target/i386/meson.build | 5 +----
target/i386/whpx/meson.build | 4 ++++
target/i386/{ => whpx}/whp-dispatch.h | 0
target/i386/{ => whpx}/whpx-all.c | 0
target/i386/{ => whpx}/whpx-cpus.c | 0
target/i386/{ => whpx}/whpx-cpus.h | 0
7 files changed, 6 insertions(+), 8 deletions(-)
create mode 100644 target/i386/whpx/meson.build
rename target/i386/{ => whpx}/whp-dispatch.h (100%)
rename target/i386/{ => whpx}/whpx-all.c (100%)
rename target/i386/{ => whpx}/whpx-cpus.c (100%)
rename target/i386/{ => whpx}/whpx-cpus.h (100%)
diff --git a/MAINTAINERS b/MAINTAINERS
index 5b3eced829..20e079f40c 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -451,10 +451,7 @@ F: include/sysemu/hvf.h
WHPX CPUs
M: Sunil Muthuswamy <sunilmut@microsoft.com>
S: Supported
-F: target/i386/whpx-all.c
-F: target/i386/whpx-cpus.c
-F: target/i386/whp-dispatch.h
-F: accel/stubs/whpx-stub.c
+F: target/i386/whpx/
F: include/sysemu/whpx.h
Guest CPU Cores (Xen)
diff --git a/target/i386/meson.build b/target/i386/meson.build
index 0209542a8a..62cd042915 100644
--- a/target/i386/meson.build
+++ b/target/i386/meson.build
@@ -27,10 +27,6 @@ i386_softmmu_ss.add(files(
'machine.c',
'monitor.c',
))
-i386_softmmu_ss.add(when: 'CONFIG_WHPX', if_true: files(
- 'whpx-all.c',
- 'whpx-cpus.c',
-))
i386_softmmu_ss.add(when: 'CONFIG_HAX', if_true: files(
'hax-all.c',
'hax-mem.c',
@@ -40,6 +36,7 @@ i386_softmmu_ss.add(when: ['CONFIG_HAX', 'CONFIG_POSIX'], if_true: files('hax-po
i386_softmmu_ss.add(when: ['CONFIG_HAX', 'CONFIG_WIN32'], if_true: files('hax-windows.c'))
subdir('kvm')
+subdir('whpx')
subdir('hvf')
target_arch += {'i386': i386_ss}
diff --git a/target/i386/whpx/meson.build b/target/i386/whpx/meson.build
new file mode 100644
index 0000000000..94a72c8efc
--- /dev/null
+++ b/target/i386/whpx/meson.build
@@ -0,0 +1,4 @@
+i386_softmmu_ss.add(when: 'CONFIG_WHPX', if_true: files(
+ 'whpx-all.c',
+ 'whpx-cpus.c',
+))
diff --git a/target/i386/whp-dispatch.h b/target/i386/whpx/whp-dispatch.h
similarity index 100%
rename from target/i386/whp-dispatch.h
rename to target/i386/whpx/whp-dispatch.h
diff --git a/target/i386/whpx-all.c b/target/i386/whpx/whpx-all.c
similarity index 100%
rename from target/i386/whpx-all.c
rename to target/i386/whpx/whpx-all.c
diff --git a/target/i386/whpx-cpus.c b/target/i386/whpx/whpx-cpus.c
similarity index 100%
rename from target/i386/whpx-cpus.c
rename to target/i386/whpx/whpx-cpus.c
diff --git a/target/i386/whpx-cpus.h b/target/i386/whpx/whpx-cpus.h
similarity index 100%
rename from target/i386/whpx-cpus.h
rename to target/i386/whpx/whpx-cpus.h
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 06/32] i386: move hax accel files into hax/
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (4 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 05/32] i386: move whpx accel files into whpx/ Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 9:22 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 07/32] i386: hvf: remove stale MAINTAINERS entry for old hvf stubs Claudio Fontana
` (28 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
MAINTAINERS | 2 +-
target/i386/{ => hax}/hax-all.c | 0
target/i386/{ => hax}/hax-cpus.c | 0
target/i386/{ => hax}/hax-cpus.h | 0
target/i386/{ => hax}/hax-i386.h | 6 +++---
target/i386/{ => hax}/hax-interface.h | 0
target/i386/{ => hax}/hax-mem.c | 0
target/i386/{ => hax}/hax-posix.c | 0
target/i386/{ => hax}/hax-posix.h | 0
target/i386/{ => hax}/hax-windows.c | 0
target/i386/{ => hax}/hax-windows.h | 0
target/i386/hax/meson.build | 7 +++++++
target/i386/meson.build | 8 +-------
13 files changed, 12 insertions(+), 11 deletions(-)
rename target/i386/{ => hax}/hax-all.c (100%)
rename target/i386/{ => hax}/hax-cpus.c (100%)
rename target/i386/{ => hax}/hax-cpus.h (100%)
rename target/i386/{ => hax}/hax-i386.h (95%)
rename target/i386/{ => hax}/hax-interface.h (100%)
rename target/i386/{ => hax}/hax-mem.c (100%)
rename target/i386/{ => hax}/hax-posix.c (100%)
rename target/i386/{ => hax}/hax-posix.h (100%)
rename target/i386/{ => hax}/hax-windows.c (100%)
rename target/i386/{ => hax}/hax-windows.h (100%)
create mode 100644 target/i386/hax/meson.build
diff --git a/MAINTAINERS b/MAINTAINERS
index 20e079f40c..448593c904 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -491,7 +491,7 @@ W: https://github.com/intel/haxm/issues
S: Maintained
F: accel/stubs/hax-stub.c
F: include/sysemu/hax.h
-F: target/i386/hax-*
+F: target/i386/hax/
Hosts
-----
diff --git a/target/i386/hax-all.c b/target/i386/hax/hax-all.c
similarity index 100%
rename from target/i386/hax-all.c
rename to target/i386/hax/hax-all.c
diff --git a/target/i386/hax-cpus.c b/target/i386/hax/hax-cpus.c
similarity index 100%
rename from target/i386/hax-cpus.c
rename to target/i386/hax/hax-cpus.c
diff --git a/target/i386/hax-cpus.h b/target/i386/hax/hax-cpus.h
similarity index 100%
rename from target/i386/hax-cpus.h
rename to target/i386/hax/hax-cpus.h
diff --git a/target/i386/hax-i386.h b/target/i386/hax/hax-i386.h
similarity index 95%
rename from target/i386/hax-i386.h
rename to target/i386/hax/hax-i386.h
index 48c4abe14e..efbb346238 100644
--- a/target/i386/hax-i386.h
+++ b/target/i386/hax/hax-i386.h
@@ -84,13 +84,13 @@ void hax_memory_init(void);
#ifdef CONFIG_POSIX
-#include "target/i386/hax-posix.h"
+#include "hax-posix.h"
#endif
#ifdef CONFIG_WIN32
-#include "target/i386/hax-windows.h"
+#include "hax-windows.h"
#endif
-#include "target/i386/hax-interface.h"
+#include "hax-interface.h"
#endif
diff --git a/target/i386/hax-interface.h b/target/i386/hax/hax-interface.h
similarity index 100%
rename from target/i386/hax-interface.h
rename to target/i386/hax/hax-interface.h
diff --git a/target/i386/hax-mem.c b/target/i386/hax/hax-mem.c
similarity index 100%
rename from target/i386/hax-mem.c
rename to target/i386/hax/hax-mem.c
diff --git a/target/i386/hax-posix.c b/target/i386/hax/hax-posix.c
similarity index 100%
rename from target/i386/hax-posix.c
rename to target/i386/hax/hax-posix.c
diff --git a/target/i386/hax-posix.h b/target/i386/hax/hax-posix.h
similarity index 100%
rename from target/i386/hax-posix.h
rename to target/i386/hax/hax-posix.h
diff --git a/target/i386/hax-windows.c b/target/i386/hax/hax-windows.c
similarity index 100%
rename from target/i386/hax-windows.c
rename to target/i386/hax/hax-windows.c
diff --git a/target/i386/hax-windows.h b/target/i386/hax/hax-windows.h
similarity index 100%
rename from target/i386/hax-windows.h
rename to target/i386/hax/hax-windows.h
diff --git a/target/i386/hax/meson.build b/target/i386/hax/meson.build
new file mode 100644
index 0000000000..77ea431b30
--- /dev/null
+++ b/target/i386/hax/meson.build
@@ -0,0 +1,7 @@
+i386_softmmu_ss.add(when: 'CONFIG_HAX', if_true: files(
+ 'hax-all.c',
+ 'hax-mem.c',
+ 'hax-cpus.c',
+))
+i386_softmmu_ss.add(when: ['CONFIG_HAX', 'CONFIG_POSIX'], if_true: files('hax-posix.c'))
+i386_softmmu_ss.add(when: ['CONFIG_HAX', 'CONFIG_WIN32'], if_true: files('hax-windows.c'))
diff --git a/target/i386/meson.build b/target/i386/meson.build
index 62cd042915..284d52ab81 100644
--- a/target/i386/meson.build
+++ b/target/i386/meson.build
@@ -27,15 +27,9 @@ i386_softmmu_ss.add(files(
'machine.c',
'monitor.c',
))
-i386_softmmu_ss.add(when: 'CONFIG_HAX', if_true: files(
- 'hax-all.c',
- 'hax-mem.c',
- 'hax-cpus.c',
-))
-i386_softmmu_ss.add(when: ['CONFIG_HAX', 'CONFIG_POSIX'], if_true: files('hax-posix.c'))
-i386_softmmu_ss.add(when: ['CONFIG_HAX', 'CONFIG_WIN32'], if_true: files('hax-windows.c'))
subdir('kvm')
+subdir('hax')
subdir('whpx')
subdir('hvf')
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 07/32] i386: hvf: remove stale MAINTAINERS entry for old hvf stubs
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (5 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 06/32] i386: move hax accel files into hax/ Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 9:22 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 08/32] i386: move TCG accel files into tcg/ Claudio Fontana
` (27 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Roman Bolshakov <r.bolshakov@yadro.com>
---
MAINTAINERS | 1 -
1 file changed, 1 deletion(-)
diff --git a/MAINTAINERS b/MAINTAINERS
index 448593c904..f53f2678d8 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -444,7 +444,6 @@ M: Cameron Esfahani <dirty@apple.com>
M: Roman Bolshakov <r.bolshakov@yadro.com>
W: https://wiki.qemu.org/Features/HVF
S: Maintained
-F: accel/stubs/hvf-stub.c
F: target/i386/hvf/
F: include/sysemu/hvf.h
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 08/32] i386: move TCG accel files into tcg/
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (6 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 07/32] i386: hvf: remove stale MAINTAINERS entry for old hvf stubs Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 9:30 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 09/32] i386: move cpu dump out of helper.c into cpu-dump.c Claudio Fontana
` (26 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
target/i386/meson.build | 14 +-------------
target/i386/{ => tcg}/bpt_helper.c | 0
target/i386/{ => tcg}/cc_helper.c | 0
target/i386/{ => tcg}/excp_helper.c | 0
target/i386/{ => tcg}/fpu_helper.c | 0
target/i386/{ => tcg}/int_helper.c | 0
target/i386/{ => tcg}/mem_helper.c | 0
target/i386/tcg/meson.build | 13 +++++++++++++
target/i386/{ => tcg}/misc_helper.c | 0
target/i386/{ => tcg}/mpx_helper.c | 0
target/i386/{ => tcg}/seg_helper.c | 0
target/i386/{ => tcg}/smm_helper.c | 0
target/i386/{ => tcg}/svm_helper.c | 0
target/i386/{ => tcg}/tcg-stub.c | 0
target/i386/{ => tcg}/translate.c | 0
15 files changed, 14 insertions(+), 13 deletions(-)
rename target/i386/{ => tcg}/bpt_helper.c (100%)
rename target/i386/{ => tcg}/cc_helper.c (100%)
rename target/i386/{ => tcg}/excp_helper.c (100%)
rename target/i386/{ => tcg}/fpu_helper.c (100%)
rename target/i386/{ => tcg}/int_helper.c (100%)
rename target/i386/{ => tcg}/mem_helper.c (100%)
create mode 100644 target/i386/tcg/meson.build
rename target/i386/{ => tcg}/misc_helper.c (100%)
rename target/i386/{ => tcg}/mpx_helper.c (100%)
rename target/i386/{ => tcg}/seg_helper.c (100%)
rename target/i386/{ => tcg}/smm_helper.c (100%)
rename target/i386/{ => tcg}/svm_helper.c (100%)
rename target/i386/{ => tcg}/tcg-stub.c (100%)
rename target/i386/{ => tcg}/translate.c (100%)
diff --git a/target/i386/meson.build b/target/i386/meson.build
index 284d52ab81..750471c9f3 100644
--- a/target/i386/meson.build
+++ b/target/i386/meson.build
@@ -5,19 +5,6 @@ i386_ss.add(files(
'helper.c',
'xsave_helper.c',
))
-i386_ss.add(when: 'CONFIG_TCG', if_true: files(
- 'bpt_helper.c',
- 'cc_helper.c',
- 'excp_helper.c',
- 'fpu_helper.c',
- 'int_helper.c',
- 'mem_helper.c',
- 'misc_helper.c',
- 'mpx_helper.c',
- 'seg_helper.c',
- 'smm_helper.c',
- 'svm_helper.c',
- 'translate.c'), if_false: files('tcg-stub.c'))
i386_ss.add(when: 'CONFIG_SEV', if_true: files('sev.c'), if_false: files('sev-stub.c'))
i386_softmmu_ss = ss.source_set()
@@ -32,6 +19,7 @@ subdir('kvm')
subdir('hax')
subdir('whpx')
subdir('hvf')
+subdir('tcg')
target_arch += {'i386': i386_ss}
target_softmmu_arch += {'i386': i386_softmmu_ss}
diff --git a/target/i386/bpt_helper.c b/target/i386/tcg/bpt_helper.c
similarity index 100%
rename from target/i386/bpt_helper.c
rename to target/i386/tcg/bpt_helper.c
diff --git a/target/i386/cc_helper.c b/target/i386/tcg/cc_helper.c
similarity index 100%
rename from target/i386/cc_helper.c
rename to target/i386/tcg/cc_helper.c
diff --git a/target/i386/excp_helper.c b/target/i386/tcg/excp_helper.c
similarity index 100%
rename from target/i386/excp_helper.c
rename to target/i386/tcg/excp_helper.c
diff --git a/target/i386/fpu_helper.c b/target/i386/tcg/fpu_helper.c
similarity index 100%
rename from target/i386/fpu_helper.c
rename to target/i386/tcg/fpu_helper.c
diff --git a/target/i386/int_helper.c b/target/i386/tcg/int_helper.c
similarity index 100%
rename from target/i386/int_helper.c
rename to target/i386/tcg/int_helper.c
diff --git a/target/i386/mem_helper.c b/target/i386/tcg/mem_helper.c
similarity index 100%
rename from target/i386/mem_helper.c
rename to target/i386/tcg/mem_helper.c
diff --git a/target/i386/tcg/meson.build b/target/i386/tcg/meson.build
new file mode 100644
index 0000000000..02794226c2
--- /dev/null
+++ b/target/i386/tcg/meson.build
@@ -0,0 +1,13 @@
+i386_ss.add(when: 'CONFIG_TCG', if_true: files(
+ 'bpt_helper.c',
+ 'cc_helper.c',
+ 'excp_helper.c',
+ 'fpu_helper.c',
+ 'int_helper.c',
+ 'mem_helper.c',
+ 'misc_helper.c',
+ 'mpx_helper.c',
+ 'seg_helper.c',
+ 'smm_helper.c',
+ 'svm_helper.c',
+ 'translate.c'), if_false: files('tcg-stub.c'))
diff --git a/target/i386/misc_helper.c b/target/i386/tcg/misc_helper.c
similarity index 100%
rename from target/i386/misc_helper.c
rename to target/i386/tcg/misc_helper.c
diff --git a/target/i386/mpx_helper.c b/target/i386/tcg/mpx_helper.c
similarity index 100%
rename from target/i386/mpx_helper.c
rename to target/i386/tcg/mpx_helper.c
diff --git a/target/i386/seg_helper.c b/target/i386/tcg/seg_helper.c
similarity index 100%
rename from target/i386/seg_helper.c
rename to target/i386/tcg/seg_helper.c
diff --git a/target/i386/smm_helper.c b/target/i386/tcg/smm_helper.c
similarity index 100%
rename from target/i386/smm_helper.c
rename to target/i386/tcg/smm_helper.c
diff --git a/target/i386/svm_helper.c b/target/i386/tcg/svm_helper.c
similarity index 100%
rename from target/i386/svm_helper.c
rename to target/i386/tcg/svm_helper.c
diff --git a/target/i386/tcg-stub.c b/target/i386/tcg/tcg-stub.c
similarity index 100%
rename from target/i386/tcg-stub.c
rename to target/i386/tcg/tcg-stub.c
diff --git a/target/i386/translate.c b/target/i386/tcg/translate.c
similarity index 100%
rename from target/i386/translate.c
rename to target/i386/tcg/translate.c
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 09/32] i386: move cpu dump out of helper.c into cpu-dump.c
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (7 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 08/32] i386: move TCG accel files into tcg/ Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 9:59 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 10/32] i386: move TCG cpu class initialization out of helper.c Claudio Fontana
` (25 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
target/i386/cpu-dump.c | 537 ++++++++++++++++++++++++++++++++++++++++
target/i386/cpu.h | 1 +
target/i386/helper.c | 514 --------------------------------------
target/i386/meson.build | 1 +
4 files changed, 539 insertions(+), 514 deletions(-)
create mode 100644 target/i386/cpu-dump.c
diff --git a/target/i386/cpu-dump.c b/target/i386/cpu-dump.c
new file mode 100644
index 0000000000..aac21f1f60
--- /dev/null
+++ b/target/i386/cpu-dump.c
@@ -0,0 +1,537 @@
+/*
+ * i386 CPU dump to FILE
+ *
+ * Copyright (c) 2003 Fabrice Bellard
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "qemu/osdep.h"
+#include "cpu.h"
+#include "qemu/qemu-print.h"
+#ifndef CONFIG_USER_ONLY
+#include "hw/i386/apic_internal.h"
+#endif
+
+/***********************************************************/
+/* x86 debug */
+
+static const char *cc_op_str[CC_OP_NB] = {
+ "DYNAMIC",
+ "EFLAGS",
+
+ "MULB",
+ "MULW",
+ "MULL",
+ "MULQ",
+
+ "ADDB",
+ "ADDW",
+ "ADDL",
+ "ADDQ",
+
+ "ADCB",
+ "ADCW",
+ "ADCL",
+ "ADCQ",
+
+ "SUBB",
+ "SUBW",
+ "SUBL",
+ "SUBQ",
+
+ "SBBB",
+ "SBBW",
+ "SBBL",
+ "SBBQ",
+
+ "LOGICB",
+ "LOGICW",
+ "LOGICL",
+ "LOGICQ",
+
+ "INCB",
+ "INCW",
+ "INCL",
+ "INCQ",
+
+ "DECB",
+ "DECW",
+ "DECL",
+ "DECQ",
+
+ "SHLB",
+ "SHLW",
+ "SHLL",
+ "SHLQ",
+
+ "SARB",
+ "SARW",
+ "SARL",
+ "SARQ",
+
+ "BMILGB",
+ "BMILGW",
+ "BMILGL",
+ "BMILGQ",
+
+ "ADCX",
+ "ADOX",
+ "ADCOX",
+
+ "CLR",
+};
+
+static void
+cpu_x86_dump_seg_cache(CPUX86State *env, FILE *f,
+ const char *name, struct SegmentCache *sc)
+{
+#ifdef TARGET_X86_64
+ if (env->hflags & HF_CS64_MASK) {
+ qemu_fprintf(f, "%-3s=%04x %016" PRIx64 " %08x %08x", name,
+ sc->selector, sc->base, sc->limit,
+ sc->flags & 0x00ffff00);
+ } else
+#endif
+ {
+ qemu_fprintf(f, "%-3s=%04x %08x %08x %08x", name, sc->selector,
+ (uint32_t)sc->base, sc->limit,
+ sc->flags & 0x00ffff00);
+ }
+
+ if (!(env->hflags & HF_PE_MASK) || !(sc->flags & DESC_P_MASK))
+ goto done;
+
+ qemu_fprintf(f, " DPL=%d ",
+ (sc->flags & DESC_DPL_MASK) >> DESC_DPL_SHIFT);
+ if (sc->flags & DESC_S_MASK) {
+ if (sc->flags & DESC_CS_MASK) {
+ qemu_fprintf(f, (sc->flags & DESC_L_MASK) ? "CS64" :
+ ((sc->flags & DESC_B_MASK) ? "CS32" : "CS16"));
+ qemu_fprintf(f, " [%c%c", (sc->flags & DESC_C_MASK) ? 'C' : '-',
+ (sc->flags & DESC_R_MASK) ? 'R' : '-');
+ } else {
+ qemu_fprintf(f, (sc->flags & DESC_B_MASK
+ || env->hflags & HF_LMA_MASK)
+ ? "DS " : "DS16");
+ qemu_fprintf(f, " [%c%c", (sc->flags & DESC_E_MASK) ? 'E' : '-',
+ (sc->flags & DESC_W_MASK) ? 'W' : '-');
+ }
+ qemu_fprintf(f, "%c]", (sc->flags & DESC_A_MASK) ? 'A' : '-');
+ } else {
+ static const char *sys_type_name[2][16] = {
+ { /* 32 bit mode */
+ "Reserved", "TSS16-avl", "LDT", "TSS16-busy",
+ "CallGate16", "TaskGate", "IntGate16", "TrapGate16",
+ "Reserved", "TSS32-avl", "Reserved", "TSS32-busy",
+ "CallGate32", "Reserved", "IntGate32", "TrapGate32"
+ },
+ { /* 64 bit mode */
+ "<hiword>", "Reserved", "LDT", "Reserved", "Reserved",
+ "Reserved", "Reserved", "Reserved", "Reserved",
+ "TSS64-avl", "Reserved", "TSS64-busy", "CallGate64",
+ "Reserved", "IntGate64", "TrapGate64"
+ }
+ };
+ qemu_fprintf(f, "%s",
+ sys_type_name[(env->hflags & HF_LMA_MASK) ? 1 : 0]
+ [(sc->flags & DESC_TYPE_MASK) >> DESC_TYPE_SHIFT]);
+ }
+done:
+ qemu_fprintf(f, "\n");
+}
+
+#ifndef CONFIG_USER_ONLY
+
+/* ARRAY_SIZE check is not required because
+ * DeliveryMode(dm) has a size of 3 bit.
+ */
+static inline const char *dm2str(uint32_t dm)
+{
+ static const char *str[] = {
+ "Fixed",
+ "...",
+ "SMI",
+ "...",
+ "NMI",
+ "INIT",
+ "...",
+ "ExtINT"
+ };
+ return str[dm];
+}
+
+static void dump_apic_lvt(const char *name, uint32_t lvt, bool is_timer)
+{
+ uint32_t dm = (lvt & APIC_LVT_DELIV_MOD) >> APIC_LVT_DELIV_MOD_SHIFT;
+ qemu_printf("%s\t 0x%08x %s %-5s %-6s %-7s %-12s %-6s",
+ name, lvt,
+ lvt & APIC_LVT_INT_POLARITY ? "active-lo" : "active-hi",
+ lvt & APIC_LVT_LEVEL_TRIGGER ? "level" : "edge",
+ lvt & APIC_LVT_MASKED ? "masked" : "",
+ lvt & APIC_LVT_DELIV_STS ? "pending" : "",
+ !is_timer ?
+ "" : lvt & APIC_LVT_TIMER_PERIODIC ?
+ "periodic" : lvt & APIC_LVT_TIMER_TSCDEADLINE ?
+ "tsc-deadline" : "one-shot",
+ dm2str(dm));
+ if (dm != APIC_DM_NMI) {
+ qemu_printf(" (vec %u)\n", lvt & APIC_VECTOR_MASK);
+ } else {
+ qemu_printf("\n");
+ }
+}
+
+/* ARRAY_SIZE check is not required because
+ * destination shorthand has a size of 2 bit.
+ */
+static inline const char *shorthand2str(uint32_t shorthand)
+{
+ const char *str[] = {
+ "no-shorthand", "self", "all-self", "all"
+ };
+ return str[shorthand];
+}
+
+static inline uint8_t divider_conf(uint32_t divide_conf)
+{
+ uint8_t divide_val = ((divide_conf & 0x8) >> 1) | (divide_conf & 0x3);
+
+ return divide_val == 7 ? 1 : 2 << divide_val;
+}
+
+static inline void mask2str(char *str, uint32_t val, uint8_t size)
+{
+ while (size--) {
+ *str++ = (val >> size) & 1 ? '1' : '0';
+ }
+ *str = 0;
+}
+
+#define MAX_LOGICAL_APIC_ID_MASK_SIZE 16
+
+static void dump_apic_icr(APICCommonState *s, CPUX86State *env)
+{
+ uint32_t icr = s->icr[0], icr2 = s->icr[1];
+ uint8_t dest_shorthand = \
+ (icr & APIC_ICR_DEST_SHORT) >> APIC_ICR_DEST_SHORT_SHIFT;
+ bool logical_mod = icr & APIC_ICR_DEST_MOD;
+ char apic_id_str[MAX_LOGICAL_APIC_ID_MASK_SIZE + 1];
+ uint32_t dest_field;
+ bool x2apic;
+
+ qemu_printf("ICR\t 0x%08x %s %s %s %s\n",
+ icr,
+ logical_mod ? "logical" : "physical",
+ icr & APIC_ICR_TRIGGER_MOD ? "level" : "edge",
+ icr & APIC_ICR_LEVEL ? "assert" : "de-assert",
+ shorthand2str(dest_shorthand));
+
+ qemu_printf("ICR2\t 0x%08x", icr2);
+ if (dest_shorthand != 0) {
+ qemu_printf("\n");
+ return;
+ }
+ x2apic = env->features[FEAT_1_ECX] & CPUID_EXT_X2APIC;
+ dest_field = x2apic ? icr2 : icr2 >> APIC_ICR_DEST_SHIFT;
+
+ if (!logical_mod) {
+ if (x2apic) {
+ qemu_printf(" cpu %u (X2APIC ID)\n", dest_field);
+ } else {
+ qemu_printf(" cpu %u (APIC ID)\n",
+ dest_field & APIC_LOGDEST_XAPIC_ID);
+ }
+ return;
+ }
+
+ if (s->dest_mode == 0xf) { /* flat mode */
+ mask2str(apic_id_str, icr2 >> APIC_ICR_DEST_SHIFT, 8);
+ qemu_printf(" mask %s (APIC ID)\n", apic_id_str);
+ } else if (s->dest_mode == 0) { /* cluster mode */
+ if (x2apic) {
+ mask2str(apic_id_str, dest_field & APIC_LOGDEST_X2APIC_ID, 16);
+ qemu_printf(" cluster %u mask %s (X2APIC ID)\n",
+ dest_field >> APIC_LOGDEST_X2APIC_SHIFT, apic_id_str);
+ } else {
+ mask2str(apic_id_str, dest_field & APIC_LOGDEST_XAPIC_ID, 4);
+ qemu_printf(" cluster %u mask %s (APIC ID)\n",
+ dest_field >> APIC_LOGDEST_XAPIC_SHIFT, apic_id_str);
+ }
+ }
+}
+
+static void dump_apic_interrupt(const char *name, uint32_t *ireg_tab,
+ uint32_t *tmr_tab)
+{
+ int i, empty = true;
+
+ qemu_printf("%s\t ", name);
+ for (i = 0; i < 256; i++) {
+ if (apic_get_bit(ireg_tab, i)) {
+ qemu_printf("%u%s ", i,
+ apic_get_bit(tmr_tab, i) ? "(level)" : "");
+ empty = false;
+ }
+ }
+ qemu_printf("%s\n", empty ? "(none)" : "");
+}
+
+void x86_cpu_dump_local_apic_state(CPUState *cs, int flags)
+{
+ X86CPU *cpu = X86_CPU(cs);
+ APICCommonState *s = APIC_COMMON(cpu->apic_state);
+ if (!s) {
+ qemu_printf("local apic state not available\n");
+ return;
+ }
+ uint32_t *lvt = s->lvt;
+
+ qemu_printf("dumping local APIC state for CPU %-2u\n\n",
+ CPU(cpu)->cpu_index);
+ dump_apic_lvt("LVT0", lvt[APIC_LVT_LINT0], false);
+ dump_apic_lvt("LVT1", lvt[APIC_LVT_LINT1], false);
+ dump_apic_lvt("LVTPC", lvt[APIC_LVT_PERFORM], false);
+ dump_apic_lvt("LVTERR", lvt[APIC_LVT_ERROR], false);
+ dump_apic_lvt("LVTTHMR", lvt[APIC_LVT_THERMAL], false);
+ dump_apic_lvt("LVTT", lvt[APIC_LVT_TIMER], true);
+
+ qemu_printf("Timer\t DCR=0x%x (divide by %u) initial_count = %u"
+ " current_count = %u\n",
+ s->divide_conf & APIC_DCR_MASK,
+ divider_conf(s->divide_conf),
+ s->initial_count, apic_get_current_count(s));
+
+ qemu_printf("SPIV\t 0x%08x APIC %s, focus=%s, spurious vec %u\n",
+ s->spurious_vec,
+ s->spurious_vec & APIC_SPURIO_ENABLED ? "enabled" : "disabled",
+ s->spurious_vec & APIC_SPURIO_FOCUS ? "on" : "off",
+ s->spurious_vec & APIC_VECTOR_MASK);
+
+ dump_apic_icr(s, &cpu->env);
+
+ qemu_printf("ESR\t 0x%08x\n", s->esr);
+
+ dump_apic_interrupt("ISR", s->isr, s->tmr);
+ dump_apic_interrupt("IRR", s->irr, s->tmr);
+
+ qemu_printf("\nAPR 0x%02x TPR 0x%02x DFR 0x%02x LDR 0x%02x",
+ s->arb_id, s->tpr, s->dest_mode, s->log_dest);
+ if (s->dest_mode == 0) {
+ qemu_printf("(cluster %u: id %u)",
+ s->log_dest >> APIC_LOGDEST_XAPIC_SHIFT,
+ s->log_dest & APIC_LOGDEST_XAPIC_ID);
+ }
+ qemu_printf(" PPR 0x%02x\n", apic_get_ppr(s));
+}
+#else
+void x86_cpu_dump_local_apic_state(CPUState *cs, int flags)
+{
+}
+#endif /* !CONFIG_USER_ONLY */
+
+#define DUMP_CODE_BYTES_TOTAL 50
+#define DUMP_CODE_BYTES_BACKWARD 20
+
+void x86_cpu_dump_state(CPUState *cs, FILE *f, int flags)
+{
+ X86CPU *cpu = X86_CPU(cs);
+ CPUX86State *env = &cpu->env;
+ int eflags, i, nb;
+ char cc_op_name[32];
+ static const char *seg_name[6] = { "ES", "CS", "SS", "DS", "FS", "GS" };
+
+ eflags = cpu_compute_eflags(env);
+#ifdef TARGET_X86_64
+ if (env->hflags & HF_CS64_MASK) {
+ qemu_fprintf(f, "RAX=%016" PRIx64 " RBX=%016" PRIx64 " RCX=%016" PRIx64 " RDX=%016" PRIx64 "\n"
+ "RSI=%016" PRIx64 " RDI=%016" PRIx64 " RBP=%016" PRIx64 " RSP=%016" PRIx64 "\n"
+ "R8 =%016" PRIx64 " R9 =%016" PRIx64 " R10=%016" PRIx64 " R11=%016" PRIx64 "\n"
+ "R12=%016" PRIx64 " R13=%016" PRIx64 " R14=%016" PRIx64 " R15=%016" PRIx64 "\n"
+ "RIP=%016" PRIx64 " RFL=%08x [%c%c%c%c%c%c%c] CPL=%d II=%d A20=%d SMM=%d HLT=%d\n",
+ env->regs[R_EAX],
+ env->regs[R_EBX],
+ env->regs[R_ECX],
+ env->regs[R_EDX],
+ env->regs[R_ESI],
+ env->regs[R_EDI],
+ env->regs[R_EBP],
+ env->regs[R_ESP],
+ env->regs[8],
+ env->regs[9],
+ env->regs[10],
+ env->regs[11],
+ env->regs[12],
+ env->regs[13],
+ env->regs[14],
+ env->regs[15],
+ env->eip, eflags,
+ eflags & DF_MASK ? 'D' : '-',
+ eflags & CC_O ? 'O' : '-',
+ eflags & CC_S ? 'S' : '-',
+ eflags & CC_Z ? 'Z' : '-',
+ eflags & CC_A ? 'A' : '-',
+ eflags & CC_P ? 'P' : '-',
+ eflags & CC_C ? 'C' : '-',
+ env->hflags & HF_CPL_MASK,
+ (env->hflags >> HF_INHIBIT_IRQ_SHIFT) & 1,
+ (env->a20_mask >> 20) & 1,
+ (env->hflags >> HF_SMM_SHIFT) & 1,
+ cs->halted);
+ } else
+#endif
+ {
+ qemu_fprintf(f, "EAX=%08x EBX=%08x ECX=%08x EDX=%08x\n"
+ "ESI=%08x EDI=%08x EBP=%08x ESP=%08x\n"
+ "EIP=%08x EFL=%08x [%c%c%c%c%c%c%c] CPL=%d II=%d A20=%d SMM=%d HLT=%d\n",
+ (uint32_t)env->regs[R_EAX],
+ (uint32_t)env->regs[R_EBX],
+ (uint32_t)env->regs[R_ECX],
+ (uint32_t)env->regs[R_EDX],
+ (uint32_t)env->regs[R_ESI],
+ (uint32_t)env->regs[R_EDI],
+ (uint32_t)env->regs[R_EBP],
+ (uint32_t)env->regs[R_ESP],
+ (uint32_t)env->eip, eflags,
+ eflags & DF_MASK ? 'D' : '-',
+ eflags & CC_O ? 'O' : '-',
+ eflags & CC_S ? 'S' : '-',
+ eflags & CC_Z ? 'Z' : '-',
+ eflags & CC_A ? 'A' : '-',
+ eflags & CC_P ? 'P' : '-',
+ eflags & CC_C ? 'C' : '-',
+ env->hflags & HF_CPL_MASK,
+ (env->hflags >> HF_INHIBIT_IRQ_SHIFT) & 1,
+ (env->a20_mask >> 20) & 1,
+ (env->hflags >> HF_SMM_SHIFT) & 1,
+ cs->halted);
+ }
+
+ for(i = 0; i < 6; i++) {
+ cpu_x86_dump_seg_cache(env, f, seg_name[i], &env->segs[i]);
+ }
+ cpu_x86_dump_seg_cache(env, f, "LDT", &env->ldt);
+ cpu_x86_dump_seg_cache(env, f, "TR", &env->tr);
+
+#ifdef TARGET_X86_64
+ if (env->hflags & HF_LMA_MASK) {
+ qemu_fprintf(f, "GDT= %016" PRIx64 " %08x\n",
+ env->gdt.base, env->gdt.limit);
+ qemu_fprintf(f, "IDT= %016" PRIx64 " %08x\n",
+ env->idt.base, env->idt.limit);
+ qemu_fprintf(f, "CR0=%08x CR2=%016" PRIx64 " CR3=%016" PRIx64 " CR4=%08x\n",
+ (uint32_t)env->cr[0],
+ env->cr[2],
+ env->cr[3],
+ (uint32_t)env->cr[4]);
+ for(i = 0; i < 4; i++)
+ qemu_fprintf(f, "DR%d=%016" PRIx64 " ", i, env->dr[i]);
+ qemu_fprintf(f, "\nDR6=%016" PRIx64 " DR7=%016" PRIx64 "\n",
+ env->dr[6], env->dr[7]);
+ } else
+#endif
+ {
+ qemu_fprintf(f, "GDT= %08x %08x\n",
+ (uint32_t)env->gdt.base, env->gdt.limit);
+ qemu_fprintf(f, "IDT= %08x %08x\n",
+ (uint32_t)env->idt.base, env->idt.limit);
+ qemu_fprintf(f, "CR0=%08x CR2=%08x CR3=%08x CR4=%08x\n",
+ (uint32_t)env->cr[0],
+ (uint32_t)env->cr[2],
+ (uint32_t)env->cr[3],
+ (uint32_t)env->cr[4]);
+ for(i = 0; i < 4; i++) {
+ qemu_fprintf(f, "DR%d=" TARGET_FMT_lx " ", i, env->dr[i]);
+ }
+ qemu_fprintf(f, "\nDR6=" TARGET_FMT_lx " DR7=" TARGET_FMT_lx "\n",
+ env->dr[6], env->dr[7]);
+ }
+ if (flags & CPU_DUMP_CCOP) {
+ if ((unsigned)env->cc_op < CC_OP_NB)
+ snprintf(cc_op_name, sizeof(cc_op_name), "%s", cc_op_str[env->cc_op]);
+ else
+ snprintf(cc_op_name, sizeof(cc_op_name), "[%d]", env->cc_op);
+#ifdef TARGET_X86_64
+ if (env->hflags & HF_CS64_MASK) {
+ qemu_fprintf(f, "CCS=%016" PRIx64 " CCD=%016" PRIx64 " CCO=%-8s\n",
+ env->cc_src, env->cc_dst,
+ cc_op_name);
+ } else
+#endif
+ {
+ qemu_fprintf(f, "CCS=%08x CCD=%08x CCO=%-8s\n",
+ (uint32_t)env->cc_src, (uint32_t)env->cc_dst,
+ cc_op_name);
+ }
+ }
+ qemu_fprintf(f, "EFER=%016" PRIx64 "\n", env->efer);
+ if (flags & CPU_DUMP_FPU) {
+ int fptag;
+ fptag = 0;
+ for(i = 0; i < 8; i++) {
+ fptag |= ((!env->fptags[i]) << i);
+ }
+ update_mxcsr_from_sse_status(env);
+ qemu_fprintf(f, "FCW=%04x FSW=%04x [ST=%d] FTW=%02x MXCSR=%08x\n",
+ env->fpuc,
+ (env->fpus & ~0x3800) | (env->fpstt & 0x7) << 11,
+ env->fpstt,
+ fptag,
+ env->mxcsr);
+ for(i=0;i<8;i++) {
+ CPU_LDoubleU u;
+ u.d = env->fpregs[i].d;
+ qemu_fprintf(f, "FPR%d=%016" PRIx64 " %04x",
+ i, u.l.lower, u.l.upper);
+ if ((i & 1) == 1)
+ qemu_fprintf(f, "\n");
+ else
+ qemu_fprintf(f, " ");
+ }
+ if (env->hflags & HF_CS64_MASK)
+ nb = 16;
+ else
+ nb = 8;
+ for(i=0;i<nb;i++) {
+ qemu_fprintf(f, "XMM%02d=%08x%08x%08x%08x",
+ i,
+ env->xmm_regs[i].ZMM_L(3),
+ env->xmm_regs[i].ZMM_L(2),
+ env->xmm_regs[i].ZMM_L(1),
+ env->xmm_regs[i].ZMM_L(0));
+ if ((i & 1) == 1)
+ qemu_fprintf(f, "\n");
+ else
+ qemu_fprintf(f, " ");
+ }
+ }
+ if (flags & CPU_DUMP_CODE) {
+ target_ulong base = env->segs[R_CS].base + env->eip;
+ target_ulong offs = MIN(env->eip, DUMP_CODE_BYTES_BACKWARD);
+ uint8_t code;
+ char codestr[3];
+
+ qemu_fprintf(f, "Code=");
+ for (i = 0; i < DUMP_CODE_BYTES_TOTAL; i++) {
+ if (cpu_memory_rw_debug(cs, base - offs + i, &code, 1, 0) == 0) {
+ snprintf(codestr, sizeof(codestr), "%02x", code);
+ } else {
+ snprintf(codestr, sizeof(codestr), "??");
+ }
+ qemu_fprintf(f, "%s%s%s%s", i > 0 ? " " : "",
+ i == offs ? "<" : "", codestr, i == offs ? ">" : "");
+ }
+ qemu_fprintf(f, "\n");
+ }
+}
diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index 9ecda75aec..d6ed45c5d7 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -2232,6 +2232,7 @@ void enable_compat_apic_id_mode(void);
#define APIC_DEFAULT_ADDRESS 0xfee00000
#define APIC_SPACE_SIZE 0x100000
+/* cpu-dump.c */
void x86_cpu_dump_local_apic_state(CPUState *cs, int flags);
/* cpu.c */
diff --git a/target/i386/helper.c b/target/i386/helper.c
index a1b3367ab2..6e7e0f507c 100644
--- a/target/i386/helper.c
+++ b/target/i386/helper.c
@@ -21,8 +21,6 @@
#include "qapi/qapi-events-run-state.h"
#include "cpu.h"
#include "exec/exec-all.h"
-#include "qemu/qemu-print.h"
-#include "sysemu/kvm.h"
#include "sysemu/runstate.h"
#include "kvm/kvm_i386.h"
#ifndef CONFIG_USER_ONLY
@@ -88,518 +86,6 @@ int cpu_x86_support_mca_broadcast(CPUX86State *env)
return 0;
}
-/***********************************************************/
-/* x86 debug */
-
-static const char *cc_op_str[CC_OP_NB] = {
- "DYNAMIC",
- "EFLAGS",
-
- "MULB",
- "MULW",
- "MULL",
- "MULQ",
-
- "ADDB",
- "ADDW",
- "ADDL",
- "ADDQ",
-
- "ADCB",
- "ADCW",
- "ADCL",
- "ADCQ",
-
- "SUBB",
- "SUBW",
- "SUBL",
- "SUBQ",
-
- "SBBB",
- "SBBW",
- "SBBL",
- "SBBQ",
-
- "LOGICB",
- "LOGICW",
- "LOGICL",
- "LOGICQ",
-
- "INCB",
- "INCW",
- "INCL",
- "INCQ",
-
- "DECB",
- "DECW",
- "DECL",
- "DECQ",
-
- "SHLB",
- "SHLW",
- "SHLL",
- "SHLQ",
-
- "SARB",
- "SARW",
- "SARL",
- "SARQ",
-
- "BMILGB",
- "BMILGW",
- "BMILGL",
- "BMILGQ",
-
- "ADCX",
- "ADOX",
- "ADCOX",
-
- "CLR",
-};
-
-static void
-cpu_x86_dump_seg_cache(CPUX86State *env, FILE *f,
- const char *name, struct SegmentCache *sc)
-{
-#ifdef TARGET_X86_64
- if (env->hflags & HF_CS64_MASK) {
- qemu_fprintf(f, "%-3s=%04x %016" PRIx64 " %08x %08x", name,
- sc->selector, sc->base, sc->limit,
- sc->flags & 0x00ffff00);
- } else
-#endif
- {
- qemu_fprintf(f, "%-3s=%04x %08x %08x %08x", name, sc->selector,
- (uint32_t)sc->base, sc->limit,
- sc->flags & 0x00ffff00);
- }
-
- if (!(env->hflags & HF_PE_MASK) || !(sc->flags & DESC_P_MASK))
- goto done;
-
- qemu_fprintf(f, " DPL=%d ",
- (sc->flags & DESC_DPL_MASK) >> DESC_DPL_SHIFT);
- if (sc->flags & DESC_S_MASK) {
- if (sc->flags & DESC_CS_MASK) {
- qemu_fprintf(f, (sc->flags & DESC_L_MASK) ? "CS64" :
- ((sc->flags & DESC_B_MASK) ? "CS32" : "CS16"));
- qemu_fprintf(f, " [%c%c", (sc->flags & DESC_C_MASK) ? 'C' : '-',
- (sc->flags & DESC_R_MASK) ? 'R' : '-');
- } else {
- qemu_fprintf(f, (sc->flags & DESC_B_MASK
- || env->hflags & HF_LMA_MASK)
- ? "DS " : "DS16");
- qemu_fprintf(f, " [%c%c", (sc->flags & DESC_E_MASK) ? 'E' : '-',
- (sc->flags & DESC_W_MASK) ? 'W' : '-');
- }
- qemu_fprintf(f, "%c]", (sc->flags & DESC_A_MASK) ? 'A' : '-');
- } else {
- static const char *sys_type_name[2][16] = {
- { /* 32 bit mode */
- "Reserved", "TSS16-avl", "LDT", "TSS16-busy",
- "CallGate16", "TaskGate", "IntGate16", "TrapGate16",
- "Reserved", "TSS32-avl", "Reserved", "TSS32-busy",
- "CallGate32", "Reserved", "IntGate32", "TrapGate32"
- },
- { /* 64 bit mode */
- "<hiword>", "Reserved", "LDT", "Reserved", "Reserved",
- "Reserved", "Reserved", "Reserved", "Reserved",
- "TSS64-avl", "Reserved", "TSS64-busy", "CallGate64",
- "Reserved", "IntGate64", "TrapGate64"
- }
- };
- qemu_fprintf(f, "%s",
- sys_type_name[(env->hflags & HF_LMA_MASK) ? 1 : 0]
- [(sc->flags & DESC_TYPE_MASK) >> DESC_TYPE_SHIFT]);
- }
-done:
- qemu_fprintf(f, "\n");
-}
-
-#ifndef CONFIG_USER_ONLY
-
-/* ARRAY_SIZE check is not required because
- * DeliveryMode(dm) has a size of 3 bit.
- */
-static inline const char *dm2str(uint32_t dm)
-{
- static const char *str[] = {
- "Fixed",
- "...",
- "SMI",
- "...",
- "NMI",
- "INIT",
- "...",
- "ExtINT"
- };
- return str[dm];
-}
-
-static void dump_apic_lvt(const char *name, uint32_t lvt, bool is_timer)
-{
- uint32_t dm = (lvt & APIC_LVT_DELIV_MOD) >> APIC_LVT_DELIV_MOD_SHIFT;
- qemu_printf("%s\t 0x%08x %s %-5s %-6s %-7s %-12s %-6s",
- name, lvt,
- lvt & APIC_LVT_INT_POLARITY ? "active-lo" : "active-hi",
- lvt & APIC_LVT_LEVEL_TRIGGER ? "level" : "edge",
- lvt & APIC_LVT_MASKED ? "masked" : "",
- lvt & APIC_LVT_DELIV_STS ? "pending" : "",
- !is_timer ?
- "" : lvt & APIC_LVT_TIMER_PERIODIC ?
- "periodic" : lvt & APIC_LVT_TIMER_TSCDEADLINE ?
- "tsc-deadline" : "one-shot",
- dm2str(dm));
- if (dm != APIC_DM_NMI) {
- qemu_printf(" (vec %u)\n", lvt & APIC_VECTOR_MASK);
- } else {
- qemu_printf("\n");
- }
-}
-
-/* ARRAY_SIZE check is not required because
- * destination shorthand has a size of 2 bit.
- */
-static inline const char *shorthand2str(uint32_t shorthand)
-{
- const char *str[] = {
- "no-shorthand", "self", "all-self", "all"
- };
- return str[shorthand];
-}
-
-static inline uint8_t divider_conf(uint32_t divide_conf)
-{
- uint8_t divide_val = ((divide_conf & 0x8) >> 1) | (divide_conf & 0x3);
-
- return divide_val == 7 ? 1 : 2 << divide_val;
-}
-
-static inline void mask2str(char *str, uint32_t val, uint8_t size)
-{
- while (size--) {
- *str++ = (val >> size) & 1 ? '1' : '0';
- }
- *str = 0;
-}
-
-#define MAX_LOGICAL_APIC_ID_MASK_SIZE 16
-
-static void dump_apic_icr(APICCommonState *s, CPUX86State *env)
-{
- uint32_t icr = s->icr[0], icr2 = s->icr[1];
- uint8_t dest_shorthand = \
- (icr & APIC_ICR_DEST_SHORT) >> APIC_ICR_DEST_SHORT_SHIFT;
- bool logical_mod = icr & APIC_ICR_DEST_MOD;
- char apic_id_str[MAX_LOGICAL_APIC_ID_MASK_SIZE + 1];
- uint32_t dest_field;
- bool x2apic;
-
- qemu_printf("ICR\t 0x%08x %s %s %s %s\n",
- icr,
- logical_mod ? "logical" : "physical",
- icr & APIC_ICR_TRIGGER_MOD ? "level" : "edge",
- icr & APIC_ICR_LEVEL ? "assert" : "de-assert",
- shorthand2str(dest_shorthand));
-
- qemu_printf("ICR2\t 0x%08x", icr2);
- if (dest_shorthand != 0) {
- qemu_printf("\n");
- return;
- }
- x2apic = env->features[FEAT_1_ECX] & CPUID_EXT_X2APIC;
- dest_field = x2apic ? icr2 : icr2 >> APIC_ICR_DEST_SHIFT;
-
- if (!logical_mod) {
- if (x2apic) {
- qemu_printf(" cpu %u (X2APIC ID)\n", dest_field);
- } else {
- qemu_printf(" cpu %u (APIC ID)\n",
- dest_field & APIC_LOGDEST_XAPIC_ID);
- }
- return;
- }
-
- if (s->dest_mode == 0xf) { /* flat mode */
- mask2str(apic_id_str, icr2 >> APIC_ICR_DEST_SHIFT, 8);
- qemu_printf(" mask %s (APIC ID)\n", apic_id_str);
- } else if (s->dest_mode == 0) { /* cluster mode */
- if (x2apic) {
- mask2str(apic_id_str, dest_field & APIC_LOGDEST_X2APIC_ID, 16);
- qemu_printf(" cluster %u mask %s (X2APIC ID)\n",
- dest_field >> APIC_LOGDEST_X2APIC_SHIFT, apic_id_str);
- } else {
- mask2str(apic_id_str, dest_field & APIC_LOGDEST_XAPIC_ID, 4);
- qemu_printf(" cluster %u mask %s (APIC ID)\n",
- dest_field >> APIC_LOGDEST_XAPIC_SHIFT, apic_id_str);
- }
- }
-}
-
-static void dump_apic_interrupt(const char *name, uint32_t *ireg_tab,
- uint32_t *tmr_tab)
-{
- int i, empty = true;
-
- qemu_printf("%s\t ", name);
- for (i = 0; i < 256; i++) {
- if (apic_get_bit(ireg_tab, i)) {
- qemu_printf("%u%s ", i,
- apic_get_bit(tmr_tab, i) ? "(level)" : "");
- empty = false;
- }
- }
- qemu_printf("%s\n", empty ? "(none)" : "");
-}
-
-void x86_cpu_dump_local_apic_state(CPUState *cs, int flags)
-{
- X86CPU *cpu = X86_CPU(cs);
- APICCommonState *s = APIC_COMMON(cpu->apic_state);
- if (!s) {
- qemu_printf("local apic state not available\n");
- return;
- }
- uint32_t *lvt = s->lvt;
-
- qemu_printf("dumping local APIC state for CPU %-2u\n\n",
- CPU(cpu)->cpu_index);
- dump_apic_lvt("LVT0", lvt[APIC_LVT_LINT0], false);
- dump_apic_lvt("LVT1", lvt[APIC_LVT_LINT1], false);
- dump_apic_lvt("LVTPC", lvt[APIC_LVT_PERFORM], false);
- dump_apic_lvt("LVTERR", lvt[APIC_LVT_ERROR], false);
- dump_apic_lvt("LVTTHMR", lvt[APIC_LVT_THERMAL], false);
- dump_apic_lvt("LVTT", lvt[APIC_LVT_TIMER], true);
-
- qemu_printf("Timer\t DCR=0x%x (divide by %u) initial_count = %u"
- " current_count = %u\n",
- s->divide_conf & APIC_DCR_MASK,
- divider_conf(s->divide_conf),
- s->initial_count, apic_get_current_count(s));
-
- qemu_printf("SPIV\t 0x%08x APIC %s, focus=%s, spurious vec %u\n",
- s->spurious_vec,
- s->spurious_vec & APIC_SPURIO_ENABLED ? "enabled" : "disabled",
- s->spurious_vec & APIC_SPURIO_FOCUS ? "on" : "off",
- s->spurious_vec & APIC_VECTOR_MASK);
-
- dump_apic_icr(s, &cpu->env);
-
- qemu_printf("ESR\t 0x%08x\n", s->esr);
-
- dump_apic_interrupt("ISR", s->isr, s->tmr);
- dump_apic_interrupt("IRR", s->irr, s->tmr);
-
- qemu_printf("\nAPR 0x%02x TPR 0x%02x DFR 0x%02x LDR 0x%02x",
- s->arb_id, s->tpr, s->dest_mode, s->log_dest);
- if (s->dest_mode == 0) {
- qemu_printf("(cluster %u: id %u)",
- s->log_dest >> APIC_LOGDEST_XAPIC_SHIFT,
- s->log_dest & APIC_LOGDEST_XAPIC_ID);
- }
- qemu_printf(" PPR 0x%02x\n", apic_get_ppr(s));
-}
-#else
-void x86_cpu_dump_local_apic_state(CPUState *cs, int flags)
-{
-}
-#endif /* !CONFIG_USER_ONLY */
-
-#define DUMP_CODE_BYTES_TOTAL 50
-#define DUMP_CODE_BYTES_BACKWARD 20
-
-void x86_cpu_dump_state(CPUState *cs, FILE *f, int flags)
-{
- X86CPU *cpu = X86_CPU(cs);
- CPUX86State *env = &cpu->env;
- int eflags, i, nb;
- char cc_op_name[32];
- static const char *seg_name[6] = { "ES", "CS", "SS", "DS", "FS", "GS" };
-
- eflags = cpu_compute_eflags(env);
-#ifdef TARGET_X86_64
- if (env->hflags & HF_CS64_MASK) {
- qemu_fprintf(f, "RAX=%016" PRIx64 " RBX=%016" PRIx64 " RCX=%016" PRIx64 " RDX=%016" PRIx64 "\n"
- "RSI=%016" PRIx64 " RDI=%016" PRIx64 " RBP=%016" PRIx64 " RSP=%016" PRIx64 "\n"
- "R8 =%016" PRIx64 " R9 =%016" PRIx64 " R10=%016" PRIx64 " R11=%016" PRIx64 "\n"
- "R12=%016" PRIx64 " R13=%016" PRIx64 " R14=%016" PRIx64 " R15=%016" PRIx64 "\n"
- "RIP=%016" PRIx64 " RFL=%08x [%c%c%c%c%c%c%c] CPL=%d II=%d A20=%d SMM=%d HLT=%d\n",
- env->regs[R_EAX],
- env->regs[R_EBX],
- env->regs[R_ECX],
- env->regs[R_EDX],
- env->regs[R_ESI],
- env->regs[R_EDI],
- env->regs[R_EBP],
- env->regs[R_ESP],
- env->regs[8],
- env->regs[9],
- env->regs[10],
- env->regs[11],
- env->regs[12],
- env->regs[13],
- env->regs[14],
- env->regs[15],
- env->eip, eflags,
- eflags & DF_MASK ? 'D' : '-',
- eflags & CC_O ? 'O' : '-',
- eflags & CC_S ? 'S' : '-',
- eflags & CC_Z ? 'Z' : '-',
- eflags & CC_A ? 'A' : '-',
- eflags & CC_P ? 'P' : '-',
- eflags & CC_C ? 'C' : '-',
- env->hflags & HF_CPL_MASK,
- (env->hflags >> HF_INHIBIT_IRQ_SHIFT) & 1,
- (env->a20_mask >> 20) & 1,
- (env->hflags >> HF_SMM_SHIFT) & 1,
- cs->halted);
- } else
-#endif
- {
- qemu_fprintf(f, "EAX=%08x EBX=%08x ECX=%08x EDX=%08x\n"
- "ESI=%08x EDI=%08x EBP=%08x ESP=%08x\n"
- "EIP=%08x EFL=%08x [%c%c%c%c%c%c%c] CPL=%d II=%d A20=%d SMM=%d HLT=%d\n",
- (uint32_t)env->regs[R_EAX],
- (uint32_t)env->regs[R_EBX],
- (uint32_t)env->regs[R_ECX],
- (uint32_t)env->regs[R_EDX],
- (uint32_t)env->regs[R_ESI],
- (uint32_t)env->regs[R_EDI],
- (uint32_t)env->regs[R_EBP],
- (uint32_t)env->regs[R_ESP],
- (uint32_t)env->eip, eflags,
- eflags & DF_MASK ? 'D' : '-',
- eflags & CC_O ? 'O' : '-',
- eflags & CC_S ? 'S' : '-',
- eflags & CC_Z ? 'Z' : '-',
- eflags & CC_A ? 'A' : '-',
- eflags & CC_P ? 'P' : '-',
- eflags & CC_C ? 'C' : '-',
- env->hflags & HF_CPL_MASK,
- (env->hflags >> HF_INHIBIT_IRQ_SHIFT) & 1,
- (env->a20_mask >> 20) & 1,
- (env->hflags >> HF_SMM_SHIFT) & 1,
- cs->halted);
- }
-
- for(i = 0; i < 6; i++) {
- cpu_x86_dump_seg_cache(env, f, seg_name[i], &env->segs[i]);
- }
- cpu_x86_dump_seg_cache(env, f, "LDT", &env->ldt);
- cpu_x86_dump_seg_cache(env, f, "TR", &env->tr);
-
-#ifdef TARGET_X86_64
- if (env->hflags & HF_LMA_MASK) {
- qemu_fprintf(f, "GDT= %016" PRIx64 " %08x\n",
- env->gdt.base, env->gdt.limit);
- qemu_fprintf(f, "IDT= %016" PRIx64 " %08x\n",
- env->idt.base, env->idt.limit);
- qemu_fprintf(f, "CR0=%08x CR2=%016" PRIx64 " CR3=%016" PRIx64 " CR4=%08x\n",
- (uint32_t)env->cr[0],
- env->cr[2],
- env->cr[3],
- (uint32_t)env->cr[4]);
- for(i = 0; i < 4; i++)
- qemu_fprintf(f, "DR%d=%016" PRIx64 " ", i, env->dr[i]);
- qemu_fprintf(f, "\nDR6=%016" PRIx64 " DR7=%016" PRIx64 "\n",
- env->dr[6], env->dr[7]);
- } else
-#endif
- {
- qemu_fprintf(f, "GDT= %08x %08x\n",
- (uint32_t)env->gdt.base, env->gdt.limit);
- qemu_fprintf(f, "IDT= %08x %08x\n",
- (uint32_t)env->idt.base, env->idt.limit);
- qemu_fprintf(f, "CR0=%08x CR2=%08x CR3=%08x CR4=%08x\n",
- (uint32_t)env->cr[0],
- (uint32_t)env->cr[2],
- (uint32_t)env->cr[3],
- (uint32_t)env->cr[4]);
- for(i = 0; i < 4; i++) {
- qemu_fprintf(f, "DR%d=" TARGET_FMT_lx " ", i, env->dr[i]);
- }
- qemu_fprintf(f, "\nDR6=" TARGET_FMT_lx " DR7=" TARGET_FMT_lx "\n",
- env->dr[6], env->dr[7]);
- }
- if (flags & CPU_DUMP_CCOP) {
- if ((unsigned)env->cc_op < CC_OP_NB)
- snprintf(cc_op_name, sizeof(cc_op_name), "%s", cc_op_str[env->cc_op]);
- else
- snprintf(cc_op_name, sizeof(cc_op_name), "[%d]", env->cc_op);
-#ifdef TARGET_X86_64
- if (env->hflags & HF_CS64_MASK) {
- qemu_fprintf(f, "CCS=%016" PRIx64 " CCD=%016" PRIx64 " CCO=%-8s\n",
- env->cc_src, env->cc_dst,
- cc_op_name);
- } else
-#endif
- {
- qemu_fprintf(f, "CCS=%08x CCD=%08x CCO=%-8s\n",
- (uint32_t)env->cc_src, (uint32_t)env->cc_dst,
- cc_op_name);
- }
- }
- qemu_fprintf(f, "EFER=%016" PRIx64 "\n", env->efer);
- if (flags & CPU_DUMP_FPU) {
- int fptag;
- fptag = 0;
- for(i = 0; i < 8; i++) {
- fptag |= ((!env->fptags[i]) << i);
- }
- update_mxcsr_from_sse_status(env);
- qemu_fprintf(f, "FCW=%04x FSW=%04x [ST=%d] FTW=%02x MXCSR=%08x\n",
- env->fpuc,
- (env->fpus & ~0x3800) | (env->fpstt & 0x7) << 11,
- env->fpstt,
- fptag,
- env->mxcsr);
- for(i=0;i<8;i++) {
- CPU_LDoubleU u;
- u.d = env->fpregs[i].d;
- qemu_fprintf(f, "FPR%d=%016" PRIx64 " %04x",
- i, u.l.lower, u.l.upper);
- if ((i & 1) == 1)
- qemu_fprintf(f, "\n");
- else
- qemu_fprintf(f, " ");
- }
- if (env->hflags & HF_CS64_MASK)
- nb = 16;
- else
- nb = 8;
- for(i=0;i<nb;i++) {
- qemu_fprintf(f, "XMM%02d=%08x%08x%08x%08x",
- i,
- env->xmm_regs[i].ZMM_L(3),
- env->xmm_regs[i].ZMM_L(2),
- env->xmm_regs[i].ZMM_L(1),
- env->xmm_regs[i].ZMM_L(0));
- if ((i & 1) == 1)
- qemu_fprintf(f, "\n");
- else
- qemu_fprintf(f, " ");
- }
- }
- if (flags & CPU_DUMP_CODE) {
- target_ulong base = env->segs[R_CS].base + env->eip;
- target_ulong offs = MIN(env->eip, DUMP_CODE_BYTES_BACKWARD);
- uint8_t code;
- char codestr[3];
-
- qemu_fprintf(f, "Code=");
- for (i = 0; i < DUMP_CODE_BYTES_TOTAL; i++) {
- if (cpu_memory_rw_debug(cs, base - offs + i, &code, 1, 0) == 0) {
- snprintf(codestr, sizeof(codestr), "%02x", code);
- } else {
- snprintf(codestr, sizeof(codestr), "??");
- }
- qemu_fprintf(f, "%s%s%s%s", i > 0 ? " " : "",
- i == offs ? "<" : "", codestr, i == offs ? ">" : "");
- }
- qemu_fprintf(f, "\n");
- }
-}
-
/***********************************************************/
/* x86 mmu */
/* XXX: add PGE support */
diff --git a/target/i386/meson.build b/target/i386/meson.build
index 750471c9f3..c4bf20b319 100644
--- a/target/i386/meson.build
+++ b/target/i386/meson.build
@@ -4,6 +4,7 @@ i386_ss.add(files(
'gdbstub.c',
'helper.c',
'xsave_helper.c',
+ 'cpu-dump.c',
))
i386_ss.add(when: 'CONFIG_SEV', if_true: files('sev.c'), if_false: files('sev-stub.c'))
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 10/32] i386: move TCG cpu class initialization out of helper.c
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (8 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 09/32] i386: move cpu dump out of helper.c into cpu-dump.c Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 10:23 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 11/32] tcg: cpu_exec_{enter,exit} helpers Claudio Fontana
` (24 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
target/i386/cpu.c | 33 ++++------
target/i386/cpu.h | 97 ++---------------------------
target/i386/helper-tcg.h | 112 ++++++++++++++++++++++++++++++++++
target/i386/helper.c | 23 -------
target/i386/meson.build | 1 +
target/i386/tcg-cpu.c | 71 +++++++++++++++++++++
target/i386/tcg-cpu.h | 15 +++++
target/i386/tcg/bpt_helper.c | 1 +
target/i386/tcg/cc_helper.c | 1 +
target/i386/tcg/excp_helper.c | 1 +
target/i386/tcg/fpu_helper.c | 33 +++++-----
target/i386/tcg/int_helper.c | 1 +
target/i386/tcg/mem_helper.c | 1 +
target/i386/tcg/misc_helper.c | 1 +
target/i386/tcg/mpx_helper.c | 1 +
target/i386/tcg/seg_helper.c | 1 +
target/i386/tcg/smm_helper.c | 2 +
target/i386/tcg/svm_helper.c | 1 +
target/i386/tcg/translate.c | 1 +
19 files changed, 244 insertions(+), 153 deletions(-)
create mode 100644 target/i386/helper-tcg.h
create mode 100644 target/i386/tcg-cpu.c
create mode 100644 target/i386/tcg-cpu.h
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index b9bd249c8f..3462d0143f 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -24,6 +24,8 @@
#include "qemu/qemu-print.h"
#include "cpu.h"
+#include "tcg-cpu.h"
+#include "helper-tcg.h"
#include "exec/exec-all.h"
#include "sysemu/kvm.h"
#include "sysemu/reset.h"
@@ -1495,7 +1497,8 @@ static inline uint64_t x86_cpu_xsave_components(X86CPU *cpu)
cpu->env.features[FEAT_XSAVE_COMP_LO];
}
-const char *get_register_name_32(unsigned int reg)
+/* Return name of 32-bit register, from a R_* constant */
+static const char *get_register_name_32(unsigned int reg)
{
if (reg >= CPU_NB_REGS32) {
return NULL;
@@ -7012,13 +7015,6 @@ static void x86_cpu_set_pc(CPUState *cs, vaddr value)
cpu->env.eip = value;
}
-static void x86_cpu_synchronize_from_tb(CPUState *cs, TranslationBlock *tb)
-{
- X86CPU *cpu = X86_CPU(cs);
-
- cpu->env.eip = tb->pc - tb->cs_base;
-}
-
int x86_cpu_pending_interrupt(CPUState *cs, int interrupt_request)
{
X86CPU *cpu = X86_CPU(cs);
@@ -7252,17 +7248,18 @@ static void x86_cpu_common_class_init(ObjectClass *oc, void *data)
cc->class_by_name = x86_cpu_class_by_name;
cc->parse_features = x86_cpu_parse_featurestr;
cc->has_work = x86_cpu_has_work;
+
#ifdef CONFIG_TCG
- cc->do_interrupt = x86_cpu_do_interrupt;
- cc->cpu_exec_interrupt = x86_cpu_exec_interrupt;
-#endif
+ tcg_cpu_common_class_init(cc);
+#endif /* CONFIG_TCG */
+
cc->dump_state = x86_cpu_dump_state;
cc->set_pc = x86_cpu_set_pc;
- cc->synchronize_from_tb = x86_cpu_synchronize_from_tb;
cc->gdb_read_register = x86_cpu_gdb_read_register;
cc->gdb_write_register = x86_cpu_gdb_write_register;
cc->get_arch_id = x86_cpu_get_arch_id;
cc->get_paging_enabled = x86_cpu_get_paging_enabled;
+
#ifndef CONFIG_USER_ONLY
cc->asidx_from_attrs = x86_asidx_from_attrs;
cc->get_memory_mapping = x86_cpu_get_memory_mapping;
@@ -7273,7 +7270,8 @@ static void x86_cpu_common_class_init(ObjectClass *oc, void *data)
cc->write_elf32_note = x86_cpu_write_elf32_note;
cc->write_elf32_qemunote = x86_cpu_write_elf32_qemunote;
cc->vmsd = &vmstate_x86_cpu;
-#endif
+#endif /* !CONFIG_USER_ONLY */
+
cc->gdb_arch_name = x86_gdb_arch_name;
#ifdef TARGET_X86_64
cc->gdb_core_xml_file = "i386-64bit.xml";
@@ -7281,15 +7279,6 @@ static void x86_cpu_common_class_init(ObjectClass *oc, void *data)
#else
cc->gdb_core_xml_file = "i386-32bit.xml";
cc->gdb_num_core_regs = 50;
-#endif
-#if defined(CONFIG_TCG) && !defined(CONFIG_USER_ONLY)
- cc->debug_excp_handler = breakpoint_handler;
-#endif
- cc->cpu_exec_enter = x86_cpu_exec_enter;
- cc->cpu_exec_exit = x86_cpu_exec_exit;
-#ifdef CONFIG_TCG
- cc->tcg_initialize = tcg_x86_init;
- cc->tlb_fill = x86_cpu_tlb_fill;
#endif
cc->disas_set_info = x86_disas_set_info;
diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index d6ed45c5d7..a0d64613dc 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -31,9 +31,6 @@
#define KVM_HAVE_MCE_INJECTION 1
-/* Maximum instruction code size */
-#define TARGET_MAX_INSN_SIZE 16
-
/* support for self modifying code even if the modified instruction is
close to the modifying instruction */
#define TARGET_HAS_PRECISE_SMC
@@ -1037,6 +1034,12 @@ typedef uint64_t FeatureWordArray[FEATURE_WORDS];
* using this information. Condition codes are not generated if they
* are only needed for conditional branches.
*/
+
+#define CC_DST (env->cc_dst)
+#define CC_SRC (env->cc_src)
+#define CC_SRC2 (env->cc_src2)
+#define CC_OP (env->cc_op)
+
typedef enum {
CC_OP_DYNAMIC, /* must use dynamic code to get cc_op */
CC_OP_EFLAGS, /* all cc are explicitly computed, CC_SRC = flags */
@@ -1765,12 +1768,6 @@ struct X86CPU {
extern VMStateDescription vmstate_x86_cpu;
#endif
-/**
- * x86_cpu_do_interrupt:
- * @cpu: vCPU the interrupt is to be handled by.
- */
-void x86_cpu_do_interrupt(CPUState *cpu);
-bool x86_cpu_exec_interrupt(CPUState *cpu, int int_req);
int x86_cpu_pending_interrupt(CPUState *cs, int interrupt_request);
int x86_cpu_write_elf64_note(WriteCoreDumpFunction f, CPUState *cpu,
@@ -1793,9 +1790,6 @@ hwaddr x86_cpu_get_phys_page_attrs_debug(CPUState *cpu, vaddr addr,
int x86_cpu_gdb_read_register(CPUState *cpu, GByteArray *buf, int reg);
int x86_cpu_gdb_write_register(CPUState *cpu, uint8_t *buf, int reg);
-void x86_cpu_exec_enter(CPUState *cpu);
-void x86_cpu_exec_exit(CPUState *cpu);
-
void x86_cpu_list(void);
int cpu_x86_support_mca_broadcast(CPUX86State *env);
@@ -1920,9 +1914,6 @@ void host_cpuid(uint32_t function, uint32_t count,
void host_vendor_fms(char *vendor, int *family, int *model, int *stepping);
/* helper.c */
-bool x86_cpu_tlb_fill(CPUState *cs, vaddr address, int size,
- MMUAccessType access_type, int mmu_idx,
- bool probe, uintptr_t retaddr);
void x86_cpu_set_a20(X86CPU *cpu, int a20_state);
#ifndef CONFIG_USER_ONLY
@@ -1947,8 +1938,6 @@ void x86_stl_phys(CPUState *cs, hwaddr addr, uint32_t val);
void x86_stq_phys(CPUState *cs, hwaddr addr, uint64_t val);
#endif
-void breakpoint_handler(CPUState *cs);
-
/* will be suppressed */
void cpu_x86_update_cr0(CPUX86State *env, uint32_t new_cr0);
void cpu_x86_update_cr3(CPUX86State *env, target_ulong new_cr3);
@@ -1958,16 +1947,6 @@ void cpu_x86_update_dr7(CPUX86State *env, uint32_t new_dr7);
/* hw/pc.c */
uint64_t cpu_get_tsc(CPUX86State *env);
-/* XXX: This value should match the one returned by CPUID
- * and in exec.c */
-# if defined(TARGET_X86_64)
-# define TCG_PHYS_ADDR_BITS 40
-# else
-# define TCG_PHYS_ADDR_BITS 36
-# endif
-
-#define PHYS_ADDR_MASK MAKE_64BIT_MASK(0, TCG_PHYS_ADDR_BITS)
-
#define X86_CPU_TYPE_SUFFIX "-" TYPE_X86_CPU
#define X86_CPU_TYPE_NAME(name) (name X86_CPU_TYPE_SUFFIX)
#define CPU_RESOLVING_TYPE TYPE_X86_CPU
@@ -1999,30 +1978,6 @@ static inline int cpu_mmu_index_kernel(CPUX86State *env)
? MMU_KNOSMAP_IDX : MMU_KSMAP_IDX;
}
-#define CC_DST (env->cc_dst)
-#define CC_SRC (env->cc_src)
-#define CC_SRC2 (env->cc_src2)
-#define CC_OP (env->cc_op)
-
-/* n must be a constant to be efficient */
-static inline target_long lshift(target_long x, int n)
-{
- if (n >= 0) {
- return x << n;
- } else {
- return x >> (-n);
- }
-}
-
-/* float macros */
-#define FT0 (env->ft0)
-#define ST0 (env->fpregs[env->fpstt].d)
-#define ST(n) (env->fpregs[(env->fpstt + (n)) & 7].d)
-#define ST1 ST(1)
-
-/* translate.c */
-void tcg_x86_init(void);
-
typedef CPUX86State CPUArchState;
typedef X86CPU ArchCPU;
@@ -2052,19 +2007,6 @@ void cpu_x86_inject_mce(Monitor *mon, X86CPU *cpu, int bank,
uint64_t status, uint64_t mcg_status, uint64_t addr,
uint64_t misc, int flags);
-/* excp_helper.c */
-void QEMU_NORETURN raise_exception(CPUX86State *env, int exception_index);
-void QEMU_NORETURN raise_exception_ra(CPUX86State *env, int exception_index,
- uintptr_t retaddr);
-void QEMU_NORETURN raise_exception_err(CPUX86State *env, int exception_index,
- int error_code);
-void QEMU_NORETURN raise_exception_err_ra(CPUX86State *env, int exception_index,
- int error_code, uintptr_t retaddr);
-void QEMU_NORETURN raise_interrupt(CPUX86State *nenv, int intno, int is_int,
- int error_code, int next_eip_addend);
-
-/* cc_helper.c */
-extern const uint8_t parity_table[256];
uint32_t cpu_cc_compute_all(CPUX86State *env1, int op);
static inline uint32_t cpu_compute_eflags(CPUX86State *env)
@@ -2076,18 +2018,6 @@ static inline uint32_t cpu_compute_eflags(CPUX86State *env)
return eflags;
}
-/* NOTE: the translator must set DisasContext.cc_op to CC_OP_EFLAGS
- * after generating a call to a helper that uses this.
