qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v26 00/20] i386 cleanup PART 2
@ 2021-03-01  8:54 Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 01/20] target/i386: Rename helper_fldt, helper_fstt Claudio Fontana
                   ` (21 more replies)
  0 siblings, 22 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

v25 -> v26:
* i386: separate fpu_helper into user and sysemu parts
 - removed the splitting of the user mode into their own user/fpu_helper.c,
   seems not worth it.

v24 -> v25:

* i386: separate fpu_helper into user and sysemu parts

 - added 2 preliminary patches to the series (from Richard)
 - rebased on those

* i386: move TCG btp_helper into sysemu/

 - fixed title typo (Alex)
 - nested #ifdef more easily (Richard)

v23 -> v24:

* i386: gdbstub: only write CR0/CR2/CR3/EFER for sysemu                                                                                     
 - remove additional #ifdef TARGET_X86_64
 - split in two patches, so it is easier to understand.

v22 -> v23:

* i386: move TCG btp_helper into sysemu/
 - extended the #ifndef CONFIG_USER_ONLY to entire else of
   if (cpl != 0).

* i386: split misc helper into user and sysemu parts
 - added g_assert_not_reached() and changed user file name to -stubs.

* i386: separate fpu_helper into user and sysemu parts
 - removed unused return value
 - added comment abut issues with current cpu_x86_fsave.

* i386: split off sysemu part of cpu.c
 - rename cpu-softmmu.c to cpu-sysemu.c
 - fixed two mispelled comments, and add two comments
   in the headers of cpu.c and cpu-sysemu.c to describe them

* i386: gdbstub: only write CR0/CR2/CR3/EFER for sysemu
 - defined some aux functions to reduce repeated code

* i386: make cpu_load_efer sysemu-only
 - move the function to helper.c, remove "inline"

v21 -> v22: replace "softmmu" with "sysemu"

v20 -> v21:

* meson: add target_user_arch
  - add hexagon

v19 -> v20:

* add new patch to make gdbstub only write certain registers for softmmu.
  In particular, CR0, CR2, CR3 and EFER should not be changed under
  CONFIG_USER_ONLY. (Paolo)

* add new patch to make cpu_load_efer softmmu-only (Paolo)

* i386: split svm_helper into softmmu and stub-only user

  - fixed commit message spelling (Eric)

  - mention in commit message that this reproduces the existing stubs,
    but really everything that requires s->cpl == 0 should be impossible
    to trigger from user-mode, and we could find a way to assert that
    more consistently.

v18 -> v19:

* i386: split smm helper (softmmu)
  - add g_assert_not_reached and cpu_abort for invalid states in
    CONFIG_USER_ONLY (Paolo)

* i386: move TCG btp_helper into softmmu/
  - for CONFIG_USER_ONLY, assert that the hidden IOBPT flags are not set
    while attempting to generate io_bpt helpers.
    Theory to verify (Paolo)

* i386: slit svm_helper into softmmu and stub-only user
  - added XXX in the commit message to highlight the question about
    whether the same check should be done controlling access to
    cpu_load_efer() and state of the hidden SVME flag. (Paolo)

v17 -> v18:

* meson: add target_user_arch

 - add target_user_arch to all targets which build user.
   Otherwise meson complains about missing key for archs without it.
   (Paolo)

* wrap a few gen_helper_ calls around ifndef CONFIG_USER_ONLY.
  This would need a look from someone like Alex or Richard I think,
  as potentially we could remove even more code I think around the
  gen_helper_ calls for CONFIG_USER_ONLY.

  In the current master code, we have empty helpers for user mode,
  but still we generate the preamble code, temporary variables etc,
  just to then call a helper_() function that does nothing.

  In particular I am referring to patches:

  i386: split tcg btp_helper into softmmu and user parts
        DEF_HELPER_FLAGS_3(set_dr, TCG_CALL_NO_WG, void, env, int, tl)
        DEF_HELPER_FLAGS_4(bpt_io, TCG_CALL_NO_WG, void, env, i32, i32, tl)
        gen_bpt_io
        gen_helper_set_dr(cpu_env, s->tmp2_i32, s->T0);

  i386: split smm helper (softmmu)
        DEF_HELPER_1(rsm, void, env)
	gen_helper_rsm(cpu_env);

  (Alex, Richard?)

* removed suffixes from user/ and softmmu/ modules
  (Alex, Philippe).
  Where possible, removed user stubs entirely.
  Renamed the leftover svm_helper stubs to user/svm_stubs.c

* cleaned up lefover unnecessary header files and squashed them.
 

v16 -> v17: changed to RFC

* tcg_ops are already in master, removed from the series

* i386: split cpu accelerators from cpu.c, using AccelCPUClass:
  removed spurious ; and added spacing before/after functions (Richard)

* added new patches as RFC for the next steps, introducing target-specific
  user-mode specific meson variables, and applied to i386/tcg as an
  example, in order to gather feedback.

v15 -> v16:

* cpu: Move synchronize_from_tb() to tcg_ops:
  - adjusted comments (Alex)

* cpu: tcg_ops: move to tcg-cpu-ops.h, keep a pointer in CPUClass:
  - remove forward decl. of AccelCPUClass, should be in a later patch. (Alex)
  - simplified comment about tcg_ops in struct CPUClass (Alex)
  - remove obsolete comment about ARM blocking TCGCPUOps from being const.
    (Alex)

* accel: replace struct CpusAccel with AccelOpsClass:
  - reworded commit message to be clearer about the objective (Alex)

* accel: introduce AccelCPUClass extending CPUClass
  - reworded commit message to be clearer about the objective (Alex)

* hw/core/cpu: call qemu_init_vcpu in cpu_common_realizefn:
  - dropped this patch (Alex, Philippe)

  will try again later, also in the context of:
  https://www.mail-archive.com/qemu-devel@nongnu.org/msg686480.html

* accel: introduce new accessor functions
  - squashed comments in previous patch introducing accel-cpu.h. (Philippe)

* accel-cpu: make cpu_realizefn return a bool
  - split in two patches, separating the change to the phys_bits check
    (Philippe)

v14 -> v15:

* change the TcgCpuOperations so that all fields of the struct are
  defined unconditionally, as per original patch by Eduardo,
  before moving them to a separate struct referenced by a pointer
  (Richard, Eduardo).

* changed (c) year to 2021

* added a patch to make accel_cpu->cpu_realizefn return bool, and
  adapt host_cpu, kvm_cpu, hvf_cpu and tcg_cpu in i386 accordingly.
  Ultimately, consistently moving to a pattern of realize functions
  returning bool will require a rework of all devices.

v13 -> v14: rebased on latest master.
v12 -> v13: rebased on latest master.

v11 -> v12: reordered patches and improved tcg_ops

* reordered all TcgCpuOperations stuff so it is at the beginning

* added patches for ARM-specific tcg ops
  debug_check_watchpoint and adjust_watchpoint_address

* added a patch that puts a forward declared pointer in the struct,
  so as to reduce the change of misuse between common_ss and specific_ss code,
  and tidy up as a consequence all targets, by defining dedicated structs.

v10 -> v11: split off PART 2,

no further changes to PART 2 other than the split.

v9 -> v10: minor tweaks and fixes

* in "i386: split cpu accelerators from cpu.c",

use kvm/kvm-cpu.c, hvf/hvf-cpu.c, tcg/tcg-cpu.c.
Easier to understand compared to editing multiple cpu.c files,
and matches the header files if needed (kvm-cpu.h).

* in "accel: replace struct CpusAccel with AccelOpsClass",

make it a bit more consistent, by naming the files defining
the AccelOpsClass types "...-accel-ops.c" instead of the old
naming "...-cpus.c".

* in "cpu: move cc->transaction_failed to tcg_ops",

protect with CONFIG_TCG the use of tcg_ops for hw/misc/jazz.c,

 #include "exec/memattrs.h" (Philippe, Eduardo)

* in "cpu: Move synchronize_from_tb() to tcg_ops",

 #include "hw/core/cpu.h" (Philippe, Eduardo)

do not remove the comment about struct TcgCpuOperations (Philippe)

* in "accel/tcg: split TCG-only code from cpu_exec_realizefn",

invert tcg_target_initialized set order (Alex)

* in "i386: move TCG cpu class initialization out of helper.c",

extract helper-tcg.h, tcg-cpu.c, and tcg-cpu.h directly into
tcg/, avoiding the extra move later to tcg/ (Alex)



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 (18):
  i386: split cpu accelerators from cpu.c, using AccelCPUClass
  cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn
  accel: introduce new accessor functions
  target/i386: fix host_cpu_adjust_phys_bits error handling
  accel-cpu: make cpu_realizefn return a bool
  meson: add target_user_arch
  i386: split off sysemu-only functionality in tcg-cpu
  i386: split smm helper (sysemu)
  i386: split tcg excp_helper into sysemu and user parts
  i386: move TCG bpt_helper into sysemu/
  i386: split misc helper user stubs and sysemu part
  i386: separate fpu_helper sysemu-only parts
  i386: split svm_helper into sysemu and stub-only user
  i386: split seg_helper into user-only and sysemu parts
  i386: split off sysemu part of cpu.c
  target/i386: gdbstub: introduce aux functions to read/write CS64 regs
  target/i386: gdbstub: only write CR0/CR2/CR3/EFER for sysemu
  i386: make cpu_load_efer sysemu-only

Richard Henderson (2):
  target/i386: Rename helper_fldt, helper_fstt
  target/i386: Split out do_fsave, do_frstor, do_fxsave, do_fxrstor

 meson.build                               |   5 +
 include/hw/core/accel-cpu.h               |   2 +-
 include/qemu/accel.h                      |  13 +
 target/i386/cpu-internal.h                |  70 ++
 target/i386/cpu.h                         |  43 +-
 target/i386/helper.h                      |  11 +
 target/i386/host-cpu.h                    |  19 +
 target/i386/kvm/kvm-cpu.h                 |  41 ++
 target/i386/tcg/helper-tcg.h              |   8 +
 target/i386/tcg/seg_helper.h              |  66 ++
 target/i386/tcg/tcg-cpu.h                 |  21 +-
 accel/accel-common.c                      |  19 +
 cpu.c                                     |   5 +-
 hw/i386/pc_piix.c                         |   1 +
 target/i386/cpu-sysemu.c                  | 352 ++++++++++
 target/i386/cpu.c                         | 782 ++--------------------
 target/i386/gdbstub.c                     | 165 ++---
 target/i386/helper.c                      |  13 +
 target/i386/host-cpu.c                    | 204 ++++++
 target/i386/hvf/hvf-cpu.c                 |  68 ++
 target/i386/kvm/kvm-cpu.c                 | 151 +++++
 target/i386/kvm/kvm.c                     |   3 +-
 target/i386/tcg/bpt_helper.c              | 276 --------
 target/i386/tcg/excp_helper.c             | 572 ----------------
 target/i386/tcg/fpu_helper.c              | 122 ++--
 target/i386/tcg/misc_helper.c             | 463 -------------
 target/i386/tcg/seg_helper.c              | 237 +------
 target/i386/tcg/sysemu/bpt_helper.c       | 293 ++++++++
 target/i386/tcg/sysemu/excp_helper.c      | 581 ++++++++++++++++
 target/i386/tcg/sysemu/fpu_helper.c       |  57 ++
 target/i386/tcg/sysemu/misc_helper.c      | 438 ++++++++++++
 target/i386/tcg/sysemu/seg_helper.c       | 125 ++++
 target/i386/tcg/{ => sysemu}/smm_helper.c |  19 +-
 target/i386/tcg/{ => sysemu}/svm_helper.c |  62 +-
 target/i386/tcg/sysemu/tcg-cpu.c          |  83 +++
 target/i386/tcg/tcg-cpu.c                 |  50 +-
 target/i386/tcg/translate.c               |  13 +-
 target/i386/tcg/user/excp_helper.c        |  39 ++
 target/i386/tcg/user/misc_stubs.c         |  75 +++
 target/i386/tcg/user/seg_helper.c         | 109 +++
 target/i386/tcg/user/svm_stubs.c          |  76 +++
 MAINTAINERS                               |   2 +-
 target/alpha/meson.build                  |   3 +
 target/arm/meson.build                    |   2 +
 target/cris/meson.build                   |   3 +
 target/hexagon/meson.build                |   3 +
 target/hppa/meson.build                   |   3 +
 target/i386/hvf/meson.build               |   1 +
 target/i386/kvm/meson.build               |   7 +-
 target/i386/meson.build                   |   9 +-
 target/i386/tcg/meson.build               |   5 +-
 target/i386/tcg/sysemu/meson.build        |  10 +
 target/i386/tcg/user/meson.build          |   6 +
 target/m68k/meson.build                   |   3 +
 target/microblaze/meson.build             |   3 +
 target/mips/meson.build                   |   3 +
 target/nios2/meson.build                  |   3 +
 target/openrisc/meson.build               |   3 +
 target/ppc/meson.build                    |   3 +
 target/riscv/meson.build                  |   3 +
 target/s390x/meson.build                  |   3 +
 target/sh4/meson.build                    |   3 +
 target/sparc/meson.build                  |   3 +
 target/tilegx/meson.build                 |   3 +
 target/tricore/meson.build                |   3 +
 target/xtensa/meson.build                 |   3 +
 66 files changed, 3266 insertions(+), 2579 deletions(-)
 create mode 100644 target/i386/cpu-internal.h
 create mode 100644 target/i386/host-cpu.h
 create mode 100644 target/i386/kvm/kvm-cpu.h
 create mode 100644 target/i386/tcg/seg_helper.h
 create mode 100644 target/i386/cpu-sysemu.c
 create mode 100644 target/i386/host-cpu.c
 create mode 100644 target/i386/hvf/hvf-cpu.c
 create mode 100644 target/i386/kvm/kvm-cpu.c
 create mode 100644 target/i386/tcg/sysemu/bpt_helper.c
 create mode 100644 target/i386/tcg/sysemu/excp_helper.c
 create mode 100644 target/i386/tcg/sysemu/fpu_helper.c
 create mode 100644 target/i386/tcg/sysemu/misc_helper.c
 create mode 100644 target/i386/tcg/sysemu/seg_helper.c
 rename target/i386/tcg/{ => sysemu}/smm_helper.c (98%)
 rename target/i386/tcg/{ => sysemu}/svm_helper.c (96%)
 create mode 100644 target/i386/tcg/sysemu/tcg-cpu.c
 create mode 100644 target/i386/tcg/user/excp_helper.c
 create mode 100644 target/i386/tcg/user/misc_stubs.c
 create mode 100644 target/i386/tcg/user/seg_helper.c
 create mode 100644 target/i386/tcg/user/svm_stubs.c
 create mode 100644 target/i386/tcg/sysemu/meson.build
 create mode 100644 target/i386/tcg/user/meson.build

-- 
2.26.2



^ permalink raw reply	[flat|nested] 40+ messages in thread

* [PATCH v26 01/20] target/i386: Rename helper_fldt, helper_fstt
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 02/20] target/i386: Split out do_fsave, do_frstor, do_fxsave, do_fxrstor Claudio Fontana
                   ` (20 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Philippe Mathieu-Daudé,
	qemu-devel, Roman Bolshakov, Claudio Fontana

From: Richard Henderson <richard.henderson@linaro.org>

Change the prefix from "helper" to "do".  The former should be
reserved for those functions that are called from TCG; the latter
is in use within the file already for those functions that are
called from the helper functions, adding a "retaddr" argument.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Claudio Fontana <cfontana@suse.de>
Tested-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
---
 target/i386/tcg/fpu_helper.c | 19 +++++++++----------
 1 file changed, 9 insertions(+), 10 deletions(-)

diff --git a/target/i386/tcg/fpu_helper.c b/target/i386/tcg/fpu_helper.c
index 60ed93520a..3d9b192901 100644
--- a/target/i386/tcg/fpu_helper.c
+++ b/target/i386/tcg/fpu_helper.c
@@ -117,8 +117,7 @@ static inline void fpop(CPUX86State *env)
     env->fpstt = (env->fpstt + 1) & 7;
 }
 
-static inline floatx80 helper_fldt(CPUX86State *env, target_ulong ptr,
-                                   uintptr_t retaddr)
+static floatx80 do_fldt(CPUX86State *env, target_ulong ptr, uintptr_t retaddr)
 {
     CPU_LDoubleU temp;
 