- */
-static inline void cpu_load_eflags(CPUX86State *env, int eflags,
- int update_mask)
-{
- CC_SRC = eflags & (CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
- CC_OP = CC_OP_EFLAGS;
- env->df = 1 - (2 * ((eflags >> 10) & 1));
- env->eflags = (env->eflags & ~update_mask) |
- (eflags & update_mask) | 0x2;
-}
/* load efer and update the corresponding hflags. XXX: do consistency
checks with cpuid bits? */
@@ -2176,16 +2106,6 @@ void helper_lock_init(void);
/* svm_helper.c */
void cpu_svm_check_intercept_param(CPUX86State *env1, uint32_t type,
uint64_t param, uintptr_t retaddr);
-void QEMU_NORETURN cpu_vmexit(CPUX86State *nenv, uint32_t exit_code,
- uint64_t exit_info_1, uintptr_t retaddr);
-void do_vmexit(CPUX86State *env, uint32_t exit_code, uint64_t exit_info_1);
-
-/* seg_helper.c */
-void do_interrupt_x86_hardirq(CPUX86State *env, int intno, int is_hw);
-
-/* smm_helper.c */
-void do_smm_enter(X86CPU *cpu);
-
/* apic.c */
void cpu_report_tpr_access(CPUX86State *env, TPRAccess access);
void apic_handle_tpr_access_report(DeviceState *d, target_ulong ip,
@@ -2224,11 +2144,6 @@ typedef int X86CPUVersion;
*/
void x86_cpu_set_default_version(X86CPUVersion version);
-/* Return name of 32-bit register, from a R_* constant */
-const char *get_register_name_32(unsigned int reg);
-
-void enable_compat_apic_id_mode(void);
-
#define APIC_DEFAULT_ADDRESS 0xfee00000
#define APIC_SPACE_SIZE 0x100000
diff --git a/target/i386/helper-tcg.h b/target/i386/helper-tcg.h
new file mode 100644
index 0000000000..57b4391a7d
--- /dev/null
+++ b/target/i386/helper-tcg.h
@@ -0,0 +1,112 @@
+/*
+ * TCG specific prototypes for helpers
+ *
+ * Copyright (c) 2003 Fabrice Bellard
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef I386_HELPER_TCG_H
+#define I386_HELPER_TCG_H
+
+#include "exec/exec-all.h"
+
+/* Maximum instruction code size */
+#define TARGET_MAX_INSN_SIZE 16
+
+/*
+ * XXX: This value should match the one returned by CPUID
+ * and in exec.c
+ */
+# if defined(TARGET_X86_64)
+# define TCG_PHYS_ADDR_BITS 40
+# else
+# define TCG_PHYS_ADDR_BITS 36
+# endif
+
+#define PHYS_ADDR_MASK MAKE_64BIT_MASK(0, TCG_PHYS_ADDR_BITS)
+
+/**
+ * x86_cpu_do_interrupt:
+ * @cpu: vCPU the interrupt is to be handled by.
+ */
+void x86_cpu_do_interrupt(CPUState *cpu);
+bool x86_cpu_exec_interrupt(CPUState *cpu, int int_req);
+
+/* helper.c */
+bool x86_cpu_tlb_fill(CPUState *cs, vaddr address, int size,
+ MMUAccessType access_type, int mmu_idx,
+ bool probe, uintptr_t retaddr);
+
+void breakpoint_handler(CPUState *cs);
+
+/* n must be a constant to be efficient */
+static inline target_long lshift(target_long x, int n)
+{
+ if (n >= 0) {
+ return x << n;
+ } else {
+ return x >> (-n);
+ }
+}
+
+/* float macros */
+#define FT0 (env->ft0)
+#define ST0 (env->fpregs[env->fpstt].d)
+#define ST(n) (env->fpregs[(env->fpstt + (n)) & 7].d)
+#define ST1 ST(1)
+
+/* translate.c */
+void tcg_x86_init(void);
+
+/* excp_helper.c */
+void QEMU_NORETURN raise_exception(CPUX86State *env, int exception_index);
+void QEMU_NORETURN raise_exception_ra(CPUX86State *env, int exception_index,
+ uintptr_t retaddr);
+void QEMU_NORETURN raise_exception_err(CPUX86State *env, int exception_index,
+ int error_code);
+void QEMU_NORETURN raise_exception_err_ra(CPUX86State *env, int exception_index,
+ int error_code, uintptr_t retaddr);
+void QEMU_NORETURN raise_interrupt(CPUX86State *nenv, int intno, int is_int,
+ int error_code, int next_eip_addend);
+
+/* cc_helper.c */
+extern const uint8_t parity_table[256];
+
+/*
+ * NOTE: the translator must set DisasContext.cc_op to CC_OP_EFLAGS
+ * after generating a call to a helper that uses this.
+ */
+static inline void cpu_load_eflags(CPUX86State *env, int eflags,
+ int update_mask)
+{
+ CC_SRC = eflags & (CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
+ CC_OP = CC_OP_EFLAGS;
+ env->df = 1 - (2 * ((eflags >> 10) & 1));
+ env->eflags = (env->eflags & ~update_mask) |
+ (eflags & update_mask) | 0x2;
+}
+
+/* svm_helper.c */
+void QEMU_NORETURN cpu_vmexit(CPUX86State *nenv, uint32_t exit_code,
+ uint64_t exit_info_1, uintptr_t retaddr);
+void do_vmexit(CPUX86State *env, uint32_t exit_code, uint64_t exit_info_1);
+
+/* seg_helper.c */
+void do_interrupt_x86_hardirq(CPUX86State *env, int intno, int is_hw);
+
+/* smm_helper.c */
+void do_smm_enter(X86CPU *cpu);
+
+#endif /* I386_HELPER_TCG_H */
diff --git a/target/i386/helper.c b/target/i386/helper.c
index 6e7e0f507c..6bb0c53182 100644
--- a/target/i386/helper.c
+++ b/target/i386/helper.c
@@ -24,10 +24,8 @@
#include "sysemu/runstate.h"
#include "kvm/kvm_i386.h"
#ifndef CONFIG_USER_ONLY
-#include "sysemu/tcg.h"
#include "sysemu/hw_accel.h"
#include "monitor/monitor.h"
-#include "hw/i386/apic_internal.h"
#endif
void cpu_sync_bndcs_hflags(CPUX86State *env)
@@ -572,27 +570,6 @@ void do_cpu_sipi(X86CPU *cpu)
}
#endif
-/* Frob eflags into and out of the CPU temporary format. */
-
-void x86_cpu_exec_enter(CPUState *cs)
-{
- X86CPU *cpu = X86_CPU(cs);
- CPUX86State *env = &cpu->env;
-
- CC_SRC = env->eflags & (CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
- env->df = 1 - (2 * ((env->eflags >> 10) & 1));
- CC_OP = CC_OP_EFLAGS;
- env->eflags &= ~(DF_MASK | CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
-}
-
-void x86_cpu_exec_exit(CPUState *cs)
-{
- X86CPU *cpu = X86_CPU(cs);
- CPUX86State *env = &cpu->env;
-
- env->eflags = cpu_compute_eflags(env);
-}
-
#ifndef CONFIG_USER_ONLY
uint8_t x86_ldub_phys(CPUState *cs, hwaddr addr)
{
diff --git a/target/i386/meson.build b/target/i386/meson.build
index c4bf20b319..9c20208e5a 100644
--- a/target/i386/meson.build
+++ b/target/i386/meson.build
@@ -6,6 +6,7 @@ i386_ss.add(files(
'xsave_helper.c',
'cpu-dump.c',
))
+i386_ss.add(when: 'CONFIG_TCG', if_true: files('tcg-cpu.c'))
i386_ss.add(when: 'CONFIG_SEV', if_true: files('sev.c'), if_false: files('sev-stub.c'))
i386_softmmu_ss = ss.source_set()
diff --git a/target/i386/tcg-cpu.c b/target/i386/tcg-cpu.c
new file mode 100644
index 0000000000..628dd29fe7
--- /dev/null
+++ b/target/i386/tcg-cpu.c
@@ -0,0 +1,71 @@
+/*
+ * i386 TCG cpu class initialization
+ *
+ * Copyright (c) 2003 Fabrice Bellard
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "qemu/osdep.h"
+#include "cpu.h"
+#include "tcg-cpu.h"
+#include "exec/exec-all.h"
+#include "sysemu/runstate.h"
+#include "helper-tcg.h"
+
+#if !defined(CONFIG_USER_ONLY)
+#include "hw/i386/apic.h"
+#endif
+
+/* Frob eflags into and out of the CPU temporary format. */
+
+static void x86_cpu_exec_enter(CPUState *cs)
+{
+ X86CPU *cpu = X86_CPU(cs);
+ CPUX86State *env = &cpu->env;
+
+ CC_SRC = env->eflags & (CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
+ env->df = 1 - (2 * ((env->eflags >> 10) & 1));
+ CC_OP = CC_OP_EFLAGS;
+ env->eflags &= ~(DF_MASK | CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
+}
+
+static void x86_cpu_exec_exit(CPUState *cs)
+{
+ X86CPU *cpu = X86_CPU(cs);
+ CPUX86State *env = &cpu->env;
+
+ env->eflags = cpu_compute_eflags(env);
+}
+
+static void x86_cpu_synchronize_from_tb(CPUState *cs, TranslationBlock *tb)
+{
+ X86CPU *cpu = X86_CPU(cs);
+
+ cpu->env.eip = tb->pc - tb->cs_base;
+}
+
+void tcg_cpu_common_class_init(CPUClass *cc)
+{
+ cc->do_interrupt = x86_cpu_do_interrupt;
+ cc->cpu_exec_interrupt = x86_cpu_exec_interrupt;
+ cc->synchronize_from_tb = x86_cpu_synchronize_from_tb;
+ cc->cpu_exec_enter = x86_cpu_exec_enter;
+ cc->cpu_exec_exit = x86_cpu_exec_exit;
+ cc->tcg_initialize = tcg_x86_init;
+ cc->tlb_fill = x86_cpu_tlb_fill;
+#ifndef CONFIG_USER_ONLY
+ cc->debug_excp_handler = breakpoint_handler;
+#endif
+}
diff --git a/target/i386/tcg-cpu.h b/target/i386/tcg-cpu.h
new file mode 100644
index 0000000000..81f02e562e
--- /dev/null
+++ b/target/i386/tcg-cpu.h
@@ -0,0 +1,15 @@
+/*
+ * i386 TCG CPU class initialization
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#ifndef TCG_CPU_H
+#define TCG_CPU_H
+
+void tcg_cpu_common_class_init(CPUClass *cc);
+
+#endif /* TCG_CPU_H */
diff --git a/target/i386/tcg/bpt_helper.c b/target/i386/tcg/bpt_helper.c
index e6cc2921e2..979230ac12 100644
--- a/target/i386/tcg/bpt_helper.c
+++ b/target/i386/tcg/bpt_helper.c
@@ -21,6 +21,7 @@
#include "cpu.h"
#include "exec/exec-all.h"
#include "exec/helper-proto.h"
+#include "helper-tcg.h"
#ifndef CONFIG_USER_ONLY
diff --git a/target/i386/tcg/cc_helper.c b/target/i386/tcg/cc_helper.c
index 924dd3cd57..cc7ea9e8b9 100644
--- a/target/i386/tcg/cc_helper.c
+++ b/target/i386/tcg/cc_helper.c
@@ -20,6 +20,7 @@
#include "qemu/osdep.h"
#include "cpu.h"
#include "exec/helper-proto.h"
+#include "helper-tcg.h"
const uint8_t parity_table[256] = {
CC_P, 0, 0, CC_P, 0, CC_P, CC_P, 0,
diff --git a/target/i386/tcg/excp_helper.c b/target/i386/tcg/excp_helper.c
index 191471749f..a0f44431fe 100644
--- a/target/i386/tcg/excp_helper.c
+++ b/target/i386/tcg/excp_helper.c
@@ -23,6 +23,7 @@
#include "qemu/log.h"
#include "sysemu/runstate.h"
#include "exec/helper-proto.h"
+#include "helper-tcg.h"
void helper_raise_interrupt(CPUX86State *env, int intno, int next_eip_addend)
{
diff --git a/target/i386/tcg/fpu_helper.c b/target/i386/tcg/fpu_helper.c
index 03b35443a6..13f31b6ac7 100644
--- a/target/i386/tcg/fpu_helper.c
+++ b/target/i386/tcg/fpu_helper.c
@@ -26,6 +26,7 @@
#include "exec/cpu_ldst.h"
#include "fpu/softfloat.h"
#include "fpu/softfloat-macros.h"
+#include "helper-tcg.h"
#ifdef CONFIG_SOFTMMU
#include "hw/irq.h"
@@ -2986,23 +2987,21 @@ void update_mxcsr_status(CPUX86State *env)
void update_mxcsr_from_sse_status(CPUX86State *env)
{
- if (tcg_enabled()) {
- uint8_t flags = get_float_exception_flags(&env->sse_status);
- /*
- * The MXCSR denormal flag has opposite semantics to
- * float_flag_input_denormal (the softfloat code sets that flag
- * only when flushing input denormals to zero, but SSE sets it
- * only when not flushing them to zero), so is not converted
- * here.
- */
- env->mxcsr |= ((flags & float_flag_invalid ? FPUS_IE : 0) |
- (flags & float_flag_divbyzero ? FPUS_ZE : 0) |
- (flags & float_flag_overflow ? FPUS_OE : 0) |
- (flags & float_flag_underflow ? FPUS_UE : 0) |
- (flags & float_flag_inexact ? FPUS_PE : 0) |
- (flags & float_flag_output_denormal ? FPUS_UE | FPUS_PE :
- 0));
- }
+ uint8_t flags = get_float_exception_flags(&env->sse_status);
+ /*
+ * The MXCSR denormal flag has opposite semantics to
+ * float_flag_input_denormal (the softfloat code sets that flag
+ * only when flushing input denormals to zero, but SSE sets it
+ * only when not flushing them to zero), so is not converted
+ * here.
+ */
+ env->mxcsr |= ((flags & float_flag_invalid ? FPUS_IE : 0) |
+ (flags & float_flag_divbyzero ? FPUS_ZE : 0) |
+ (flags & float_flag_overflow ? FPUS_OE : 0) |
+ (flags & float_flag_underflow ? FPUS_UE : 0) |
+ (flags & float_flag_inexact ? FPUS_PE : 0) |
+ (flags & float_flag_output_denormal ? FPUS_UE | FPUS_PE :
+ 0));
}
void helper_update_mxcsr(CPUX86State *env)
diff --git a/target/i386/tcg/int_helper.c b/target/i386/tcg/int_helper.c
index 4f89436b53..87fa7280ee 100644
--- a/target/i386/tcg/int_helper.c
+++ b/target/i386/tcg/int_helper.c
@@ -24,6 +24,7 @@
#include "exec/helper-proto.h"
#include "qapi/error.h"
#include "qemu/guest-random.h"
+#include "helper-tcg.h"
//#define DEBUG_MULDIV
diff --git a/target/i386/tcg/mem_helper.c b/target/i386/tcg/mem_helper.c
index 21ca3e3e88..e5cd2de1bf 100644
--- a/target/i386/tcg/mem_helper.c
+++ b/target/i386/tcg/mem_helper.c
@@ -25,6 +25,7 @@
#include "qemu/int128.h"
#include "qemu/atomic128.h"
#include "tcg/tcg.h"
+#include "helper-tcg.h"
void helper_cmpxchg8b_unlocked(CPUX86State *env, target_ulong a0)
{
diff --git a/target/i386/tcg/misc_helper.c b/target/i386/tcg/misc_helper.c
index ae259d9145..c99370e5e3 100644
--- a/target/i386/tcg/misc_helper.c
+++ b/target/i386/tcg/misc_helper.c
@@ -24,6 +24,7 @@
#include "exec/exec-all.h"
#include "exec/cpu_ldst.h"
#include "exec/address-spaces.h"
+#include "helper-tcg.h"
void helper_outb(CPUX86State *env, uint32_t port, uint32_t data)
{
diff --git a/target/i386/tcg/mpx_helper.c b/target/i386/tcg/mpx_helper.c
index fd966174b4..22423eedcd 100644
--- a/target/i386/tcg/mpx_helper.c
+++ b/target/i386/tcg/mpx_helper.c
@@ -22,6 +22,7 @@
#include "exec/helper-proto.h"
#include "exec/cpu_ldst.h"
#include "exec/exec-all.h"
+#include "helper-tcg.h"
void helper_bndck(CPUX86State *env, uint32_t fail)
diff --git a/target/i386/tcg/seg_helper.c b/target/i386/tcg/seg_helper.c
index 09b6554660..ed3e04a187 100644
--- a/target/i386/tcg/seg_helper.c
+++ b/target/i386/tcg/seg_helper.c
@@ -25,6 +25,7 @@
#include "exec/exec-all.h"
#include "exec/cpu_ldst.h"
#include "exec/log.h"
+#include "helper-tcg.h"
//#define DEBUG_PCALL
diff --git a/target/i386/tcg/smm_helper.c b/target/i386/tcg/smm_helper.c
index d20e8edfdf..62d027abd3 100644
--- a/target/i386/tcg/smm_helper.c
+++ b/target/i386/tcg/smm_helper.c
@@ -22,6 +22,8 @@
#include "cpu.h"
#include "exec/helper-proto.h"
#include "exec/log.h"
+#include "helper-tcg.h"
+
/* SMM support */
diff --git a/target/i386/tcg/svm_helper.c b/target/i386/tcg/svm_helper.c
index 38931586e5..097bb9b83d 100644
--- a/target/i386/tcg/svm_helper.c
+++ b/target/i386/tcg/svm_helper.c
@@ -22,6 +22,7 @@
#include "exec/helper-proto.h"
#include "exec/exec-all.h"
#include "exec/cpu_ldst.h"
+#include "helper-tcg.h"
/* Secure Virtual Machine helpers */
diff --git a/target/i386/tcg/translate.c b/target/i386/tcg/translate.c
index 4c57307e42..5988ea0289 100644
--- a/target/i386/tcg/translate.c
+++ b/target/i386/tcg/translate.c
@@ -28,6 +28,7 @@
#include "exec/helper-proto.h"
#include "exec/helper-gen.h"
+#include "helper-tcg.h"
#include "trace-tcg.h"
#include "exec/log.h"
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 11/32] tcg: cpu_exec_{enter,exit} helpers
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (9 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 10/32] i386: move TCG cpu class initialization out of helper.c Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 10:33 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 12/32] tcg: make CPUClass.cpu_exec_* optional Claudio Fontana
` (23 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Colin Xu, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Claudio Fontana
From: Eduardo Habkost <ehabkost@redhat.com>
Move invocation of CPUClass.cpu_exec_*() to separate helpers,
to make it easier to refactor that code later.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
accel/tcg/cpu-exec.c | 23 ++++++++++++++++++-----
1 file changed, 18 insertions(+), 5 deletions(-)
diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
index 58aea605d8..8d31145ad2 100644
--- a/accel/tcg/cpu-exec.c
+++ b/accel/tcg/cpu-exec.c
@@ -236,9 +236,22 @@ static void cpu_exec_nocache(CPUState *cpu, int max_cycles,
}
#endif
+static void cpu_exec_enter(CPUState *cpu)
+{
+ CPUClass *cc = CPU_GET_CLASS(cpu);
+
+ cc->cpu_exec_enter(cpu);
+}
+
+static void cpu_exec_exit(CPUState *cpu)
+{
+ CPUClass *cc = CPU_GET_CLASS(cpu);
+
+ cc->cpu_exec_exit(cpu);
+}
+
void cpu_exec_step_atomic(CPUState *cpu)
{
- CPUClass *cc = CPU_GET_CLASS(cpu);
TranslationBlock *tb;
target_ulong cs_base, pc;
uint32_t flags;
@@ -257,11 +270,11 @@ void cpu_exec_step_atomic(CPUState *cpu)
/* Since we got here, we know that parallel_cpus must be true. */
parallel_cpus = false;
- cc->cpu_exec_enter(cpu);
+ cpu_exec_enter(cpu);
/* execute the generated code */
trace_exec_tb(tb, pc);
cpu_tb_exec(cpu, tb);
- cc->cpu_exec_exit(cpu);
+ cpu_exec_exit(cpu);
} else {
/*
* The mmap_lock is dropped by tb_gen_code if it runs out of
@@ -713,7 +726,7 @@ int cpu_exec(CPUState *cpu)
rcu_read_lock();
- cc->cpu_exec_enter(cpu);
+ cpu_exec_enter(cpu);
/* Calculate difference between guest clock and host clock.
* This delay includes the delay of the last cycle, so
@@ -775,7 +788,7 @@ int cpu_exec(CPUState *cpu)
}
}
- cc->cpu_exec_exit(cpu);
+ cpu_exec_exit(cpu);
rcu_read_unlock();
return ret;
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 12/32] tcg: make CPUClass.cpu_exec_* optional
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (10 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 11/32] tcg: cpu_exec_{enter,exit} helpers Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 10:36 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 13/32] tcg: Make CPUClass.debug_excp_handler optional Claudio Fontana
` (22 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Colin Xu, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Claudio Fontana
From: Eduardo Habkost <ehabkost@redhat.com>
This will let us simplify the code that initializes CPU class
methods, when we move cpu_exec_*() to a separate struct.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
accel/tcg/cpu-exec.c | 11 ++++++++---
1 file changed, 8 insertions(+), 3 deletions(-)
diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
index 8d31145ad2..890b88861a 100644
--- a/accel/tcg/cpu-exec.c
+++ b/accel/tcg/cpu-exec.c
@@ -240,14 +240,18 @@ static void cpu_exec_enter(CPUState *cpu)
{
CPUClass *cc = CPU_GET_CLASS(cpu);
- cc->cpu_exec_enter(cpu);
+ if (cc->cpu_exec_enter) {
+ cc->cpu_exec_enter(cpu);
+ }
}
static void cpu_exec_exit(CPUState *cpu)
{
CPUClass *cc = CPU_GET_CLASS(cpu);
- cc->cpu_exec_exit(cpu);
+ if (cc->cpu_exec_exit) {
+ cc->cpu_exec_exit(cpu);
+ }
}
void cpu_exec_step_atomic(CPUState *cpu)
@@ -619,7 +623,8 @@ static inline bool cpu_handle_interrupt(CPUState *cpu,
True when it is, and we should restart on a new TB,
and via longjmp via cpu_loop_exit. */
else {
- if (cc->cpu_exec_interrupt(cpu, interrupt_request)) {
+ if (cc->cpu_exec_interrupt &&
+ cc->cpu_exec_interrupt(cpu, interrupt_request)) {
if (need_replay_interrupt(interrupt_request)) {
replay_interrupt();
}
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 13/32] tcg: Make CPUClass.debug_excp_handler optional
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (11 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 12/32] tcg: make CPUClass.cpu_exec_* optional Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 10:37 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 14/32] cpu: Remove unnecessary noop methods Claudio Fontana
` (21 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Colin Xu, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Claudio Fontana
From: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
accel/tcg/cpu-exec.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
index 890b88861a..64cba89356 100644
--- a/accel/tcg/cpu-exec.c
+++ b/accel/tcg/cpu-exec.c
@@ -482,7 +482,9 @@ static inline void cpu_handle_debug_exception(CPUState *cpu)
}
}
- cc->debug_excp_handler(cpu);
+ if (cc->debug_excp_handler) {
+ cc->debug_excp_handler(cpu);
+ }
}
static inline bool cpu_handle_exception(CPUState *cpu, int *ret)
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 14/32] cpu: Remove unnecessary noop methods
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (12 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 13/32] tcg: Make CPUClass.debug_excp_handler optional Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 10:38 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 15/32] cpu: Introduce TCGCpuOperations struct Claudio Fontana
` (20 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Colin Xu, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Claudio Fontana
From: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
hw/core/cpu.c | 13 -------------
1 file changed, 13 deletions(-)
diff --git a/hw/core/cpu.c b/hw/core/cpu.c
index 576fa1d7ba..994a12cb35 100644
--- a/hw/core/cpu.c
+++ b/hw/core/cpu.c
@@ -199,15 +199,6 @@ static bool cpu_common_virtio_is_big_endian(CPUState *cpu)
return target_words_bigendian();
}
-static void cpu_common_noop(CPUState *cpu)
-{
-}
-
-static bool cpu_common_exec_interrupt(CPUState *cpu, int int_req)
-{
- return false;
-}
-
#if !defined(CONFIG_USER_ONLY)
GuestPanicInformation *cpu_get_crash_info(CPUState *cpu)
{
@@ -410,11 +401,7 @@ static void cpu_class_init(ObjectClass *klass, void *data)
k->gdb_read_register = cpu_common_gdb_read_register;
k->gdb_write_register = cpu_common_gdb_write_register;
k->virtio_is_big_endian = cpu_common_virtio_is_big_endian;
- k->debug_excp_handler = cpu_common_noop;
k->debug_check_watchpoint = cpu_common_debug_check_watchpoint;
- k->cpu_exec_enter = cpu_common_noop;
- k->cpu_exec_exit = cpu_common_noop;
- k->cpu_exec_interrupt = cpu_common_exec_interrupt;
k->adjust_watchpoint_address = cpu_adjust_watchpoint_address;
set_bit(DEVICE_CATEGORY_CPU, dc->categories);
dc->realize = cpu_common_realizefn;
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 15/32] cpu: Introduce TCGCpuOperations struct
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (13 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 14/32] cpu: Remove unnecessary noop methods Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 10:39 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 16/32] target/riscv: remove CONFIG_TCG, as it is always TCG Claudio Fontana
` (19 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Colin Xu, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Claudio Fontana
From: Eduardo Habkost <ehabkost@redhat.com>
The TCG-specific CPU methods will be moved to a separate struct,
to make it easier to move accel-specific code outside generic CPU
code in the future. Start by moving tcg_initialize().
The new CPUClass.tcg_opts field may eventually become a pointer,
but keep it an embedded struct for now, to make code conversion
easier.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
[claudio: make the tcg code build for CONFIG_TCG only]
Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
---
MAINTAINERS | 1 +
cpu.c | 6 +++++-
include/hw/core/cpu.h | 9 ++++++++-
include/hw/core/tcg-cpu-ops.h | 25 +++++++++++++++++++++++++
target/alpha/cpu.c | 2 +-
target/arm/cpu.c | 2 +-
target/avr/cpu.c | 2 +-
target/cris/cpu.c | 12 ++++++------
target/hppa/cpu.c | 2 +-
target/i386/tcg-cpu.c | 2 +-
target/lm32/cpu.c | 2 +-
target/m68k/cpu.c | 2 +-
target/microblaze/cpu.c | 2 +-
target/mips/cpu.c | 2 +-
target/moxie/cpu.c | 2 +-
target/nios2/cpu.c | 2 +-
target/openrisc/cpu.c | 2 +-
target/ppc/translate_init.c.inc | 2 +-
target/riscv/cpu.c | 2 +-
target/rx/cpu.c | 2 +-
target/s390x/cpu.c | 2 +-
target/sh4/cpu.c | 2 +-
target/sparc/cpu.c | 2 +-
target/tilegx/cpu.c | 2 +-
target/tricore/cpu.c | 2 +-
target/unicore32/cpu.c | 2 +-
target/xtensa/cpu.c | 2 +-
27 files changed, 67 insertions(+), 30 deletions(-)
create mode 100644 include/hw/core/tcg-cpu-ops.h
diff --git a/MAINTAINERS b/MAINTAINERS
index f53f2678d8..d876f504a6 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -1535,6 +1535,7 @@ F: qapi/machine.json
F: qapi/machine-target.json
F: include/hw/boards.h
F: include/hw/core/cpu.h
+F: include/hw/core/tcg-cpu-ops.h
F: include/hw/cpu/cluster.h
F: include/sysemu/numa.h
T: git https://github.com/ehabkost/qemu.git machine-next
diff --git a/cpu.c b/cpu.c
index 0be5dcb6f3..27ad096cc4 100644
--- a/cpu.c
+++ b/cpu.c
@@ -174,14 +174,18 @@ void cpu_exec_initfn(CPUState *cpu)
void cpu_exec_realizefn(CPUState *cpu, Error **errp)
{
CPUClass *cc = CPU_GET_CLASS(cpu);
+#ifdef CONFIG_TCG
static bool tcg_target_initialized;
+#endif /* CONFIG_TCG */
cpu_list_add(cpu);
+#ifdef CONFIG_TCG
if (tcg_enabled() && !tcg_target_initialized) {
tcg_target_initialized = true;
- cc->tcg_initialize();
+ cc->tcg_ops.initialize();
}
+#endif /* CONFIG_TCG */
tlb_init(cpu);
qemu_plugin_vcpu_init_hook(cpu);
diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
index 3d92c967ff..c93b08a0fb 100644
--- a/include/hw/core/cpu.h
+++ b/include/hw/core/cpu.h
@@ -76,6 +76,10 @@ typedef struct CPUWatchpoint CPUWatchpoint;
struct TranslationBlock;
+#ifdef CONFIG_TCG
+#include "tcg-cpu-ops.h"
+#endif /* CONFIG_TCG */
+
/**
* CPUClass:
* @class_by_name: Callback to map -cpu command line model name to an
@@ -221,12 +225,15 @@ struct CPUClass {
void (*disas_set_info)(CPUState *cpu, disassemble_info *info);
vaddr (*adjust_watchpoint_address)(CPUState *cpu, vaddr addr, int len);
- void (*tcg_initialize)(void);
const char *deprecation_note;
/* Keep non-pointer data at the end to minimize holes. */
int gdb_num_core_regs;
bool gdb_stop_before_watchpoint;
+
+#ifdef CONFIG_TCG
+ TcgCpuOperations tcg_ops;
+#endif /* CONFIG_TCG */
};
/*
diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
new file mode 100644
index 0000000000..4475ef0996
--- /dev/null
+++ b/include/hw/core/tcg-cpu-ops.h
@@ -0,0 +1,25 @@
+/*
+ * TCG-Specific operations that are not meaningful for hardware accelerators
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#ifndef TCG_CPU_OPS_H
+#define TCG_CPU_OPS_H
+
+/**
+ * struct TcgCpuOperations: TCG operations specific to a CPU class
+ */
+typedef struct TcgCpuOperations {
+ /**
+ * @initialize: Initalize TCG state
+ *
+ * Called when the first CPU is realized.
+ */
+ void (*initialize)(void);
+} TcgCpuOperations;
+
+#endif /* TCG_CPU_OPS_H */
diff --git a/target/alpha/cpu.c b/target/alpha/cpu.c
index b3fd6643e8..d66f0351a9 100644
--- a/target/alpha/cpu.c
+++ b/target/alpha/cpu.c
@@ -231,7 +231,7 @@ static void alpha_cpu_class_init(ObjectClass *oc, void *data)
dc->vmsd = &vmstate_alpha_cpu;
#endif
cc->disas_set_info = alpha_cpu_disas_set_info;
- cc->tcg_initialize = alpha_translate_init;
+ cc->tcg_ops.initialize = alpha_translate_init;
cc->gdb_num_core_regs = 67;
}
diff --git a/target/arm/cpu.c b/target/arm/cpu.c
index 07492e9f9a..1fa9382a7c 100644
--- a/target/arm/cpu.c
+++ b/target/arm/cpu.c
@@ -2261,7 +2261,7 @@ static void arm_cpu_class_init(ObjectClass *oc, void *data)
cc->gdb_stop_before_watchpoint = true;
cc->disas_set_info = arm_disas_set_info;
#ifdef CONFIG_TCG
- cc->tcg_initialize = arm_translate_init;
+ cc->tcg_ops.initialize = arm_translate_init;
cc->tlb_fill = arm_cpu_tlb_fill;
cc->debug_excp_handler = arm_debug_excp_handler;
cc->debug_check_watchpoint = arm_debug_check_watchpoint;
diff --git a/target/avr/cpu.c b/target/avr/cpu.c
index 5d9c4ad5bf..94306a2aa0 100644
--- a/target/avr/cpu.c
+++ b/target/avr/cpu.c
@@ -206,7 +206,7 @@ static void avr_cpu_class_init(ObjectClass *oc, void *data)
cc->tlb_fill = avr_cpu_tlb_fill;
cc->vmsd = &vms_avr_cpu;
cc->disas_set_info = avr_cpu_disas_set_info;
- cc->tcg_initialize = avr_cpu_tcg_init;
+ cc->tcg_ops.initialize = avr_cpu_tcg_init;
cc->synchronize_from_tb = avr_cpu_synchronize_from_tb;
cc->gdb_read_register = avr_cpu_gdb_read_register;
cc->gdb_write_register = avr_cpu_gdb_write_register;
diff --git a/target/cris/cpu.c b/target/cris/cpu.c
index cff6b9eabf..4328f8e6c9 100644
--- a/target/cris/cpu.c
+++ b/target/cris/cpu.c
@@ -201,7 +201,7 @@ static void crisv8_cpu_class_init(ObjectClass *oc, void *data)
ccc->vr = 8;
cc->do_interrupt = crisv10_cpu_do_interrupt;
cc->gdb_read_register = crisv10_cpu_gdb_read_register;
- cc->tcg_initialize = cris_initialize_crisv10_tcg;
+ cc->tcg_ops.initialize = cris_initialize_crisv10_tcg;
}
static void crisv9_cpu_class_init(ObjectClass *oc, void *data)
@@ -212,7 +212,7 @@ static void crisv9_cpu_class_init(ObjectClass *oc, void *data)
ccc->vr = 9;
cc->do_interrupt = crisv10_cpu_do_interrupt;
cc->gdb_read_register = crisv10_cpu_gdb_read_register;
- cc->tcg_initialize = cris_initialize_crisv10_tcg;
+ cc->tcg_ops.initialize = cris_initialize_crisv10_tcg;
}
static void crisv10_cpu_class_init(ObjectClass *oc, void *data)
@@ -223,7 +223,7 @@ static void crisv10_cpu_class_init(ObjectClass *oc, void *data)
ccc->vr = 10;
cc->do_interrupt = crisv10_cpu_do_interrupt;
cc->gdb_read_register = crisv10_cpu_gdb_read_register;
- cc->tcg_initialize = cris_initialize_crisv10_tcg;
+ cc->tcg_ops.initialize = cris_initialize_crisv10_tcg;
}
static void crisv11_cpu_class_init(ObjectClass *oc, void *data)
@@ -234,7 +234,7 @@ static void crisv11_cpu_class_init(ObjectClass *oc, void *data)
ccc->vr = 11;
cc->do_interrupt = crisv10_cpu_do_interrupt;
cc->gdb_read_register = crisv10_cpu_gdb_read_register;
- cc->tcg_initialize = cris_initialize_crisv10_tcg;
+ cc->tcg_ops.initialize = cris_initialize_crisv10_tcg;
}
static void crisv17_cpu_class_init(ObjectClass *oc, void *data)
@@ -245,7 +245,7 @@ static void crisv17_cpu_class_init(ObjectClass *oc, void *data)
ccc->vr = 17;
cc->do_interrupt = crisv10_cpu_do_interrupt;
cc->gdb_read_register = crisv10_cpu_gdb_read_register;
- cc->tcg_initialize = cris_initialize_crisv10_tcg;
+ cc->tcg_ops.initialize = cris_initialize_crisv10_tcg;
}
static void crisv32_cpu_class_init(ObjectClass *oc, void *data)
@@ -284,7 +284,7 @@ static void cris_cpu_class_init(ObjectClass *oc, void *data)
cc->gdb_stop_before_watchpoint = true;
cc->disas_set_info = cris_disas_set_info;
- cc->tcg_initialize = cris_initialize_tcg;
+ cc->tcg_ops.initialize = cris_initialize_tcg;
}
#define DEFINE_CRIS_CPU_TYPE(cpu_model, initfn) \
diff --git a/target/hppa/cpu.c b/target/hppa/cpu.c
index 71b6aca45d..4c778966c2 100644
--- a/target/hppa/cpu.c
+++ b/target/hppa/cpu.c
@@ -153,7 +153,7 @@ static void hppa_cpu_class_init(ObjectClass *oc, void *data)
#endif
cc->do_unaligned_access = hppa_cpu_do_unaligned_access;
cc->disas_set_info = hppa_cpu_disas_set_info;
- cc->tcg_initialize = hppa_translate_init;
+ cc->tcg_ops.initialize = hppa_translate_init;
cc->gdb_num_core_regs = 128;
}
diff --git a/target/i386/tcg-cpu.c b/target/i386/tcg-cpu.c
index 628dd29fe7..1f2a3e881a 100644
--- a/target/i386/tcg-cpu.c
+++ b/target/i386/tcg-cpu.c
@@ -63,7 +63,7 @@ void tcg_cpu_common_class_init(CPUClass *cc)
cc->synchronize_from_tb = x86_cpu_synchronize_from_tb;
cc->cpu_exec_enter = x86_cpu_exec_enter;
cc->cpu_exec_exit = x86_cpu_exec_exit;
- cc->tcg_initialize = tcg_x86_init;
+ cc->tcg_ops.initialize = tcg_x86_init;
cc->tlb_fill = x86_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->debug_excp_handler = breakpoint_handler;
diff --git a/target/lm32/cpu.c b/target/lm32/cpu.c
index c50ad5fa15..ef795b81a4 100644
--- a/target/lm32/cpu.c
+++ b/target/lm32/cpu.c
@@ -237,7 +237,7 @@ static void lm32_cpu_class_init(ObjectClass *oc, void *data)
cc->gdb_stop_before_watchpoint = true;
cc->debug_excp_handler = lm32_debug_excp_handler;
cc->disas_set_info = lm32_cpu_disas_set_info;
- cc->tcg_initialize = lm32_translate_init;
+ cc->tcg_ops.initialize = lm32_translate_init;
}
#define DEFINE_LM32_CPU_TYPE(cpu_model, initfn) \
diff --git a/target/m68k/cpu.c b/target/m68k/cpu.c
index 72c545149e..b66d86c353 100644
--- a/target/m68k/cpu.c
+++ b/target/m68k/cpu.c
@@ -289,7 +289,7 @@ static void m68k_cpu_class_init(ObjectClass *c, void *data)
cc->get_phys_page_debug = m68k_cpu_get_phys_page_debug;
#endif
cc->disas_set_info = m68k_cpu_disas_set_info;
- cc->tcg_initialize = m68k_tcg_init;
+ cc->tcg_ops.initialize = m68k_tcg_init;
cc->gdb_num_core_regs = 18;
diff --git a/target/microblaze/cpu.c b/target/microblaze/cpu.c
index 9b2482159d..bc10518fa3 100644
--- a/target/microblaze/cpu.c
+++ b/target/microblaze/cpu.c
@@ -335,7 +335,7 @@ static void mb_cpu_class_init(ObjectClass *oc, void *data)
cc->gdb_num_core_regs = 32 + 27;
cc->disas_set_info = mb_disas_set_info;
- cc->tcg_initialize = mb_tcg_init;
+ cc->tcg_ops.initialize = mb_tcg_init;
}
static const TypeInfo mb_cpu_type_info = {
diff --git a/target/mips/cpu.c b/target/mips/cpu.c
index 76d50b00b4..bc48573763 100644
--- a/target/mips/cpu.c
+++ b/target/mips/cpu.c
@@ -249,7 +249,7 @@ static void mips_cpu_class_init(ObjectClass *c, void *data)
#endif
cc->disas_set_info = mips_cpu_disas_set_info;
#ifdef CONFIG_TCG
- cc->tcg_initialize = mips_tcg_init;
+ cc->tcg_ops.initialize = mips_tcg_init;
cc->tlb_fill = mips_cpu_tlb_fill;
#endif
diff --git a/target/moxie/cpu.c b/target/moxie/cpu.c
index 6e0443ccb7..224cfc8361 100644
--- a/target/moxie/cpu.c
+++ b/target/moxie/cpu.c
@@ -116,7 +116,7 @@ static void moxie_cpu_class_init(ObjectClass *oc, void *data)
cc->vmsd = &vmstate_moxie_cpu;
#endif
cc->disas_set_info = moxie_cpu_disas_set_info;
- cc->tcg_initialize = moxie_translate_init;
+ cc->tcg_ops.initialize = moxie_translate_init;
}
static void moxielite_initfn(Object *obj)
diff --git a/target/nios2/cpu.c b/target/nios2/cpu.c
index 8f7011fcb9..29c9c6f634 100644
--- a/target/nios2/cpu.c
+++ b/target/nios2/cpu.c
@@ -205,7 +205,7 @@ static void nios2_cpu_class_init(ObjectClass *oc, void *data)
cc->gdb_read_register = nios2_cpu_gdb_read_register;
cc->gdb_write_register = nios2_cpu_gdb_write_register;
cc->gdb_num_core_regs = 49;
- cc->tcg_initialize = nios2_tcg_init;
+ cc->tcg_ops.initialize = nios2_tcg_init;
}
static const TypeInfo nios2_cpu_type_info = {
diff --git a/target/openrisc/cpu.c b/target/openrisc/cpu.c
index 5528c0918f..e442f4f97c 100644
--- a/target/openrisc/cpu.c
+++ b/target/openrisc/cpu.c
@@ -166,7 +166,7 @@ static void openrisc_cpu_class_init(ObjectClass *oc, void *data)
dc->vmsd = &vmstate_openrisc_cpu;
#endif
cc->gdb_num_core_regs = 32 + 3;
- cc->tcg_initialize = openrisc_translate_init;
+ cc->tcg_ops.initialize = openrisc_translate_init;
cc->disas_set_info = openrisc_disas_set_info;
}
diff --git a/target/ppc/translate_init.c.inc b/target/ppc/translate_init.c.inc
index 78cc8f043b..9a6932b774 100644
--- a/target/ppc/translate_init.c.inc
+++ b/target/ppc/translate_init.c.inc
@@ -10945,7 +10945,7 @@ static void ppc_cpu_class_init(ObjectClass *oc, void *data)
cc->virtio_is_big_endian = ppc_cpu_is_big_endian;
#endif
#ifdef CONFIG_TCG
- cc->tcg_initialize = ppc_translate_init;
+ cc->tcg_ops.initialize = ppc_translate_init;
cc->tlb_fill = ppc_cpu_tlb_fill;
#endif
#ifndef CONFIG_USER_ONLY
diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index 6a0264fc6b..a52e0ce466 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -562,7 +562,7 @@ static void riscv_cpu_class_init(ObjectClass *c, void *data)
cc->vmsd = &vmstate_riscv_cpu;
#endif
#ifdef CONFIG_TCG
- cc->tcg_initialize = riscv_translate_init;
+ cc->tcg_ops.initialize = riscv_translate_init;
cc->tlb_fill = riscv_cpu_tlb_fill;
#endif
device_class_set_props(dc, riscv_cpu_properties);
diff --git a/target/rx/cpu.c b/target/rx/cpu.c
index 23ee17a701..a701a09b11 100644
--- a/target/rx/cpu.c
+++ b/target/rx/cpu.c
@@ -194,7 +194,7 @@ static void rx_cpu_class_init(ObjectClass *klass, void *data)
cc->gdb_write_register = rx_cpu_gdb_write_register;
cc->get_phys_page_debug = rx_cpu_get_phys_page_debug;
cc->disas_set_info = rx_cpu_disas_set_info;
- cc->tcg_initialize = rx_translate_init;
+ cc->tcg_ops.initialize = rx_translate_init;
cc->tlb_fill = rx_cpu_tlb_fill;
cc->gdb_num_core_regs = 26;
diff --git a/target/s390x/cpu.c b/target/s390x/cpu.c
index 7b66718c44..697b94ff7b 100644
--- a/target/s390x/cpu.c
+++ b/target/s390x/cpu.c
@@ -512,7 +512,7 @@ static void s390_cpu_class_init(ObjectClass *oc, void *data)
#endif
cc->disas_set_info = s390_cpu_disas_set_info;
#ifdef CONFIG_TCG
- cc->tcg_initialize = s390x_translate_init;
+ cc->tcg_ops.initialize = s390x_translate_init;
cc->tlb_fill = s390_cpu_tlb_fill;
#endif
diff --git a/target/sh4/cpu.c b/target/sh4/cpu.c
index 3c68021c56..bdc5c9d90b 100644
--- a/target/sh4/cpu.c
+++ b/target/sh4/cpu.c
@@ -231,7 +231,7 @@ static void superh_cpu_class_init(ObjectClass *oc, void *data)
cc->get_phys_page_debug = superh_cpu_get_phys_page_debug;
#endif
cc->disas_set_info = superh_cpu_disas_set_info;
- cc->tcg_initialize = sh4_translate_init;
+ cc->tcg_ops.initialize = sh4_translate_init;
cc->gdb_num_core_regs = 59;
diff --git a/target/sparc/cpu.c b/target/sparc/cpu.c
index 701e794eac..07e48b86d1 100644
--- a/target/sparc/cpu.c
+++ b/target/sparc/cpu.c
@@ -879,7 +879,7 @@ static void sparc_cpu_class_init(ObjectClass *oc, void *data)
cc->vmsd = &vmstate_sparc_cpu;
#endif
cc->disas_set_info = cpu_sparc_disas_set_info;
- cc->tcg_initialize = sparc_tcg_init;
+ cc->tcg_ops.initialize = sparc_tcg_init;
#if defined(TARGET_SPARC64) && !defined(TARGET_ABI32)
cc->gdb_num_core_regs = 86;
diff --git a/target/tilegx/cpu.c b/target/tilegx/cpu.c
index 1fee87c094..cd24d0eb9d 100644
--- a/target/tilegx/cpu.c
+++ b/target/tilegx/cpu.c
@@ -153,7 +153,7 @@ static void tilegx_cpu_class_init(ObjectClass *oc, void *data)
cc->set_pc = tilegx_cpu_set_pc;
cc->tlb_fill = tilegx_cpu_tlb_fill;
cc->gdb_num_core_regs = 0;
- cc->tcg_initialize = tilegx_tcg_init;
+ cc->tcg_ops.initialize = tilegx_tcg_init;
}
static const TypeInfo tilegx_cpu_type_info = {
diff --git a/target/tricore/cpu.c b/target/tricore/cpu.c
index 2f2e5b029f..78b2925955 100644
--- a/target/tricore/cpu.c
+++ b/target/tricore/cpu.c
@@ -164,7 +164,7 @@ static void tricore_cpu_class_init(ObjectClass *c, void *data)
cc->set_pc = tricore_cpu_set_pc;
cc->synchronize_from_tb = tricore_cpu_synchronize_from_tb;
cc->get_phys_page_debug = tricore_cpu_get_phys_page_debug;
- cc->tcg_initialize = tricore_tcg_init;
+ cc->tcg_ops.initialize = tricore_tcg_init;
cc->tlb_fill = tricore_cpu_tlb_fill;
}
diff --git a/target/unicore32/cpu.c b/target/unicore32/cpu.c
index b27fb9689f..226bf4226e 100644
--- a/target/unicore32/cpu.c
+++ b/target/unicore32/cpu.c
@@ -137,7 +137,7 @@ static void uc32_cpu_class_init(ObjectClass *oc, void *data)
cc->set_pc = uc32_cpu_set_pc;
cc->tlb_fill = uc32_cpu_tlb_fill;
cc->get_phys_page_debug = uc32_cpu_get_phys_page_debug;
- cc->tcg_initialize = uc32_translate_init;
+ cc->tcg_ops.initialize = uc32_translate_init;
dc->vmsd = &vmstate_uc32_cpu;
}
diff --git a/target/xtensa/cpu.c b/target/xtensa/cpu.c
index 88a32268a1..5a6f5bf88b 100644
--- a/target/xtensa/cpu.c
+++ b/target/xtensa/cpu.c
@@ -209,7 +209,7 @@ static void xtensa_cpu_class_init(ObjectClass *oc, void *data)
#endif
cc->debug_excp_handler = xtensa_breakpoint_handler;
cc->disas_set_info = xtensa_cpu_disas_set_info;
- cc->tcg_initialize = xtensa_translate_init;
+ cc->tcg_ops.initialize = xtensa_translate_init;
dc->vmsd = &vmstate_xtensa_cpu;
}
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 16/32] target/riscv: remove CONFIG_TCG, as it is always TCG
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (14 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 15/32] cpu: Introduce TCGCpuOperations struct Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-08 22:10 ` Alistair Francis
2020-12-08 19:48 ` [RFC v9 17/32] accel/tcg: split TCG-only code from cpu_exec_realizefn Claudio Fontana
` (18 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
for now only TCG is allowed as an accelerator for riscv,
so remove the CONFIG_TCG use.
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
target/riscv/cpu.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index a52e0ce466..27dd1645c9 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -561,10 +561,9 @@ static void riscv_cpu_class_init(ObjectClass *c, void *data)
/* For now, mark unmigratable: */
cc->vmsd = &vmstate_riscv_cpu;
#endif
-#ifdef CONFIG_TCG
cc->tcg_ops.initialize = riscv_translate_init;
cc->tlb_fill = riscv_cpu_tlb_fill;
-#endif
+
device_class_set_props(dc, riscv_cpu_properties);
}
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 17/32] accel/tcg: split TCG-only code from cpu_exec_realizefn
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (15 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 16/32] target/riscv: remove CONFIG_TCG, as it is always TCG Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 10:42 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 18/32] cpu: Move synchronize_from_tb() to tcg_ops Claudio Fontana
` (17 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
move away TCG-only code, make it compile only on TCG.
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
accel/tcg/cpu-exec.c | 28 +++++++++++++++++
cpu.c | 70 ++++++++++++++++++++-----------------------
hw/core/cpu.c | 6 +++-
include/hw/core/cpu.h | 8 +++++
4 files changed, 74 insertions(+), 38 deletions(-)
diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
index 64cba89356..436dfbf155 100644
--- a/accel/tcg/cpu-exec.c
+++ b/accel/tcg/cpu-exec.c
@@ -801,6 +801,34 @@ int cpu_exec(CPUState *cpu)
return ret;
}
+void tcg_exec_realizefn(CPUState *cpu, Error **errp)
+{
+ static bool tcg_target_initialized;
+ CPUClass *cc = CPU_GET_CLASS(cpu);
+
+ if (!tcg_target_initialized) {
+ tcg_target_initialized = true;
+ cc->tcg_ops.initialize();
+ }
+ tlb_init(cpu);
+ qemu_plugin_vcpu_init_hook(cpu);
+
+#ifndef CONFIG_USER_ONLY
+ tcg_iommu_init_notifier_list(cpu);
+#endif /* !CONFIG_USER_ONLY */
+}
+
+/* undo the initializations in reverse order */
+void tcg_exec_unrealizefn(CPUState *cpu)
+{
+#ifndef CONFIG_USER_ONLY
+ tcg_iommu_free_notifier_list(cpu);
+#endif /* !CONFIG_USER_ONLY */
+
+ qemu_plugin_vcpu_exit_hook(cpu);
+ tlb_destroy(cpu);
+}
+
#ifndef CONFIG_USER_ONLY
void dump_drift_info(void)
diff --git a/cpu.c b/cpu.c
index 27ad096cc4..5cc8f181be 100644
--- a/cpu.c
+++ b/cpu.c
@@ -124,13 +124,35 @@ const VMStateDescription vmstate_cpu_common = {
};
#endif
+void cpu_exec_realizefn(CPUState *cpu, Error **errp)
+{
+ CPUClass *cc = CPU_GET_CLASS(cpu);
+
+ cpu_list_add(cpu);
+
+#ifdef CONFIG_TCG
+ /* NB: errp parameter is unused currently */
+ if (tcg_enabled()) {
+ tcg_exec_realizefn(cpu, errp);
+ }
+#endif /* CONFIG_TCG */
+
+#ifdef CONFIG_USER_ONLY
+ assert(cc->vmsd == NULL);
+#else
+ if (qdev_get_vmsd(DEVICE(cpu)) == NULL) {
+ vmstate_register(NULL, cpu->cpu_index, &vmstate_cpu_common, cpu);
+ }
+ if (cc->vmsd != NULL) {
+ vmstate_register(NULL, cpu->cpu_index, cc->vmsd, cpu);
+ }
+#endif /* CONFIG_USER_ONLY */
+}
+
void cpu_exec_unrealizefn(CPUState *cpu)
{
CPUClass *cc = CPU_GET_CLASS(cpu);
- tlb_destroy(cpu);
- cpu_list_remove(cpu);
-
#ifdef CONFIG_USER_ONLY
assert(cc->vmsd == NULL);
#else
@@ -140,8 +162,15 @@ void cpu_exec_unrealizefn(CPUState *cpu)
if (qdev_get_vmsd(DEVICE(cpu)) == NULL) {
vmstate_unregister(NULL, &vmstate_cpu_common, cpu);
}
- tcg_iommu_free_notifier_list(cpu);
#endif
+#ifdef CONFIG_TCG
+ /* NB: errp parameter is unused currently */
+ if (tcg_enabled()) {
+ tcg_exec_unrealizefn(cpu);
+ }
+#endif /* CONFIG_TCG */
+
+ cpu_list_remove(cpu);
}
Property cpu_common_props[] = {
@@ -171,39 +200,6 @@ void cpu_exec_initfn(CPUState *cpu)
#endif
}
-void cpu_exec_realizefn(CPUState *cpu, Error **errp)
-{
- CPUClass *cc = CPU_GET_CLASS(cpu);
-#ifdef CONFIG_TCG
- static bool tcg_target_initialized;
-#endif /* CONFIG_TCG */
-
- cpu_list_add(cpu);
-
-#ifdef CONFIG_TCG
- if (tcg_enabled() && !tcg_target_initialized) {
- tcg_target_initialized = true;
- cc->tcg_ops.initialize();
- }
-#endif /* CONFIG_TCG */
- tlb_init(cpu);
-
- qemu_plugin_vcpu_init_hook(cpu);
-
-#ifdef CONFIG_USER_ONLY
- assert(cc->vmsd == NULL);
-#else /* !CONFIG_USER_ONLY */
- if (qdev_get_vmsd(DEVICE(cpu)) == NULL) {
- vmstate_register(NULL, cpu->cpu_index, &vmstate_cpu_common, cpu);
- }
- if (cc->vmsd != NULL) {
- vmstate_register(NULL, cpu->cpu_index, cc->vmsd, cpu);
- }
-
- tcg_iommu_init_notifier_list(cpu);
-#endif
-}
-
const char *parse_cpu_option(const char *cpu_option)
{
ObjectClass *oc;
diff --git a/hw/core/cpu.c b/hw/core/cpu.c
index 994a12cb35..1f04aab16b 100644
--- a/hw/core/cpu.c
+++ b/hw/core/cpu.c
@@ -199,6 +199,10 @@ static bool cpu_common_virtio_is_big_endian(CPUState *cpu)
return target_words_bigendian();
}
+/*
+ * XXX the following #if is always true because this is a common_ss
+ * module, so target CONFIG_* is never defined.
+ */
#if !defined(CONFIG_USER_ONLY)
GuestPanicInformation *cpu_get_crash_info(CPUState *cpu)
{
@@ -340,9 +344,9 @@ static void cpu_common_realizefn(DeviceState *dev, Error **errp)
static void cpu_common_unrealizefn(DeviceState *dev)
{
CPUState *cpu = CPU(dev);
+
/* NOTE: latest generic point before the cpu is fully unrealized */
trace_fini_vcpu(cpu);
- qemu_plugin_vcpu_exit_hook(cpu);
cpu_exec_unrealizefn(cpu);
}
diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
index c93b08a0fb..ea648d52ad 100644
--- a/include/hw/core/cpu.h
+++ b/include/hw/core/cpu.h
@@ -1119,10 +1119,18 @@ AddressSpace *cpu_get_address_space(CPUState *cpu, int asidx);
void QEMU_NORETURN cpu_abort(CPUState *cpu, const char *fmt, ...)
GCC_FMT_ATTR(2, 3);
extern Property cpu_common_props[];
+
+/* $(top_srcdir)/cpu.c */
void cpu_exec_initfn(CPUState *cpu);
void cpu_exec_realizefn(CPUState *cpu, Error **errp);
void cpu_exec_unrealizefn(CPUState *cpu);
+#ifdef CONFIG_TCG
+/* accel/tcg/cpu-exec.c */
+void tcg_exec_realizefn(CPUState *cpu, Error **errp);
+void tcg_exec_unrealizefn(CPUState *cpu);
+#endif /* CONFIG_TCG */
+
/**
* target_words_bigendian:
* Returns true if the (default) endianness of the target is big endian,
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 18/32] cpu: Move synchronize_from_tb() to tcg_ops
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (16 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 17/32] accel/tcg: split TCG-only code from cpu_exec_realizefn Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 9:27 ` Philippe Mathieu-Daudé
2020-12-09 10:50 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 19/32] cpu: Move cpu_exec_* " Claudio Fontana
` (16 subsequent siblings)
34 siblings, 2 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Colin Xu, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Claudio Fontana
From: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
[claudio: wrapped in CONFIG_TCG]
Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
---
accel/tcg/cpu-exec.c | 4 ++--
include/hw/core/cpu.h | 8 --------
include/hw/core/tcg-cpu-ops.h | 14 +++++++++++---
target/arm/cpu.c | 2 +-
target/avr/cpu.c | 2 +-
target/hppa/cpu.c | 2 +-
target/i386/tcg-cpu.c | 2 +-
target/microblaze/cpu.c | 2 +-
target/mips/cpu.c | 2 +-
target/riscv/cpu.c | 2 +-
target/rx/cpu.c | 2 +-
target/sh4/cpu.c | 2 +-
target/sparc/cpu.c | 2 +-
target/tricore/cpu.c | 2 +-
14 files changed, 24 insertions(+), 24 deletions(-)
diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
index 436dfbf155..776415d47e 100644
--- a/accel/tcg/cpu-exec.c
+++ b/accel/tcg/cpu-exec.c
@@ -192,8 +192,8 @@ static inline tcg_target_ulong cpu_tb_exec(CPUState *cpu, TranslationBlock *itb)
TARGET_FMT_lx "] %s\n",
last_tb->tc.ptr, last_tb->pc,
lookup_symbol(last_tb->pc));
- if (cc->synchronize_from_tb) {
- cc->synchronize_from_tb(cpu, last_tb);
+ if (cc->tcg_ops.synchronize_from_tb) {
+ cc->tcg_ops.synchronize_from_tb(cpu, last_tb);
} else {
assert(cc->set_pc);
cc->set_pc(cpu, last_tb->pc);
diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
index ea648d52ad..83007d262c 100644
--- a/include/hw/core/cpu.h
+++ b/include/hw/core/cpu.h
@@ -110,13 +110,6 @@ struct TranslationBlock;
* If the target behaviour here is anything other than "set
* the PC register to the value passed in" then the target must
* also implement the synchronize_from_tb hook.
- * @synchronize_from_tb: Callback for synchronizing state from a TCG
- * #TranslationBlock. This is called when we abandon execution
- * of a TB before starting it, and must set all parts of the CPU
- * state which the previous TB in the chain may not have updated.
- * This always includes at least the program counter; some targets
- * will need to do more. If this hook is not implemented then the
- * default is to call @set_pc(tb->pc).
* @tlb_fill: Callback for handling a softmmu tlb miss or user-only
* address fault. For system mode, if the access is valid, call
* tlb_set_page and return true; if the access is invalid, and
@@ -193,7 +186,6 @@ struct CPUClass {
void (*get_memory_mapping)(CPUState *cpu, MemoryMappingList *list,
Error **errp);
void (*set_pc)(CPUState *cpu, vaddr value);
- void (*synchronize_from_tb)(CPUState *cpu, struct TranslationBlock *tb);
bool (*tlb_fill)(CPUState *cpu, vaddr address, int size,
MMUAccessType access_type, int mmu_idx,
bool probe, uintptr_t retaddr);
diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
index 4475ef0996..109291ac52 100644
--- a/include/hw/core/tcg-cpu-ops.h
+++ b/include/hw/core/tcg-cpu-ops.h
@@ -10,9 +10,6 @@
#ifndef TCG_CPU_OPS_H
#define TCG_CPU_OPS_H
-/**
- * struct TcgCpuOperations: TCG operations specific to a CPU class
- */
typedef struct TcgCpuOperations {
/**
* @initialize: Initalize TCG state
@@ -20,6 +17,17 @@ typedef struct TcgCpuOperations {
* Called when the first CPU is realized.
*/
void (*initialize)(void);
+ /**
+ * @synchronize_from_tb: Synchronize state from a TCG #TranslationBlock
+ *
+ * This is called when we abandon execution of a TB before
+ * starting it, and must set all parts of the CPU state which
+ * the previous TB in the chain may not have updated. This
+ * will need to do more. If this hook is not implemented then
+ * the default is to call
+ * @set_pc(tb->pc).
+ */
+ void (*synchronize_from_tb)(CPUState *cpu, struct TranslationBlock *tb);
} TcgCpuOperations;
#endif /* TCG_CPU_OPS_H */
diff --git a/target/arm/cpu.c b/target/arm/cpu.c
index 1fa9382a7c..2f9e27c248 100644
--- a/target/arm/cpu.c
+++ b/target/arm/cpu.c
@@ -2242,7 +2242,6 @@ static void arm_cpu_class_init(ObjectClass *oc, void *data)
cc->cpu_exec_interrupt = arm_cpu_exec_interrupt;
cc->dump_state = arm_cpu_dump_state;
cc->set_pc = arm_cpu_set_pc;
- cc->synchronize_from_tb = arm_cpu_synchronize_from_tb;
cc->gdb_read_register = arm_cpu_gdb_read_register;
cc->gdb_write_register = arm_cpu_gdb_write_register;
#ifndef CONFIG_USER_ONLY
@@ -2262,6 +2261,7 @@ static void arm_cpu_class_init(ObjectClass *oc, void *data)
cc->disas_set_info = arm_disas_set_info;
#ifdef CONFIG_TCG
cc->tcg_ops.initialize = arm_translate_init;
+ cc->tcg_ops.synchronize_from_tb = arm_cpu_synchronize_from_tb;
cc->tlb_fill = arm_cpu_tlb_fill;
cc->debug_excp_handler = arm_debug_excp_handler;
cc->debug_check_watchpoint = arm_debug_check_watchpoint;
diff --git a/target/avr/cpu.c b/target/avr/cpu.c
index 94306a2aa0..f753c15768 100644
--- a/target/avr/cpu.c
+++ b/target/avr/cpu.c
@@ -207,7 +207,7 @@ static void avr_cpu_class_init(ObjectClass *oc, void *data)
cc->vmsd = &vms_avr_cpu;
cc->disas_set_info = avr_cpu_disas_set_info;
cc->tcg_ops.initialize = avr_cpu_tcg_init;
- cc->synchronize_from_tb = avr_cpu_synchronize_from_tb;
+ cc->tcg_ops.synchronize_from_tb = avr_cpu_synchronize_from_tb;
cc->gdb_read_register = avr_cpu_gdb_read_register;
cc->gdb_write_register = avr_cpu_gdb_write_register;
cc->gdb_num_core_regs = 35;
diff --git a/target/hppa/cpu.c b/target/hppa/cpu.c
index 4c778966c2..12a09e93ae 100644
--- a/target/hppa/cpu.c
+++ b/target/hppa/cpu.c
@@ -143,7 +143,7 @@ static void hppa_cpu_class_init(ObjectClass *oc, void *data)
cc->cpu_exec_interrupt = hppa_cpu_exec_interrupt;
cc->dump_state = hppa_cpu_dump_state;
cc->set_pc = hppa_cpu_set_pc;
- cc->synchronize_from_tb = hppa_cpu_synchronize_from_tb;
+ cc->tcg_ops.synchronize_from_tb = hppa_cpu_synchronize_from_tb;
cc->gdb_read_register = hppa_cpu_gdb_read_register;
cc->gdb_write_register = hppa_cpu_gdb_write_register;
cc->tlb_fill = hppa_cpu_tlb_fill;
diff --git a/target/i386/tcg-cpu.c b/target/i386/tcg-cpu.c
index 1f2a3e881a..d1414e2970 100644
--- a/target/i386/tcg-cpu.c
+++ b/target/i386/tcg-cpu.c
@@ -60,7 +60,7 @@ void tcg_cpu_common_class_init(CPUClass *cc)
{
cc->do_interrupt = x86_cpu_do_interrupt;
cc->cpu_exec_interrupt = x86_cpu_exec_interrupt;
- cc->synchronize_from_tb = x86_cpu_synchronize_from_tb;
+ cc->tcg_ops.synchronize_from_tb = x86_cpu_synchronize_from_tb;
cc->cpu_exec_enter = x86_cpu_exec_enter;
cc->cpu_exec_exit = x86_cpu_exec_exit;
cc->tcg_ops.initialize = tcg_x86_init;
diff --git a/target/microblaze/cpu.c b/target/microblaze/cpu.c
index bc10518fa3..97d94d9c27 100644
--- a/target/microblaze/cpu.c
+++ b/target/microblaze/cpu.c
@@ -322,7 +322,7 @@ static void mb_cpu_class_init(ObjectClass *oc, void *data)
cc->cpu_exec_interrupt = mb_cpu_exec_interrupt;
cc->dump_state = mb_cpu_dump_state;
cc->set_pc = mb_cpu_set_pc;
- cc->synchronize_from_tb = mb_cpu_synchronize_from_tb;
+ cc->tcg_ops.synchronize_from_tb = mb_cpu_synchronize_from_tb;
cc->gdb_read_register = mb_cpu_gdb_read_register;
cc->gdb_write_register = mb_cpu_gdb_write_register;
cc->tlb_fill = mb_cpu_tlb_fill;
diff --git a/target/mips/cpu.c b/target/mips/cpu.c
index bc48573763..71c488f3dd 100644
--- a/target/mips/cpu.c
+++ b/target/mips/cpu.c
@@ -238,7 +238,6 @@ static void mips_cpu_class_init(ObjectClass *c, void *data)
cc->cpu_exec_interrupt = mips_cpu_exec_interrupt;
cc->dump_state = mips_cpu_dump_state;
cc->set_pc = mips_cpu_set_pc;
- cc->synchronize_from_tb = mips_cpu_synchronize_from_tb;
cc->gdb_read_register = mips_cpu_gdb_read_register;
cc->gdb_write_register = mips_cpu_gdb_write_register;
#ifndef CONFIG_USER_ONLY
@@ -250,6 +249,7 @@ static void mips_cpu_class_init(ObjectClass *c, void *data)
cc->disas_set_info = mips_cpu_disas_set_info;
#ifdef CONFIG_TCG
cc->tcg_ops.initialize = mips_tcg_init;
+ cc->tcg_ops.synchronize_from_tb = mips_cpu_synchronize_from_tb;
cc->tlb_fill = mips_cpu_tlb_fill;
#endif
diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index 27dd1645c9..a9c30879d3 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -543,7 +543,7 @@ static void riscv_cpu_class_init(ObjectClass *c, void *data)
cc->cpu_exec_interrupt = riscv_cpu_exec_interrupt;
cc->dump_state = riscv_cpu_dump_state;
cc->set_pc = riscv_cpu_set_pc;
- cc->synchronize_from_tb = riscv_cpu_synchronize_from_tb;
+ cc->tcg_ops.synchronize_from_tb = riscv_cpu_synchronize_from_tb;
cc->gdb_read_register = riscv_cpu_gdb_read_register;
cc->gdb_write_register = riscv_cpu_gdb_write_register;
cc->gdb_num_core_regs = 33;
diff --git a/target/rx/cpu.c b/target/rx/cpu.c
index a701a09b11..d03c4e0b05 100644
--- a/target/rx/cpu.c
+++ b/target/rx/cpu.c
@@ -189,7 +189,7 @@ static void rx_cpu_class_init(ObjectClass *klass, void *data)
cc->cpu_exec_interrupt = rx_cpu_exec_interrupt;
cc->dump_state = rx_cpu_dump_state;
cc->set_pc = rx_cpu_set_pc;
- cc->synchronize_from_tb = rx_cpu_synchronize_from_tb;
+ cc->tcg_ops.synchronize_from_tb = rx_cpu_synchronize_from_tb;
cc->gdb_read_register = rx_cpu_gdb_read_register;
cc->gdb_write_register = rx_cpu_gdb_write_register;
cc->get_phys_page_debug = rx_cpu_get_phys_page_debug;
diff --git a/target/sh4/cpu.c b/target/sh4/cpu.c
index bdc5c9d90b..a33025b5c8 100644
--- a/target/sh4/cpu.c
+++ b/target/sh4/cpu.c
@@ -222,7 +222,7 @@ static void superh_cpu_class_init(ObjectClass *oc, void *data)
cc->cpu_exec_interrupt = superh_cpu_exec_interrupt;
cc->dump_state = superh_cpu_dump_state;
cc->set_pc = superh_cpu_set_pc;
- cc->synchronize_from_tb = superh_cpu_synchronize_from_tb;
+ cc->tcg_ops.synchronize_from_tb = superh_cpu_synchronize_from_tb;
cc->gdb_read_register = superh_cpu_gdb_read_register;
cc->gdb_write_register = superh_cpu_gdb_write_register;
cc->tlb_fill = superh_cpu_tlb_fill;
diff --git a/target/sparc/cpu.c b/target/sparc/cpu.c
index 07e48b86d1..baf6c5b587 100644
--- a/target/sparc/cpu.c
+++ b/target/sparc/cpu.c
@@ -868,7 +868,7 @@ static void sparc_cpu_class_init(ObjectClass *oc, void *data)
cc->memory_rw_debug = sparc_cpu_memory_rw_debug;
#endif
cc->set_pc = sparc_cpu_set_pc;
- cc->synchronize_from_tb = sparc_cpu_synchronize_from_tb;
+ cc->tcg_ops.synchronize_from_tb = sparc_cpu_synchronize_from_tb;
cc->gdb_read_register = sparc_cpu_gdb_read_register;
cc->gdb_write_register = sparc_cpu_gdb_write_register;
cc->tlb_fill = sparc_cpu_tlb_fill;
diff --git a/target/tricore/cpu.c b/target/tricore/cpu.c
index 78b2925955..5edf96c600 100644
--- a/target/tricore/cpu.c
+++ b/target/tricore/cpu.c
@@ -162,7 +162,7 @@ static void tricore_cpu_class_init(ObjectClass *c, void *data)
cc->dump_state = tricore_cpu_dump_state;
cc->set_pc = tricore_cpu_set_pc;
- cc->synchronize_from_tb = tricore_cpu_synchronize_from_tb;
+ cc->tcg_ops.synchronize_from_tb = tricore_cpu_synchronize_from_tb;
cc->get_phys_page_debug = tricore_cpu_get_phys_page_debug;
cc->tcg_ops.initialize = tricore_tcg_init;
cc->tlb_fill = tricore_cpu_tlb_fill;
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 19/32] cpu: Move cpu_exec_* to tcg_ops
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (17 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 18/32] cpu: Move synchronize_from_tb() to tcg_ops Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 9:28 ` Philippe Mathieu-Daudé
2020-12-09 11:16 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 20/32] cpu: Move tlb_fill " Claudio Fontana
` (15 subsequent siblings)
34 siblings, 2 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Colin Xu, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Claudio Fontana
From: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
[claudio: wrapped in CONFIG_TCG]
Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
---
accel/tcg/cpu-exec.c | 12 ++++++------
include/hw/core/cpu.h | 6 ------
include/hw/core/tcg-cpu-ops.h | 9 +++++++++
target/alpha/cpu.c | 3 ++-
target/arm/cpu.c | 2 +-
target/arm/cpu64.c | 5 ++++-
target/arm/cpu_tcg.c | 5 ++++-
target/avr/cpu.c | 2 +-
target/cris/cpu.c | 2 +-
target/hppa/cpu.c | 2 +-
target/i386/tcg-cpu.c | 6 +++---
target/lm32/cpu.c | 2 +-
target/m68k/cpu.c | 2 +-
target/microblaze/cpu.c | 2 +-
target/mips/cpu.c | 2 +-
target/nios2/cpu.c | 2 +-
target/openrisc/cpu.c | 2 +-
target/ppc/translate_init.c.inc | 10 +++++-----
target/riscv/cpu.c | 2 +-
target/rx/cpu.c | 2 +-
target/s390x/cpu.c | 2 +-
target/sh4/cpu.c | 2 +-
target/sparc/cpu.c | 2 +-
target/tilegx/cpu.c | 2 +-
target/unicore32/cpu.c | 2 +-
target/xtensa/cpu.c | 2 +-
26 files changed, 51 insertions(+), 41 deletions(-)
diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
index 776415d47e..edb349835e 100644
--- a/accel/tcg/cpu-exec.c
+++ b/accel/tcg/cpu-exec.c
@@ -240,8 +240,8 @@ static void cpu_exec_enter(CPUState *cpu)
{
CPUClass *cc = CPU_GET_CLASS(cpu);
- if (cc->cpu_exec_enter) {
- cc->cpu_exec_enter(cpu);
+ if (cc->tcg_ops.cpu_exec_enter) {
+ cc->tcg_ops.cpu_exec_enter(cpu);
}
}
@@ -249,8 +249,8 @@ static void cpu_exec_exit(CPUState *cpu)
{
CPUClass *cc = CPU_GET_CLASS(cpu);
- if (cc->cpu_exec_exit) {
- cc->cpu_exec_exit(cpu);
+ if (cc->tcg_ops.cpu_exec_exit) {
+ cc->tcg_ops.cpu_exec_exit(cpu);
}
}
@@ -625,8 +625,8 @@ static inline bool cpu_handle_interrupt(CPUState *cpu,
True when it is, and we should restart on a new TB,
and via longjmp via cpu_loop_exit. */
else {
- if (cc->cpu_exec_interrupt &&
- cc->cpu_exec_interrupt(cpu, interrupt_request)) {
+ if (cc->tcg_ops.cpu_exec_interrupt &&
+ cc->tcg_ops.cpu_exec_interrupt(cpu, interrupt_request)) {
if (need_replay_interrupt(interrupt_request)) {
replay_interrupt();
}
diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
index 83007d262c..52142e9094 100644
--- a/include/hw/core/cpu.h
+++ b/include/hw/core/cpu.h
@@ -146,9 +146,6 @@ struct TranslationBlock;
* @gdb_get_dynamic_xml: Callback to return dynamically generated XML for the
* gdb stub. Returns a pointer to the XML contents for the specified XML file
* or NULL if the CPU doesn't have a dynamically generated content for it.
- * @cpu_exec_enter: Callback for cpu_exec preparation.
- * @cpu_exec_exit: Callback for cpu_exec cleanup.
- * @cpu_exec_interrupt: Callback for processing interrupts in cpu_exec.
* @disas_set_info: Setup architecture specific components of disassembly info
* @adjust_watchpoint_address: Perform a target-specific adjustment to an
* address before attempting to match it against watchpoints.
@@ -211,9 +208,6 @@ struct CPUClass {
const char *gdb_core_xml_file;
gchar * (*gdb_arch_name)(CPUState *cpu);
const char * (*gdb_get_dynamic_xml)(CPUState *cpu, const char *xmlname);
- void (*cpu_exec_enter)(CPUState *cpu);
- void (*cpu_exec_exit)(CPUState *cpu);
- bool (*cpu_exec_interrupt)(CPUState *cpu, int interrupt_request);
void (*disas_set_info)(CPUState *cpu, disassemble_info *info);
vaddr (*adjust_watchpoint_address)(CPUState *cpu, vaddr addr, int len);
diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
index 109291ac52..e12f32919b 100644
--- a/include/hw/core/tcg-cpu-ops.h
+++ b/include/hw/core/tcg-cpu-ops.h
@@ -10,6 +10,9 @@
#ifndef TCG_CPU_OPS_H
#define TCG_CPU_OPS_H
+/**
+ * struct TcgCpuOperations: TCG operations specific to a CPU class
+ */
typedef struct TcgCpuOperations {
/**
* @initialize: Initalize TCG state
@@ -28,6 +31,12 @@ typedef struct TcgCpuOperations {
* @set_pc(tb->pc).
*/
void (*synchronize_from_tb)(CPUState *cpu, struct TranslationBlock *tb);
+ /** @cpu_exec_enter: Callback for cpu_exec preparation */
+ void (*cpu_exec_enter)(CPUState *cpu);
+ /** @cpu_exec_exit: Callback for cpu_exec cleanup */
+ void (*cpu_exec_exit)(CPUState *cpu);
+ /** @cpu_exec_interrupt: Callback for processing interrupts in cpu_exec */
+ bool (*cpu_exec_interrupt)(CPUState *cpu, int interrupt_request);
} TcgCpuOperations;
#endif /* TCG_CPU_OPS_H */
diff --git a/target/alpha/cpu.c b/target/alpha/cpu.c
index d66f0351a9..4f206c154d 100644
--- a/target/alpha/cpu.c
+++ b/target/alpha/cpu.c
@@ -218,7 +218,6 @@ static void alpha_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = alpha_cpu_class_by_name;
cc->has_work = alpha_cpu_has_work;
cc->do_interrupt = alpha_cpu_do_interrupt;
- cc->cpu_exec_interrupt = alpha_cpu_exec_interrupt;
cc->dump_state = alpha_cpu_dump_state;
cc->set_pc = alpha_cpu_set_pc;
cc->gdb_read_register = alpha_cpu_gdb_read_register;
@@ -234,6 +233,8 @@ static void alpha_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.initialize = alpha_translate_init;
cc->gdb_num_core_regs = 67;
+
+ cc->tcg_ops.cpu_exec_interrupt = alpha_cpu_exec_interrupt;
}
#define DEFINE_ALPHA_CPU_TYPE(base_type, cpu_model, initfn) \
diff --git a/target/arm/cpu.c b/target/arm/cpu.c
index 2f9e27c248..470b7b48c3 100644
--- a/target/arm/cpu.c
+++ b/target/arm/cpu.c
@@ -2239,7 +2239,6 @@ static void arm_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = arm_cpu_class_by_name;
cc->has_work = arm_cpu_has_work;
- cc->cpu_exec_interrupt = arm_cpu_exec_interrupt;
cc->dump_state = arm_cpu_dump_state;
cc->set_pc = arm_cpu_set_pc;
cc->gdb_read_register = arm_cpu_gdb_read_register;
@@ -2261,6 +2260,7 @@ static void arm_cpu_class_init(ObjectClass *oc, void *data)
cc->disas_set_info = arm_disas_set_info;
#ifdef CONFIG_TCG
cc->tcg_ops.initialize = arm_translate_init;
+ cc->tcg_ops.cpu_exec_interrupt = arm_cpu_exec_interrupt;
cc->tcg_ops.synchronize_from_tb = arm_cpu_synchronize_from_tb;
cc->tlb_fill = arm_cpu_tlb_fill;
cc->debug_excp_handler = arm_debug_excp_handler;
diff --git a/target/arm/cpu64.c b/target/arm/cpu64.c
index 649213082f..ebc53c8bdc 100644
--- a/target/arm/cpu64.c
+++ b/target/arm/cpu64.c
@@ -784,7 +784,10 @@ static void aarch64_cpu_class_init(ObjectClass *oc, void *data)
{
CPUClass *cc = CPU_CLASS(oc);
- cc->cpu_exec_interrupt = arm_cpu_exec_interrupt;
+#ifdef CONFIG_TCG
+ cc->tcg_ops.cpu_exec_interrupt = arm_cpu_exec_interrupt;
+#endif /* CONFIG_TCG */
+
cc->gdb_read_register = aarch64_cpu_gdb_read_register;
cc->gdb_write_register = aarch64_cpu_gdb_write_register;
cc->gdb_num_core_regs = 34;
diff --git a/target/arm/cpu_tcg.c b/target/arm/cpu_tcg.c
index 0013e25412..a6538d3a96 100644
--- a/target/arm/cpu_tcg.c
+++ b/target/arm/cpu_tcg.c
@@ -628,7 +628,10 @@ static void arm_v7m_class_init(ObjectClass *oc, void *data)
cc->do_interrupt = arm_v7m_cpu_do_interrupt;
#endif
- cc->cpu_exec_interrupt = arm_v7m_cpu_exec_interrupt;
+#ifdef CONFIG_TCG
+ cc->tcg_ops.cpu_exec_interrupt = arm_v7m_cpu_exec_interrupt;
+#endif /* CONFIG_TCG */
+
cc->gdb_core_xml_file = "arm-m-profile.xml";
}
diff --git a/target/avr/cpu.c b/target/avr/cpu.c
index f753c15768..277b00dbfc 100644
--- a/target/avr/cpu.c
+++ b/target/avr/cpu.c
@@ -198,7 +198,7 @@ static void avr_cpu_class_init(ObjectClass *oc, void *data)
cc->has_work = avr_cpu_has_work;
cc->do_interrupt = avr_cpu_do_interrupt;
- cc->cpu_exec_interrupt = avr_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = avr_cpu_exec_interrupt;
cc->dump_state = avr_cpu_dump_state;
cc->set_pc = avr_cpu_set_pc;
cc->memory_rw_debug = avr_cpu_memory_rw_debug;
diff --git a/target/cris/cpu.c b/target/cris/cpu.c
index 4328f8e6c9..7489fc20c8 100644
--- a/target/cris/cpu.c
+++ b/target/cris/cpu.c
@@ -269,7 +269,7 @@ static void cris_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = cris_cpu_class_by_name;
cc->has_work = cris_cpu_has_work;
cc->do_interrupt = cris_cpu_do_interrupt;
- cc->cpu_exec_interrupt = cris_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = cris_cpu_exec_interrupt;
cc->dump_state = cris_cpu_dump_state;
cc->set_pc = cris_cpu_set_pc;
cc->gdb_read_register = cris_cpu_gdb_read_register;
diff --git a/target/hppa/cpu.c b/target/hppa/cpu.c
index 12a09e93ae..61444753f2 100644
--- a/target/hppa/cpu.c
+++ b/target/hppa/cpu.c
@@ -140,7 +140,7 @@ static void hppa_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = hppa_cpu_class_by_name;
cc->has_work = hppa_cpu_has_work;
cc->do_interrupt = hppa_cpu_do_interrupt;
- cc->cpu_exec_interrupt = hppa_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = hppa_cpu_exec_interrupt;
cc->dump_state = hppa_cpu_dump_state;
cc->set_pc = hppa_cpu_set_pc;
cc->tcg_ops.synchronize_from_tb = hppa_cpu_synchronize_from_tb;
diff --git a/target/i386/tcg-cpu.c b/target/i386/tcg-cpu.c
index d1414e2970..5e0f2a2fae 100644
--- a/target/i386/tcg-cpu.c
+++ b/target/i386/tcg-cpu.c
@@ -59,10 +59,10 @@ static void x86_cpu_synchronize_from_tb(CPUState *cs, TranslationBlock *tb)
void tcg_cpu_common_class_init(CPUClass *cc)
{
cc->do_interrupt = x86_cpu_do_interrupt;
- cc->cpu_exec_interrupt = x86_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = x86_cpu_exec_interrupt;
cc->tcg_ops.synchronize_from_tb = x86_cpu_synchronize_from_tb;
- cc->cpu_exec_enter = x86_cpu_exec_enter;
- cc->cpu_exec_exit = x86_cpu_exec_exit;
+ cc->tcg_ops.cpu_exec_enter = x86_cpu_exec_enter;
+ cc->tcg_ops.cpu_exec_exit = x86_cpu_exec_exit;
cc->tcg_ops.initialize = tcg_x86_init;
cc->tlb_fill = x86_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
diff --git a/target/lm32/cpu.c b/target/lm32/cpu.c
index ef795b81a4..eea2d3e515 100644
--- a/target/lm32/cpu.c
+++ b/target/lm32/cpu.c
@@ -223,7 +223,7 @@ static void lm32_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = lm32_cpu_class_by_name;
cc->has_work = lm32_cpu_has_work;
cc->do_interrupt = lm32_cpu_do_interrupt;
- cc->cpu_exec_interrupt = lm32_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = lm32_cpu_exec_interrupt;
cc->dump_state = lm32_cpu_dump_state;
cc->set_pc = lm32_cpu_set_pc;
cc->gdb_read_register = lm32_cpu_gdb_read_register;
diff --git a/target/m68k/cpu.c b/target/m68k/cpu.c
index b66d86c353..c0fa517fc3 100644
--- a/target/m68k/cpu.c
+++ b/target/m68k/cpu.c
@@ -278,7 +278,7 @@ static void m68k_cpu_class_init(ObjectClass *c, void *data)
cc->class_by_name = m68k_cpu_class_by_name;
cc->has_work = m68k_cpu_has_work;
cc->do_interrupt = m68k_cpu_do_interrupt;
- cc->cpu_exec_interrupt = m68k_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = m68k_cpu_exec_interrupt;
cc->dump_state = m68k_cpu_dump_state;
cc->set_pc = m68k_cpu_set_pc;
cc->gdb_read_register = m68k_cpu_gdb_read_register;
diff --git a/target/microblaze/cpu.c b/target/microblaze/cpu.c
index 97d94d9c27..833d7f2d59 100644
--- a/target/microblaze/cpu.c
+++ b/target/microblaze/cpu.c
@@ -319,7 +319,7 @@ static void mb_cpu_class_init(ObjectClass *oc, void *data)
cc->has_work = mb_cpu_has_work;
cc->do_interrupt = mb_cpu_do_interrupt;
cc->do_unaligned_access = mb_cpu_do_unaligned_access;
- cc->cpu_exec_interrupt = mb_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = mb_cpu_exec_interrupt;
cc->dump_state = mb_cpu_dump_state;
cc->set_pc = mb_cpu_set_pc;
cc->tcg_ops.synchronize_from_tb = mb_cpu_synchronize_from_tb;
diff --git a/target/mips/cpu.c b/target/mips/cpu.c
index 71c488f3dd..344ee402fc 100644
--- a/target/mips/cpu.c
+++ b/target/mips/cpu.c
@@ -235,7 +235,6 @@ static void mips_cpu_class_init(ObjectClass *c, void *data)
cc->class_by_name = mips_cpu_class_by_name;
cc->has_work = mips_cpu_has_work;
cc->do_interrupt = mips_cpu_do_interrupt;
- cc->cpu_exec_interrupt = mips_cpu_exec_interrupt;
cc->dump_state = mips_cpu_dump_state;
cc->set_pc = mips_cpu_set_pc;
cc->gdb_read_register = mips_cpu_gdb_read_register;
@@ -249,6 +248,7 @@ static void mips_cpu_class_init(ObjectClass *c, void *data)
cc->disas_set_info = mips_cpu_disas_set_info;
#ifdef CONFIG_TCG
cc->tcg_ops.initialize = mips_tcg_init;
+ cc->tcg_ops.cpu_exec_interrupt = mips_cpu_exec_interrupt;
cc->tcg_ops.synchronize_from_tb = mips_cpu_synchronize_from_tb;
cc->tlb_fill = mips_cpu_tlb_fill;
#endif
diff --git a/target/nios2/cpu.c b/target/nios2/cpu.c
index 29c9c6f634..9eeb01fb5b 100644
--- a/target/nios2/cpu.c
+++ b/target/nios2/cpu.c
@@ -193,7 +193,7 @@ static void nios2_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = nios2_cpu_class_by_name;
cc->has_work = nios2_cpu_has_work;
cc->do_interrupt = nios2_cpu_do_interrupt;
- cc->cpu_exec_interrupt = nios2_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = nios2_cpu_exec_interrupt;
cc->dump_state = nios2_cpu_dump_state;
cc->set_pc = nios2_cpu_set_pc;
cc->disas_set_info = nios2_cpu_disas_set_info;
diff --git a/target/openrisc/cpu.c b/target/openrisc/cpu.c
index e442f4f97c..df8a41f956 100644
--- a/target/openrisc/cpu.c
+++ b/target/openrisc/cpu.c
@@ -155,7 +155,7 @@ static void openrisc_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = openrisc_cpu_class_by_name;
cc->has_work = openrisc_cpu_has_work;
cc->do_interrupt = openrisc_cpu_do_interrupt;
- cc->cpu_exec_interrupt = openrisc_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = openrisc_cpu_exec_interrupt;
cc->dump_state = openrisc_cpu_dump_state;
cc->set_pc = openrisc_cpu_set_pc;
cc->gdb_read_register = openrisc_cpu_gdb_read_register;
diff --git a/target/ppc/translate_init.c.inc b/target/ppc/translate_init.c.inc
index 9a6932b774..e0edbac999 100644
--- a/target/ppc/translate_init.c.inc
+++ b/target/ppc/translate_init.c.inc
@@ -10909,7 +10909,6 @@ static void ppc_cpu_class_init(ObjectClass *oc, void *data)
cc->parse_features = ppc_cpu_parse_featurestr;
cc->has_work = ppc_cpu_has_work;
cc->do_interrupt = ppc_cpu_do_interrupt;
- cc->cpu_exec_interrupt = ppc_cpu_exec_interrupt;
cc->dump_state = ppc_cpu_dump_state;
cc->dump_statistics = ppc_cpu_dump_statistics;
cc->set_pc = ppc_cpu_set_pc;
@@ -10946,12 +10945,13 @@ static void ppc_cpu_class_init(ObjectClass *oc, void *data)
#endif
#ifdef CONFIG_TCG
cc->tcg_ops.initialize = ppc_translate_init;
+ cc->tcg_ops.cpu_exec_interrupt = ppc_cpu_exec_interrupt;
cc->tlb_fill = ppc_cpu_tlb_fill;
-#endif
#ifndef CONFIG_USER_ONLY
- cc->cpu_exec_enter = ppc_cpu_exec_enter;
- cc->cpu_exec_exit = ppc_cpu_exec_exit;
-#endif
+ cc->tcg_ops.cpu_exec_enter = ppc_cpu_exec_enter;
+ cc->tcg_ops.cpu_exec_exit = ppc_cpu_exec_exit;
+#endif /* !CONFIG_USER_ONLY */
+#endif /* CONFIG_TCG */
cc->disas_set_info = ppc_disas_set_info;
diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index a9c30879d3..15d58698a6 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -540,7 +540,7 @@ static void riscv_cpu_class_init(ObjectClass *c, void *data)
cc->class_by_name = riscv_cpu_class_by_name;
cc->has_work = riscv_cpu_has_work;
cc->do_interrupt = riscv_cpu_do_interrupt;
- cc->cpu_exec_interrupt = riscv_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = riscv_cpu_exec_interrupt;
cc->dump_state = riscv_cpu_dump_state;
cc->set_pc = riscv_cpu_set_pc;
cc->tcg_ops.synchronize_from_tb = riscv_cpu_synchronize_from_tb;
diff --git a/target/rx/cpu.c b/target/rx/cpu.c
index d03c4e0b05..3ba93590d2 100644
--- a/target/rx/cpu.c
+++ b/target/rx/cpu.c
@@ -186,7 +186,7 @@ static void rx_cpu_class_init(ObjectClass *klass, void *data)
cc->class_by_name = rx_cpu_class_by_name;
cc->has_work = rx_cpu_has_work;
cc->do_interrupt = rx_cpu_do_interrupt;
- cc->cpu_exec_interrupt = rx_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = rx_cpu_exec_interrupt;
cc->dump_state = rx_cpu_dump_state;
cc->set_pc = rx_cpu_set_pc;
cc->tcg_ops.synchronize_from_tb = rx_cpu_synchronize_from_tb;
diff --git a/target/s390x/cpu.c b/target/s390x/cpu.c
index 697b94ff7b..add2f4b21f 100644
--- a/target/s390x/cpu.c
+++ b/target/s390x/cpu.c
@@ -505,7 +505,7 @@ static void s390_cpu_class_init(ObjectClass *oc, void *data)
cc->get_crash_info = s390_cpu_get_crash_info;
cc->write_elf64_note = s390_cpu_write_elf64_note;
#ifdef CONFIG_TCG
- cc->cpu_exec_interrupt = s390_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = s390_cpu_exec_interrupt;
cc->debug_excp_handler = s390x_cpu_debug_excp_handler;
cc->do_unaligned_access = s390x_cpu_do_unaligned_access;
#endif
diff --git a/target/sh4/cpu.c b/target/sh4/cpu.c
index a33025b5c8..0574194cd0 100644
--- a/target/sh4/cpu.c
+++ b/target/sh4/cpu.c
@@ -219,7 +219,7 @@ static void superh_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = superh_cpu_class_by_name;
cc->has_work = superh_cpu_has_work;
cc->do_interrupt = superh_cpu_do_interrupt;
- cc->cpu_exec_interrupt = superh_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = superh_cpu_exec_interrupt;
cc->dump_state = superh_cpu_dump_state;
cc->set_pc = superh_cpu_set_pc;
cc->tcg_ops.synchronize_from_tb = superh_cpu_synchronize_from_tb;
diff --git a/target/sparc/cpu.c b/target/sparc/cpu.c
index baf6c5b587..c559f15e14 100644
--- a/target/sparc/cpu.c
+++ b/target/sparc/cpu.c
@@ -862,7 +862,7 @@ static void sparc_cpu_class_init(ObjectClass *oc, void *data)
cc->parse_features = sparc_cpu_parse_features;
cc->has_work = sparc_cpu_has_work;
cc->do_interrupt = sparc_cpu_do_interrupt;
- cc->cpu_exec_interrupt = sparc_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = sparc_cpu_exec_interrupt;
cc->dump_state = sparc_cpu_dump_state;
#if !defined(TARGET_SPARC64) && !defined(CONFIG_USER_ONLY)
cc->memory_rw_debug = sparc_cpu_memory_rw_debug;
diff --git a/target/tilegx/cpu.c b/target/tilegx/cpu.c
index cd24d0eb9d..4c6176d26e 100644
--- a/target/tilegx/cpu.c
+++ b/target/tilegx/cpu.c
@@ -148,7 +148,7 @@ static void tilegx_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = tilegx_cpu_class_by_name;
cc->has_work = tilegx_cpu_has_work;
cc->do_interrupt = tilegx_cpu_do_interrupt;
- cc->cpu_exec_interrupt = tilegx_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = tilegx_cpu_exec_interrupt;
cc->dump_state = tilegx_cpu_dump_state;
cc->set_pc = tilegx_cpu_set_pc;
cc->tlb_fill = tilegx_cpu_tlb_fill;
diff --git a/target/unicore32/cpu.c b/target/unicore32/cpu.c
index 226bf4226e..84c3419989 100644
--- a/target/unicore32/cpu.c
+++ b/target/unicore32/cpu.c
@@ -132,7 +132,7 @@ static void uc32_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = uc32_cpu_class_by_name;
cc->has_work = uc32_cpu_has_work;
cc->do_interrupt = uc32_cpu_do_interrupt;
- cc->cpu_exec_interrupt = uc32_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = uc32_cpu_exec_interrupt;
cc->dump_state = uc32_cpu_dump_state;
cc->set_pc = uc32_cpu_set_pc;
cc->tlb_fill = uc32_cpu_tlb_fill;
diff --git a/target/xtensa/cpu.c b/target/xtensa/cpu.c
index 5a6f5bf88b..42a5e4ebe8 100644
--- a/target/xtensa/cpu.c
+++ b/target/xtensa/cpu.c
@@ -195,7 +195,7 @@ static void xtensa_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = xtensa_cpu_class_by_name;
cc->has_work = xtensa_cpu_has_work;
cc->do_interrupt = xtensa_cpu_do_interrupt;
- cc->cpu_exec_interrupt = xtensa_cpu_exec_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = xtensa_cpu_exec_interrupt;
cc->dump_state = xtensa_cpu_dump_state;
cc->set_pc = xtensa_cpu_set_pc;
cc->gdb_read_register = xtensa_cpu_gdb_read_register;
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 20/32] cpu: Move tlb_fill to tcg_ops
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (18 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 19/32] cpu: Move cpu_exec_* " Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 11:26 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 21/32] cpu: Move debug_excp_handler " Claudio Fontana
` (14 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Colin Xu, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Claudio Fontana
From: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
[claudio: wrapped in CONFIG_TCG]
Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
---
accel/tcg/cputlb.c | 6 +++---
accel/tcg/user-exec.c | 6 +++---
include/hw/core/cpu.h | 9 ---------
include/hw/core/tcg-cpu-ops.h | 12 ++++++++++++
target/alpha/cpu.c | 2 +-
target/arm/cpu.c | 2 +-
target/avr/cpu.c | 2 +-
target/cris/cpu.c | 2 +-
target/hppa/cpu.c | 2 +-
target/i386/tcg-cpu.c | 2 +-
target/lm32/cpu.c | 2 +-
target/m68k/cpu.c | 2 +-
target/microblaze/cpu.c | 2 +-
target/mips/cpu.c | 2 +-
target/moxie/cpu.c | 2 +-
target/nios2/cpu.c | 2 +-
target/openrisc/cpu.c | 2 +-
target/ppc/translate_init.c.inc | 2 +-
target/riscv/cpu.c | 2 +-
target/rx/cpu.c | 2 +-
target/s390x/cpu.c | 2 +-
target/sh4/cpu.c | 2 +-
target/sparc/cpu.c | 2 +-
target/tilegx/cpu.c | 2 +-
target/tricore/cpu.c | 2 +-
target/unicore32/cpu.c | 2 +-
target/xtensa/cpu.c | 2 +-
27 files changed, 41 insertions(+), 38 deletions(-)
diff --git a/accel/tcg/cputlb.c b/accel/tcg/cputlb.c
index 42ab79c1a5..2dc71b5528 100644
--- a/accel/tcg/cputlb.c
+++ b/accel/tcg/cputlb.c
@@ -1286,7 +1286,7 @@ static void tlb_fill(CPUState *cpu, target_ulong addr, int size,
* This is not a probe, so only valid return is success; failure
* should result in exception + longjmp to the cpu loop.
*/
- ok = cc->tlb_fill(cpu, addr, size, access_type, mmu_idx, false, retaddr);
+ ok = cc->tcg_ops.tlb_fill(cpu, addr, size, access_type, mmu_idx, false, retaddr);
assert(ok);
}
@@ -1557,8 +1557,8 @@ static int probe_access_internal(CPUArchState *env, target_ulong addr,
CPUState *cs = env_cpu(env);
CPUClass *cc = CPU_GET_CLASS(cs);
- if (!cc->tlb_fill(cs, addr, fault_size, access_type,
- mmu_idx, nonfault, retaddr)) {
+ if (!cc->tcg_ops.tlb_fill(cs, addr, fault_size, access_type,
+ mmu_idx, nonfault, retaddr)) {
/* Non-faulting page table read failed. */
*phost = NULL;
return TLB_INVALID_MASK;
diff --git a/accel/tcg/user-exec.c b/accel/tcg/user-exec.c
index 4ebe25461a..7f53992251 100644
--- a/accel/tcg/user-exec.c
+++ b/accel/tcg/user-exec.c
@@ -186,7 +186,7 @@ static inline int handle_cpu_signal(uintptr_t pc, siginfo_t *info,
clear_helper_retaddr();
cc = CPU_GET_CLASS(cpu);
- cc->tlb_fill(cpu, address, 0, access_type, MMU_USER_IDX, false, pc);
+ cc->tcg_ops.tlb_fill(cpu, address, 0, access_type, MMU_USER_IDX, false, pc);
g_assert_not_reached();
}
@@ -216,8 +216,8 @@ static int probe_access_internal(CPUArchState *env, target_ulong addr,
} else {
CPUState *cpu = env_cpu(env);
CPUClass *cc = CPU_GET_CLASS(cpu);
- cc->tlb_fill(cpu, addr, fault_size, access_type,
- MMU_USER_IDX, false, ra);
+ cc->tcg_ops.tlb_fill(cpu, addr, fault_size, access_type,
+ MMU_USER_IDX, false, ra);
g_assert_not_reached();
}
}
diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
index 52142e9094..c82ef261c6 100644
--- a/include/hw/core/cpu.h
+++ b/include/hw/core/cpu.h
@@ -110,12 +110,6 @@ struct TranslationBlock;
* If the target behaviour here is anything other than "set
* the PC register to the value passed in" then the target must
* also implement the synchronize_from_tb hook.
- * @tlb_fill: Callback for handling a softmmu tlb miss or user-only
- * address fault. For system mode, if the access is valid, call
- * tlb_set_page and return true; if the access is invalid, and
- * probe is true, return false; otherwise raise an exception and
- * do not return. For user-only mode, always raise an exception
- * and do not return.
* @get_phys_page_debug: Callback for obtaining a physical address.
* @get_phys_page_attrs_debug: Callback for obtaining a physical address and the
* associated memory transaction attributes to use for the access.
@@ -183,9 +177,6 @@ struct CPUClass {
void (*get_memory_mapping)(CPUState *cpu, MemoryMappingList *list,
Error **errp);
void (*set_pc)(CPUState *cpu, vaddr value);
- bool (*tlb_fill)(CPUState *cpu, vaddr address, int size,
- MMUAccessType access_type, int mmu_idx,
- bool probe, uintptr_t retaddr);
hwaddr (*get_phys_page_debug)(CPUState *cpu, vaddr addr);
hwaddr (*get_phys_page_attrs_debug)(CPUState *cpu, vaddr addr,
MemTxAttrs *attrs);
diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
index e12f32919b..2ea94acca0 100644
--- a/include/hw/core/tcg-cpu-ops.h
+++ b/include/hw/core/tcg-cpu-ops.h
@@ -37,6 +37,18 @@ typedef struct TcgCpuOperations {
void (*cpu_exec_exit)(CPUState *cpu);
/** @cpu_exec_interrupt: Callback for processing interrupts in cpu_exec */
bool (*cpu_exec_interrupt)(CPUState *cpu, int interrupt_request);
+ /**
+ * @tlb_fill: Handle a softmmu tlb miss or user-only address fault
+ *
+ * For system mode, if the access is valid, call tlb_set_page
+ * and return true; if the access is invalid, and probe is
+ * true, return false; otherwise raise an exception and do
+ * not return. For user-only mode, always raise an exception
+ * and do not return.
+ */
+ bool (*tlb_fill)(CPUState *cpu, vaddr address, int size,
+ MMUAccessType access_type, int mmu_idx,
+ bool probe, uintptr_t retaddr);
} TcgCpuOperations;
#endif /* TCG_CPU_OPS_H */
diff --git a/target/alpha/cpu.c b/target/alpha/cpu.c
index 4f206c154d..0369d5a99c 100644
--- a/target/alpha/cpu.c
+++ b/target/alpha/cpu.c
@@ -222,7 +222,7 @@ static void alpha_cpu_class_init(ObjectClass *oc, void *data)
cc->set_pc = alpha_cpu_set_pc;
cc->gdb_read_register = alpha_cpu_gdb_read_register;
cc->gdb_write_register = alpha_cpu_gdb_write_register;
- cc->tlb_fill = alpha_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = alpha_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->do_transaction_failed = alpha_cpu_do_transaction_failed;
cc->do_unaligned_access = alpha_cpu_do_unaligned_access;
diff --git a/target/arm/cpu.c b/target/arm/cpu.c
index 470b7b48c3..0dac097367 100644
--- a/target/arm/cpu.c
+++ b/target/arm/cpu.c
@@ -2262,7 +2262,7 @@ static void arm_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.initialize = arm_translate_init;
cc->tcg_ops.cpu_exec_interrupt = arm_cpu_exec_interrupt;
cc->tcg_ops.synchronize_from_tb = arm_cpu_synchronize_from_tb;
- cc->tlb_fill = arm_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = arm_cpu_tlb_fill;
cc->debug_excp_handler = arm_debug_excp_handler;
cc->debug_check_watchpoint = arm_debug_check_watchpoint;
cc->do_unaligned_access = arm_cpu_do_unaligned_access;
diff --git a/target/avr/cpu.c b/target/avr/cpu.c
index 277b00dbfc..699055de7c 100644
--- a/target/avr/cpu.c
+++ b/target/avr/cpu.c
@@ -203,7 +203,7 @@ static void avr_cpu_class_init(ObjectClass *oc, void *data)
cc->set_pc = avr_cpu_set_pc;
cc->memory_rw_debug = avr_cpu_memory_rw_debug;
cc->get_phys_page_debug = avr_cpu_get_phys_page_debug;
- cc->tlb_fill = avr_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = avr_cpu_tlb_fill;
cc->vmsd = &vms_avr_cpu;
cc->disas_set_info = avr_cpu_disas_set_info;
cc->tcg_ops.initialize = avr_cpu_tcg_init;
diff --git a/target/cris/cpu.c b/target/cris/cpu.c
index 7489fc20c8..9222717f3e 100644
--- a/target/cris/cpu.c
+++ b/target/cris/cpu.c
@@ -274,7 +274,7 @@ static void cris_cpu_class_init(ObjectClass *oc, void *data)
cc->set_pc = cris_cpu_set_pc;
cc->gdb_read_register = cris_cpu_gdb_read_register;
cc->gdb_write_register = cris_cpu_gdb_write_register;
- cc->tlb_fill = cris_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = cris_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->get_phys_page_debug = cris_cpu_get_phys_page_debug;
dc->vmsd = &vmstate_cris_cpu;
diff --git a/target/hppa/cpu.c b/target/hppa/cpu.c
index 61444753f2..e2d79f954e 100644
--- a/target/hppa/cpu.c
+++ b/target/hppa/cpu.c
@@ -146,7 +146,7 @@ static void hppa_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.synchronize_from_tb = hppa_cpu_synchronize_from_tb;
cc->gdb_read_register = hppa_cpu_gdb_read_register;
cc->gdb_write_register = hppa_cpu_gdb_write_register;
- cc->tlb_fill = hppa_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = hppa_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->get_phys_page_debug = hppa_cpu_get_phys_page_debug;
dc->vmsd = &vmstate_hppa_cpu;
diff --git a/target/i386/tcg-cpu.c b/target/i386/tcg-cpu.c
index 5e0f2a2fae..8606dd6a3e 100644
--- a/target/i386/tcg-cpu.c
+++ b/target/i386/tcg-cpu.c
@@ -64,7 +64,7 @@ void tcg_cpu_common_class_init(CPUClass *cc)
cc->tcg_ops.cpu_exec_enter = x86_cpu_exec_enter;
cc->tcg_ops.cpu_exec_exit = x86_cpu_exec_exit;
cc->tcg_ops.initialize = tcg_x86_init;
- cc->tlb_fill = x86_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = x86_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->debug_excp_handler = breakpoint_handler;
#endif
diff --git a/target/lm32/cpu.c b/target/lm32/cpu.c
index eea2d3e515..76dc728858 100644
--- a/target/lm32/cpu.c
+++ b/target/lm32/cpu.c
@@ -228,7 +228,7 @@ static void lm32_cpu_class_init(ObjectClass *oc, void *data)
cc->set_pc = lm32_cpu_set_pc;
cc->gdb_read_register = lm32_cpu_gdb_read_register;
cc->gdb_write_register = lm32_cpu_gdb_write_register;
- cc->tlb_fill = lm32_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = lm32_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->get_phys_page_debug = lm32_cpu_get_phys_page_debug;
cc->vmsd = &vmstate_lm32_cpu;
diff --git a/target/m68k/cpu.c b/target/m68k/cpu.c
index c0fa517fc3..bc109faa21 100644
--- a/target/m68k/cpu.c
+++ b/target/m68k/cpu.c
@@ -283,7 +283,7 @@ static void m68k_cpu_class_init(ObjectClass *c, void *data)
cc->set_pc = m68k_cpu_set_pc;
cc->gdb_read_register = m68k_cpu_gdb_read_register;
cc->gdb_write_register = m68k_cpu_gdb_write_register;
- cc->tlb_fill = m68k_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = m68k_cpu_tlb_fill;
#if defined(CONFIG_SOFTMMU)
cc->do_transaction_failed = m68k_cpu_transaction_failed;
cc->get_phys_page_debug = m68k_cpu_get_phys_page_debug;
diff --git a/target/microblaze/cpu.c b/target/microblaze/cpu.c
index 833d7f2d59..6e660a27b8 100644
--- a/target/microblaze/cpu.c
+++ b/target/microblaze/cpu.c
@@ -325,7 +325,7 @@ static void mb_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.synchronize_from_tb = mb_cpu_synchronize_from_tb;
cc->gdb_read_register = mb_cpu_gdb_read_register;
cc->gdb_write_register = mb_cpu_gdb_write_register;
- cc->tlb_fill = mb_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = mb_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->do_transaction_failed = mb_cpu_transaction_failed;
cc->get_phys_page_debug = mb_cpu_get_phys_page_debug;
diff --git a/target/mips/cpu.c b/target/mips/cpu.c
index 344ee402fc..1f7573d319 100644
--- a/target/mips/cpu.c
+++ b/target/mips/cpu.c
@@ -250,7 +250,7 @@ static void mips_cpu_class_init(ObjectClass *c, void *data)
cc->tcg_ops.initialize = mips_tcg_init;
cc->tcg_ops.cpu_exec_interrupt = mips_cpu_exec_interrupt;
cc->tcg_ops.synchronize_from_tb = mips_cpu_synchronize_from_tb;
- cc->tlb_fill = mips_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = mips_cpu_tlb_fill;
#endif
cc->gdb_num_core_regs = 73;
diff --git a/target/moxie/cpu.c b/target/moxie/cpu.c
index 224cfc8361..1177d092c1 100644
--- a/target/moxie/cpu.c
+++ b/target/moxie/cpu.c
@@ -110,7 +110,7 @@ static void moxie_cpu_class_init(ObjectClass *oc, void *data)
cc->do_interrupt = moxie_cpu_do_interrupt;
cc->dump_state = moxie_cpu_dump_state;
cc->set_pc = moxie_cpu_set_pc;
- cc->tlb_fill = moxie_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = moxie_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->get_phys_page_debug = moxie_cpu_get_phys_page_debug;
cc->vmsd = &vmstate_moxie_cpu;
diff --git a/target/nios2/cpu.c b/target/nios2/cpu.c
index 9eeb01fb5b..a96b74b00c 100644
--- a/target/nios2/cpu.c
+++ b/target/nios2/cpu.c
@@ -197,7 +197,7 @@ static void nios2_cpu_class_init(ObjectClass *oc, void *data)
cc->dump_state = nios2_cpu_dump_state;
cc->set_pc = nios2_cpu_set_pc;
cc->disas_set_info = nios2_cpu_disas_set_info;
- cc->tlb_fill = nios2_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = nios2_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->do_unaligned_access = nios2_cpu_do_unaligned_access;
cc->get_phys_page_debug = nios2_cpu_get_phys_page_debug;
diff --git a/target/openrisc/cpu.c b/target/openrisc/cpu.c
index df8a41f956..e6d1c9764b 100644
--- a/target/openrisc/cpu.c
+++ b/target/openrisc/cpu.c
@@ -160,7 +160,7 @@ static void openrisc_cpu_class_init(ObjectClass *oc, void *data)
cc->set_pc = openrisc_cpu_set_pc;
cc->gdb_read_register = openrisc_cpu_gdb_read_register;
cc->gdb_write_register = openrisc_cpu_gdb_write_register;
- cc->tlb_fill = openrisc_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = openrisc_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->get_phys_page_debug = openrisc_cpu_get_phys_page_debug;
dc->vmsd = &vmstate_openrisc_cpu;
diff --git a/target/ppc/translate_init.c.inc b/target/ppc/translate_init.c.inc
index e0edbac999..e83000b2fb 100644
--- a/target/ppc/translate_init.c.inc
+++ b/target/ppc/translate_init.c.inc
@@ -10946,7 +10946,7 @@ static void ppc_cpu_class_init(ObjectClass *oc, void *data)
#ifdef CONFIG_TCG
cc->tcg_ops.initialize = ppc_translate_init;
cc->tcg_ops.cpu_exec_interrupt = ppc_cpu_exec_interrupt;
- cc->tlb_fill = ppc_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = ppc_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->tcg_ops.cpu_exec_enter = ppc_cpu_exec_enter;
cc->tcg_ops.cpu_exec_exit = ppc_cpu_exec_exit;
diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index 15d58698a6..31a43f7bfb 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -562,7 +562,7 @@ static void riscv_cpu_class_init(ObjectClass *c, void *data)
cc->vmsd = &vmstate_riscv_cpu;
#endif
cc->tcg_ops.initialize = riscv_translate_init;
- cc->tlb_fill = riscv_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = riscv_cpu_tlb_fill;
device_class_set_props(dc, riscv_cpu_properties);
}
diff --git a/target/rx/cpu.c b/target/rx/cpu.c
index 3ba93590d2..c815533223 100644
--- a/target/rx/cpu.c
+++ b/target/rx/cpu.c
@@ -195,7 +195,7 @@ static void rx_cpu_class_init(ObjectClass *klass, void *data)
cc->get_phys_page_debug = rx_cpu_get_phys_page_debug;
cc->disas_set_info = rx_cpu_disas_set_info;
cc->tcg_ops.initialize = rx_translate_init;
- cc->tlb_fill = rx_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = rx_cpu_tlb_fill;
cc->gdb_num_core_regs = 26;
cc->gdb_core_xml_file = "rx-core.xml";
diff --git a/target/s390x/cpu.c b/target/s390x/cpu.c
index add2f4b21f..6cd2b30192 100644
--- a/target/s390x/cpu.c
+++ b/target/s390x/cpu.c
@@ -513,7 +513,7 @@ static void s390_cpu_class_init(ObjectClass *oc, void *data)
cc->disas_set_info = s390_cpu_disas_set_info;
#ifdef CONFIG_TCG
cc->tcg_ops.initialize = s390x_translate_init;
- cc->tlb_fill = s390_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = s390_cpu_tlb_fill;
#endif
cc->gdb_num_core_regs = S390_NUM_CORE_REGS;
diff --git a/target/sh4/cpu.c b/target/sh4/cpu.c
index 0574194cd0..7a9019edec 100644
--- a/target/sh4/cpu.c
+++ b/target/sh4/cpu.c
@@ -225,7 +225,7 @@ static void superh_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.synchronize_from_tb = superh_cpu_synchronize_from_tb;
cc->gdb_read_register = superh_cpu_gdb_read_register;
cc->gdb_write_register = superh_cpu_gdb_write_register;
- cc->tlb_fill = superh_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = superh_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->do_unaligned_access = superh_cpu_do_unaligned_access;
cc->get_phys_page_debug = superh_cpu_get_phys_page_debug;
diff --git a/target/sparc/cpu.c b/target/sparc/cpu.c
index c559f15e14..760e0ea92c 100644
--- a/target/sparc/cpu.c
+++ b/target/sparc/cpu.c
@@ -871,7 +871,7 @@ static void sparc_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.synchronize_from_tb = sparc_cpu_synchronize_from_tb;
cc->gdb_read_register = sparc_cpu_gdb_read_register;
cc->gdb_write_register = sparc_cpu_gdb_write_register;
- cc->tlb_fill = sparc_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = sparc_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->do_transaction_failed = sparc_cpu_do_transaction_failed;
cc->do_unaligned_access = sparc_cpu_do_unaligned_access;
diff --git a/target/tilegx/cpu.c b/target/tilegx/cpu.c
index 4c6176d26e..75b3a4bae3 100644
--- a/target/tilegx/cpu.c
+++ b/target/tilegx/cpu.c
@@ -151,7 +151,7 @@ static void tilegx_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.cpu_exec_interrupt = tilegx_cpu_exec_interrupt;
cc->dump_state = tilegx_cpu_dump_state;
cc->set_pc = tilegx_cpu_set_pc;
- cc->tlb_fill = tilegx_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = tilegx_cpu_tlb_fill;
cc->gdb_num_core_regs = 0;
cc->tcg_ops.initialize = tilegx_tcg_init;
}
diff --git a/target/tricore/cpu.c b/target/tricore/cpu.c
index 5edf96c600..89a14f81d7 100644
--- a/target/tricore/cpu.c
+++ b/target/tricore/cpu.c
@@ -165,7 +165,7 @@ static void tricore_cpu_class_init(ObjectClass *c, void *data)
cc->tcg_ops.synchronize_from_tb = tricore_cpu_synchronize_from_tb;
cc->get_phys_page_debug = tricore_cpu_get_phys_page_debug;
cc->tcg_ops.initialize = tricore_tcg_init;
- cc->tlb_fill = tricore_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = tricore_cpu_tlb_fill;
}
#define DEFINE_TRICORE_CPU_TYPE(cpu_model, initfn) \
diff --git a/target/unicore32/cpu.c b/target/unicore32/cpu.c
index 84c3419989..a57d315d2f 100644
--- a/target/unicore32/cpu.c
+++ b/target/unicore32/cpu.c
@@ -135,7 +135,7 @@ static void uc32_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.cpu_exec_interrupt = uc32_cpu_exec_interrupt;
cc->dump_state = uc32_cpu_dump_state;
cc->set_pc = uc32_cpu_set_pc;
- cc->tlb_fill = uc32_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = uc32_cpu_tlb_fill;
cc->get_phys_page_debug = uc32_cpu_get_phys_page_debug;
cc->tcg_ops.initialize = uc32_translate_init;
dc->vmsd = &vmstate_uc32_cpu;
diff --git a/target/xtensa/cpu.c b/target/xtensa/cpu.c
index 42a5e4ebe8..e764dbeb73 100644
--- a/target/xtensa/cpu.c
+++ b/target/xtensa/cpu.c
@@ -201,7 +201,7 @@ static void xtensa_cpu_class_init(ObjectClass *oc, void *data)
cc->gdb_read_register = xtensa_cpu_gdb_read_register;
cc->gdb_write_register = xtensa_cpu_gdb_write_register;
cc->gdb_stop_before_watchpoint = true;
- cc->tlb_fill = xtensa_cpu_tlb_fill;
+ cc->tcg_ops.tlb_fill = xtensa_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->do_unaligned_access = xtensa_cpu_do_unaligned_access;
cc->get_phys_page_debug = xtensa_cpu_get_phys_page_debug;
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 21/32] cpu: Move debug_excp_handler to tcg_ops
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (19 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 20/32] cpu: Move tlb_fill " Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 11:29 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 22/32] target/arm: do not use cc->do_interrupt for KVM directly Claudio Fontana
` (13 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Colin Xu, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Claudio Fontana
From: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
accel/tcg/cpu-exec.c | 4 ++--
include/hw/core/cpu.h | 2 --
include/hw/core/tcg-cpu-ops.h | 2 ++
target/arm/cpu.c | 2 +-
target/i386/tcg-cpu.c | 2 +-
target/lm32/cpu.c | 2 +-
target/s390x/cpu.c | 2 +-
target/xtensa/cpu.c | 2 +-
8 files changed, 9 insertions(+), 9 deletions(-)
diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
index edb349835e..343ae427a0 100644
--- a/accel/tcg/cpu-exec.c
+++ b/accel/tcg/cpu-exec.c
@@ -482,8 +482,8 @@ static inline void cpu_handle_debug_exception(CPUState *cpu)
}
}
- if (cc->debug_excp_handler) {
- cc->debug_excp_handler(cpu);
+ if (cc->tcg_ops.debug_excp_handler) {
+ cc->tcg_ops.debug_excp_handler(cpu);
}
}
diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
index c82ef261c6..f438abb785 100644
--- a/include/hw/core/cpu.h
+++ b/include/hw/core/cpu.h
@@ -121,7 +121,6 @@ struct TranslationBlock;
* @gdb_write_register: Callback for letting GDB write a register.
* @debug_check_watchpoint: Callback: return true if the architectural
* watchpoint whose address has matched should really fire.
- * @debug_excp_handler: Callback for handling debug exceptions.
* @write_elf64_note: Callback for writing a CPU-specific ELF note to a
* 64-bit VM coredump.
* @write_elf32_qemunote: Callback for writing a CPU- and QEMU-specific ELF
@@ -184,7 +183,6 @@ struct CPUClass {
int (*gdb_read_register)(CPUState *cpu, GByteArray *buf, int reg);
int (*gdb_write_register)(CPUState *cpu, uint8_t *buf, int reg);
bool (*debug_check_watchpoint)(CPUState *cpu, CPUWatchpoint *wp);
- void (*debug_excp_handler)(CPUState *cpu);
int (*write_elf64_note)(WriteCoreDumpFunction f, CPUState *cpu,
int cpuid, void *opaque);
diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
index 2ea94acca0..dbbc64418c 100644
--- a/include/hw/core/tcg-cpu-ops.h
+++ b/include/hw/core/tcg-cpu-ops.h
@@ -49,6 +49,8 @@ typedef struct TcgCpuOperations {
bool (*tlb_fill)(CPUState *cpu, vaddr address, int size,
MMUAccessType access_type, int mmu_idx,
bool probe, uintptr_t retaddr);
+ /** @debug_excp_handler: Callback for handling debug exceptions */
+ void (*debug_excp_handler)(CPUState *cpu);
} TcgCpuOperations;
#endif /* TCG_CPU_OPS_H */
diff --git a/target/arm/cpu.c b/target/arm/cpu.c
index 0dac097367..20cbfaea51 100644
--- a/target/arm/cpu.c
+++ b/target/arm/cpu.c
@@ -2263,7 +2263,7 @@ static void arm_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.cpu_exec_interrupt = arm_cpu_exec_interrupt;
cc->tcg_ops.synchronize_from_tb = arm_cpu_synchronize_from_tb;
cc->tcg_ops.tlb_fill = arm_cpu_tlb_fill;
- cc->debug_excp_handler = arm_debug_excp_handler;
+ cc->tcg_ops.debug_excp_handler = arm_debug_excp_handler;
cc->debug_check_watchpoint = arm_debug_check_watchpoint;
cc->do_unaligned_access = arm_cpu_do_unaligned_access;
#if !defined(CONFIG_USER_ONLY)
diff --git a/target/i386/tcg-cpu.c b/target/i386/tcg-cpu.c
index 8606dd6a3e..38ed8bf6d3 100644
--- a/target/i386/tcg-cpu.c
+++ b/target/i386/tcg-cpu.c
@@ -66,6 +66,6 @@ void tcg_cpu_common_class_init(CPUClass *cc)
cc->tcg_ops.initialize = tcg_x86_init;
cc->tcg_ops.tlb_fill = x86_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
- cc->debug_excp_handler = breakpoint_handler;
+ cc->tcg_ops.debug_excp_handler = breakpoint_handler;
#endif
}
diff --git a/target/lm32/cpu.c b/target/lm32/cpu.c
index 76dc728858..bbe1405e32 100644
--- a/target/lm32/cpu.c
+++ b/target/lm32/cpu.c
@@ -235,7 +235,7 @@ static void lm32_cpu_class_init(ObjectClass *oc, void *data)
#endif
cc->gdb_num_core_regs = 32 + 7;
cc->gdb_stop_before_watchpoint = true;
- cc->debug_excp_handler = lm32_debug_excp_handler;
+ cc->tcg_ops.debug_excp_handler = lm32_debug_excp_handler;
cc->disas_set_info = lm32_cpu_disas_set_info;
cc->tcg_ops.initialize = lm32_translate_init;
}
diff --git a/target/s390x/cpu.c b/target/s390x/cpu.c
index 6cd2b30192..04856076b3 100644
--- a/target/s390x/cpu.c
+++ b/target/s390x/cpu.c
@@ -506,7 +506,7 @@ static void s390_cpu_class_init(ObjectClass *oc, void *data)
cc->write_elf64_note = s390_cpu_write_elf64_note;
#ifdef CONFIG_TCG
cc->tcg_ops.cpu_exec_interrupt = s390_cpu_exec_interrupt;
- cc->debug_excp_handler = s390x_cpu_debug_excp_handler;
+ cc->tcg_ops.debug_excp_handler = s390x_cpu_debug_excp_handler;
cc->do_unaligned_access = s390x_cpu_do_unaligned_access;
#endif
#endif
diff --git a/target/xtensa/cpu.c b/target/xtensa/cpu.c
index e764dbeb73..b6f13ceb32 100644
--- a/target/xtensa/cpu.c
+++ b/target/xtensa/cpu.c
@@ -207,7 +207,7 @@ static void xtensa_cpu_class_init(ObjectClass *oc, void *data)
cc->get_phys_page_debug = xtensa_cpu_get_phys_page_debug;
cc->do_transaction_failed = xtensa_cpu_do_transaction_failed;
#endif
- cc->debug_excp_handler = xtensa_breakpoint_handler;
+ cc->tcg_ops.debug_excp_handler = xtensa_breakpoint_handler;
cc->disas_set_info = xtensa_cpu_disas_set_info;
cc->tcg_ops.initialize = xtensa_translate_init;
dc->vmsd = &vmstate_xtensa_cpu;
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 22/32] target/arm: do not use cc->do_interrupt for KVM directly
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (20 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 21/32] cpu: Move debug_excp_handler " Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 11:30 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 23/32] cpu: move cc->do_interrupt to tcg_ops Claudio Fontana
` (12 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Peter Maydell, Eduardo Habkost, Paul Durrant,
Alex Bennée, Jason Wang, Marcelo Tosatti, qemu-devel,
Peter Xu, Dario Faggioli, Cameron Esfahani, haxm-team,
Claudio Fontana, Anthony Perard, Bruce Rogers, Olaf Hering,
Emilio G . Cota, Colin Xu
cc->do_interrupt is in theory a TCG callback used in accel/tcg only,
to prepare the emulated architecture to take an interrupt as defined
in the hardware specifications,
but in reality the _do_interrupt style of functions in targets are
also occasionally reused by KVM to prepare the architecture state in a
similar way where userspace code has identified that it needs to
deliver an exception to the guest.
In the case of ARM, that includes:
1) the vcpu thread got a SIGBUS indicating a memory error,
and we need to deliver a Synchronous External Abort to the guest to
let it know about the error.
2) the kernel told us about a debug exception (breakpoint, watchpoint)
but it is not for one of QEMU's own gdbstub breakpoints/watchpoints
so it must be a breakpoint the guest itself has set up, therefore
we need to deliver it to the guest.
So in order to reuse code, the same arm_do_interrupt function is used.
This is all fine, but we need to avoid calling it using the callback
registered in CPUClass, since that one is now TCG-only.
Fortunately this is easily solved by replacing calls to
CPUClass::do_interrupt() with explicit calls to arm_do_interrupt().
Signed-off-by: Claudio Fontana <cfontana@suse.de>
Cc: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
---
target/arm/helper.c | 4 ++++
target/arm/kvm64.c | 4 ++--
2 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/target/arm/helper.c b/target/arm/helper.c
index 38cd35c049..bebaabf525 100644
--- a/target/arm/helper.c
+++ b/target/arm/helper.c
@@ -9895,6 +9895,10 @@ static void handle_semihosting(CPUState *cs)
* Do any appropriate logging, handle PSCI calls, and then hand off
* to the AArch64-entry or AArch32-entry function depending on the
* target exception level's register width.
+ *
+ * Note: this is used for both TCG (as the do_interrupt tcg op),
+ * and KVM to re-inject guest debug exceptions, and to
+ * inject a Synchronous-External-Abort.
*/
void arm_cpu_do_interrupt(CPUState *cs)
{
diff --git a/target/arm/kvm64.c b/target/arm/kvm64.c
index f74bac2457..2b17e4203d 100644
--- a/target/arm/kvm64.c
+++ b/target/arm/kvm64.c
@@ -960,7 +960,7 @@ static void kvm_inject_arm_sea(CPUState *c)
env->exception.syndrome = esr;
- cc->do_interrupt(c);
+ arm_cpu_do_interrupt(c);
}
#define AARCH64_CORE_REG(x) (KVM_REG_ARM64 | KVM_REG_SIZE_U64 | \
@@ -1545,7 +1545,7 @@ bool kvm_arm_handle_debug(CPUState *cs, struct kvm_debug_exit_arch *debug_exit)
env->exception.vaddress = debug_exit->far;
env->exception.target_el = 1;
qemu_mutex_lock_iothread();
- cc->do_interrupt(cs);
+ arm_cpu_do_interrupt(cs);
qemu_mutex_unlock_iothread();
return false;
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 23/32] cpu: move cc->do_interrupt to tcg_ops
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (21 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 22/32] target/arm: do not use cc->do_interrupt for KVM directly Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 11:43 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 24/32] cpu: move cc->transaction_failed " Claudio Fontana
` (11 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
accel/tcg/cpu-exec.c | 4 ++--
include/hw/core/cpu.h | 2 --
include/hw/core/tcg-cpu-ops.h | 3 +++
target/alpha/cpu.c | 2 +-
target/arm/cpu.c | 6 ++++--
target/arm/cpu_tcg.c | 9 ++++-----
target/avr/cpu.c | 2 +-
target/avr/helper.c | 4 ++--
target/cris/cpu.c | 12 ++++++------
target/cris/helper.c | 4 ++--
target/hppa/cpu.c | 2 +-
target/i386/tcg-cpu.c | 2 +-
target/lm32/cpu.c | 2 +-
target/m68k/cpu.c | 2 +-
target/microblaze/cpu.c | 2 +-
target/mips/cpu.c | 6 ++++--
target/moxie/cpu.c | 2 +-
target/nios2/cpu.c | 2 +-
target/openrisc/cpu.c | 2 +-
target/ppc/translate_init.c.inc | 2 +-
target/riscv/cpu.c | 2 +-
target/rx/cpu.c | 2 +-
target/s390x/cpu.c | 2 +-
target/sh4/cpu.c | 2 +-
target/sparc/cpu.c | 2 +-
target/tilegx/cpu.c | 2 +-
target/unicore32/cpu.c | 2 +-
target/xtensa/cpu.c | 2 +-
28 files changed, 46 insertions(+), 42 deletions(-)
diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
index 343ae427a0..68ccaa76f8 100644
--- a/accel/tcg/cpu-exec.c
+++ b/accel/tcg/cpu-exec.c
@@ -517,7 +517,7 @@ static inline bool cpu_handle_exception(CPUState *cpu, int *ret)
loop */
#if defined(TARGET_I386)
CPUClass *cc = CPU_GET_CLASS(cpu);
- cc->do_interrupt(cpu);
+ cc->tcg_ops.do_interrupt(cpu);
#endif
*ret = cpu->exception_index;
cpu->exception_index = -1;
@@ -526,7 +526,7 @@ static inline bool cpu_handle_exception(CPUState *cpu, int *ret)
if (replay_exception()) {
CPUClass *cc = CPU_GET_CLASS(cpu);
qemu_mutex_lock_iothread();
- cc->do_interrupt(cpu);
+ cc->tcg_ops.do_interrupt(cpu);
qemu_mutex_unlock_iothread();
cpu->exception_index = -1;
diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
index f438abb785..b178549d48 100644
--- a/include/hw/core/cpu.h
+++ b/include/hw/core/cpu.h
@@ -87,7 +87,6 @@ struct TranslationBlock;
* @parse_features: Callback to parse command line arguments.
* @reset_dump_flags: #CPUDumpFlags to use for reset logging.
* @has_work: Callback for checking if there is work to do.
- * @do_interrupt: Callback for interrupt handling.
* @do_unaligned_access: Callback for unaligned access handling, if
* the target defines #TARGET_ALIGNED_ONLY.
* @do_transaction_failed: Callback for handling failed memory transactions
@@ -157,7 +156,6 @@ struct CPUClass {
int reset_dump_flags;
bool (*has_work)(CPUState *cpu);
- void (*do_interrupt)(CPUState *cpu);
void (*do_unaligned_access)(CPUState *cpu, vaddr addr,
MMUAccessType access_type,
int mmu_idx, uintptr_t retaddr);
diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
index dbbc64418c..6b9b72e785 100644
--- a/include/hw/core/tcg-cpu-ops.h
+++ b/include/hw/core/tcg-cpu-ops.h
@@ -37,6 +37,9 @@ typedef struct TcgCpuOperations {
void (*cpu_exec_exit)(CPUState *cpu);
/** @cpu_exec_interrupt: Callback for processing interrupts in cpu_exec */
bool (*cpu_exec_interrupt)(CPUState *cpu, int interrupt_request);
+ /** @do_interrupt: Callback for interrupt handling. */
+ void (*do_interrupt)(CPUState *cpu);
+
/**
* @tlb_fill: Handle a softmmu tlb miss or user-only address fault
*
diff --git a/target/alpha/cpu.c b/target/alpha/cpu.c
index 0369d5a99c..fde5825ebc 100644
--- a/target/alpha/cpu.c
+++ b/target/alpha/cpu.c
@@ -217,7 +217,7 @@ static void alpha_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = alpha_cpu_class_by_name;
cc->has_work = alpha_cpu_has_work;
- cc->do_interrupt = alpha_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = alpha_cpu_do_interrupt;
cc->dump_state = alpha_cpu_dump_state;
cc->set_pc = alpha_cpu_set_pc;
cc->gdb_read_register = alpha_cpu_gdb_read_register;
diff --git a/target/arm/cpu.c b/target/arm/cpu.c
index 20cbfaea51..826cf8351b 100644
--- a/target/arm/cpu.c
+++ b/target/arm/cpu.c
@@ -535,6 +535,7 @@ static inline bool arm_excp_unmasked(CPUState *cs, unsigned int excp_idx,
return unmasked || pstate_unmasked;
}
+#ifdef CONFIG_TCG
bool arm_cpu_exec_interrupt(CPUState *cs, int interrupt_request)
{
CPUClass *cc = CPU_GET_CLASS(cs);
@@ -584,9 +585,10 @@ bool arm_cpu_exec_interrupt(CPUState *cs, int interrupt_request)
found:
cs->exception_index = excp_idx;
env->exception.target_el = target_el;
- cc->do_interrupt(cs);
+ cc->tcg_ops.do_interrupt(cs);
return true;
}
+#endif /* CONFIG_TCG */
void arm_cpu_update_virq(ARMCPU *cpu)
{
@@ -2244,7 +2246,6 @@ static void arm_cpu_class_init(ObjectClass *oc, void *data)
cc->gdb_read_register = arm_cpu_gdb_read_register;
cc->gdb_write_register = arm_cpu_gdb_write_register;
#ifndef CONFIG_USER_ONLY
- cc->do_interrupt = arm_cpu_do_interrupt;
cc->get_phys_page_attrs_debug = arm_cpu_get_phys_page_attrs_debug;
cc->asidx_from_attrs = arm_asidx_from_attrs;
cc->vmsd = &vmstate_arm_cpu;
@@ -2269,6 +2270,7 @@ static void arm_cpu_class_init(ObjectClass *oc, void *data)
#if !defined(CONFIG_USER_ONLY)
cc->do_transaction_failed = arm_cpu_do_transaction_failed;
cc->adjust_watchpoint_address = arm_adjust_watchpoint_address;
+ cc->tcg_ops.do_interrupt = arm_cpu_do_interrupt;
#endif /* CONFIG_TCG && !CONFIG_USER_ONLY */
#endif
}
diff --git a/target/arm/cpu_tcg.c b/target/arm/cpu_tcg.c
index a6538d3a96..e64f18a5ee 100644
--- a/target/arm/cpu_tcg.c
+++ b/target/arm/cpu_tcg.c
@@ -33,7 +33,7 @@ static bool arm_v7m_cpu_exec_interrupt(CPUState *cs, int interrupt_request)
if (interrupt_request & CPU_INTERRUPT_HARD
&& (armv7m_nvic_can_take_pending_exception(env->nvic))) {
cs->exception_index = EXCP_IRQ;
- cc->do_interrupt(cs);
+ cc->tcg_ops.do_interrupt(cs);
ret = true;
}
return ret;
@@ -624,12 +624,11 @@ static void arm_v7m_class_init(ObjectClass *oc, void *data)
CPUClass *cc = CPU_CLASS(oc);
acc->info = data;
-#ifndef CONFIG_USER_ONLY
- cc->do_interrupt = arm_v7m_cpu_do_interrupt;
-#endif
-
#ifdef CONFIG_TCG
cc->tcg_ops.cpu_exec_interrupt = arm_v7m_cpu_exec_interrupt;
+#ifndef CONFIG_USER_ONLY
+ cc->tcg_ops.do_interrupt = arm_v7m_cpu_do_interrupt;
+#endif
#endif /* CONFIG_TCG */
cc->gdb_core_xml_file = "arm-m-profile.xml";
diff --git a/target/avr/cpu.c b/target/avr/cpu.c
index 699055de7c..5a5ae68444 100644
--- a/target/avr/cpu.c
+++ b/target/avr/cpu.c
@@ -197,7 +197,7 @@ static void avr_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = avr_cpu_class_by_name;
cc->has_work = avr_cpu_has_work;
- cc->do_interrupt = avr_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = avr_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = avr_cpu_exec_interrupt;
cc->dump_state = avr_cpu_dump_state;
cc->set_pc = avr_cpu_set_pc;
diff --git a/target/avr/helper.c b/target/avr/helper.c
index d96d14372b..69d3b6181f 100644
--- a/target/avr/helper.c
+++ b/target/avr/helper.c
@@ -34,7 +34,7 @@ bool avr_cpu_exec_interrupt(CPUState *cs, int interrupt_request)
if (interrupt_request & CPU_INTERRUPT_RESET) {
if (cpu_interrupts_enabled(env)) {
cs->exception_index = EXCP_RESET;
- cc->do_interrupt(cs);
+ cc->tcg_ops.do_interrupt(cs);
cs->interrupt_request &= ~CPU_INTERRUPT_RESET;
@@ -45,7 +45,7 @@ bool avr_cpu_exec_interrupt(CPUState *cs, int interrupt_request)
if (cpu_interrupts_enabled(env) && env->intsrc != 0) {
int index = ctz32(env->intsrc);
cs->exception_index = EXCP_INT(index);
- cc->do_interrupt(cs);
+ cc->tcg_ops.do_interrupt(cs);
env->intsrc &= env->intsrc - 1; /* clear the interrupt */
cs->interrupt_request &= ~CPU_INTERRUPT_HARD;
diff --git a/target/cris/cpu.c b/target/cris/cpu.c
index 9222717f3e..b65743e8ca 100644
--- a/target/cris/cpu.c
+++ b/target/cris/cpu.c
@@ -199,7 +199,7 @@ static void crisv8_cpu_class_init(ObjectClass *oc, void *data)
CRISCPUClass *ccc = CRIS_CPU_CLASS(oc);
ccc->vr = 8;
- cc->do_interrupt = crisv10_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = crisv10_cpu_do_interrupt;
cc->gdb_read_register = crisv10_cpu_gdb_read_register;
cc->tcg_ops.initialize = cris_initialize_crisv10_tcg;
}
@@ -210,7 +210,7 @@ static void crisv9_cpu_class_init(ObjectClass *oc, void *data)
CRISCPUClass *ccc = CRIS_CPU_CLASS(oc);
ccc->vr = 9;
- cc->do_interrupt = crisv10_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = crisv10_cpu_do_interrupt;
cc->gdb_read_register = crisv10_cpu_gdb_read_register;
cc->tcg_ops.initialize = cris_initialize_crisv10_tcg;
}
@@ -221,7 +221,7 @@ static void crisv10_cpu_class_init(ObjectClass *oc, void *data)
CRISCPUClass *ccc = CRIS_CPU_CLASS(oc);
ccc->vr = 10;
- cc->do_interrupt = crisv10_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = crisv10_cpu_do_interrupt;
cc->gdb_read_register = crisv10_cpu_gdb_read_register;
cc->tcg_ops.initialize = cris_initialize_crisv10_tcg;
}
@@ -232,7 +232,7 @@ static void crisv11_cpu_class_init(ObjectClass *oc, void *data)
CRISCPUClass *ccc = CRIS_CPU_CLASS(oc);
ccc->vr = 11;
- cc->do_interrupt = crisv10_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = crisv10_cpu_do_interrupt;
cc->gdb_read_register = crisv10_cpu_gdb_read_register;
cc->tcg_ops.initialize = cris_initialize_crisv10_tcg;
}
@@ -243,7 +243,7 @@ static void crisv17_cpu_class_init(ObjectClass *oc, void *data)
CRISCPUClass *ccc = CRIS_CPU_CLASS(oc);
ccc->vr = 17;
- cc->do_interrupt = crisv10_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = crisv10_cpu_do_interrupt;
cc->gdb_read_register = crisv10_cpu_gdb_read_register;
cc->tcg_ops.initialize = cris_initialize_crisv10_tcg;
}
@@ -268,7 +268,7 @@ static void cris_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = cris_cpu_class_by_name;
cc->has_work = cris_cpu_has_work;
- cc->do_interrupt = cris_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = cris_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = cris_cpu_exec_interrupt;
cc->dump_state = cris_cpu_dump_state;
cc->set_pc = cris_cpu_set_pc;
diff --git a/target/cris/helper.c b/target/cris/helper.c
index ed45c3d9b7..1f4d6f7d45 100644
--- a/target/cris/helper.c
+++ b/target/cris/helper.c
@@ -299,7 +299,7 @@ bool cris_cpu_exec_interrupt(CPUState *cs, int interrupt_request)
&& (env->pregs[PR_CCS] & I_FLAG)
&& !env->locked_irq) {
cs->exception_index = EXCP_IRQ;
- cc->do_interrupt(cs);
+ cc->tcg_ops.do_interrupt(cs);
ret = true;
}
if (interrupt_request & CPU_INTERRUPT_NMI) {
@@ -311,7 +311,7 @@ bool cris_cpu_exec_interrupt(CPUState *cs, int interrupt_request)
}
if ((env->pregs[PR_CCS] & m_flag_archval)) {
cs->exception_index = EXCP_NMI;
- cc->do_interrupt(cs);
+ cc->tcg_ops.do_interrupt(cs);
ret = true;
}
}
diff --git a/target/hppa/cpu.c b/target/hppa/cpu.c
index e2d79f954e..cce6ae6213 100644
--- a/target/hppa/cpu.c
+++ b/target/hppa/cpu.c
@@ -139,7 +139,7 @@ static void hppa_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = hppa_cpu_class_by_name;
cc->has_work = hppa_cpu_has_work;
- cc->do_interrupt = hppa_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = hppa_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = hppa_cpu_exec_interrupt;
cc->dump_state = hppa_cpu_dump_state;
cc->set_pc = hppa_cpu_set_pc;
diff --git a/target/i386/tcg-cpu.c b/target/i386/tcg-cpu.c
index 38ed8bf6d3..d2dd521612 100644
--- a/target/i386/tcg-cpu.c
+++ b/target/i386/tcg-cpu.c
@@ -58,7 +58,7 @@ static void x86_cpu_synchronize_from_tb(CPUState *cs, TranslationBlock *tb)
void tcg_cpu_common_class_init(CPUClass *cc)
{
- cc->do_interrupt = x86_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = x86_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = x86_cpu_exec_interrupt;
cc->tcg_ops.synchronize_from_tb = x86_cpu_synchronize_from_tb;
cc->tcg_ops.cpu_exec_enter = x86_cpu_exec_enter;
diff --git a/target/lm32/cpu.c b/target/lm32/cpu.c
index bbe1405e32..fb3761b749 100644
--- a/target/lm32/cpu.c
+++ b/target/lm32/cpu.c
@@ -222,7 +222,7 @@ static void lm32_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = lm32_cpu_class_by_name;
cc->has_work = lm32_cpu_has_work;
- cc->do_interrupt = lm32_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = lm32_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = lm32_cpu_exec_interrupt;
cc->dump_state = lm32_cpu_dump_state;
cc->set_pc = lm32_cpu_set_pc;
diff --git a/target/m68k/cpu.c b/target/m68k/cpu.c
index bc109faa21..3e84de772c 100644
--- a/target/m68k/cpu.c
+++ b/target/m68k/cpu.c
@@ -277,7 +277,7 @@ static void m68k_cpu_class_init(ObjectClass *c, void *data)
cc->class_by_name = m68k_cpu_class_by_name;
cc->has_work = m68k_cpu_has_work;
- cc->do_interrupt = m68k_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = m68k_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = m68k_cpu_exec_interrupt;
cc->dump_state = m68k_cpu_dump_state;
cc->set_pc = m68k_cpu_set_pc;
diff --git a/target/microblaze/cpu.c b/target/microblaze/cpu.c
index 6e660a27b8..4f95248b2e 100644
--- a/target/microblaze/cpu.c
+++ b/target/microblaze/cpu.c
@@ -317,7 +317,7 @@ static void mb_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = mb_cpu_class_by_name;
cc->has_work = mb_cpu_has_work;
- cc->do_interrupt = mb_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = mb_cpu_do_interrupt;
cc->do_unaligned_access = mb_cpu_do_unaligned_access;
cc->tcg_ops.cpu_exec_interrupt = mb_cpu_exec_interrupt;
cc->dump_state = mb_cpu_dump_state;
diff --git a/target/mips/cpu.c b/target/mips/cpu.c
index 1f7573d319..dd5fb78e6c 100644
--- a/target/mips/cpu.c
+++ b/target/mips/cpu.c
@@ -234,7 +234,6 @@ static void mips_cpu_class_init(ObjectClass *c, void *data)
cc->class_by_name = mips_cpu_class_by_name;
cc->has_work = mips_cpu_has_work;
- cc->do_interrupt = mips_cpu_do_interrupt;
cc->dump_state = mips_cpu_dump_state;
cc->set_pc = mips_cpu_set_pc;
cc->gdb_read_register = mips_cpu_gdb_read_register;
@@ -251,7 +250,10 @@ static void mips_cpu_class_init(ObjectClass *c, void *data)
cc->tcg_ops.cpu_exec_interrupt = mips_cpu_exec_interrupt;
cc->tcg_ops.synchronize_from_tb = mips_cpu_synchronize_from_tb;
cc->tcg_ops.tlb_fill = mips_cpu_tlb_fill;
-#endif
+#ifndef CONFIG_USER_ONLY
+ cc->tcg_ops.do_interrupt = mips_cpu_do_interrupt;
+#endif /* CONFIG_USER_ONLY */
+#endif /* CONFIG_TCG */
cc->gdb_num_core_regs = 73;
cc->gdb_stop_before_watchpoint = true;
diff --git a/target/moxie/cpu.c b/target/moxie/cpu.c
index 1177d092c1..36bef4d357 100644
--- a/target/moxie/cpu.c
+++ b/target/moxie/cpu.c
@@ -107,7 +107,7 @@ static void moxie_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = moxie_cpu_class_by_name;
cc->has_work = moxie_cpu_has_work;
- cc->do_interrupt = moxie_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = moxie_cpu_do_interrupt;
cc->dump_state = moxie_cpu_dump_state;
cc->set_pc = moxie_cpu_set_pc;
cc->tcg_ops.tlb_fill = moxie_cpu_tlb_fill;
diff --git a/target/nios2/cpu.c b/target/nios2/cpu.c
index a96b74b00c..2b959f0e49 100644
--- a/target/nios2/cpu.c
+++ b/target/nios2/cpu.c
@@ -192,7 +192,7 @@ static void nios2_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = nios2_cpu_class_by_name;
cc->has_work = nios2_cpu_has_work;
- cc->do_interrupt = nios2_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = nios2_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = nios2_cpu_exec_interrupt;
cc->dump_state = nios2_cpu_dump_state;
cc->set_pc = nios2_cpu_set_pc;
diff --git a/target/openrisc/cpu.c b/target/openrisc/cpu.c
index e6d1c9764b..27105c5c09 100644
--- a/target/openrisc/cpu.c
+++ b/target/openrisc/cpu.c
@@ -154,7 +154,7 @@ static void openrisc_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = openrisc_cpu_class_by_name;
cc->has_work = openrisc_cpu_has_work;
- cc->do_interrupt = openrisc_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = openrisc_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = openrisc_cpu_exec_interrupt;
cc->dump_state = openrisc_cpu_dump_state;
cc->set_pc = openrisc_cpu_set_pc;
diff --git a/target/ppc/translate_init.c.inc b/target/ppc/translate_init.c.inc
index e83000b2fb..0471da0d08 100644
--- a/target/ppc/translate_init.c.inc
+++ b/target/ppc/translate_init.c.inc
@@ -10908,7 +10908,6 @@ static void ppc_cpu_class_init(ObjectClass *oc, void *data)
pcc->parent_parse_features = cc->parse_features;
cc->parse_features = ppc_cpu_parse_featurestr;
cc->has_work = ppc_cpu_has_work;
- cc->do_interrupt = ppc_cpu_do_interrupt;
cc->dump_state = ppc_cpu_dump_state;
cc->dump_statistics = ppc_cpu_dump_statistics;
cc->set_pc = ppc_cpu_set_pc;
@@ -10946,6 +10945,7 @@ static void ppc_cpu_class_init(ObjectClass *oc, void *data)
#ifdef CONFIG_TCG
cc->tcg_ops.initialize = ppc_translate_init;
cc->tcg_ops.cpu_exec_interrupt = ppc_cpu_exec_interrupt;
+ cc->tcg_ops.do_interrupt = ppc_cpu_do_interrupt;
cc->tcg_ops.tlb_fill = ppc_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->tcg_ops.cpu_exec_enter = ppc_cpu_exec_enter;
diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index 31a43f7bfb..6a815569cc 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -539,7 +539,7 @@ static void riscv_cpu_class_init(ObjectClass *c, void *data)
cc->class_by_name = riscv_cpu_class_by_name;
cc->has_work = riscv_cpu_has_work;
- cc->do_interrupt = riscv_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = riscv_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = riscv_cpu_exec_interrupt;
cc->dump_state = riscv_cpu_dump_state;
cc->set_pc = riscv_cpu_set_pc;
diff --git a/target/rx/cpu.c b/target/rx/cpu.c
index c815533223..dc7d1c3c57 100644
--- a/target/rx/cpu.c
+++ b/target/rx/cpu.c
@@ -185,7 +185,7 @@ static void rx_cpu_class_init(ObjectClass *klass, void *data)
cc->class_by_name = rx_cpu_class_by_name;
cc->has_work = rx_cpu_has_work;
- cc->do_interrupt = rx_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = rx_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = rx_cpu_exec_interrupt;
cc->dump_state = rx_cpu_dump_state;
cc->set_pc = rx_cpu_set_pc;
diff --git a/target/s390x/cpu.c b/target/s390x/cpu.c
index 04856076b3..b838bd61a4 100644
--- a/target/s390x/cpu.c
+++ b/target/s390x/cpu.c
@@ -493,7 +493,7 @@ static void s390_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = s390_cpu_class_by_name,
cc->has_work = s390_cpu_has_work;
#ifdef CONFIG_TCG
- cc->do_interrupt = s390_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = s390_cpu_do_interrupt;
#endif
cc->dump_state = s390_cpu_dump_state;
cc->set_pc = s390_cpu_set_pc;
diff --git a/target/sh4/cpu.c b/target/sh4/cpu.c
index 7a9019edec..ff835d4bc1 100644
--- a/target/sh4/cpu.c
+++ b/target/sh4/cpu.c
@@ -218,7 +218,7 @@ static void superh_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = superh_cpu_class_by_name;
cc->has_work = superh_cpu_has_work;
- cc->do_interrupt = superh_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = superh_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = superh_cpu_exec_interrupt;
cc->dump_state = superh_cpu_dump_state;
cc->set_pc = superh_cpu_set_pc;
diff --git a/target/sparc/cpu.c b/target/sparc/cpu.c
index 760e0ea92c..3f23aa5962 100644
--- a/target/sparc/cpu.c
+++ b/target/sparc/cpu.c
@@ -861,7 +861,7 @@ static void sparc_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = sparc_cpu_class_by_name;
cc->parse_features = sparc_cpu_parse_features;
cc->has_work = sparc_cpu_has_work;
- cc->do_interrupt = sparc_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = sparc_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = sparc_cpu_exec_interrupt;
cc->dump_state = sparc_cpu_dump_state;
#if !defined(TARGET_SPARC64) && !defined(CONFIG_USER_ONLY)
diff --git a/target/tilegx/cpu.c b/target/tilegx/cpu.c
index 75b3a4bae3..7d4ead4ef1 100644
--- a/target/tilegx/cpu.c
+++ b/target/tilegx/cpu.c
@@ -147,7 +147,7 @@ static void tilegx_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = tilegx_cpu_class_by_name;
cc->has_work = tilegx_cpu_has_work;
- cc->do_interrupt = tilegx_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = tilegx_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = tilegx_cpu_exec_interrupt;
cc->dump_state = tilegx_cpu_dump_state;
cc->set_pc = tilegx_cpu_set_pc;
diff --git a/target/unicore32/cpu.c b/target/unicore32/cpu.c
index a57d315d2f..e27ffc571a 100644
--- a/target/unicore32/cpu.c
+++ b/target/unicore32/cpu.c
@@ -131,7 +131,7 @@ static void uc32_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = uc32_cpu_class_by_name;
cc->has_work = uc32_cpu_has_work;
- cc->do_interrupt = uc32_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = uc32_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = uc32_cpu_exec_interrupt;
cc->dump_state = uc32_cpu_dump_state;
cc->set_pc = uc32_cpu_set_pc;
diff --git a/target/xtensa/cpu.c b/target/xtensa/cpu.c
index b6f13ceb32..3ff025f0fe 100644
--- a/target/xtensa/cpu.c
+++ b/target/xtensa/cpu.c
@@ -194,7 +194,7 @@ static void xtensa_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = xtensa_cpu_class_by_name;
cc->has_work = xtensa_cpu_has_work;
- cc->do_interrupt = xtensa_cpu_do_interrupt;
+ cc->tcg_ops.do_interrupt = xtensa_cpu_do_interrupt;
cc->tcg_ops.cpu_exec_interrupt = xtensa_cpu_exec_interrupt;
cc->dump_state = xtensa_cpu_dump_state;
cc->set_pc = xtensa_cpu_set_pc;
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 24/32] cpu: move cc->transaction_failed to tcg_ops
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (22 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 23/32] cpu: move cc->do_interrupt to tcg_ops Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 9:31 ` Philippe Mathieu-Daudé
2020-12-09 12:03 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 25/32] cpu: move do_unaligned_access " Claudio Fontana
` (10 subsequent siblings)
34 siblings, 2 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
hw/mips/jazz.c | 4 ++--
include/hw/core/cpu.h | 18 +++++++-----------
include/hw/core/tcg-cpu-ops.h | 9 +++++++++
target/alpha/cpu.c | 2 +-
target/arm/cpu.c | 4 ++--
target/m68k/cpu.c | 2 +-
target/microblaze/cpu.c | 2 +-
target/mips/cpu.c | 2 +-
target/riscv/cpu.c | 2 +-
target/sparc/cpu.c | 2 +-
target/xtensa/cpu.c | 2 +-
11 files changed, 27 insertions(+), 22 deletions(-)
diff --git a/hw/mips/jazz.c b/hw/mips/jazz.c
index 71448f72ac..53e195b406 100644
--- a/hw/mips/jazz.c
+++ b/hw/mips/jazz.c
@@ -204,8 +204,8 @@ static void mips_jazz_init(MachineState *machine,
* memory region that catches all memory accesses, as we do on Malta.