@@ -127,8 +126,8 @@ static inline floatx80 helper_fldt(CPUX86State *env, target_ulong ptr,
     return temp.d;
 }
 
-static inline void helper_fstt(CPUX86State *env, floatx80 f, target_ulong ptr,
-                               uintptr_t retaddr)
+static void do_fstt(CPUX86State *env, floatx80 f, target_ulong ptr,
+                    uintptr_t retaddr)
 {
     CPU_LDoubleU temp;
 
@@ -405,14 +404,14 @@ void helper_fldt_ST0(CPUX86State *env, target_ulong ptr)
     int new_fpstt;
 
     new_fpstt = (env->fpstt - 1) & 7;
-    env->fpregs[new_fpstt].d = helper_fldt(env, ptr, GETPC());
+    env->fpregs[new_fpstt].d = do_fldt(env, ptr, GETPC());
     env->fpstt = new_fpstt;
     env->fptags[new_fpstt] = 0; /* validate stack entry */
 }
 
 void helper_fstt_ST0(CPUX86State *env, target_ulong ptr)
 {
-    helper_fstt(env, ST0, ptr, GETPC());
+    do_fstt(env, ST0, ptr, GETPC());
 }
 
 void helper_fpush(CPUX86State *env)
@@ -2468,7 +2467,7 @@ void helper_fsave(CPUX86State *env, target_ulong ptr, int data32)
     ptr += (14 << data32);
     for (i = 0; i < 8; i++) {
         tmp = ST(i);
-        helper_fstt(env, tmp, ptr, GETPC());
+        do_fstt(env, tmp, ptr, GETPC());
         ptr += 10;
     }
 
@@ -2495,7 +2494,7 @@ void helper_frstor(CPUX86State *env, target_ulong ptr, int data32)
     ptr += (14 << data32);
 
     for (i = 0; i < 8; i++) {
-        tmp = helper_fldt(env, ptr, GETPC());
+        tmp = do_fldt(env, ptr, GETPC());
         ST(i) = tmp;
         ptr += 10;
     }
@@ -2539,7 +2538,7 @@ static void do_xsave_fpu(CPUX86State *env, target_ulong ptr, uintptr_t ra)
     addr = ptr + XO(legacy.fpregs);
     for (i = 0; i < 8; i++) {
         floatx80 tmp = ST(i);
-        helper_fstt(env, tmp, addr, ra);
+        do_fstt(env, tmp, addr, ra);
         addr += 16;
     }
 }
@@ -2703,7 +2702,7 @@ static void do_xrstor_fpu(CPUX86State *env, target_ulong ptr, uintptr_t ra)
 
     addr = ptr + XO(legacy.fpregs);
     for (i = 0; i < 8; i++) {
-        floatx80 tmp = helper_fldt(env, addr, ra);
+        floatx80 tmp = do_fldt(env, addr, ra);
         ST(i) = tmp;
         addr += 16;
     }
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 02/20] target/i386: Split out do_fsave, do_frstor, do_fxsave, do_fxrstor
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 01/20] target/i386: Rename helper_fldt, helper_fstt Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 03/20] i386: split cpu accelerators from cpu.c, using AccelCPUClass Claudio Fontana
                   ` (19 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Philippe Mathieu-Daudé,
	qemu-devel, Roman Bolshakov, Claudio Fontana

From: Richard Henderson <richard.henderson@linaro.org>

The helper_* functions must use GETPC() to unwind from TCG.
The cpu_x86_* functions cannot, and directly calling the
helper_* functions is a bug.  Split out new functions that
perform the work and can be used by both.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Claudio Fontana <cfontana@suse.de>
Tested-by: Claudio Fontana <cfontana@suse.de>
---
 target/i386/tcg/fpu_helper.c | 66 +++++++++++++++++++++++-------------
 1 file changed, 42 insertions(+), 24 deletions(-)

diff --git a/target/i386/tcg/fpu_helper.c b/target/i386/tcg/fpu_helper.c
index 3d9b192901..20e4d2e715 100644
--- a/target/i386/tcg/fpu_helper.c
+++ b/target/i386/tcg/fpu_helper.c
@@ -2457,17 +2457,18 @@ void helper_fldenv(CPUX86State *env, target_ulong ptr, int data32)
     do_fldenv(env, ptr, data32, GETPC());
 }
 
-void helper_fsave(CPUX86State *env, target_ulong ptr, int data32)
+static void do_fsave(CPUX86State *env, target_ulong ptr, int data32,
+                     uintptr_t retaddr)
 {
     floatx80 tmp;
     int i;
 
-    do_fstenv(env, ptr, data32, GETPC());
+    do_fstenv(env, ptr, data32, retaddr);
 
     ptr += (14 << data32);
     for (i = 0; i < 8; i++) {
         tmp = ST(i);
-        do_fstt(env, tmp, ptr, GETPC());
+        do_fstt(env, tmp, ptr, retaddr);
         ptr += 10;
     }
 
@@ -2485,30 +2486,41 @@ void helper_fsave(CPUX86State *env, target_ulong ptr, int data32)
     env->fptags[7] = 1;
 }
 
+void helper_fsave(CPUX86State *env, target_ulong ptr, int data32)
+{
+    do_fsave(env, ptr, data32, GETPC());
+}
+
+static void do_frstor(CPUX86State *env, target_ulong ptr, int data32,
+                      uintptr_t retaddr)
+{
+    floatx80 tmp;
+    int i;
+
+    do_fldenv(env, ptr, data32, retaddr);
+    ptr += (14 << data32);
+
+    for (i = 0; i < 8; i++) {
+        tmp = do_fldt(env, ptr, retaddr);
+        ST(i) = tmp;
+        ptr += 10;
+    }
+}
+
 void helper_frstor(CPUX86State *env, target_ulong ptr, int data32)
 {
-    floatx80 tmp;
-    int i;
-
-    do_fldenv(env, ptr, data32, GETPC());
-    ptr += (14 << data32);
-
-    for (i = 0; i < 8; i++) {
-        tmp = do_fldt(env, ptr, GETPC());
-        ST(i) = tmp;
-        ptr += 10;
-    }
+    do_frstor(env, ptr, data32, GETPC());
 }
 
 #if defined(CONFIG_USER_ONLY)
 void cpu_x86_fsave(CPUX86State *env, target_ulong ptr, int data32)
 {
-    helper_fsave(env, ptr, data32);
+    do_fsave(env, ptr, data32, 0);
 }
 
 void cpu_x86_frstor(CPUX86State *env, target_ulong ptr, int data32)
 {
-    helper_frstor(env, ptr, data32);
+    do_frstor(env, ptr, data32, 0);
 }
 #endif
 
@@ -2593,10 +2605,8 @@ static void do_xsave_pkru(CPUX86State *env, target_ulong ptr, uintptr_t ra)
     cpu_stq_data_ra(env, ptr, env->pkru, ra);
 }
 
-void helper_fxsave(CPUX86State *env, target_ulong ptr)
+static void do_fxsave(CPUX86State *env, target_ulong ptr, uintptr_t ra)
 {
-    uintptr_t ra = GETPC();
-
     /* The operand must be 16 byte aligned */
     if (ptr & 0xf) {
         raise_exception_ra(env, EXCP0D_GPF, ra);
@@ -2615,6 +2625,11 @@ void helper_fxsave(CPUX86State *env, target_ulong ptr)
     }
 }
 
+void helper_fxsave(CPUX86State *env, target_ulong ptr)
+{
+    do_fxsave(env, ptr, GETPC());
+}
+
 static uint64_t get_xinuse(CPUX86State *env)
 {
     uint64_t inuse = -1;
@@ -2757,10 +2772,8 @@ static void do_xrstor_pkru(CPUX86State *env, target_ulong ptr, uintptr_t ra)
     env->pkru = cpu_ldq_data_ra(env, ptr, ra);
 }
 
-void helper_fxrstor(CPUX86State *env, target_ulong ptr)
+static void do_fxrstor(CPUX86State *env, target_ulong ptr, uintptr_t ra)
 {
-    uintptr_t ra = GETPC();
-
     /* The operand must be 16 byte aligned */
     if (ptr & 0xf) {
         raise_exception_ra(env, EXCP0D_GPF, ra);
@@ -2779,15 +2792,20 @@ void helper_fxrstor(CPUX86State *env, target_ulong ptr)
     }
 }
 
+void helper_fxrstor(CPUX86State *env, target_ulong ptr)
+{
+    do_fxrstor(env, ptr, GETPC());
+}
+
 #if defined(CONFIG_USER_ONLY)
 void cpu_x86_fxsave(CPUX86State *env, target_ulong ptr)
 {
-    helper_fxsave(env, ptr);
+    do_fxsave(env, ptr, 0);
 }
 
 void cpu_x86_fxrstor(CPUX86State *env, target_ulong ptr)
 {
-    helper_fxrstor(env, ptr);
+    do_fxrstor(env, ptr, 0);
 }
 #endif
 
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 03/20] i386: split cpu accelerators from cpu.c, using AccelCPUClass
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 01/20] target/i386: Rename helper_fldt, helper_fstt Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 02/20] target/i386: Split out do_fsave, do_frstor, do_fxsave, do_fxrstor Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 04/20] cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn Claudio Fontana
                   ` (18 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

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/kvm-cpu.c    KVM x86 AccelCPUClass
hvf/hvf-cpu.c    HVF x86 AccelCPUClass
tcg/tcg-cpu.c    TCG x86 AccelCPUClass

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/cpu.h           |  20 +-
 target/i386/host-cpu.h      |  19 ++
 target/i386/kvm/kvm-cpu.h   |  41 ++++
 target/i386/tcg/tcg-cpu.h   |  15 --
 hw/i386/pc_piix.c           |   1 +
 target/i386/cpu.c           | 390 ++++--------------------------------
 target/i386/host-cpu.c      | 201 +++++++++++++++++++
 target/i386/hvf/hvf-cpu.c   |  68 +++++++
 target/i386/kvm/kvm-cpu.c   | 151 ++++++++++++++
 target/i386/kvm/kvm.c       |   3 +-
 target/i386/tcg/tcg-cpu.c   | 113 ++++++++++-
 MAINTAINERS                 |   2 +-
 target/i386/hvf/meson.build |   1 +
 target/i386/kvm/meson.build |   7 +-
 target/i386/meson.build     |   6 +-
 15 files changed, 651 insertions(+), 387 deletions(-)
 create mode 100644 target/i386/host-cpu.h
 create mode 100644 target/i386/kvm/kvm-cpu.h
 delete mode 100644 target/i386/tcg/tcg-cpu.h
 create mode 100644 target/i386/host-cpu.c
 create mode 100644 target/i386/hvf/hvf-cpu.c
 create mode 100644 target/i386/kvm/kvm-cpu.c

diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index 8be39cfb62..c8a84a9033 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -1925,13 +1925,20 @@ 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);
@@ -2136,17 +2143,6 @@ 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.h b/target/i386/host-cpu.h
new file mode 100644
index 0000000000..b47bc0943f
--- /dev/null
+++ b/target/i386/host-cpu.h
@@ -0,0 +1,19 @@
+/*
+ * x86 host CPU type initialization and host CPU functions
+ *
+ * Copyright 2021 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/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/tcg/tcg-cpu.h b/target/i386/tcg/tcg-cpu.h
deleted file mode 100644
index 81f02e562e..0000000000
--- a/target/i386/tcg/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/hw/i386/pc_piix.c b/hw/i386/pc_piix.c
index 2904b40163..7772880c04 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 6a53446e6a..648e41791f 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -22,38 +22,25 @@
 #include "qemu/cutils.h"
 #include "qemu/bitops.h"
 #include "qemu/qemu-print.h"
-
 #include "cpu.h"
-#include "tcg/tcg-cpu.h"
 #include "tcg/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 "sysemu/whpx.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
@@ -595,8 +582,8 @@ 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++) {
@@ -1589,25 +1576,6 @@ 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
@@ -1632,10 +1600,6 @@ 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;
@@ -4237,32 +4201,6 @@ 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" },
-    { "kvm-msi-ext-dest-id", "off" },
-    { "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
@@ -4304,61 +4242,6 @@ 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),
@@ -4381,62 +4264,25 @@ 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);
-        }
-        object_property_set_bool(OBJECT(cpu), "host-phys-bits", 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 = {
@@ -4446,31 +4292,6 @@ 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);
@@ -5189,7 +5010,7 @@ 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++) {
@@ -5236,8 +5057,6 @@ 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
@@ -5264,20 +5083,6 @@ 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");
-        } else if (kvm_irqchip_is_split() && kvm_enable_x2apic()) {
-            x86_cpu_change_kvm_default("kvm-msi-ext-dest-id", "on");
-        }
-
-        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,
@@ -5287,15 +5092,12 @@ 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);
 
@@ -6326,53 +6128,12 @@ 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) {
@@ -6684,33 +6445,22 @@ static void x86_cpu_hyperv_realize(X86CPU *cpu)
 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) {
+        cc->accel_cpu->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) {
@@ -6762,39 +6512,7 @@ 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);
@@ -6802,8 +6520,8 @@ 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;
@@ -6875,33 +6593,6 @@ 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);
 
     /*
@@ -7101,6 +6792,8 @@ 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;
 
     env->nr_dies = 1;
@@ -7148,6 +6841,11 @@ 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) {
+        cc->accel_cpu->cpu_instance_init(CPU(obj));
+    }
 }
 
 static int64_t x86_cpu_get_arch_id(CPUState *cs)
@@ -7405,11 +7103,6 @@ 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;
@@ -7520,9 +7213,6 @@ 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/host-cpu.c b/target/i386/host-cpu.c
new file mode 100644
index 0000000000..9cfe56ce41
--- /dev/null
+++ b/target/i386/host-cpu.c
@@ -0,0 +1,201 @@
+/*
+ * x86 host CPU functions, and "host" cpu type initialization
+ *
+ * Copyright 2021 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;
+
+    /* Use max host physical address bits if -cpu max option is applied */
+    object_property_set_bool(OBJECT(cpu), "host-phys-bits", true, &error_abort);
+
+    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/hvf/hvf-cpu.c b/target/i386/hvf/hvf-cpu.c
new file mode 100644
index 0000000000..8fbc423888
--- /dev/null
+++ b/target/i386/hvf/hvf-cpu.c
@@ -0,0 +1,68 @@
+/*
+ * x86 HVF CPU type initialization
+ *
+ * Copyright 2021 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/kvm/kvm-cpu.c b/target/i386/kvm/kvm-cpu.c
new file mode 100644
index 0000000000..c23bbe6c50
--- /dev/null
+++ b/target/i386/kvm/kvm-cpu.c
@@ -0,0 +1,151 @@
+/*
+ * x86 KVM CPU type initialization
+ *
+ * Copyright 2021 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" },
+    { "kvm-msi-ext-dest-id", "off" },
+    { "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");
+    } else if (kvm_irqchip_is_split() && kvm_enable_x2apic()) {
+        x86_cpu_change_kvm_default("kvm-msi-ext-dest-id", "on");
+    }
+
+    /* 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.c b/target/i386/kvm/kvm.c
index 0b5755e42b..5064f904c5 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"
@@ -288,7 +289,7 @@ 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/tcg/tcg-cpu.c b/target/i386/tcg/tcg-cpu.c
index 1e125d2175..1d3d6d1c6a 100644
--- a/target/i386/tcg/tcg-cpu.c
+++ b/target/i386/tcg/tcg-cpu.c
@@ -19,13 +19,14 @@
 
 #include "qemu/osdep.h"
 #include "cpu.h"
-#include "tcg-cpu.h"
-#include "exec/exec-all.h"
-#include "sysemu/runstate.h"
 #include "helper-tcg.h"
+#include "qemu/accel.h"
+#include "hw/core/accel-cpu.h"
 
-#if !defined(CONFIG_USER_ONLY)
-#include "hw/i386/apic.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.  */
@@ -72,7 +73,107 @@ static struct TCGCPUOps x86_tcg_ops = {
 #endif /* !CONFIG_USER_ONLY */
 };
 
-void tcg_cpu_common_class_init(CPUClass *cc)
+static void tcg_cpu_class_init(CPUClass *cc)
 {
     cc->tcg_ops = &x86_tcg_ops;
 }
+
+#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 */
+
+/*
+ * 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/MAINTAINERS b/MAINTAINERS
index 9b2aa18e1f..ac7b0d6c6a 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -350,7 +350,7 @@ 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/target/i386/hvf/meson.build b/target/i386/hvf/meson.build
index e9eb5a5da8..d253d5fd10 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',
+  'hvf-cpu.c',
 ))
diff --git a/target/i386/kvm/meson.build b/target/i386/kvm/meson.build
index 1d66559187..0a533411ca 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',
+  'kvm-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 c4bf20b319..fd24479590 100644
--- a/target/i386/meson.build
+++ b/target/i386/meson.build
@@ -6,7 +6,11 @@ i386_ss.add(files(
   'xsave_helper.c',
   'cpu-dump.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(
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 04/20] cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (2 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 03/20] i386: split cpu accelerators from cpu.c, using AccelCPUClass Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 05/20] accel: introduce new accessor functions Claudio Fontana
                   ` (17 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

move the call to accel_cpu->cpu_realizefn 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>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 cpu.c             |  6 ++++++
 target/i386/cpu.c | 20 +++++++-------------
 2 files changed, 13 insertions(+), 13 deletions(-)

diff --git a/cpu.c b/cpu.c
index bfbe5a66f9..ba5d272c1e 100644
--- a/cpu.c
+++ b/cpu.c
@@ -36,6 +36,7 @@
 #include "sysemu/replay.h"
 #include "exec/translate-all.h"
 #include "exec/log.h"
+#include "hw/core/accel-cpu.h"
 
 uintptr_t qemu_host_page_size;
 intptr_t qemu_host_page_mask;
@@ -130,6 +131,11 @@ void cpu_exec_realizefn(CPUState *cpu, Error **errp)
 
     cpu_list_add(cpu);
 
+    if (cc->accel_cpu) {
+        /* NB: errp parameter is unused currently */
+        cc->accel_cpu->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 648e41791f..6e2b5d7e59 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -6445,16 +6445,19 @@ static void x86_cpu_hyperv_realize(X86CPU *cpu)
 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) {
-        cc->accel_cpu->cpu_realizefn(cs, errp);
+    /* Process Hyper-V enlightenments */
+    x86_cpu_hyperv_realize(cpu);
+
+    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()) {
@@ -6570,15 +6573,6 @@ static void x86_cpu_realizefn(DeviceState *dev, Error **errp)
         env->cache_info_amd.l3_cache = &legacy_l3_cache;
     }
 
-    /* Process Hyper-V enlightenments */
-    x86_cpu_hyperv_realize(cpu);
-
-    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] 40+ messages in thread

* [PATCH v26 05/20] accel: introduce new accessor functions
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (3 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 04/20] cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 06/20] target/i386: fix host_cpu_adjust_phys_bits error handling Claudio Fontana
                   ` (16 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

avoid open coding the accesses to cpu->accel_cpu interfaces,
and instead introduce:

accel_cpu_instance_init,
accel_cpu_realizefn

to be used by the targets/ initfn code,
and by cpu_exec_realizefn respectively.

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 include/qemu/accel.h | 13 +++++++++++++
 accel/accel-common.c | 19 +++++++++++++++++++
 cpu.c                |  6 +-----
 target/i386/cpu.c    |  9 ++-------
 4 files changed, 35 insertions(+), 12 deletions(-)

diff --git a/include/qemu/accel.h b/include/qemu/accel.h
index b9d6d69eb8..da0c8ab523 100644
--- a/include/qemu/accel.h
+++ b/include/qemu/accel.h
@@ -78,4 +78,17 @@ int accel_init_machine(AccelState *accel, MachineState *ms);
 void accel_setup_post(MachineState *ms);
 #endif /* !CONFIG_USER_ONLY */
 
+/**
+ * accel_cpu_instance_init:
+ * @cpu: The CPU that needs to do accel-specific object initializations.
+ */
+void accel_cpu_instance_init(CPUState *cpu);
+
+/**
+ * accel_cpu_realizefn:
+ * @cpu: The CPU that needs to call accel-specific cpu realization.
+ * @errp: currently unused.
+ */
+void accel_cpu_realizefn(CPUState *cpu, Error **errp);
+
 #endif /* QEMU_ACCEL_H */
diff --git a/accel/accel-common.c b/accel/accel-common.c
index 9901b0531c..0f6fb4fb66 100644
--- a/accel/accel-common.c
+++ b/accel/accel-common.c
@@ -89,6 +89,25 @@ void accel_init_interfaces(AccelClass *ac)
     accel_init_cpu_interfaces(ac);
 }
 
+void accel_cpu_instance_init(CPUState *cpu)
+{
+    CPUClass *cc = CPU_GET_CLASS(cpu);
+
+    if (cc->accel_cpu && cc->accel_cpu->cpu_instance_init) {
+        cc->accel_cpu->cpu_instance_init(cpu);
+    }
+}
+
+void accel_cpu_realizefn(CPUState *cpu, Error **errp)
+{
+    CPUClass *cc = CPU_GET_CLASS(cpu);
+
+    if (cc->accel_cpu && cc->accel_cpu->cpu_realizefn) {
+        /* NB: errp parameter is unused currently */
+        cc->accel_cpu->cpu_realizefn(cpu, errp);
+    }
+}
+
 static const TypeInfo accel_cpu_type = {
     .name = TYPE_ACCEL_CPU,
     .parent = TYPE_OBJECT,
diff --git a/cpu.c b/cpu.c
index ba5d272c1e..25e6fbfa2c 100644
--- a/cpu.c
+++ b/cpu.c
@@ -130,11 +130,7 @@ void cpu_exec_realizefn(CPUState *cpu, Error **errp)
     CPUClass *cc = CPU_GET_CLASS(cpu);
 
     cpu_list_add(cpu);
-
-    if (cc->accel_cpu) {
-        /* NB: errp parameter is unused currently */
-        cc->accel_cpu->cpu_realizefn(cpu, errp);
-    }
+    accel_cpu_realizefn(cpu, errp);
 
 #ifdef CONFIG_TCG
     /* NB: errp parameter is unused currently */
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 6e2b5d7e59..14e2a60ee5 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 "sysemu/whpx.h"
 #include "kvm/kvm_i386.h"
@@ -6786,8 +6785,6 @@ 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;
 
     env->nr_dies = 1;
@@ -6836,10 +6833,8 @@ 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) {
-        cc->accel_cpu->cpu_instance_init(CPU(obj));
-    }
+    /* if required, do accelerator-specific cpu initializations */
+    accel_cpu_instance_init(CPU(obj));
 }
 
 static int64_t x86_cpu_get_arch_id(CPUState *cs)
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 06/20] target/i386: fix host_cpu_adjust_phys_bits error handling
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (4 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 05/20] accel: introduce new accessor functions Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 07/20] accel-cpu: make cpu_realizefn return a bool Claudio Fontana
                   ` (15 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

move the check for phys_bits outside of host_cpu_adjust_phys_bits,
because otherwise it is impossible to return an error condition
explicitly.

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/host-cpu.c | 22 ++++++++++++----------
 1 file changed, 12 insertions(+), 10 deletions(-)

diff --git a/target/i386/host-cpu.c b/target/i386/host-cpu.c
index 9cfe56ce41..d07d41c34c 100644
--- a/target/i386/host-cpu.c
+++ b/target/i386/host-cpu.c
@@ -50,7 +50,7 @@ static void host_cpu_enable_cpu_pm(X86CPU *cpu)
     env->features[FEAT_1_ECX] |= CPUID_EXT_MONITOR;
 }
 
-static uint32_t host_cpu_adjust_phys_bits(X86CPU *cpu, Error **errp)
+static uint32_t host_cpu_adjust_phys_bits(X86CPU *cpu)
 {
     uint32_t host_phys_bits = host_cpu_phys_bits();
     uint32_t phys_bits = cpu->phys_bits;
@@ -77,14 +77,6 @@ static uint32_t host_cpu_adjust_phys_bits(X86CPU *cpu, Error **errp)
         }
     }
 
-    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;
 }
 
@@ -97,7 +89,17 @@ void host_cpu_realizefn(CPUState *cs, Error **errp)
         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);
+        uint32_t phys_bits = host_cpu_adjust_phys_bits(cpu);
+
+        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;
+        }
+        cpu->phys_bits = phys_bits;
     }
 }
 
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 07/20] accel-cpu: make cpu_realizefn return a bool
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (5 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 06/20] target/i386: fix host_cpu_adjust_phys_bits error handling Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 08/20] meson: add target_user_arch Claudio Fontana
                   ` (14 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

overall, all devices' realize functions take an Error **errp, but return void.

hw/core/qdev.c code, which realizes devices, therefore does:

local_err = NULL;
dc->realize(dev, &local_err);
if (local_err != NULL) {
    goto fail;
}

However, we can improve at least accel_cpu to return a meaningful bool value.

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 include/hw/core/accel-cpu.h | 2 +-
 include/qemu/accel.h        | 2 +-
 target/i386/host-cpu.h      | 2 +-
 accel/accel-common.c        | 6 +++---
 cpu.c                       | 5 +++--
 target/i386/host-cpu.c      | 5 +++--
 target/i386/kvm/kvm-cpu.c   | 4 ++--
 target/i386/tcg/tcg-cpu.c   | 6 ++++--
 8 files changed, 18 insertions(+), 14 deletions(-)

diff --git a/include/hw/core/accel-cpu.h b/include/hw/core/accel-cpu.h
index 24a6697412..5dbfd79955 100644
--- a/include/hw/core/accel-cpu.h
+++ b/include/hw/core/accel-cpu.h
@@ -32,7 +32,7 @@ typedef struct AccelCPUClass {
 
     void (*cpu_class_init)(CPUClass *cc);
     void (*cpu_instance_init)(CPUState *cpu);
-    void (*cpu_realizefn)(CPUState *cpu, Error **errp);
+    bool (*cpu_realizefn)(CPUState *cpu, Error **errp);
 } AccelCPUClass;
 
 #endif /* ACCEL_CPU_H */
diff --git a/include/qemu/accel.h b/include/qemu/accel.h
index da0c8ab523..4f4c283f6f 100644
--- a/include/qemu/accel.h
+++ b/include/qemu/accel.h
@@ -89,6 +89,6 @@ void accel_cpu_instance_init(CPUState *cpu);
  * @cpu: The CPU that needs to call accel-specific cpu realization.
  * @errp: currently unused.
  */
-void accel_cpu_realizefn(CPUState *cpu, Error **errp);
+bool accel_cpu_realizefn(CPUState *cpu, Error **errp);
 
 #endif /* QEMU_ACCEL_H */
diff --git a/target/i386/host-cpu.h b/target/i386/host-cpu.h
index b47bc0943f..6a9bc918ba 100644
--- a/target/i386/host-cpu.h
+++ b/target/i386/host-cpu.h
@@ -12,7 +12,7 @@
 
 void host_cpu_instance_init(X86CPU *cpu);
 void host_cpu_max_instance_init(X86CPU *cpu);
-void host_cpu_realizefn(CPUState *cs, Error **errp);
+bool host_cpu_realizefn(CPUState *cs, Error **errp);
 
 void host_cpu_vendor_fms(char *vendor, int *family, int *model, int *stepping);
 
diff --git a/accel/accel-common.c b/accel/accel-common.c
index 0f6fb4fb66..d77c09d7b5 100644
--- a/accel/accel-common.c
+++ b/accel/accel-common.c
@@ -98,14 +98,14 @@ void accel_cpu_instance_init(CPUState *cpu)
     }
 }
 
-void accel_cpu_realizefn(CPUState *cpu, Error **errp)
+bool accel_cpu_realizefn(CPUState *cpu, Error **errp)
 {
     CPUClass *cc = CPU_GET_CLASS(cpu);
 
     if (cc->accel_cpu && cc->accel_cpu->cpu_realizefn) {
-        /* NB: errp parameter is unused currently */
-        cc->accel_cpu->cpu_realizefn(cpu, errp);
+        return cc->accel_cpu->cpu_realizefn(cpu, errp);
     }
+    return true;
 }
 
 static const TypeInfo accel_cpu_type = {
diff --git a/cpu.c b/cpu.c
index 25e6fbfa2c..34a0484bf4 100644
--- a/cpu.c
+++ b/cpu.c
@@ -130,8 +130,9 @@ void cpu_exec_realizefn(CPUState *cpu, Error **errp)
     CPUClass *cc = CPU_GET_CLASS(cpu);
 
     cpu_list_add(cpu);
-    accel_cpu_realizefn(cpu, errp);
-
+    if (!accel_cpu_realizefn(cpu, errp)) {
+        return;
+    }
 #ifdef CONFIG_TCG
     /* NB: errp parameter is unused currently */
     if (tcg_enabled()) {
diff --git a/target/i386/host-cpu.c b/target/i386/host-cpu.c
index d07d41c34c..4ea9e354ea 100644
--- a/target/i386/host-cpu.c
+++ b/target/i386/host-cpu.c
@@ -80,7 +80,7 @@ static uint32_t host_cpu_adjust_phys_bits(X86CPU *cpu)
     return phys_bits;
 }
 
-void host_cpu_realizefn(CPUState *cs, Error **errp)
+bool host_cpu_realizefn(CPUState *cs, Error **errp)
 {
     X86CPU *cpu = X86_CPU(cs);
     CPUX86State *env = &cpu->env;
@@ -97,10 +97,11 @@ void host_cpu_realizefn(CPUState *cs, Error **errp)
             error_setg(errp, "phys-bits should be between 32 and %u "
                        " (but is %u)",
                        TARGET_PHYS_ADDR_SPACE_BITS, phys_bits);
-            return;
+            return false;
         }
         cpu->phys_bits = phys_bits;
     }
+    return true;
 }
 
 #define CPUID_MODEL_ID_SZ 48
diff --git a/target/i386/kvm/kvm-cpu.c b/target/i386/kvm/kvm-cpu.c
index c23bbe6c50..c660ad4293 100644
--- a/target/i386/kvm/kvm-cpu.c
+++ b/target/i386/kvm/kvm-cpu.c
@@ -18,7 +18,7 @@
 #include "kvm_i386.h"
 #include "hw/core/accel-cpu.h"
 
-static void kvm_cpu_realizefn(CPUState *cs, Error **errp)
+static bool kvm_cpu_realizefn(CPUState *cs, Error **errp)
 {
     X86CPU *cpu = X86_CPU(cs);
     CPUX86State *env = &cpu->env;
@@ -41,7 +41,7 @@ static void kvm_cpu_realizefn(CPUState *cs, Error **errp)
                                                    MSR_IA32_UCODE_REV);
         }
     }
-    host_cpu_realizefn(cs, errp);
+    return host_cpu_realizefn(cs, errp);
 }
 
 /*
diff --git a/target/i386/tcg/tcg-cpu.c b/target/i386/tcg/tcg-cpu.c
index 1d3d6d1c6a..23e1f5f0c3 100644
--- a/target/i386/tcg/tcg-cpu.c
+++ b/target/i386/tcg/tcg-cpu.c
@@ -96,7 +96,7 @@ static void x86_cpu_machine_done(Notifier *n, void *unused)
     }
 }
 
-static void tcg_cpu_realizefn(CPUState *cs, Error **errp)
+static bool tcg_cpu_realizefn(CPUState *cs, Error **errp)
 {
     X86CPU *cpu = X86_CPU(cs);
 
@@ -132,12 +132,14 @@ static void tcg_cpu_realizefn(CPUState *cs, Error **errp)
     /* ... 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);
+    return true;
 }
 
 #else /* CONFIG_USER_ONLY */
 
-static void tcg_cpu_realizefn(CPUState *cs, Error **errp)
+static bool tcg_cpu_realizefn(CPUState *cs, Error **errp)
 {
+    return true;
 }
 
 #endif /* !CONFIG_USER_ONLY */
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 08/20] meson: add target_user_arch
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (6 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 07/20] accel-cpu: make cpu_realizefn return a bool Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 09/20] i386: split off sysemu-only functionality in tcg-cpu Claudio Fontana
                   ` (13 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

the lack of target_user_arch makes it hard to fully leverage the
build system in order to separate user code from sysemu code.

Provide it, so that we can avoid the proliferation of #ifdef
in target code.