*/
cc = CPU_GET_CLASS(cpu);
- real_do_transaction_failed = cc->do_transaction_failed;
- cc->do_transaction_failed = mips_jazz_do_transaction_failed;
+ real_do_transaction_failed = cc->tcg_ops.do_transaction_failed;
+ cc->tcg_ops.do_transaction_failed = mips_jazz_do_transaction_failed;
/* allocate RAM */
memory_region_add_subregion(address_space, 0, machine->ram);
diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
index b178549d48..0c6c17fb48 100644
--- a/include/hw/core/cpu.h
+++ b/include/hw/core/cpu.h
@@ -89,8 +89,6 @@ struct TranslationBlock;
* @has_work: Callback for checking if there is work to do.
* @do_unaligned_access: Callback for unaligned access handling, if
* the target defines #TARGET_ALIGNED_ONLY.
- * @do_transaction_failed: Callback for handling failed memory transactions
- * (ie bus faults or external aborts; not MMU faults)
* @virtio_is_big_endian: Callback to return %true if a CPU which supports
* runtime configurable endianness is currently big-endian. Non-configurable
* CPUs can use the default implementation of this method. This method should
@@ -159,10 +157,6 @@ struct CPUClass {
void (*do_unaligned_access)(CPUState *cpu, vaddr addr,
MMUAccessType access_type,
int mmu_idx, uintptr_t retaddr);
- void (*do_transaction_failed)(CPUState *cpu, hwaddr physaddr, vaddr addr,
- unsigned size, MMUAccessType access_type,
- int mmu_idx, MemTxAttrs attrs,
- MemTxResult response, uintptr_t retaddr);
bool (*virtio_is_big_endian)(CPUState *cpu);
int (*memory_rw_debug)(CPUState *cpu, vaddr addr,
uint8_t *buf, int len, bool is_write);
@@ -848,7 +842,7 @@ static inline void cpu_unaligned_access(CPUState *cpu, vaddr addr,
cc->do_unaligned_access(cpu, addr, access_type, mmu_idx, retaddr);
}
-
+#ifdef CONFIG_TCG
static inline void cpu_transaction_failed(CPUState *cpu, hwaddr physaddr,
vaddr addr, unsigned size,
MMUAccessType access_type,
@@ -858,12 +852,14 @@ static inline void cpu_transaction_failed(CPUState *cpu, hwaddr physaddr,
{
CPUClass *cc = CPU_GET_CLASS(cpu);
- if (!cpu->ignore_memory_transaction_failures && cc->do_transaction_failed) {
- cc->do_transaction_failed(cpu, physaddr, addr, size, access_type,
- mmu_idx, attrs, response, retaddr);
+ if (!cpu->ignore_memory_transaction_failures &&
+ cc->tcg_ops.do_transaction_failed) {
+ cc->tcg_ops.do_transaction_failed(cpu, physaddr, addr, size, access_type,
+ mmu_idx, attrs, response, retaddr);
}
}
-#endif
+#endif /* CONFIG_TCG */
+#endif /* CONFIG_SOFTMMU */
#endif /* NEED_CPU_H */
diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
index 6b9b72e785..0cc0927738 100644
--- a/include/hw/core/tcg-cpu-ops.h
+++ b/include/hw/core/tcg-cpu-ops.h
@@ -40,6 +40,15 @@ typedef struct TcgCpuOperations {
/** @do_interrupt: Callback for interrupt handling. */
void (*do_interrupt)(CPUState *cpu);
+ /**
+ * @do_transaction_failed: Callback for handling failed memory transactions
+ * (ie bus faults or external aborts; not MMU faults)
+ */
+ void (*do_transaction_failed)(CPUState *cpu, hwaddr physaddr, vaddr addr,
+ unsigned size, MMUAccessType access_type,
+ int mmu_idx, MemTxAttrs attrs,
+ MemTxResult response, uintptr_t retaddr);
+
/**
* @tlb_fill: Handle a softmmu tlb miss or user-only address fault
*
diff --git a/target/alpha/cpu.c b/target/alpha/cpu.c
index fde5825ebc..3e651b246f 100644
--- a/target/alpha/cpu.c
+++ b/target/alpha/cpu.c
@@ -224,7 +224,7 @@ static void alpha_cpu_class_init(ObjectClass *oc, void *data)
cc->gdb_write_register = alpha_cpu_gdb_write_register;
cc->tcg_ops.tlb_fill = alpha_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
- cc->do_transaction_failed = alpha_cpu_do_transaction_failed;
+ cc->tcg_ops.do_transaction_failed = alpha_cpu_do_transaction_failed;
cc->do_unaligned_access = alpha_cpu_do_unaligned_access;
cc->get_phys_page_debug = alpha_cpu_get_phys_page_debug;
dc->vmsd = &vmstate_alpha_cpu;
diff --git a/target/arm/cpu.c b/target/arm/cpu.c
index 826cf8351b..5e9202ed86 100644
--- a/target/arm/cpu.c
+++ b/target/arm/cpu.c
@@ -2268,11 +2268,11 @@ static void arm_cpu_class_init(ObjectClass *oc, void *data)
cc->debug_check_watchpoint = arm_debug_check_watchpoint;
cc->do_unaligned_access = arm_cpu_do_unaligned_access;
#if !defined(CONFIG_USER_ONLY)
- cc->do_transaction_failed = arm_cpu_do_transaction_failed;
+ cc->tcg_ops.do_transaction_failed = arm_cpu_do_transaction_failed;
cc->adjust_watchpoint_address = arm_adjust_watchpoint_address;
cc->tcg_ops.do_interrupt = arm_cpu_do_interrupt;
#endif /* CONFIG_TCG && !CONFIG_USER_ONLY */
-#endif
+#endif /* CONFIG_TCG */
}
#ifdef CONFIG_KVM
diff --git a/target/m68k/cpu.c b/target/m68k/cpu.c
index 3e84de772c..3f60c99865 100644
--- a/target/m68k/cpu.c
+++ b/target/m68k/cpu.c
@@ -285,7 +285,7 @@ static void m68k_cpu_class_init(ObjectClass *c, void *data)
cc->gdb_write_register = m68k_cpu_gdb_write_register;
cc->tcg_ops.tlb_fill = m68k_cpu_tlb_fill;
#if defined(CONFIG_SOFTMMU)
- cc->do_transaction_failed = m68k_cpu_transaction_failed;
+ cc->tcg_ops.do_transaction_failed = m68k_cpu_transaction_failed;
cc->get_phys_page_debug = m68k_cpu_get_phys_page_debug;
#endif
cc->disas_set_info = m68k_cpu_disas_set_info;
diff --git a/target/microblaze/cpu.c b/target/microblaze/cpu.c
index 4f95248b2e..fa57a324dc 100644
--- a/target/microblaze/cpu.c
+++ b/target/microblaze/cpu.c
@@ -327,7 +327,7 @@ static void mb_cpu_class_init(ObjectClass *oc, void *data)
cc->gdb_write_register = mb_cpu_gdb_write_register;
cc->tcg_ops.tlb_fill = mb_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
- cc->do_transaction_failed = mb_cpu_transaction_failed;
+ cc->tcg_ops.do_transaction_failed = mb_cpu_transaction_failed;
cc->get_phys_page_debug = mb_cpu_get_phys_page_debug;
dc->vmsd = &vmstate_mb_cpu;
#endif
diff --git a/target/mips/cpu.c b/target/mips/cpu.c
index dd5fb78e6c..259bb791f7 100644
--- a/target/mips/cpu.c
+++ b/target/mips/cpu.c
@@ -239,7 +239,6 @@ static void mips_cpu_class_init(ObjectClass *c, void *data)
cc->gdb_read_register = mips_cpu_gdb_read_register;
cc->gdb_write_register = mips_cpu_gdb_write_register;
#ifndef CONFIG_USER_ONLY
- cc->do_transaction_failed = mips_cpu_do_transaction_failed;
cc->do_unaligned_access = mips_cpu_do_unaligned_access;
cc->get_phys_page_debug = mips_cpu_get_phys_page_debug;
cc->vmsd = &vmstate_mips_cpu;
@@ -252,6 +251,7 @@ static void mips_cpu_class_init(ObjectClass *c, void *data)
cc->tcg_ops.tlb_fill = mips_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->tcg_ops.do_interrupt = mips_cpu_do_interrupt;
+ cc->tcg_ops.do_transaction_failed = mips_cpu_do_transaction_failed;
#endif /* CONFIG_USER_ONLY */
#endif /* CONFIG_TCG */
diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index 6a815569cc..7dfd8d7339 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -555,7 +555,7 @@ static void riscv_cpu_class_init(ObjectClass *c, void *data)
cc->gdb_stop_before_watchpoint = true;
cc->disas_set_info = riscv_cpu_disas_set_info;
#ifndef CONFIG_USER_ONLY
- cc->do_transaction_failed = riscv_cpu_do_transaction_failed;
+ cc->tcg_ops.do_transaction_failed = riscv_cpu_do_transaction_failed;
cc->do_unaligned_access = riscv_cpu_do_unaligned_access;
cc->get_phys_page_debug = riscv_cpu_get_phys_page_debug;
/* For now, mark unmigratable: */
diff --git a/target/sparc/cpu.c b/target/sparc/cpu.c
index 3f23aa5962..009d0f07c3 100644
--- a/target/sparc/cpu.c
+++ b/target/sparc/cpu.c
@@ -873,7 +873,7 @@ static void sparc_cpu_class_init(ObjectClass *oc, void *data)
cc->gdb_write_register = sparc_cpu_gdb_write_register;
cc->tcg_ops.tlb_fill = sparc_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
- cc->do_transaction_failed = sparc_cpu_do_transaction_failed;
+ cc->tcg_ops.do_transaction_failed = sparc_cpu_do_transaction_failed;
cc->do_unaligned_access = sparc_cpu_do_unaligned_access;
cc->get_phys_page_debug = sparc_cpu_get_phys_page_debug;
cc->vmsd = &vmstate_sparc_cpu;
diff --git a/target/xtensa/cpu.c b/target/xtensa/cpu.c
index 3ff025f0fe..fc52fde696 100644
--- a/target/xtensa/cpu.c
+++ b/target/xtensa/cpu.c
@@ -205,7 +205,7 @@ static void xtensa_cpu_class_init(ObjectClass *oc, void *data)
#ifndef CONFIG_USER_ONLY
cc->do_unaligned_access = xtensa_cpu_do_unaligned_access;
cc->get_phys_page_debug = xtensa_cpu_get_phys_page_debug;
- cc->do_transaction_failed = xtensa_cpu_do_transaction_failed;
+ cc->tcg_ops.do_transaction_failed = xtensa_cpu_do_transaction_failed;
#endif
cc->tcg_ops.debug_excp_handler = xtensa_breakpoint_handler;
cc->disas_set_info = xtensa_cpu_disas_set_info;
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 25/32] cpu: move do_unaligned_access to tcg_ops
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (23 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 24/32] cpu: move cc->transaction_failed " Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 12:47 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 26/32] accel: extend AccelState and AccelClass to user-mode Claudio Fontana
` (9 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
make it consistently SOFTMMU-only.
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
include/hw/core/cpu.h | 17 +++--------------
include/hw/core/tcg-cpu-ops.h | 7 +++++++
target/alpha/cpu.c | 2 +-
target/arm/cpu.c | 2 +-
target/hppa/cpu.c | 4 +++-
target/microblaze/cpu.c | 2 +-
target/mips/cpu.c | 3 ++-
target/nios2/cpu.c | 2 +-
target/ppc/translate_init.c.inc | 2 +-
target/riscv/cpu.c | 2 +-
target/s390x/cpu.c | 2 +-
target/sh4/cpu.c | 2 +-
target/sparc/cpu.c | 2 +-
target/xtensa/cpu.c | 2 +-
14 files changed, 25 insertions(+), 26 deletions(-)
diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
index 0c6c17fb48..db54223983 100644
--- a/include/hw/core/cpu.h
+++ b/include/hw/core/cpu.h
@@ -87,8 +87,6 @@ struct TranslationBlock;
* @parse_features: Callback to parse command line arguments.
* @reset_dump_flags: #CPUDumpFlags to use for reset logging.
* @has_work: Callback for checking if there is work to do.
- * @do_unaligned_access: Callback for unaligned access handling, if
- * the target defines #TARGET_ALIGNED_ONLY.
* @virtio_is_big_endian: Callback to return %true if a CPU which supports
* runtime configurable endianness is currently big-endian. Non-configurable
* CPUs can use the default implementation of this method. This method should
@@ -154,9 +152,6 @@ struct CPUClass {
int reset_dump_flags;
bool (*has_work)(CPUState *cpu);
- void (*do_unaligned_access)(CPUState *cpu, vaddr addr,
- MMUAccessType access_type,
- int mmu_idx, uintptr_t retaddr);
bool (*virtio_is_big_endian)(CPUState *cpu);
int (*memory_rw_debug)(CPUState *cpu, vaddr addr,
uint8_t *buf, int len, bool is_write);
@@ -831,18 +826,15 @@ CPUState *cpu_by_arch_id(int64_t id);
void cpu_interrupt(CPUState *cpu, int mask);
-#ifdef NEED_CPU_H
-
-#ifdef CONFIG_SOFTMMU
+#if !defined(CONFIG_USER_ONLY) && defined(CONFIG_TCG)
static inline void cpu_unaligned_access(CPUState *cpu, vaddr addr,
MMUAccessType access_type,
int mmu_idx, uintptr_t retaddr)
{
CPUClass *cc = CPU_GET_CLASS(cpu);
- cc->do_unaligned_access(cpu, addr, access_type, mmu_idx, retaddr);
+ cc->tcg_ops.do_unaligned_access(cpu, addr, access_type, mmu_idx, retaddr);
}
-#ifdef CONFIG_TCG
static inline void cpu_transaction_failed(CPUState *cpu, hwaddr physaddr,
vaddr addr, unsigned size,
MMUAccessType access_type,
@@ -858,10 +850,7 @@ static inline void cpu_transaction_failed(CPUState *cpu, hwaddr physaddr,
mmu_idx, attrs, response, retaddr);
}
}
-#endif /* CONFIG_TCG */
-#endif /* CONFIG_SOFTMMU */
-
-#endif /* NEED_CPU_H */
+#endif /* !CONFIG_USER_ONLY && CONFIG_TCG */
/**
* cpu_set_pc:
diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
index 0cc0927738..b53b6122ed 100644
--- a/include/hw/core/tcg-cpu-ops.h
+++ b/include/hw/core/tcg-cpu-ops.h
@@ -48,6 +48,13 @@ typedef struct TcgCpuOperations {
unsigned size, MMUAccessType access_type,
int mmu_idx, MemTxAttrs attrs,
MemTxResult response, uintptr_t retaddr);
+ /**
+ * @do_unaligned_access: Callback for unaligned access handling, if
+ * the target defines #TARGET_ALIGNED_ONLY.
+ */
+ void (*do_unaligned_access)(CPUState *cpu, vaddr addr,
+ MMUAccessType access_type,
+ int mmu_idx, uintptr_t retaddr);
/**
* @tlb_fill: Handle a softmmu tlb miss or user-only address fault
diff --git a/target/alpha/cpu.c b/target/alpha/cpu.c
index 3e651b246f..0b5f5a78e3 100644
--- a/target/alpha/cpu.c
+++ b/target/alpha/cpu.c
@@ -225,7 +225,7 @@ static void alpha_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.tlb_fill = alpha_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->tcg_ops.do_transaction_failed = alpha_cpu_do_transaction_failed;
- cc->do_unaligned_access = alpha_cpu_do_unaligned_access;
+ cc->tcg_ops.do_unaligned_access = alpha_cpu_do_unaligned_access;
cc->get_phys_page_debug = alpha_cpu_get_phys_page_debug;
dc->vmsd = &vmstate_alpha_cpu;
#endif
diff --git a/target/arm/cpu.c b/target/arm/cpu.c
index 5e9202ed86..9a8c1dc2e1 100644
--- a/target/arm/cpu.c
+++ b/target/arm/cpu.c
@@ -2266,9 +2266,9 @@ static void arm_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.tlb_fill = arm_cpu_tlb_fill;
cc->tcg_ops.debug_excp_handler = arm_debug_excp_handler;
cc->debug_check_watchpoint = arm_debug_check_watchpoint;
- cc->do_unaligned_access = arm_cpu_do_unaligned_access;
#if !defined(CONFIG_USER_ONLY)
cc->tcg_ops.do_transaction_failed = arm_cpu_do_transaction_failed;
+ cc->tcg_ops.do_unaligned_access = arm_cpu_do_unaligned_access;
cc->adjust_watchpoint_address = arm_adjust_watchpoint_address;
cc->tcg_ops.do_interrupt = arm_cpu_do_interrupt;
#endif /* CONFIG_TCG && !CONFIG_USER_ONLY */
diff --git a/target/hppa/cpu.c b/target/hppa/cpu.c
index cce6ae6213..0985b3661f 100644
--- a/target/hppa/cpu.c
+++ b/target/hppa/cpu.c
@@ -70,6 +70,7 @@ static void hppa_cpu_disas_set_info(CPUState *cs, disassemble_info *info)
info->print_insn = print_insn_hppa;
}
+#ifndef CONFIG_USER_ONLY
static void hppa_cpu_do_unaligned_access(CPUState *cs, vaddr addr,
MMUAccessType access_type,
int mmu_idx, uintptr_t retaddr)
@@ -86,6 +87,7 @@ static void hppa_cpu_do_unaligned_access(CPUState *cs, vaddr addr,
cpu_loop_exit_restore(cs, retaddr);
}
+#endif /* CONFIG_USER_ONLY */
static void hppa_cpu_realizefn(DeviceState *dev, Error **errp)
{
@@ -149,9 +151,9 @@ static void hppa_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.tlb_fill = hppa_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->get_phys_page_debug = hppa_cpu_get_phys_page_debug;
+ cc->tcg_ops.do_unaligned_access = hppa_cpu_do_unaligned_access;
dc->vmsd = &vmstate_hppa_cpu;
#endif
- cc->do_unaligned_access = hppa_cpu_do_unaligned_access;
cc->disas_set_info = hppa_cpu_disas_set_info;
cc->tcg_ops.initialize = hppa_translate_init;
diff --git a/target/microblaze/cpu.c b/target/microblaze/cpu.c
index fa57a324dc..395f4a300f 100644
--- a/target/microblaze/cpu.c
+++ b/target/microblaze/cpu.c
@@ -318,7 +318,6 @@ static void mb_cpu_class_init(ObjectClass *oc, void *data)
cc->class_by_name = mb_cpu_class_by_name;
cc->has_work = mb_cpu_has_work;
cc->tcg_ops.do_interrupt = mb_cpu_do_interrupt;
- cc->do_unaligned_access = mb_cpu_do_unaligned_access;
cc->tcg_ops.cpu_exec_interrupt = mb_cpu_exec_interrupt;
cc->dump_state = mb_cpu_dump_state;
cc->set_pc = mb_cpu_set_pc;
@@ -328,6 +327,7 @@ static void mb_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.tlb_fill = mb_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->tcg_ops.do_transaction_failed = mb_cpu_transaction_failed;
+ cc->tcg_ops.do_unaligned_access = mb_cpu_do_unaligned_access;
cc->get_phys_page_debug = mb_cpu_get_phys_page_debug;
dc->vmsd = &vmstate_mb_cpu;
#endif
diff --git a/target/mips/cpu.c b/target/mips/cpu.c
index 259bb791f7..236d0d707b 100644
--- a/target/mips/cpu.c
+++ b/target/mips/cpu.c
@@ -239,7 +239,6 @@ static void mips_cpu_class_init(ObjectClass *c, void *data)
cc->gdb_read_register = mips_cpu_gdb_read_register;
cc->gdb_write_register = mips_cpu_gdb_write_register;
#ifndef CONFIG_USER_ONLY
- cc->do_unaligned_access = mips_cpu_do_unaligned_access;
cc->get_phys_page_debug = mips_cpu_get_phys_page_debug;
cc->vmsd = &vmstate_mips_cpu;
#endif
@@ -252,6 +251,8 @@ static void mips_cpu_class_init(ObjectClass *c, void *data)
#ifndef CONFIG_USER_ONLY
cc->tcg_ops.do_interrupt = mips_cpu_do_interrupt;
cc->tcg_ops.do_transaction_failed = mips_cpu_do_transaction_failed;
+ cc->tcg_ops.do_unaligned_access = mips_cpu_do_unaligned_access;
+
#endif /* CONFIG_USER_ONLY */
#endif /* CONFIG_TCG */
diff --git a/target/nios2/cpu.c b/target/nios2/cpu.c
index 2b959f0e49..059eea8c94 100644
--- a/target/nios2/cpu.c
+++ b/target/nios2/cpu.c
@@ -199,7 +199,7 @@ static void nios2_cpu_class_init(ObjectClass *oc, void *data)
cc->disas_set_info = nios2_cpu_disas_set_info;
cc->tcg_ops.tlb_fill = nios2_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
- cc->do_unaligned_access = nios2_cpu_do_unaligned_access;
+ cc->tcg_ops.do_unaligned_access = nios2_cpu_do_unaligned_access;
cc->get_phys_page_debug = nios2_cpu_get_phys_page_debug;
#endif
cc->gdb_read_register = nios2_cpu_gdb_read_register;
diff --git a/target/ppc/translate_init.c.inc b/target/ppc/translate_init.c.inc
index 0471da0d08..18ce389947 100644
--- a/target/ppc/translate_init.c.inc
+++ b/target/ppc/translate_init.c.inc
@@ -10913,7 +10913,6 @@ static void ppc_cpu_class_init(ObjectClass *oc, void *data)
cc->set_pc = ppc_cpu_set_pc;
cc->gdb_read_register = ppc_cpu_gdb_read_register;
cc->gdb_write_register = ppc_cpu_gdb_write_register;
- cc->do_unaligned_access = ppc_cpu_do_unaligned_access;
#ifndef CONFIG_USER_ONLY
cc->get_phys_page_debug = ppc_cpu_get_phys_page_debug;
cc->vmsd = &vmstate_ppc_cpu;
@@ -10950,6 +10949,7 @@ static void ppc_cpu_class_init(ObjectClass *oc, void *data)
#ifndef CONFIG_USER_ONLY
cc->tcg_ops.cpu_exec_enter = ppc_cpu_exec_enter;
cc->tcg_ops.cpu_exec_exit = ppc_cpu_exec_exit;
+ cc->tcg_ops.do_unaligned_access = ppc_cpu_do_unaligned_access;
#endif /* !CONFIG_USER_ONLY */
#endif /* CONFIG_TCG */
diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index 7dfd8d7339..e5626862c2 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -556,7 +556,7 @@ static void riscv_cpu_class_init(ObjectClass *c, void *data)
cc->disas_set_info = riscv_cpu_disas_set_info;
#ifndef CONFIG_USER_ONLY
cc->tcg_ops.do_transaction_failed = riscv_cpu_do_transaction_failed;
- cc->do_unaligned_access = riscv_cpu_do_unaligned_access;
+ cc->tcg_ops.do_unaligned_access = riscv_cpu_do_unaligned_access;
cc->get_phys_page_debug = riscv_cpu_get_phys_page_debug;
/* For now, mark unmigratable: */
cc->vmsd = &vmstate_riscv_cpu;
diff --git a/target/s390x/cpu.c b/target/s390x/cpu.c
index b838bd61a4..86f654fd6b 100644
--- a/target/s390x/cpu.c
+++ b/target/s390x/cpu.c
@@ -507,7 +507,7 @@ static void s390_cpu_class_init(ObjectClass *oc, void *data)
#ifdef CONFIG_TCG
cc->tcg_ops.cpu_exec_interrupt = s390_cpu_exec_interrupt;
cc->tcg_ops.debug_excp_handler = s390x_cpu_debug_excp_handler;
- cc->do_unaligned_access = s390x_cpu_do_unaligned_access;
+ cc->tcg_ops.do_unaligned_access = s390x_cpu_do_unaligned_access;
#endif
#endif
cc->disas_set_info = s390_cpu_disas_set_info;
diff --git a/target/sh4/cpu.c b/target/sh4/cpu.c
index ff835d4bc1..fbd5f42675 100644
--- a/target/sh4/cpu.c
+++ b/target/sh4/cpu.c
@@ -227,7 +227,7 @@ static void superh_cpu_class_init(ObjectClass *oc, void *data)
cc->gdb_write_register = superh_cpu_gdb_write_register;
cc->tcg_ops.tlb_fill = superh_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
- cc->do_unaligned_access = superh_cpu_do_unaligned_access;
+ cc->tcg_ops.do_unaligned_access = superh_cpu_do_unaligned_access;
cc->get_phys_page_debug = superh_cpu_get_phys_page_debug;
#endif
cc->disas_set_info = superh_cpu_disas_set_info;
diff --git a/target/sparc/cpu.c b/target/sparc/cpu.c
index 009d0f07c3..3b53ef2390 100644
--- a/target/sparc/cpu.c
+++ b/target/sparc/cpu.c
@@ -874,7 +874,7 @@ static void sparc_cpu_class_init(ObjectClass *oc, void *data)
cc->tcg_ops.tlb_fill = sparc_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
cc->tcg_ops.do_transaction_failed = sparc_cpu_do_transaction_failed;
- cc->do_unaligned_access = sparc_cpu_do_unaligned_access;
+ cc->tcg_ops.do_unaligned_access = sparc_cpu_do_unaligned_access;
cc->get_phys_page_debug = sparc_cpu_get_phys_page_debug;
cc->vmsd = &vmstate_sparc_cpu;
#endif
diff --git a/target/xtensa/cpu.c b/target/xtensa/cpu.c
index fc52fde696..4b6381569f 100644
--- a/target/xtensa/cpu.c
+++ b/target/xtensa/cpu.c
@@ -203,7 +203,7 @@ static void xtensa_cpu_class_init(ObjectClass *oc, void *data)
cc->gdb_stop_before_watchpoint = true;
cc->tcg_ops.tlb_fill = xtensa_cpu_tlb_fill;
#ifndef CONFIG_USER_ONLY
- cc->do_unaligned_access = xtensa_cpu_do_unaligned_access;
+ cc->tcg_ops.do_unaligned_access = xtensa_cpu_do_unaligned_access;
cc->get_phys_page_debug = xtensa_cpu_get_phys_page_debug;
cc->tcg_ops.do_transaction_failed = xtensa_cpu_do_transaction_failed;
#endif
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 26/32] accel: extend AccelState and AccelClass to user-mode
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (24 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 25/32] cpu: move do_unaligned_access " Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 12:51 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 27/32] accel: replace struct CpusAccel with AccelOpsClass Claudio Fontana
` (8 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
MAINTAINERS | 2 +-
accel/accel-common.c | 50 ++++++++++++++++++++++++++++++
accel/{accel.c => accel-softmmu.c} | 27 ++--------------
accel/accel-user.c | 24 ++++++++++++++
accel/meson.build | 4 ++-
accel/qtest/qtest.c | 2 +-
accel/tcg/meson.build | 2 +-
accel/tcg/tcg-all.c | 13 ++++++--
accel/xen/xen-all.c | 2 +-
bsd-user/main.c | 6 +++-
include/hw/boards.h | 2 +-
include/{sysemu => qemu}/accel.h | 14 +++++----
include/sysemu/hvf.h | 2 +-
include/sysemu/kvm.h | 2 +-
include/sysemu/kvm_int.h | 2 +-
linux-user/main.c | 6 +++-
softmmu/memory.c | 2 +-
softmmu/qtest.c | 2 +-
softmmu/vl.c | 2 +-
target/i386/hax/hax-all.c | 2 +-
target/i386/hvf/hvf-i386.h | 2 +-
target/i386/hvf/hvf.c | 2 +-
target/i386/hvf/x86_task.c | 2 +-
target/i386/whpx/whpx-all.c | 2 +-
24 files changed, 124 insertions(+), 52 deletions(-)
create mode 100644 accel/accel-common.c
rename accel/{accel.c => accel-softmmu.c} (75%)
create mode 100644 accel/accel-user.c
rename include/{sysemu => qemu}/accel.h (95%)
diff --git a/MAINTAINERS b/MAINTAINERS
index d876f504a6..6235dd3a9f 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -434,7 +434,7 @@ Overall
M: Richard Henderson <richard.henderson@linaro.org>
R: Paolo Bonzini <pbonzini@redhat.com>
S: Maintained
-F: include/sysemu/accel.h
+F: include/qemu/accel.h
F: accel/accel.c
F: accel/Makefile.objs
F: accel/stubs/Makefile.objs
diff --git a/accel/accel-common.c b/accel/accel-common.c
new file mode 100644
index 0000000000..ddec8cb5ae
--- /dev/null
+++ b/accel/accel-common.c
@@ -0,0 +1,50 @@
+/*
+ * QEMU accel class, components common to system emulation and user mode
+ *
+ * Copyright (c) 2003-2008 Fabrice Bellard
+ * Copyright (c) 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include "qemu/osdep.h"
+#include "qemu/accel.h"
+
+static const TypeInfo accel_type = {
+ .name = TYPE_ACCEL,
+ .parent = TYPE_OBJECT,
+ .class_size = sizeof(AccelClass),
+ .instance_size = sizeof(AccelState),
+};
+
+/* Lookup AccelClass from opt_name. Returns NULL if not found */
+AccelClass *accel_find(const char *opt_name)
+{
+ char *class_name = g_strdup_printf(ACCEL_CLASS_NAME("%s"), opt_name);
+ AccelClass *ac = ACCEL_CLASS(object_class_by_name(class_name));
+ g_free(class_name);
+ return ac;
+}
+
+static void register_accel_types(void)
+{
+ type_register_static(&accel_type);
+}
+
+type_init(register_accel_types);
diff --git a/accel/accel.c b/accel/accel-softmmu.c
similarity index 75%
rename from accel/accel.c
rename to accel/accel-softmmu.c
index cb555e3b06..f89da8f9d1 100644
--- a/accel/accel.c
+++ b/accel/accel-softmmu.c
@@ -1,5 +1,5 @@
/*
- * QEMU System Emulator, accelerator interfaces
+ * QEMU accel class, system emulation components
*
* Copyright (c) 2003-2008 Fabrice Bellard
* Copyright (c) 2014 Red Hat Inc.
@@ -24,28 +24,12 @@
*/
#include "qemu/osdep.h"
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "hw/boards.h"
#include "sysemu/arch_init.h"
#include "sysemu/sysemu.h"
#include "qom/object.h"
-static const TypeInfo accel_type = {
- .name = TYPE_ACCEL,
- .parent = TYPE_OBJECT,
- .class_size = sizeof(AccelClass),
- .instance_size = sizeof(AccelState),
-};
-
-/* Lookup AccelClass from opt_name. Returns NULL if not found */
-AccelClass *accel_find(const char *opt_name)
-{
- char *class_name = g_strdup_printf(ACCEL_CLASS_NAME("%s"), opt_name);
- AccelClass *ac = ACCEL_CLASS(object_class_by_name(class_name));
- g_free(class_name);
- return ac;
-}
-
int accel_init_machine(AccelState *accel, MachineState *ms)
{
AccelClass *acc = ACCEL_GET_CLASS(accel);
@@ -76,10 +60,3 @@ void accel_setup_post(MachineState *ms)
acc->setup_post(ms, accel);
}
}
-
-static void register_accel_types(void)
-{
- type_register_static(&accel_type);
-}
-
-type_init(register_accel_types);
diff --git a/accel/accel-user.c b/accel/accel-user.c
new file mode 100644
index 0000000000..26bdda6236
--- /dev/null
+++ b/accel/accel-user.c
@@ -0,0 +1,24 @@
+/*
+ * QEMU accel class, user-mode components
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#include "qemu/osdep.h"
+#include "qemu/accel.h"
+
+AccelState *current_accel(void)
+{
+ static AccelState *accel;
+
+ if (!accel) {
+ AccelClass *ac = accel_find("tcg");
+
+ g_assert(ac != NULL);
+ accel = ACCEL(object_new_with_class(OBJECT_CLASS(ac)));
+ }
+ return accel;
+}
diff --git a/accel/meson.build b/accel/meson.build
index b26cca227a..b44ba30c86 100644
--- a/accel/meson.build
+++ b/accel/meson.build
@@ -1,4 +1,6 @@
-softmmu_ss.add(files('accel.c'))
+specific_ss.add(files('accel-common.c'))
+softmmu_ss.add(files('accel-softmmu.c'))
+user_ss.add(files('accel-user.c'))
subdir('qtest')
subdir('kvm')
diff --git a/accel/qtest/qtest.c b/accel/qtest/qtest.c
index b282cea5cf..b4e731cb2b 100644
--- a/accel/qtest/qtest.c
+++ b/accel/qtest/qtest.c
@@ -17,7 +17,7 @@
#include "qemu/module.h"
#include "qemu/option.h"
#include "qemu/config-file.h"
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "sysemu/qtest.h"
#include "sysemu/cpus.h"
#include "sysemu/cpu-timers.h"
diff --git a/accel/tcg/meson.build b/accel/tcg/meson.build
index f39aab0a0c..424d9bb1fc 100644
--- a/accel/tcg/meson.build
+++ b/accel/tcg/meson.build
@@ -1,5 +1,6 @@
tcg_ss = ss.source_set()
tcg_ss.add(files(
+ 'tcg-all.c',
'cpu-exec-common.c',
'cpu-exec.c',
'tcg-runtime-gvec.c',
@@ -13,7 +14,6 @@ tcg_ss.add(when: 'CONFIG_PLUGIN', if_true: [files('plugin-gen.c'), libdl])
specific_ss.add_all(when: 'CONFIG_TCG', if_true: tcg_ss)
specific_ss.add(when: ['CONFIG_SOFTMMU', 'CONFIG_TCG'], if_true: files(
- 'tcg-all.c',
'cputlb.c',
'tcg-cpus.c',
'tcg-cpus-mttcg.c',
diff --git a/accel/tcg/tcg-all.c b/accel/tcg/tcg-all.c
index 1ac0b76515..7125d0cc29 100644
--- a/accel/tcg/tcg-all.c
+++ b/accel/tcg/tcg-all.c
@@ -30,9 +30,12 @@
#include "tcg/tcg.h"
#include "qapi/error.h"
#include "qemu/error-report.h"
-#include "hw/boards.h"
+#include "qemu/accel.h"
#include "qapi/qapi-builtin-visit.h"
+
+#ifndef CONFIG_USER_ONLY
#include "tcg-cpus.h"
+#endif /* CONFIG_USER_ONLY */
struct TCGState {
AccelState parent_obj;
@@ -106,8 +109,12 @@ static int tcg_init(MachineState *ms)
mttcg_enabled = s->mttcg_enabled;
/*
- * Initialize TCG regions
+ * Initialize TCG regions only for softmmu.
+ *
+ * This needs to be done later for user mode, because the prologue
+ * generation needs to be delayed so that GUEST_BASE is already set.
*/
+#ifndef CONFIG_USER_ONLY
tcg_region_init();
if (mttcg_enabled) {
@@ -117,6 +124,8 @@ static int tcg_init(MachineState *ms)
} else {
cpus_register_accel(&tcg_cpus_rr);
}
+#endif /* !CONFIG_USER_ONLY */
+
return 0;
}
diff --git a/accel/xen/xen-all.c b/accel/xen/xen-all.c
index 878a4089d9..594aaf6b49 100644
--- a/accel/xen/xen-all.c
+++ b/accel/xen/xen-all.c
@@ -15,7 +15,7 @@
#include "hw/xen/xen-legacy-backend.h"
#include "hw/xen/xen_pt.h"
#include "chardev/char.h"
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "sysemu/cpus.h"
#include "sysemu/xen.h"
#include "sysemu/runstate.h"
diff --git a/bsd-user/main.c b/bsd-user/main.c
index 0a918e8f74..ff295bcb29 100644
--- a/bsd-user/main.c
+++ b/bsd-user/main.c
@@ -20,6 +20,7 @@
#include "qemu/osdep.h"
#include "qemu-common.h"
#include "qemu/units.h"
+#include "qemu/accel.h"
#include "sysemu/tcg.h"
#include "qemu-version.h"
#include <machine/trap.h>
@@ -908,8 +909,11 @@ int main(int argc, char **argv)
}
/* init tcg before creating CPUs and to get qemu_host_page_size */
- tcg_exec_init(0);
+ {
+ AccelClass *ac = ACCEL_GET_CLASS(current_accel());
+ ac->init_machine(NULL);
+ }
cpu_type = parse_cpu_option(cpu_model);
cpu = cpu_create(cpu_type);
env = cpu->env_ptr;
diff --git a/include/hw/boards.h b/include/hw/boards.h
index a49e3a6b44..b754504afe 100644
--- a/include/hw/boards.h
+++ b/include/hw/boards.h
@@ -6,7 +6,7 @@
#include "exec/memory.h"
#include "sysemu/hostmem.h"
#include "sysemu/blockdev.h"
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "qapi/qapi-types-machine.h"
#include "qemu/module.h"
#include "qom/object.h"
diff --git a/include/sysemu/accel.h b/include/qemu/accel.h
similarity index 95%
rename from include/sysemu/accel.h
rename to include/qemu/accel.h
index e08b8ab8fa..fac4a18703 100644
--- a/include/sysemu/accel.h
+++ b/include/qemu/accel.h
@@ -20,8 +20,8 @@
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
* THE SOFTWARE.
*/
-#ifndef HW_ACCEL_H
-#define HW_ACCEL_H
+#ifndef QEMU_ACCEL_H
+#define QEMU_ACCEL_H
#include "qom/object.h"
#include "exec/hwaddr.h"
@@ -37,8 +37,8 @@ typedef struct AccelClass {
/*< public >*/
const char *name;
-#ifndef CONFIG_USER_ONLY
int (*init_machine)(MachineState *ms);
+#ifndef CONFIG_USER_ONLY
void (*setup_post)(MachineState *ms, AccelState *accel);
bool (*has_memory)(MachineState *ms, AddressSpace *as,
hwaddr start_addr, hwaddr size);
@@ -67,11 +67,13 @@ typedef struct AccelClass {
OBJECT_GET_CLASS(AccelClass, (obj), TYPE_ACCEL)
AccelClass *accel_find(const char *opt_name);
+AccelState *current_accel(void);
+
+#ifndef CONFIG_USER_ONLY
int accel_init_machine(AccelState *accel, MachineState *ms);
/* Called just before os_setup_post (ie just before drop OS privs) */
void accel_setup_post(MachineState *ms);
+#endif /* !CONFIG_USER_ONLY */
-AccelState *current_accel(void);
-
-#endif
+#endif /* QEMU_ACCEL_H */
diff --git a/include/sysemu/hvf.h b/include/sysemu/hvf.h
index f893768df9..c98636bc81 100644
--- a/include/sysemu/hvf.h
+++ b/include/sysemu/hvf.h
@@ -13,7 +13,7 @@
#ifndef HVF_H
#define HVF_H
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "qom/object.h"
#ifdef CONFIG_HVF
diff --git a/include/sysemu/kvm.h b/include/sysemu/kvm.h
index bb5d5cf497..739682f3c3 100644
--- a/include/sysemu/kvm.h
+++ b/include/sysemu/kvm.h
@@ -17,7 +17,7 @@
#include "qemu/queue.h"
#include "hw/core/cpu.h"
#include "exec/memattrs.h"
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "qom/object.h"
#ifdef NEED_CPU_H
diff --git a/include/sysemu/kvm_int.h b/include/sysemu/kvm_int.h
index 65740806da..ccb8869f01 100644
--- a/include/sysemu/kvm_int.h
+++ b/include/sysemu/kvm_int.h
@@ -10,7 +10,7 @@
#define QEMU_KVM_INT_H
#include "exec/memory.h"
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "sysemu/kvm.h"
typedef struct KVMSlot
diff --git a/linux-user/main.c b/linux-user/main.c
index 24d1eb73ad..5c059a8445 100644
--- a/linux-user/main.c
+++ b/linux-user/main.c
@@ -20,6 +20,7 @@
#include "qemu/osdep.h"
#include "qemu-common.h"
#include "qemu/units.h"
+#include "qemu/accel.h"
#include "sysemu/tcg.h"
#include "qemu-version.h"
#include <sys/syscall.h>
@@ -703,8 +704,11 @@ int main(int argc, char **argv, char **envp)
cpu_type = parse_cpu_option(cpu_model);
/* init tcg before creating CPUs and to get qemu_host_page_size */
- tcg_exec_init(0);
+ {
+ AccelClass *ac = ACCEL_GET_CLASS(current_accel());
+ ac->init_machine(NULL);
+ }
cpu = cpu_create(cpu_type);
env = cpu->env_ptr;
cpu_reset(cpu);
diff --git a/softmmu/memory.c b/softmmu/memory.c
index 11ca94d037..9484ae9503 100644
--- a/softmmu/memory.c
+++ b/softmmu/memory.c
@@ -32,7 +32,7 @@
#include "sysemu/kvm.h"
#include "sysemu/runstate.h"
#include "sysemu/tcg.h"
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "hw/boards.h"
#include "migration/vmstate.h"
diff --git a/softmmu/qtest.c b/softmmu/qtest.c
index 7965dc9a16..130c366615 100644
--- a/softmmu/qtest.c
+++ b/softmmu/qtest.c
@@ -20,7 +20,7 @@
#include "exec/ioport.h"
#include "exec/memory.h"
#include "hw/irq.h"
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "sysemu/cpu-timers.h"
#include "qemu/config-file.h"
#include "qemu/option.h"
diff --git a/softmmu/vl.c b/softmmu/vl.c
index e6e0ad5a92..bc20c526d2 100644
--- a/softmmu/vl.c
+++ b/softmmu/vl.c
@@ -40,7 +40,7 @@
#include "qemu/error-report.h"
#include "qemu/sockets.h"
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "hw/usb.h"
#include "hw/isa/isa.h"
#include "hw/scsi/scsi.h"
diff --git a/target/i386/hax/hax-all.c b/target/i386/hax/hax-all.c
index fecfe8cd6e..d7f4bb44a7 100644
--- a/target/i386/hax/hax-all.c
+++ b/target/i386/hax/hax-all.c
@@ -28,7 +28,7 @@
#include "exec/address-spaces.h"
#include "qemu-common.h"
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "sysemu/reset.h"
#include "sysemu/runstate.h"
#include "hw/boards.h"
diff --git a/target/i386/hvf/hvf-i386.h b/target/i386/hvf/hvf-i386.h
index e0edffd077..50b914fd67 100644
--- a/target/i386/hvf/hvf-i386.h
+++ b/target/i386/hvf/hvf-i386.h
@@ -16,7 +16,7 @@
#ifndef HVF_I386_H
#define HVF_I386_H
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "sysemu/hvf.h"
#include "cpu.h"
#include "x86.h"
diff --git a/target/i386/hvf/hvf.c b/target/i386/hvf/hvf.c
index ed9356565c..ffc9efa40f 100644
--- a/target/i386/hvf/hvf.c
+++ b/target/i386/hvf/hvf.c
@@ -69,7 +69,7 @@
#include "exec/address-spaces.h"
#include "hw/i386/apic_internal.h"
#include "qemu/main-loop.h"
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "target/i386/cpu.h"
#include "hvf-cpus.h"
diff --git a/target/i386/hvf/x86_task.c b/target/i386/hvf/x86_task.c
index 6f04478b3a..d66dfd7669 100644
--- a/target/i386/hvf/x86_task.c
+++ b/target/i386/hvf/x86_task.c
@@ -28,7 +28,7 @@
#include "hw/i386/apic_internal.h"
#include "qemu/main-loop.h"
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "target/i386/cpu.h"
// TODO: taskswitch handling
diff --git a/target/i386/whpx/whpx-all.c b/target/i386/whpx/whpx-all.c
index f4f3e33eac..ee6b606194 100644
--- a/target/i386/whpx/whpx-all.c
+++ b/target/i386/whpx/whpx-all.c
@@ -13,7 +13,7 @@
#include "exec/address-spaces.h"
#include "exec/ioport.h"
#include "qemu-common.h"
-#include "sysemu/accel.h"
+#include "qemu/accel.h"
#include "sysemu/whpx.h"
#include "sysemu/cpus.h"
#include "sysemu/runstate.h"
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 27/32] accel: replace struct CpusAccel with AccelOpsClass
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (25 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 26/32] accel: extend AccelState and AccelClass to user-mode Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-09 12:54 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 28/32] accel: introduce AccelCPUClass extending CPUClass Claudio Fontana
` (7 subsequent siblings)
34 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
centralize the registration of the cpus.c module
accelerator operations in accel/accel-softmmu.c
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
MAINTAINERS | 3 ++-
accel/accel-common.c | 11 +++++++++
accel/accel-softmmu.c | 43 +++++++++++++++++++++++++++++++---
accel/accel-softmmu.h | 15 ++++++++++++
accel/kvm/kvm-all.c | 2 --
accel/kvm/kvm-cpus.c | 26 ++++++++++++++++-----
accel/kvm/kvm-cpus.h | 2 --
accel/qtest/qtest.c | 23 +++++++++++++-----
accel/tcg/tcg-all.c | 12 ----------
accel/tcg/tcg-cpus-icount.c | 11 +--------
accel/tcg/tcg-cpus-icount.h | 2 ++
accel/tcg/tcg-cpus-mttcg.c | 12 +++-------
accel/tcg/tcg-cpus-mttcg.h | 19 +++++++++++++++
accel/tcg/tcg-cpus-rr.c | 7 ------
accel/tcg/tcg-cpus.c | 43 ++++++++++++++++++++++++++++++++++
accel/tcg/tcg-cpus.h | 4 ----
accel/xen/xen-all.c | 22 ++++++++++++------
bsd-user/main.c | 3 ++-
include/qemu/accel.h | 2 ++
include/sysemu/accel-ops.h | 45 ++++++++++++++++++++++++++++++++++++
include/sysemu/cpus.h | 26 ++++-----------------
linux-user/main.c | 1 +
softmmu/cpus.c | 12 +++++-----
softmmu/vl.c | 6 +++--
target/i386/hax/hax-all.c | 3 ---
target/i386/hax/hax-cpus.c | 29 +++++++++++++++++------
target/i386/hax/hax-cpus.h | 2 --
target/i386/hvf/hvf-cpus.c | 27 +++++++++++++++++-----
target/i386/hvf/hvf-cpus.h | 2 --
target/i386/hvf/hvf.c | 1 -
target/i386/whpx/whpx-all.c | 2 --
target/i386/whpx/whpx-cpus.c | 29 +++++++++++++++++------
target/i386/whpx/whpx-cpus.h | 2 --
33 files changed, 318 insertions(+), 131 deletions(-)
create mode 100644 accel/accel-softmmu.h
create mode 100644 accel/tcg/tcg-cpus-mttcg.h
create mode 100644 include/sysemu/accel-ops.h
diff --git a/MAINTAINERS b/MAINTAINERS
index 6235dd3a9f..8f0e773a47 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -435,7 +435,8 @@ M: Richard Henderson <richard.henderson@linaro.org>
R: Paolo Bonzini <pbonzini@redhat.com>
S: Maintained
F: include/qemu/accel.h
-F: accel/accel.c
+F: include/sysemu/accel-ops.h
+F: accel/accel-*.c
F: accel/Makefile.objs
F: accel/stubs/Makefile.objs
diff --git a/accel/accel-common.c b/accel/accel-common.c
index ddec8cb5ae..6b59873419 100644
--- a/accel/accel-common.c
+++ b/accel/accel-common.c
@@ -26,6 +26,10 @@
#include "qemu/osdep.h"
#include "qemu/accel.h"
+#ifndef CONFIG_USER_ONLY
+#include "accel-softmmu.h"
+#endif /* !CONFIG_USER_ONLY */
+
static const TypeInfo accel_type = {
.name = TYPE_ACCEL,
.parent = TYPE_OBJECT,
@@ -42,6 +46,13 @@ AccelClass *accel_find(const char *opt_name)
return ac;
}
+void accel_init_interfaces(AccelClass *ac)
+{
+#ifndef CONFIG_USER_ONLY
+ accel_init_ops_interfaces(ac);
+#endif /* !CONFIG_USER_ONLY */
+}
+
static void register_accel_types(void)
{
type_register_static(&accel_type);
diff --git a/accel/accel-softmmu.c b/accel/accel-softmmu.c
index f89da8f9d1..2d15d3f2f4 100644
--- a/accel/accel-softmmu.c
+++ b/accel/accel-softmmu.c
@@ -26,9 +26,9 @@
#include "qemu/osdep.h"
#include "qemu/accel.h"
#include "hw/boards.h"
-#include "sysemu/arch_init.h"
-#include "sysemu/sysemu.h"
-#include "qom/object.h"
+#include "sysemu/cpus.h"
+
+#include "accel-softmmu.h"
int accel_init_machine(AccelState *accel, MachineState *ms)
{
@@ -60,3 +60,40 @@ void accel_setup_post(MachineState *ms)
acc->setup_post(ms, accel);
}
}
+
+/* initialize the arch-independent accel operation interfaces */
+void accel_init_ops_interfaces(AccelClass *ac)
+{
+ const char *ac_name;
+ char *ops_name;
+ AccelOpsClass *ops;
+
+ ac_name = object_class_get_name(OBJECT_CLASS(ac));
+ g_assert(ac_name != NULL);
+
+ ops_name = g_strdup_printf("%s" ACCEL_OPS_SUFFIX, ac_name);
+ ops = ACCEL_OPS_CLASS(object_class_by_name(ops_name));
+ g_free(ops_name);
+
+ /*
+ * all accelerators need to define ops, providing at least a mandatory
+ * non-NULL create_vcpu_thread operation.
+ */
+ g_assert(ops != NULL);
+ if (ops->ops_init) {
+ ops->ops_init(ops);
+ }
+ cpus_register_accel(ops);
+}
+
+static const TypeInfo accel_ops_type_info = {
+ .name = TYPE_ACCEL_OPS,
+ .parent = TYPE_OBJECT,
+ .abstract = true,
+ .class_size = sizeof(AccelOpsClass),
+};
+static void accel_softmmu_register_types(void)
+{
+ type_register_static(&accel_ops_type_info);
+}
+type_init(accel_softmmu_register_types);
diff --git a/accel/accel-softmmu.h b/accel/accel-softmmu.h
new file mode 100644
index 0000000000..2877b5c234
--- /dev/null
+++ b/accel/accel-softmmu.h
@@ -0,0 +1,15 @@
+/*
+ * QEMU System Emulation accel internal functions
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#ifndef ACCEL_SOFTMMU_H
+#define ACCEL_SOFTMMU_H
+
+void accel_init_ops_interfaces(AccelClass *ac);
+
+#endif /* ACCEL_SOFTMMU_H */
diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c
index baaa54249d..18be3cd113 100644
--- a/accel/kvm/kvm-all.c
+++ b/accel/kvm/kvm-all.c
@@ -2253,8 +2253,6 @@ static int kvm_init(MachineState *ms)
ret = ram_block_discard_disable(true);
assert(!ret);
}
-
- cpus_register_accel(&kvm_cpus);
return 0;
err:
diff --git a/accel/kvm/kvm-cpus.c b/accel/kvm/kvm-cpus.c
index d809b1e74c..fc9dda46ae 100644
--- a/accel/kvm/kvm-cpus.c
+++ b/accel/kvm/kvm-cpus.c
@@ -74,11 +74,25 @@ static void kvm_start_vcpu_thread(CPUState *cpu)
cpu, QEMU_THREAD_JOINABLE);
}
-const CpusAccel kvm_cpus = {
- .create_vcpu_thread = kvm_start_vcpu_thread,
+static void kvm_cpus_class_init(ObjectClass *oc, void *data)
+{
+ AccelOpsClass *ops = ACCEL_OPS_CLASS(oc);
- .synchronize_post_reset = kvm_cpu_synchronize_post_reset,
- .synchronize_post_init = kvm_cpu_synchronize_post_init,
- .synchronize_state = kvm_cpu_synchronize_state,
- .synchronize_pre_loadvm = kvm_cpu_synchronize_pre_loadvm,
+ ops->create_vcpu_thread = kvm_start_vcpu_thread;
+ ops->synchronize_post_reset = kvm_cpu_synchronize_post_reset;
+ ops->synchronize_post_init = kvm_cpu_synchronize_post_init;
+ ops->synchronize_state = kvm_cpu_synchronize_state;
+ ops->synchronize_pre_loadvm = kvm_cpu_synchronize_pre_loadvm;
};
+static const TypeInfo kvm_cpus_type_info = {
+ .name = ACCEL_OPS_NAME("kvm"),
+
+ .parent = TYPE_ACCEL_OPS,
+ .class_init = kvm_cpus_class_init,
+ .abstract = true,
+};
+static void kvm_cpus_register_types(void)
+{
+ type_register_static(&kvm_cpus_type_info);
+}
+type_init(kvm_cpus_register_types);
diff --git a/accel/kvm/kvm-cpus.h b/accel/kvm/kvm-cpus.h
index 3df732b816..bf0bd1bee4 100644
--- a/accel/kvm/kvm-cpus.h
+++ b/accel/kvm/kvm-cpus.h
@@ -12,8 +12,6 @@
#include "sysemu/cpus.h"
-extern const CpusAccel kvm_cpus;
-
int kvm_init_vcpu(CPUState *cpu, Error **errp);
int kvm_cpu_exec(CPUState *cpu);
void kvm_destroy_vcpu(CPUState *cpu);
diff --git a/accel/qtest/qtest.c b/accel/qtest/qtest.c
index b4e731cb2b..68d86de30f 100644
--- a/accel/qtest/qtest.c
+++ b/accel/qtest/qtest.c
@@ -25,14 +25,8 @@
#include "qemu/main-loop.h"
#include "hw/core/cpu.h"
-const CpusAccel qtest_cpus = {
- .create_vcpu_thread = dummy_start_vcpu_thread,
- .get_virtual_clock = qtest_get_virtual_clock,
-};
-
static int qtest_init_accel(MachineState *ms)
{
- cpus_register_accel(&qtest_cpus);
return 0;
}
@@ -52,9 +46,26 @@ static const TypeInfo qtest_accel_type = {
.class_init = qtest_accel_class_init,
};
+static void qtest_cpus_class_init(ObjectClass *oc, void *data)
+{
+ AccelOpsClass *ops = ACCEL_OPS_CLASS(oc);
+
+ ops->create_vcpu_thread = dummy_start_vcpu_thread;
+ ops->get_virtual_clock = qtest_get_virtual_clock;
+};
+
+static const TypeInfo qtest_cpus_type_info = {
+ .name = ACCEL_OPS_NAME("qtest"),
+
+ .parent = TYPE_ACCEL_OPS,
+ .class_init = qtest_cpus_class_init,
+ .abstract = true,
+};
+
static void qtest_type_init(void)
{
type_register_static(&qtest_accel_type);
+ type_register_static(&qtest_cpus_type_info);
}
type_init(qtest_type_init);
diff --git a/accel/tcg/tcg-all.c b/accel/tcg/tcg-all.c
index 7125d0cc29..f0d97b4b21 100644
--- a/accel/tcg/tcg-all.c
+++ b/accel/tcg/tcg-all.c
@@ -33,10 +33,6 @@
#include "qemu/accel.h"
#include "qapi/qapi-builtin-visit.h"
-#ifndef CONFIG_USER_ONLY
-#include "tcg-cpus.h"
-#endif /* CONFIG_USER_ONLY */
-
struct TCGState {
AccelState parent_obj;
@@ -116,14 +112,6 @@ static int tcg_init(MachineState *ms)
*/
#ifndef CONFIG_USER_ONLY
tcg_region_init();
-
- if (mttcg_enabled) {
- cpus_register_accel(&tcg_cpus_mttcg);
- } else if (icount_enabled()) {
- cpus_register_accel(&tcg_cpus_icount);
- } else {
- cpus_register_accel(&tcg_cpus_rr);
- }
#endif /* !CONFIG_USER_ONLY */
return 0;
diff --git a/accel/tcg/tcg-cpus-icount.c b/accel/tcg/tcg-cpus-icount.c
index 9f45432275..5445b4d545 100644
--- a/accel/tcg/tcg-cpus-icount.c
+++ b/accel/tcg/tcg-cpus-icount.c
@@ -125,7 +125,7 @@ void icount_process_data(CPUState *cpu)
replay_mutex_unlock();
}
-static void icount_handle_interrupt(CPUState *cpu, int mask)
+void icount_handle_interrupt(CPUState *cpu, int mask)
{
int old_mask = cpu->interrupt_request;
@@ -136,12 +136,3 @@ static void icount_handle_interrupt(CPUState *cpu, int mask)
cpu_abort(cpu, "Raised interrupt while not in I/O function");
}
}
-
-const CpusAccel tcg_cpus_icount = {
- .create_vcpu_thread = rr_start_vcpu_thread,
- .kick_vcpu_thread = rr_kick_vcpu_thread,
-
- .handle_interrupt = icount_handle_interrupt,
- .get_virtual_clock = icount_get,
- .get_elapsed_ticks = icount_get,
-};
diff --git a/accel/tcg/tcg-cpus-icount.h b/accel/tcg/tcg-cpus-icount.h
index b695939dfa..d884aa2aaa 100644
--- a/accel/tcg/tcg-cpus-icount.h
+++ b/accel/tcg/tcg-cpus-icount.h
@@ -14,4 +14,6 @@ void icount_handle_deadline(void);
void icount_prepare_for_run(CPUState *cpu);
void icount_process_data(CPUState *cpu);
+void icount_handle_interrupt(CPUState *cpu, int mask);
+
#endif /* TCG_CPUS_ICOUNT_H */
diff --git a/accel/tcg/tcg-cpus-mttcg.c b/accel/tcg/tcg-cpus-mttcg.c
index 9c3767d260..dabf5ed42e 100644
--- a/accel/tcg/tcg-cpus-mttcg.c
+++ b/accel/tcg/tcg-cpus-mttcg.c
@@ -33,6 +33,7 @@
#include "hw/boards.h"
#include "tcg-cpus.h"
+#include "tcg-cpus-mttcg.h"
/*
* In the multi-threaded case each vCPU has its own thread. The TLS
@@ -103,12 +104,12 @@ static void *mttcg_cpu_thread_fn(void *arg)
return NULL;
}
-static void mttcg_kick_vcpu_thread(CPUState *cpu)
+void mttcg_kick_vcpu_thread(CPUState *cpu)
{
cpu_exit(cpu);
}
-static void mttcg_start_vcpu_thread(CPUState *cpu)
+void mttcg_start_vcpu_thread(CPUState *cpu)
{
char thread_name[VCPU_THREAD_NAME_SIZE];
@@ -131,10 +132,3 @@ static void mttcg_start_vcpu_thread(CPUState *cpu)
cpu->hThread = qemu_thread_get_handle(cpu->thread);
#endif
}
-
-const CpusAccel tcg_cpus_mttcg = {
- .create_vcpu_thread = mttcg_start_vcpu_thread,
- .kick_vcpu_thread = mttcg_kick_vcpu_thread,
-
- .handle_interrupt = tcg_cpus_handle_interrupt,
-};
diff --git a/accel/tcg/tcg-cpus-mttcg.h b/accel/tcg/tcg-cpus-mttcg.h
new file mode 100644
index 0000000000..0af91dd3b3
--- /dev/null
+++ b/accel/tcg/tcg-cpus-mttcg.h
@@ -0,0 +1,19 @@
+/*
+ * QEMU TCG Multi Threaded vCPUs implementation
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#ifndef TCG_CPUS_MTTCG_H
+#define TCG_CPUS_MTTCG_H
+
+/* kick MTTCG vCPU thread */
+void mttcg_kick_vcpu_thread(CPUState *cpu);
+
+/* start an mttcg vCPU thread */
+void mttcg_start_vcpu_thread(CPUState *cpu);
+
+#endif /* TCG_CPUS_MTTCG_H */
diff --git a/accel/tcg/tcg-cpus-rr.c b/accel/tcg/tcg-cpus-rr.c
index 0181d2e4eb..802c57bb60 100644
--- a/accel/tcg/tcg-cpus-rr.c
+++ b/accel/tcg/tcg-cpus-rr.c
@@ -296,10 +296,3 @@ void rr_start_vcpu_thread(CPUState *cpu)
cpu->created = true;
}
}
-
-const CpusAccel tcg_cpus_rr = {
- .create_vcpu_thread = rr_start_vcpu_thread,
- .kick_vcpu_thread = rr_kick_vcpu_thread,
-
- .handle_interrupt = tcg_cpus_handle_interrupt,
-};
diff --git a/accel/tcg/tcg-cpus.c b/accel/tcg/tcg-cpus.c
index e335f9f155..38a58ab271 100644
--- a/accel/tcg/tcg-cpus.c
+++ b/accel/tcg/tcg-cpus.c
@@ -35,6 +35,9 @@
#include "hw/boards.h"
#include "tcg-cpus.h"
+#include "tcg-cpus-mttcg.h"
+#include "tcg-cpus-rr.h"
+#include "tcg-cpus-icount.h"
/* common functionality among all TCG variants */
@@ -80,3 +83,43 @@ void tcg_cpus_handle_interrupt(CPUState *cpu, int mask)
qatomic_set(&cpu_neg(cpu)->icount_decr.u16.high, -1);
}
}
+
+static void tcg_cpus_ops_init(AccelOpsClass *ops)
+{
+ if (qemu_tcg_mttcg_enabled()) {
+ ops->create_vcpu_thread = mttcg_start_vcpu_thread;
+ ops->kick_vcpu_thread = mttcg_kick_vcpu_thread;
+ ops->handle_interrupt = tcg_cpus_handle_interrupt;
+
+ } else if (icount_enabled()) {
+ ops->create_vcpu_thread = rr_start_vcpu_thread;
+ ops->kick_vcpu_thread = rr_kick_vcpu_thread;
+ ops->handle_interrupt = icount_handle_interrupt;
+ ops->get_virtual_clock = icount_get;
+ ops->get_elapsed_ticks = icount_get;
+
+ } else {
+ ops->create_vcpu_thread = rr_start_vcpu_thread;
+ ops->kick_vcpu_thread = rr_kick_vcpu_thread;
+ ops->handle_interrupt = tcg_cpus_handle_interrupt;
+ }
+}
+
+static void tcg_cpus_class_init(ObjectClass *oc, void *data)
+{
+ AccelOpsClass *ops = ACCEL_OPS_CLASS(oc);
+
+ ops->ops_init = tcg_cpus_ops_init;
+};
+static const TypeInfo tcg_cpus_type_info = {
+ .name = ACCEL_OPS_NAME("tcg"),
+
+ .parent = TYPE_ACCEL_OPS,
+ .class_init = tcg_cpus_class_init,
+ .abstract = true,
+};
+static void tcg_cpus_register_types(void)
+{
+ type_register_static(&tcg_cpus_type_info);
+}
+type_init(tcg_cpus_register_types);
diff --git a/accel/tcg/tcg-cpus.h b/accel/tcg/tcg-cpus.h
index d6893a32f8..923cbace12 100644
--- a/accel/tcg/tcg-cpus.h
+++ b/accel/tcg/tcg-cpus.h
@@ -14,10 +14,6 @@
#include "sysemu/cpus.h"
-extern const CpusAccel tcg_cpus_mttcg;
-extern const CpusAccel tcg_cpus_icount;
-extern const CpusAccel tcg_cpus_rr;
-
void tcg_cpus_destroy(CPUState *cpu);
int tcg_cpus_exec(CPUState *cpu);
void tcg_cpus_handle_interrupt(CPUState *cpu, int mask);
diff --git a/accel/xen/xen-all.c b/accel/xen/xen-all.c
index 594aaf6b49..cd5aa2b96e 100644
--- a/accel/xen/xen-all.c
+++ b/accel/xen/xen-all.c
@@ -154,10 +154,6 @@ static void xen_setup_post(MachineState *ms, AccelState *accel)
}
}
-const CpusAccel xen_cpus = {
- .create_vcpu_thread = dummy_start_vcpu_thread,
-};
-
static int xen_init(MachineState *ms)
{
MachineClass *mc = MACHINE_GET_CLASS(ms);
@@ -185,9 +181,6 @@ static int xen_init(MachineState *ms)
* opt out of system RAM being allocated by generic code
*/
mc->default_ram_id = NULL;
-
- cpus_register_accel(&xen_cpus);
-
return 0;
}
@@ -222,9 +215,24 @@ static const TypeInfo xen_accel_type = {
.class_init = xen_accel_class_init,
};
+static void xen_cpus_class_init(ObjectClass *oc, void *data)
+{
+ AccelOpsClass *ops = ACCEL_OPS_CLASS(oc);
+
+ ops->create_vcpu_thread = dummy_start_vcpu_thread;
+};
+static const TypeInfo xen_cpus_type_info = {
+ .name = ACCEL_OPS_NAME("xen"),
+
+ .parent = TYPE_ACCEL_OPS,
+ .class_init = xen_cpus_class_init,
+ .abstract = true,
+};
+
static void xen_type_init(void)
{
type_register_static(&xen_accel_type);
+ type_register_static(&xen_cpus_type_info);
}
type_init(xen_type_init);
diff --git a/bsd-user/main.c b/bsd-user/main.c
index ff295bcb29..a68ce5f446 100644
--- a/bsd-user/main.c
+++ b/bsd-user/main.c
@@ -908,13 +908,14 @@ int main(int argc, char **argv)
#endif
}
+ cpu_type = parse_cpu_option(cpu_model);
/* init tcg before creating CPUs and to get qemu_host_page_size */
{
AccelClass *ac = ACCEL_GET_CLASS(current_accel());
ac->init_machine(NULL);
+ accel_init_interfaces(ac);
}
- cpu_type = parse_cpu_option(cpu_model);
cpu = cpu_create(cpu_type);
env = cpu->env_ptr;
#if defined(TARGET_SPARC) || defined(TARGET_PPC)
diff --git a/include/qemu/accel.h b/include/qemu/accel.h
index fac4a18703..b9d6d69eb8 100644
--- a/include/qemu/accel.h
+++ b/include/qemu/accel.h
@@ -69,6 +69,8 @@ typedef struct AccelClass {
AccelClass *accel_find(const char *opt_name);
AccelState *current_accel(void);
+void accel_init_interfaces(AccelClass *ac);
+
#ifndef CONFIG_USER_ONLY
int accel_init_machine(AccelState *accel, MachineState *ms);
diff --git a/include/sysemu/accel-ops.h b/include/sysemu/accel-ops.h
new file mode 100644
index 0000000000..6102d2f80d
--- /dev/null
+++ b/include/sysemu/accel-ops.h
@@ -0,0 +1,45 @@
+/*
+ * Accelerator OPS, used for cpus.c module
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#ifndef ACCEL_OPS_H
+#define ACCEL_OPS_H
+
+#include "qom/object.h"
+
+#define ACCEL_OPS_SUFFIX "-ops"
+#define TYPE_ACCEL_OPS "accel" ACCEL_OPS_SUFFIX
+#define ACCEL_OPS_NAME(name) (name "-" TYPE_ACCEL_OPS)
+
+typedef struct AccelOpsClass AccelOpsClass;
+DECLARE_CLASS_CHECKERS(AccelOpsClass, ACCEL_OPS, TYPE_ACCEL_OPS)
+
+/* cpus.c operations interface */
+struct AccelOpsClass {
+ /*< private >*/
+ ObjectClass parent_class;
+ /*< public >*/
+
+ /* initialization function called when accel is chosen */
+ void (*ops_init)(AccelOpsClass *ops);
+
+ void (*create_vcpu_thread)(CPUState *cpu); /* MANDATORY NON-NULL */
+ void (*kick_vcpu_thread)(CPUState *cpu);
+
+ void (*synchronize_post_reset)(CPUState *cpu);
+ void (*synchronize_post_init)(CPUState *cpu);
+ void (*synchronize_state)(CPUState *cpu);
+ void (*synchronize_pre_loadvm)(CPUState *cpu);
+
+ void (*handle_interrupt)(CPUState *cpu, int mask);
+
+ int64_t (*get_virtual_clock)(void);
+ int64_t (*get_elapsed_ticks)(void);
+};
+
+#endif /* ACCEL_OPS_H */
diff --git a/include/sysemu/cpus.h b/include/sysemu/cpus.h
index e8156728c6..2cd74392e0 100644
--- a/include/sysemu/cpus.h
+++ b/include/sysemu/cpus.h
@@ -2,30 +2,14 @@
#define QEMU_CPUS_H
#include "qemu/timer.h"
+#include "sysemu/accel-ops.h"
-/* cpus.c */
+/* register accel-specific operations */
+void cpus_register_accel(const AccelOpsClass *i);
-/* CPU execution threads */
+/* accel/dummy-cpus.c */
-typedef struct CpusAccel {
- void (*create_vcpu_thread)(CPUState *cpu); /* MANDATORY */
- void (*kick_vcpu_thread)(CPUState *cpu);
-
- void (*synchronize_post_reset)(CPUState *cpu);
- void (*synchronize_post_init)(CPUState *cpu);
- void (*synchronize_state)(CPUState *cpu);
- void (*synchronize_pre_loadvm)(CPUState *cpu);
-
- void (*handle_interrupt)(CPUState *cpu, int mask);
-
- int64_t (*get_virtual_clock)(void);
- int64_t (*get_elapsed_ticks)(void);
-} CpusAccel;
-
-/* register accel-specific cpus interface implementation */
-void cpus_register_accel(const CpusAccel *i);
-
-/* Create a dummy vcpu for CpusAccel->create_vcpu_thread */
+/* Create a dummy vcpu for AccelOpsClass->create_vcpu_thread */
void dummy_start_vcpu_thread(CPUState *);
/* interface available for cpus accelerator threads */
diff --git a/linux-user/main.c b/linux-user/main.c
index 5c059a8445..e28d016c6f 100644
--- a/linux-user/main.c
+++ b/linux-user/main.c
@@ -708,6 +708,7 @@ int main(int argc, char **argv, char **envp)
AccelClass *ac = ACCEL_GET_CLASS(current_accel());
ac->init_machine(NULL);
+ accel_init_interfaces(ac);
}
cpu = cpu_create(cpu_type);
env = cpu->env_ptr;
diff --git a/softmmu/cpus.c b/softmmu/cpus.c
index e46ac68ad0..659617e7ef 100644
--- a/softmmu/cpus.c
+++ b/softmmu/cpus.c
@@ -127,7 +127,7 @@ void hw_error(const char *fmt, ...)