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>

[claudio: added changes for new target hexagon]

Signed-off-by: Claudio Fontana <cfontana@suse.de>
---
 meson.build                   | 5 +++++
 target/alpha/meson.build      | 3 +++
 target/arm/meson.build        | 2 ++
 target/cris/meson.build       | 3 +++
 target/hexagon/meson.build    | 3 +++
 target/hppa/meson.build       | 3 +++
 target/i386/meson.build       | 2 ++
 target/m68k/meson.build       | 3 +++
 target/microblaze/meson.build | 3 +++
 target/mips/meson.build       | 3 +++
 target/nios2/meson.build      | 3 +++
 target/openrisc/meson.build   | 3 +++
 target/ppc/meson.build        | 3 +++
 target/riscv/meson.build      | 3 +++
 target/s390x/meson.build      | 3 +++
 target/sh4/meson.build        | 3 +++
 target/sparc/meson.build      | 3 +++
 target/tilegx/meson.build     | 3 +++
 target/tricore/meson.build    | 3 +++
 target/xtensa/meson.build     | 3 +++
 20 files changed, 60 insertions(+)

diff --git a/meson.build b/meson.build
index 05a67c20d9..5be4e5f38c 100644
--- a/meson.build
+++ b/meson.build
@@ -1735,6 +1735,7 @@ modules = {}
 hw_arch = {}
 target_arch = {}
 target_softmmu_arch = {}
+target_user_arch = {}
 
 ###############
 # Trace files #
@@ -2132,6 +2133,10 @@ foreach target : target_dirs
     abi = config_target['TARGET_ABI_DIR']
     target_type='user'
     qemu_target_name = 'qemu-' + target_name
+    t = target_user_arch[arch].apply(config_target, strict: false)
+    arch_srcs += t.sources()
+    arch_deps += t.dependencies()
+
     if 'CONFIG_LINUX_USER' in config_target
       base_dir = 'linux-user'
       target_inc += include_directories('linux-user/host/' / config_host['ARCH'])
diff --git a/target/alpha/meson.build b/target/alpha/meson.build
index 1aec55abb4..1b0555d3ee 100644
--- a/target/alpha/meson.build
+++ b/target/alpha/meson.build
@@ -14,5 +14,8 @@ alpha_ss.add(files(
 alpha_softmmu_ss = ss.source_set()
 alpha_softmmu_ss.add(files('machine.c'))
 
+alpha_user_ss = ss.source_set()
+
 target_arch += {'alpha': alpha_ss}
 target_softmmu_arch += {'alpha': alpha_softmmu_ss}
+target_user_arch += {'alpha': alpha_user_ss}
diff --git a/target/arm/meson.build b/target/arm/meson.build
index 15b936c101..a96af5ee1b 100644
--- a/target/arm/meson.build
+++ b/target/arm/meson.build
@@ -53,6 +53,8 @@ arm_softmmu_ss.add(files(
   'monitor.c',
   'psci.c',
 ))
+arm_user_ss = ss.source_set()
 
 target_arch += {'arm': arm_ss}
 target_softmmu_arch += {'arm': arm_softmmu_ss}
+target_user_arch += {'arm': arm_user_ss}
diff --git a/target/cris/meson.build b/target/cris/meson.build
index 67c3793c85..7fd81e0348 100644
--- a/target/cris/meson.build
+++ b/target/cris/meson.build
@@ -10,5 +10,8 @@ cris_ss.add(files(
 cris_softmmu_ss = ss.source_set()
 cris_softmmu_ss.add(files('mmu.c', 'machine.c'))
 
+cris_user_ss = ss.source_set()
+
 target_arch += {'cris': cris_ss}
 target_softmmu_arch += {'cris': cris_softmmu_ss}
+target_user_arch += {'cris': cris_user_ss}
diff --git a/target/hexagon/meson.build b/target/hexagon/meson.build
index 15318a6fa7..e92d45400d 100644
--- a/target/hexagon/meson.build
+++ b/target/hexagon/meson.build
@@ -188,4 +188,7 @@ hexagon_ss.add(files(
     'conv_emu.c',
 ))
 
+hexagon_user_ss = ss.source_set()
+
 target_arch += {'hexagon': hexagon_ss}
+target_user_arch += {'hexagon': hexagon_user_ss}
diff --git a/target/hppa/meson.build b/target/hppa/meson.build
index 8a7ff82efc..85ad314671 100644
--- a/target/hppa/meson.build
+++ b/target/hppa/meson.build
@@ -15,5 +15,8 @@ hppa_ss.add(files(
 hppa_softmmu_ss = ss.source_set()
 hppa_softmmu_ss.add(files('machine.c'))
 
+hppa_user_ss = ss.source_set()
+
 target_arch += {'hppa': hppa_ss}
 target_softmmu_arch += {'hppa': hppa_softmmu_ss}
+target_user_arch += {'hppa': hppa_user_ss}
diff --git a/target/i386/meson.build b/target/i386/meson.build
index fd24479590..cac26a4581 100644
--- a/target/i386/meson.build
+++ b/target/i386/meson.build
@@ -19,6 +19,7 @@ i386_softmmu_ss.add(files(
   'machine.c',
   'monitor.c',
 ))
+i386_user_ss = ss.source_set()
 
 subdir('kvm')
 subdir('hax')
@@ -28,3 +29,4 @@ subdir('tcg')
 
 target_arch += {'i386': i386_ss}
 target_softmmu_arch += {'i386': i386_softmmu_ss}
+target_user_arch += {'i386': i386_user_ss}
diff --git a/target/m68k/meson.build b/target/m68k/meson.build
index 05cd9fbd1e..b507682684 100644
--- a/target/m68k/meson.build
+++ b/target/m68k/meson.build
@@ -13,5 +13,8 @@ m68k_ss.add(files(
 m68k_softmmu_ss = ss.source_set()
 m68k_softmmu_ss.add(files('monitor.c'))
 
+m68k_user_ss = ss.source_set()
+
 target_arch += {'m68k': m68k_ss}
 target_softmmu_arch += {'m68k': m68k_softmmu_ss}
+target_user_arch += {'m68k': m68k_user_ss}
diff --git a/target/microblaze/meson.build b/target/microblaze/meson.build
index 05ee0ec163..52d8fcb0a3 100644
--- a/target/microblaze/meson.build
+++ b/target/microblaze/meson.build
@@ -16,5 +16,8 @@ microblaze_softmmu_ss.add(files(
   'machine.c',
 ))
 
+microblaze_user_ss = ss.source_set()
+
 target_arch += {'microblaze': microblaze_ss}
 target_softmmu_arch += {'microblaze': microblaze_softmmu_ss}
+target_user_arch += {'microblaze': microblaze_user_ss}
diff --git a/target/mips/meson.build b/target/mips/meson.build
index 9741545440..30843c1521 100644
--- a/target/mips/meson.build
+++ b/target/mips/meson.build
@@ -36,5 +36,8 @@ mips_softmmu_ss.add(when: 'CONFIG_TCG', if_true: files(
   'cp0_helper.c',
 ))
 
+mips_user_ss = ss.source_set()
+
 target_arch += {'mips': mips_ss}
 target_softmmu_arch += {'mips': mips_softmmu_ss}
+target_user_arch += {'mips': mips_user_ss}
diff --git a/target/nios2/meson.build b/target/nios2/meson.build
index e643917db1..00367056fa 100644
--- a/target/nios2/meson.build
+++ b/target/nios2/meson.build
@@ -11,5 +11,8 @@ nios2_ss.add(files(
 nios2_softmmu_ss = ss.source_set()
 nios2_softmmu_ss.add(files('monitor.c'))
 
+nios2_user_ss = ss.source_set()
+
 target_arch += {'nios2': nios2_ss}
 target_softmmu_arch += {'nios2': nios2_softmmu_ss}
+target_user_arch += {'nios2': nios2_user_ss}
diff --git a/target/openrisc/meson.build b/target/openrisc/meson.build
index 9774a58306..794a9e8161 100644
--- a/target/openrisc/meson.build
+++ b/target/openrisc/meson.build
@@ -19,5 +19,8 @@ openrisc_ss.add(files(
 openrisc_softmmu_ss = ss.source_set()
 openrisc_softmmu_ss.add(files('machine.c'))
 
+openrisc_user_ss = ss.source_set()
+
 target_arch += {'openrisc': openrisc_ss}
 target_softmmu_arch += {'openrisc': openrisc_softmmu_ss}
+target_user_arch += {'openrisc': openrisc_user_ss}
diff --git a/target/ppc/meson.build b/target/ppc/meson.build
index bbfef90e08..cdd69bf989 100644
--- a/target/ppc/meson.build
+++ b/target/ppc/meson.build
@@ -33,5 +33,8 @@ ppc_softmmu_ss.add(when: 'TARGET_PPC64', if_true: files(
   'mmu-radix64.c',
 ))
 
+ppc_user_ss = ss.source_set()
+
 target_arch += {'ppc': ppc_ss}
 target_softmmu_arch += {'ppc': ppc_softmmu_ss}
+target_user_arch += {'ppc': ppc_user_ss}
diff --git a/target/riscv/meson.build b/target/riscv/meson.build
index 14a5c62dac..a55851336b 100644
--- a/target/riscv/meson.build
+++ b/target/riscv/meson.build
@@ -31,5 +31,8 @@ riscv_softmmu_ss.add(files(
   'machine.c'
 ))
 
+riscv_user_ss = ss.source_set()
+
 target_arch += {'riscv': riscv_ss}
 target_softmmu_arch += {'riscv': riscv_softmmu_ss}
+target_user_arch += {'riscv': riscv_user_ss}
diff --git a/target/s390x/meson.build b/target/s390x/meson.build
index c42eadb7d2..1219f64112 100644
--- a/target/s390x/meson.build
+++ b/target/s390x/meson.build
@@ -58,5 +58,8 @@ if host_machine.cpu_family() == 's390x' and cc.has_link_argument('-Wl,--s390-pgs
                        if_true: declare_dependency(link_args: ['-Wl,--s390-pgste']))
 endif
 
+s390x_user_ss = ss.source_set()
+
 target_arch += {'s390x': s390x_ss}
 target_softmmu_arch += {'s390x': s390x_softmmu_ss}
+target_user_arch += {'s390x': s390x_user_ss}
diff --git a/target/sh4/meson.build b/target/sh4/meson.build
index 56a57576da..5a05729bc1 100644
--- a/target/sh4/meson.build
+++ b/target/sh4/meson.build
@@ -10,5 +10,8 @@ sh4_ss.add(files(
 sh4_softmmu_ss = ss.source_set()
 sh4_softmmu_ss.add(files('monitor.c'))
 
+sh4_user_ss = ss.source_set()
+
 target_arch += {'sh4': sh4_ss}
 target_softmmu_arch += {'sh4': sh4_softmmu_ss}
+target_user_arch += {'sh4': sh4_user_ss}
diff --git a/target/sparc/meson.build b/target/sparc/meson.build
index a3638b9503..cc77a77064 100644
--- a/target/sparc/meson.build
+++ b/target/sparc/meson.build
@@ -19,5 +19,8 @@ sparc_softmmu_ss.add(files(
   'monitor.c',
 ))
 
+sparc_user_ss = ss.source_set()
+
 target_arch += {'sparc': sparc_ss}
 target_softmmu_arch += {'sparc': sparc_softmmu_ss}
+target_user_arch += {'sparc': sparc_user_ss}
diff --git a/target/tilegx/meson.build b/target/tilegx/meson.build
index 678590439c..23dab8b9a1 100644
--- a/target/tilegx/meson.build
+++ b/target/tilegx/meson.build
@@ -9,5 +9,8 @@ tilegx_ss.add(zlib)
 
 tilegx_softmmu_ss = ss.source_set()
 
+tilegx_user_ss = ss.source_set()
+
 target_arch += {'tilegx': tilegx_ss}
 target_softmmu_arch += {'tilegx': tilegx_softmmu_ss}
+target_user_arch += {'tilegx': tilegx_user_ss}
diff --git a/target/tricore/meson.build b/target/tricore/meson.build
index 0ccc829517..7086ae1a22 100644
--- a/target/tricore/meson.build
+++ b/target/tricore/meson.build
@@ -11,5 +11,8 @@ tricore_ss.add(zlib)
 
 tricore_softmmu_ss = ss.source_set()
 
+tricore_user_ss = ss.source_set()
+
 target_arch += {'tricore': tricore_ss}
 target_softmmu_arch += {'tricore': tricore_softmmu_ss}
+target_user_arch += {'tricore': tricore_user_ss}
diff --git a/target/xtensa/meson.build b/target/xtensa/meson.build
index dd750a977e..949b2c8334 100644
--- a/target/xtensa/meson.build
+++ b/target/xtensa/meson.build
@@ -28,5 +28,8 @@ xtensa_softmmu_ss.add(files(
   'xtensa-semi.c',
 ))
 
+xtensa_user_ss = ss.source_set()
+
 target_arch += {'xtensa': xtensa_ss}
 target_softmmu_arch += {'xtensa': xtensa_softmmu_ss}
+target_user_arch += {'xtensa': xtensa_user_ss}
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 09/20] i386: split off sysemu-only functionality in tcg-cpu
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (7 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 08/20] meson: add target_user_arch Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 10/20] i386: split smm helper (sysemu) Claudio Fontana
                   ` (12 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/tcg/tcg-cpu.h          | 24 +++++++++
 target/i386/tcg/sysemu/tcg-cpu.c   | 83 ++++++++++++++++++++++++++++++
 target/i386/tcg/tcg-cpu.c          | 75 ++-------------------------
 target/i386/tcg/meson.build        |  3 ++
 target/i386/tcg/sysemu/meson.build |  3 ++
 target/i386/tcg/user/meson.build   |  2 +
 6 files changed, 119 insertions(+), 71 deletions(-)
 create mode 100644 target/i386/tcg/tcg-cpu.h
 create mode 100644 target/i386/tcg/sysemu/tcg-cpu.c
 create mode 100644 target/i386/tcg/sysemu/meson.build
 create mode 100644 target/i386/tcg/user/meson.build

diff --git a/target/i386/tcg/tcg-cpu.h b/target/i386/tcg/tcg-cpu.h
new file mode 100644
index 0000000000..36bd300af0
--- /dev/null
+++ b/target/i386/tcg/tcg-cpu.h
@@ -0,0 +1,24 @@
+/*
+ * i386 TCG cpu class initialization 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 TCG_CPU_H
+#define TCG_CPU_H
+
+bool tcg_cpu_realizefn(CPUState *cs, Error **errp);
+
+#endif /* TCG_CPU_H */
diff --git a/target/i386/tcg/sysemu/tcg-cpu.c b/target/i386/tcg/sysemu/tcg-cpu.c
new file mode 100644
index 0000000000..c223c0fe9b
--- /dev/null
+++ b/target/i386/tcg/sysemu/tcg-cpu.c
@@ -0,0 +1,83 @@
+/*
+ * i386 TCG cpu class initialization functions specific to sysemu
+ *
+ *  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/helper-tcg.h"
+
+#include "sysemu/sysemu.h"
+#include "qemu/units.h"
+#include "exec/address-spaces.h"
+
+#include "tcg/tcg-cpu.h"
+
+static void tcg_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);
+    }
+}
+
+bool 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 = tcg_cpu_machine_done;
+    qemu_add_machine_init_done_notifier(&cpu->machine_done);
+    return true;
+}
diff --git a/target/i386/tcg/tcg-cpu.c b/target/i386/tcg/tcg-cpu.c
index 23e1f5f0c3..e311f52855 100644
--- a/target/i386/tcg/tcg-cpu.c
+++ b/target/i386/tcg/tcg-cpu.c
@@ -23,11 +23,7 @@
 #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
+#include "tcg-cpu.h"
 
 /* Frob eflags into and out of the CPU temporary format.  */
 
@@ -78,72 +74,6 @@ static void tcg_cpu_class_init(CPUClass *cc)
     cc->tcg_ops = &x86_tcg_ops;
 }
 
-#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 bool 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);
-    return true;
-}
-
-#else /* CONFIG_USER_ONLY */
-
-static bool tcg_cpu_realizefn(CPUState *cs, Error **errp)
-{
-    return true;
-}
-
-#endif /* !CONFIG_USER_ONLY */
-
 /*
  * TCG-specific defaults that override all CPU models when using TCG
  */
@@ -163,7 +93,10 @@ static void tcg_cpu_accel_class_init(ObjectClass *oc, void *data)
 {
     AccelCPUClass *acc = ACCEL_CPU_CLASS(oc);
 
+#ifndef CONFIG_USER_ONLY
     acc->cpu_realizefn = tcg_cpu_realizefn;
+#endif /* CONFIG_USER_ONLY */
+
     acc->cpu_class_init = tcg_cpu_class_init;
     acc->cpu_instance_init = tcg_cpu_instance_init;
 }
diff --git a/target/i386/tcg/meson.build b/target/i386/tcg/meson.build
index 6a1a73cdbf..320bcd1e46 100644
--- a/target/i386/tcg/meson.build
+++ b/target/i386/tcg/meson.build
@@ -12,3 +12,6 @@ i386_ss.add(when: 'CONFIG_TCG', if_true: files(
   'svm_helper.c',
   'tcg-cpu.c',
   'translate.c'), if_false: files('tcg-stub.c'))
+
+subdir('sysemu')
+subdir('user')
diff --git a/target/i386/tcg/sysemu/meson.build b/target/i386/tcg/sysemu/meson.build
new file mode 100644
index 0000000000..4ab30cc32e
--- /dev/null
+++ b/target/i386/tcg/sysemu/meson.build
@@ -0,0 +1,3 @@
+i386_softmmu_ss.add(when: ['CONFIG_TCG', 'CONFIG_SOFTMMU'], if_true: files(
+  'tcg-cpu.c',
+))
diff --git a/target/i386/tcg/user/meson.build b/target/i386/tcg/user/meson.build
new file mode 100644
index 0000000000..7aecc53155
--- /dev/null
+++ b/target/i386/tcg/user/meson.build
@@ -0,0 +1,2 @@
+i386_user_ss.add(when: ['CONFIG_TCG', 'CONFIG_USER_ONLY'], if_true: files(
+))
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 10/20] i386: split smm helper (sysemu)
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (8 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 09/20] i386: split off sysemu-only functionality in tcg-cpu Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 11/20] i386: split tcg excp_helper into sysemu and user parts Claudio Fontana
                   ` (11 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

smm is only really useful for sysemu, split in two modules
around the CONFIG_USER_ONLY, in order to remove the ifdef
and use the build system instead.

add cpu_abort() when detecting attempts to enter SMM mode via
SMI interrupt in user-mode, and assert that the cpu is not
in SMM mode while translating RSM instructions.