/*
* The chosen accelerator is supposed to register this.
*/
-static const CpusAccel *cpus_accel;
+static const AccelOpsClass *cpus_accel;
void cpu_synchronize_all_states(void)
{
@@ -593,11 +593,11 @@ void cpu_remove_sync(CPUState *cpu)
qemu_mutex_lock_iothread();
}
-void cpus_register_accel(const CpusAccel *ca)
+void cpus_register_accel(const AccelOpsClass *ops)
{
- assert(ca != NULL);
- assert(ca->create_vcpu_thread != NULL); /* mandatory */
- cpus_accel = ca;
+ assert(ops != NULL);
+ assert(ops->create_vcpu_thread != NULL); /* mandatory */
+ cpus_accel = ops;
}
void qemu_init_vcpu(CPUState *cpu)
@@ -617,7 +617,7 @@ void qemu_init_vcpu(CPUState *cpu)
cpu_address_space_init(cpu, 0, "cpu-memory", cpu->memory);
}
- /* accelerators all implement the CpusAccel interface */
+ /* accelerators all implement the AccelOpsClass */
g_assert(cpus_accel != NULL && cpus_accel->create_vcpu_thread != NULL);
cpus_accel->create_vcpu_thread(cpu);
diff --git a/softmmu/vl.c b/softmmu/vl.c
index bc20c526d2..b97708300e 100644
--- a/softmmu/vl.c
+++ b/softmmu/vl.c
@@ -2558,7 +2558,7 @@ static bool object_create_initial(const char *type, QemuOpts *opts)
}
/* Memory allocation by backends needs to be done
- * after configure_accelerator() (due to the tcg_enabled()
+ * after do_configure_accelerator() (due to the tcg_enabled()
* checks at memory_region_init_*()).
*
* Also, allocation of large amounts of memory may delay
@@ -4186,7 +4186,7 @@ void qemu_init(int argc, char **argv, char **envp)
*
* Machine compat properties: object_set_machine_compat_props().
* Accelerator compat props: object_set_accelerator_compat_props(),
- * called from configure_accelerator().
+ * called from do_configure_accelerator().
*/
if (!qtest_enabled() && machine_class->deprecation_reason) {
@@ -4321,6 +4321,8 @@ void qemu_init(int argc, char **argv, char **envp)
if (cpu_option) {
current_machine->cpu_type = parse_cpu_option(cpu_option);
}
+ /* NB: for machine none cpu_type could STILL be NULL here! */
+ accel_init_interfaces(ACCEL_GET_CLASS(current_machine->accelerator));
if (current_machine->ram_memdev_id) {
Object *backend;
diff --git a/target/i386/hax/hax-all.c b/target/i386/hax/hax-all.c
index d7f4bb44a7..ec3c426223 100644
--- a/target/i386/hax/hax-all.c
+++ b/target/i386/hax/hax-all.c
@@ -364,9 +364,6 @@ static int hax_accel_init(MachineState *ms)
!ret ? "working" : "not working",
!ret ? "fast virt" : "emulation");
}
- if (ret == 0) {
- cpus_register_accel(&hax_cpus);
- }
return ret;
}
diff --git a/target/i386/hax/hax-cpus.c b/target/i386/hax/hax-cpus.c
index f72c85bd49..2f8424388d 100644
--- a/target/i386/hax/hax-cpus.c
+++ b/target/i386/hax/hax-cpus.c
@@ -74,12 +74,27 @@ static void hax_start_vcpu_thread(CPUState *cpu)
#endif
}
-const CpusAccel hax_cpus = {
- .create_vcpu_thread = hax_start_vcpu_thread,
- .kick_vcpu_thread = hax_kick_vcpu_thread,
+static void hax_cpus_class_init(ObjectClass *oc, void *data)
+{
+ AccelOpsClass *ops = ACCEL_OPS_CLASS(oc);
- .synchronize_post_reset = hax_cpu_synchronize_post_reset,
- .synchronize_post_init = hax_cpu_synchronize_post_init,
- .synchronize_state = hax_cpu_synchronize_state,
- .synchronize_pre_loadvm = hax_cpu_synchronize_pre_loadvm,
+ ops->create_vcpu_thread = hax_start_vcpu_thread;
+ ops->kick_vcpu_thread = hax_kick_vcpu_thread;
+
+ ops->synchronize_post_reset = hax_cpu_synchronize_post_reset;
+ ops->synchronize_post_init = hax_cpu_synchronize_post_init;
+ ops->synchronize_state = hax_cpu_synchronize_state;
+ ops->synchronize_pre_loadvm = hax_cpu_synchronize_pre_loadvm;
+};
+static const TypeInfo hax_cpus_type_info = {
+ .name = ACCEL_OPS_NAME("hax"),
+
+ .parent = TYPE_ACCEL_OPS,
+ .class_init = hax_cpus_class_init,
+ .abstract = true,
};
+static void hax_cpus_register_types(void)
+{
+ type_register_static(&hax_cpus_type_info);
+}
+type_init(hax_cpus_register_types);
diff --git a/target/i386/hax/hax-cpus.h b/target/i386/hax/hax-cpus.h
index ee8ab7a631..c7698519cd 100644
--- a/target/i386/hax/hax-cpus.h
+++ b/target/i386/hax/hax-cpus.h
@@ -12,8 +12,6 @@
#include "sysemu/cpus.h"
-extern const CpusAccel hax_cpus;
-
#include "hax-interface.h"
#include "hax-i386.h"
diff --git a/target/i386/hvf/hvf-cpus.c b/target/i386/hvf/hvf-cpus.c
index 817b3d7452..dd022a84c4 100644
--- a/target/i386/hvf/hvf-cpus.c
+++ b/target/i386/hvf/hvf-cpus.c
@@ -121,11 +121,26 @@ static void hvf_start_vcpu_thread(CPUState *cpu)
cpu, QEMU_THREAD_JOINABLE);
}
-const CpusAccel hvf_cpus = {
- .create_vcpu_thread = hvf_start_vcpu_thread,
+static void hvf_cpus_class_init(ObjectClass *oc, void *data)
+{
+ AccelOpsClass *ops = ACCEL_OPS_CLASS(oc);
- .synchronize_post_reset = hvf_cpu_synchronize_post_reset,
- .synchronize_post_init = hvf_cpu_synchronize_post_init,
- .synchronize_state = hvf_cpu_synchronize_state,
- .synchronize_pre_loadvm = hvf_cpu_synchronize_pre_loadvm,
+ ops->create_vcpu_thread = hvf_start_vcpu_thread;
+
+ ops->synchronize_post_reset = hvf_cpu_synchronize_post_reset;
+ ops->synchronize_post_init = hvf_cpu_synchronize_post_init;
+ ops->synchronize_state = hvf_cpu_synchronize_state;
+ ops->synchronize_pre_loadvm = hvf_cpu_synchronize_pre_loadvm;
+};
+static const TypeInfo hvf_cpus_type_info = {
+ .name = ACCEL_OPS_NAME("hvf"),
+
+ .parent = TYPE_ACCEL_OPS,
+ .class_init = hvf_cpus_class_init,
+ .abstract = true,
};
+static void hvf_cpus_register_types(void)
+{
+ type_register_static(&hvf_cpus_type_info);
+}
+type_init(hvf_cpus_register_types);
diff --git a/target/i386/hvf/hvf-cpus.h b/target/i386/hvf/hvf-cpus.h
index ced31b82c0..8f992da168 100644
--- a/target/i386/hvf/hvf-cpus.h
+++ b/target/i386/hvf/hvf-cpus.h
@@ -12,8 +12,6 @@
#include "sysemu/cpus.h"
-extern const CpusAccel hvf_cpus;
-
int hvf_init_vcpu(CPUState *);
int hvf_vcpu_exec(CPUState *);
void hvf_cpu_synchronize_state(CPUState *);
diff --git a/target/i386/hvf/hvf.c b/target/i386/hvf/hvf.c
index ffc9efa40f..bd94bb5243 100644
--- a/target/i386/hvf/hvf.c
+++ b/target/i386/hvf/hvf.c
@@ -887,7 +887,6 @@ static int hvf_accel_init(MachineState *ms)
hvf_state = s;
memory_listener_register(&hvf_memory_listener, &address_space_memory);
- cpus_register_accel(&hvf_cpus);
return 0;
}
diff --git a/target/i386/whpx/whpx-all.c b/target/i386/whpx/whpx-all.c
index ee6b606194..90adae9af7 100644
--- a/target/i386/whpx/whpx-all.c
+++ b/target/i386/whpx/whpx-all.c
@@ -1642,8 +1642,6 @@ static int whpx_accel_init(MachineState *ms)
whpx_memory_init();
- cpus_register_accel(&whpx_cpus);
-
printf("Windows Hypervisor Platform accelerator is operational\n");
return 0;
diff --git a/target/i386/whpx/whpx-cpus.c b/target/i386/whpx/whpx-cpus.c
index d9bd5a2d36..8e82974de7 100644
--- a/target/i386/whpx/whpx-cpus.c
+++ b/target/i386/whpx/whpx-cpus.c
@@ -85,12 +85,27 @@ static void whpx_kick_vcpu_thread(CPUState *cpu)
}
}
-const CpusAccel whpx_cpus = {
- .create_vcpu_thread = whpx_start_vcpu_thread,
- .kick_vcpu_thread = whpx_kick_vcpu_thread,
+static void whpx_cpus_class_init(ObjectClass *oc, void *data)
+{
+ AccelOpsClass *ops = ACCEL_OPS_CLASS(oc);
- .synchronize_post_reset = whpx_cpu_synchronize_post_reset,
- .synchronize_post_init = whpx_cpu_synchronize_post_init,
- .synchronize_state = whpx_cpu_synchronize_state,
- .synchronize_pre_loadvm = whpx_cpu_synchronize_pre_loadvm,
+ ops->create_vcpu_thread = whpx_start_vcpu_thread;
+ ops->kick_vcpu_thread = whpx_kick_vcpu_thread;
+
+ ops->synchronize_post_reset = whpx_cpu_synchronize_post_reset;
+ ops->synchronize_post_init = whpx_cpu_synchronize_post_init;
+ ops->synchronize_state = whpx_cpu_synchronize_state;
+ ops->synchronize_pre_loadvm = whpx_cpu_synchronize_pre_loadvm;
+};
+static const TypeInfo whpx_cpus_type_info = {
+ .name = ACCEL_OPS_NAME("whpx"),
+
+ .parent = TYPE_ACCEL_OPS,
+ .class_init = whpx_cpus_class_init,
+ .abstract = true,
};
+static void whpx_cpus_register_types(void)
+{
+ type_register_static(&whpx_cpus_type_info);
+}
+type_init(whpx_cpus_register_types);
diff --git a/target/i386/whpx/whpx-cpus.h b/target/i386/whpx/whpx-cpus.h
index bdb367d1d0..2dee6d61ea 100644
--- a/target/i386/whpx/whpx-cpus.h
+++ b/target/i386/whpx/whpx-cpus.h
@@ -12,8 +12,6 @@
#include "sysemu/cpus.h"
-extern const CpusAccel whpx_cpus;
-
int whpx_init_vcpu(CPUState *cpu);
int whpx_vcpu_exec(CPUState *cpu);
void whpx_destroy_vcpu(CPUState *cpu);
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 28/32] accel: introduce AccelCPUClass extending CPUClass
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (26 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 27/32] accel: replace struct CpusAccel with AccelOpsClass Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-08 19:48 ` [RFC v9 29/32] i386: split cpu accelerators from cpu.c, using AccelCPUClass Claudio Fontana
` (6 subsequent siblings)
34 siblings, 0 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
add a new optional interface to CPUClass,
which allows accelerators to extend the CPUClass
with additional accelerator-specific initializations.
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
MAINTAINERS | 1 +
accel/accel-common.c | 44 +++++++++++++++++++++++++++++++++++++
include/hw/core/accel-cpu.h | 25 +++++++++++++++++++++
include/hw/core/cpu.h | 13 +++++++++++
4 files changed, 83 insertions(+)
create mode 100644 include/hw/core/accel-cpu.h
diff --git a/MAINTAINERS b/MAINTAINERS
index 8f0e773a47..f084d73f6b 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -436,6 +436,7 @@ R: Paolo Bonzini <pbonzini@redhat.com>
S: Maintained
F: include/qemu/accel.h
F: include/sysemu/accel-ops.h
+F: include/hw/core/accel-cpu.h
F: accel/accel-*.c
F: accel/Makefile.objs
F: accel/stubs/Makefile.objs
diff --git a/accel/accel-common.c b/accel/accel-common.c
index 6b59873419..ef73c761fc 100644
--- a/accel/accel-common.c
+++ b/accel/accel-common.c
@@ -26,6 +26,9 @@
#include "qemu/osdep.h"
#include "qemu/accel.h"
+#include "cpu.h"
+#include "hw/core/accel-cpu.h"
+
#ifndef CONFIG_USER_ONLY
#include "accel-softmmu.h"
#endif /* !CONFIG_USER_ONLY */
@@ -46,16 +49,57 @@ AccelClass *accel_find(const char *opt_name)
return ac;
}
+static void accel_init_cpu_int_aux(ObjectClass *klass, void *opaque)
+{
+ CPUClass *cc = CPU_CLASS(klass);
+ AccelCPUClass *accel_cpu_interface = opaque;
+
+ cc->accel_cpu_interface = accel_cpu_interface;
+ if (accel_cpu_interface->cpu_class_init) {
+ accel_cpu_interface->cpu_class_init(cc);
+ }
+}
+
+/* initialize the arch-specific accel CpuClass interfaces */
+static void accel_init_cpu_interfaces(AccelClass *ac)
+{
+ const char *ac_name; /* AccelClass name */
+ char *acc_name; /* AccelCPUClass name */
+ ObjectClass *acc; /* AccelCPUClass */
+
+ ac_name = object_class_get_name(OBJECT_CLASS(ac));
+ g_assert(ac_name != NULL);
+
+ acc_name = g_strdup_printf("%s-%s", ac_name, CPU_RESOLVING_TYPE);
+ acc = object_class_by_name(acc_name);
+ g_free(acc_name);
+
+ if (acc) {
+ object_class_foreach(accel_init_cpu_int_aux,
+ CPU_RESOLVING_TYPE, false, acc);
+ }
+}
+
void accel_init_interfaces(AccelClass *ac)
{
#ifndef CONFIG_USER_ONLY
accel_init_ops_interfaces(ac);
#endif /* !CONFIG_USER_ONLY */
+
+ accel_init_cpu_interfaces(ac);
}
+static const TypeInfo accel_cpu_type = {
+ .name = TYPE_ACCEL_CPU,
+ .parent = TYPE_OBJECT,
+ .abstract = true,
+ .class_size = sizeof(AccelCPUClass),
+};
+
static void register_accel_types(void)
{
type_register_static(&accel_type);
+ type_register_static(&accel_cpu_type);
}
type_init(register_accel_types);
diff --git a/include/hw/core/accel-cpu.h b/include/hw/core/accel-cpu.h
new file mode 100644
index 0000000000..dce08a9100
--- /dev/null
+++ b/include/hw/core/accel-cpu.h
@@ -0,0 +1,25 @@
+/*
+ * Accelerator interface, specializes CPUClass
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#ifndef ACCEL_CPU_H
+#define ACCEL_CPU_H
+
+/*
+ * these defines cannot be in cpu.h, because we are using
+ * CPU_RESOLVING_TYPE here.
+ * Use this header to define your accelerator-specific
+ * cpu-specific accelerator interfaces.
+ */
+
+#define TYPE_ACCEL_CPU "accel-" CPU_RESOLVING_TYPE
+#define ACCEL_CPU_NAME(name) (name "-" TYPE_ACCEL_CPU)
+typedef struct AccelCPUClass AccelCPUClass;
+DECLARE_CLASS_CHECKERS(AccelCPUClass, ACCEL_CPU, TYPE_ACCEL_CPU)
+
+#endif /* ACCEL_CPU_H */
diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
index db54223983..97e1dd8279 100644
--- a/include/hw/core/cpu.h
+++ b/include/hw/core/cpu.h
@@ -76,6 +76,17 @@ typedef struct CPUWatchpoint CPUWatchpoint;
struct TranslationBlock;
+/* see also accel-cpu.h */
+typedef struct AccelCPUClass {
+ /*< private >*/
+ ObjectClass parent_class;
+ /*< public >*/
+
+ void (*cpu_class_init)(CPUClass *cc);
+ void (*cpu_instance_init)(CPUState *cpu);
+ void (*cpu_realizefn)(CPUState *cpu, Error **errp);
+} AccelCPUClass;
+
#ifdef CONFIG_TCG
#include "tcg-cpu-ops.h"
#endif /* CONFIG_TCG */
@@ -196,6 +207,8 @@ struct CPUClass {
#ifdef CONFIG_TCG
TcgCpuOperations tcg_ops;
#endif /* CONFIG_TCG */
+
+ AccelCPUClass *accel_cpu_interface;
};
/*
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 29/32] i386: split cpu accelerators from cpu.c, using AccelCPUClass
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (27 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 28/32] accel: introduce AccelCPUClass extending CPUClass Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-08 19:48 ` [RFC v9 30/32] cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn Claudio Fontana
` (5 subsequent siblings)
34 siblings, 0 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
i386 is the first user of AccelCPUClass, allowing to split
cpu.c into:
cpu.c cpuid and common x86 cpu functionality
host-cpu.c host x86 cpu functions and "host" cpu type
kvm/cpu.c KVM x86 AccelCPUClass
hvf/cpu.c HVF x86 AccelCPUClass
tcg/cpu.c TCG x86 AccelCPUClass
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
MAINTAINERS | 2 +-
hw/i386/pc_piix.c | 1 +
target/i386/cpu.c | 386 ++++--------------------------------
target/i386/cpu.h | 20 +-
target/i386/host-cpu.c | 198 ++++++++++++++++++
target/i386/host-cpu.h | 19 ++
target/i386/hvf/cpu.c | 65 ++++++
target/i386/hvf/meson.build | 1 +
target/i386/kvm/cpu.c | 148 ++++++++++++++
target/i386/kvm/kvm-cpu.h | 41 ++++
target/i386/kvm/kvm.c | 3 +-
target/i386/kvm/meson.build | 7 +-
target/i386/meson.build | 8 +-
target/i386/tcg-cpu.c | 71 -------
target/i386/tcg-cpu.h | 15 --
target/i386/tcg/cpu.c | 173 ++++++++++++++++
target/i386/tcg/meson.build | 3 +-
17 files changed, 711 insertions(+), 450 deletions(-)
create mode 100644 target/i386/host-cpu.c
create mode 100644 target/i386/host-cpu.h
create mode 100644 target/i386/hvf/cpu.c
create mode 100644 target/i386/kvm/cpu.c
create mode 100644 target/i386/kvm/kvm-cpu.h
delete mode 100644 target/i386/tcg-cpu.c
delete mode 100644 target/i386/tcg-cpu.h
create mode 100644 target/i386/tcg/cpu.c
diff --git a/MAINTAINERS b/MAINTAINERS
index f084d73f6b..4f3f1e8b18 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -337,7 +337,7 @@ M: Paolo Bonzini <pbonzini@redhat.com>
M: Richard Henderson <richard.henderson@linaro.org>
M: Eduardo Habkost <ehabkost@redhat.com>
S: Maintained
-F: target/i386/
+F: target/i386/tcg/
F: tests/tcg/i386/
F: tests/tcg/x86_64/
F: hw/i386/
diff --git a/hw/i386/pc_piix.c b/hw/i386/pc_piix.c
index 13d1628f13..d3f013f3a1 100644
--- a/hw/i386/pc_piix.c
+++ b/hw/i386/pc_piix.c
@@ -64,6 +64,7 @@
#include "hw/hyperv/vmbus-bridge.h"
#include "hw/mem/nvdimm.h"
#include "hw/i386/acpi-build.h"
+#include "kvm/kvm-cpu.h"
#define MAX_IDE_BUS 2
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 3462d0143f..27fba3b003 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -22,37 +22,24 @@
#include "qemu/cutils.h"
#include "qemu/bitops.h"
#include "qemu/qemu-print.h"
-
#include "cpu.h"
-#include "tcg-cpu.h"
#include "helper-tcg.h"
#include "exec/exec-all.h"
#include "sysemu/kvm.h"
#include "sysemu/reset.h"
#include "sysemu/hvf.h"
-#include "sysemu/cpus.h"
+#include "hw/core/accel-cpu.h"
#include "sysemu/xen.h"
#include "kvm/kvm_i386.h"
#include "sev_i386.h"
-
-#include "qemu/error-report.h"
#include "qemu/module.h"
-#include "qemu/option.h"
-#include "qemu/config-file.h"
-#include "qapi/error.h"
#include "qapi/qapi-visit-machine.h"
#include "qapi/qapi-visit-run-state.h"
#include "qapi/qmp/qdict.h"
#include "qapi/qmp/qerror.h"
-#include "qapi/visitor.h"
#include "qom/qom-qobject.h"
-#include "sysemu/arch_init.h"
#include "qapi/qapi-commands-machine-target.h"
-
#include "standard-headers/asm-x86/kvm_para.h"
-
-#include "sysemu/sysemu.h"
-#include "sysemu/tcg.h"
#include "hw/qdev-properties.h"
#include "hw/i386/topology.h"
#ifndef CONFIG_USER_ONLY
@@ -594,8 +581,8 @@ static CPUCacheInfo legacy_l3_cache = {
#define INTEL_PT_CYCLE_BITMAP 0x1fff /* Support 0,2^(0~11) */
#define INTEL_PT_PSB_BITMAP (0x003f << 16) /* Support 2K,4K,8K,16K,32K,64K */
-static void x86_cpu_vendor_words2str(char *dst, uint32_t vendor1,
- uint32_t vendor2, uint32_t vendor3)
+void x86_cpu_vendor_words2str(char *dst, uint32_t vendor1,
+ uint32_t vendor2, uint32_t vendor3)
{
int i;
for (i = 0; i < 4; i++) {
@@ -1563,25 +1550,6 @@ void host_cpuid(uint32_t function, uint32_t count,
*edx = vec[3];
}
-void host_vendor_fms(char *vendor, int *family, int *model, int *stepping)
-{
- uint32_t eax, ebx, ecx, edx;
-
- host_cpuid(0x0, 0, &eax, &ebx, &ecx, &edx);
- x86_cpu_vendor_words2str(vendor, ebx, edx, ecx);
-
- host_cpuid(0x1, 0, &eax, &ebx, &ecx, &edx);
- if (family) {
- *family = ((eax >> 8) & 0x0F) + ((eax >> 20) & 0xFF);
- }
- if (model) {
- *model = ((eax >> 4) & 0x0F) | ((eax & 0xF0000) >> 12);
- }
- if (stepping) {
- *stepping = eax & 0x0F;
- }
-}
-
/* CPU class name definitions: */
/* Return type name for a given CPU model name
@@ -1606,10 +1574,6 @@ static char *x86_cpu_class_get_model_name(X86CPUClass *cc)
strlen(class_name) - strlen(X86_CPU_TYPE_SUFFIX));
}
-typedef struct PropValue {
- const char *prop, *value;
-} PropValue;
-
typedef struct X86CPUVersionDefinition {
X86CPUVersion version;
const char *alias;
@@ -4106,31 +4070,6 @@ static X86CPUDefinition builtin_x86_defs[] = {
},
};
-/* KVM-specific features that are automatically added/removed
- * from all CPU models when KVM is enabled.
- */
-static PropValue kvm_default_props[] = {
- { "kvmclock", "on" },
- { "kvm-nopiodelay", "on" },
- { "kvm-asyncpf", "on" },
- { "kvm-steal-time", "on" },
- { "kvm-pv-eoi", "on" },
- { "kvmclock-stable-bit", "on" },
- { "x2apic", "on" },
- { "acpi", "off" },
- { "monitor", "off" },
- { "svm", "off" },
- { NULL, NULL },
-};
-
-/* TCG-specific defaults that override all CPU models when using TCG
- */
-static PropValue tcg_default_props[] = {
- { "vme", "off" },
- { NULL, NULL },
-};
-
-
/*
* We resolve CPU model aliases using -v1 when using "-machine
* none", but this is just for compatibility while libvirt isn't
@@ -4172,61 +4111,6 @@ static X86CPUVersion x86_cpu_model_resolve_version(const X86CPUModel *model)
return v;
}
-void x86_cpu_change_kvm_default(const char *prop, const char *value)
-{
- PropValue *pv;
- for (pv = kvm_default_props; pv->prop; pv++) {
- if (!strcmp(pv->prop, prop)) {
- pv->value = value;
- break;
- }
- }
-
- /* It is valid to call this function only for properties that
- * are already present in the kvm_default_props table.
- */
- assert(pv->prop);
-}
-
-static bool lmce_supported(void)
-{
- uint64_t mce_cap = 0;
-
-#ifdef CONFIG_KVM
- if (kvm_ioctl(kvm_state, KVM_X86_GET_MCE_CAP_SUPPORTED, &mce_cap) < 0) {
- return false;
- }
-#endif
-
- return !!(mce_cap & MCG_LMCE_P);
-}
-
-#define CPUID_MODEL_ID_SZ 48
-
-/**
- * cpu_x86_fill_model_id:
- * Get CPUID model ID string from host CPU.
- *
- * @str should have at least CPUID_MODEL_ID_SZ bytes
- *
- * The function does NOT add a null terminator to the string
- * automatically.
- */
-static int cpu_x86_fill_model_id(char *str)
-{
- uint32_t eax = 0, ebx = 0, ecx = 0, edx = 0;
- int i;
-
- for (i = 0; i < 3; i++) {
- host_cpuid(0x80000002 + i, 0, &eax, &ebx, &ecx, &edx);
- memcpy(str + i * 16 + 0, &eax, 4);
- memcpy(str + i * 16 + 4, &ebx, 4);
- memcpy(str + i * 16 + 8, &ecx, 4);
- memcpy(str + i * 16 + 12, &edx, 4);
- }
- return 0;
-}
-
static Property max_x86_cpu_properties[] = {
DEFINE_PROP_BOOL("migratable", X86CPU, migratable, true),
DEFINE_PROP_BOOL("host-cache-info", X86CPU, cache_info_passthrough, false),
@@ -4249,61 +4133,25 @@ static void max_x86_cpu_class_init(ObjectClass *oc, void *data)
static void max_x86_cpu_initfn(Object *obj)
{
X86CPU *cpu = X86_CPU(obj);
- CPUX86State *env = &cpu->env;
- KVMState *s = kvm_state;
/* We can't fill the features array here because we don't know yet if
* "migratable" is true or false.
*/
cpu->max_features = true;
-
- if (accel_uses_host_cpuid()) {
- char vendor[CPUID_VENDOR_SZ + 1] = { 0 };
- char model_id[CPUID_MODEL_ID_SZ + 1] = { 0 };
- int family, model, stepping;
-
- host_vendor_fms(vendor, &family, &model, &stepping);
- cpu_x86_fill_model_id(model_id);
-
- object_property_set_str(OBJECT(cpu), "vendor", vendor, &error_abort);
- object_property_set_int(OBJECT(cpu), "family", family, &error_abort);
- object_property_set_int(OBJECT(cpu), "model", model, &error_abort);
- object_property_set_int(OBJECT(cpu), "stepping", stepping,
- &error_abort);
- object_property_set_str(OBJECT(cpu), "model-id", model_id,
- &error_abort);
-
- if (kvm_enabled()) {
- env->cpuid_min_level =
- kvm_arch_get_supported_cpuid(s, 0x0, 0, R_EAX);
- env->cpuid_min_xlevel =
- kvm_arch_get_supported_cpuid(s, 0x80000000, 0, R_EAX);
- env->cpuid_min_xlevel2 =
- kvm_arch_get_supported_cpuid(s, 0xC0000000, 0, R_EAX);
- } else {
- env->cpuid_min_level =
- hvf_get_supported_cpuid(0x0, 0, R_EAX);
- env->cpuid_min_xlevel =
- hvf_get_supported_cpuid(0x80000000, 0, R_EAX);
- env->cpuid_min_xlevel2 =
- hvf_get_supported_cpuid(0xC0000000, 0, R_EAX);
- }
-
- if (lmce_supported()) {
- object_property_set_bool(OBJECT(cpu), "lmce", true, &error_abort);
- }
- } else {
- object_property_set_str(OBJECT(cpu), "vendor", CPUID_VENDOR_AMD,
- &error_abort);
- object_property_set_int(OBJECT(cpu), "family", 6, &error_abort);
- object_property_set_int(OBJECT(cpu), "model", 6, &error_abort);
- object_property_set_int(OBJECT(cpu), "stepping", 3, &error_abort);
- object_property_set_str(OBJECT(cpu), "model-id",
- "QEMU TCG CPU version " QEMU_HW_VERSION,
- &error_abort);
- }
-
object_property_set_bool(OBJECT(cpu), "pmu", true, &error_abort);
+
+ /*
+ * these defaults are used for TCG and all other accelerators
+ * besides KVM and HVF, which overwrite these values
+ */
+ object_property_set_str(OBJECT(cpu), "vendor", CPUID_VENDOR_AMD,
+ &error_abort);
+ object_property_set_int(OBJECT(cpu), "family", 6, &error_abort);
+ object_property_set_int(OBJECT(cpu), "model", 6, &error_abort);
+ object_property_set_int(OBJECT(cpu), "stepping", 3, &error_abort);
+ object_property_set_str(OBJECT(cpu), "model-id",
+ "QEMU TCG CPU version " QEMU_HW_VERSION,
+ &error_abort);
}
static const TypeInfo max_x86_cpu_type_info = {
@@ -4313,31 +4161,6 @@ static const TypeInfo max_x86_cpu_type_info = {
.class_init = max_x86_cpu_class_init,
};
-#if defined(CONFIG_KVM) || defined(CONFIG_HVF)
-static void host_x86_cpu_class_init(ObjectClass *oc, void *data)
-{
- X86CPUClass *xcc = X86_CPU_CLASS(oc);
-
- xcc->host_cpuid_required = true;
- xcc->ordering = 8;
-
-#if defined(CONFIG_KVM)
- xcc->model_description =
- "KVM processor with all supported host features ";
-#elif defined(CONFIG_HVF)
- xcc->model_description =
- "HVF processor with all supported host features ";
-#endif
-}
-
-static const TypeInfo host_x86_cpu_type_info = {
- .name = X86_CPU_TYPE_NAME("host"),
- .parent = X86_CPU_TYPE_NAME("max"),
- .class_init = host_x86_cpu_class_init,
-};
-
-#endif
-
static char *feature_word_description(FeatureWordInfo *f, uint32_t bit)
{
assert(f->type == CPUID_FEATURE_WORD || f->type == MSR_FEATURE_WORD);
@@ -5063,7 +4886,7 @@ static uint64_t x86_cpu_get_supported_feature_word(FeatureWord w,
return r;
}
-static void x86_cpu_apply_props(X86CPU *cpu, PropValue *props)
+void x86_cpu_apply_props(X86CPU *cpu, PropValue *props)
{
PropValue *pv;
for (pv = props; pv->prop; pv++) {
@@ -5110,8 +4933,6 @@ static void x86_cpu_load_model(X86CPU *cpu, X86CPUModel *model)
{
X86CPUDefinition *def = model->cpudef;
CPUX86State *env = &cpu->env;
- const char *vendor;
- char host_vendor[CPUID_VENDOR_SZ + 1];
FeatureWord w;
/*NOTE: any property set by this function should be returned by
@@ -5138,18 +4959,6 @@ static void x86_cpu_load_model(X86CPU *cpu, X86CPUModel *model)
/* legacy-cache defaults to 'off' if CPU model provides cache info */
cpu->legacy_cache = !def->cache_info;
- /* Special cases not set in the X86CPUDefinition structs: */
- /* TODO: in-kernel irqchip for hvf */
- if (kvm_enabled()) {
- if (!kvm_irqchip_in_kernel()) {
- x86_cpu_change_kvm_default("x2apic", "off");
- }
-
- x86_cpu_apply_props(cpu, kvm_default_props);
- } else if (tcg_enabled()) {
- x86_cpu_apply_props(cpu, tcg_default_props);
- }
-
env->features[FEAT_1_ECX] |= CPUID_EXT_HYPERVISOR;
/* sysenter isn't supported in compatibility mode on AMD,
@@ -5159,15 +4968,12 @@ static void x86_cpu_load_model(X86CPU *cpu, X86CPUModel *model)
* KVM's sysenter/syscall emulation in compatibility mode and
* when doing cross vendor migration
*/
- vendor = def->vendor;
- if (accel_uses_host_cpuid()) {
- uint32_t ebx = 0, ecx = 0, edx = 0;
- host_cpuid(0, 0, NULL, &ebx, &ecx, &edx);
- x86_cpu_vendor_words2str(host_vendor, ebx, edx, ecx);
- vendor = host_vendor;
- }
- object_property_set_str(OBJECT(cpu), "vendor", vendor, &error_abort);
+ /*
+ * vendor property is set here but then overloaded with the
+ * host cpu vendor for KVM and HVF.
+ */
+ object_property_set_str(OBJECT(cpu), "vendor", def->vendor, &error_abort);
x86_cpu_apply_version_props(cpu, model);
@@ -6192,53 +5998,12 @@ static void x86_cpu_apic_realize(X86CPU *cpu, Error **errp)
apic_mmio_map_once = true;
}
}
-
-static void x86_cpu_machine_done(Notifier *n, void *unused)
-{
- X86CPU *cpu = container_of(n, X86CPU, machine_done);
- MemoryRegion *smram =
- (MemoryRegion *) object_resolve_path("/machine/smram", NULL);
-
- if (smram) {
- cpu->smram = g_new(MemoryRegion, 1);
- memory_region_init_alias(cpu->smram, OBJECT(cpu), "smram",
- smram, 0, 4 * GiB);
- memory_region_set_enabled(cpu->smram, true);
- memory_region_add_subregion_overlap(cpu->cpu_as_root, 0, cpu->smram, 1);
- }
-}
#else
static void x86_cpu_apic_realize(X86CPU *cpu, Error **errp)
{
}
#endif
-/* Note: Only safe for use on x86(-64) hosts */
-static uint32_t x86_host_phys_bits(void)
-{
- uint32_t eax;
- uint32_t host_phys_bits;
-
- host_cpuid(0x80000000, 0, &eax, NULL, NULL, NULL);
- if (eax >= 0x80000008) {
- host_cpuid(0x80000008, 0, &eax, NULL, NULL, NULL);
- /* Note: According to AMD doc 25481 rev 2.34 they have a field
- * at 23:16 that can specify a maximum physical address bits for
- * the guest that can override this value; but I've not seen
- * anything with that set.
- */
- host_phys_bits = eax & 0xff;
- } else {
- /* It's an odd 64 bit machine that doesn't have the leaf for
- * physical address bits; fall back to 36 that's most older
- * Intel.
- */
- host_phys_bits = 36;
- }
-
- return host_phys_bits;
-}
-
static void x86_cpu_adjust_level(X86CPU *cpu, uint32_t *min, uint32_t value)
{
if (*min < value) {
@@ -6515,33 +6280,22 @@ static void x86_cpu_filter_features(X86CPU *cpu, bool verbose)
static void x86_cpu_realizefn(DeviceState *dev, Error **errp)
{
CPUState *cs = CPU(dev);
+ CPUClass *cc = CPU_GET_CLASS(cs);
X86CPU *cpu = X86_CPU(dev);
X86CPUClass *xcc = X86_CPU_GET_CLASS(dev);
CPUX86State *env = &cpu->env;
Error *local_err = NULL;
static bool ht_warned;
- if (xcc->host_cpuid_required) {
- if (!accel_uses_host_cpuid()) {
- g_autofree char *name = x86_cpu_class_get_model_name(xcc);
- error_setg(&local_err, "CPU model '%s' requires KVM", name);
- goto out;
- }
+ /* The accelerator realizefn needs to be called first. */
+ if (cc->accel_cpu_interface) {
+ cc->accel_cpu_interface->cpu_realizefn(cs, errp);
}
- if (cpu->max_features && accel_uses_host_cpuid()) {
- if (enable_cpu_pm) {
- host_cpuid(5, 0, &cpu->mwait.eax, &cpu->mwait.ebx,
- &cpu->mwait.ecx, &cpu->mwait.edx);
- env->features[FEAT_1_ECX] |= CPUID_EXT_MONITOR;
- if (kvm_enabled() && kvm_has_waitpkg()) {
- env->features[FEAT_7_0_ECX] |= CPUID_7_0_ECX_WAITPKG;
- }
- }
- if (kvm_enabled() && cpu->ucode_rev == 0) {
- cpu->ucode_rev = kvm_arch_get_supported_msr_feature(kvm_state,
- MSR_IA32_UCODE_REV);
- }
+ if (xcc->host_cpuid_required && !accel_uses_host_cpuid()) {
+ g_autofree char *name = x86_cpu_class_get_model_name(xcc);
+ error_setg(&local_err, "CPU model '%s' requires KVM or HVF", name);
+ goto out;
}
if (cpu->ucode_rev == 0) {
@@ -6593,39 +6347,7 @@ static void x86_cpu_realizefn(DeviceState *dev, Error **errp)
* consumer AMD devices but nothing else.
*/
if (env->features[FEAT_8000_0001_EDX] & CPUID_EXT2_LM) {
- if (accel_uses_host_cpuid()) {
- uint32_t host_phys_bits = x86_host_phys_bits();
- static bool warned;
-
- /* Print a warning if the user set it to a value that's not the
- * host value.
- */
- if (cpu->phys_bits != host_phys_bits && cpu->phys_bits != 0 &&
- !warned) {
- warn_report("Host physical bits (%u)"
- " does not match phys-bits property (%u)",
- host_phys_bits, cpu->phys_bits);
- warned = true;
- }
-
- if (cpu->host_phys_bits) {
- /* The user asked for us to use the host physical bits */
- cpu->phys_bits = host_phys_bits;
- if (cpu->host_phys_bits_limit &&
- cpu->phys_bits > cpu->host_phys_bits_limit) {
- cpu->phys_bits = cpu->host_phys_bits_limit;
- }
- }
-
- if (cpu->phys_bits &&
- (cpu->phys_bits > TARGET_PHYS_ADDR_SPACE_BITS ||
- cpu->phys_bits < 32)) {
- error_setg(errp, "phys-bits should be between 32 and %u "
- " (but is %u)",
- TARGET_PHYS_ADDR_SPACE_BITS, cpu->phys_bits);
- return;
- }
- } else {
+ if (!accel_uses_host_cpuid()) {
if (cpu->phys_bits && cpu->phys_bits != TCG_PHYS_ADDR_BITS) {
error_setg(errp, "TCG only supports phys-bits=%u",
TCG_PHYS_ADDR_BITS);
@@ -6633,8 +6355,8 @@ static void x86_cpu_realizefn(DeviceState *dev, Error **errp)
}
}
/* 0 means it was not explicitly set by the user (or by machine
- * compat_props or by the host code above). In this case, the default
- * is the value used by TCG (40).
+ * compat_props or by the host code in host-cpu.c).
+ * In this case, the default is the value used by TCG (40).
*/
if (cpu->phys_bits == 0) {
cpu->phys_bits = TCG_PHYS_ADDR_BITS;
@@ -6704,33 +6426,6 @@ static void x86_cpu_realizefn(DeviceState *dev, Error **errp)
mce_init(cpu);
-#ifndef CONFIG_USER_ONLY
- if (tcg_enabled()) {
- cpu->cpu_as_mem = g_new(MemoryRegion, 1);
- cpu->cpu_as_root = g_new(MemoryRegion, 1);
-
- /* Outer container... */
- memory_region_init(cpu->cpu_as_root, OBJECT(cpu), "memory", ~0ull);
- memory_region_set_enabled(cpu->cpu_as_root, true);
-
- /* ... with two regions inside: normal system memory with low
- * priority, and...
- */
- memory_region_init_alias(cpu->cpu_as_mem, OBJECT(cpu), "memory",
- get_system_memory(), 0, ~0ull);
- memory_region_add_subregion_overlap(cpu->cpu_as_root, 0, cpu->cpu_as_mem, 0);
- memory_region_set_enabled(cpu->cpu_as_mem, true);
-
- cs->num_ases = 2;
- cpu_address_space_init(cs, 0, "cpu-memory", cs->memory);
- cpu_address_space_init(cs, 1, "cpu-smm", cpu->cpu_as_root);
-
- /* ... SMRAM with higher priority, linked from /machine/smram. */
- cpu->machine_done.notify = x86_cpu_machine_done;
- qemu_add_machine_init_done_notifier(&cpu->machine_done);
- }
-#endif
-
qemu_init_vcpu(cs);
/*
@@ -6936,6 +6631,8 @@ static void x86_cpu_initfn(Object *obj)
{
X86CPU *cpu = X86_CPU(obj);
X86CPUClass *xcc = X86_CPU_GET_CLASS(obj);
+ CPUClass *cc = CPU_CLASS(xcc);
+
CPUX86State *env = &cpu->env;
FeatureWord w;
@@ -6992,6 +6689,11 @@ static void x86_cpu_initfn(Object *obj)
if (xcc->model) {
x86_cpu_load_model(cpu, xcc->model);
}
+
+ /* if required, do the accelerator-specific cpu initialization */
+ if (cc->accel_cpu_interface) {
+ cc->accel_cpu_interface->cpu_instance_init(CPU(obj));
+ }
}
static int64_t x86_cpu_get_arch_id(CPUState *cs)
@@ -7248,11 +6950,6 @@ static void x86_cpu_common_class_init(ObjectClass *oc, void *data)
cc->class_by_name = x86_cpu_class_by_name;
cc->parse_features = x86_cpu_parse_featurestr;
cc->has_work = x86_cpu_has_work;
-
-#ifdef CONFIG_TCG
- tcg_cpu_common_class_init(cc);
-#endif /* CONFIG_TCG */
-
cc->dump_state = x86_cpu_dump_state;
cc->set_pc = x86_cpu_set_pc;
cc->gdb_read_register = x86_cpu_gdb_read_register;
@@ -7357,9 +7054,6 @@ static void x86_cpu_register_types(void)
}
type_register_static(&max_x86_cpu_type_info);
type_register_static(&x86_base_cpu_type_info);
-#if defined(CONFIG_KVM) || defined(CONFIG_HVF)
- type_register_static(&host_x86_cpu_type_info);
-#endif
}
type_init(x86_cpu_register_types)
diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index a0d64613dc..b3e39fc631 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -1905,13 +1905,20 @@ int cpu_x86_signal_handler(int host_signum, void *pinfo,
void *puc);
/* cpu.c */
+void x86_cpu_vendor_words2str(char *dst, uint32_t vendor1,
+ uint32_t vendor2, uint32_t vendor3);
+typedef struct PropValue {
+ const char *prop, *value;
+} PropValue;
+void x86_cpu_apply_props(X86CPU *cpu, PropValue *props);
+
+/* cpu.c other functions (cpuid) */
void cpu_x86_cpuid(CPUX86State *env, uint32_t index, uint32_t count,
uint32_t *eax, uint32_t *ebx,
uint32_t *ecx, uint32_t *edx);
void cpu_clear_apic_feature(CPUX86State *env);
void host_cpuid(uint32_t function, uint32_t count,
uint32_t *eax, uint32_t *ebx, uint32_t *ecx, uint32_t *edx);
-void host_vendor_fms(char *vendor, int *family, int *model, int *stepping);
/* helper.c */
void x86_cpu_set_a20(X86CPU *cpu, int a20_state);
@@ -2111,17 +2118,6 @@ void cpu_report_tpr_access(CPUX86State *env, TPRAccess access);
void apic_handle_tpr_access_report(DeviceState *d, target_ulong ip,
TPRAccess access);
-
-/* Change the value of a KVM-specific default
- *
- * If value is NULL, no default will be set and the original
- * value from the CPU model table will be kept.
- *
- * It is valid to call this function only for properties that
- * are already present in the kvm_default_props table.
- */
-void x86_cpu_change_kvm_default(const char *prop, const char *value);
-
/* Special values for X86CPUVersion: */
/* Resolve to latest CPU version */
diff --git a/target/i386/host-cpu.c b/target/i386/host-cpu.c
new file mode 100644
index 0000000000..3ce2bc9a84
--- /dev/null
+++ b/target/i386/host-cpu.c
@@ -0,0 +1,198 @@
+/*
+ * x86 host CPU functions, and "host" cpu type initialization
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#include "qemu/osdep.h"
+#include "cpu.h"
+#include "host-cpu.h"
+#include "qapi/error.h"
+#include "sysemu/sysemu.h"
+
+/* Note: Only safe for use on x86(-64) hosts */
+static uint32_t host_cpu_phys_bits(void)
+{
+ uint32_t eax;
+ uint32_t host_phys_bits;
+
+ host_cpuid(0x80000000, 0, &eax, NULL, NULL, NULL);
+ if (eax >= 0x80000008) {
+ host_cpuid(0x80000008, 0, &eax, NULL, NULL, NULL);
+ /*
+ * Note: According to AMD doc 25481 rev 2.34 they have a field
+ * at 23:16 that can specify a maximum physical address bits for
+ * the guest that can override this value; but I've not seen
+ * anything with that set.
+ */
+ host_phys_bits = eax & 0xff;
+ } else {
+ /*
+ * It's an odd 64 bit machine that doesn't have the leaf for
+ * physical address bits; fall back to 36 that's most older
+ * Intel.
+ */
+ host_phys_bits = 36;
+ }
+
+ return host_phys_bits;
+}
+
+static void host_cpu_enable_cpu_pm(X86CPU *cpu)
+{
+ CPUX86State *env = &cpu->env;
+
+ host_cpuid(5, 0, &cpu->mwait.eax, &cpu->mwait.ebx,
+ &cpu->mwait.ecx, &cpu->mwait.edx);
+ env->features[FEAT_1_ECX] |= CPUID_EXT_MONITOR;
+}
+
+static uint32_t host_cpu_adjust_phys_bits(X86CPU *cpu, Error **errp)
+{
+ uint32_t host_phys_bits = host_cpu_phys_bits();
+ uint32_t phys_bits = cpu->phys_bits;
+ static bool warned;
+
+ /*
+ * Print a warning if the user set it to a value that's not the
+ * host value.
+ */
+ if (phys_bits != host_phys_bits && phys_bits != 0 &&
+ !warned) {
+ warn_report("Host physical bits (%u)"
+ " does not match phys-bits property (%u)",
+ host_phys_bits, phys_bits);
+ warned = true;
+ }
+
+ if (cpu->host_phys_bits) {
+ /* The user asked for us to use the host physical bits */
+ phys_bits = host_phys_bits;
+ if (cpu->host_phys_bits_limit &&
+ phys_bits > cpu->host_phys_bits_limit) {
+ phys_bits = cpu->host_phys_bits_limit;
+ }
+ }
+
+ if (phys_bits &&
+ (phys_bits > TARGET_PHYS_ADDR_SPACE_BITS ||
+ phys_bits < 32)) {
+ error_setg(errp, "phys-bits should be between 32 and %u "
+ " (but is %u)",
+ TARGET_PHYS_ADDR_SPACE_BITS, phys_bits);
+ }
+
+ return phys_bits;
+}
+
+void host_cpu_realizefn(CPUState *cs, Error **errp)
+{
+ X86CPU *cpu = X86_CPU(cs);
+ CPUX86State *env = &cpu->env;
+
+ if (cpu->max_features && enable_cpu_pm) {
+ host_cpu_enable_cpu_pm(cpu);
+ }
+ if (env->features[FEAT_8000_0001_EDX] & CPUID_EXT2_LM) {
+ cpu->phys_bits = host_cpu_adjust_phys_bits(cpu, errp);
+ }
+}
+
+#define CPUID_MODEL_ID_SZ 48
+/**
+ * cpu_x86_fill_model_id:
+ * Get CPUID model ID string from host CPU.
+ *
+ * @str should have at least CPUID_MODEL_ID_SZ bytes
+ *
+ * The function does NOT add a null terminator to the string
+ * automatically.
+ */
+static int host_cpu_fill_model_id(char *str)
+{
+ uint32_t eax = 0, ebx = 0, ecx = 0, edx = 0;
+ int i;
+
+ for (i = 0; i < 3; i++) {
+ host_cpuid(0x80000002 + i, 0, &eax, &ebx, &ecx, &edx);
+ memcpy(str + i * 16 + 0, &eax, 4);
+ memcpy(str + i * 16 + 4, &ebx, 4);
+ memcpy(str + i * 16 + 8, &ecx, 4);
+ memcpy(str + i * 16 + 12, &edx, 4);
+ }
+ return 0;
+}
+
+void host_cpu_vendor_fms(char *vendor, int *family, int *model, int *stepping)
+{
+ uint32_t eax, ebx, ecx, edx;
+
+ host_cpuid(0x0, 0, &eax, &ebx, &ecx, &edx);
+ x86_cpu_vendor_words2str(vendor, ebx, edx, ecx);
+
+ host_cpuid(0x1, 0, &eax, &ebx, &ecx, &edx);
+ if (family) {
+ *family = ((eax >> 8) & 0x0F) + ((eax >> 20) & 0xFF);
+ }
+ if (model) {
+ *model = ((eax >> 4) & 0x0F) | ((eax & 0xF0000) >> 12);
+ }
+ if (stepping) {
+ *stepping = eax & 0x0F;
+ }
+}
+
+void host_cpu_instance_init(X86CPU *cpu)
+{
+ uint32_t ebx = 0, ecx = 0, edx = 0;
+ char vendor[CPUID_VENDOR_SZ + 1];
+
+ host_cpuid(0, 0, NULL, &ebx, &ecx, &edx);
+ x86_cpu_vendor_words2str(vendor, ebx, edx, ecx);
+
+ object_property_set_str(OBJECT(cpu), "vendor", vendor, &error_abort);
+}
+
+void host_cpu_max_instance_init(X86CPU *cpu)
+{
+ char vendor[CPUID_VENDOR_SZ + 1] = { 0 };
+ char model_id[CPUID_MODEL_ID_SZ + 1] = { 0 };
+ int family, model, stepping;
+
+ host_cpu_vendor_fms(vendor, &family, &model, &stepping);
+ host_cpu_fill_model_id(model_id);
+
+ object_property_set_str(OBJECT(cpu), "vendor", vendor, &error_abort);
+ object_property_set_int(OBJECT(cpu), "family", family, &error_abort);
+ object_property_set_int(OBJECT(cpu), "model", model, &error_abort);
+ object_property_set_int(OBJECT(cpu), "stepping", stepping,
+ &error_abort);
+ object_property_set_str(OBJECT(cpu), "model-id", model_id,
+ &error_abort);
+}
+
+static void host_cpu_class_init(ObjectClass *oc, void *data)
+{
+ X86CPUClass *xcc = X86_CPU_CLASS(oc);
+
+ xcc->host_cpuid_required = true;
+ xcc->ordering = 8;
+ xcc->model_description =
+ g_strdup_printf("processor with all supported host features ");
+}
+
+static const TypeInfo host_cpu_type_info = {
+ .name = X86_CPU_TYPE_NAME("host"),
+ .parent = X86_CPU_TYPE_NAME("max"),
+ .class_init = host_cpu_class_init,
+};
+
+static void host_cpu_type_init(void)
+{
+ type_register_static(&host_cpu_type_info);
+}
+
+type_init(host_cpu_type_init);
diff --git a/target/i386/host-cpu.h b/target/i386/host-cpu.h
new file mode 100644
index 0000000000..d1f2644422
--- /dev/null
+++ b/target/i386/host-cpu.h
@@ -0,0 +1,19 @@
+/*
+ * x86 host CPU type initialization and host CPU functions
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#ifndef HOST_CPU_H
+#define HOST_CPU_H
+
+void host_cpu_instance_init(X86CPU *cpu);
+void host_cpu_max_instance_init(X86CPU *cpu);
+void host_cpu_realizefn(CPUState *cs, Error **errp);
+
+void host_cpu_vendor_fms(char *vendor, int *family, int *model, int *stepping);
+
+#endif /* HOST_CPU_H */
diff --git a/target/i386/hvf/cpu.c b/target/i386/hvf/cpu.c
new file mode 100644
index 0000000000..d6579571f1
--- /dev/null
+++ b/target/i386/hvf/cpu.c
@@ -0,0 +1,65 @@
+/*
+ * x86 HVF CPU type initialization
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#include "qemu/osdep.h"
+#include "cpu.h"
+#include "host-cpu.h"
+#include "qapi/error.h"
+#include "sysemu/sysemu.h"
+#include "hw/boards.h"
+#include "sysemu/hvf.h"
+#include "hw/core/accel-cpu.h"
+
+static void hvf_cpu_max_instance_init(X86CPU *cpu)
+{
+ CPUX86State *env = &cpu->env;
+
+ host_cpu_max_instance_init(cpu);
+
+ env->cpuid_min_level =
+ hvf_get_supported_cpuid(0x0, 0, R_EAX);
+ env->cpuid_min_xlevel =
+ hvf_get_supported_cpuid(0x80000000, 0, R_EAX);
+ env->cpuid_min_xlevel2 =
+ hvf_get_supported_cpuid(0xC0000000, 0, R_EAX);
+}
+
+static void hvf_cpu_instance_init(CPUState *cs)
+{
+ X86CPU *cpu = X86_CPU(cs);
+
+ host_cpu_instance_init(cpu);
+
+ /* Special cases not set in the X86CPUDefinition structs: */
+ /* TODO: in-kernel irqchip for hvf */
+
+ if (cpu->max_features) {
+ hvf_cpu_max_instance_init(cpu);
+ }
+}
+
+static void hvf_cpu_accel_class_init(ObjectClass *oc, void *data)
+{
+ AccelCPUClass *acc = ACCEL_CPU_CLASS(oc);
+
+ acc->cpu_realizefn = host_cpu_realizefn;
+ acc->cpu_instance_init = hvf_cpu_instance_init;
+};
+static const TypeInfo hvf_cpu_accel_type_info = {
+ .name = ACCEL_CPU_NAME("hvf"),
+
+ .parent = TYPE_ACCEL_CPU,
+ .class_init = hvf_cpu_accel_class_init,
+ .abstract = true,
+};
+static void hvf_cpu_accel_register_types(void)
+{
+ type_register_static(&hvf_cpu_accel_type_info);
+}
+type_init(hvf_cpu_accel_register_types);
diff --git a/target/i386/hvf/meson.build b/target/i386/hvf/meson.build
index 409c9a3f14..a7fba5724c 100644
--- a/target/i386/hvf/meson.build
+++ b/target/i386/hvf/meson.build
@@ -10,4 +10,5 @@ i386_softmmu_ss.add(when: [hvf, 'CONFIG_HVF'], if_true: files(
'x86_mmu.c',
'x86_task.c',
'x86hvf.c',
+ 'cpu.c',
))
diff --git a/target/i386/kvm/cpu.c b/target/i386/kvm/cpu.c
new file mode 100644
index 0000000000..adc5120cf6
--- /dev/null
+++ b/target/i386/kvm/cpu.c
@@ -0,0 +1,148 @@
+/*
+ * x86 KVM CPU type initialization
+ *
+ * Copyright 2020 SUSE LLC
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#include "qemu/osdep.h"
+#include "cpu.h"
+#include "host-cpu.h"
+#include "kvm-cpu.h"
+#include "qapi/error.h"
+#include "sysemu/sysemu.h"
+#include "hw/boards.h"
+
+#include "kvm_i386.h"
+#include "hw/core/accel-cpu.h"
+
+static void kvm_cpu_realizefn(CPUState *cs, Error **errp)
+{
+ X86CPU *cpu = X86_CPU(cs);
+ CPUX86State *env = &cpu->env;
+
+ /*
+ * The realize order is important, since x86_cpu_realize() checks if
+ * nothing else has been set by the user (or by accelerators) in
+ * cpu->ucode_rev and cpu->phys_bits.
+ *
+ * realize order:
+ * kvm_cpu -> host_cpu -> x86_cpu
+ */
+ if (cpu->max_features) {
+ if (enable_cpu_pm && kvm_has_waitpkg()) {
+ env->features[FEAT_7_0_ECX] |= CPUID_7_0_ECX_WAITPKG;
+ }
+ if (cpu->ucode_rev == 0) {
+ cpu->ucode_rev =
+ kvm_arch_get_supported_msr_feature(kvm_state,
+ MSR_IA32_UCODE_REV);
+ }
+ }
+ host_cpu_realizefn(cs, errp);
+}
+
+/*
+ * KVM-specific features that are automatically added/removed
+ * from all CPU models when KVM is enabled.
+ */
+static PropValue kvm_default_props[] = {
+ { "kvmclock", "on" },
+ { "kvm-nopiodelay", "on" },
+ { "kvm-asyncpf", "on" },
+ { "kvm-steal-time", "on" },
+ { "kvm-pv-eoi", "on" },
+ { "kvmclock-stable-bit", "on" },
+ { "x2apic", "on" },
+ { "acpi", "off" },
+ { "monitor", "off" },
+ { "svm", "off" },
+ { NULL, NULL },
+};
+
+void x86_cpu_change_kvm_default(const char *prop, const char *value)
+{
+ PropValue *pv;
+ for (pv = kvm_default_props; pv->prop; pv++) {
+ if (!strcmp(pv->prop, prop)) {
+ pv->value = value;
+ break;
+ }
+ }
+
+ /*
+ * It is valid to call this function only for properties that
+ * are already present in the kvm_default_props table.
+ */
+ assert(pv->prop);
+}
+
+static bool lmce_supported(void)
+{
+ uint64_t mce_cap = 0;
+
+ if (kvm_ioctl(kvm_state, KVM_X86_GET_MCE_CAP_SUPPORTED, &mce_cap) < 0) {
+ return false;
+ }
+ return !!(mce_cap & MCG_LMCE_P);
+}
+
+static void kvm_cpu_max_instance_init(X86CPU *cpu)
+{
+ CPUX86State *env = &cpu->env;
+ KVMState *s = kvm_state;
+
+ host_cpu_max_instance_init(cpu);
+
+ if (lmce_supported()) {
+ object_property_set_bool(OBJECT(cpu), "lmce", true, &error_abort);
+ }
+
+ env->cpuid_min_level =
+ kvm_arch_get_supported_cpuid(s, 0x0, 0, R_EAX);
+ env->cpuid_min_xlevel =
+ kvm_arch_get_supported_cpuid(s, 0x80000000, 0, R_EAX);
+ env->cpuid_min_xlevel2 =
+ kvm_arch_get_supported_cpuid(s, 0xC0000000, 0, R_EAX);
+}
+
+static void kvm_cpu_instance_init(CPUState *cs)
+{
+ X86CPU *cpu = X86_CPU(cs);
+
+ host_cpu_instance_init(cpu);
+
+ if (!kvm_irqchip_in_kernel()) {
+ x86_cpu_change_kvm_default("x2apic", "off");
+ }
+
+ /* Special cases not set in the X86CPUDefinition structs: */
+
+ x86_cpu_apply_props(cpu, kvm_default_props);
+
+ if (cpu->max_features) {
+ kvm_cpu_max_instance_init(cpu);
+ }
+}
+
+static void kvm_cpu_accel_class_init(ObjectClass *oc, void *data)
+{
+ AccelCPUClass *acc = ACCEL_CPU_CLASS(oc);
+
+ acc->cpu_realizefn = kvm_cpu_realizefn;
+ acc->cpu_instance_init = kvm_cpu_instance_init;
+}
+static const TypeInfo kvm_cpu_accel_type_info = {
+ .name = ACCEL_CPU_NAME("kvm"),
+
+ .parent = TYPE_ACCEL_CPU,
+ .class_init = kvm_cpu_accel_class_init,
+ .abstract = true,
+};
+static void kvm_cpu_accel_register_types(void)
+{
+ type_register_static(&kvm_cpu_accel_type_info);
+}
+type_init(kvm_cpu_accel_register_types);
diff --git a/target/i386/kvm/kvm-cpu.h b/target/i386/kvm/kvm-cpu.h
new file mode 100644
index 0000000000..e858ca21e5
--- /dev/null
+++ b/target/i386/kvm/kvm-cpu.h
@@ -0,0 +1,41 @@
+/*
+ * i386 KVM CPU type and functions
+ *
+ * Copyright (c) 2003 Fabrice Bellard
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef KVM_CPU_H
+#define KVM_CPU_H
+
+#ifdef CONFIG_KVM
+/*
+ * Change the value of a KVM-specific default
+ *
+ * If value is NULL, no default will be set and the original
+ * value from the CPU model table will be kept.
+ *
+ * It is valid to call this function only for properties that
+ * are already present in the kvm_default_props table.
+ */
+void x86_cpu_change_kvm_default(const char *prop, const char *value);
+
+#else /* !CONFIG_KVM */
+
+#define x86_cpu_change_kvm_default(a, b)
+
+#endif /* CONFIG_KVM */
+
+#endif /* KVM_CPU_H */
diff --git a/target/i386/kvm/kvm.c b/target/i386/kvm/kvm.c
index a2934dda02..35c86fdba6 100644
--- a/target/i386/kvm/kvm.c
+++ b/target/i386/kvm/kvm.c
@@ -22,6 +22,7 @@
#include "standard-headers/asm-x86/kvm_para.h"
#include "cpu.h"
+#include "host-cpu.h"
#include "sysemu/sysemu.h"
#include "sysemu/hw_accel.h"
#include "sysemu/kvm_int.h"
@@ -285,7 +286,7 @@ static bool host_tsx_broken(void)
int family, model, stepping;\
char vendor[CPUID_VENDOR_SZ + 1];
- host_vendor_fms(vendor, &family, &model, &stepping);
+ host_cpu_vendor_fms(vendor, &family, &model, &stepping);
/* Check if we are running on a Haswell host known to have broken TSX */
return !strcmp(vendor, CPUID_VENDOR_INTEL) &&
diff --git a/target/i386/kvm/meson.build b/target/i386/kvm/meson.build
index 1d66559187..0bc3724eb3 100644
--- a/target/i386/kvm/meson.build
+++ b/target/i386/kvm/meson.build
@@ -1,3 +1,8 @@
i386_ss.add(when: 'CONFIG_KVM', if_false: files('kvm-stub.c'))
-i386_softmmu_ss.add(when: 'CONFIG_KVM', if_true: files('kvm.c'))
+
+i386_softmmu_ss.add(when: 'CONFIG_KVM', if_true: files(
+ 'kvm.c',
+ 'cpu.c',
+))
+
i386_softmmu_ss.add(when: 'CONFIG_HYPERV', if_true: files('hyperv.c'), if_false: files('hyperv-stub.c'))
diff --git a/target/i386/meson.build b/target/i386/meson.build
index 9c20208e5a..4e6e915e7f 100644
--- a/target/i386/meson.build
+++ b/target/i386/meson.build
@@ -6,8 +6,12 @@ i386_ss.add(files(
'xsave_helper.c',
'cpu-dump.c',
))
-i386_ss.add(when: 'CONFIG_TCG', if_true: files('tcg-cpu.c'))
-i386_ss.add(when: 'CONFIG_SEV', if_true: files('sev.c'), if_false: files('sev-stub.c'))
+
+i386_ss.add(when: 'CONFIG_SEV', if_true: files('host-cpu.c', 'sev.c'), if_false: files('sev-stub.c'))
+
+# x86 cpu type
+i386_ss.add(when: 'CONFIG_KVM', if_true: files('host-cpu.c'))
+i386_ss.add(when: 'CONFIG_HVF', if_true: files('host-cpu.c'))
i386_softmmu_ss = ss.source_set()
i386_softmmu_ss.add(files(
diff --git a/target/i386/tcg-cpu.c b/target/i386/tcg-cpu.c
deleted file mode 100644
index d2dd521612..0000000000
--- a/target/i386/tcg-cpu.c
+++ /dev/null
@@ -1,71 +0,0 @@
-/*
- * i386 TCG cpu class initialization
- *
- * Copyright (c) 2003 Fabrice Bellard
- *
- * This library is free software; you can redistribute it and/or
- * modify it under the terms of the GNU Lesser General Public
- * License as published by the Free Software Foundation; either
- * version 2 of the License, or (at your option) any later version.
- *
- * This library is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- * Lesser General Public License for more details.
- *
- * You should have received a copy of the GNU Lesser General Public
- * License along with this library; if not, see <http://www.gnu.org/licenses/>.