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/helper.h                      |  4 ++++
 target/i386/tcg/seg_helper.c              |  4 ++++
 target/i386/tcg/{ => sysemu}/smm_helper.c | 19 ++-----------------
 target/i386/tcg/translate.c               |  5 +++++
 target/i386/tcg/meson.build               |  1 -
 target/i386/tcg/sysemu/meson.build        |  1 +
 6 files changed, 16 insertions(+), 18 deletions(-)
 rename target/i386/tcg/{ => sysemu}/smm_helper.c (98%)

diff --git a/target/i386/helper.h b/target/i386/helper.h
index c2ae2f7e61..8ffda4cdc6 100644
--- a/target/i386/helper.h
+++ b/target/i386/helper.h
@@ -70,7 +70,11 @@ DEF_HELPER_1(clac, void, env)
 DEF_HELPER_1(stac, void, env)
 DEF_HELPER_3(boundw, void, env, tl, int)
 DEF_HELPER_3(boundl, void, env, tl, int)
+
+#ifndef CONFIG_USER_ONLY
 DEF_HELPER_1(rsm, void, env)
+#endif /* !CONFIG_USER_ONLY */
+
 DEF_HELPER_2(into, void, env, int)
 DEF_HELPER_2(cmpxchg8b_unlocked, void, env, tl)
 DEF_HELPER_2(cmpxchg8b, void, env, tl)
diff --git a/target/i386/tcg/seg_helper.c b/target/i386/tcg/seg_helper.c
index 180d47f0e9..d04fbdd7cd 100644
--- a/target/i386/tcg/seg_helper.c
+++ b/target/i386/tcg/seg_helper.c
@@ -1351,7 +1351,11 @@ bool x86_cpu_exec_interrupt(CPUState *cs, int interrupt_request)
     case CPU_INTERRUPT_SMI:
         cpu_svm_check_intercept_param(env, SVM_EXIT_SMI, 0, 0);
         cs->interrupt_request &= ~CPU_INTERRUPT_SMI;
+#ifdef CONFIG_USER_ONLY
+        cpu_abort(CPU(cpu), "SMI interrupt: cannot enter SMM in user-mode");
+#else
         do_smm_enter(cpu);
+#endif /* CONFIG_USER_ONLY */
         break;
     case CPU_INTERRUPT_NMI:
         cpu_svm_check_intercept_param(env, SVM_EXIT_NMI, 0, 0);
diff --git a/target/i386/tcg/smm_helper.c b/target/i386/tcg/sysemu/smm_helper.c
similarity index 98%
rename from target/i386/tcg/smm_helper.c
rename to target/i386/tcg/sysemu/smm_helper.c
index 62d027abd3..a45b5651c3 100644
--- a/target/i386/tcg/smm_helper.c
+++ b/target/i386/tcg/sysemu/smm_helper.c
@@ -1,5 +1,5 @@
 /*
- *  x86 SMM helpers
+ *  x86 SMM helpers (sysemu-only)
  *
  *  Copyright (c) 2003 Fabrice Bellard
  *
@@ -18,27 +18,14 @@
  */
 
 #include "qemu/osdep.h"
-#include "qemu/main-loop.h"
 #include "cpu.h"
 #include "exec/helper-proto.h"
 #include "exec/log.h"
-#include "helper-tcg.h"
+#include "tcg/helper-tcg.h"
 
 
 /* SMM support */
 
-#if defined(CONFIG_USER_ONLY)
-
-void do_smm_enter(X86CPU *cpu)
-{
-}
-
-void helper_rsm(CPUX86State *env)
-{
-}
-
-#else
-
 #ifdef TARGET_X86_64
 #define SMM_REVISION_ID 0x00020064
 #else
@@ -330,5 +317,3 @@ void helper_rsm(CPUX86State *env)
     qemu_log_mask(CPU_LOG_INT, "SMM: after RSM\n");
     log_cpu_state_mask(CPU_LOG_INT, CPU(cpu), CPU_DUMP_CCOP);
 }
-
-#endif /* !CONFIG_USER_ONLY */
diff --git a/target/i386/tcg/translate.c b/target/i386/tcg/translate.c
index af1faf9342..b882041ef0 100644
--- a/target/i386/tcg/translate.c
+++ b/target/i386/tcg/translate.c
@@ -8319,9 +8319,14 @@ static target_ulong disas_insn(DisasContext *s, CPUState *cpu)
         gen_svm_check_intercept(s, pc_start, SVM_EXIT_RSM);
         if (!(s->flags & HF_SMM_MASK))
             goto illegal_op;
+#ifdef CONFIG_USER_ONLY
+        /* we should not be in SMM mode */
+        g_assert_not_reached();
+#else
         gen_update_cc_op(s);
         gen_jmp_im(s, s->pc - s->cs_base);
         gen_helper_rsm(cpu_env);
+#endif /* CONFIG_USER_ONLY */
         gen_eob(s);
         break;
     case 0x1b8: /* SSE4.2 popcnt */
diff --git a/target/i386/tcg/meson.build b/target/i386/tcg/meson.build
index 320bcd1e46..449d9719ef 100644
--- a/target/i386/tcg/meson.build
+++ b/target/i386/tcg/meson.build
@@ -8,7 +8,6 @@ i386_ss.add(when: 'CONFIG_TCG', if_true: files(
   'misc_helper.c',
   'mpx_helper.c',
   'seg_helper.c',
-  'smm_helper.c',
   'svm_helper.c',
   'tcg-cpu.c',
   'translate.c'), if_false: files('tcg-stub.c'))
diff --git a/target/i386/tcg/sysemu/meson.build b/target/i386/tcg/sysemu/meson.build
index 4ab30cc32e..35ba16dc3d 100644
--- a/target/i386/tcg/sysemu/meson.build
+++ b/target/i386/tcg/sysemu/meson.build
@@ -1,3 +1,4 @@
 i386_softmmu_ss.add(when: ['CONFIG_TCG', 'CONFIG_SOFTMMU'], if_true: files(
   'tcg-cpu.c',
+  'smm_helper.c',
 ))
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 11/20] i386: split tcg excp_helper into sysemu and user parts
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (9 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 10/20] i386: split smm helper (sysemu) Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 12/20] i386: move TCG bpt_helper into sysemu/ Claudio Fontana
                   ` (10 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/tcg/excp_helper.c        | 572 --------------------------
 target/i386/tcg/sysemu/excp_helper.c | 581 +++++++++++++++++++++++++++
 target/i386/tcg/user/excp_helper.c   |  39 ++
 target/i386/tcg/sysemu/meson.build   |   1 +
 target/i386/tcg/user/meson.build     |   1 +
 5 files changed, 622 insertions(+), 572 deletions(-)
 create mode 100644 target/i386/tcg/sysemu/excp_helper.c
 create mode 100644 target/i386/tcg/user/excp_helper.c