- */
-
-#include "qemu/osdep.h"
-#include "cpu.h"
-#include "tcg-cpu.h"
-#include "exec/exec-all.h"
-#include "sysemu/runstate.h"
-#include "helper-tcg.h"
-
-#if !defined(CONFIG_USER_ONLY)
-#include "hw/i386/apic.h"
-#endif
-
-/* Frob eflags into and out of the CPU temporary format. */
-
-static void x86_cpu_exec_enter(CPUState *cs)
-{
- X86CPU *cpu = X86_CPU(cs);
- CPUX86State *env = &cpu->env;
-
- CC_SRC = env->eflags & (CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
- env->df = 1 - (2 * ((env->eflags >> 10) & 1));
- CC_OP = CC_OP_EFLAGS;
- env->eflags &= ~(DF_MASK | CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
-}
-
-static void x86_cpu_exec_exit(CPUState *cs)
-{
- X86CPU *cpu = X86_CPU(cs);
- CPUX86State *env = &cpu->env;
-
- env->eflags = cpu_compute_eflags(env);
-}
-
-static void x86_cpu_synchronize_from_tb(CPUState *cs, TranslationBlock *tb)
-{
- X86CPU *cpu = X86_CPU(cs);
-
- cpu->env.eip = tb->pc - tb->cs_base;
-}
-
-void tcg_cpu_common_class_init(CPUClass *cc)
-{
- cc->tcg_ops.do_interrupt = x86_cpu_do_interrupt;
- cc->tcg_ops.cpu_exec_interrupt = x86_cpu_exec_interrupt;
- cc->tcg_ops.synchronize_from_tb = x86_cpu_synchronize_from_tb;
- cc->tcg_ops.cpu_exec_enter = x86_cpu_exec_enter;
- cc->tcg_ops.cpu_exec_exit = x86_cpu_exec_exit;
- cc->tcg_ops.initialize = tcg_x86_init;
- cc->tcg_ops.tlb_fill = x86_cpu_tlb_fill;
-#ifndef CONFIG_USER_ONLY
- cc->tcg_ops.debug_excp_handler = breakpoint_handler;
-#endif
-}
diff --git a/target/i386/tcg-cpu.h b/target/i386/tcg-cpu.h
deleted file mode 100644
index 81f02e562e..0000000000
--- a/target/i386/tcg-cpu.h
+++ /dev/null
@@ -1,15 +0,0 @@
-/*
- * i386 TCG CPU class initialization
- *
- * Copyright 2020 SUSE LLC
- *
- * This work is licensed under the terms of the GNU GPL, version 2 or later.
- * See the COPYING file in the top-level directory.
- */
-
-#ifndef TCG_CPU_H
-#define TCG_CPU_H
-
-void tcg_cpu_common_class_init(CPUClass *cc);
-
-#endif /* TCG_CPU_H */
diff --git a/target/i386/tcg/cpu.c b/target/i386/tcg/cpu.c
new file mode 100644
index 0000000000..45262de5fb
--- /dev/null
+++ b/target/i386/tcg/cpu.c
@@ -0,0 +1,173 @@
+/*
+ * i386 TCG cpu class initialization
+ *
+ * Copyright (c) 2003 Fabrice Bellard
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "qemu/osdep.h"
+#include "cpu.h"
+#include "helper-tcg.h"
+#include "qemu/accel.h"
+#include "hw/core/accel-cpu.h"
+
+#ifndef CONFIG_USER_ONLY
+#include "sysemu/sysemu.h"
+#include "qemu/units.h"
+#include "exec/address-spaces.h"
+#endif
+
+/* Frob eflags into and out of the CPU temporary format. */
+
+static void x86_cpu_exec_enter(CPUState *cs)
+{
+ X86CPU *cpu = X86_CPU(cs);
+ CPUX86State *env = &cpu->env;
+
+ CC_SRC = env->eflags & (CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
+ env->df = 1 - (2 * ((env->eflags >> 10) & 1));
+ CC_OP = CC_OP_EFLAGS;
+ env->eflags &= ~(DF_MASK | CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
+}
+
+static void x86_cpu_exec_exit(CPUState *cs)
+{
+ X86CPU *cpu = X86_CPU(cs);
+ CPUX86State *env = &cpu->env;
+
+ env->eflags = cpu_compute_eflags(env);
+}
+
+static void x86_cpu_synchronize_from_tb(CPUState *cs, TranslationBlock *tb)
+{
+ X86CPU *cpu = X86_CPU(cs);
+
+ cpu->env.eip = tb->pc - tb->cs_base;
+}
+
+#ifndef CONFIG_USER_ONLY
+
+static void x86_cpu_machine_done(Notifier *n, void *unused)
+{
+ X86CPU *cpu = container_of(n, X86CPU, machine_done);
+ MemoryRegion *smram =
+ (MemoryRegion *) object_resolve_path("/machine/smram", NULL);
+
+ if (smram) {
+ cpu->smram = g_new(MemoryRegion, 1);
+ memory_region_init_alias(cpu->smram, OBJECT(cpu), "smram",
+ smram, 0, 4 * GiB);
+ memory_region_set_enabled(cpu->smram, true);
+ memory_region_add_subregion_overlap(cpu->cpu_as_root, 0,
+ cpu->smram, 1);
+ }
+}
+
+static void tcg_cpu_realizefn(CPUState *cs, Error **errp)
+{
+ X86CPU *cpu = X86_CPU(cs);
+
+ /*
+ * The realize order is important, since x86_cpu_realize() checks if
+ * nothing else has been set by the user (or by accelerators) in
+ * cpu->ucode_rev and cpu->phys_bits, and the memory regions
+ * initialized here are needed for the vcpu initialization.
+ *
+ * realize order:
+ * tcg_cpu -> host_cpu -> x86_cpu
+ */
+ cpu->cpu_as_mem = g_new(MemoryRegion, 1);
+ cpu->cpu_as_root = g_new(MemoryRegion, 1);
+
+ /* Outer container... */
+ memory_region_init(cpu->cpu_as_root, OBJECT(cpu), "memory", ~0ull);
+ memory_region_set_enabled(cpu->cpu_as_root, true);
+
+ /*
+ * ... with two regions inside: normal system memory with low
+ * priority, and...
+ */
+ memory_region_init_alias(cpu->cpu_as_mem, OBJECT(cpu), "memory",
+ get_system_memory(), 0, ~0ull);
+ memory_region_add_subregion_overlap(cpu->cpu_as_root, 0, cpu->cpu_as_mem, 0);
+ memory_region_set_enabled(cpu->cpu_as_mem, true);
+
+ cs->num_ases = 2;
+ cpu_address_space_init(cs, 0, "cpu-memory", cs->memory);
+ cpu_address_space_init(cs, 1, "cpu-smm", cpu->cpu_as_root);
+
+ /* ... SMRAM with higher priority, linked from /machine/smram. */
+ cpu->machine_done.notify = x86_cpu_machine_done;
+ qemu_add_machine_init_done_notifier(&cpu->machine_done);
+}
+
+#else /* CONFIG_USER_ONLY */
+
+static void tcg_cpu_realizefn(CPUState *cs, Error **errp)
+{
+}
+
+#endif /* !CONFIG_USER_ONLY */
+
+
+static void tcg_cpu_class_init(CPUClass *cc)
+{
+ cc->tcg_ops.do_interrupt = x86_cpu_do_interrupt;
+ cc->tcg_ops.cpu_exec_interrupt = x86_cpu_exec_interrupt;
+ cc->tcg_ops.synchronize_from_tb = x86_cpu_synchronize_from_tb;
+ cc->tcg_ops.cpu_exec_enter = x86_cpu_exec_enter;
+ cc->tcg_ops.cpu_exec_exit = x86_cpu_exec_exit;
+ cc->tcg_ops.initialize = tcg_x86_init;
+ cc->tcg_ops.tlb_fill = x86_cpu_tlb_fill;
+#ifndef CONFIG_USER_ONLY
+ cc->tcg_ops.debug_excp_handler = breakpoint_handler;
+#endif /* !CONFIG_USER_ONLY */
+}
+
+/*
+ * TCG-specific defaults that override all CPU models when using TCG
+ */
+static PropValue tcg_default_props[] = {
+ { "vme", "off" },
+ { NULL, NULL },
+};
+
+static void tcg_cpu_instance_init(CPUState *cs)
+{
+ X86CPU *cpu = X86_CPU(cs);
+ /* Special cases not set in the X86CPUDefinition structs: */
+ x86_cpu_apply_props(cpu, tcg_default_props);
+}
+
+static void tcg_cpu_accel_class_init(ObjectClass *oc, void *data)
+{
+ AccelCPUClass *acc = ACCEL_CPU_CLASS(oc);
+
+ acc->cpu_realizefn = tcg_cpu_realizefn;
+ acc->cpu_class_init = tcg_cpu_class_init;
+ acc->cpu_instance_init = tcg_cpu_instance_init;
+}
+static const TypeInfo tcg_cpu_accel_type_info = {
+ .name = ACCEL_CPU_NAME("tcg"),
+
+ .parent = TYPE_ACCEL_CPU,
+ .class_init = tcg_cpu_accel_class_init,
+ .abstract = true,
+};
+static void tcg_cpu_accel_register_types(void)
+{
+ type_register_static(&tcg_cpu_accel_type_info);
+}
+type_init(tcg_cpu_accel_register_types);
diff --git a/target/i386/tcg/meson.build b/target/i386/tcg/meson.build
index 02794226c2..9e439df9c7 100644
--- a/target/i386/tcg/meson.build
+++ b/target/i386/tcg/meson.build
@@ -10,4 +10,5 @@ i386_ss.add(when: 'CONFIG_TCG', if_true: files(
'seg_helper.c',
'smm_helper.c',
'svm_helper.c',
- 'translate.c'), if_false: files('tcg-stub.c'))
+ 'translate.c',
+ 'cpu.c'), if_false: files('tcg-stub.c'))
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 30/32] cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (28 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 29/32] i386: split cpu accelerators from cpu.c, using AccelCPUClass Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-08 19:48 ` [RFC v9 31/32] hw/core/cpu: call qemu_init_vcpu in cpu_common_realizefn Claudio Fontana
` (4 subsequent siblings)
34 siblings, 0 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
move the call to the accel_cpu_interface method to the general
cpu_exec_realizefn from target/i386, so it does not need to be
called for every target explicitly as we enable more targets.
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
cpu.c | 5 +++++
target/i386/cpu.c | 15 ++++-----------
2 files changed, 9 insertions(+), 11 deletions(-)
diff --git a/cpu.c b/cpu.c
index 5cc8f181be..a59a909cfe 100644
--- a/cpu.c
+++ b/cpu.c
@@ -130,6 +130,11 @@ void cpu_exec_realizefn(CPUState *cpu, Error **errp)
cpu_list_add(cpu);
+ if (cc->accel_cpu_interface) {
+ /* NB: errp parameter is unused currently */
+ cc->accel_cpu_interface->cpu_realizefn(cpu, errp);
+ }
+
#ifdef CONFIG_TCG
/* NB: errp parameter is unused currently */
if (tcg_enabled()) {
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 27fba3b003..25998c0122 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -6280,16 +6280,16 @@ static void x86_cpu_filter_features(X86CPU *cpu, bool verbose)
static void x86_cpu_realizefn(DeviceState *dev, Error **errp)
{
CPUState *cs = CPU(dev);
- CPUClass *cc = CPU_GET_CLASS(cs);
X86CPU *cpu = X86_CPU(dev);
X86CPUClass *xcc = X86_CPU_GET_CLASS(dev);
CPUX86State *env = &cpu->env;
Error *local_err = NULL;
static bool ht_warned;
- /* The accelerator realizefn needs to be called first. */
- if (cc->accel_cpu_interface) {
- cc->accel_cpu_interface->cpu_realizefn(cs, errp);
+ cpu_exec_realizefn(cs, &local_err);
+ if (local_err != NULL) {
+ error_propagate(errp, local_err);
+ return;
}
if (xcc->host_cpuid_required && !accel_uses_host_cpuid()) {
@@ -6405,13 +6405,6 @@ static void x86_cpu_realizefn(DeviceState *dev, Error **errp)
env->cache_info_amd.l3_cache = &legacy_l3_cache;
}
-
- cpu_exec_realizefn(cs, &local_err);
- if (local_err != NULL) {
- error_propagate(errp, local_err);
- return;
- }
-
#ifndef CONFIG_USER_ONLY
MachineState *ms = MACHINE(qdev_get_machine());
qemu_register_reset(x86_cpu_machine_reset_cb, cpu);
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 31/32] hw/core/cpu: call qemu_init_vcpu in cpu_common_realizefn
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (29 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 30/32] cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-08 19:48 ` [RFC v9 32/32] cpu: introduce cpu_accel_instance_init Claudio Fontana
` (3 subsequent siblings)
34 siblings, 0 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
move the call to qemu_init_vcpu inside cpu_common_realizefn,
so it does not need to be done explicitly in each target cpu.
This makes it a little bit better, but still the way realize
is done continues to be bad; ideally the cpu_list_add would
be done in common_cpu, and in this case we could avoid even
more additional calls in target/xxx/cpu.c,
but this cannot happen because target cpu code, plugins, etc
now all rely on cpu->index, since no particular order was
defined previously, so we are stuck with the freak call order
for the target cpu realizefn.
After this patch the target/xxx/cpu.c realizefn body is now:
void mycpu_realizefn(DeviceState *dev, Error **errp)
{
/* ... */
cpu_exec_realizefn(CPU_STATE(dev), errp);
/* ... anything that needs done pre-qemu_vcpu_init */
xcc->parent_realize(dev, errp); /* does qemu_vcpu_init */
/* ... anything that needs to be done after qemu_vcpu_init */
}
Note: better do some testing for all targets for this.
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
hw/core/cpu.c | 2 ++
target/alpha/cpu.c | 5 +----
target/arm/cpu.c | 4 +---
target/avr/cpu.c | 3 +--
target/cris/cpu.c | 2 --
target/hppa/cpu.c | 1 -
target/i386/cpu.c | 5 +----
target/lm32/cpu.c | 3 ---
target/m68k/cpu.c | 2 --
target/microblaze/cpu.c | 9 +++------
target/mips/cpu.c | 2 --
target/moxie/cpu.c | 4 +---
target/nios2/cpu.c | 4 +---
target/openrisc/cpu.c | 4 +---
target/ppc/translate_init.c.inc | 5 ++---
target/riscv/cpu.c | 8 +++-----
target/rx/cpu.c | 8 +++-----
target/s390x/cpu.c | 3 +--
target/sh4/cpu.c | 2 --
target/sparc/cpu.c | 4 +---
target/tilegx/cpu.c | 2 --
target/tricore/cpu.c | 2 --
target/unicore32/cpu.c | 6 +-----
target/xtensa/cpu.c | 2 --
24 files changed, 23 insertions(+), 69 deletions(-)
diff --git a/hw/core/cpu.c b/hw/core/cpu.c
index 1f04aab16b..f41c009e6c 100644
--- a/hw/core/cpu.c
+++ b/hw/core/cpu.c
@@ -318,6 +318,8 @@ static void cpu_common_realizefn(DeviceState *dev, Error **errp)
CPUState *cpu = CPU(dev);
Object *machine = qdev_get_machine();
+ qemu_init_vcpu(cpu);
+
/* qdev_get_machine() can return something that's not TYPE_MACHINE
* if this is one of the user-only emulators; in that case there's
* no need to check the ignore_memory_transaction_failures board flag.
diff --git a/target/alpha/cpu.c b/target/alpha/cpu.c
index 0b5f5a78e3..d43abbbac1 100644
--- a/target/alpha/cpu.c
+++ b/target/alpha/cpu.c
@@ -56,18 +56,15 @@ static void alpha_cpu_disas_set_info(CPUState *cpu, disassemble_info *info)
static void alpha_cpu_realizefn(DeviceState *dev, Error **errp)
{
- CPUState *cs = CPU(dev);
AlphaCPUClass *acc = ALPHA_CPU_GET_CLASS(dev);
Error *local_err = NULL;
- cpu_exec_realizefn(cs, &local_err);
+ cpu_exec_realizefn(CPU(dev), &local_err);
if (local_err != NULL) {
error_propagate(errp, local_err);
return;
}
- qemu_init_vcpu(cs);
-
acc->parent_realize(dev, errp);
}
diff --git a/target/arm/cpu.c b/target/arm/cpu.c
index 9a8c1dc2e1..6366c6af59 100644
--- a/target/arm/cpu.c
+++ b/target/arm/cpu.c
@@ -1859,10 +1859,8 @@ static void arm_cpu_realizefn(DeviceState *dev, Error **errp)
}
}
- qemu_init_vcpu(cs);
- cpu_reset(cs);
-
acc->parent_realize(dev, errp);
+ cpu_reset(cs);
}
static ObjectClass *arm_cpu_class_by_name(const char *cpu_model)
diff --git a/target/avr/cpu.c b/target/avr/cpu.c
index 5a5ae68444..43f91f650d 100644
--- a/target/avr/cpu.c
+++ b/target/avr/cpu.c
@@ -97,10 +97,9 @@ static void avr_cpu_realizefn(DeviceState *dev, Error **errp)
error_propagate(errp, local_err);
return;
}
- qemu_init_vcpu(cs);
- cpu_reset(cs);
mcc->parent_realize(dev, errp);
+ cpu_reset(cs);
}
static void avr_cpu_set_int(void *opaque, int irq, int level)
diff --git a/target/cris/cpu.c b/target/cris/cpu.c
index b65743e8ca..42a497d9fe 100644
--- a/target/cris/cpu.c
+++ b/target/cris/cpu.c
@@ -135,8 +135,6 @@ static void cris_cpu_realizefn(DeviceState *dev, Error **errp)
}
cpu_reset(cs);
- qemu_init_vcpu(cs);
-
ccc->parent_realize(dev, errp);
}
diff --git a/target/hppa/cpu.c b/target/hppa/cpu.c
index 0985b3661f..979714c62e 100644
--- a/target/hppa/cpu.c
+++ b/target/hppa/cpu.c
@@ -101,7 +101,6 @@ static void hppa_cpu_realizefn(DeviceState *dev, Error **errp)
return;
}
- qemu_init_vcpu(cs);
acc->parent_realize(dev, errp);
#ifndef CONFIG_USER_ONLY
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 25998c0122..aabb8edfe0 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -6419,8 +6419,7 @@ static void x86_cpu_realizefn(DeviceState *dev, Error **errp)
mce_init(cpu);
- qemu_init_vcpu(cs);
-
+ xcc->parent_realize(dev, &local_err);
/*
* Most Intel and certain AMD CPUs support hyperthreading. Even though QEMU
* fixes this issue by adjusting CPUID_0000_0001_EBX and CPUID_8000_0008_ECX
@@ -6447,8 +6446,6 @@ static void x86_cpu_realizefn(DeviceState *dev, Error **errp)
}
cpu_reset(cs);
- xcc->parent_realize(dev, &local_err);
-
out:
if (local_err != NULL) {
error_propagate(errp, local_err);
diff --git a/target/lm32/cpu.c b/target/lm32/cpu.c
index fb3761b749..99196c22a4 100644
--- a/target/lm32/cpu.c
+++ b/target/lm32/cpu.c
@@ -133,9 +133,6 @@ static void lm32_cpu_realizefn(DeviceState *dev, Error **errp)
}
cpu_reset(cs);
-
- qemu_init_vcpu(cs);
-
lcc->parent_realize(dev, errp);
}
diff --git a/target/m68k/cpu.c b/target/m68k/cpu.c
index 3f60c99865..41d3b2c11e 100644
--- a/target/m68k/cpu.c
+++ b/target/m68k/cpu.c
@@ -248,8 +248,6 @@ static void m68k_cpu_realizefn(DeviceState *dev, Error **errp)
m68k_cpu_init_gdb(cpu);
cpu_reset(cs);
- qemu_init_vcpu(cs);
-
mcc->parent_realize(dev, errp);
}
diff --git a/target/microblaze/cpu.c b/target/microblaze/cpu.c
index 395f4a300f..6e068b2ae0 100644
--- a/target/microblaze/cpu.c
+++ b/target/microblaze/cpu.c
@@ -145,15 +145,14 @@ static void mb_disas_set_info(CPUState *cpu, disassemble_info *info)
static void mb_cpu_realizefn(DeviceState *dev, Error **errp)
{
- CPUState *cs = CPU(dev);
MicroBlazeCPUClass *mcc = MICROBLAZE_CPU_GET_CLASS(dev);
- MicroBlazeCPU *cpu = MICROBLAZE_CPU(cs);
+ MicroBlazeCPU *cpu = MICROBLAZE_CPU(dev);
uint8_t version_code = 0;
const char *version;
int i = 0;
Error *local_err = NULL;
- cpu_exec_realizefn(cs, &local_err);
+ cpu_exec_realizefn(CPU(dev), &local_err);
if (local_err != NULL) {
error_propagate(errp, local_err);
return;
@@ -165,7 +164,7 @@ static void mb_cpu_realizefn(DeviceState *dev, Error **errp)
return;
}
- qemu_init_vcpu(cs);
+ mcc->parent_realize(dev, errp);
version = cpu->cfg.version ? cpu->cfg.version : DEFAULT_CPU_VERSION;
for (i = 0; mb_cpu_lookup[i].name && version; i++) {
@@ -231,8 +230,6 @@ static void mb_cpu_realizefn(DeviceState *dev, Error **errp)
cpu->cfg.mmu_tlb_access = 3;
cpu->cfg.mmu_zones = 16;
cpu->cfg.addr_mask = MAKE_64BIT_MASK(0, cpu->cfg.addr_size);
-
- mcc->parent_realize(dev, errp);
}
static void mb_cpu_initfn(Object *obj)
diff --git a/target/mips/cpu.c b/target/mips/cpu.c
index 236d0d707b..68cc62c08e 100644
--- a/target/mips/cpu.c
+++ b/target/mips/cpu.c
@@ -182,8 +182,6 @@ static void mips_cpu_realizefn(DeviceState *dev, Error **errp)
cpu_mips_realize_env(&cpu->env);
cpu_reset(cs);
- qemu_init_vcpu(cs);
-
mcc->parent_realize(dev, errp);
}
diff --git a/target/moxie/cpu.c b/target/moxie/cpu.c
index 36bef4d357..4dfd2936af 100644
--- a/target/moxie/cpu.c
+++ b/target/moxie/cpu.c
@@ -66,10 +66,8 @@ static void moxie_cpu_realizefn(DeviceState *dev, Error **errp)
return;
}
- qemu_init_vcpu(cs);
- cpu_reset(cs);
-
mcc->parent_realize(dev, errp);
+ cpu_reset(cs);
}
static void moxie_cpu_initfn(Object *obj)
diff --git a/target/nios2/cpu.c b/target/nios2/cpu.c
index 059eea8c94..46440d6ad8 100644
--- a/target/nios2/cpu.c
+++ b/target/nios2/cpu.c
@@ -92,10 +92,8 @@ static void nios2_cpu_realizefn(DeviceState *dev, Error **errp)
return;
}
- qemu_init_vcpu(cs);
- cpu_reset(cs);
-
ncc->parent_realize(dev, errp);
+ cpu_reset(cs);
}
static bool nios2_cpu_exec_interrupt(CPUState *cs, int interrupt_request)
diff --git a/target/openrisc/cpu.c b/target/openrisc/cpu.c
index 27105c5c09..e16b66fd14 100644
--- a/target/openrisc/cpu.c
+++ b/target/openrisc/cpu.c
@@ -77,10 +77,8 @@ static void openrisc_cpu_realizefn(DeviceState *dev, Error **errp)
return;
}
- qemu_init_vcpu(cs);
- cpu_reset(cs);
-
occ->parent_realize(dev, errp);
+ cpu_reset(cs);
}
static void openrisc_cpu_initfn(Object *obj)
diff --git a/target/ppc/translate_init.c.inc b/target/ppc/translate_init.c.inc
index 18ce389947..637fb389db 100644
--- a/target/ppc/translate_init.c.inc
+++ b/target/ppc/translate_init.c.inc
@@ -10093,7 +10093,7 @@ static int ppc_fixup_cpu(PowerPCCPU *cpu)
return 0;
}
-static void ppc_cpu_realize(DeviceState *dev, Error **errp)
+static void ppc_cpu_realizefn(DeviceState *dev, Error **errp)
{
CPUState *cs = CPU(dev);
PowerPCCPU *cpu = POWERPC_CPU(dev);
@@ -10143,7 +10143,6 @@ static void ppc_cpu_realize(DeviceState *dev, Error **errp)
gdb_register_coprocessor(cs, gdb_get_spr_reg, gdb_set_spr_reg,
pcc->gdb_num_sprs, "power-spr.xml", 0);
#endif
- qemu_init_vcpu(cs);
pcc->parent_realize(dev, errp);
@@ -10894,7 +10893,7 @@ static void ppc_cpu_class_init(ObjectClass *oc, void *data)
CPUClass *cc = CPU_CLASS(oc);
DeviceClass *dc = DEVICE_CLASS(oc);
- device_class_set_parent_realize(dc, ppc_cpu_realize,
+ device_class_set_parent_realize(dc, ppc_cpu_realizefn,
&pcc->parent_realize);
device_class_set_parent_unrealize(dc, ppc_cpu_unrealize,
&pcc->parent_unrealize);
diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index e5626862c2..6b4eadfdeb 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -341,7 +341,7 @@ static void riscv_cpu_disas_set_info(CPUState *s, disassemble_info *info)
#endif
}
-static void riscv_cpu_realize(DeviceState *dev, Error **errp)
+static void riscv_cpu_realizefn(DeviceState *dev, Error **errp)
{
CPUState *cs = CPU(dev);
RISCVCPU *cpu = RISCV_CPU(dev);
@@ -486,10 +486,8 @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp)
riscv_cpu_register_gdb_regs_for_features(cs);
- qemu_init_vcpu(cs);
- cpu_reset(cs);
-
mcc->parent_realize(dev, errp);
+ cpu_reset(cs);
}
static void riscv_cpu_init(Object *obj)
@@ -532,7 +530,7 @@ static void riscv_cpu_class_init(ObjectClass *c, void *data)
CPUClass *cc = CPU_CLASS(c);
DeviceClass *dc = DEVICE_CLASS(c);
- device_class_set_parent_realize(dc, riscv_cpu_realize,
+ device_class_set_parent_realize(dc, riscv_cpu_realizefn,
&mcc->parent_realize);
device_class_set_parent_reset(dc, riscv_cpu_reset, &mcc->parent_reset);
diff --git a/target/rx/cpu.c b/target/rx/cpu.c
index dc7d1c3c57..75ee2866dc 100644
--- a/target/rx/cpu.c
+++ b/target/rx/cpu.c
@@ -105,7 +105,7 @@ static ObjectClass *rx_cpu_class_by_name(const char *cpu_model)
return oc;
}
-static void rx_cpu_realize(DeviceState *dev, Error **errp)
+static void rx_cpu_realizefn(DeviceState *dev, Error **errp)
{
CPUState *cs = CPU(dev);
RXCPUClass *rcc = RX_CPU_GET_CLASS(dev);
@@ -117,10 +117,8 @@ static void rx_cpu_realize(DeviceState *dev, Error **errp)
return;
}
- qemu_init_vcpu(cs);
- cpu_reset(cs);
-
rcc->parent_realize(dev, errp);
+ cpu_reset(cs);
}
static void rx_cpu_set_irq(void *opaque, int no, int request)
@@ -178,7 +176,7 @@ static void rx_cpu_class_init(ObjectClass *klass, void *data)
CPUClass *cc = CPU_CLASS(klass);
RXCPUClass *rcc = RX_CPU_CLASS(klass);
- device_class_set_parent_realize(dc, rx_cpu_realize,
+ device_class_set_parent_realize(dc, rx_cpu_realizefn,
&rcc->parent_realize);
device_class_set_parent_reset(dc, rx_cpu_reset,
&rcc->parent_reset);
diff --git a/target/s390x/cpu.c b/target/s390x/cpu.c
index 86f654fd6b..d67ae110af 100644
--- a/target/s390x/cpu.c
+++ b/target/s390x/cpu.c
@@ -232,8 +232,8 @@ static void s390_cpu_realizefn(DeviceState *dev, Error **errp)
qemu_register_reset(s390_cpu_machine_reset_cb, cpu);
#endif
s390_cpu_gdb_init(cs);
- qemu_init_vcpu(cs);
+ scc->parent_realize(dev, &err);
/*
* KVM requires the initial CPU reset ioctl to be executed on the target
* CPU thread. CPU hotplug under single-threaded TCG will not work with
@@ -246,7 +246,6 @@ static void s390_cpu_realizefn(DeviceState *dev, Error **errp)
cpu_reset(cs);
}
- scc->parent_realize(dev, &err);
out:
error_propagate(errp, err);
}
diff --git a/target/sh4/cpu.c b/target/sh4/cpu.c
index fbd5f42675..b9380cce83 100644
--- a/target/sh4/cpu.c
+++ b/target/sh4/cpu.c
@@ -185,8 +185,6 @@ static void superh_cpu_realizefn(DeviceState *dev, Error **errp)
}
cpu_reset(cs);
- qemu_init_vcpu(cs);
-
scc->parent_realize(dev, errp);
}
diff --git a/target/sparc/cpu.c b/target/sparc/cpu.c
index 3b53ef2390..0b604372d3 100644
--- a/target/sparc/cpu.c
+++ b/target/sparc/cpu.c
@@ -738,9 +738,9 @@ static void sparc_cpu_realizefn(DeviceState *dev, Error **errp)
{
CPUState *cs = CPU(dev);
SPARCCPUClass *scc = SPARC_CPU_GET_CLASS(dev);
- Error *local_err = NULL;
SPARCCPU *cpu = SPARC_CPU(dev);
CPUSPARCState *env = &cpu->env;
+ Error *local_err = NULL;
#if defined(CONFIG_USER_ONLY)
if ((env->def.features & CPU_FEATURE_FLOAT)) {
@@ -768,8 +768,6 @@ static void sparc_cpu_realizefn(DeviceState *dev, Error **errp)
return;
}
- qemu_init_vcpu(cs);
-
scc->parent_realize(dev, errp);
}
diff --git a/target/tilegx/cpu.c b/target/tilegx/cpu.c
index 7d4ead4ef1..01db5bdbcf 100644
--- a/target/tilegx/cpu.c
+++ b/target/tilegx/cpu.c
@@ -93,8 +93,6 @@ static void tilegx_cpu_realizefn(DeviceState *dev, Error **errp)
}
cpu_reset(cs);
- qemu_init_vcpu(cs);
-
tcc->parent_realize(dev, errp);
}
diff --git a/target/tricore/cpu.c b/target/tricore/cpu.c
index 89a14f81d7..ac57239f87 100644
--- a/target/tricore/cpu.c
+++ b/target/tricore/cpu.c
@@ -93,8 +93,6 @@ static void tricore_cpu_realizefn(DeviceState *dev, Error **errp)
set_feature(env, TRICORE_FEATURE_13);
}
cpu_reset(cs);
- qemu_init_vcpu(cs);
-
tcc->parent_realize(dev, errp);
}
diff --git a/target/unicore32/cpu.c b/target/unicore32/cpu.c
index e27ffc571a..fe4ae21700 100644
--- a/target/unicore32/cpu.c
+++ b/target/unicore32/cpu.c
@@ -84,18 +84,14 @@ static void uc32_any_cpu_initfn(Object *obj)
static void uc32_cpu_realizefn(DeviceState *dev, Error **errp)
{
- CPUState *cs = CPU(dev);
UniCore32CPUClass *ucc = UNICORE32_CPU_GET_CLASS(dev);
Error *local_err = NULL;
- cpu_exec_realizefn(cs, &local_err);
+ cpu_exec_realizefn(CPU(dev), &local_err);
if (local_err != NULL) {
error_propagate(errp, local_err);
return;
}
-
- qemu_init_vcpu(cs);
-
ucc->parent_realize(dev, errp);
}
diff --git a/target/xtensa/cpu.c b/target/xtensa/cpu.c
index 4b6381569f..7387d90cc6 100644
--- a/target/xtensa/cpu.c
+++ b/target/xtensa/cpu.c
@@ -153,8 +153,6 @@ static void xtensa_cpu_realizefn(DeviceState *dev, Error **errp)
cs->gdb_num_regs = xcc->config->gdb_regmap.num_regs;
- qemu_init_vcpu(cs);
-
xcc->parent_realize(dev, errp);
}
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* [RFC v9 32/32] cpu: introduce cpu_accel_instance_init
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (30 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 31/32] hw/core/cpu: call qemu_init_vcpu in cpu_common_realizefn Claudio Fontana
@ 2020-12-08 19:48 ` Claudio Fontana
2020-12-08 20:00 ` [RFC v9 00/22] i386 cleanup Philippe Mathieu-Daudé
` (2 subsequent siblings)
34 siblings, 0 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 19:48 UTC (permalink / raw)
To: Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy, Philippe Mathieu-Daudé
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Claudio Fontana, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota, Colin Xu
centralize the calls to cpu->accel_cpu_interface
Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
hw/core/cpu.c | 9 +++++++++
include/hw/core/cpu.h | 6 ++++++
target/i386/cpu.c | 9 ++-------
3 files changed, 17 insertions(+), 7 deletions(-)
diff --git a/hw/core/cpu.c b/hw/core/cpu.c
index f41c009e6c..873cf5e4ef 100644
--- a/hw/core/cpu.c
+++ b/hw/core/cpu.c
@@ -242,6 +242,15 @@ void cpu_reset(CPUState *cpu)
trace_guest_cpu_reset(cpu);
}
+void cpu_accel_instance_init(CPUState *cpu)
+{
+ CPUClass *cc = CPU_GET_CLASS(cpu);
+
+ if (cc->accel_cpu_interface) {
+ cc->accel_cpu_interface->cpu_instance_init(cpu);
+ }
+}
+
static void cpu_common_reset(DeviceState *dev)
{
CPUState *cpu = CPU(dev);
diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
index 97e1dd8279..cc05c8fc96 100644
--- a/include/hw/core/cpu.h
+++ b/include/hw/core/cpu.h
@@ -664,6 +664,12 @@ void cpu_list_remove(CPUState *cpu);
*/
void cpu_reset(CPUState *cpu);
+/**
+ * cpu_accel_instance_init:
+ * @cpu: The CPU that needs to do accel-specific object initializations.
+ */
+void cpu_accel_instance_init(CPUState *cpu);
+
/**
* cpu_class_by_name:
* @typename: The CPU base type.
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index aabb8edfe0..e2f16a1f37 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -28,7 +28,6 @@
#include "sysemu/kvm.h"
#include "sysemu/reset.h"
#include "sysemu/hvf.h"
-#include "hw/core/accel-cpu.h"
#include "sysemu/xen.h"
#include "kvm/kvm_i386.h"
#include "sev_i386.h"
@@ -6621,8 +6620,6 @@ static void x86_cpu_initfn(Object *obj)
{
X86CPU *cpu = X86_CPU(obj);
X86CPUClass *xcc = X86_CPU_GET_CLASS(obj);
- CPUClass *cc = CPU_CLASS(xcc);
-
CPUX86State *env = &cpu->env;
FeatureWord w;
@@ -6680,10 +6677,8 @@ static void x86_cpu_initfn(Object *obj)
x86_cpu_load_model(cpu, xcc->model);
}
- /* if required, do the accelerator-specific cpu initialization */
- if (cc->accel_cpu_interface) {
- cc->accel_cpu_interface->cpu_instance_init(CPU(obj));
- }
+ /* if required, do accelerator-specific cpu initializations */
+ cpu_accel_instance_init(CPU(obj));
}
static int64_t x86_cpu_get_arch_id(CPUState *cs)
--
2.26.2
^ permalink raw reply related [flat|nested] 87+ messages in thread
* Re: [RFC v9 00/22] i386 cleanup
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (31 preceding siblings ...)
2020-12-08 19:48 ` [RFC v9 32/32] cpu: introduce cpu_accel_instance_init Claudio Fontana
@ 2020-12-08 20:00 ` Philippe Mathieu-Daudé
2020-12-08 22:15 ` Claudio Fontana
2020-12-08 22:00 ` no-reply
2020-12-09 12:58 ` Alex Bennée
34 siblings, 1 reply; 87+ messages in thread
From: Philippe Mathieu-Daudé @ 2020-12-08 20:00 UTC (permalink / raw)
To: Claudio Fontana, Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy
Cc: Laurent Vivier, Peter Maydell, Eduardo Habkost, Paul Durrant,
Jason Wang, Marcelo Tosatti, qemu-devel, Peter Xu,
Dario Faggioli, Cameron Esfahani, haxm-team, Colin Xu,
Anthony Perard, Bruce Rogers, Olaf Hering, Emilio G . Cota
On 12/8/20 8:48 PM, Claudio Fontana wrote:
> v8 -> v9: move additional methods to CPUClass->tcg_ops
>
> do_unaligned_access, transaction_failed and do_interrupt.
>
> do_interrupt is a bit tricky, as the same code is reused
> (albeit not usually directly) for KVM under certain odd conditions.
>
> Change arm, as the only user of do_interrupt callback for KVM,
> to instead call the target function directly arm_do_interrupt.
Please use scripts/git.orderfile for such refactors please ;)
...
> MAINTAINERS | 19 +-
> accel/accel-common.c | 105 +++++
> accel/{accel.c => accel-softmmu.c} | 60 +--
> accel/accel-softmmu.h | 15 +
> accel/accel-user.c | 24 ++
> accel/kvm/kvm-all.c | 2 -
> accel/kvm/kvm-cpus.c | 26 +-
> accel/kvm/kvm-cpus.h | 2 -
> accel/meson.build | 4 +-
> accel/qtest/qtest.c | 25 +-
> accel/tcg/cpu-exec.c | 70 +++-
> accel/tcg/cputlb.c | 6 +-
> accel/tcg/meson.build | 9 +-
> accel/tcg/tcg-all.c | 14 +-
> accel/tcg/tcg-cpus-icount.c | 138 +++++++
> accel/tcg/tcg-cpus-icount.h | 19 +
> accel/tcg/tcg-cpus-mttcg.c | 134 +++++++
> accel/tcg/tcg-cpus-mttcg.h | 19 +
> accel/tcg/tcg-cpus-rr.c | 298 ++++++++++++++
> accel/tcg/tcg-cpus-rr.h | 21 +
> accel/tcg/tcg-cpus.c | 539 +++-----------------------
> accel/tcg/tcg-cpus.h | 8 +-
> accel/tcg/user-exec.c | 6 +-
> accel/xen/xen-all.c | 24 +-
> bsd-user/main.c | 11 +-
> cpu.c | 71 ++--
> hw/core/cpu.c | 30 +-
> hw/i386/fw_cfg.c | 2 +-
> hw/i386/intel_iommu.c | 2 +-
> hw/i386/kvm/apic.c | 2 +-
> hw/i386/kvm/clock.c | 2 +-
> hw/i386/microvm.c | 2 +-
> hw/i386/pc.c | 2 +-
> hw/i386/pc_piix.c | 1 +
> hw/i386/x86.c | 2 +-
> hw/mips/jazz.c | 4 +-
> include/hw/boards.h | 2 +-
> include/hw/core/accel-cpu.h | 25 ++
> include/hw/core/cpu.h | 92 ++---
> include/hw/core/tcg-cpu-ops.h | 75 ++++
> include/{sysemu => qemu}/accel.h | 16 +-
> include/sysemu/accel-ops.h | 45 +++
> include/sysemu/cpus.h | 26 +-
> include/sysemu/hvf.h | 2 +-
> include/sysemu/kvm.h | 2 +-
> include/sysemu/kvm_int.h | 2 +-
> linux-user/main.c | 7 +-
> meson.build | 1 +
> softmmu/cpus.c | 12 +-
> softmmu/icount.c | 2 +-
> softmmu/memory.c | 2 +-
> softmmu/qtest.c | 2 +-
> softmmu/vl.c | 8 +-
> target/alpha/cpu.c | 18 +-
> target/arm/cpu.c | 26 +-
> target/arm/cpu64.c | 5 +-
> target/arm/cpu_tcg.c | 8 +-
> target/arm/helper.c | 4 +
> target/arm/kvm64.c | 4 +-
> target/avr/cpu.c | 13 +-
> target/avr/helper.c | 4 +-
> target/cris/cpu.c | 30 +-
> target/cris/helper.c | 4 +-
> target/hppa/cpu.c | 15 +-
> target/i386/cpu-dump.c | 537 +++++++++++++++++++++++++
> target/i386/cpu.c | 418 ++------------------
> target/i386/cpu.h | 120 +-----
> target/i386/{ => hax}/hax-all.c | 5 +-
> target/i386/{ => hax}/hax-cpus.c | 29 +-
> target/i386/{ => hax}/hax-cpus.h | 2 -
> target/i386/{ => hax}/hax-i386.h | 6 +-
> target/i386/{ => hax}/hax-interface.h | 0
> target/i386/{ => hax}/hax-mem.c | 0
> target/i386/{ => hax}/hax-posix.c | 0
> target/i386/{ => hax}/hax-posix.h | 0
> target/i386/{ => hax}/hax-windows.c | 0
> target/i386/{ => hax}/hax-windows.h | 0
> target/i386/hax/meson.build | 7 +
> target/i386/helper-tcg.h | 112 ++++++
> target/i386/helper.c | 539 +-------------------------
> target/i386/host-cpu.c | 198 ++++++++++
> target/i386/host-cpu.h | 19 +
> target/i386/hvf/cpu.c | 65 ++++
> target/i386/hvf/hvf-cpus.c | 27 +-
> target/i386/hvf/hvf-cpus.h | 2 -
> target/i386/hvf/hvf-i386.h | 2 +-
> target/i386/hvf/hvf.c | 3 +-
> target/i386/hvf/meson.build | 1 +
> target/i386/hvf/x86_task.c | 2 +-
> target/i386/kvm/cpu.c | 148 +++++++
> target/i386/{ => kvm}/hyperv-proto.h | 0
> target/i386/{ => kvm}/hyperv-stub.c | 0
> target/i386/{ => kvm}/hyperv.c | 0
> target/i386/{ => kvm}/hyperv.h | 0
> target/i386/kvm/kvm-cpu.h | 41 ++
> target/i386/{ => kvm}/kvm-stub.c | 0
> target/i386/{ => kvm}/kvm.c | 3 +-
> target/i386/{ => kvm}/kvm_i386.h | 0
> target/i386/kvm/meson.build | 8 +
> target/i386/kvm/trace-events | 7 +
> target/i386/kvm/trace.h | 1 +
> target/i386/machine.c | 4 +-
> target/i386/meson.build | 39 +-
> target/i386/{ => tcg}/bpt_helper.c | 1 +
> target/i386/{ => tcg}/cc_helper.c | 1 +
> target/i386/tcg/cpu.c | 173 +++++++++
> target/i386/{ => tcg}/excp_helper.c | 1 +
> target/i386/{ => tcg}/fpu_helper.c | 33 +-
> target/i386/{ => tcg}/int_helper.c | 1 +
> target/i386/{ => tcg}/mem_helper.c | 1 +
> target/i386/tcg/meson.build | 14 +
> target/i386/{ => tcg}/misc_helper.c | 1 +
> target/i386/{ => tcg}/mpx_helper.c | 1 +
> target/i386/{ => tcg}/seg_helper.c | 1 +
> target/i386/{ => tcg}/smm_helper.c | 2 +
> target/i386/{ => tcg}/svm_helper.c | 1 +
> target/i386/{ => tcg}/tcg-stub.c | 0
> target/i386/{ => tcg}/translate.c | 1 +
> target/i386/trace-events | 6 -
> target/i386/whpx/meson.build | 4 +
> target/i386/{ => whpx}/whp-dispatch.h | 0
> target/i386/{ => whpx}/whpx-all.c | 4 +-
> target/i386/{ => whpx}/whpx-cpus.c | 29 +-
> target/i386/{ => whpx}/whpx-cpus.h | 2 -
> target/lm32/cpu.c | 13 +-
> target/m68k/cpu.c | 12 +-
> target/microblaze/cpu.c | 23 +-
> target/mips/cpu.c | 21 +-
> target/moxie/cpu.c | 10 +-
> target/nios2/cpu.c | 14 +-
> target/openrisc/cpu.c | 12 +-
> target/ppc/translate_init.c.inc | 23 +-
> target/riscv/cpu.c | 25 +-
> target/rx/cpu.c | 18 +-
> target/s390x/cpu.c | 15 +-
> target/sh4/cpu.c | 14 +-
> target/sparc/cpu.c | 18 +-
> target/tilegx/cpu.c | 10 +-
> target/tricore/cpu.c | 8 +-
> target/unicore32/cpu.c | 14 +-
> target/xtensa/cpu.c | 16 +-
> 141 files changed, 3007 insertions(+), 2051 deletions(-)
...
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 00/22] i386 cleanup
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (32 preceding siblings ...)
2020-12-08 20:00 ` [RFC v9 00/22] i386 cleanup Philippe Mathieu-Daudé
@ 2020-12-08 22:00 ` no-reply
2020-12-09 10:22 ` Alex Bennée
2020-12-09 12:58 ` Alex Bennée
34 siblings, 1 reply; 87+ messages in thread
From: no-reply @ 2020-12-08 22:00 UTC (permalink / raw)
To: cfontana
Cc: peter.maydell, paul, jasowang, qemu-devel, peterx, haxm-team,
colin.xu, ohering, sstabellini, brogers, cota, cfontana,
anthony.perard, philmd, lvivier, thuth, ehabkost,
richard.henderson, dirty, dfaggioli, r.bolshakov, sunilmut,
mtosatti, wenchao.wang, pbonzini
Patchew URL: https://patchew.org/QEMU/20201208194839.31305-1-cfontana@suse.de/
Hi,
This series seems to have some coding style problems. See output below for
more information:
Type: series
Message-id: 20201208194839.31305-1-cfontana@suse.de
Subject: [RFC v9 00/22] i386 cleanup
=== TEST SCRIPT BEGIN ===
#!/bin/bash
git rev-parse base > /dev/null || exit 0
git config --local diff.renamelimit 0
git config --local diff.renames True
git config --local diff.algorithm histogram
./scripts/checkpatch.pl --mailback base..
=== TEST SCRIPT END ===
Updating 3c8cf5a9c21ff8782164d1def7f44bd888713384
From https://github.com/patchew-project/qemu
* [new tag] patchew/20201208194839.31305-1-cfontana@suse.de -> patchew/20201208194839.31305-1-cfontana@suse.de
Switched to a new branch 'test'
1df468e cpu: introduce cpu_accel_instance_init
273e6bc hw/core/cpu: call qemu_init_vcpu in cpu_common_realizefn
6c8c6a4 cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn
c7ecc11 i386: split cpu accelerators from cpu.c, using AccelCPUClass
739ef2b accel: introduce AccelCPUClass extending CPUClass
706469e accel: replace struct CpusAccel with AccelOpsClass
c3aa222 accel: extend AccelState and AccelClass to user-mode
c45684a cpu: move do_unaligned_access to tcg_ops
bca539a cpu: move cc->transaction_failed to tcg_ops
74feabd cpu: move cc->do_interrupt to tcg_ops
f22d896 target/arm: do not use cc->do_interrupt for KVM directly
9a74429 cpu: Move debug_excp_handler to tcg_ops
bf5b24b cpu: Move tlb_fill to tcg_ops
481a06e cpu: Move cpu_exec_* to tcg_ops
af72917 cpu: Move synchronize_from_tb() to tcg_ops
f9b393d accel/tcg: split TCG-only code from cpu_exec_realizefn
2019e4f target/riscv: remove CONFIG_TCG, as it is always TCG
53b3f4e cpu: Introduce TCGCpuOperations struct
8a0c2e9 cpu: Remove unnecessary noop methods
043b89c tcg: Make CPUClass.debug_excp_handler optional
b271ab4 tcg: make CPUClass.cpu_exec_* optional
ac15b62 tcg: cpu_exec_{enter,exit} helpers
0560b09 i386: move TCG cpu class initialization out of helper.c
36ae5f1 i386: move cpu dump out of helper.c into cpu-dump.c
ef2d207 i386: move TCG accel files into tcg/
9d7a737 i386: hvf: remove stale MAINTAINERS entry for old hvf stubs
42deca7 i386: move hax accel files into hax/
7dfc2be i386: move whpx accel files into whpx/
4d4c885 i386: move kvm accel files into kvm/
d9966d5 accel/tcg: rename tcg-cpus functions to match module name
a8f0cbf accel/tcg: split tcg_start_vcpu_thread
5e46bf0 accel/tcg: split CpusAccel into three TCG variants
=== OUTPUT BEGIN ===
1/32 Checking commit 5e46bf019fb7 (accel/tcg: split CpusAccel into three TCG variants)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#57:
new file mode 100644
total: 0 errors, 1 warnings, 1156 lines checked
Patch 1/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
2/32 Checking commit a8f0cbf6b610 (accel/tcg: split tcg_start_vcpu_thread)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#103:
deleted file mode 100644
total: 0 errors, 1 warnings, 201 lines checked
Patch 2/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
3/32 Checking commit d9966d571c7e (accel/tcg: rename tcg-cpus functions to match module name)
4/32 Checking commit 4d4c885ff790 (i386: move kvm accel files into kvm/)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#167:
rename from target/i386/hyperv-proto.h
total: 0 errors, 1 warnings, 150 lines checked
Patch 4/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
5/32 Checking commit 7dfc2bea3c57 (i386: move whpx accel files into whpx/)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#50:
new file mode 100644
total: 0 errors, 1 warnings, 32 lines checked
Patch 5/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
6/32 Checking commit 42deca72cda4 (i386: move hax accel files into hax/)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#25:
rename from target/i386/hax-all.c
total: 0 errors, 1 warnings, 47 lines checked
Patch 6/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
7/32 Checking commit 9d7a7374d252 (i386: hvf: remove stale MAINTAINERS entry for old hvf stubs)
8/32 Checking commit ef2d207c5f81 (i386: move TCG accel files into tcg/)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#44:
rename from target/i386/bpt_helper.c
total: 0 errors, 1 warnings, 39 lines checked
Patch 8/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
9/32 Checking commit 36ae5f191d8c (i386: move cpu dump out of helper.c into cpu-dump.c)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#11:
new file mode 100644
ERROR: braces {} are necessary for all arms of this statement
#128: FILE: target/i386/cpu-dump.c:113:
+ if (!(env->hflags & HF_PE_MASK) || !(sc->flags & DESC_P_MASK))
[...]
WARNING: Block comments use a leading /* on a separate line
#172: FILE: target/i386/cpu-dump.c:157:
+/* ARRAY_SIZE check is not required because
WARNING: Block comments use a leading /* on a separate line
#211: FILE: target/i386/cpu-dump.c:196:
+/* ARRAY_SIZE check is not required because
ERROR: line over 90 characters
#373: FILE: target/i386/cpu-dump.c:358:
+ qemu_fprintf(f, "RAX=%016" PRIx64 " RBX=%016" PRIx64 " RCX=%016" PRIx64 " RDX=%016" PRIx64 "\n"
ERROR: line over 90 characters
#374: FILE: target/i386/cpu-dump.c:359:
+ "RSI=%016" PRIx64 " RDI=%016" PRIx64 " RBP=%016" PRIx64 " RSP=%016" PRIx64 "\n"
ERROR: line over 90 characters
#375: FILE: target/i386/cpu-dump.c:360:
+ "R8 =%016" PRIx64 " R9 =%016" PRIx64 " R10=%016" PRIx64 " R11=%016" PRIx64 "\n"
ERROR: line over 90 characters
#376: FILE: target/i386/cpu-dump.c:361:
+ "R12=%016" PRIx64 " R13=%016" PRIx64 " R14=%016" PRIx64 " R15=%016" PRIx64 "\n"
ERROR: line over 90 characters
#377: FILE: target/i386/cpu-dump.c:362:
+ "RIP=%016" PRIx64 " RFL=%08x [%c%c%c%c%c%c%c] CPL=%d II=%d A20=%d SMM=%d HLT=%d\n",
ERROR: space required before the open parenthesis '('
#436: FILE: target/i386/cpu-dump.c:421:
+ for(i = 0; i < 6; i++) {
WARNING: line over 80 characters
#448: FILE: target/i386/cpu-dump.c:433:
+ qemu_fprintf(f, "CR0=%08x CR2=%016" PRIx64 " CR3=%016" PRIx64 " CR4=%08x\n",
ERROR: space required before the open parenthesis '('
#453: FILE: target/i386/cpu-dump.c:438:
+ for(i = 0; i < 4; i++)
ERROR: braces {} are necessary for all arms of this statement
#453: FILE: target/i386/cpu-dump.c:438:
+ for(i = 0; i < 4; i++)
[...]
ERROR: space required before the open parenthesis '('
#469: FILE: target/i386/cpu-dump.c:454:
+ for(i = 0; i < 4; i++) {
ERROR: braces {} are necessary for all arms of this statement
#476: FILE: target/i386/cpu-dump.c:461:
+ if ((unsigned)env->cc_op < CC_OP_NB)
[...]
+ else
[...]
WARNING: line over 80 characters
#477: FILE: target/i386/cpu-dump.c:462:
+ snprintf(cc_op_name, sizeof(cc_op_name), "%s", cc_op_str[env->cc_op]);
ERROR: space required before the open parenthesis '('
#497: FILE: target/i386/cpu-dump.c:482:
+ for(i = 0; i < 8; i++) {
ERROR: spaces required around that '=' (ctx:VxV)
#507: FILE: target/i386/cpu-dump.c:492:
+ for(i=0;i<8;i++) {
^
ERROR: space required after that ';' (ctx:VxV)
#507: FILE: target/i386/cpu-dump.c:492:
+ for(i=0;i<8;i++) {
^
ERROR: spaces required around that '<' (ctx:VxV)
#507: FILE: target/i386/cpu-dump.c:492:
+ for(i=0;i<8;i++) {
^
ERROR: space required after that ';' (ctx:VxV)
#507: FILE: target/i386/cpu-dump.c:492:
+ for(i=0;i<8;i++) {
^
ERROR: space required before the open parenthesis '('
#507: FILE: target/i386/cpu-dump.c:492:
+ for(i=0;i<8;i++) {
ERROR: braces {} are necessary for all arms of this statement
#512: FILE: target/i386/cpu-dump.c:497:
+ if ((i & 1) == 1)
[...]
+ else
[...]
ERROR: braces {} are necessary for all arms of this statement
#517: FILE: target/i386/cpu-dump.c:502:
+ if (env->hflags & HF_CS64_MASK)
[...]
+ else
[...]
ERROR: spaces required around that '=' (ctx:VxV)
#521: FILE: target/i386/cpu-dump.c:506:
+ for(i=0;i<nb;i++) {
^
ERROR: space required after that ';' (ctx:VxV)
#521: FILE: target/i386/cpu-dump.c:506:
+ for(i=0;i<nb;i++) {
^
ERROR: spaces required around that '<' (ctx:VxV)
#521: FILE: target/i386/cpu-dump.c:506:
+ for(i=0;i<nb;i++) {
^
ERROR: space required after that ';' (ctx:VxV)
#521: FILE: target/i386/cpu-dump.c:506:
+ for(i=0;i<nb;i++) {
^
ERROR: space required before the open parenthesis '('
#521: FILE: target/i386/cpu-dump.c:506:
+ for(i=0;i<nb;i++) {
ERROR: braces {} are necessary for all arms of this statement
#528: FILE: target/i386/cpu-dump.c:513:
+ if ((i & 1) == 1)
[...]
+ else
[...]
total: 25 errors, 5 warnings, 1077 lines checked
Patch 9/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
10/32 Checking commit 0560b09d8481 (i386: move TCG cpu class initialization out of helper.c)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#282:
new file mode 100644
total: 0 errors, 1 warnings, 610 lines checked
Patch 10/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
11/32 Checking commit ac15b624a91f (tcg: cpu_exec_{enter,exit} helpers)
12/32 Checking commit b271ab48c45b (tcg: make CPUClass.cpu_exec_* optional)
13/32 Checking commit 043b89c9c315 (tcg: Make CPUClass.debug_excp_handler optional)
14/32 Checking commit 8a0c2e9a0e2c (cpu: Remove unnecessary noop methods)
15/32 Checking commit 53b3f4e5b390 (cpu: Introduce TCGCpuOperations struct)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#92:
new file mode 100644
total: 0 errors, 1 warnings, 301 lines checked
Patch 15/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
16/32 Checking commit 2019e4f802bb (target/riscv: remove CONFIG_TCG, as it is always TCG)
17/32 Checking commit f9b393d26c70 (accel/tcg: split TCG-only code from cpu_exec_realizefn)
18/32 Checking commit af729170f4b8 (cpu: Move synchronize_from_tb() to tcg_ops)
19/32 Checking commit 481a06e3b020 (cpu: Move cpu_exec_* to tcg_ops)
20/32 Checking commit bf5b24bcd22f (cpu: Move tlb_fill to tcg_ops)
WARNING: line over 80 characters
#22: FILE: accel/tcg/cputlb.c:1289:
+ ok = cc->tcg_ops.tlb_fill(cpu, addr, size, access_type, mmu_idx, false, retaddr);
total: 0 errors, 1 warnings, 259 lines checked
Patch 20/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
21/32 Checking commit 9a74429f2029 (cpu: Move debug_excp_handler to tcg_ops)
22/32 Checking commit f22d896967f8 (target/arm: do not use cc->do_interrupt for KVM directly)
23/32 Checking commit 74feabd5bae5 (cpu: move cc->do_interrupt to tcg_ops)
24/32 Checking commit bca539a6251e (cpu: move cc->transaction_failed to tcg_ops)
WARNING: line over 80 characters
#67: FILE: include/hw/core/cpu.h:857:
+ cc->tcg_ops.do_transaction_failed(cpu, physaddr, addr, size, access_type,
total: 0 errors, 1 warnings, 144 lines checked
Patch 24/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
25/32 Checking commit c45684ac5ad9 (cpu: move do_unaligned_access to tcg_ops)
26/32 Checking commit c3aa2224bd71 (accel: extend AccelState and AccelClass to user-mode)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#24:
new file mode 100644
total: 0 errors, 1 warnings, 359 lines checked
Patch 26/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
27/32 Checking commit 706469ebcc15 (accel: replace struct CpusAccel with AccelOpsClass)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#115:
new file mode 100644
total: 0 errors, 1 warnings, 691 lines checked
Patch 27/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
28/32 Checking commit 739ef2bd13aa (accel: introduce AccelCPUClass extending CPUClass)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#99:
new file mode 100644
total: 0 errors, 1 warnings, 123 lines checked
Patch 28/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
29/32 Checking commit c7ecc117664f (i386: split cpu accelerators from cpu.c, using AccelCPUClass)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#664:
new file mode 100644
WARNING: line over 80 characters
#1435: FILE: target/i386/tcg/cpu.c:104:
+ memory_region_add_subregion_overlap(cpu->cpu_as_root, 0, cpu->cpu_as_mem, 0);
total: 0 errors, 2 warnings, 1296 lines checked
Patch 29/32 has style problems, please review. If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
30/32 Checking commit 6c8c6a4d84dd (cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn)
31/32 Checking commit 273e6bca6404 (hw/core/cpu: call qemu_init_vcpu in cpu_common_realizefn)
32/32 Checking commit 1df468eafd63 (cpu: introduce cpu_accel_instance_init)
=== OUTPUT END ===
Test command exited with code: 1
The full log is available at
http://patchew.org/logs/20201208194839.31305-1-cfontana@suse.de/testing.checkpatch/?type=message.
---
Email generated automatically by Patchew [https://patchew.org/].
Please send your feedback to patchew-devel@redhat.com
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 16/32] target/riscv: remove CONFIG_TCG, as it is always TCG
2020-12-08 19:48 ` [RFC v9 16/32] target/riscv: remove CONFIG_TCG, as it is always TCG Claudio Fontana
@ 2020-12-08 22:10 ` Alistair Francis
0 siblings, 0 replies; 87+ messages in thread
From: Alistair Francis @ 2020-12-08 22:10 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel@nongnu.org Developers,
Peter Xu, haxm-team, Colin Xu, Olaf Hering, Stefano Stabellini,
Bruce Rogers, Emilio G . Cota, Anthony Perard,
Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
On Tue, Dec 8, 2020 at 12:52 PM Claudio Fontana <cfontana@suse.de> wrote:
>
> for now only TCG is allowed as an accelerator for riscv,
> so remove the CONFIG_TCG use.
>
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Alistair
> ---
> target/riscv/cpu.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
> index a52e0ce466..27dd1645c9 100644
> --- a/target/riscv/cpu.c
> +++ b/target/riscv/cpu.c
> @@ -561,10 +561,9 @@ static void riscv_cpu_class_init(ObjectClass *c, void *data)
> /* For now, mark unmigratable: */
> cc->vmsd = &vmstate_riscv_cpu;
> #endif
> -#ifdef CONFIG_TCG
> cc->tcg_ops.initialize = riscv_translate_init;
> cc->tlb_fill = riscv_cpu_tlb_fill;
> -#endif
> +
> device_class_set_props(dc, riscv_cpu_properties);
> }
>
> --
> 2.26.2
>
>
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 00/22] i386 cleanup
2020-12-08 20:00 ` [RFC v9 00/22] i386 cleanup Philippe Mathieu-Daudé
@ 2020-12-08 22:15 ` Claudio Fontana
2020-12-09 8:47 ` Paolo Bonzini
0 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-08 22:15 UTC (permalink / raw)
To: Philippe Mathieu-Daudé,
Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy
Cc: Laurent Vivier, Peter Maydell, Eduardo Habkost, Paul Durrant,
Jason Wang, Marcelo Tosatti, qemu-devel, Peter Xu,
Dario Faggioli, Emilio G . Cota, haxm-team, Cameron Esfahani,
Anthony Perard, Bruce Rogers, Olaf Hering, Colin Xu
On 12/8/20 9:00 PM, Philippe Mathieu-Daudé wrote:
> On 12/8/20 8:48 PM, Claudio Fontana wrote:
>> v8 -> v9: move additional methods to CPUClass->tcg_ops
>>
>> do_unaligned_access, transaction_failed and do_interrupt.
>>
>> do_interrupt is a bit tricky, as the same code is reused
>> (albeit not usually directly) for KVM under certain odd conditions.
>>
>> Change arm, as the only user of do_interrupt callback for KVM,
>> to instead call the target function directly arm_do_interrupt.
>
> Please use scripts/git.orderfile for such refactors please ;)
Aye aye!
If you'd like me to respin now, just let me know.
Ciao,
Claudio
>
> ...
>
>> MAINTAINERS | 19 +-
>> accel/accel-common.c | 105 +++++
>> accel/{accel.c => accel-softmmu.c} | 60 +--
>> accel/accel-softmmu.h | 15 +
>> accel/accel-user.c | 24 ++
>> accel/kvm/kvm-all.c | 2 -
>> accel/kvm/kvm-cpus.c | 26 +-
>> accel/kvm/kvm-cpus.h | 2 -
>> accel/meson.build | 4 +-
>> accel/qtest/qtest.c | 25 +-
>> accel/tcg/cpu-exec.c | 70 +++-
>> accel/tcg/cputlb.c | 6 +-
>> accel/tcg/meson.build | 9 +-
>> accel/tcg/tcg-all.c | 14 +-
>> accel/tcg/tcg-cpus-icount.c | 138 +++++++
>> accel/tcg/tcg-cpus-icount.h | 19 +
>> accel/tcg/tcg-cpus-mttcg.c | 134 +++++++
>> accel/tcg/tcg-cpus-mttcg.h | 19 +
>> accel/tcg/tcg-cpus-rr.c | 298 ++++++++++++++
>> accel/tcg/tcg-cpus-rr.h | 21 +
>> accel/tcg/tcg-cpus.c | 539 +++-----------------------
>> accel/tcg/tcg-cpus.h | 8 +-
>> accel/tcg/user-exec.c | 6 +-
>> accel/xen/xen-all.c | 24 +-
>> bsd-user/main.c | 11 +-
>> cpu.c | 71 ++--
>> hw/core/cpu.c | 30 +-
>> hw/i386/fw_cfg.c | 2 +-
>> hw/i386/intel_iommu.c | 2 +-
>> hw/i386/kvm/apic.c | 2 +-
>> hw/i386/kvm/clock.c | 2 +-
>> hw/i386/microvm.c | 2 +-
>> hw/i386/pc.c | 2 +-
>> hw/i386/pc_piix.c | 1 +
>> hw/i386/x86.c | 2 +-
>> hw/mips/jazz.c | 4 +-
>> include/hw/boards.h | 2 +-
>> include/hw/core/accel-cpu.h | 25 ++
>> include/hw/core/cpu.h | 92 ++---
>> include/hw/core/tcg-cpu-ops.h | 75 ++++
>> include/{sysemu => qemu}/accel.h | 16 +-
>> include/sysemu/accel-ops.h | 45 +++
>> include/sysemu/cpus.h | 26 +-
>> include/sysemu/hvf.h | 2 +-
>> include/sysemu/kvm.h | 2 +-
>> include/sysemu/kvm_int.h | 2 +-
>> linux-user/main.c | 7 +-
>> meson.build | 1 +
>> softmmu/cpus.c | 12 +-
>> softmmu/icount.c | 2 +-
>> softmmu/memory.c | 2 +-
>> softmmu/qtest.c | 2 +-
>> softmmu/vl.c | 8 +-
>> target/alpha/cpu.c | 18 +-
>> target/arm/cpu.c | 26 +-
>> target/arm/cpu64.c | 5 +-
>> target/arm/cpu_tcg.c | 8 +-
>> target/arm/helper.c | 4 +
>> target/arm/kvm64.c | 4 +-
>> target/avr/cpu.c | 13 +-
>> target/avr/helper.c | 4 +-
>> target/cris/cpu.c | 30 +-
>> target/cris/helper.c | 4 +-
>> target/hppa/cpu.c | 15 +-
>> target/i386/cpu-dump.c | 537 +++++++++++++++++++++++++
>> target/i386/cpu.c | 418 ++------------------
>> target/i386/cpu.h | 120 +-----
>> target/i386/{ => hax}/hax-all.c | 5 +-
>> target/i386/{ => hax}/hax-cpus.c | 29 +-
>> target/i386/{ => hax}/hax-cpus.h | 2 -
>> target/i386/{ => hax}/hax-i386.h | 6 +-
>> target/i386/{ => hax}/hax-interface.h | 0
>> target/i386/{ => hax}/hax-mem.c | 0
>> target/i386/{ => hax}/hax-posix.c | 0
>> target/i386/{ => hax}/hax-posix.h | 0
>> target/i386/{ => hax}/hax-windows.c | 0
>> target/i386/{ => hax}/hax-windows.h | 0
>> target/i386/hax/meson.build | 7 +
>> target/i386/helper-tcg.h | 112 ++++++
>> target/i386/helper.c | 539 +-------------------------
>> target/i386/host-cpu.c | 198 ++++++++++
>> target/i386/host-cpu.h | 19 +
>> target/i386/hvf/cpu.c | 65 ++++
>> target/i386/hvf/hvf-cpus.c | 27 +-
>> target/i386/hvf/hvf-cpus.h | 2 -
>> target/i386/hvf/hvf-i386.h | 2 +-
>> target/i386/hvf/hvf.c | 3 +-
>> target/i386/hvf/meson.build | 1 +
>> target/i386/hvf/x86_task.c | 2 +-
>> target/i386/kvm/cpu.c | 148 +++++++
>> target/i386/{ => kvm}/hyperv-proto.h | 0
>> target/i386/{ => kvm}/hyperv-stub.c | 0
>> target/i386/{ => kvm}/hyperv.c | 0
>> target/i386/{ => kvm}/hyperv.h | 0
>> target/i386/kvm/kvm-cpu.h | 41 ++
>> target/i386/{ => kvm}/kvm-stub.c | 0
>> target/i386/{ => kvm}/kvm.c | 3 +-
>> target/i386/{ => kvm}/kvm_i386.h | 0
>> target/i386/kvm/meson.build | 8 +
>> target/i386/kvm/trace-events | 7 +
>> target/i386/kvm/trace.h | 1 +
>> target/i386/machine.c | 4 +-
>> target/i386/meson.build | 39 +-
>> target/i386/{ => tcg}/bpt_helper.c | 1 +
>> target/i386/{ => tcg}/cc_helper.c | 1 +
>> target/i386/tcg/cpu.c | 173 +++++++++
>> target/i386/{ => tcg}/excp_helper.c | 1 +
>> target/i386/{ => tcg}/fpu_helper.c | 33 +-
>> target/i386/{ => tcg}/int_helper.c | 1 +
>> target/i386/{ => tcg}/mem_helper.c | 1 +
>> target/i386/tcg/meson.build | 14 +
>> target/i386/{ => tcg}/misc_helper.c | 1 +
>> target/i386/{ => tcg}/mpx_helper.c | 1 +
>> target/i386/{ => tcg}/seg_helper.c | 1 +
>> target/i386/{ => tcg}/smm_helper.c | 2 +
>> target/i386/{ => tcg}/svm_helper.c | 1 +
>> target/i386/{ => tcg}/tcg-stub.c | 0
>> target/i386/{ => tcg}/translate.c | 1 +
>> target/i386/trace-events | 6 -
>> target/i386/whpx/meson.build | 4 +
>> target/i386/{ => whpx}/whp-dispatch.h | 0
>> target/i386/{ => whpx}/whpx-all.c | 4 +-
>> target/i386/{ => whpx}/whpx-cpus.c | 29 +-
>> target/i386/{ => whpx}/whpx-cpus.h | 2 -
>> target/lm32/cpu.c | 13 +-
>> target/m68k/cpu.c | 12 +-
>> target/microblaze/cpu.c | 23 +-
>> target/mips/cpu.c | 21 +-
>> target/moxie/cpu.c | 10 +-
>> target/nios2/cpu.c | 14 +-
>> target/openrisc/cpu.c | 12 +-
>> target/ppc/translate_init.c.inc | 23 +-
>> target/riscv/cpu.c | 25 +-
>> target/rx/cpu.c | 18 +-
>> target/s390x/cpu.c | 15 +-
>> target/sh4/cpu.c | 14 +-
>> target/sparc/cpu.c | 18 +-
>> target/tilegx/cpu.c | 10 +-
>> target/tricore/cpu.c | 8 +-
>> target/unicore32/cpu.c | 14 +-
>> target/xtensa/cpu.c | 16 +-
>> 141 files changed, 3007 insertions(+), 2051 deletions(-)
> ...
>
>
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 01/32] accel/tcg: split CpusAccel into three TCG variants
2020-12-08 19:48 ` [RFC v9 01/32] accel/tcg: split CpusAccel into three TCG variants Claudio Fontana
@ 2020-12-09 8:34 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 8:34 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Philippe Mathieu-Daudé,
Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> split up the CpusAccel tcg_cpus into three TCG variants:
>
> tcg_cpus_rr (single threaded, round robin cpus)
> tcg_cpus_icount (same as rr, but with instruction counting enabled)
> tcg_cpus_mttcg (multi-threaded cpus)
>
<snip>
> diff --git a/accel/tcg/tcg-cpus-mttcg.c b/accel/tcg/tcg-cpus-mttcg.c
> new file mode 100644
> index 0000000000..dac724fc85
> --- /dev/null
> +++ b/accel/tcg/tcg-cpus-mttcg.c
> @@ -0,0 +1,117 @@
<snip>
> +
> +/*
> + * In the multi-threaded case each vCPU has its own thread. The TLS
> + * variable current_cpu can be used deep in the code to find the
> + * current CPUState for a given thread.
> + */
> +
<snip>
> +
> +/*
> + * In the multi-threaded case each vCPU has its own thread. The TLS
> + * variable current_cpu can be used deep in the code to find the
> + * current CPUState for a given thread.
> + */
nit: This is overly repetitive - I think we can drop the duplication in
the header. If anything the comment should probably be re-formulated at
the current_cpu definition in cpus-common.
<snip>
Otherwise nice clean-up ;-)
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 00/22] i386 cleanup
2020-12-08 22:15 ` Claudio Fontana
@ 2020-12-09 8:47 ` Paolo Bonzini
0 siblings, 0 replies; 87+ messages in thread
From: Paolo Bonzini @ 2020-12-09 8:47 UTC (permalink / raw)
To: Claudio Fontana, Philippe Mathieu-Daudé,
Thomas Huth, Richard Henderson, Stefano Stabellini, Wenchao Wang,
Roman Bolshakov, Sunil Muthuswamy
Cc: Laurent Vivier, Peter Maydell, Eduardo Habkost, Paul Durrant,
Jason Wang, Marcelo Tosatti, qemu-devel, Peter Xu,
Dario Faggioli, Emilio G . Cota, haxm-team, Cameron Esfahani,
Anthony Perard, Bruce Rogers, Olaf Hering, Colin Xu
On 08/12/20 23:15, Claudio Fontana wrote:
> On 12/8/20 9:00 PM, Philippe Mathieu-Daudé wrote:
>> On 12/8/20 8:48 PM, Claudio Fontana wrote:
>>> v8 -> v9: move additional methods to CPUClass->tcg_ops
>>>
>>> do_unaligned_access, transaction_failed and do_interrupt.
>>>
>>> do_interrupt is a bit tricky, as the same code is reused
>>> (albeit not usually directly) for KVM under certain odd conditions.
>>>
>>> Change arm, as the only user of do_interrupt callback for KVM,
>>> to instead call the target function directly arm_do_interrupt.
>>
>> Please use scripts/git.orderfile for such refactors please ;)
>
> Aye aye!
> If you'd like me to respin now, just let me know.
No, not needed thanks!
Paolo
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 02/32] accel/tcg: split tcg_start_vcpu_thread
2020-12-08 19:48 ` [RFC v9 02/32] accel/tcg: split tcg_start_vcpu_thread Claudio Fontana
@ 2020-12-09 9:03 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 9:03 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> after the initial split into 3 tcg variants, we proceed to also
> split tcg_start_vcpu_thread.
>
> We actually split it in 2 this time, since the icount variant
> just uses the round robin function.
>
> Suggested-by: Richard Henderson <richard.henderson@linaro.org>
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
> ---
> accel/tcg/tcg-all.c | 5 ++++
> accel/tcg/tcg-cpus-icount.c | 2 +-
> accel/tcg/tcg-cpus-mttcg.c | 29 +++++++++++++++++--
> accel/tcg/tcg-cpus-mttcg.h | 21 --------------
> accel/tcg/tcg-cpus-rr.c | 39 +++++++++++++++++++++++--
> accel/tcg/tcg-cpus-rr.h | 3 +-
> accel/tcg/tcg-cpus.c | 58 -------------------------------------
> accel/tcg/tcg-cpus.h | 1 -
> 8 files changed, 71 insertions(+), 87 deletions(-)
> delete mode 100644 accel/tcg/tcg-cpus-mttcg.h
>
> diff --git a/accel/tcg/tcg-all.c b/accel/tcg/tcg-all.c
> index e42a028043..1ac0b76515 100644
> --- a/accel/tcg/tcg-all.c
> +++ b/accel/tcg/tcg-all.c
> @@ -105,6 +105,11 @@ static int tcg_init(MachineState *ms)
> tcg_exec_init(s->tb_size * 1024 * 1024);
> mttcg_enabled = s->mttcg_enabled;
>
> + /*
> + * Initialize TCG regions
> + */
> + tcg_region_init();
> +
It might be worth noting in the commit log this is OK because the
accelerator is now guaranteed to be initialised so you know the details
of MTTCG when initialising the regions.
> if (mttcg_enabled) {
> cpus_register_accel(&tcg_cpus_mttcg);
> } else if (icount_enabled()) {
> diff --git a/accel/tcg/tcg-cpus-icount.c b/accel/tcg/tcg-cpus-icount.c
> index d3af3afb6d..82dbe2cacf 100644
<snip>
> diff --git a/accel/tcg/tcg-cpus-mttcg.h b/accel/tcg/tcg-cpus-mttcg.h
> deleted file mode 100644
> index d1bd771f49..0000000000
> --- a/accel/tcg/tcg-cpus-mttcg.h
> +++ /dev/null
> @@ -1,21 +0,0 @@
> -/*
> - * QEMU TCG Multi Threaded vCPUs implementation
> - *
> - * Copyright 2020 SUSE LLC
> - *
> - * This work is licensed under the terms of the GNU GPL, version 2 or later.
> - * See the COPYING file in the top-level directory.
> - */
> -
> -#ifndef TCG_CPUS_MTTCG_H
> -#define TCG_CPUS_MTTCG_H
> -
> -/*
> - * In the multi-threaded case each vCPU has its own thread. The TLS
> - * variable current_cpu can be used deep in the code to find the
> - * current CPUState for a given thread.
> - */
Ahh gone now ;-)
> -
> -void *tcg_cpu_thread_fn(void *arg);
> -
> -#endif /* TCG_CPUS_MTTCG_H */
> diff --git a/accel/tcg/tcg-cpus-rr.c b/accel/tcg/tcg-cpus-rr.c
> index ad50a3765f..f3b262bec7 100644
> --- a/accel/tcg/tcg-cpus-rr.c
> +++ b/accel/tcg/tcg-cpus-rr.c
> @@ -144,7 +144,7 @@ static void deal_with_unplugged_cpus(void)
> * elsewhere.
> */
>
> -void *tcg_rr_cpu_thread_fn(void *arg)
> +static void *tcg_rr_cpu_thread_fn(void *arg)
> {
> CPUState *cpu = arg;
>
> @@ -262,8 +262,43 @@ void *tcg_rr_cpu_thread_fn(void *arg)
> return NULL;
> }
>
> +void rr_start_vcpu_thread(CPUState *cpu)
> +{
> + char thread_name[VCPU_THREAD_NAME_SIZE];
> + static QemuCond *single_tcg_halt_cond;
> + static QemuThread *single_tcg_cpu_thread;
> +
> + g_assert(tcg_enabled());
> + parallel_cpus = false;
> +
> + if (!single_tcg_cpu_thread) {
> + cpu->thread = g_malloc0(sizeof(QemuThread));
> + cpu->halt_cond = g_malloc0(sizeof(QemuCond));
> + qemu_cond_init(cpu->halt_cond);
> +
> + /* share a single thread for all cpus with TCG */
> + snprintf(thread_name, VCPU_THREAD_NAME_SIZE, "ALL CPUs/TCG");
> + qemu_thread_create(cpu->thread, thread_name,
> + tcg_rr_cpu_thread_fn,
> + cpu, QEMU_THREAD_JOINABLE);
> +
> + single_tcg_halt_cond = cpu->halt_cond;
> + single_tcg_cpu_thread = cpu->thread;
> +#ifdef _WIN32
> + cpu->hThread = qemu_thread_get_handle(cpu->thread);
> +#endif
> + } else {
> + /* we share the thread */
> + cpu->thread = single_tcg_cpu_thread;
> + cpu->halt_cond = single_tcg_halt_cond;
> + cpu->thread_id = first_cpu->thread_id;
> + cpu->can_do_io = 1;
> + cpu->created = true;
I was wonder if we should be duplicating hThread here but on closer
examination it looks like it's only used by HAX/WHPX accelerators so
maybe this is a candidate for accelerator specific CPU data?
Otherwise LGTM:
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 03/32] accel/tcg: rename tcg-cpus functions to match module name
2020-12-08 19:48 ` [RFC v9 03/32] accel/tcg: rename tcg-cpus functions to match module name Claudio Fontana
@ 2020-12-09 9:10 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 9:10 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Philippe Mathieu-Daudé,
Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
> ---
> accel/tcg/tcg-cpus-icount.c | 24 ++++++------
> accel/tcg/tcg-cpus-icount.h | 6 +--
> accel/tcg/tcg-cpus-mttcg.c | 10 ++---
> accel/tcg/tcg-cpus-rr.c | 74 ++++++++++++++++++-------------------
> accel/tcg/tcg-cpus-rr.h | 2 +-
> accel/tcg/tcg-cpus.c | 6 +--
> accel/tcg/tcg-cpus.h | 6 +--
> 7 files changed, 64 insertions(+), 64 deletions(-)
>
> diff --git a/accel/tcg/tcg-cpus-icount.c b/accel/tcg/tcg-cpus-icount.c
> index 82dbe2cacf..9f45432275 100644
> --- a/accel/tcg/tcg-cpus-icount.c
> +++ b/accel/tcg/tcg-cpus-icount.c
> @@ -36,7 +36,7 @@
> #include "tcg-cpus-icount.h"
> #include "tcg-cpus-rr.h"
>
> -static int64_t tcg_get_icount_limit(void)
> +static int64_t icount_get_limit(void)
> {
> int64_t deadline;
>
> @@ -68,37 +68,37 @@ static int64_t tcg_get_icount_limit(void)
> }
> }
>
> -static void notify_aio_contexts(void)
> +static void icount_notify_aio_contexts(void)
> {
> /* Wake up other AioContexts. */
> qemu_clock_notify(QEMU_CLOCK_VIRTUAL);
> qemu_clock_run_timers(QEMU_CLOCK_VIRTUAL);
> }
nit: for static functions it's probably overkill but whatever...
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 04/32] i386: move kvm accel files into kvm/
2020-12-08 19:48 ` [RFC v9 04/32] i386: move kvm accel files into kvm/ Claudio Fontana
@ 2020-12-09 9:17 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 9:17 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 05/32] i386: move whpx accel files into whpx/
2020-12-08 19:48 ` [RFC v9 05/32] i386: move whpx accel files into whpx/ Claudio Fontana
@ 2020-12-09 9:21 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 9:21 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> ---
> MAINTAINERS | 5 +----
> target/i386/meson.build | 5 +----
> target/i386/whpx/meson.build | 4 ++++
> target/i386/{ => whpx}/whp-dispatch.h | 0
> target/i386/{ => whpx}/whpx-all.c | 0
> target/i386/{ => whpx}/whpx-cpus.c | 0
> target/i386/{ => whpx}/whpx-cpus.h | 0
> 7 files changed, 6 insertions(+), 8 deletions(-)
> create mode 100644 target/i386/whpx/meson.build
> rename target/i386/{ => whpx}/whp-dispatch.h (100%)
> rename target/i386/{ => whpx}/whpx-all.c (100%)
> rename target/i386/{ => whpx}/whpx-cpus.c (100%)
> rename target/i386/{ => whpx}/whpx-cpus.h (100%)
>
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 5b3eced829..20e079f40c 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -451,10 +451,7 @@ F: include/sysemu/hvf.h
> WHPX CPUs
> M: Sunil Muthuswamy <sunilmut@microsoft.com>
> S: Supported
> -F: target/i386/whpx-all.c
> -F: target/i386/whpx-cpus.c
> -F: target/i386/whp-dispatch.h
> -F: accel/stubs/whpx-stub.c
heh - we never had a stubs...
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 06/32] i386: move hax accel files into hax/
2020-12-08 19:48 ` [RFC v9 06/32] i386: move hax accel files into hax/ Claudio Fontana
@ 2020-12-09 9:22 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 9:22 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 07/32] i386: hvf: remove stale MAINTAINERS entry for old hvf stubs
2020-12-08 19:48 ` [RFC v9 07/32] i386: hvf: remove stale MAINTAINERS entry for old hvf stubs Claudio Fontana
@ 2020-12-09 9:22 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 9:22 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> Reviewed-by: Roman Bolshakov <r.bolshakov@yadro.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 18/32] cpu: Move synchronize_from_tb() to tcg_ops
2020-12-08 19:48 ` [RFC v9 18/32] cpu: Move synchronize_from_tb() to tcg_ops Claudio Fontana
@ 2020-12-09 9:27 ` Philippe Mathieu-Daudé
2020-12-09 14:33 ` Claudio Fontana
2020-12-09 10:50 ` Alex Bennée
1 sibling, 1 reply; 87+ messages in thread
From: Philippe Mathieu-Daudé @ 2020-12-09 9:27 UTC (permalink / raw)
To: Claudio Fontana, Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Colin Xu, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota
On 12/8/20 8:48 PM, Claudio Fontana wrote:
> From: Eduardo Habkost <ehabkost@redhat.com>
>
> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
> [claudio: wrapped in CONFIG_TCG]
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> ---
...
> diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
> index 4475ef0996..109291ac52 100644
> --- a/include/hw/core/tcg-cpu-ops.h
> +++ b/include/hw/core/tcg-cpu-ops.h
> @@ -10,9 +10,6 @@
> #ifndef TCG_CPU_OPS_H
> #define TCG_CPU_OPS_H
Missing for TranslationBlock:
#include "hw/core/cpu.h"
>
> -/**
> - * struct TcgCpuOperations: TCG operations specific to a CPU class
> - */
Why remove the comment?
> typedef struct TcgCpuOperations {
> /**
> * @initialize: Initalize TCG state
> @@ -20,6 +17,17 @@ typedef struct TcgCpuOperations {
> * Called when the first CPU is realized.
> */
> void (*initialize)(void);
> + /**
> + * @synchronize_from_tb: Synchronize state from a TCG #TranslationBlock
> + *
> + * This is called when we abandon execution of a TB before
> + * starting it, and must set all parts of the CPU state which
> + * the previous TB in the chain may not have updated. This
> + * will need to do more. If this hook is not implemented then
> + * the default is to call
> + * @set_pc(tb->pc).
> + */
> + void (*synchronize_from_tb)(CPUState *cpu, struct TranslationBlock *tb);
> } TcgCpuOperations;
...
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 19/32] cpu: Move cpu_exec_* to tcg_ops
2020-12-08 19:48 ` [RFC v9 19/32] cpu: Move cpu_exec_* " Claudio Fontana
@ 2020-12-09 9:28 ` Philippe Mathieu-Daudé
2020-12-09 11:02 ` Claudio Fontana
2020-12-09 11:16 ` Alex Bennée
1 sibling, 1 reply; 87+ messages in thread
From: Philippe Mathieu-Daudé @ 2020-12-09 9:28 UTC (permalink / raw)
To: Claudio Fontana, Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Colin Xu, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota
On 12/8/20 8:48 PM, Claudio Fontana wrote:
> From: Eduardo Habkost <ehabkost@redhat.com>
>
> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
> [claudio: wrapped in CONFIG_TCG]
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> ---
...
> diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
> index 109291ac52..e12f32919b 100644
> --- a/include/hw/core/tcg-cpu-ops.h
> +++ b/include/hw/core/tcg-cpu-ops.h
> @@ -10,6 +10,9 @@
> #ifndef TCG_CPU_OPS_H
> #define TCG_CPU_OPS_H
>
> +/**
> + * struct TcgCpuOperations: TCG operations specific to a CPU class
> + */
Oh, added back ;)
> typedef struct TcgCpuOperations {
> /**
> * @initialize: Initalize TCG state
> @@ -28,6 +31,12 @@ typedef struct TcgCpuOperations {
> * @set_pc(tb->pc).
> */
> void (*synchronize_from_tb)(CPUState *cpu, struct TranslationBlock *tb);
> + /** @cpu_exec_enter: Callback for cpu_exec preparation */
> + void (*cpu_exec_enter)(CPUState *cpu);
> + /** @cpu_exec_exit: Callback for cpu_exec cleanup */
> + void (*cpu_exec_exit)(CPUState *cpu);
> + /** @cpu_exec_interrupt: Callback for processing interrupts in cpu_exec */
> + bool (*cpu_exec_interrupt)(CPUState *cpu, int interrupt_request);
> } TcgCpuOperations;
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 08/32] i386: move TCG accel files into tcg/
2020-12-08 19:48 ` [RFC v9 08/32] i386: move TCG accel files into tcg/ Claudio Fontana
@ 2020-12-09 9:30 ` Alex Bennée
2020-12-09 11:05 ` Claudio Fontana
0 siblings, 1 reply; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 9:30 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> ---
> target/i386/meson.build | 14 +-------------
> target/i386/{ => tcg}/bpt_helper.c | 0
> target/i386/{ => tcg}/cc_helper.c | 0
> target/i386/{ => tcg}/excp_helper.c | 0
> target/i386/{ => tcg}/fpu_helper.c | 0
> target/i386/{ => tcg}/int_helper.c | 0
> target/i386/{ => tcg}/mem_helper.c | 0
> target/i386/tcg/meson.build | 13 +++++++++++++
> target/i386/{ => tcg}/misc_helper.c | 0
> target/i386/{ => tcg}/mpx_helper.c | 0
> target/i386/{ => tcg}/seg_helper.c | 0
> target/i386/{ => tcg}/smm_helper.c | 0
> target/i386/{ => tcg}/svm_helper.c | 0
> target/i386/{ => tcg}/tcg-stub.c | 0
> target/i386/{ => tcg}/translate.c | 0
> 15 files changed, 14 insertions(+), 13 deletions(-)
> rename target/i386/{ => tcg}/bpt_helper.c (100%)
> rename target/i386/{ => tcg}/cc_helper.c (100%)
> rename target/i386/{ => tcg}/excp_helper.c (100%)
> rename target/i386/{ => tcg}/fpu_helper.c (100%)
> rename target/i386/{ => tcg}/int_helper.c (100%)
> rename target/i386/{ => tcg}/mem_helper.c (100%)
> create mode 100644 target/i386/tcg/meson.build
> rename target/i386/{ => tcg}/misc_helper.c (100%)
> rename target/i386/{ => tcg}/mpx_helper.c (100%)
> rename target/i386/{ => tcg}/seg_helper.c (100%)
> rename target/i386/{ => tcg}/smm_helper.c (100%)
> rename target/i386/{ => tcg}/svm_helper.c (100%)
> rename target/i386/{ => tcg}/tcg-stub.c (100%)
> rename target/i386/{ => tcg}/translate.c (100%)
I noticed target/i386/helper-tcg.h is still in the main directory. It
maybe be a straggler that is harder to separate - if so might be worth
mentioning in the commit message. Otherwise:
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 24/32] cpu: move cc->transaction_failed to tcg_ops
2020-12-08 19:48 ` [RFC v9 24/32] cpu: move cc->transaction_failed " Claudio Fontana
@ 2020-12-09 9:31 ` Philippe Mathieu-Daudé
2020-12-09 14:43 ` Claudio Fontana
2020-12-09 12:03 ` Alex Bennée
1 sibling, 1 reply; 87+ messages in thread
From: Philippe Mathieu-Daudé @ 2020-12-09 9:31 UTC (permalink / raw)
To: Claudio Fontana, Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Colin Xu, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota
On 12/8/20 8:48 PM, Claudio Fontana wrote:
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> ---
...
> diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
> index 6b9b72e785..0cc0927738 100644
> --- a/include/hw/core/tcg-cpu-ops.h
> +++ b/include/hw/core/tcg-cpu-ops.h
Missing (for MemTxAttrs/MemTxResult):
#include "exec/memattrs.h"
> @@ -40,6 +40,15 @@ typedef struct TcgCpuOperations {
> /** @do_interrupt: Callback for interrupt handling. */
> void (*do_interrupt)(CPUState *cpu);
>
> + /**
> + * @do_transaction_failed: Callback for handling failed memory transactions
> + * (ie bus faults or external aborts; not MMU faults)
> + */
> + void (*do_transaction_failed)(CPUState *cpu, hwaddr physaddr, vaddr addr,
> + unsigned size, MMUAccessType access_type,
> + int mmu_idx, MemTxAttrs attrs,
> + MemTxResult response, uintptr_t retaddr);
> +
...
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 09/32] i386: move cpu dump out of helper.c into cpu-dump.c
2020-12-08 19:48 ` [RFC v9 09/32] i386: move cpu dump out of helper.c into cpu-dump.c Claudio Fontana
@ 2020-12-09 9:59 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 9:59 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 00/22] i386 cleanup
2020-12-08 22:00 ` no-reply
@ 2020-12-09 10:22 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 10:22 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, paul, jasowang, peterx, haxm-team, colin.xu,
ohering, sstabellini, brogers, cota, cfontana, anthony.perard,
philmd, lvivier, thuth, ehabkost, richard.henderson, dirty,
dfaggioli, r.bolshakov, sunilmut, mtosatti, wenchao.wang,
pbonzini
no-reply@patchew.org writes:
> Patchew URL: https://patchew.org/QEMU/20201208194839.31305-1-cfontana@suse.de/
>
>
>
> Hi,
>
> This series seems to have some coding style problems. See output below for
> more information:
Also replicated in:
https://gitlab.com/stsquad/qemu/-/jobs/899057106
>
> Type: series
> Message-id: 20201208194839.31305-1-cfontana@suse.de
> Subject: [RFC v9 00/22] i386 cleanup
>
> === TEST SCRIPT BEGIN ===
> #!/bin/bash
> git rev-parse base > /dev/null || exit 0
> git config --local diff.renamelimit 0
> git config --local diff.renames True
> git config --local diff.algorithm histogram
> ./scripts/checkpatch.pl --mailback base..
> === TEST SCRIPT END ===
>
> Updating 3c8cf5a9c21ff8782164d1def7f44bd888713384
> From https://github.com/patchew-project/qemu
> * [new tag] patchew/20201208194839.31305-1-cfontana@suse.de -> patchew/20201208194839.31305-1-cfontana@suse.de
> Switched to a new branch 'test'
> 1df468e cpu: introduce cpu_accel_instance_init
> 273e6bc hw/core/cpu: call qemu_init_vcpu in cpu_common_realizefn
> 6c8c6a4 cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn
> c7ecc11 i386: split cpu accelerators from cpu.c, using AccelCPUClass
> 739ef2b accel: introduce AccelCPUClass extending CPUClass
> 706469e accel: replace struct CpusAccel with AccelOpsClass
> c3aa222 accel: extend AccelState and AccelClass to user-mode
> c45684a cpu: move do_unaligned_access to tcg_ops
> bca539a cpu: move cc->transaction_failed to tcg_ops
> 74feabd cpu: move cc->do_interrupt to tcg_ops
> f22d896 target/arm: do not use cc->do_interrupt for KVM directly
> 9a74429 cpu: Move debug_excp_handler to tcg_ops
> bf5b24b cpu: Move tlb_fill to tcg_ops
> 481a06e cpu: Move cpu_exec_* to tcg_ops
> af72917 cpu: Move synchronize_from_tb() to tcg_ops
> f9b393d accel/tcg: split TCG-only code from cpu_exec_realizefn
> 2019e4f target/riscv: remove CONFIG_TCG, as it is always TCG
> 53b3f4e cpu: Introduce TCGCpuOperations struct
> 8a0c2e9 cpu: Remove unnecessary noop methods
> 043b89c tcg: Make CPUClass.debug_excp_handler optional
> b271ab4 tcg: make CPUClass.cpu_exec_* optional
> ac15b62 tcg: cpu_exec_{enter,exit} helpers
> 0560b09 i386: move TCG cpu class initialization out of helper.c
> 36ae5f1 i386: move cpu dump out of helper.c into cpu-dump.c
> ef2d207 i386: move TCG accel files into tcg/
> 9d7a737 i386: hvf: remove stale MAINTAINERS entry for old hvf stubs
> 42deca7 i386: move hax accel files into hax/
> 7dfc2be i386: move whpx accel files into whpx/
> 4d4c885 i386: move kvm accel files into kvm/
> d9966d5 accel/tcg: rename tcg-cpus functions to match module name
> a8f0cbf accel/tcg: split tcg_start_vcpu_thread
> 5e46bf0 accel/tcg: split CpusAccel into three TCG variants
>
> === OUTPUT BEGIN ===
> 1/32 Checking commit 5e46bf019fb7 (accel/tcg: split CpusAccel into three TCG variants)
> WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
> #57:
> new file mode 100644
>
> total: 0 errors, 1 warnings, 1156 lines checked
>
> Patch 1/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 2/32 Checking commit a8f0cbf6b610 (accel/tcg: split tcg_start_vcpu_thread)
> WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
> #103:
> deleted file mode 100644
>
> total: 0 errors, 1 warnings, 201 lines checked
>
> Patch 2/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 3/32 Checking commit d9966d571c7e (accel/tcg: rename tcg-cpus functions to match module name)
> 4/32 Checking commit 4d4c885ff790 (i386: move kvm accel files into kvm/)
> WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
> #167:
> rename from target/i386/hyperv-proto.h
>
> total: 0 errors, 1 warnings, 150 lines checked
>
> Patch 4/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 5/32 Checking commit 7dfc2bea3c57 (i386: move whpx accel files into whpx/)
> WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
> #50:
> new file mode 100644
>
> total: 0 errors, 1 warnings, 32 lines checked
>
> Patch 5/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 6/32 Checking commit 42deca72cda4 (i386: move hax accel files into hax/)
> WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
> #25:
> rename from target/i386/hax-all.c
>
> total: 0 errors, 1 warnings, 47 lines checked
>
> Patch 6/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 7/32 Checking commit 9d7a7374d252 (i386: hvf: remove stale MAINTAINERS entry for old hvf stubs)
> 8/32 Checking commit ef2d207c5f81 (i386: move TCG accel files into tcg/)
> WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
> #44:
> rename from target/i386/bpt_helper.c
>
> total: 0 errors, 1 warnings, 39 lines checked
>
> Patch 8/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 9/32 Checking commit 36ae5f191d8c (i386: move cpu dump out of helper.c into cpu-dump.c)
> WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
> #11:
> new file mode 100644
>
> ERROR: braces {} are necessary for all arms of this statement
> #128: FILE: target/i386/cpu-dump.c:113:
> + if (!(env->hflags & HF_PE_MASK) || !(sc->flags & DESC_P_MASK))
> [...]
>
> WARNING: Block comments use a leading /* on a separate line
> #172: FILE: target/i386/cpu-dump.c:157:
> +/* ARRAY_SIZE check is not required because
>
> WARNING: Block comments use a leading /* on a separate line
> #211: FILE: target/i386/cpu-dump.c:196:
> +/* ARRAY_SIZE check is not required because
>
> ERROR: line over 90 characters
> #373: FILE: target/i386/cpu-dump.c:358:
> + qemu_fprintf(f, "RAX=%016" PRIx64 " RBX=%016" PRIx64 " RCX=%016" PRIx64 " RDX=%016" PRIx64 "\n"
>
> ERROR: line over 90 characters
> #374: FILE: target/i386/cpu-dump.c:359:
> + "RSI=%016" PRIx64 " RDI=%016" PRIx64 " RBP=%016" PRIx64 " RSP=%016" PRIx64 "\n"
>
> ERROR: line over 90 characters
> #375: FILE: target/i386/cpu-dump.c:360:
> + "R8 =%016" PRIx64 " R9 =%016" PRIx64 " R10=%016" PRIx64 " R11=%016" PRIx64 "\n"
>
> ERROR: line over 90 characters
> #376: FILE: target/i386/cpu-dump.c:361:
> + "R12=%016" PRIx64 " R13=%016" PRIx64 " R14=%016" PRIx64 " R15=%016" PRIx64 "\n"
>
> ERROR: line over 90 characters
> #377: FILE: target/i386/cpu-dump.c:362:
> + "RIP=%016" PRIx64 " RFL=%08x [%c%c%c%c%c%c%c] CPL=%d II=%d A20=%d SMM=%d HLT=%d\n",
>
> ERROR: space required before the open parenthesis '('
> #436: FILE: target/i386/cpu-dump.c:421:
> + for(i = 0; i < 6; i++) {
>
> WARNING: line over 80 characters
> #448: FILE: target/i386/cpu-dump.c:433:
> + qemu_fprintf(f, "CR0=%08x CR2=%016" PRIx64 " CR3=%016" PRIx64 " CR4=%08x\n",
>
> ERROR: space required before the open parenthesis '('
> #453: FILE: target/i386/cpu-dump.c:438:
> + for(i = 0; i < 4; i++)
>
> ERROR: braces {} are necessary for all arms of this statement
> #453: FILE: target/i386/cpu-dump.c:438:
> + for(i = 0; i < 4; i++)
> [...]
>
> ERROR: space required before the open parenthesis '('
> #469: FILE: target/i386/cpu-dump.c:454:
> + for(i = 0; i < 4; i++) {
>
> ERROR: braces {} are necessary for all arms of this statement
> #476: FILE: target/i386/cpu-dump.c:461:
> + if ((unsigned)env->cc_op < CC_OP_NB)
> [...]
> + else
> [...]
>
> WARNING: line over 80 characters
> #477: FILE: target/i386/cpu-dump.c:462:
> + snprintf(cc_op_name, sizeof(cc_op_name), "%s", cc_op_str[env->cc_op]);
>
> ERROR: space required before the open parenthesis '('
> #497: FILE: target/i386/cpu-dump.c:482:
> + for(i = 0; i < 8; i++) {
>
> ERROR: spaces required around that '=' (ctx:VxV)
> #507: FILE: target/i386/cpu-dump.c:492:
> + for(i=0;i<8;i++) {
> ^
>
> ERROR: space required after that ';' (ctx:VxV)
> #507: FILE: target/i386/cpu-dump.c:492:
> + for(i=0;i<8;i++) {
> ^
>
> ERROR: spaces required around that '<' (ctx:VxV)
> #507: FILE: target/i386/cpu-dump.c:492:
> + for(i=0;i<8;i++) {
> ^
>
> ERROR: space required after that ';' (ctx:VxV)
> #507: FILE: target/i386/cpu-dump.c:492:
> + for(i=0;i<8;i++) {
> ^
>
> ERROR: space required before the open parenthesis '('
> #507: FILE: target/i386/cpu-dump.c:492:
> + for(i=0;i<8;i++) {
>
> ERROR: braces {} are necessary for all arms of this statement
> #512: FILE: target/i386/cpu-dump.c:497:
> + if ((i & 1) == 1)
> [...]
> + else
> [...]
>
> ERROR: braces {} are necessary for all arms of this statement
> #517: FILE: target/i386/cpu-dump.c:502:
> + if (env->hflags & HF_CS64_MASK)
> [...]
> + else
> [...]
>
> ERROR: spaces required around that '=' (ctx:VxV)
> #521: FILE: target/i386/cpu-dump.c:506:
> + for(i=0;i<nb;i++) {
> ^
>
> ERROR: space required after that ';' (ctx:VxV)
> #521: FILE: target/i386/cpu-dump.c:506:
> + for(i=0;i<nb;i++) {
> ^
>
> ERROR: spaces required around that '<' (ctx:VxV)
> #521: FILE: target/i386/cpu-dump.c:506:
> + for(i=0;i<nb;i++) {
> ^
>
> ERROR: space required after that ';' (ctx:VxV)
> #521: FILE: target/i386/cpu-dump.c:506:
> + for(i=0;i<nb;i++) {
> ^
>
> ERROR: space required before the open parenthesis '('
> #521: FILE: target/i386/cpu-dump.c:506:
> + for(i=0;i<nb;i++) {
>
> ERROR: braces {} are necessary for all arms of this statement
> #528: FILE: target/i386/cpu-dump.c:513:
> + if ((i & 1) == 1)
> [...]
> + else
> [...]
>
> total: 25 errors, 5 warnings, 1077 lines checked
>
> Patch 9/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
>
> 10/32 Checking commit 0560b09d8481 (i386: move TCG cpu class initialization out of helper.c)
> WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
> #282:
> new file mode 100644
>
> total: 0 errors, 1 warnings, 610 lines checked
>
> Patch 10/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 11/32 Checking commit ac15b624a91f (tcg: cpu_exec_{enter,exit} helpers)
> 12/32 Checking commit b271ab48c45b (tcg: make CPUClass.cpu_exec_* optional)
> 13/32 Checking commit 043b89c9c315 (tcg: Make CPUClass.debug_excp_handler optional)
> 14/32 Checking commit 8a0c2e9a0e2c (cpu: Remove unnecessary noop methods)
> 15/32 Checking commit 53b3f4e5b390 (cpu: Introduce TCGCpuOperations struct)
> WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
> #92:
> new file mode 100644
>
> total: 0 errors, 1 warnings, 301 lines checked
>
> Patch 15/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 16/32 Checking commit 2019e4f802bb (target/riscv: remove CONFIG_TCG, as it is always TCG)
> 17/32 Checking commit f9b393d26c70 (accel/tcg: split TCG-only code from cpu_exec_realizefn)
> 18/32 Checking commit af729170f4b8 (cpu: Move synchronize_from_tb() to tcg_ops)
> 19/32 Checking commit 481a06e3b020 (cpu: Move cpu_exec_* to tcg_ops)
> 20/32 Checking commit bf5b24bcd22f (cpu: Move tlb_fill to tcg_ops)
> WARNING: line over 80 characters
> #22: FILE: accel/tcg/cputlb.c:1289:
> + ok = cc->tcg_ops.tlb_fill(cpu, addr, size, access_type, mmu_idx, false, retaddr);
>
> total: 0 errors, 1 warnings, 259 lines checked
>
> Patch 20/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 21/32 Checking commit 9a74429f2029 (cpu: Move debug_excp_handler to tcg_ops)
> 22/32 Checking commit f22d896967f8 (target/arm: do not use cc->do_interrupt for KVM directly)
> 23/32 Checking commit 74feabd5bae5 (cpu: move cc->do_interrupt to tcg_ops)
> 24/32 Checking commit bca539a6251e (cpu: move cc->transaction_failed to tcg_ops)
> WARNING: line over 80 characters
> #67: FILE: include/hw/core/cpu.h:857:
> + cc->tcg_ops.do_transaction_failed(cpu, physaddr, addr, size, access_type,
>
> total: 0 errors, 1 warnings, 144 lines checked
>
> Patch 24/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 25/32 Checking commit c45684ac5ad9 (cpu: move do_unaligned_access to tcg_ops)
> 26/32 Checking commit c3aa2224bd71 (accel: extend AccelState and AccelClass to user-mode)
> WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
> #24:
> new file mode 100644
>
> total: 0 errors, 1 warnings, 359 lines checked
>
> Patch 26/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 27/32 Checking commit 706469ebcc15 (accel: replace struct CpusAccel with AccelOpsClass)
> WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
> #115:
> new file mode 100644
>
> total: 0 errors, 1 warnings, 691 lines checked
>
> Patch 27/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 28/32 Checking commit 739ef2bd13aa (accel: introduce AccelCPUClass extending CPUClass)
> WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
> #99:
> new file mode 100644
>
> total: 0 errors, 1 warnings, 123 lines checked
>
> Patch 28/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 29/32 Checking commit c7ecc117664f (i386: split cpu accelerators from cpu.c, using AccelCPUClass)
> WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
> #664:
> new file mode 100644
>
> WARNING: line over 80 characters
> #1435: FILE: target/i386/tcg/cpu.c:104:
> + memory_region_add_subregion_overlap(cpu->cpu_as_root, 0, cpu->cpu_as_mem, 0);
>
> total: 0 errors, 2 warnings, 1296 lines checked
>
> Patch 29/32 has style problems, please review. If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 30/32 Checking commit 6c8c6a4d84dd (cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn)
> 31/32 Checking commit 273e6bca6404 (hw/core/cpu: call qemu_init_vcpu in cpu_common_realizefn)
> 32/32 Checking commit 1df468eafd63 (cpu: introduce cpu_accel_instance_init)
> === OUTPUT END ===
>
> Test command exited with code: 1
>
>
> The full log is available at
> http://patchew.org/logs/20201208194839.31305-1-cfontana@suse.de/testing.checkpatch/?type=message.
> ---
> Email generated automatically by Patchew [https://patchew.org/].
> Please send your feedback to patchew-devel@redhat.com
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 10/32] i386: move TCG cpu class initialization out of helper.c
2020-12-08 19:48 ` [RFC v9 10/32] i386: move TCG cpu class initialization out of helper.c Claudio Fontana
@ 2020-12-09 10:23 ` Alex Bennée
2020-12-09 11:19 ` Claudio Fontana
0 siblings, 1 reply; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 10:23 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> ---
> target/i386/cpu.c | 33 ++++------
> target/i386/cpu.h | 97 ++---------------------------
> target/i386/helper-tcg.h | 112 ++++++++++++++++++++++++++++++++++
> target/i386/helper.c | 23 -------
> target/i386/meson.build | 1 +
> target/i386/tcg-cpu.c | 71 +++++++++++++++++++++
> target/i386/tcg-cpu.h | 15 +++++
> target/i386/tcg/bpt_helper.c | 1 +
> target/i386/tcg/cc_helper.c | 1 +
> target/i386/tcg/excp_helper.c | 1 +
> target/i386/tcg/fpu_helper.c | 33 +++++-----
> target/i386/tcg/int_helper.c | 1 +
> target/i386/tcg/mem_helper.c | 1 +
> target/i386/tcg/misc_helper.c | 1 +
> target/i386/tcg/mpx_helper.c | 1 +
> target/i386/tcg/seg_helper.c | 1 +
> target/i386/tcg/smm_helper.c | 2 +
> target/i386/tcg/svm_helper.c | 1 +
> target/i386/tcg/translate.c | 1 +
> 19 files changed, 244 insertions(+), 153 deletions(-)
> create mode 100644 target/i386/helper-tcg.h
> create mode 100644 target/i386/tcg-cpu.c
> create mode 100644 target/i386/tcg-cpu.h
>
> diff --git a/target/i386/cpu.c b/target/i386/cpu.c
> index b9bd249c8f..3462d0143f 100644
> --- a/target/i386/cpu.c
> +++ b/target/i386/cpu.c
> @@ -24,6 +24,8 @@
> #include "qemu/qemu-print.h"
>
> #include "cpu.h"
> +#include "tcg-cpu.h"
> +#include "helper-tcg.h"
> #include "exec/exec-all.h"
> #include "sysemu/kvm.h"
> #include "sysemu/reset.h"
> @@ -1495,7 +1497,8 @@ static inline uint64_t x86_cpu_xsave_components(X86CPU *cpu)
> cpu->env.features[FEAT_XSAVE_COMP_LO];
> }
>
> -const char *get_register_name_32(unsigned int reg)
> +/* Return name of 32-bit register, from a R_* constant */
> +static const char *get_register_name_32(unsigned int reg)
> {
> if (reg >= CPU_NB_REGS32) {
> return NULL;
> @@ -7012,13 +7015,6 @@ static void x86_cpu_set_pc(CPUState *cs, vaddr value)
> cpu->env.eip = value;
> }
>
> -static void x86_cpu_synchronize_from_tb(CPUState *cs, TranslationBlock *tb)
> -{
> - X86CPU *cpu = X86_CPU(cs);
> -
> - cpu->env.eip = tb->pc - tb->cs_base;
> -}
> -
> int x86_cpu_pending_interrupt(CPUState *cs, int interrupt_request)
> {
> X86CPU *cpu = X86_CPU(cs);
> @@ -7252,17 +7248,18 @@ static void x86_cpu_common_class_init(ObjectClass *oc, void *data)
> cc->class_by_name = x86_cpu_class_by_name;
> cc->parse_features = x86_cpu_parse_featurestr;
> cc->has_work = x86_cpu_has_work;
> +
> #ifdef CONFIG_TCG
> - cc->do_interrupt = x86_cpu_do_interrupt;
> - cc->cpu_exec_interrupt = x86_cpu_exec_interrupt;
> -#endif
> + tcg_cpu_common_class_init(cc);
Are we likely to have clashing names here as other arches get converted?
tcg_x86_cpu_common_class_init or x86_cpu_common_class_init?
<snip>
> diff --git a/target/i386/tcg-cpu.c b/target/i386/tcg-cpu.c
> new file mode 100644
> index 0000000000..628dd29fe7
> --- /dev/null
> +++ b/target/i386/tcg-cpu.c
> @@ -0,0 +1,71 @@
<snip>
> +
> +void tcg_cpu_common_class_init(CPUClass *cc)
> +{
> + cc->do_interrupt = x86_cpu_do_interrupt;
> + cc->cpu_exec_interrupt = x86_cpu_exec_interrupt;
> + cc->synchronize_from_tb = x86_cpu_synchronize_from_tb;
> + cc->cpu_exec_enter = x86_cpu_exec_enter;
> + cc->cpu_exec_exit = x86_cpu_exec_exit;
> + cc->tcg_initialize = tcg_x86_init;
> + cc->tlb_fill = x86_cpu_tlb_fill;
> +#ifndef CONFIG_USER_ONLY
> + cc->debug_excp_handler = breakpoint_handler;
> +#endif
> +}
Oh I see this moves down to target/i386/tcg/ eventually. Maybe we should
just create the new file in place so it's easier to follow the changes
as we convert to a proper abstracted class?
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 11/32] tcg: cpu_exec_{enter,exit} helpers
2020-12-08 19:48 ` [RFC v9 11/32] tcg: cpu_exec_{enter,exit} helpers Claudio Fontana
@ 2020-12-09 10:33 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 10:33 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> From: Eduardo Habkost <ehabkost@redhat.com>
>
> Move invocation of CPUClass.cpu_exec_*() to separate helpers,
> to make it easier to refactor that code later.
>
> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 12/32] tcg: make CPUClass.cpu_exec_* optional
2020-12-08 19:48 ` [RFC v9 12/32] tcg: make CPUClass.cpu_exec_* optional Claudio Fontana
@ 2020-12-09 10:36 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 10:36 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> From: Eduardo Habkost <ehabkost@redhat.com>
>
> This will let us simplify the code that initializes CPU class
> methods, when we move cpu_exec_*() to a separate struct.
> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 13/32] tcg: Make CPUClass.debug_excp_handler optional
2020-12-08 19:48 ` [RFC v9 13/32] tcg: Make CPUClass.debug_excp_handler optional Claudio Fontana
@ 2020-12-09 10:37 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 10:37 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> From: Eduardo Habkost <ehabkost@redhat.com>
>
> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 14/32] cpu: Remove unnecessary noop methods
2020-12-08 19:48 ` [RFC v9 14/32] cpu: Remove unnecessary noop methods Claudio Fontana
@ 2020-12-09 10:38 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 10:38 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> From: Eduardo Habkost <ehabkost@redhat.com>
>
> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 15/32] cpu: Introduce TCGCpuOperations struct
2020-12-08 19:48 ` [RFC v9 15/32] cpu: Introduce TCGCpuOperations struct Claudio Fontana
@ 2020-12-09 10:39 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 10:39 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> From: Eduardo Habkost <ehabkost@redhat.com>
>
> The TCG-specific CPU methods will be moved to a separate struct,
> to make it easier to move accel-specific code outside generic CPU
> code in the future. Start by moving tcg_initialize().
>
> The new CPUClass.tcg_opts field may eventually become a pointer,
> but keep it an embedded struct for now, to make code conversion
> easier.
>
> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
>
> [claudio: make the tcg code build for CONFIG_TCG only]
>
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 17/32] accel/tcg: split TCG-only code from cpu_exec_realizefn
2020-12-08 19:48 ` [RFC v9 17/32] accel/tcg: split TCG-only code from cpu_exec_realizefn Claudio Fontana
@ 2020-12-09 10:42 ` Alex Bennée
2020-12-09 11:22 ` Claudio Fontana
0 siblings, 1 reply; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 10:42 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> move away TCG-only code, make it compile only on TCG.
>
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> ---
> accel/tcg/cpu-exec.c | 28 +++++++++++++++++
> cpu.c | 70 ++++++++++++++++++++-----------------------
> hw/core/cpu.c | 6 +++-
> include/hw/core/cpu.h | 8 +++++
> 4 files changed, 74 insertions(+), 38 deletions(-)
>
> diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
> index 64cba89356..436dfbf155 100644
> --- a/accel/tcg/cpu-exec.c
> +++ b/accel/tcg/cpu-exec.c
> @@ -801,6 +801,34 @@ int cpu_exec(CPUState *cpu)
> return ret;
> }
>
> +void tcg_exec_realizefn(CPUState *cpu, Error **errp)
> +{
> + static bool tcg_target_initialized;
> + CPUClass *cc = CPU_GET_CLASS(cpu);
> +
> + if (!tcg_target_initialized) {
> + tcg_target_initialized = true;
> + cc->tcg_ops.initialize();
nit: it makes no difference but stylistically it makes sense to set
tcg_target_initialized after we have in fact initialised.
Also we've dropped the tcg_enabled() check, if indeed it will always be
true should we not assert it to ensure the statement for tcg_exec_init
remains the case: "Must be called before using the QEMU cpus."
Otherwise LGTM:
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 18/32] cpu: Move synchronize_from_tb() to tcg_ops
2020-12-08 19:48 ` [RFC v9 18/32] cpu: Move synchronize_from_tb() to tcg_ops Claudio Fontana
2020-12-09 9:27 ` Philippe Mathieu-Daudé
@ 2020-12-09 10:50 ` Alex Bennée
2020-12-09 14:46 ` Eduardo Habkost
1 sibling, 1 reply; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 10:50 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> From: Eduardo Habkost <ehabkost@redhat.com>
>
> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
> [claudio: wrapped in CONFIG_TCG]
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> ---
> accel/tcg/cpu-exec.c | 4 ++--
> include/hw/core/cpu.h | 8 --------
> include/hw/core/tcg-cpu-ops.h | 14 +++++++++++---
> target/arm/cpu.c | 2 +-
> target/avr/cpu.c | 2 +-
> target/hppa/cpu.c | 2 +-
> target/i386/tcg-cpu.c | 2 +-
> target/microblaze/cpu.c | 2 +-
> target/mips/cpu.c | 2 +-
> target/riscv/cpu.c | 2 +-
> target/rx/cpu.c | 2 +-
> target/sh4/cpu.c | 2 +-
> target/sparc/cpu.c | 2 +-
> target/tricore/cpu.c | 2 +-
> 14 files changed, 24 insertions(+), 24 deletions(-)
>
> diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
> index 436dfbf155..776415d47e 100644
> --- a/accel/tcg/cpu-exec.c
> +++ b/accel/tcg/cpu-exec.c
> @@ -192,8 +192,8 @@ static inline tcg_target_ulong cpu_tb_exec(CPUState *cpu, TranslationBlock *itb)
> TARGET_FMT_lx "] %s\n",
> last_tb->tc.ptr, last_tb->pc,
> lookup_symbol(last_tb->pc));
> - if (cc->synchronize_from_tb) {
> - cc->synchronize_from_tb(cpu, last_tb);
> + if (cc->tcg_ops.synchronize_from_tb) {
> + cc->tcg_ops.synchronize_from_tb(cpu, last_tb);
> } else {
> assert(cc->set_pc);
> cc->set_pc(cpu, last_tb->pc);
> diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
> index ea648d52ad..83007d262c 100644
> --- a/include/hw/core/cpu.h
> +++ b/include/hw/core/cpu.h
> @@ -110,13 +110,6 @@ struct TranslationBlock;
> * If the target behaviour here is anything other than "set
> * the PC register to the value passed in" then the target must
> * also implement the synchronize_from_tb hook.
> - * @synchronize_from_tb: Callback for synchronizing state from a TCG
> - * #TranslationBlock. This is called when we abandon execution
> - * of a TB before starting it, and must set all parts of the CPU
> - * state which the previous TB in the chain may not have updated.
> - * This always includes at least the program counter; some targets
> - * will need to do more. If this hook is not implemented then the
> - * default is to call @set_pc(tb->pc).
> * @tlb_fill: Callback for handling a softmmu tlb miss or user-only
> * address fault. For system mode, if the access is valid, call
> * tlb_set_page and return true; if the access is invalid, and
> @@ -193,7 +186,6 @@ struct CPUClass {
> void (*get_memory_mapping)(CPUState *cpu, MemoryMappingList *list,
> Error **errp);
> void (*set_pc)(CPUState *cpu, vaddr value);
> - void (*synchronize_from_tb)(CPUState *cpu, struct TranslationBlock *tb);
> bool (*tlb_fill)(CPUState *cpu, vaddr address, int size,
> MMUAccessType access_type, int mmu_idx,
> bool probe, uintptr_t retaddr);
> diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
> index 4475ef0996..109291ac52 100644
> --- a/include/hw/core/tcg-cpu-ops.h
> +++ b/include/hw/core/tcg-cpu-ops.h
> @@ -10,9 +10,6 @@
> #ifndef TCG_CPU_OPS_H
> #define TCG_CPU_OPS_H
>
> -/**
> - * struct TcgCpuOperations: TCG operations specific to a CPU class
> - */
> typedef struct TcgCpuOperations {
> /**
> * @initialize: Initalize TCG state
> @@ -20,6 +17,17 @@ typedef struct TcgCpuOperations {
> * Called when the first CPU is realized.
> */
> void (*initialize)(void);
> + /**
> + * @synchronize_from_tb: Synchronize state from a TCG #TranslationBlock
> + *
> + * This is called when we abandon execution of a TB before
> + * starting it, and must set all parts of the CPU state which
> + * the previous TB in the chain may not have updated. This
> + * will need to do more. If this hook is not implemented then
> + * the default is to call
> + * @set_pc(tb->pc).
> + */
We could clean-up the comment here to better fit the javadoc style API
documentation format, i.e.:
/**
* synchronize_from_tb: Synchronize current TCG execution state
* @cpu: current cpu
* @tb: about to be executed #TranslationBlock
*
* This is called when we abandon execution of a TB before
* starting it, and must set all parts of the CPU state which
* the previous TB in the chain may not have updated. This
* will need to do more. If this hook is not implemented then
* the default is to call @set_pc(tb->pc).
*/
<snip>
Otherwise:
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 19/32] cpu: Move cpu_exec_* to tcg_ops
2020-12-09 9:28 ` Philippe Mathieu-Daudé
@ 2020-12-09 11:02 ` Claudio Fontana
0 siblings, 0 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-09 11:02 UTC (permalink / raw)
To: Philippe Mathieu-Daudé,
Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Cameron Esfahani, haxm-team, Colin Xu, Anthony Perard,
Bruce Rogers, Olaf Hering, Emilio G . Cota
On 12/9/20 10:28 AM, Philippe Mathieu-Daudé wrote:
> On 12/8/20 8:48 PM, Claudio Fontana wrote:
>> From: Eduardo Habkost <ehabkost@redhat.com>
>>
>> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
>> [claudio: wrapped in CONFIG_TCG]
>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
>> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
>> ---
> ...
>
>> diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
>> index 109291ac52..e12f32919b 100644
>> --- a/include/hw/core/tcg-cpu-ops.h
>> +++ b/include/hw/core/tcg-cpu-ops.h
>> @@ -10,6 +10,9 @@
>> #ifndef TCG_CPU_OPS_H
>> #define TCG_CPU_OPS_H
>>
>> +/**
>> + * struct TcgCpuOperations: TCG operations specific to a CPU class
>> + */
>
> Oh, added back ;)
Fixed this "back and forth", thanks!
Claudio
>
>> typedef struct TcgCpuOperations {
>> /**
>> * @initialize: Initalize TCG state
>> @@ -28,6 +31,12 @@ typedef struct TcgCpuOperations {
>> * @set_pc(tb->pc).
>> */
>> void (*synchronize_from_tb)(CPUState *cpu, struct TranslationBlock *tb);
>> + /** @cpu_exec_enter: Callback for cpu_exec preparation */
>> + void (*cpu_exec_enter)(CPUState *cpu);
>> + /** @cpu_exec_exit: Callback for cpu_exec cleanup */
>> + void (*cpu_exec_exit)(CPUState *cpu);
>> + /** @cpu_exec_interrupt: Callback for processing interrupts in cpu_exec */
>> + bool (*cpu_exec_interrupt)(CPUState *cpu, int interrupt_request);
>> } TcgCpuOperations;
>
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 08/32] i386: move TCG accel files into tcg/
2020-12-09 9:30 ` Alex Bennée
@ 2020-12-09 11:05 ` Claudio Fontana
0 siblings, 0 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-09 11:05 UTC (permalink / raw)
To: Alex Bennée
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
On 12/9/20 10:30 AM, Alex Bennée wrote:
>
> Claudio Fontana <cfontana@suse.de> writes:
>
>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
>> ---
>> target/i386/meson.build | 14 +-------------
>> target/i386/{ => tcg}/bpt_helper.c | 0
>> target/i386/{ => tcg}/cc_helper.c | 0
>> target/i386/{ => tcg}/excp_helper.c | 0
>> target/i386/{ => tcg}/fpu_helper.c | 0
>> target/i386/{ => tcg}/int_helper.c | 0
>> target/i386/{ => tcg}/mem_helper.c | 0
>> target/i386/tcg/meson.build | 13 +++++++++++++
>> target/i386/{ => tcg}/misc_helper.c | 0
>> target/i386/{ => tcg}/mpx_helper.c | 0
>> target/i386/{ => tcg}/seg_helper.c | 0
>> target/i386/{ => tcg}/smm_helper.c | 0
>> target/i386/{ => tcg}/svm_helper.c | 0
>> target/i386/{ => tcg}/tcg-stub.c | 0
>> target/i386/{ => tcg}/translate.c | 0
>> 15 files changed, 14 insertions(+), 13 deletions(-)
>> rename target/i386/{ => tcg}/bpt_helper.c (100%)
>> rename target/i386/{ => tcg}/cc_helper.c (100%)
>> rename target/i386/{ => tcg}/excp_helper.c (100%)
>> rename target/i386/{ => tcg}/fpu_helper.c (100%)
>> rename target/i386/{ => tcg}/int_helper.c (100%)
>> rename target/i386/{ => tcg}/mem_helper.c (100%)
>> create mode 100644 target/i386/tcg/meson.build
>> rename target/i386/{ => tcg}/misc_helper.c (100%)
>> rename target/i386/{ => tcg}/mpx_helper.c (100%)
>> rename target/i386/{ => tcg}/seg_helper.c (100%)
>> rename target/i386/{ => tcg}/smm_helper.c (100%)
>> rename target/i386/{ => tcg}/svm_helper.c (100%)
>> rename target/i386/{ => tcg}/tcg-stub.c (100%)
>> rename target/i386/{ => tcg}/translate.c (100%)
>
> I noticed target/i386/helper-tcg.h is still in the main directory. It
This is something to fix in
i386: move TCG cpu class initialization out of helper.c
will do, thanks!
> maybe be a straggler that is harder to separate - if so might be worth
> mentioning in the commit message. Otherwise:
>
> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
>
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 19/32] cpu: Move cpu_exec_* to tcg_ops
2020-12-08 19:48 ` [RFC v9 19/32] cpu: Move cpu_exec_* " Claudio Fontana
2020-12-09 9:28 ` Philippe Mathieu-Daudé
@ 2020-12-09 11:16 ` Alex Bennée
1 sibling, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 11:16 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> From: Eduardo Habkost <ehabkost@redhat.com>
>
> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
> [claudio: wrapped in CONFIG_TCG]
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 10/32] i386: move TCG cpu class initialization out of helper.c
2020-12-09 10:23 ` Alex Bennée
@ 2020-12-09 11:19 ` Claudio Fontana
0 siblings, 0 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-09 11:19 UTC (permalink / raw)
To: Alex Bennée
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
On 12/9/20 11:23 AM, Alex Bennée wrote:
>
> Claudio Fontana <cfontana@suse.de> writes:
>
>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
>> ---
>> target/i386/cpu.c | 33 ++++------
>> target/i386/cpu.h | 97 ++---------------------------
>> target/i386/helper-tcg.h | 112 ++++++++++++++++++++++++++++++++++
>> target/i386/helper.c | 23 -------
>> target/i386/meson.build | 1 +
>> target/i386/tcg-cpu.c | 71 +++++++++++++++++++++
>> target/i386/tcg-cpu.h | 15 +++++
>> target/i386/tcg/bpt_helper.c | 1 +
>> target/i386/tcg/cc_helper.c | 1 +
>> target/i386/tcg/excp_helper.c | 1 +
>> target/i386/tcg/fpu_helper.c | 33 +++++-----
>> target/i386/tcg/int_helper.c | 1 +
>> target/i386/tcg/mem_helper.c | 1 +
>> target/i386/tcg/misc_helper.c | 1 +
>> target/i386/tcg/mpx_helper.c | 1 +
>> target/i386/tcg/seg_helper.c | 1 +
>> target/i386/tcg/smm_helper.c | 2 +
>> target/i386/tcg/svm_helper.c | 1 +
>> target/i386/tcg/translate.c | 1 +
>> 19 files changed, 244 insertions(+), 153 deletions(-)
>> create mode 100644 target/i386/helper-tcg.h
>> create mode 100644 target/i386/tcg-cpu.c
>> create mode 100644 target/i386/tcg-cpu.h
>>
>> diff --git a/target/i386/cpu.c b/target/i386/cpu.c
>> index b9bd249c8f..3462d0143f 100644
>> --- a/target/i386/cpu.c
>> +++ b/target/i386/cpu.c
>> @@ -24,6 +24,8 @@
>> #include "qemu/qemu-print.h"
>>
>> #include "cpu.h"
>> +#include "tcg-cpu.h"
>> +#include "helper-tcg.h"
>> #include "exec/exec-all.h"
>> #include "sysemu/kvm.h"
>> #include "sysemu/reset.h"
>> @@ -1495,7 +1497,8 @@ static inline uint64_t x86_cpu_xsave_components(X86CPU *cpu)
>> cpu->env.features[FEAT_XSAVE_COMP_LO];
>> }
>>
>> -const char *get_register_name_32(unsigned int reg)
>> +/* Return name of 32-bit register, from a R_* constant */
>> +static const char *get_register_name_32(unsigned int reg)
>> {
>> if (reg >= CPU_NB_REGS32) {
>> return NULL;
>> @@ -7012,13 +7015,6 @@ static void x86_cpu_set_pc(CPUState *cs, vaddr value)
>> cpu->env.eip = value;
>> }
>>
>> -static void x86_cpu_synchronize_from_tb(CPUState *cs, TranslationBlock *tb)
>> -{
>> - X86CPU *cpu = X86_CPU(cs);
>> -
>> - cpu->env.eip = tb->pc - tb->cs_base;
>> -}
>> -
>> int x86_cpu_pending_interrupt(CPUState *cs, int interrupt_request)
>> {
>> X86CPU *cpu = X86_CPU(cs);
>> @@ -7252,17 +7248,18 @@ static void x86_cpu_common_class_init(ObjectClass *oc, void *data)
>> cc->class_by_name = x86_cpu_class_by_name;
>> cc->parse_features = x86_cpu_parse_featurestr;
>> cc->has_work = x86_cpu_has_work;
>> +
>> #ifdef CONFIG_TCG
>> - cc->do_interrupt = x86_cpu_do_interrupt;
>> - cc->cpu_exec_interrupt = x86_cpu_exec_interrupt;
>> -#endif
>> + tcg_cpu_common_class_init(cc);
>
> Are we likely to have clashing names here as other arches get converted?
> tcg_x86_cpu_common_class_init or x86_cpu_common_class_init?
well, not in the same binary I presume? But I can add an x86 in front.
>
> <snip>
>> diff --git a/target/i386/tcg-cpu.c b/target/i386/tcg-cpu.c
>> new file mode 100644
>> index 0000000000..628dd29fe7
>> --- /dev/null
>> +++ b/target/i386/tcg-cpu.c
>> @@ -0,0 +1,71 @@
> <snip>
>> +
>> +void tcg_cpu_common_class_init(CPUClass *cc)
>> +{
>> + cc->do_interrupt = x86_cpu_do_interrupt;
>> + cc->cpu_exec_interrupt = x86_cpu_exec_interrupt;
>> + cc->synchronize_from_tb = x86_cpu_synchronize_from_tb;
>> + cc->cpu_exec_enter = x86_cpu_exec_enter;
>> + cc->cpu_exec_exit = x86_cpu_exec_exit;
>> + cc->tcg_initialize = tcg_x86_init;
>> + cc->tlb_fill = x86_cpu_tlb_fill;
>> +#ifndef CONFIG_USER_ONLY
>> + cc->debug_excp_handler = breakpoint_handler;
>> +#endif
>> +}
>
> Oh I see this moves down to target/i386/tcg/ eventually. Maybe we should
> just create the new file in place so it's easier to follow the changes
> as we convert to a proper abstracted class?
>
Yes, will do.
Thanks,
C
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 17/32] accel/tcg: split TCG-only code from cpu_exec_realizefn
2020-12-09 10:42 ` Alex Bennée
@ 2020-12-09 11:22 ` Claudio Fontana
0 siblings, 0 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-09 11:22 UTC (permalink / raw)
To: Alex Bennée
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
On 12/9/20 11:42 AM, Alex Bennée wrote:
>
> Claudio Fontana <cfontana@suse.de> writes:
>
>> move away TCG-only code, make it compile only on TCG.
>>
>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
>> ---
>> accel/tcg/cpu-exec.c | 28 +++++++++++++++++
>> cpu.c | 70 ++++++++++++++++++++-----------------------
>> hw/core/cpu.c | 6 +++-
>> include/hw/core/cpu.h | 8 +++++
>> 4 files changed, 74 insertions(+), 38 deletions(-)
>>
>> diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
>> index 64cba89356..436dfbf155 100644
>> --- a/accel/tcg/cpu-exec.c
>> +++ b/accel/tcg/cpu-exec.c
>> @@ -801,6 +801,34 @@ int cpu_exec(CPUState *cpu)
>> return ret;
>> }
>>
>> +void tcg_exec_realizefn(CPUState *cpu, Error **errp)
>> +{
>> + static bool tcg_target_initialized;
>> + CPUClass *cc = CPU_GET_CLASS(cpu);
>> +
>> + if (!tcg_target_initialized) {
>> + tcg_target_initialized = true;
>> + cc->tcg_ops.initialize();
>
> nit: it makes no difference but stylistically it makes sense to set
> tcg_target_initialized after we have in fact initialised.
>
> Also we've dropped the tcg_enabled() check,
The tcg_enabled() check should be there, not here but in cpu_exec_realizefn,
if (tcg_enabled()) {
tcg_exec_realizefn(...)
}
will check, thanks!
> if indeed it will always be
> true should we not assert it to ensure the statement for tcg_exec_init
> remains the case: "Must be called before using the QEMU cpus."
>
> Otherwise LGTM:
>
> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
>
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 20/32] cpu: Move tlb_fill to tcg_ops
2020-12-08 19:48 ` [RFC v9 20/32] cpu: Move tlb_fill " Claudio Fontana
@ 2020-12-09 11:26 ` Alex Bennée
2020-12-09 14:38 ` Claudio Fontana
0 siblings, 1 reply; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 11:26 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> From: Eduardo Habkost <ehabkost@redhat.com>
>
> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
> [claudio: wrapped in CONFIG_TCG]
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> ---
> accel/tcg/cputlb.c | 6 +++---
> accel/tcg/user-exec.c | 6 +++---
> include/hw/core/cpu.h | 9 ---------
> include/hw/core/tcg-cpu-ops.h | 12 ++++++++++++
> target/alpha/cpu.c | 2 +-
> target/arm/cpu.c | 2 +-
> target/avr/cpu.c | 2 +-
> target/cris/cpu.c | 2 +-
> target/hppa/cpu.c | 2 +-
> target/i386/tcg-cpu.c | 2 +-
> target/lm32/cpu.c | 2 +-
> target/m68k/cpu.c | 2 +-
> target/microblaze/cpu.c | 2 +-
> target/mips/cpu.c | 2 +-
> target/moxie/cpu.c | 2 +-
> target/nios2/cpu.c | 2 +-
> target/openrisc/cpu.c | 2 +-
> target/ppc/translate_init.c.inc | 2 +-
> target/riscv/cpu.c | 2 +-
> target/rx/cpu.c | 2 +-
> target/s390x/cpu.c | 2 +-
> target/sh4/cpu.c | 2 +-
> target/sparc/cpu.c | 2 +-
> target/tilegx/cpu.c | 2 +-
> target/tricore/cpu.c | 2 +-
> target/unicore32/cpu.c | 2 +-
> target/xtensa/cpu.c | 2 +-
> 27 files changed, 41 insertions(+), 38 deletions(-)
>
> diff --git a/accel/tcg/cputlb.c b/accel/tcg/cputlb.c
> index 42ab79c1a5..2dc71b5528 100644
> --- a/accel/tcg/cputlb.c
> +++ b/accel/tcg/cputlb.c
> @@ -1286,7 +1286,7 @@ static void tlb_fill(CPUState *cpu, target_ulong addr, int size,
> * This is not a probe, so only valid return is success; failure
> * should result in exception + longjmp to the cpu loop.
> */
> - ok = cc->tlb_fill(cpu, addr, size, access_type, mmu_idx, false, retaddr);
> + ok = cc->tcg_ops.tlb_fill(cpu, addr, size, access_type, mmu_idx, false, retaddr);
> assert(ok);
> }
>
> @@ -1557,8 +1557,8 @@ static int probe_access_internal(CPUArchState *env, target_ulong addr,
> CPUState *cs = env_cpu(env);
> CPUClass *cc = CPU_GET_CLASS(cs);
>
> - if (!cc->tlb_fill(cs, addr, fault_size, access_type,
> - mmu_idx, nonfault, retaddr)) {
> + if (!cc->tcg_ops.tlb_fill(cs, addr, fault_size, access_type,
> + mmu_idx, nonfault, retaddr)) {
> /* Non-faulting page table read failed. */
> *phost = NULL;
> return TLB_INVALID_MASK;
> diff --git a/accel/tcg/user-exec.c b/accel/tcg/user-exec.c
> index 4ebe25461a..7f53992251 100644
> --- a/accel/tcg/user-exec.c
> +++ b/accel/tcg/user-exec.c
> @@ -186,7 +186,7 @@ static inline int handle_cpu_signal(uintptr_t pc, siginfo_t *info,
> clear_helper_retaddr();
>
> cc = CPU_GET_CLASS(cpu);
> - cc->tlb_fill(cpu, address, 0, access_type, MMU_USER_IDX, false, pc);
> + cc->tcg_ops.tlb_fill(cpu, address, 0, access_type, MMU_USER_IDX, false, pc);
> g_assert_not_reached();
> }
>
> @@ -216,8 +216,8 @@ static int probe_access_internal(CPUArchState *env, target_ulong addr,
> } else {
> CPUState *cpu = env_cpu(env);
> CPUClass *cc = CPU_GET_CLASS(cpu);
> - cc->tlb_fill(cpu, addr, fault_size, access_type,
> - MMU_USER_IDX, false, ra);
> + cc->tcg_ops.tlb_fill(cpu, addr, fault_size, access_type,
> + MMU_USER_IDX, false, ra);
> g_assert_not_reached();
> }
> }
> diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
> index 52142e9094..c82ef261c6 100644
> --- a/include/hw/core/cpu.h
> +++ b/include/hw/core/cpu.h
> @@ -110,12 +110,6 @@ struct TranslationBlock;
> * If the target behaviour here is anything other than "set
> * the PC register to the value passed in" then the target must
> * also implement the synchronize_from_tb hook.
> - * @tlb_fill: Callback for handling a softmmu tlb miss or user-only
> - * address fault. For system mode, if the access is valid, call
> - * tlb_set_page and return true; if the access is invalid, and
> - * probe is true, return false; otherwise raise an exception and
> - * do not return. For user-only mode, always raise an exception
> - * and do not return.
> * @get_phys_page_debug: Callback for obtaining a physical address.
> * @get_phys_page_attrs_debug: Callback for obtaining a physical address and the
> * associated memory transaction attributes to use for the access.
> @@ -183,9 +177,6 @@ struct CPUClass {
> void (*get_memory_mapping)(CPUState *cpu, MemoryMappingList *list,
> Error **errp);
> void (*set_pc)(CPUState *cpu, vaddr value);
> - bool (*tlb_fill)(CPUState *cpu, vaddr address, int size,
> - MMUAccessType access_type, int mmu_idx,
> - bool probe, uintptr_t retaddr);
> hwaddr (*get_phys_page_debug)(CPUState *cpu, vaddr addr);
> hwaddr (*get_phys_page_attrs_debug)(CPUState *cpu, vaddr addr,
> MemTxAttrs *attrs);
> diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
> index e12f32919b..2ea94acca0 100644
> --- a/include/hw/core/tcg-cpu-ops.h
> +++ b/include/hw/core/tcg-cpu-ops.h
> @@ -37,6 +37,18 @@ typedef struct TcgCpuOperations {
> void (*cpu_exec_exit)(CPUState *cpu);
> /** @cpu_exec_interrupt: Callback for processing interrupts in cpu_exec */
> bool (*cpu_exec_interrupt)(CPUState *cpu, int interrupt_request);
> + /**
> + * @tlb_fill: Handle a softmmu tlb miss or user-only address fault
> + *
> + * For system mode, if the access is valid, call tlb_set_page
> + * and return true; if the access is invalid, and probe is
> + * true, return false; otherwise raise an exception and do
> + * not return. For user-only mode, always raise an exception
> + * and do not return.
> + */
> + bool (*tlb_fill)(CPUState *cpu, vaddr address, int size,
> + MMUAccessType access_type, int mmu_idx,
> + bool probe, uintptr_t retaddr);
As per previous patch, here is a chance to clean-up the comment.
Otherwise:
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 21/32] cpu: Move debug_excp_handler to tcg_ops
2020-12-08 19:48 ` [RFC v9 21/32] cpu: Move debug_excp_handler " Claudio Fontana
@ 2020-12-09 11:29 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 11:29 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> From: Eduardo Habkost <ehabkost@redhat.com>
>
> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 22/32] target/arm: do not use cc->do_interrupt for KVM directly
2020-12-08 19:48 ` [RFC v9 22/32] target/arm: do not use cc->do_interrupt for KVM directly Claudio Fontana
@ 2020-12-09 11:30 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 11:30 UTC (permalink / raw)
To: Claudio Fontana
Cc: Peter Maydell, Paul Durrant, Jason Wang, qemu-devel, Peter Xu,
haxm-team, Colin Xu, Olaf Hering, Stefano Stabellini,
Bruce Rogers, Emilio G . Cota, Anthony Perard,
Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> cc->do_interrupt is in theory a TCG callback used in accel/tcg only,
> to prepare the emulated architecture to take an interrupt as defined
> in the hardware specifications,
>
> but in reality the _do_interrupt style of functions in targets are
> also occasionally reused by KVM to prepare the architecture state in a
> similar way where userspace code has identified that it needs to
> deliver an exception to the guest.
>
> In the case of ARM, that includes:
>
> 1) the vcpu thread got a SIGBUS indicating a memory error,
> and we need to deliver a Synchronous External Abort to the guest to
> let it know about the error.
> 2) the kernel told us about a debug exception (breakpoint, watchpoint)
> but it is not for one of QEMU's own gdbstub breakpoints/watchpoints
> so it must be a breakpoint the guest itself has set up, therefore
> we need to deliver it to the guest.
>
> So in order to reuse code, the same arm_do_interrupt function is used.
> This is all fine, but we need to avoid calling it using the callback
> registered in CPUClass, since that one is now TCG-only.
>
> Fortunately this is easily solved by replacing calls to
> CPUClass::do_interrupt() with explicit calls to arm_do_interrupt().
My ultra-modern gcc10 gentoo box is quick to point out:
64.c.o -c ../../target/arm/kvm64.c
../../target/arm/kvm64.c: In function ‘kvm_inject_arm_sea’:
../../target/arm/kvm64.c:947:15: error: unused variable ‘cc’ [-Werror=unused-variable]
947 | CPUClass *cc = CPU_GET_CLASS(c);
| ^~
../../target/arm/kvm64.c: In function ‘kvm_arm_handle_debug’:
../../target/arm/kvm64.c:1494:15: error: unused variable ‘cc’ [-Werror=unused-variable]
1494 | CPUClass *cc = CPU_GET_CLASS(cs);
| ^~
cc1: all warnings being treated as errors
>
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> Cc: Peter Maydell <peter.maydell@linaro.org>
> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> ---
> target/arm/helper.c | 4 ++++
> target/arm/kvm64.c | 4 ++--
> 2 files changed, 6 insertions(+), 2 deletions(-)
>
> diff --git a/target/arm/helper.c b/target/arm/helper.c
> index 38cd35c049..bebaabf525 100644
> --- a/target/arm/helper.c
> +++ b/target/arm/helper.c
> @@ -9895,6 +9895,10 @@ static void handle_semihosting(CPUState *cs)
> * Do any appropriate logging, handle PSCI calls, and then hand off
> * to the AArch64-entry or AArch32-entry function depending on the
> * target exception level's register width.
> + *
> + * Note: this is used for both TCG (as the do_interrupt tcg op),
> + * and KVM to re-inject guest debug exceptions, and to
> + * inject a Synchronous-External-Abort.
> */
> void arm_cpu_do_interrupt(CPUState *cs)
> {
> diff --git a/target/arm/kvm64.c b/target/arm/kvm64.c
> index f74bac2457..2b17e4203d 100644
> --- a/target/arm/kvm64.c
> +++ b/target/arm/kvm64.c
> @@ -960,7 +960,7 @@ static void kvm_inject_arm_sea(CPUState *c)
>
> env->exception.syndrome = esr;
>
> - cc->do_interrupt(c);
> + arm_cpu_do_interrupt(c);
> }
>
> #define AARCH64_CORE_REG(x) (KVM_REG_ARM64 | KVM_REG_SIZE_U64 | \
> @@ -1545,7 +1545,7 @@ bool kvm_arm_handle_debug(CPUState *cs, struct kvm_debug_exit_arch *debug_exit)
> env->exception.vaddress = debug_exit->far;
> env->exception.target_el = 1;
> qemu_mutex_lock_iothread();
> - cc->do_interrupt(cs);
> + arm_cpu_do_interrupt(cs);
> qemu_mutex_unlock_iothread();
>
> return false;
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 23/32] cpu: move cc->do_interrupt to tcg_ops
2020-12-08 19:48 ` [RFC v9 23/32] cpu: move cc->do_interrupt to tcg_ops Claudio Fontana
@ 2020-12-09 11:43 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 11:43 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 24/32] cpu: move cc->transaction_failed to tcg_ops
2020-12-08 19:48 ` [RFC v9 24/32] cpu: move cc->transaction_failed " Claudio Fontana
2020-12-09 9:31 ` Philippe Mathieu-Daudé
@ 2020-12-09 12:03 ` Alex Bennée
1 sibling, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 12:03 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 25/32] cpu: move do_unaligned_access to tcg_ops
2020-12-08 19:48 ` [RFC v9 25/32] cpu: move do_unaligned_access " Claudio Fontana
@ 2020-12-09 12:47 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 12:47 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> make it consistently SOFTMMU-only.