diff --git a/target/i386/tcg/excp_helper.c b/target/i386/tcg/excp_helper.c
index b7d6259e4a..0183f3932e 100644
--- a/target/i386/tcg/excp_helper.c
+++ b/target/i386/tcg/excp_helper.c
@@ -137,575 +137,3 @@ void raise_exception_ra(CPUX86State *env, int exception_index, uintptr_t retaddr
 {
     raise_interrupt2(env, exception_index, 0, 0, 0, retaddr);
 }
-
-#if !defined(CONFIG_USER_ONLY)
-static hwaddr get_hphys(CPUState *cs, hwaddr gphys, MMUAccessType access_type,
-                        int *prot)
-{
-    CPUX86State *env = &X86_CPU(cs)->env;
-    uint64_t rsvd_mask = PG_HI_RSVD_MASK;
-    uint64_t ptep, pte;
-    uint64_t exit_info_1 = 0;
-    target_ulong pde_addr, pte_addr;
-    uint32_t page_offset;
-    int page_size;
-
-    if (likely(!(env->hflags2 & HF2_NPT_MASK))) {
-        return gphys;
-    }
-
-    if (!(env->nested_pg_mode & SVM_NPT_NXE)) {
-        rsvd_mask |= PG_NX_MASK;
-    }
-
-    if (env->nested_pg_mode & SVM_NPT_PAE) {
-        uint64_t pde, pdpe;
-        target_ulong pdpe_addr;
-
-#ifdef TARGET_X86_64
-        if (env->nested_pg_mode & SVM_NPT_LMA) {
-            uint64_t pml5e;
-            uint64_t pml4e_addr, pml4e;
-
-            pml5e = env->nested_cr3;
-            ptep = PG_NX_MASK | PG_USER_MASK | PG_RW_MASK;
-
-            pml4e_addr = (pml5e & PG_ADDRESS_MASK) +
-                    (((gphys >> 39) & 0x1ff) << 3);
-            pml4e = x86_ldq_phys(cs, pml4e_addr);
-            if (!(pml4e & PG_PRESENT_MASK)) {
-                goto do_fault;
-            }
-            if (pml4e & (rsvd_mask | PG_PSE_MASK)) {
-                goto do_fault_rsvd;
-            }
-            if (!(pml4e & PG_ACCESSED_MASK)) {
-                pml4e |= PG_ACCESSED_MASK;
-                x86_stl_phys_notdirty(cs, pml4e_addr, pml4e);
-            }
-            ptep &= pml4e ^ PG_NX_MASK;
-            pdpe_addr = (pml4e & PG_ADDRESS_MASK) +
-                    (((gphys >> 30) & 0x1ff) << 3);
-            pdpe = x86_ldq_phys(cs, pdpe_addr);
-            if (!(pdpe & PG_PRESENT_MASK)) {
-                goto do_fault;
-            }
-            if (pdpe & rsvd_mask) {
-                goto do_fault_rsvd;
-            }
-            ptep &= pdpe ^ PG_NX_MASK;
-            if (!(pdpe & PG_ACCESSED_MASK)) {
-                pdpe |= PG_ACCESSED_MASK;
-                x86_stl_phys_notdirty(cs, pdpe_addr, pdpe);
-            }
-            if (pdpe & PG_PSE_MASK) {
-                /* 1 GB page */
-                page_size = 1024 * 1024 * 1024;
-                pte_addr = pdpe_addr;
-                pte = pdpe;
-                goto do_check_protect;
-            }
-        } else
-#endif
-        {
-            pdpe_addr = (env->nested_cr3 & ~0x1f) + ((gphys >> 27) & 0x18);
-            pdpe = x86_ldq_phys(cs, pdpe_addr);
-            if (!(pdpe & PG_PRESENT_MASK)) {
-                goto do_fault;
-            }
-            rsvd_mask |= PG_HI_USER_MASK;
-            if (pdpe & (rsvd_mask | PG_NX_MASK)) {
-                goto do_fault_rsvd;
-            }
-            ptep = PG_NX_MASK | PG_USER_MASK | PG_RW_MASK;
-        }
-
-        pde_addr = (pdpe & PG_ADDRESS_MASK) + (((gphys >> 21) & 0x1ff) << 3);
-        pde = x86_ldq_phys(cs, pde_addr);
-        if (!(pde & PG_PRESENT_MASK)) {
-            goto do_fault;
-        }
-        if (pde & rsvd_mask) {
-            goto do_fault_rsvd;
-        }
-        ptep &= pde ^ PG_NX_MASK;
-        if (pde & PG_PSE_MASK) {
-            /* 2 MB page */
-            page_size = 2048 * 1024;
-            pte_addr = pde_addr;
-            pte = pde;
-            goto do_check_protect;
-        }
-        /* 4 KB page */
-        if (!(pde & PG_ACCESSED_MASK)) {
-            pde |= PG_ACCESSED_MASK;
-            x86_stl_phys_notdirty(cs, pde_addr, pde);
-        }
-        pte_addr = (pde & PG_ADDRESS_MASK) + (((gphys >> 12) & 0x1ff) << 3);
-        pte = x86_ldq_phys(cs, pte_addr);
-        if (!(pte & PG_PRESENT_MASK)) {
-            goto do_fault;
-        }
-        if (pte & rsvd_mask) {
-            goto do_fault_rsvd;
-        }
-        /* combine pde and pte nx, user and rw protections */
-        ptep &= pte ^ PG_NX_MASK;
-        page_size = 4096;
-    } else {
-        uint32_t pde;
-
-        /* page directory entry */
-        pde_addr = (env->nested_cr3 & ~0xfff) + ((gphys >> 20) & 0xffc);
-        pde = x86_ldl_phys(cs, pde_addr);
-        if (!(pde & PG_PRESENT_MASK)) {
-            goto do_fault;
-        }
-        ptep = pde | PG_NX_MASK;
-
-        /* if host cr4 PSE bit is set, then we use a 4MB page */
-        if ((pde & PG_PSE_MASK) && (env->nested_pg_mode & SVM_NPT_PSE)) {
-            page_size = 4096 * 1024;
-            pte_addr = pde_addr;
-
-            /* Bits 20-13 provide bits 39-32 of the address, bit 21 is reserved.
-             * Leave bits 20-13 in place for setting accessed/dirty bits below.
-             */
-            pte = pde | ((pde & 0x1fe000LL) << (32 - 13));
-            rsvd_mask = 0x200000;
-            goto do_check_protect_pse36;
-        }
-
-        if (!(pde & PG_ACCESSED_MASK)) {
-            pde |= PG_ACCESSED_MASK;
-            x86_stl_phys_notdirty(cs, pde_addr, pde);
-        }
-
-        /* page directory entry */
-        pte_addr = (pde & ~0xfff) + ((gphys >> 10) & 0xffc);
-        pte = x86_ldl_phys(cs, pte_addr);
-        if (!(pte & PG_PRESENT_MASK)) {
-            goto do_fault;
-        }
-        /* combine pde and pte user and rw protections */
-        ptep &= pte | PG_NX_MASK;
-        page_size = 4096;
-        rsvd_mask = 0;
-    }
-
- do_check_protect:
-    rsvd_mask |= (page_size - 1) & PG_ADDRESS_MASK & ~PG_PSE_PAT_MASK;
- do_check_protect_pse36:
-    if (pte & rsvd_mask) {
-        goto do_fault_rsvd;
-    }
-    ptep ^= PG_NX_MASK;
-
-    if (!(ptep & PG_USER_MASK)) {
-        goto do_fault_protect;
-    }
-    if (ptep & PG_NX_MASK) {
-        if (access_type == MMU_INST_FETCH) {
-            goto do_fault_protect;
-        }
-        *prot &= ~PAGE_EXEC;
-    }
-    if (!(ptep & PG_RW_MASK)) {
-        if (access_type == MMU_DATA_STORE) {
-            goto do_fault_protect;
-        }
-        *prot &= ~PAGE_WRITE;
-    }
-
-    pte &= PG_ADDRESS_MASK & ~(page_size - 1);
-    page_offset = gphys & (page_size - 1);
-    return pte + page_offset;
-
- do_fault_rsvd:
-    exit_info_1 |= SVM_NPTEXIT_RSVD;
- do_fault_protect:
-    exit_info_1 |= SVM_NPTEXIT_P;
- do_fault:
-    x86_stq_phys(cs, env->vm_vmcb + offsetof(struct vmcb, control.exit_info_2),
-                 gphys);
-    exit_info_1 |= SVM_NPTEXIT_US;
-    if (access_type == MMU_DATA_STORE) {
-        exit_info_1 |= SVM_NPTEXIT_RW;
-    } else if (access_type == MMU_INST_FETCH) {
-        exit_info_1 |= SVM_NPTEXIT_ID;
-    }
-    if (prot) {
-        exit_info_1 |= SVM_NPTEXIT_GPA;
-    } else { /* page table access */
-        exit_info_1 |= SVM_NPTEXIT_GPT;
-    }
-    cpu_vmexit(env, SVM_EXIT_NPF, exit_info_1, env->retaddr);
-}
-
-/* return value:
- * -1 = cannot handle fault
- * 0  = nothing more to do
- * 1  = generate PF fault
- */
-static int handle_mmu_fault(CPUState *cs, vaddr addr, int size,
-                            int is_write1, int mmu_idx)
-{
-    X86CPU *cpu = X86_CPU(cs);
-    CPUX86State *env = &cpu->env;
-    uint64_t ptep, pte;
-    int32_t a20_mask;
-    target_ulong pde_addr, pte_addr;
-    int error_code = 0;
-    int is_dirty, prot, page_size, is_write, is_user;
-    hwaddr paddr;
-    uint64_t rsvd_mask = PG_HI_RSVD_MASK;
-    uint32_t page_offset;
-    target_ulong vaddr;
-    uint32_t pkr;
-
-    is_user = mmu_idx == MMU_USER_IDX;
-#if defined(DEBUG_MMU)
-    printf("MMU fault: addr=%" VADDR_PRIx " w=%d u=%d eip=" TARGET_FMT_lx "\n",
-           addr, is_write1, is_user, env->eip);
-#endif
-    is_write = is_write1 & 1;
-
-    a20_mask = x86_get_a20_mask(env);
-    if (!(env->cr[0] & CR0_PG_MASK)) {
-        pte = addr;
-#ifdef TARGET_X86_64
-        if (!(env->hflags & HF_LMA_MASK)) {
-            /* Without long mode we can only address 32bits in real mode */
-            pte = (uint32_t)pte;
-        }
-#endif
-        prot = PAGE_READ | PAGE_WRITE | PAGE_EXEC;
-        page_size = 4096;
-        goto do_mapping;
-    }
-
-    if (!(env->efer & MSR_EFER_NXE)) {
-        rsvd_mask |= PG_NX_MASK;
-    }
-
-    if (env->cr[4] & CR4_PAE_MASK) {
-        uint64_t pde, pdpe;
-        target_ulong pdpe_addr;
-
-#ifdef TARGET_X86_64
-        if (env->hflags & HF_LMA_MASK) {
-            bool la57 = env->cr[4] & CR4_LA57_MASK;
-            uint64_t pml5e_addr, pml5e;
-            uint64_t pml4e_addr, pml4e;
-            int32_t sext;
-
-            /* test virtual address sign extension */
-            sext = la57 ? (int64_t)addr >> 56 : (int64_t)addr >> 47;
-            if (sext != 0 && sext != -1) {
-                env->error_code = 0;
-                cs->exception_index = EXCP0D_GPF;
-                return 1;
-            }
-
-            if (la57) {
-                pml5e_addr = ((env->cr[3] & ~0xfff) +
-                        (((addr >> 48) & 0x1ff) << 3)) & a20_mask;
-                pml5e_addr = get_hphys(cs, pml5e_addr, MMU_DATA_STORE, NULL);
-                pml5e = x86_ldq_phys(cs, pml5e_addr);
-                if (!(pml5e & PG_PRESENT_MASK)) {
-                    goto do_fault;
-                }
-                if (pml5e & (rsvd_mask | PG_PSE_MASK)) {
-                    goto do_fault_rsvd;
-                }
-                if (!(pml5e & PG_ACCESSED_MASK)) {
-                    pml5e |= PG_ACCESSED_MASK;
-                    x86_stl_phys_notdirty(cs, pml5e_addr, pml5e);
-                }
-                ptep = pml5e ^ PG_NX_MASK;
-            } else {
-                pml5e = env->cr[3];
-                ptep = PG_NX_MASK | PG_USER_MASK | PG_RW_MASK;
-            }
-
-            pml4e_addr = ((pml5e & PG_ADDRESS_MASK) +
-                    (((addr >> 39) & 0x1ff) << 3)) & a20_mask;
-            pml4e_addr = get_hphys(cs, pml4e_addr, MMU_DATA_STORE, false);
-            pml4e = x86_ldq_phys(cs, pml4e_addr);
-            if (!(pml4e & PG_PRESENT_MASK)) {
-                goto do_fault;
-            }
-            if (pml4e & (rsvd_mask | PG_PSE_MASK)) {
-                goto do_fault_rsvd;
-            }
-            if (!(pml4e & PG_ACCESSED_MASK)) {
-                pml4e |= PG_ACCESSED_MASK;
-                x86_stl_phys_notdirty(cs, pml4e_addr, pml4e);
-            }
-            ptep &= pml4e ^ PG_NX_MASK;
-            pdpe_addr = ((pml4e & PG_ADDRESS_MASK) + (((addr >> 30) & 0x1ff) << 3)) &
-                a20_mask;
-            pdpe_addr = get_hphys(cs, pdpe_addr, MMU_DATA_STORE, NULL);
-            pdpe = x86_ldq_phys(cs, pdpe_addr);
-            if (!(pdpe & PG_PRESENT_MASK)) {
-                goto do_fault;
-            }
-            if (pdpe & rsvd_mask) {
-                goto do_fault_rsvd;
-            }
-            ptep &= pdpe ^ PG_NX_MASK;
-            if (!(pdpe & PG_ACCESSED_MASK)) {
-                pdpe |= PG_ACCESSED_MASK;
-                x86_stl_phys_notdirty(cs, pdpe_addr, pdpe);
-            }
-            if (pdpe & PG_PSE_MASK) {
-                /* 1 GB page */
-                page_size = 1024 * 1024 * 1024;
-                pte_addr = pdpe_addr;
-                pte = pdpe;
-                goto do_check_protect;
-            }
-        } else
-#endif
-        {
-            /* XXX: load them when cr3 is loaded ? */
-            pdpe_addr = ((env->cr[3] & ~0x1f) + ((addr >> 27) & 0x18)) &
-                a20_mask;
-            pdpe_addr = get_hphys(cs, pdpe_addr, MMU_DATA_STORE, false);
-            pdpe = x86_ldq_phys(cs, pdpe_addr);
-            if (!(pdpe & PG_PRESENT_MASK)) {
-                goto do_fault;
-            }
-            rsvd_mask |= PG_HI_USER_MASK;
-            if (pdpe & (rsvd_mask | PG_NX_MASK)) {
-                goto do_fault_rsvd;
-            }
-            ptep = PG_NX_MASK | PG_USER_MASK | PG_RW_MASK;
-        }
-
-        pde_addr = ((pdpe & PG_ADDRESS_MASK) + (((addr >> 21) & 0x1ff) << 3)) &
-            a20_mask;
-        pde_addr = get_hphys(cs, pde_addr, MMU_DATA_STORE, NULL);
-        pde = x86_ldq_phys(cs, pde_addr);
-        if (!(pde & PG_PRESENT_MASK)) {
-            goto do_fault;
-        }
-        if (pde & rsvd_mask) {
-            goto do_fault_rsvd;
-        }
-        ptep &= pde ^ PG_NX_MASK;
-        if (pde & PG_PSE_MASK) {
-            /* 2 MB page */
-            page_size = 2048 * 1024;
-            pte_addr = pde_addr;
-            pte = pde;
-            goto do_check_protect;
-        }
-        /* 4 KB page */
-        if (!(pde & PG_ACCESSED_MASK)) {
-            pde |= PG_ACCESSED_MASK;
-            x86_stl_phys_notdirty(cs, pde_addr, pde);
-        }
-        pte_addr = ((pde & PG_ADDRESS_MASK) + (((addr >> 12) & 0x1ff) << 3)) &
-            a20_mask;
-        pte_addr = get_hphys(cs, pte_addr, MMU_DATA_STORE, NULL);
-        pte = x86_ldq_phys(cs, pte_addr);
-        if (!(pte & PG_PRESENT_MASK)) {
-            goto do_fault;
-        }
-        if (pte & rsvd_mask) {
-            goto do_fault_rsvd;
-        }
-        /* combine pde and pte nx, user and rw protections */
-        ptep &= pte ^ PG_NX_MASK;
-        page_size = 4096;
-    } else {
-        uint32_t pde;
-
-        /* page directory entry */
-        pde_addr = ((env->cr[3] & ~0xfff) + ((addr >> 20) & 0xffc)) &
-            a20_mask;
-        pde_addr = get_hphys(cs, pde_addr, MMU_DATA_STORE, NULL);
-        pde = x86_ldl_phys(cs, pde_addr);
-        if (!(pde & PG_PRESENT_MASK)) {
-            goto do_fault;
-        }
-        ptep = pde | PG_NX_MASK;
-
-        /* if PSE bit is set, then we use a 4MB page */
-        if ((pde & PG_PSE_MASK) && (env->cr[4] & CR4_PSE_MASK)) {
-            page_size = 4096 * 1024;
-            pte_addr = pde_addr;
-
-            /* Bits 20-13 provide bits 39-32 of the address, bit 21 is reserved.
-             * Leave bits 20-13 in place for setting accessed/dirty bits below.
-             */
-            pte = pde | ((pde & 0x1fe000LL) << (32 - 13));
-            rsvd_mask = 0x200000;
-            goto do_check_protect_pse36;
-        }
-
-        if (!(pde & PG_ACCESSED_MASK)) {
-            pde |= PG_ACCESSED_MASK;
-            x86_stl_phys_notdirty(cs, pde_addr, pde);
-        }
-
-        /* page directory entry */
-        pte_addr = ((pde & ~0xfff) + ((addr >> 10) & 0xffc)) &
-            a20_mask;
-        pte_addr = get_hphys(cs, pte_addr, MMU_DATA_STORE, NULL);
-        pte = x86_ldl_phys(cs, pte_addr);
-        if (!(pte & PG_PRESENT_MASK)) {
-            goto do_fault;
-        }
-        /* combine pde and pte user and rw protections */
-        ptep &= pte | PG_NX_MASK;
-        page_size = 4096;
-        rsvd_mask = 0;
-    }
-
-do_check_protect:
-    rsvd_mask |= (page_size - 1) & PG_ADDRESS_MASK & ~PG_PSE_PAT_MASK;
-do_check_protect_pse36:
-    if (pte & rsvd_mask) {
-        goto do_fault_rsvd;
-    }
-    ptep ^= PG_NX_MASK;
-
-    /* can the page can be put in the TLB?  prot will tell us */
-    if (is_user && !(ptep & PG_USER_MASK)) {
-        goto do_fault_protect;
-    }
-
-    prot = 0;
-    if (mmu_idx != MMU_KSMAP_IDX || !(ptep & PG_USER_MASK)) {
-        prot |= PAGE_READ;
-        if ((ptep & PG_RW_MASK) || (!is_user && !(env->cr[0] & CR0_WP_MASK))) {
-            prot |= PAGE_WRITE;
-        }
-    }
-    if (!(ptep & PG_NX_MASK) &&
-        (mmu_idx == MMU_USER_IDX ||
-         !((env->cr[4] & CR4_SMEP_MASK) && (ptep & PG_USER_MASK)))) {
-        prot |= PAGE_EXEC;
-    }
-
-    if (!(env->hflags & HF_LMA_MASK)) {
-        pkr = 0;
-    } else if (ptep & PG_USER_MASK) {
-        pkr = env->cr[4] & CR4_PKE_MASK ? env->pkru : 0;
-    } else {
-        pkr = env->cr[4] & CR4_PKS_MASK ? env->pkrs : 0;
-    }
-    if (pkr) {
-        uint32_t pk = (pte & PG_PKRU_MASK) >> PG_PKRU_BIT;
-        uint32_t pkr_ad = (pkr >> pk * 2) & 1;
-        uint32_t pkr_wd = (pkr >> pk * 2) & 2;
-        uint32_t pkr_prot = PAGE_READ | PAGE_WRITE | PAGE_EXEC;
-
-        if (pkr_ad) {
-            pkr_prot &= ~(PAGE_READ | PAGE_WRITE);
-        } else if (pkr_wd && (is_user || env->cr[0] & CR0_WP_MASK)) {
-            pkr_prot &= ~PAGE_WRITE;
-        }
-
-        prot &= pkr_prot;
-        if ((pkr_prot & (1 << is_write1)) == 0) {
-            assert(is_write1 != 2);
-            error_code |= PG_ERROR_PK_MASK;
-            goto do_fault_protect;
-        }
-    }
-
-    if ((prot & (1 << is_write1)) == 0) {
-        goto do_fault_protect;
-    }
-
-    /* yes, it can! */
-    is_dirty = is_write && !(pte & PG_DIRTY_MASK);
-    if (!(pte & PG_ACCESSED_MASK) || is_dirty) {
-        pte |= PG_ACCESSED_MASK;
-        if (is_dirty) {
-            pte |= PG_DIRTY_MASK;
-        }
-        x86_stl_phys_notdirty(cs, pte_addr, pte);
-    }
-
-    if (!(pte & PG_DIRTY_MASK)) {
-        /* only set write access if already dirty... otherwise wait
-           for dirty access */
-        assert(!is_write);
-        prot &= ~PAGE_WRITE;
-    }
-
- do_mapping:
-    pte = pte & a20_mask;
-
-    /* align to page_size */
-    pte &= PG_ADDRESS_MASK & ~(page_size - 1);
-    page_offset = addr & (page_size - 1);
-    paddr = get_hphys(cs, pte + page_offset, is_write1, &prot);
-
-    /* Even if 4MB pages, we map only one 4KB page in the cache to
-       avoid filling it too fast */
-    vaddr = addr & TARGET_PAGE_MASK;
-    paddr &= TARGET_PAGE_MASK;
-
-    assert(prot & (1 << is_write1));
-    tlb_set_page_with_attrs(cs, vaddr, paddr, cpu_get_mem_attrs(env),
-                            prot, mmu_idx, page_size);
-    return 0;
- do_fault_rsvd:
-    error_code |= PG_ERROR_RSVD_MASK;
- do_fault_protect:
-    error_code |= PG_ERROR_P_MASK;
- do_fault:
-    error_code |= (is_write << PG_ERROR_W_BIT);
-    if (is_user)
-        error_code |= PG_ERROR_U_MASK;
-    if (is_write1 == 2 &&
-        (((env->efer & MSR_EFER_NXE) &&
-          (env->cr[4] & CR4_PAE_MASK)) ||
-         (env->cr[4] & CR4_SMEP_MASK)))
-        error_code |= PG_ERROR_I_D_MASK;
-    if (env->intercept_exceptions & (1 << EXCP0E_PAGE)) {
-        /* cr2 is not modified in case of exceptions */
-        x86_stq_phys(cs,
-                 env->vm_vmcb + offsetof(struct vmcb, control.exit_info_2),
-                 addr);
-    } else {
-        env->cr[2] = addr;
-    }
-    env->error_code = error_code;
-    cs->exception_index = EXCP0E_PAGE;
-    return 1;
-}
-#endif
-
-bool x86_cpu_tlb_fill(CPUState *cs, vaddr addr, int size,
-                      MMUAccessType access_type, int mmu_idx,
-                      bool probe, uintptr_t retaddr)
-{
-    X86CPU *cpu = X86_CPU(cs);
-    CPUX86State *env = &cpu->env;
-
-#ifdef CONFIG_USER_ONLY
-    /* user mode only emulation */
-    env->cr[2] = addr;
-    env->error_code = (access_type == MMU_DATA_STORE) << PG_ERROR_W_BIT;
-    env->error_code |= PG_ERROR_U_MASK;
-    cs->exception_index = EXCP0E_PAGE;
-    env->exception_is_int = 0;
-    env->exception_next_eip = -1;
-    cpu_loop_exit_restore(cs, retaddr);
-#else
-    env->retaddr = retaddr;
-    if (handle_mmu_fault(cs, addr, size, access_type, mmu_idx)) {
-        /* FIXME: On error in get_hphys we have already jumped out.  */
-        g_assert(!probe);
-        raise_exception_err_ra(env, cs->exception_index,
-                               env->error_code, retaddr);
-    }
-    return true;
-#endif
-}
diff --git a/target/i386/tcg/sysemu/excp_helper.c b/target/i386/tcg/sysemu/excp_helper.c
new file mode 100644
index 0000000000..37e6eead05
--- /dev/null
+++ b/target/i386/tcg/sysemu/excp_helper.c
@@ -0,0 +1,581 @@
+/*
+ *  x86 exception helpers - sysemu code
+ *
+ *  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 2.1 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/helper-tcg.h"
+
+static hwaddr get_hphys(CPUState *cs, hwaddr gphys, MMUAccessType access_type,
+                        int *prot)
+{
+    CPUX86State *env = &X86_CPU(cs)->env;
+    uint64_t rsvd_mask = PG_HI_RSVD_MASK;
+    uint64_t ptep, pte;
+    uint64_t exit_info_1 = 0;
+    target_ulong pde_addr, pte_addr;
+    uint32_t page_offset;
+    int page_size;
+
+    if (likely(!(env->hflags2 & HF2_NPT_MASK))) {
+        return gphys;
+    }
+
+    if (!(env->nested_pg_mode & SVM_NPT_NXE)) {
+        rsvd_mask |= PG_NX_MASK;
+    }
+
+    if (env->nested_pg_mode & SVM_NPT_PAE) {
+        uint64_t pde, pdpe;
+        target_ulong pdpe_addr;
+
+#ifdef TARGET_X86_64
+        if (env->nested_pg_mode & SVM_NPT_LMA) {
+            uint64_t pml5e;
+            uint64_t pml4e_addr, pml4e;
+
+            pml5e = env->nested_cr3;
+            ptep = PG_NX_MASK | PG_USER_MASK | PG_RW_MASK;
+
+            pml4e_addr = (pml5e & PG_ADDRESS_MASK) +
+                    (((gphys >> 39) & 0x1ff) << 3);
+            pml4e = x86_ldq_phys(cs, pml4e_addr);
+            if (!(pml4e & PG_PRESENT_MASK)) {
+                goto do_fault;
+            }
+            if (pml4e & (rsvd_mask | PG_PSE_MASK)) {
+                goto do_fault_rsvd;
+            }
+            if (!(pml4e & PG_ACCESSED_MASK)) {
+                pml4e |= PG_ACCESSED_MASK;
+                x86_stl_phys_notdirty(cs, pml4e_addr, pml4e);
+            }
+            ptep &= pml4e ^ PG_NX_MASK;
+            pdpe_addr = (pml4e & PG_ADDRESS_MASK) +
+                    (((gphys >> 30) & 0x1ff) << 3);
+            pdpe = x86_ldq_phys(cs, pdpe_addr);
+            if (!(pdpe & PG_PRESENT_MASK)) {
+                goto do_fault;
+            }
+            if (pdpe & rsvd_mask) {
+                goto do_fault_rsvd;
+            }
+            ptep &= pdpe ^ PG_NX_MASK;
+            if (!(pdpe & PG_ACCESSED_MASK)) {
+                pdpe |= PG_ACCESSED_MASK;
+                x86_stl_phys_notdirty(cs, pdpe_addr, pdpe);
+            }
+            if (pdpe & PG_PSE_MASK) {
+                /* 1 GB page */
+                page_size = 1024 * 1024 * 1024;
+                pte_addr = pdpe_addr;
+                pte = pdpe;
+                goto do_check_protect;
+            }
+        } else
+#endif
+        {
+            pdpe_addr = (env->nested_cr3 & ~0x1f) + ((gphys >> 27) & 0x18);
+            pdpe = x86_ldq_phys(cs, pdpe_addr);
+            if (!(pdpe & PG_PRESENT_MASK)) {
+                goto do_fault;
+            }
+            rsvd_mask |= PG_HI_USER_MASK;
+            if (pdpe & (rsvd_mask | PG_NX_MASK)) {
+                goto do_fault_rsvd;
+            }
+            ptep = PG_NX_MASK | PG_USER_MASK | PG_RW_MASK;
+        }
+
+        pde_addr = (pdpe & PG_ADDRESS_MASK) + (((gphys >> 21) & 0x1ff) << 3);
+        pde = x86_ldq_phys(cs, pde_addr);
+        if (!(pde & PG_PRESENT_MASK)) {
+            goto do_fault;
+        }
+        if (pde & rsvd_mask) {
+            goto do_fault_rsvd;
+        }
+        ptep &= pde ^ PG_NX_MASK;
+        if (pde & PG_PSE_MASK) {
+            /* 2 MB page */
+            page_size = 2048 * 1024;
+            pte_addr = pde_addr;
+            pte = pde;
+            goto do_check_protect;
+        }
+        /* 4 KB page */
+        if (!(pde & PG_ACCESSED_MASK)) {
+            pde |= PG_ACCESSED_MASK;
+            x86_stl_phys_notdirty(cs, pde_addr, pde);
+        }
+        pte_addr = (pde & PG_ADDRESS_MASK) + (((gphys >> 12) & 0x1ff) << 3);
+        pte = x86_ldq_phys(cs, pte_addr);
+        if (!(pte & PG_PRESENT_MASK)) {
+            goto do_fault;
+        }
+        if (pte & rsvd_mask) {
+            goto do_fault_rsvd;
+        }
+        /* combine pde and pte nx, user and rw protections */
+        ptep &= pte ^ PG_NX_MASK;
+        page_size = 4096;
+    } else {
+        uint32_t pde;
+
+        /* page directory entry */
+        pde_addr = (env->nested_cr3 & ~0xfff) + ((gphys >> 20) & 0xffc);
+        pde = x86_ldl_phys(cs, pde_addr);
+        if (!(pde & PG_PRESENT_MASK)) {
+            goto do_fault;
+        }
+        ptep = pde | PG_NX_MASK;
+
+        /* if host cr4 PSE bit is set, then we use a 4MB page */
+        if ((pde & PG_PSE_MASK) && (env->nested_pg_mode & SVM_NPT_PSE)) {
+            page_size = 4096 * 1024;
+            pte_addr = pde_addr;
+
+            /* Bits 20-13 provide bits 39-32 of the address, bit 21 is reserved.
+             * Leave bits 20-13 in place for setting accessed/dirty bits below.
+             */
+            pte = pde | ((pde & 0x1fe000LL) << (32 - 13));
+            rsvd_mask = 0x200000;
+            goto do_check_protect_pse36;
+        }
+
+        if (!(pde & PG_ACCESSED_MASK)) {
+            pde |= PG_ACCESSED_MASK;
+            x86_stl_phys_notdirty(cs, pde_addr, pde);
+        }
+
+        /* page directory entry */
+        pte_addr = (pde & ~0xfff) + ((gphys >> 10) & 0xffc);
+        pte = x86_ldl_phys(cs, pte_addr);
+        if (!(pte & PG_PRESENT_MASK)) {
+            goto do_fault;
+        }
+        /* combine pde and pte user and rw protections */
+        ptep &= pte | PG_NX_MASK;
+        page_size = 4096;
+        rsvd_mask = 0;
+    }
+
+ do_check_protect:
+    rsvd_mask |= (page_size - 1) & PG_ADDRESS_MASK & ~PG_PSE_PAT_MASK;
+ do_check_protect_pse36:
+    if (pte & rsvd_mask) {
+        goto do_fault_rsvd;
+    }
+    ptep ^= PG_NX_MASK;
+
+    if (!(ptep & PG_USER_MASK)) {
+        goto do_fault_protect;
+    }
+    if (ptep & PG_NX_MASK) {
+        if (access_type == MMU_INST_FETCH) {
+            goto do_fault_protect;
+        }
+        *prot &= ~PAGE_EXEC;
+    }
+    if (!(ptep & PG_RW_MASK)) {
+        if (access_type == MMU_DATA_STORE) {
+            goto do_fault_protect;
+        }
+        *prot &= ~PAGE_WRITE;
+    }
+
+    pte &= PG_ADDRESS_MASK & ~(page_size - 1);
+    page_offset = gphys & (page_size - 1);
+    return pte + page_offset;
+
+ do_fault_rsvd:
+    exit_info_1 |= SVM_NPTEXIT_RSVD;
+ do_fault_protect:
+    exit_info_1 |= SVM_NPTEXIT_P;
+ do_fault:
+    x86_stq_phys(cs, env->vm_vmcb + offsetof(struct vmcb, control.exit_info_2),
+                 gphys);
+    exit_info_1 |= SVM_NPTEXIT_US;
+    if (access_type == MMU_DATA_STORE) {
+        exit_info_1 |= SVM_NPTEXIT_RW;
+    } else if (access_type == MMU_INST_FETCH) {
+        exit_info_1 |= SVM_NPTEXIT_ID;
+    }
+    if (prot) {
+        exit_info_1 |= SVM_NPTEXIT_GPA;
+    } else { /* page table access */
+        exit_info_1 |= SVM_NPTEXIT_GPT;
+    }
+    cpu_vmexit(env, SVM_EXIT_NPF, exit_info_1, env->retaddr);
+}
+
+/* return value:
+ * -1 = cannot handle fault
+ * 0  = nothing more to do
+ * 1  = generate PF fault
+ */
+static int handle_mmu_fault(CPUState *cs, vaddr addr, int size,
+                            int is_write1, int mmu_idx)
+{
+    X86CPU *cpu = X86_CPU(cs);
+    CPUX86State *env = &cpu->env;
+    uint64_t ptep, pte;
+    int32_t a20_mask;
+    target_ulong pde_addr, pte_addr;
+    int error_code = 0;
+    int is_dirty, prot, page_size, is_write, is_user;
+    hwaddr paddr;
+    uint64_t rsvd_mask = PG_HI_RSVD_MASK;
+    uint32_t page_offset;
+    target_ulong vaddr;
+    uint32_t pkr;
+
+    is_user = mmu_idx == MMU_USER_IDX;
+#if defined(DEBUG_MMU)
+    printf("MMU fault: addr=%" VADDR_PRIx " w=%d u=%d eip=" TARGET_FMT_lx "\n",
+           addr, is_write1, is_user, env->eip);
+#endif
+    is_write = is_write1 & 1;
+
+    a20_mask = x86_get_a20_mask(env);
+    if (!(env->cr[0] & CR0_PG_MASK)) {
+        pte = addr;
+#ifdef TARGET_X86_64
+        if (!(env->hflags & HF_LMA_MASK)) {
+            /* Without long mode we can only address 32bits in real mode */
+            pte = (uint32_t)pte;
+        }
+#endif
+        prot = PAGE_READ | PAGE_WRITE | PAGE_EXEC;
+        page_size = 4096;
+        goto do_mapping;
+    }
+
+    if (!(env->efer & MSR_EFER_NXE)) {
+        rsvd_mask |= PG_NX_MASK;
+    }
+
+    if (env->cr[4] & CR4_PAE_MASK) {
+        uint64_t pde, pdpe;
+        target_ulong pdpe_addr;
+
+#ifdef TARGET_X86_64
+        if (env->hflags & HF_LMA_MASK) {
+            bool la57 = env->cr[4] & CR4_LA57_MASK;
+            uint64_t pml5e_addr, pml5e;
+            uint64_t pml4e_addr, pml4e;
+            int32_t sext;
+
+            /* test virtual address sign extension */
+            sext = la57 ? (int64_t)addr >> 56 : (int64_t)addr >> 47;
+            if (sext != 0 && sext != -1) {
+                env->error_code = 0;
+                cs->exception_index = EXCP0D_GPF;
+                return 1;
+            }
+
+            if (la57) {
+                pml5e_addr = ((env->cr[3] & ~0xfff) +
+                        (((addr >> 48) & 0x1ff) << 3)) & a20_mask;
+                pml5e_addr = get_hphys(cs, pml5e_addr, MMU_DATA_STORE, NULL);
+                pml5e = x86_ldq_phys(cs, pml5e_addr);
+                if (!(pml5e & PG_PRESENT_MASK)) {
+                    goto do_fault;
+                }
+                if (pml5e & (rsvd_mask | PG_PSE_MASK)) {
+                    goto do_fault_rsvd;
+                }
+                if (!(pml5e & PG_ACCESSED_MASK)) {
+                    pml5e |= PG_ACCESSED_MASK;
+                    x86_stl_phys_notdirty(cs, pml5e_addr, pml5e);
+                }
+                ptep = pml5e ^ PG_NX_MASK;
+            } else {
+                pml5e = env->cr[3];
+                ptep = PG_NX_MASK | PG_USER_MASK | PG_RW_MASK;
+            }
+
+            pml4e_addr = ((pml5e & PG_ADDRESS_MASK) +
+                    (((addr >> 39) & 0x1ff) << 3)) & a20_mask;
+            pml4e_addr = get_hphys(cs, pml4e_addr, MMU_DATA_STORE, false);
+            pml4e = x86_ldq_phys(cs, pml4e_addr);
+            if (!(pml4e & PG_PRESENT_MASK)) {
+                goto do_fault;
+            }
+            if (pml4e & (rsvd_mask | PG_PSE_MASK)) {
+                goto do_fault_rsvd;
+            }
+            if (!(pml4e & PG_ACCESSED_MASK)) {
+                pml4e |= PG_ACCESSED_MASK;
+                x86_stl_phys_notdirty(cs, pml4e_addr, pml4e);
+            }
+            ptep &= pml4e ^ PG_NX_MASK;
+            pdpe_addr = ((pml4e & PG_ADDRESS_MASK) + (((addr >> 30) & 0x1ff) << 3)) &
+                a20_mask;
+            pdpe_addr = get_hphys(cs, pdpe_addr, MMU_DATA_STORE, NULL);
+            pdpe = x86_ldq_phys(cs, pdpe_addr);
+            if (!(pdpe & PG_PRESENT_MASK)) {
+                goto do_fault;
+            }
+            if (pdpe & rsvd_mask) {
+                goto do_fault_rsvd;
+            }
+            ptep &= pdpe ^ PG_NX_MASK;
+            if (!(pdpe & PG_ACCESSED_MASK)) {
+                pdpe |= PG_ACCESSED_MASK;
+                x86_stl_phys_notdirty(cs, pdpe_addr, pdpe);
+            }
+            if (pdpe & PG_PSE_MASK) {
+                /* 1 GB page */
+                page_size = 1024 * 1024 * 1024;
+                pte_addr = pdpe_addr;
+                pte = pdpe;
+                goto do_check_protect;
+            }
+        } else
+#endif
+        {
+            /* XXX: load them when cr3 is loaded ? */
+            pdpe_addr = ((env->cr[3] & ~0x1f) + ((addr >> 27) & 0x18)) &
+                a20_mask;
+            pdpe_addr = get_hphys(cs, pdpe_addr, MMU_DATA_STORE, false);
+            pdpe = x86_ldq_phys(cs, pdpe_addr);
+            if (!(pdpe & PG_PRESENT_MASK)) {
+                goto do_fault;
+            }
+            rsvd_mask |= PG_HI_USER_MASK;
+            if (pdpe & (rsvd_mask | PG_NX_MASK)) {
+                goto do_fault_rsvd;
+            }
+            ptep = PG_NX_MASK | PG_USER_MASK | PG_RW_MASK;
+        }
+
+        pde_addr = ((pdpe & PG_ADDRESS_MASK) + (((addr >> 21) & 0x1ff) << 3)) &
+            a20_mask;
+        pde_addr = get_hphys(cs, pde_addr, MMU_DATA_STORE, NULL);
+        pde = x86_ldq_phys(cs, pde_addr);
+        if (!(pde & PG_PRESENT_MASK)) {
+            goto do_fault;
+        }
+        if (pde & rsvd_mask) {
+            goto do_fault_rsvd;
+        }
+        ptep &= pde ^ PG_NX_MASK;
+        if (pde & PG_PSE_MASK) {
+            /* 2 MB page */
+            page_size = 2048 * 1024;
+            pte_addr = pde_addr;
+            pte = pde;
+            goto do_check_protect;
+        }
+        /* 4 KB page */
+        if (!(pde & PG_ACCESSED_MASK)) {
+            pde |= PG_ACCESSED_MASK;
+            x86_stl_phys_notdirty(cs, pde_addr, pde);
+        }
+        pte_addr = ((pde & PG_ADDRESS_MASK) + (((addr >> 12) & 0x1ff) << 3)) &
+            a20_mask;
+        pte_addr = get_hphys(cs, pte_addr, MMU_DATA_STORE, NULL);
+        pte = x86_ldq_phys(cs, pte_addr);
+        if (!(pte & PG_PRESENT_MASK)) {
+            goto do_fault;
+        }
+        if (pte & rsvd_mask) {
+            goto do_fault_rsvd;
+        }
+        /* combine pde and pte nx, user and rw protections */
+        ptep &= pte ^ PG_NX_MASK;
+        page_size = 4096;
+    } else {
+        uint32_t pde;
+
+        /* page directory entry */
+        pde_addr = ((env->cr[3] & ~0xfff) + ((addr >> 20) & 0xffc)) &
+            a20_mask;
+        pde_addr = get_hphys(cs, pde_addr, MMU_DATA_STORE, NULL);
+        pde = x86_ldl_phys(cs, pde_addr);
+        if (!(pde & PG_PRESENT_MASK)) {
+            goto do_fault;
+        }
+        ptep = pde | PG_NX_MASK;
+
+        /* if PSE bit is set, then we use a 4MB page */
+        if ((pde & PG_PSE_MASK) && (env->cr[4] & CR4_PSE_MASK)) {
+            page_size = 4096 * 1024;
+            pte_addr = pde_addr;
+
+            /* Bits 20-13 provide bits 39-32 of the address, bit 21 is reserved.
+             * Leave bits 20-13 in place for setting accessed/dirty bits below.
+             */
+            pte = pde | ((pde & 0x1fe000LL) << (32 - 13));
+            rsvd_mask = 0x200000;
+            goto do_check_protect_pse36;
+        }
+
+        if (!(pde & PG_ACCESSED_MASK)) {
+            pde |= PG_ACCESSED_MASK;
+            x86_stl_phys_notdirty(cs, pde_addr, pde);
+        }
+
+        /* page directory entry */
+        pte_addr = ((pde & ~0xfff) + ((addr >> 10) & 0xffc)) &
+            a20_mask;
+        pte_addr = get_hphys(cs, pte_addr, MMU_DATA_STORE, NULL);
+        pte = x86_ldl_phys(cs, pte_addr);
+        if (!(pte & PG_PRESENT_MASK)) {
+            goto do_fault;
+        }
+        /* combine pde and pte user and rw protections */
+        ptep &= pte | PG_NX_MASK;
+        page_size = 4096;
+        rsvd_mask = 0;
+    }
+
+do_check_protect:
+    rsvd_mask |= (page_size - 1) & PG_ADDRESS_MASK & ~PG_PSE_PAT_MASK;
+do_check_protect_pse36:
+    if (pte & rsvd_mask) {
+        goto do_fault_rsvd;
+    }
+    ptep ^= PG_NX_MASK;
+
+    /* can the page can be put in the TLB?  prot will tell us */
+    if (is_user && !(ptep & PG_USER_MASK)) {
+        goto do_fault_protect;
+    }
+
+    prot = 0;
+    if (mmu_idx != MMU_KSMAP_IDX || !(ptep & PG_USER_MASK)) {
+        prot |= PAGE_READ;
+        if ((ptep & PG_RW_MASK) || (!is_user && !(env->cr[0] & CR0_WP_MASK))) {
+            prot |= PAGE_WRITE;
+        }
+    }
+    if (!(ptep & PG_NX_MASK) &&
+        (mmu_idx == MMU_USER_IDX ||
+         !((env->cr[4] & CR4_SMEP_MASK) && (ptep & PG_USER_MASK)))) {
+        prot |= PAGE_EXEC;
+    }
+
+    if (!(env->hflags & HF_LMA_MASK)) {
+        pkr = 0;
+    } else if (ptep & PG_USER_MASK) {
+        pkr = env->cr[4] & CR4_PKE_MASK ? env->pkru : 0;
+    } else {
+        pkr = env->cr[4] & CR4_PKS_MASK ? env->pkrs : 0;
+    }
+    if (pkr) {
+        uint32_t pk = (pte & PG_PKRU_MASK) >> PG_PKRU_BIT;
+        uint32_t pkr_ad = (pkr >> pk * 2) & 1;
+        uint32_t pkr_wd = (pkr >> pk * 2) & 2;
+        uint32_t pkr_prot = PAGE_READ | PAGE_WRITE | PAGE_EXEC;
+
+        if (pkr_ad) {
+            pkr_prot &= ~(PAGE_READ | PAGE_WRITE);
+        } else if (pkr_wd && (is_user || env->cr[0] & CR0_WP_MASK)) {
+            pkr_prot &= ~PAGE_WRITE;
+        }
+
+        prot &= pkr_prot;
+        if ((pkr_prot & (1 << is_write1)) == 0) {
+            assert(is_write1 != 2);
+            error_code |= PG_ERROR_PK_MASK;
+            goto do_fault_protect;
+        }
+    }
+
+    if ((prot & (1 << is_write1)) == 0) {
+        goto do_fault_protect;
+    }
+
+    /* yes, it can! */
+    is_dirty = is_write && !(pte & PG_DIRTY_MASK);
+    if (!(pte & PG_ACCESSED_MASK) || is_dirty) {
+        pte |= PG_ACCESSED_MASK;
+        if (is_dirty) {
+            pte |= PG_DIRTY_MASK;
+        }
+        x86_stl_phys_notdirty(cs, pte_addr, pte);
+    }
+
+    if (!(pte & PG_DIRTY_MASK)) {
+        /* only set write access if already dirty... otherwise wait
+           for dirty access */
+        assert(!is_write);
+        prot &= ~PAGE_WRITE;
+    }
+
+ do_mapping:
+    pte = pte & a20_mask;
+
+    /* align to page_size */
+    pte &= PG_ADDRESS_MASK & ~(page_size - 1);
+    page_offset = addr & (page_size - 1);
+    paddr = get_hphys(cs, pte + page_offset, is_write1, &prot);
+
+    /* Even if 4MB pages, we map only one 4KB page in the cache to
+       avoid filling it too fast */
+    vaddr = addr & TARGET_PAGE_MASK;
+    paddr &= TARGET_PAGE_MASK;
+
+    assert(prot & (1 << is_write1));
+    tlb_set_page_with_attrs(cs, vaddr, paddr, cpu_get_mem_attrs(env),
+                            prot, mmu_idx, page_size);
+    return 0;
+ do_fault_rsvd:
+    error_code |= PG_ERROR_RSVD_MASK;
+ do_fault_protect:
+    error_code |= PG_ERROR_P_MASK;
+ do_fault:
+    error_code |= (is_write << PG_ERROR_W_BIT);
+    if (is_user)
+        error_code |= PG_ERROR_U_MASK;
+    if (is_write1 == 2 &&
+        (((env->efer & MSR_EFER_NXE) &&
+          (env->cr[4] & CR4_PAE_MASK)) ||
+         (env->cr[4] & CR4_SMEP_MASK)))
+        error_code |= PG_ERROR_I_D_MASK;
+    if (env->intercept_exceptions & (1 << EXCP0E_PAGE)) {
+        /* cr2 is not modified in case of exceptions */
+        x86_stq_phys(cs,
+                 env->vm_vmcb + offsetof(struct vmcb, control.exit_info_2),
+                 addr);
+    } else {
+        env->cr[2] = addr;
+    }
+    env->error_code = error_code;
+    cs->exception_index = EXCP0E_PAGE;
+    return 1;
+}
+
+bool x86_cpu_tlb_fill(CPUState *cs, vaddr addr, int size,
+                      MMUAccessType access_type, int mmu_idx,
+                      bool probe, uintptr_t retaddr)
+{
+    X86CPU *cpu = X86_CPU(cs);
+    CPUX86State *env = &cpu->env;
+
+    env->retaddr = retaddr;
+    if (handle_mmu_fault(cs, addr, size, access_type, mmu_idx)) {
+        /* FIXME: On error in get_hphys we have already jumped out.  */
+        g_assert(!probe);
+        raise_exception_err_ra(env, cs->exception_index,
+                               env->error_code, retaddr);
+    }
+    return true;
+}
diff --git a/target/i386/tcg/user/excp_helper.c b/target/i386/tcg/user/excp_helper.c
new file mode 100644
index 0000000000..a89b5228fd
--- /dev/null
+++ b/target/i386/tcg/user/excp_helper.c
@@ -0,0 +1,39 @@
+/*
+ *  x86 exception helpers - user-mode specific code
+ *
+ *  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 2.1 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 "exec/exec-all.h"
+#include "tcg/helper-tcg.h"
+
+bool x86_cpu_tlb_fill(CPUState *cs, vaddr addr, int size,
+                      MMUAccessType access_type, int mmu_idx,
+                      bool probe, uintptr_t retaddr)
+{
+    X86CPU *cpu = X86_CPU(cs);
+    CPUX86State *env = &cpu->env;
+
+    env->cr[2] = addr;
+    env->error_code = (access_type == MMU_DATA_STORE) << PG_ERROR_W_BIT;
+    env->error_code |= PG_ERROR_U_MASK;
+    cs->exception_index = EXCP0E_PAGE;
+    env->exception_is_int = 0;
+    env->exception_next_eip = -1;
+    cpu_loop_exit_restore(cs, retaddr);
+}
diff --git a/target/i386/tcg/sysemu/meson.build b/target/i386/tcg/sysemu/meson.build
index 35ba16dc3d..6d0a0a0fee 100644
--- a/target/i386/tcg/sysemu/meson.build
+++ b/target/i386/tcg/sysemu/meson.build
@@ -1,4 +1,5 @@
 i386_softmmu_ss.add(when: ['CONFIG_TCG', 'CONFIG_SOFTMMU'], if_true: files(
   'tcg-cpu.c',
   'smm_helper.c',
+  'excp_helper.c',
 ))
diff --git a/target/i386/tcg/user/meson.build b/target/i386/tcg/user/meson.build
index 7aecc53155..e0ef0f02e2 100644
--- a/target/i386/tcg/user/meson.build
+++ b/target/i386/tcg/user/meson.build
@@ -1,2 +1,3 @@
 i386_user_ss.add(when: ['CONFIG_TCG', 'CONFIG_USER_ONLY'], if_true: files(
+  'excp_helper.c',
 ))
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 12/20] i386: move TCG bpt_helper into sysemu/
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (10 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 11/20] i386: split tcg excp_helper into sysemu and user parts Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 13/20] i386: split misc helper user stubs and sysemu part Claudio Fontana
                   ` (9 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