>
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 26/32] accel: extend AccelState and AccelClass to user-mode
2020-12-08 19:48 ` [RFC v9 26/32] accel: extend AccelState and AccelClass to user-mode Claudio Fontana
@ 2020-12-09 12:51 ` Alex Bennée
2020-12-09 12:58 ` Claudio Fontana
0 siblings, 1 reply; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 12:51 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> ---
> MAINTAINERS | 2 +-
> accel/accel-common.c | 50 ++++++++++++++++++++++++++++++
> accel/{accel.c => accel-softmmu.c} | 27 ++--------------
> accel/accel-user.c | 24 ++++++++++++++
> accel/meson.build | 4 ++-
> accel/qtest/qtest.c | 2 +-
> accel/tcg/meson.build | 2 +-
> accel/tcg/tcg-all.c | 13 ++++++--
> accel/xen/xen-all.c | 2 +-
> bsd-user/main.c | 6 +++-
> include/hw/boards.h | 2 +-
> include/{sysemu => qemu}/accel.h | 14 +++++----
> include/sysemu/hvf.h | 2 +-
> include/sysemu/kvm.h | 2 +-
> include/sysemu/kvm_int.h | 2 +-
> linux-user/main.c | 6 +++-
> softmmu/memory.c | 2 +-
> softmmu/qtest.c | 2 +-
> softmmu/vl.c | 2 +-
> target/i386/hax/hax-all.c | 2 +-
> target/i386/hvf/hvf-i386.h | 2 +-
> target/i386/hvf/hvf.c | 2 +-
> target/i386/hvf/x86_task.c | 2 +-
> target/i386/whpx/whpx-all.c | 2 +-
> 24 files changed, 124 insertions(+), 52 deletions(-)
> create mode 100644 accel/accel-common.c
> rename accel/{accel.c => accel-softmmu.c} (75%)
> create mode 100644 accel/accel-user.c
> rename include/{sysemu => qemu}/accel.h (95%)
>
> diff --git a/MAINTAINERS b/MAINTAINERS
> index d876f504a6..6235dd3a9f 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -434,7 +434,7 @@ Overall
> M: Richard Henderson <richard.henderson@linaro.org>
> R: Paolo Bonzini <pbonzini@redhat.com>
> S: Maintained
> -F: include/sysemu/accel.h
> +F: include/qemu/accel.h
> F: accel/accel.c
> F: accel/Makefile.objs
> F: accel/stubs/Makefile.objs
> diff --git a/accel/accel-common.c b/accel/accel-common.c
> new file mode 100644
> index 0000000000..ddec8cb5ae
> --- /dev/null
> +++ b/accel/accel-common.c
> @@ -0,0 +1,50 @@
> +/*
> + * QEMU accel class, components common to system emulation and user mode
> + *
> + * Copyright (c) 2003-2008 Fabrice Bellard
> + * Copyright (c) 2014 Red Hat Inc.
> + *
> + * Permission is hereby granted, free of charge, to any person obtaining a copy
> + * of this software and associated documentation files (the "Software"), to deal
> + * in the Software without restriction, including without limitation the rights
> + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
> + * copies of the Software, and to permit persons to whom the Software is
> + * furnished to do so, subject to the following conditions:
> + *
> + * The above copyright notice and this permission notice shall be included in
> + * all copies or substantial portions of the Software.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
> + * THE SOFTWARE.
> + */
> +
> +#include "qemu/osdep.h"
> +#include "qemu/accel.h"
> +
> +static const TypeInfo accel_type = {
> + .name = TYPE_ACCEL,
> + .parent = TYPE_OBJECT,
> + .class_size = sizeof(AccelClass),
> + .instance_size = sizeof(AccelState),
> +};
> +
> +/* Lookup AccelClass from opt_name. Returns NULL if not found */
> +AccelClass *accel_find(const char *opt_name)
> +{
> + char *class_name = g_strdup_printf(ACCEL_CLASS_NAME("%s"), opt_name);
> + AccelClass *ac = ACCEL_CLASS(object_class_by_name(class_name));
> + g_free(class_name);
> + return ac;
> +}
> +
> +static void register_accel_types(void)
> +{
> + type_register_static(&accel_type);
> +}
> +
> +type_init(register_accel_types);
> diff --git a/accel/accel.c b/accel/accel-softmmu.c
> similarity index 75%
> rename from accel/accel.c
> rename to accel/accel-softmmu.c
> index cb555e3b06..f89da8f9d1 100644
> --- a/accel/accel.c
> +++ b/accel/accel-softmmu.c
> @@ -1,5 +1,5 @@
> /*
> - * QEMU System Emulator, accelerator interfaces
> + * QEMU accel class, system emulation components
> *
> * Copyright (c) 2003-2008 Fabrice Bellard
> * Copyright (c) 2014 Red Hat Inc.
> @@ -24,28 +24,12 @@
> */
>
> #include "qemu/osdep.h"
> -#include "sysemu/accel.h"
> +#include "qemu/accel.h"
> #include "hw/boards.h"
> #include "sysemu/arch_init.h"
> #include "sysemu/sysemu.h"
> #include "qom/object.h"
>
> -static const TypeInfo accel_type = {
> - .name = TYPE_ACCEL,
> - .parent = TYPE_OBJECT,
> - .class_size = sizeof(AccelClass),
> - .instance_size = sizeof(AccelState),
> -};
> -
> -/* Lookup AccelClass from opt_name. Returns NULL if not found */
> -AccelClass *accel_find(const char *opt_name)
> -{
> - char *class_name = g_strdup_printf(ACCEL_CLASS_NAME("%s"), opt_name);
> - AccelClass *ac = ACCEL_CLASS(object_class_by_name(class_name));
> - g_free(class_name);
> - return ac;
> -}
> -
> int accel_init_machine(AccelState *accel, MachineState *ms)
> {
> AccelClass *acc = ACCEL_GET_CLASS(accel);
> @@ -76,10 +60,3 @@ void accel_setup_post(MachineState *ms)
> acc->setup_post(ms, accel);
> }
> }
> -
> -static void register_accel_types(void)
> -{
> - type_register_static(&accel_type);
> -}
> -
> -type_init(register_accel_types);
> diff --git a/accel/accel-user.c b/accel/accel-user.c
> new file mode 100644
> index 0000000000..26bdda6236
> --- /dev/null
> +++ b/accel/accel-user.c
> @@ -0,0 +1,24 @@
> +/*
> + * QEMU accel class, user-mode components
> + *
> + * Copyright 2020 SUSE LLC
> + *
> + * This work is licensed under the terms of the GNU GPL, version 2 or later.
> + * See the COPYING file in the top-level directory.
> + */
> +
> +#include "qemu/osdep.h"
> +#include "qemu/accel.h"
> +
> +AccelState *current_accel(void)
> +{
> + static AccelState *accel;
> +
> + if (!accel) {
> + AccelClass *ac = accel_find("tcg");
> +
> + g_assert(ac != NULL);
> + accel = ACCEL(object_new_with_class(OBJECT_CLASS(ac)));
> + }
> + return accel;
> +}
> diff --git a/accel/meson.build b/accel/meson.build
> index b26cca227a..b44ba30c86 100644
> --- a/accel/meson.build
> +++ b/accel/meson.build
> @@ -1,4 +1,6 @@
> -softmmu_ss.add(files('accel.c'))
> +specific_ss.add(files('accel-common.c'))
> +softmmu_ss.add(files('accel-softmmu.c'))
> +user_ss.add(files('accel-user.c'))
>
> subdir('qtest')
> subdir('kvm')
> diff --git a/accel/qtest/qtest.c b/accel/qtest/qtest.c
> index b282cea5cf..b4e731cb2b 100644
> --- a/accel/qtest/qtest.c
> +++ b/accel/qtest/qtest.c
> @@ -17,7 +17,7 @@
> #include "qemu/module.h"
> #include "qemu/option.h"
> #include "qemu/config-file.h"
> -#include "sysemu/accel.h"
> +#include "qemu/accel.h"
> #include "sysemu/qtest.h"
> #include "sysemu/cpus.h"
> #include "sysemu/cpu-timers.h"
> diff --git a/accel/tcg/meson.build b/accel/tcg/meson.build
> index f39aab0a0c..424d9bb1fc 100644
> --- a/accel/tcg/meson.build
> +++ b/accel/tcg/meson.build
> @@ -1,5 +1,6 @@
> tcg_ss = ss.source_set()
> tcg_ss.add(files(
> + 'tcg-all.c',
> 'cpu-exec-common.c',
> 'cpu-exec.c',
> 'tcg-runtime-gvec.c',
> @@ -13,7 +14,6 @@ tcg_ss.add(when: 'CONFIG_PLUGIN', if_true: [files('plugin-gen.c'), libdl])
> specific_ss.add_all(when: 'CONFIG_TCG', if_true: tcg_ss)
>
> specific_ss.add(when: ['CONFIG_SOFTMMU', 'CONFIG_TCG'], if_true: files(
> - 'tcg-all.c',
> 'cputlb.c',
> 'tcg-cpus.c',
> 'tcg-cpus-mttcg.c',
> diff --git a/accel/tcg/tcg-all.c b/accel/tcg/tcg-all.c
> index 1ac0b76515..7125d0cc29 100644
> --- a/accel/tcg/tcg-all.c
> +++ b/accel/tcg/tcg-all.c
> @@ -30,9 +30,12 @@
> #include "tcg/tcg.h"
> #include "qapi/error.h"
> #include "qemu/error-report.h"
> -#include "hw/boards.h"
> +#include "qemu/accel.h"
> #include "qapi/qapi-builtin-visit.h"
> +
> +#ifndef CONFIG_USER_ONLY
> #include "tcg-cpus.h"
> +#endif /* CONFIG_USER_ONLY */
>
> struct TCGState {
> AccelState parent_obj;
> @@ -106,8 +109,12 @@ static int tcg_init(MachineState *ms)
> mttcg_enabled = s->mttcg_enabled;
>
> /*
> - * Initialize TCG regions
> + * Initialize TCG regions only for softmmu.
> + *
> + * This needs to be done later for user mode, because the prologue
> + * generation needs to be delayed so that GUEST_BASE is already set.
> */
> +#ifndef CONFIG_USER_ONLY
> tcg_region_init();
This implies we broke bisectability with:
accel/tcg: split tcg_start_vcpu_thread
<snip>
Otherwise with that fixed:
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 27/32] accel: replace struct CpusAccel with AccelOpsClass
2020-12-08 19:48 ` [RFC v9 27/32] accel: replace struct CpusAccel with AccelOpsClass Claudio Fontana
@ 2020-12-09 12:54 ` Alex Bennée
2020-12-09 14:50 ` Claudio Fontana
2020-12-09 17:28 ` Claudio Fontana
0 siblings, 2 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 12:54 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> centralize the registration of the cpus.c module
> accelerator operations in accel/accel-softmmu.c
>
> Signed-off-by: Claudio Fontana <cfontana@suse.de>
<snip>
> diff --git a/accel/tcg/tcg-cpus.c b/accel/tcg/tcg-cpus.c
> index e335f9f155..38a58ab271 100644
> --- a/accel/tcg/tcg-cpus.c
> +++ b/accel/tcg/tcg-cpus.c
> @@ -35,6 +35,9 @@
> #include "hw/boards.h"
>
> #include "tcg-cpus.h"
> +#include "tcg-cpus-mttcg.h"
> +#include "tcg-cpus-rr.h"
> +#include "tcg-cpus-icount.h"
>
> /* common functionality among all TCG variants */
>
> @@ -80,3 +83,43 @@ void tcg_cpus_handle_interrupt(CPUState *cpu, int mask)
> qatomic_set(&cpu_neg(cpu)->icount_decr.u16.high, -1);
> }
> }
> +
> +static void tcg_cpus_ops_init(AccelOpsClass *ops)
> +{
> + if (qemu_tcg_mttcg_enabled()) {
> + ops->create_vcpu_thread = mttcg_start_vcpu_thread;
> + ops->kick_vcpu_thread = mttcg_kick_vcpu_thread;
> + ops->handle_interrupt = tcg_cpus_handle_interrupt;
> +
> + } else if (icount_enabled()) {
> + ops->create_vcpu_thread = rr_start_vcpu_thread;
> + ops->kick_vcpu_thread = rr_kick_vcpu_thread;
> + ops->handle_interrupt = icount_handle_interrupt;
> + ops->get_virtual_clock = icount_get;
> + ops->get_elapsed_ticks = icount_get;
> +
> + } else {
> + ops->create_vcpu_thread = rr_start_vcpu_thread;
> + ops->kick_vcpu_thread = rr_kick_vcpu_thread;
> + ops->handle_interrupt = tcg_cpus_handle_interrupt;
> + }
> +}
Aren't we going backwards here by having a global function aware of the
different accelerator types rather than encapsulating this is the
particular accelerator machinery?
<snip>
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 26/32] accel: extend AccelState and AccelClass to user-mode
2020-12-09 12:51 ` Alex Bennée
@ 2020-12-09 12:58 ` Claudio Fontana
0 siblings, 0 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-09 12:58 UTC (permalink / raw)
To: Alex Bennée
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
On 12/9/20 1:51 PM, Alex Bennée wrote:
>
> Claudio Fontana <cfontana@suse.de> writes:
>
>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
>> ---
>> MAINTAINERS | 2 +-
>> accel/accel-common.c | 50 ++++++++++++++++++++++++++++++
>> accel/{accel.c => accel-softmmu.c} | 27 ++--------------
>> accel/accel-user.c | 24 ++++++++++++++
>> accel/meson.build | 4 ++-
>> accel/qtest/qtest.c | 2 +-
>> accel/tcg/meson.build | 2 +-
>> accel/tcg/tcg-all.c | 13 ++++++--
>> accel/xen/xen-all.c | 2 +-
>> bsd-user/main.c | 6 +++-
>> include/hw/boards.h | 2 +-
>> include/{sysemu => qemu}/accel.h | 14 +++++----
>> include/sysemu/hvf.h | 2 +-
>> include/sysemu/kvm.h | 2 +-
>> include/sysemu/kvm_int.h | 2 +-
>> linux-user/main.c | 6 +++-
>> softmmu/memory.c | 2 +-
>> softmmu/qtest.c | 2 +-
>> softmmu/vl.c | 2 +-
>> target/i386/hax/hax-all.c | 2 +-
>> target/i386/hvf/hvf-i386.h | 2 +-
>> target/i386/hvf/hvf.c | 2 +-
>> target/i386/hvf/x86_task.c | 2 +-
>> target/i386/whpx/whpx-all.c | 2 +-
>> 24 files changed, 124 insertions(+), 52 deletions(-)
>> create mode 100644 accel/accel-common.c
>> rename accel/{accel.c => accel-softmmu.c} (75%)
>> create mode 100644 accel/accel-user.c
>> rename include/{sysemu => qemu}/accel.h (95%)
>>
>> diff --git a/MAINTAINERS b/MAINTAINERS
>> index d876f504a6..6235dd3a9f 100644
>> --- a/MAINTAINERS
>> +++ b/MAINTAINERS
>> @@ -434,7 +434,7 @@ Overall
>> M: Richard Henderson <richard.henderson@linaro.org>
>> R: Paolo Bonzini <pbonzini@redhat.com>
>> S: Maintained
>> -F: include/sysemu/accel.h
>> +F: include/qemu/accel.h
>> F: accel/accel.c
>> F: accel/Makefile.objs
>> F: accel/stubs/Makefile.objs
>> diff --git a/accel/accel-common.c b/accel/accel-common.c
>> new file mode 100644
>> index 0000000000..ddec8cb5ae
>> --- /dev/null
>> +++ b/accel/accel-common.c
>> @@ -0,0 +1,50 @@
>> +/*
>> + * QEMU accel class, components common to system emulation and user mode
>> + *
>> + * Copyright (c) 2003-2008 Fabrice Bellard
>> + * Copyright (c) 2014 Red Hat Inc.
>> + *
>> + * Permission is hereby granted, free of charge, to any person obtaining a copy
>> + * of this software and associated documentation files (the "Software"), to deal
>> + * in the Software without restriction, including without limitation the rights
>> + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
>> + * copies of the Software, and to permit persons to whom the Software is
>> + * furnished to do so, subject to the following conditions:
>> + *
>> + * The above copyright notice and this permission notice shall be included in
>> + * all copies or substantial portions of the Software.
>> + *
>> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
>> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
>> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
>> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
>> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
>> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
>> + * THE SOFTWARE.
>> + */
>> +
>> +#include "qemu/osdep.h"
>> +#include "qemu/accel.h"
>> +
>> +static const TypeInfo accel_type = {
>> + .name = TYPE_ACCEL,
>> + .parent = TYPE_OBJECT,
>> + .class_size = sizeof(AccelClass),
>> + .instance_size = sizeof(AccelState),
>> +};
>> +
>> +/* Lookup AccelClass from opt_name. Returns NULL if not found */
>> +AccelClass *accel_find(const char *opt_name)
>> +{
>> + char *class_name = g_strdup_printf(ACCEL_CLASS_NAME("%s"), opt_name);
>> + AccelClass *ac = ACCEL_CLASS(object_class_by_name(class_name));
>> + g_free(class_name);
>> + return ac;
>> +}
>> +
>> +static void register_accel_types(void)
>> +{
>> + type_register_static(&accel_type);
>> +}
>> +
>> +type_init(register_accel_types);
>> diff --git a/accel/accel.c b/accel/accel-softmmu.c
>> similarity index 75%
>> rename from accel/accel.c
>> rename to accel/accel-softmmu.c
>> index cb555e3b06..f89da8f9d1 100644
>> --- a/accel/accel.c
>> +++ b/accel/accel-softmmu.c
>> @@ -1,5 +1,5 @@
>> /*
>> - * QEMU System Emulator, accelerator interfaces
>> + * QEMU accel class, system emulation components
>> *
>> * Copyright (c) 2003-2008 Fabrice Bellard
>> * Copyright (c) 2014 Red Hat Inc.
>> @@ -24,28 +24,12 @@
>> */
>>
>> #include "qemu/osdep.h"
>> -#include "sysemu/accel.h"
>> +#include "qemu/accel.h"
>> #include "hw/boards.h"
>> #include "sysemu/arch_init.h"
>> #include "sysemu/sysemu.h"
>> #include "qom/object.h"
>>
>> -static const TypeInfo accel_type = {
>> - .name = TYPE_ACCEL,
>> - .parent = TYPE_OBJECT,
>> - .class_size = sizeof(AccelClass),
>> - .instance_size = sizeof(AccelState),
>> -};
>> -
>> -/* Lookup AccelClass from opt_name. Returns NULL if not found */
>> -AccelClass *accel_find(const char *opt_name)
>> -{
>> - char *class_name = g_strdup_printf(ACCEL_CLASS_NAME("%s"), opt_name);
>> - AccelClass *ac = ACCEL_CLASS(object_class_by_name(class_name));
>> - g_free(class_name);
>> - return ac;
>> -}
>> -
>> int accel_init_machine(AccelState *accel, MachineState *ms)
>> {
>> AccelClass *acc = ACCEL_GET_CLASS(accel);
>> @@ -76,10 +60,3 @@ void accel_setup_post(MachineState *ms)
>> acc->setup_post(ms, accel);
>> }
>> }
>> -
>> -static void register_accel_types(void)
>> -{
>> - type_register_static(&accel_type);
>> -}
>> -
>> -type_init(register_accel_types);
>> diff --git a/accel/accel-user.c b/accel/accel-user.c
>> new file mode 100644
>> index 0000000000..26bdda6236
>> --- /dev/null
>> +++ b/accel/accel-user.c
>> @@ -0,0 +1,24 @@
>> +/*
>> + * QEMU accel class, user-mode components
>> + *
>> + * Copyright 2020 SUSE LLC
>> + *
>> + * This work is licensed under the terms of the GNU GPL, version 2 or later.
>> + * See the COPYING file in the top-level directory.
>> + */
>> +
>> +#include "qemu/osdep.h"
>> +#include "qemu/accel.h"
>> +
>> +AccelState *current_accel(void)
>> +{
>> + static AccelState *accel;
>> +
>> + if (!accel) {
>> + AccelClass *ac = accel_find("tcg");
>> +
>> + g_assert(ac != NULL);
>> + accel = ACCEL(object_new_with_class(OBJECT_CLASS(ac)));
>> + }
>> + return accel;
>> +}
>> diff --git a/accel/meson.build b/accel/meson.build
>> index b26cca227a..b44ba30c86 100644
>> --- a/accel/meson.build
>> +++ b/accel/meson.build
>> @@ -1,4 +1,6 @@
>> -softmmu_ss.add(files('accel.c'))
>> +specific_ss.add(files('accel-common.c'))
>> +softmmu_ss.add(files('accel-softmmu.c'))
>> +user_ss.add(files('accel-user.c'))
>>
>> subdir('qtest')
>> subdir('kvm')
>> diff --git a/accel/qtest/qtest.c b/accel/qtest/qtest.c
>> index b282cea5cf..b4e731cb2b 100644
>> --- a/accel/qtest/qtest.c
>> +++ b/accel/qtest/qtest.c
>> @@ -17,7 +17,7 @@
>> #include "qemu/module.h"
>> #include "qemu/option.h"
>> #include "qemu/config-file.h"
>> -#include "sysemu/accel.h"
>> +#include "qemu/accel.h"
>> #include "sysemu/qtest.h"
>> #include "sysemu/cpus.h"
>> #include "sysemu/cpu-timers.h"
>> diff --git a/accel/tcg/meson.build b/accel/tcg/meson.build
>> index f39aab0a0c..424d9bb1fc 100644
>> --- a/accel/tcg/meson.build
>> +++ b/accel/tcg/meson.build
>> @@ -1,5 +1,6 @@
>> tcg_ss = ss.source_set()
>> tcg_ss.add(files(
>> + 'tcg-all.c',
>> 'cpu-exec-common.c',
>> 'cpu-exec.c',
>> 'tcg-runtime-gvec.c',
>> @@ -13,7 +14,6 @@ tcg_ss.add(when: 'CONFIG_PLUGIN', if_true: [files('plugin-gen.c'), libdl])
>> specific_ss.add_all(when: 'CONFIG_TCG', if_true: tcg_ss)
>>
>> specific_ss.add(when: ['CONFIG_SOFTMMU', 'CONFIG_TCG'], if_true: files(
>> - 'tcg-all.c',
>> 'cputlb.c',
>> 'tcg-cpus.c',
>> 'tcg-cpus-mttcg.c',
>> diff --git a/accel/tcg/tcg-all.c b/accel/tcg/tcg-all.c
>> index 1ac0b76515..7125d0cc29 100644
>> --- a/accel/tcg/tcg-all.c
>> +++ b/accel/tcg/tcg-all.c
>> @@ -30,9 +30,12 @@
>> #include "tcg/tcg.h"
>> #include "qapi/error.h"
>> #include "qemu/error-report.h"
>> -#include "hw/boards.h"
>> +#include "qemu/accel.h"
>> #include "qapi/qapi-builtin-visit.h"
>> +
>> +#ifndef CONFIG_USER_ONLY
>> #include "tcg-cpus.h"
>> +#endif /* CONFIG_USER_ONLY */
>>
>> struct TCGState {
>> AccelState parent_obj;
>> @@ -106,8 +109,12 @@ static int tcg_init(MachineState *ms)
>> mttcg_enabled = s->mttcg_enabled;
>>
>> /*
>> - * Initialize TCG regions
>> + * Initialize TCG regions only for softmmu.
>> + *
>> + * This needs to be done later for user mode, because the prologue
>> + * generation needs to be delayed so that GUEST_BASE is already set.
>> */
>> +#ifndef CONFIG_USER_ONLY
>> tcg_region_init();
>
> This implies we broke bisectability with:
>
> accel/tcg: split tcg_start_vcpu_thread
Hi Alex,
I don't think so, this is only an issue after applying this patch,
because tcg_init is only called by softmmu code up and until this very patch,
bisect should be ok..
Ciao,
Claudio
>
> <snip>
>
> Otherwise with that fixed:
>
> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
>
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 00/22] i386 cleanup
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
` (33 preceding siblings ...)
2020-12-08 22:00 ` no-reply
@ 2020-12-09 12:58 ` Alex Bennée
2020-12-09 14:10 ` Claudio Fontana
34 siblings, 1 reply; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 12:58 UTC (permalink / raw)
To: Claudio Fontana
Cc: Peter Maydell, Paul Durrant, Jason Wang, qemu-devel, Peter Xu,
haxm-team, Colin Xu, Olaf Hering, Stefano Stabellini,
Bruce Rogers, Emilio G . Cota, Anthony Perard,
Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
<snip>
>
> Looking forward to your comments on this proposal,
OK I've completed my pass although I petered out a little just before
the changes that affect the explicit ordering which was a bit too much
for my brain to process after lunch. Some general thoughts...
I like the way this is going and having the much more modular concept
for both accelerators and target accelerator support. However to protect
the new build combinations I think we need to cover some TCG and HW
accelerator builds in the gitlab build matrix.
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 00/22] i386 cleanup
2020-12-09 12:58 ` Alex Bennée
@ 2020-12-09 14:10 ` Claudio Fontana
0 siblings, 0 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-09 14:10 UTC (permalink / raw)
To: Alex Bennée
Cc: Peter Maydell, Paul Durrant, Jason Wang, qemu-devel, Peter Xu,
haxm-team, Colin Xu, Olaf Hering, Stefano Stabellini,
Bruce Rogers, Emilio G . Cota, Anthony Perard,
Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
On 12/9/20 1:58 PM, Alex Bennée wrote:
>
> Claudio Fontana <cfontana@suse.de> writes:
>
> <snip>
>>
>> Looking forward to your comments on this proposal,
>
> OK I've completed my pass although I petered out a little just before
> the changes that affect the explicit ordering which was a bit too much
> for my brain to process after lunch. Some general thoughts...
>
> I like the way this is going and having the much more modular concept
> for both accelerators and target accelerator support. However to protect
> the new build combinations I think we need to cover some TCG and HW
> accelerator builds in the gitlab build matrix.
>
Hi Alex,
thanks a lot for your review pass,
will start addressing your comments in the next spin,
Ciao!
Claudio
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 18/32] cpu: Move synchronize_from_tb() to tcg_ops
2020-12-09 9:27 ` Philippe Mathieu-Daudé
@ 2020-12-09 14:33 ` Claudio Fontana
2020-12-09 15:06 ` Philippe Mathieu-Daudé
0 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-09 14:33 UTC (permalink / raw)
To: Philippe Mathieu-Daudé,
Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Emilio G . Cota, haxm-team, Cameron Esfahani, Anthony Perard,
Bruce Rogers, Olaf Hering, Colin Xu
Hi Philippe,
On 12/9/20 10:27 AM, Philippe Mathieu-Daudé wrote:
> On 12/8/20 8:48 PM, Claudio Fontana wrote:
>> From: Eduardo Habkost <ehabkost@redhat.com>
>>
>> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
>> [claudio: wrapped in CONFIG_TCG]
>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
>> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
>> ---
> ...
>> diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
>> index 4475ef0996..109291ac52 100644
>> --- a/include/hw/core/tcg-cpu-ops.h
>> +++ b/include/hw/core/tcg-cpu-ops.h
>> @@ -10,9 +10,6 @@
>> #ifndef TCG_CPU_OPS_H
>> #define TCG_CPU_OPS_H
>
> Missing for TranslationBlock:
>
> #include "hw/core/cpu.h"
this file though is only included from hw/core/cpu.h itself under CONFIG_TCG..
>
>>
>> -/**
>> - * struct TcgCpuOperations: TCG operations specific to a CPU class
>> - */
>
> Why remove the comment?
a mistake, will fix.
>
>> typedef struct TcgCpuOperations {
>> /**
>> * @initialize: Initalize TCG state
>> @@ -20,6 +17,17 @@ typedef struct TcgCpuOperations {
>> * Called when the first CPU is realized.
>> */
>> void (*initialize)(void);
>> + /**
>> + * @synchronize_from_tb: Synchronize state from a TCG #TranslationBlock
>> + *
>> + * This is called when we abandon execution of a TB before
>> + * starting it, and must set all parts of the CPU state which
>> + * the previous TB in the chain may not have updated. This
>> + * will need to do more. If this hook is not implemented then
>> + * the default is to call
>> + * @set_pc(tb->pc).
>> + */
>> + void (*synchronize_from_tb)(CPUState *cpu, struct TranslationBlock *tb);
>> } TcgCpuOperations;
> ...
>
>
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 20/32] cpu: Move tlb_fill to tcg_ops
2020-12-09 11:26 ` Alex Bennée
@ 2020-12-09 14:38 ` Claudio Fontana
2020-12-09 16:12 ` Alex Bennée
0 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-09 14:38 UTC (permalink / raw)
To: Alex Bennée
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
On 12/9/20 12:26 PM, Alex Bennée wrote:
>
> Claudio Fontana <cfontana@suse.de> writes:
>
>> From: Eduardo Habkost <ehabkost@redhat.com>
>>
>> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
>> [claudio: wrapped in CONFIG_TCG]
>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
>> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
>> ---
>> accel/tcg/cputlb.c | 6 +++---
>> accel/tcg/user-exec.c | 6 +++---
>> include/hw/core/cpu.h | 9 ---------
>> include/hw/core/tcg-cpu-ops.h | 12 ++++++++++++
>> target/alpha/cpu.c | 2 +-
>> target/arm/cpu.c | 2 +-
>> target/avr/cpu.c | 2 +-
>> target/cris/cpu.c | 2 +-
>> target/hppa/cpu.c | 2 +-
>> target/i386/tcg-cpu.c | 2 +-
>> target/lm32/cpu.c | 2 +-
>> target/m68k/cpu.c | 2 +-
>> target/microblaze/cpu.c | 2 +-
>> target/mips/cpu.c | 2 +-
>> target/moxie/cpu.c | 2 +-
>> target/nios2/cpu.c | 2 +-
>> target/openrisc/cpu.c | 2 +-
>> target/ppc/translate_init.c.inc | 2 +-
>> target/riscv/cpu.c | 2 +-
>> target/rx/cpu.c | 2 +-
>> target/s390x/cpu.c | 2 +-
>> target/sh4/cpu.c | 2 +-
>> target/sparc/cpu.c | 2 +-
>> target/tilegx/cpu.c | 2 +-
>> target/tricore/cpu.c | 2 +-
>> target/unicore32/cpu.c | 2 +-
>> target/xtensa/cpu.c | 2 +-
>> 27 files changed, 41 insertions(+), 38 deletions(-)
>>
>> diff --git a/accel/tcg/cputlb.c b/accel/tcg/cputlb.c
>> index 42ab79c1a5..2dc71b5528 100644
>> --- a/accel/tcg/cputlb.c
>> +++ b/accel/tcg/cputlb.c
>> @@ -1286,7 +1286,7 @@ static void tlb_fill(CPUState *cpu, target_ulong addr, int size,
>> * This is not a probe, so only valid return is success; failure
>> * should result in exception + longjmp to the cpu loop.
>> */
>> - ok = cc->tlb_fill(cpu, addr, size, access_type, mmu_idx, false, retaddr);
>> + ok = cc->tcg_ops.tlb_fill(cpu, addr, size, access_type, mmu_idx, false, retaddr);
>> assert(ok);
>> }
>>
>> @@ -1557,8 +1557,8 @@ static int probe_access_internal(CPUArchState *env, target_ulong addr,
>> CPUState *cs = env_cpu(env);
>> CPUClass *cc = CPU_GET_CLASS(cs);
>>
>> - if (!cc->tlb_fill(cs, addr, fault_size, access_type,
>> - mmu_idx, nonfault, retaddr)) {
>> + if (!cc->tcg_ops.tlb_fill(cs, addr, fault_size, access_type,
>> + mmu_idx, nonfault, retaddr)) {
>> /* Non-faulting page table read failed. */
>> *phost = NULL;
>> return TLB_INVALID_MASK;
>> diff --git a/accel/tcg/user-exec.c b/accel/tcg/user-exec.c
>> index 4ebe25461a..7f53992251 100644
>> --- a/accel/tcg/user-exec.c
>> +++ b/accel/tcg/user-exec.c
>> @@ -186,7 +186,7 @@ static inline int handle_cpu_signal(uintptr_t pc, siginfo_t *info,
>> clear_helper_retaddr();
>>
>> cc = CPU_GET_CLASS(cpu);
>> - cc->tlb_fill(cpu, address, 0, access_type, MMU_USER_IDX, false, pc);
>> + cc->tcg_ops.tlb_fill(cpu, address, 0, access_type, MMU_USER_IDX, false, pc);
>> g_assert_not_reached();
>> }
>>
>> @@ -216,8 +216,8 @@ static int probe_access_internal(CPUArchState *env, target_ulong addr,
>> } else {
>> CPUState *cpu = env_cpu(env);
>> CPUClass *cc = CPU_GET_CLASS(cpu);
>> - cc->tlb_fill(cpu, addr, fault_size, access_type,
>> - MMU_USER_IDX, false, ra);
>> + cc->tcg_ops.tlb_fill(cpu, addr, fault_size, access_type,
>> + MMU_USER_IDX, false, ra);
>> g_assert_not_reached();
>> }
>> }
>> diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
>> index 52142e9094..c82ef261c6 100644
>> --- a/include/hw/core/cpu.h
>> +++ b/include/hw/core/cpu.h
>> @@ -110,12 +110,6 @@ struct TranslationBlock;
>> * If the target behaviour here is anything other than "set
>> * the PC register to the value passed in" then the target must
>> * also implement the synchronize_from_tb hook.
>> - * @tlb_fill: Callback for handling a softmmu tlb miss or user-only
>> - * address fault. For system mode, if the access is valid, call
>> - * tlb_set_page and return true; if the access is invalid, and
>> - * probe is true, return false; otherwise raise an exception and
>> - * do not return. For user-only mode, always raise an exception
>> - * and do not return.
>> * @get_phys_page_debug: Callback for obtaining a physical address.
>> * @get_phys_page_attrs_debug: Callback for obtaining a physical address and the
>> * associated memory transaction attributes to use for the access.
>> @@ -183,9 +177,6 @@ struct CPUClass {
>> void (*get_memory_mapping)(CPUState *cpu, MemoryMappingList *list,
>> Error **errp);
>> void (*set_pc)(CPUState *cpu, vaddr value);
>> - bool (*tlb_fill)(CPUState *cpu, vaddr address, int size,
>> - MMUAccessType access_type, int mmu_idx,
>> - bool probe, uintptr_t retaddr);
>> hwaddr (*get_phys_page_debug)(CPUState *cpu, vaddr addr);
>> hwaddr (*get_phys_page_attrs_debug)(CPUState *cpu, vaddr addr,
>> MemTxAttrs *attrs);
>> diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
>> index e12f32919b..2ea94acca0 100644
>> --- a/include/hw/core/tcg-cpu-ops.h
>> +++ b/include/hw/core/tcg-cpu-ops.h
>> @@ -37,6 +37,18 @@ typedef struct TcgCpuOperations {
>> void (*cpu_exec_exit)(CPUState *cpu);
>> /** @cpu_exec_interrupt: Callback for processing interrupts in cpu_exec */
>> bool (*cpu_exec_interrupt)(CPUState *cpu, int interrupt_request);
>> + /**
>> + * @tlb_fill: Handle a softmmu tlb miss or user-only address fault
>> + *
>> + * For system mode, if the access is valid, call tlb_set_page
>> + * and return true; if the access is invalid, and probe is
>> + * true, return false; otherwise raise an exception and do
>> + * not return. For user-only mode, always raise an exception
>> + * and do not return.
>> + */
>> + bool (*tlb_fill)(CPUState *cpu, vaddr address, int size,
>> + MMUAccessType access_type, int mmu_idx,
>> + bool probe, uintptr_t retaddr);
>
> As per previous patch, here is a chance to clean-up the comment.
Could you provide the text? I think you understand this better than I do...
>
> Otherwise:
>
> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
>
Thanks!
Claudio
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 24/32] cpu: move cc->transaction_failed to tcg_ops
2020-12-09 9:31 ` Philippe Mathieu-Daudé
@ 2020-12-09 14:43 ` Claudio Fontana
2020-12-09 14:59 ` Eduardo Habkost
0 siblings, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-09 14:43 UTC (permalink / raw)
To: Philippe Mathieu-Daudé,
Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Emilio G . Cota, haxm-team, Cameron Esfahani, Anthony Perard,
Bruce Rogers, Olaf Hering, Colin Xu
On 12/9/20 10:31 AM, Philippe Mathieu-Daudé wrote:
> On 12/8/20 8:48 PM, Claudio Fontana wrote:
>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
>> ---
> ...
>> diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
>> index 6b9b72e785..0cc0927738 100644
>> --- a/include/hw/core/tcg-cpu-ops.h
>> +++ b/include/hw/core/tcg-cpu-ops.h
>
> Missing (for MemTxAttrs/MemTxResult):
>
> #include "exec/memattrs.h"
right, this is included only from cpu.h though, which already does it, should we still repeat the include here?
Thanks,
Claudio
>
>> @@ -40,6 +40,15 @@ typedef struct TcgCpuOperations {
>> /** @do_interrupt: Callback for interrupt handling. */
>> void (*do_interrupt)(CPUState *cpu);
>>
>> + /**
>> + * @do_transaction_failed: Callback for handling failed memory transactions
>> + * (ie bus faults or external aborts; not MMU faults)
>> + */
>> + void (*do_transaction_failed)(CPUState *cpu, hwaddr physaddr, vaddr addr,
>> + unsigned size, MMUAccessType access_type,
>> + int mmu_idx, MemTxAttrs attrs,
>> + MemTxResult response, uintptr_t retaddr);
>> +
> ...
>
>
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 18/32] cpu: Move synchronize_from_tb() to tcg_ops
2020-12-09 10:50 ` Alex Bennée
@ 2020-12-09 14:46 ` Eduardo Habkost
2020-12-09 15:51 ` Alex Bennée
0 siblings, 1 reply; 87+ messages in thread
From: Eduardo Habkost @ 2020-12-09 14:46 UTC (permalink / raw)
To: Alex Bennée
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Claudio Fontana, Anthony Perard,
Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Richard Henderson, Cameron Esfahani,
Dario Faggioli, Roman Bolshakov, Sunil Muthuswamy,
Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
On Wed, Dec 09, 2020 at 10:50:29AM +0000, Alex Bennée wrote:
>
> Claudio Fontana <cfontana@suse.de> writes:
>
> > From: Eduardo Habkost <ehabkost@redhat.com>
> >
> > Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
> > [claudio: wrapped in CONFIG_TCG]
> > Signed-off-by: Claudio Fontana <cfontana@suse.de>
> > Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> > ---
[...]
> > @@ -20,6 +17,17 @@ typedef struct TcgCpuOperations {
> > * Called when the first CPU is realized.
> > */
> > void (*initialize)(void);
> > + /**
> > + * @synchronize_from_tb: Synchronize state from a TCG #TranslationBlock
> > + *
> > + * This is called when we abandon execution of a TB before
> > + * starting it, and must set all parts of the CPU state which
> > + * the previous TB in the chain may not have updated. This
> > + * will need to do more. If this hook is not implemented then
> > + * the default is to call
> > + * @set_pc(tb->pc).
> > + */
>
> We could clean-up the comment here to better fit the javadoc style API
> documentation format, i.e.:
>
> /**
> * synchronize_from_tb: Synchronize current TCG execution state
> * @cpu: current cpu
> * @tb: about to be executed #TranslationBlock
> *
> * This is called when we abandon execution of a TB before
> * starting it, and must set all parts of the CPU state which
> * the previous TB in the chain may not have updated. This
> * will need to do more. If this hook is not implemented then
> * the default is to call @set_pc(tb->pc).
> */
Would this work with kernel-doc? This will be parsed as a
portion of the documentation of struct TcgCpuOperations, so I was
not expecting kernel-doc to treat the "@cpu: ..." and "@tb: ..."
lines as valid.
--
Eduardo
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 27/32] accel: replace struct CpusAccel with AccelOpsClass
2020-12-09 12:54 ` Alex Bennée
@ 2020-12-09 14:50 ` Claudio Fontana
2020-12-09 17:28 ` Claudio Fontana
1 sibling, 0 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-09 14:50 UTC (permalink / raw)
To: Alex Bennée
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
On 12/9/20 1:54 PM, Alex Bennée wrote:
>
> Claudio Fontana <cfontana@suse.de> writes:
>
>> centralize the registration of the cpus.c module
>> accelerator operations in accel/accel-softmmu.c
>>
>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> <snip>
>> diff --git a/accel/tcg/tcg-cpus.c b/accel/tcg/tcg-cpus.c
>> index e335f9f155..38a58ab271 100644
>> --- a/accel/tcg/tcg-cpus.c
>> +++ b/accel/tcg/tcg-cpus.c
>> @@ -35,6 +35,9 @@
>> #include "hw/boards.h"
>>
>> #include "tcg-cpus.h"
>> +#include "tcg-cpus-mttcg.h"
>> +#include "tcg-cpus-rr.h"
>> +#include "tcg-cpus-icount.h"
>>
>> /* common functionality among all TCG variants */
>>
>> @@ -80,3 +83,43 @@ void tcg_cpus_handle_interrupt(CPUState *cpu, int mask)
>> qatomic_set(&cpu_neg(cpu)->icount_decr.u16.high, -1);
>> }
>> }
>> +
>> +static void tcg_cpus_ops_init(AccelOpsClass *ops)
>> +{
>> + if (qemu_tcg_mttcg_enabled()) {
>> + ops->create_vcpu_thread = mttcg_start_vcpu_thread;
>> + ops->kick_vcpu_thread = mttcg_kick_vcpu_thread;
>> + ops->handle_interrupt = tcg_cpus_handle_interrupt;
>> +
>> + } else if (icount_enabled()) {
>> + ops->create_vcpu_thread = rr_start_vcpu_thread;
>> + ops->kick_vcpu_thread = rr_kick_vcpu_thread;
>> + ops->handle_interrupt = icount_handle_interrupt;
>> + ops->get_virtual_clock = icount_get;
>> + ops->get_elapsed_ticks = icount_get;
>> +
>> + } else {
>> + ops->create_vcpu_thread = rr_start_vcpu_thread;
>> + ops->kick_vcpu_thread = rr_kick_vcpu_thread;
>> + ops->handle_interrupt = tcg_cpus_handle_interrupt;
>> + }
>> +}
>
> Aren't we going backwards here by having a global function aware of the
> different accelerator types rather than encapsulating this is the
> particular accelerator machinery?
>
> <snip>
>
I don't fully understand what you mean here, which is the global function, what is the particular accelerator machinery?
Could you suggest an alternative for my understanding?
Thanks a lot,
Claudio
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 24/32] cpu: move cc->transaction_failed to tcg_ops
2020-12-09 14:43 ` Claudio Fontana
@ 2020-12-09 14:59 ` Eduardo Habkost
0 siblings, 0 replies; 87+ messages in thread
From: Eduardo Habkost @ 2020-12-09 14:59 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Richard Henderson, Cameron Esfahani,
Dario Faggioli, Roman Bolshakov, Sunil Muthuswamy,
Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
On Wed, Dec 09, 2020 at 03:43:54PM +0100, Claudio Fontana wrote:
> On 12/9/20 10:31 AM, Philippe Mathieu-Daudé wrote:
> > On 12/8/20 8:48 PM, Claudio Fontana wrote:
> >> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> >> ---
> > ...
> >> diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
> >> index 6b9b72e785..0cc0927738 100644
> >> --- a/include/hw/core/tcg-cpu-ops.h
> >> +++ b/include/hw/core/tcg-cpu-ops.h
> >
> > Missing (for MemTxAttrs/MemTxResult):
> >
> > #include "exec/memattrs.h"
>
>
> right, this is included only from cpu.h though, which already does it, should we still repeat the include here?
It doesn't hurt, and may save some headache in the future. A .c
file should be able to include tcg-cpu-ops.h directly if
necessary.
--
Eduardo
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 18/32] cpu: Move synchronize_from_tb() to tcg_ops
2020-12-09 14:33 ` Claudio Fontana
@ 2020-12-09 15:06 ` Philippe Mathieu-Daudé
0 siblings, 0 replies; 87+ messages in thread
From: Philippe Mathieu-Daudé @ 2020-12-09 15:06 UTC (permalink / raw)
To: Claudio Fontana, Paolo Bonzini, Thomas Huth, Richard Henderson,
Stefano Stabellini, Wenchao Wang, Roman Bolshakov,
Sunil Muthuswamy
Cc: Laurent Vivier, Eduardo Habkost, Paul Durrant, Jason Wang,
Marcelo Tosatti, qemu-devel, Peter Xu, Dario Faggioli,
Emilio G . Cota, haxm-team, Cameron Esfahani, Anthony Perard,
Bruce Rogers, Olaf Hering, Colin Xu
On 12/9/20 3:33 PM, Claudio Fontana wrote:
> Hi Philippe,
>
> On 12/9/20 10:27 AM, Philippe Mathieu-Daudé wrote:
>> On 12/8/20 8:48 PM, Claudio Fontana wrote:
>>> From: Eduardo Habkost <ehabkost@redhat.com>
>>>
>>> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
>>> [claudio: wrapped in CONFIG_TCG]
>>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
>>> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
>>> ---
>> ...
>>> diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
>>> index 4475ef0996..109291ac52 100644
>>> --- a/include/hw/core/tcg-cpu-ops.h
>>> +++ b/include/hw/core/tcg-cpu-ops.h
>>> @@ -10,9 +10,6 @@
>>> #ifndef TCG_CPU_OPS_H
>>> #define TCG_CPU_OPS_H
>>
>> Missing for TranslationBlock:
>>
>> #include "hw/core/cpu.h"
>
>
> this file though is only included from hw/core/cpu.h itself under CONFIG_TCG..
1/ we never know how it will be included, includes should
be self-contained IMO.
2/ my editor isn't that smart and can't figure out the types.
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 18/32] cpu: Move synchronize_from_tb() to tcg_ops
2020-12-09 14:46 ` Eduardo Habkost
@ 2020-12-09 15:51 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 15:51 UTC (permalink / raw)
To: Eduardo Habkost
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Claudio Fontana, Anthony Perard,
Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Richard Henderson, Cameron Esfahani,
Dario Faggioli, Roman Bolshakov, Sunil Muthuswamy,
Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Eduardo Habkost <ehabkost@redhat.com> writes:
> On Wed, Dec 09, 2020 at 10:50:29AM +0000, Alex Bennée wrote:
>>
>> Claudio Fontana <cfontana@suse.de> writes:
>>
>> > From: Eduardo Habkost <ehabkost@redhat.com>
>> >
>> > Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
>> > [claudio: wrapped in CONFIG_TCG]
>> > Signed-off-by: Claudio Fontana <cfontana@suse.de>
>> > Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
>> > ---
> [...]
>> > @@ -20,6 +17,17 @@ typedef struct TcgCpuOperations {
>> > * Called when the first CPU is realized.
>> > */
>> > void (*initialize)(void);
>> > + /**
>> > + * @synchronize_from_tb: Synchronize state from a TCG #TranslationBlock
>> > + *
>> > + * This is called when we abandon execution of a TB before
>> > + * starting it, and must set all parts of the CPU state which
>> > + * the previous TB in the chain may not have updated. This
>> > + * will need to do more. If this hook is not implemented then
>> > + * the default is to call
>> > + * @set_pc(tb->pc).
>> > + */
>>
>> We could clean-up the comment here to better fit the javadoc style API
>> documentation format, i.e.:
>>
>> /**
>> * synchronize_from_tb: Synchronize current TCG execution state
>> * @cpu: current cpu
>> * @tb: about to be executed #TranslationBlock
>> *
>> * This is called when we abandon execution of a TB before
>> * starting it, and must set all parts of the CPU state which
>> * the previous TB in the chain may not have updated. This
>> * will need to do more. If this hook is not implemented then
>> * the default is to call @set_pc(tb->pc).
>> */
>
> Would this work with kernel-doc? This will be parsed as a
> portion of the documentation of struct TcgCpuOperations, so I was
> not expecting kernel-doc to treat the "@cpu: ..." and "@tb: ..."
> lines as valid.
Hmm looking at the guides it seems they wouldn't be processed:
https://www.kernel.org/doc/html/v4.18/doc-guide/kernel-doc.html#structure-union-and-enumeration-documentation
I must admit I had previously just used exec-all.h as my template for
what a well formed function document comment is meant to look like. It
seems for it to work properly you need to use function pointer typedefs
and document those.
However looking at which headers we convert into docs at the moment this
is all moot as we don't process these headers at the moment. IOW ignore
this request ;-)
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 20/32] cpu: Move tlb_fill to tcg_ops
2020-12-09 14:38 ` Claudio Fontana
@ 2020-12-09 16:12 ` Alex Bennée
0 siblings, 0 replies; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 16:12 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> On 12/9/20 12:26 PM, Alex Bennée wrote:
>>
>> Claudio Fontana <cfontana@suse.de> writes:
>>
>>> From: Eduardo Habkost <ehabkost@redhat.com>
>>>
>>> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
>>> [claudio: wrapped in CONFIG_TCG]
>>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
>>> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
>>> ---
>>> accel/tcg/cputlb.c | 6 +++---
>>> accel/tcg/user-exec.c | 6 +++---
>>> include/hw/core/cpu.h | 9 ---------
>>> include/hw/core/tcg-cpu-ops.h | 12 ++++++++++++
>>> target/alpha/cpu.c | 2 +-
>>> target/arm/cpu.c | 2 +-
>>> target/avr/cpu.c | 2 +-
>>> target/cris/cpu.c | 2 +-
>>> target/hppa/cpu.c | 2 +-
>>> target/i386/tcg-cpu.c | 2 +-
>>> target/lm32/cpu.c | 2 +-
>>> target/m68k/cpu.c | 2 +-
>>> target/microblaze/cpu.c | 2 +-
>>> target/mips/cpu.c | 2 +-
>>> target/moxie/cpu.c | 2 +-
>>> target/nios2/cpu.c | 2 +-
>>> target/openrisc/cpu.c | 2 +-
>>> target/ppc/translate_init.c.inc | 2 +-
>>> target/riscv/cpu.c | 2 +-
>>> target/rx/cpu.c | 2 +-
>>> target/s390x/cpu.c | 2 +-
>>> target/sh4/cpu.c | 2 +-
>>> target/sparc/cpu.c | 2 +-
>>> target/tilegx/cpu.c | 2 +-
>>> target/tricore/cpu.c | 2 +-
>>> target/unicore32/cpu.c | 2 +-
>>> target/xtensa/cpu.c | 2 +-
>>> 27 files changed, 41 insertions(+), 38 deletions(-)
>>>
>>> diff --git a/accel/tcg/cputlb.c b/accel/tcg/cputlb.c
>>> index 42ab79c1a5..2dc71b5528 100644
>>> --- a/accel/tcg/cputlb.c
>>> +++ b/accel/tcg/cputlb.c
>>> @@ -1286,7 +1286,7 @@ static void tlb_fill(CPUState *cpu, target_ulong addr, int size,
>>> * This is not a probe, so only valid return is success; failure
>>> * should result in exception + longjmp to the cpu loop.
>>> */
>>> - ok = cc->tlb_fill(cpu, addr, size, access_type, mmu_idx, false, retaddr);
>>> + ok = cc->tcg_ops.tlb_fill(cpu, addr, size, access_type, mmu_idx, false, retaddr);
>>> assert(ok);
>>> }
>>>
>>> @@ -1557,8 +1557,8 @@ static int probe_access_internal(CPUArchState *env, target_ulong addr,
>>> CPUState *cs = env_cpu(env);
>>> CPUClass *cc = CPU_GET_CLASS(cs);
>>>
>>> - if (!cc->tlb_fill(cs, addr, fault_size, access_type,
>>> - mmu_idx, nonfault, retaddr)) {
>>> + if (!cc->tcg_ops.tlb_fill(cs, addr, fault_size, access_type,
>>> + mmu_idx, nonfault, retaddr)) {
>>> /* Non-faulting page table read failed. */
>>> *phost = NULL;
>>> return TLB_INVALID_MASK;
>>> diff --git a/accel/tcg/user-exec.c b/accel/tcg/user-exec.c
>>> index 4ebe25461a..7f53992251 100644
>>> --- a/accel/tcg/user-exec.c
>>> +++ b/accel/tcg/user-exec.c
>>> @@ -186,7 +186,7 @@ static inline int handle_cpu_signal(uintptr_t pc, siginfo_t *info,
>>> clear_helper_retaddr();
>>>
>>> cc = CPU_GET_CLASS(cpu);
>>> - cc->tlb_fill(cpu, address, 0, access_type, MMU_USER_IDX, false, pc);
>>> + cc->tcg_ops.tlb_fill(cpu, address, 0, access_type, MMU_USER_IDX, false, pc);
>>> g_assert_not_reached();
>>> }
>>>
>>> @@ -216,8 +216,8 @@ static int probe_access_internal(CPUArchState *env, target_ulong addr,
>>> } else {
>>> CPUState *cpu = env_cpu(env);
>>> CPUClass *cc = CPU_GET_CLASS(cpu);
>>> - cc->tlb_fill(cpu, addr, fault_size, access_type,
>>> - MMU_USER_IDX, false, ra);
>>> + cc->tcg_ops.tlb_fill(cpu, addr, fault_size, access_type,
>>> + MMU_USER_IDX, false, ra);
>>> g_assert_not_reached();
>>> }
>>> }
>>> diff --git a/include/hw/core/cpu.h b/include/hw/core/cpu.h
>>> index 52142e9094..c82ef261c6 100644
>>> --- a/include/hw/core/cpu.h
>>> +++ b/include/hw/core/cpu.h
>>> @@ -110,12 +110,6 @@ struct TranslationBlock;
>>> * If the target behaviour here is anything other than "set
>>> * the PC register to the value passed in" then the target must
>>> * also implement the synchronize_from_tb hook.
>>> - * @tlb_fill: Callback for handling a softmmu tlb miss or user-only
>>> - * address fault. For system mode, if the access is valid, call
>>> - * tlb_set_page and return true; if the access is invalid, and
>>> - * probe is true, return false; otherwise raise an exception and
>>> - * do not return. For user-only mode, always raise an exception
>>> - * and do not return.
>>> * @get_phys_page_debug: Callback for obtaining a physical address.
>>> * @get_phys_page_attrs_debug: Callback for obtaining a physical address and the
>>> * associated memory transaction attributes to use for the access.
>>> @@ -183,9 +177,6 @@ struct CPUClass {
>>> void (*get_memory_mapping)(CPUState *cpu, MemoryMappingList *list,
>>> Error **errp);
>>> void (*set_pc)(CPUState *cpu, vaddr value);
>>> - bool (*tlb_fill)(CPUState *cpu, vaddr address, int size,
>>> - MMUAccessType access_type, int mmu_idx,
>>> - bool probe, uintptr_t retaddr);
>>> hwaddr (*get_phys_page_debug)(CPUState *cpu, vaddr addr);
>>> hwaddr (*get_phys_page_attrs_debug)(CPUState *cpu, vaddr addr,
>>> MemTxAttrs *attrs);
>>> diff --git a/include/hw/core/tcg-cpu-ops.h b/include/hw/core/tcg-cpu-ops.h
>>> index e12f32919b..2ea94acca0 100644
>>> --- a/include/hw/core/tcg-cpu-ops.h
>>> +++ b/include/hw/core/tcg-cpu-ops.h
>>> @@ -37,6 +37,18 @@ typedef struct TcgCpuOperations {
>>> void (*cpu_exec_exit)(CPUState *cpu);
>>> /** @cpu_exec_interrupt: Callback for processing interrupts in cpu_exec */
>>> bool (*cpu_exec_interrupt)(CPUState *cpu, int interrupt_request);
>>> + /**
>>> + * @tlb_fill: Handle a softmmu tlb miss or user-only address fault
>>> + *
>>> + * For system mode, if the access is valid, call tlb_set_page
>>> + * and return true; if the access is invalid, and probe is
>>> + * true, return false; otherwise raise an exception and do
>>> + * not return. For user-only mode, always raise an exception
>>> + * and do not return.
>>> + */
>>> + bool (*tlb_fill)(CPUState *cpu, vaddr address, int size,
>>> + MMUAccessType access_type, int mmu_idx,
>>> + bool probe, uintptr_t retaddr);
>>
>> As per previous patch, here is a chance to clean-up the comment.
>
>
> Could you provide the text? I think you understand this better than I
> do...
As Eduardo pointed out the kernel-doc pass won't work with in-line
functions unless they are extracted and typedefed which seems excessive
considering we don't currently generate docs from these headers. Ignore
the request.
>
>
>>
>> Otherwise:
>>
>> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
>>
>
> Thanks!
>
> Claudio
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 27/32] accel: replace struct CpusAccel with AccelOpsClass
2020-12-09 12:54 ` Alex Bennée
2020-12-09 14:50 ` Claudio Fontana
@ 2020-12-09 17:28 ` Claudio Fontana
2020-12-09 18:30 ` Alex Bennée
1 sibling, 1 reply; 87+ messages in thread
From: Claudio Fontana @ 2020-12-09 17:28 UTC (permalink / raw)
To: Alex Bennée
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
On 12/9/20 1:54 PM, Alex Bennée wrote:
>
> Claudio Fontana <cfontana@suse.de> writes:
>
>> centralize the registration of the cpus.c module
>> accelerator operations in accel/accel-softmmu.c
>>
>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
> <snip>
>> diff --git a/accel/tcg/tcg-cpus.c b/accel/tcg/tcg-cpus.c
>> index e335f9f155..38a58ab271 100644
>> --- a/accel/tcg/tcg-cpus.c
>> +++ b/accel/tcg/tcg-cpus.c
>> @@ -35,6 +35,9 @@
>> #include "hw/boards.h"
>>
>> #include "tcg-cpus.h"
>> +#include "tcg-cpus-mttcg.h"
>> +#include "tcg-cpus-rr.h"
>> +#include "tcg-cpus-icount.h"
>>
>> /* common functionality among all TCG variants */
>>
>> @@ -80,3 +83,43 @@ void tcg_cpus_handle_interrupt(CPUState *cpu, int mask)
>> qatomic_set(&cpu_neg(cpu)->icount_decr.u16.high, -1);
>> }
>> }
>> +
>> +static void tcg_cpus_ops_init(AccelOpsClass *ops)
>> +{
>> + if (qemu_tcg_mttcg_enabled()) {
>> + ops->create_vcpu_thread = mttcg_start_vcpu_thread;
>> + ops->kick_vcpu_thread = mttcg_kick_vcpu_thread;
>> + ops->handle_interrupt = tcg_cpus_handle_interrupt;
>> +
>> + } else if (icount_enabled()) {
>> + ops->create_vcpu_thread = rr_start_vcpu_thread;
>> + ops->kick_vcpu_thread = rr_kick_vcpu_thread;
>> + ops->handle_interrupt = icount_handle_interrupt;
>> + ops->get_virtual_clock = icount_get;
>> + ops->get_elapsed_ticks = icount_get;
>> +
>> + } else {
>> + ops->create_vcpu_thread = rr_start_vcpu_thread;
>> + ops->kick_vcpu_thread = rr_kick_vcpu_thread;
>> + ops->handle_interrupt = tcg_cpus_handle_interrupt;
>> + }
>> +}
>
> Aren't we going backwards here by having a global function aware of the
> different accelerator types rather than encapsulating this is the
> particular accelerator machinery?
>
> <snip>
>
Now I got your point.
The ideal would be to have three classes. One called tcg-mttcg, one tcg-icount, one tcg-rr.
The problem: backward compatibility I think, since currently we have only one accel, "tcg".
But, hmm maybe fixable, I'll take a look.
Thanks!
Claudio
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 27/32] accel: replace struct CpusAccel with AccelOpsClass
2020-12-09 17:28 ` Claudio Fontana
@ 2020-12-09 18:30 ` Alex Bennée
2020-12-09 19:27 ` Claudio Fontana
0 siblings, 1 reply; 87+ messages in thread
From: Alex Bennée @ 2020-12-09 18:30 UTC (permalink / raw)
To: Claudio Fontana
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
Claudio Fontana <cfontana@suse.de> writes:
> On 12/9/20 1:54 PM, Alex Bennée wrote:
>>
>> Claudio Fontana <cfontana@suse.de> writes:
>>
>>> centralize the registration of the cpus.c module
>>> accelerator operations in accel/accel-softmmu.c
>>>
>>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
>> <snip>
>>> diff --git a/accel/tcg/tcg-cpus.c b/accel/tcg/tcg-cpus.c
>>> index e335f9f155..38a58ab271 100644
>>> --- a/accel/tcg/tcg-cpus.c
>>> +++ b/accel/tcg/tcg-cpus.c
>>> @@ -35,6 +35,9 @@
>>> #include "hw/boards.h"
>>>
>>> #include "tcg-cpus.h"
>>> +#include "tcg-cpus-mttcg.h"
>>> +#include "tcg-cpus-rr.h"
>>> +#include "tcg-cpus-icount.h"
>>>
>>> /* common functionality among all TCG variants */
>>>
>>> @@ -80,3 +83,43 @@ void tcg_cpus_handle_interrupt(CPUState *cpu, int mask)
>>> qatomic_set(&cpu_neg(cpu)->icount_decr.u16.high, -1);
>>> }
>>> }
>>> +
>>> +static void tcg_cpus_ops_init(AccelOpsClass *ops)
>>> +{
>>> + if (qemu_tcg_mttcg_enabled()) {
>>> + ops->create_vcpu_thread = mttcg_start_vcpu_thread;
>>> + ops->kick_vcpu_thread = mttcg_kick_vcpu_thread;
>>> + ops->handle_interrupt = tcg_cpus_handle_interrupt;
>>> +
>>> + } else if (icount_enabled()) {
>>> + ops->create_vcpu_thread = rr_start_vcpu_thread;
>>> + ops->kick_vcpu_thread = rr_kick_vcpu_thread;
>>> + ops->handle_interrupt = icount_handle_interrupt;
>>> + ops->get_virtual_clock = icount_get;
>>> + ops->get_elapsed_ticks = icount_get;
>>> +
>>> + } else {
>>> + ops->create_vcpu_thread = rr_start_vcpu_thread;
>>> + ops->kick_vcpu_thread = rr_kick_vcpu_thread;
>>> + ops->handle_interrupt = tcg_cpus_handle_interrupt;
>>> + }
>>> +}
>>
>> Aren't we going backwards here by having a global function aware of the
>> different accelerator types rather than encapsulating this is the
>> particular accelerator machinery?
>>
>> <snip>
>>
>
> Now I got your point.
>
> The ideal would be to have three classes. One called tcg-mttcg, one tcg-icount, one tcg-rr.
> The problem: backward compatibility I think, since currently we have only one accel, "tcg".
>
> But, hmm maybe fixable, I'll take a look.
Yeah I was wondering if we were going to have subclasses for each "type"
of TCG. But now I'm wondering if that even makes sense. Will we ever
want to built a TCG enabled binary that say doesn't do icount? Maybe
not - having a single AccelOpsClass which runs in 3 modes will probably
do for now.
>
> Thanks!
>
> Claudio
--
Alex Bennée
^ permalink raw reply [flat|nested] 87+ messages in thread
* Re: [RFC v9 27/32] accel: replace struct CpusAccel with AccelOpsClass
2020-12-09 18:30 ` Alex Bennée
@ 2020-12-09 19:27 ` Claudio Fontana
0 siblings, 0 replies; 87+ messages in thread
From: Claudio Fontana @ 2020-12-09 19:27 UTC (permalink / raw)
To: Alex Bennée
Cc: Paul Durrant, Jason Wang, qemu-devel, Peter Xu, haxm-team,
Colin Xu, Olaf Hering, Stefano Stabellini, Bruce Rogers,
Emilio G . Cota, Anthony Perard, Philippe Mathieu-Daudé,
Laurent Vivier, Thomas Huth, Eduardo Habkost, Richard Henderson,
Cameron Esfahani, Dario Faggioli, Roman Bolshakov,
Sunil Muthuswamy, Marcelo Tosatti, Wenchao Wang, Paolo Bonzini
On 12/9/20 7:30 PM, Alex Bennée wrote:
>
> Claudio Fontana <cfontana@suse.de> writes:
>
>> On 12/9/20 1:54 PM, Alex Bennée wrote:
>>>
>>> Claudio Fontana <cfontana@suse.de> writes:
>>>
>>>> centralize the registration of the cpus.c module
>>>> accelerator operations in accel/accel-softmmu.c
>>>>
>>>> Signed-off-by: Claudio Fontana <cfontana@suse.de>
>>> <snip>
>>>> diff --git a/accel/tcg/tcg-cpus.c b/accel/tcg/tcg-cpus.c
>>>> index e335f9f155..38a58ab271 100644
>>>> --- a/accel/tcg/tcg-cpus.c
>>>> +++ b/accel/tcg/tcg-cpus.c
>>>> @@ -35,6 +35,9 @@
>>>> #include "hw/boards.h"
>>>>
>>>> #include "tcg-cpus.h"
>>>> +#include "tcg-cpus-mttcg.h"
>>>> +#include "tcg-cpus-rr.h"
>>>> +#include "tcg-cpus-icount.h"
>>>>
>>>> /* common functionality among all TCG variants */
>>>>
>>>> @@ -80,3 +83,43 @@ void tcg_cpus_handle_interrupt(CPUState *cpu, int mask)
>>>> qatomic_set(&cpu_neg(cpu)->icount_decr.u16.high, -1);
>>>> }
>>>> }
>>>> +
>>>> +static void tcg_cpus_ops_init(AccelOpsClass *ops)
>>>> +{
>>>> + if (qemu_tcg_mttcg_enabled()) {
>>>> + ops->create_vcpu_thread = mttcg_start_vcpu_thread;
>>>> + ops->kick_vcpu_thread = mttcg_kick_vcpu_thread;
>>>> + ops->handle_interrupt = tcg_cpus_handle_interrupt;
>>>> +
>>>> + } else if (icount_enabled()) {
>>>> + ops->create_vcpu_thread = rr_start_vcpu_thread;
>>>> + ops->kick_vcpu_thread = rr_kick_vcpu_thread;
>>>> + ops->handle_interrupt = icount_handle_interrupt;
>>>> + ops->get_virtual_clock = icount_get;
>>>> + ops->get_elapsed_ticks = icount_get;
>>>> +
>>>> + } else {
>>>> + ops->create_vcpu_thread = rr_start_vcpu_thread;
>>>> + ops->kick_vcpu_thread = rr_kick_vcpu_thread;
>>>> + ops->handle_interrupt = tcg_cpus_handle_interrupt;
>>>> + }
>>>> +}
>>>
>>> Aren't we going backwards here by having a global function aware of the
>>> different accelerator types rather than encapsulating this is the
>>> particular accelerator machinery?
>>>
>>> <snip>
>>>
>>
>> Now I got your point.
>>
>> The ideal would be to have three classes. One called tcg-mttcg, one tcg-icount, one tcg-rr.
>> The problem: backward compatibility I think, since currently we have only one accel, "tcg".
>>
>> But, hmm maybe fixable, I'll take a look.
>
> Yeah I was wondering if we were going to have subclasses for each "type"
> of TCG. But now I'm wondering if that even makes sense. Will we ever
> want to built a TCG enabled binary that say doesn't do icount? Maybe
> not - having a single AccelOpsClass which runs in 3 modes will probably
> do for now.
Ok, I think you have a point there, so if you find a way we could subclass and not break backward compatibility in a non-super-convoluted way,
I am all ears, thanks a lot!
Ciao,
Claudio
^ permalink raw reply [flat|nested] 87+ messages in thread
end of thread, other threads:[~2020-12-09 20:06 UTC | newest]
Thread overview: 87+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-12-08 19:48 [RFC v9 00/22] i386 cleanup Claudio Fontana
2020-12-08 19:48 ` [RFC v9 01/32] accel/tcg: split CpusAccel into three TCG variants Claudio Fontana
2020-12-09 8:34 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 02/32] accel/tcg: split tcg_start_vcpu_thread Claudio Fontana
2020-12-09 9:03 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 03/32] accel/tcg: rename tcg-cpus functions to match module name Claudio Fontana
2020-12-09 9:10 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 04/32] i386: move kvm accel files into kvm/ Claudio Fontana
2020-12-09 9:17 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 05/32] i386: move whpx accel files into whpx/ Claudio Fontana
2020-12-09 9:21 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 06/32] i386: move hax accel files into hax/ Claudio Fontana
2020-12-09 9:22 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 07/32] i386: hvf: remove stale MAINTAINERS entry for old hvf stubs Claudio Fontana
2020-12-09 9:22 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 08/32] i386: move TCG accel files into tcg/ Claudio Fontana
2020-12-09 9:30 ` Alex Bennée
2020-12-09 11:05 ` Claudio Fontana
2020-12-08 19:48 ` [RFC v9 09/32] i386: move cpu dump out of helper.c into cpu-dump.c Claudio Fontana
2020-12-09 9:59 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 10/32] i386: move TCG cpu class initialization out of helper.c Claudio Fontana
2020-12-09 10:23 ` Alex Bennée
2020-12-09 11:19 ` Claudio Fontana
2020-12-08 19:48 ` [RFC v9 11/32] tcg: cpu_exec_{enter,exit} helpers Claudio Fontana
2020-12-09 10:33 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 12/32] tcg: make CPUClass.cpu_exec_* optional Claudio Fontana
2020-12-09 10:36 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 13/32] tcg: Make CPUClass.debug_excp_handler optional Claudio Fontana
2020-12-09 10:37 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 14/32] cpu: Remove unnecessary noop methods Claudio Fontana
2020-12-09 10:38 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 15/32] cpu: Introduce TCGCpuOperations struct Claudio Fontana
2020-12-09 10:39 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 16/32] target/riscv: remove CONFIG_TCG, as it is always TCG Claudio Fontana
2020-12-08 22:10 ` Alistair Francis
2020-12-08 19:48 ` [RFC v9 17/32] accel/tcg: split TCG-only code from cpu_exec_realizefn Claudio Fontana
2020-12-09 10:42 ` Alex Bennée
2020-12-09 11:22 ` Claudio Fontana
2020-12-08 19:48 ` [RFC v9 18/32] cpu: Move synchronize_from_tb() to tcg_ops Claudio Fontana
2020-12-09 9:27 ` Philippe Mathieu-Daudé
2020-12-09 14:33 ` Claudio Fontana
2020-12-09 15:06 ` Philippe Mathieu-Daudé
2020-12-09 10:50 ` Alex Bennée
2020-12-09 14:46 ` Eduardo Habkost
2020-12-09 15:51 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 19/32] cpu: Move cpu_exec_* " Claudio Fontana
2020-12-09 9:28 ` Philippe Mathieu-Daudé
2020-12-09 11:02 ` Claudio Fontana
2020-12-09 11:16 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 20/32] cpu: Move tlb_fill " Claudio Fontana
2020-12-09 11:26 ` Alex Bennée
2020-12-09 14:38 ` Claudio Fontana
2020-12-09 16:12 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 21/32] cpu: Move debug_excp_handler " Claudio Fontana
2020-12-09 11:29 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 22/32] target/arm: do not use cc->do_interrupt for KVM directly Claudio Fontana
2020-12-09 11:30 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 23/32] cpu: move cc->do_interrupt to tcg_ops Claudio Fontana
2020-12-09 11:43 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 24/32] cpu: move cc->transaction_failed " Claudio Fontana
2020-12-09 9:31 ` Philippe Mathieu-Daudé
2020-12-09 14:43 ` Claudio Fontana
2020-12-09 14:59 ` Eduardo Habkost
2020-12-09 12:03 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 25/32] cpu: move do_unaligned_access " Claudio Fontana
2020-12-09 12:47 ` Alex Bennée
2020-12-08 19:48 ` [RFC v9 26/32] accel: extend AccelState and AccelClass to user-mode Claudio Fontana
2020-12-09 12:51 ` Alex Bennée
2020-12-09 12:58 ` Claudio Fontana
2020-12-08 19:48 ` [RFC v9 27/32] accel: replace struct CpusAccel with AccelOpsClass Claudio Fontana
2020-12-09 12:54 ` Alex Bennée
2020-12-09 14:50 ` Claudio Fontana
2020-12-09 17:28 ` Claudio Fontana
2020-12-09 18:30 ` Alex Bennée
2020-12-09 19:27 ` Claudio Fontana
2020-12-08 19:48 ` [RFC v9 28/32] accel: introduce AccelCPUClass extending CPUClass Claudio Fontana
2020-12-08 19:48 ` [RFC v9 29/32] i386: split cpu accelerators from cpu.c, using AccelCPUClass Claudio Fontana
2020-12-08 19:48 ` [RFC v9 30/32] cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn Claudio Fontana
2020-12-08 19:48 ` [RFC v9 31/32] hw/core/cpu: call qemu_init_vcpu in cpu_common_realizefn Claudio Fontana
2020-12-08 19:48 ` [RFC v9 32/32] cpu: introduce cpu_accel_instance_init Claudio Fontana
2020-12-08 20:00 ` [RFC v9 00/22] i386 cleanup Philippe Mathieu-Daudé
2020-12-08 22:15 ` Claudio Fontana
2020-12-09 8:47 ` Paolo Bonzini
2020-12-08 22:00 ` no-reply
2020-12-09 10:22 ` Alex Bennée
2020-12-09 12:58 ` Alex Bennée
2020-12-09 14:10 ` Claudio Fontana
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.