for user-mode, assert that the hidden IOBPT flags are not set
while attempting to generate io_bpt helpers.

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/helper.h                |   7 +
 target/i386/tcg/helper-tcg.h        |   3 +
 target/i386/tcg/bpt_helper.c        | 276 --------------------------
 target/i386/tcg/sysemu/bpt_helper.c | 293 ++++++++++++++++++++++++++++
 target/i386/tcg/translate.c         |   8 +-
 target/i386/tcg/sysemu/meson.build  |   1 +
 6 files changed, 311 insertions(+), 277 deletions(-)
 create mode 100644 target/i386/tcg/sysemu/bpt_helper.c

diff --git a/target/i386/helper.h b/target/i386/helper.h
index 8ffda4cdc6..095520f81f 100644
--- a/target/i386/helper.h
+++ b/target/i386/helper.h
@@ -46,7 +46,11 @@ DEF_HELPER_2(read_crN, tl, env, int)
 DEF_HELPER_3(write_crN, void, env, int, tl)
 DEF_HELPER_2(lmsw, void, env, tl)
 DEF_HELPER_1(clts, void, env)
+
+#ifndef CONFIG_USER_ONLY
 DEF_HELPER_FLAGS_3(set_dr, TCG_CALL_NO_WG, void, env, int, tl)
+#endif /* !CONFIG_USER_ONLY */
+
 DEF_HELPER_FLAGS_2(get_dr, TCG_CALL_NO_WG, tl, env, int)
 DEF_HELPER_2(invlpg, void, env, tl)
 
@@ -100,7 +104,10 @@ DEF_HELPER_3(outw, void, env, i32, i32)
 DEF_HELPER_2(inw, tl, env, i32)
 DEF_HELPER_3(outl, void, env, i32, i32)
 DEF_HELPER_2(inl, tl, env, i32)
+
+#ifndef CONFIG_USER_ONLY
 DEF_HELPER_FLAGS_4(bpt_io, TCG_CALL_NO_WG, void, env, i32, i32, tl)
+#endif /* !CONFIG_USER_ONLY */
 
 DEF_HELPER_3(svm_check_intercept_param, void, env, i32, i64)
 DEF_HELPER_4(svm_check_io, void, env, i32, i32, i32)
diff --git a/target/i386/tcg/helper-tcg.h b/target/i386/tcg/helper-tcg.h
index c133c63555..b420b3356d 100644
--- a/target/i386/tcg/helper-tcg.h
+++ b/target/i386/tcg/helper-tcg.h
@@ -92,4 +92,7 @@ void do_interrupt_x86_hardirq(CPUX86State *env, int intno, int is_hw);
 /* smm_helper.c */
 void do_smm_enter(X86CPU *cpu);
 
+/* bpt_helper.c */
+bool check_hw_breakpoints(CPUX86State *env, bool force_dr6_update);
+
 #endif /* I386_HELPER_TCG_H */
diff --git a/target/i386/tcg/bpt_helper.c b/target/i386/tcg/bpt_helper.c
index 979230ac12..fb2a65ac9c 100644
--- a/target/i386/tcg/bpt_helper.c
+++ b/target/i386/tcg/bpt_helper.c
@@ -19,223 +19,9 @@
 
 #include "qemu/osdep.h"
 #include "cpu.h"
-#include "exec/exec-all.h"
 #include "exec/helper-proto.h"
 #include "helper-tcg.h"
 
-
-#ifndef CONFIG_USER_ONLY
-static inline bool hw_local_breakpoint_enabled(unsigned long dr7, int index)
-{
-    return (dr7 >> (index * 2)) & 1;
-}
-
-static inline bool hw_global_breakpoint_enabled(unsigned long dr7, int index)
-{
-    return (dr7 >> (index * 2)) & 2;
-
-}
-static inline bool hw_breakpoint_enabled(unsigned long dr7, int index)
-{
-    return hw_global_breakpoint_enabled(dr7, index) ||
-           hw_local_breakpoint_enabled(dr7, index);
-}
-
-static inline int hw_breakpoint_type(unsigned long dr7, int index)
-{
-    return (dr7 >> (DR7_TYPE_SHIFT + (index * 4))) & 3;
-}
-
-static inline int hw_breakpoint_len(unsigned long dr7, int index)
-{
-    int len = ((dr7 >> (DR7_LEN_SHIFT + (index * 4))) & 3);
-    return (len == 2) ? 8 : len + 1;
-}
-
-static int hw_breakpoint_insert(CPUX86State *env, int index)
-{
-    CPUState *cs = env_cpu(env);
-    target_ulong dr7 = env->dr[7];
-    target_ulong drN = env->dr[index];
-    int err = 0;
-
-    switch (hw_breakpoint_type(dr7, index)) {
-    case DR7_TYPE_BP_INST:
-        if (hw_breakpoint_enabled(dr7, index)) {
-            err = cpu_breakpoint_insert(cs, drN, BP_CPU,
-                                        &env->cpu_breakpoint[index]);
-        }
-        break;
-
-    case DR7_TYPE_IO_RW:
-        /* Notice when we should enable calls to bpt_io.  */
-        return hw_breakpoint_enabled(env->dr[7], index)
-               ? HF_IOBPT_MASK : 0;
-
-    case DR7_TYPE_DATA_WR:
-        if (hw_breakpoint_enabled(dr7, index)) {
-            err = cpu_watchpoint_insert(cs, drN,
-                                        hw_breakpoint_len(dr7, index),
-                                        BP_CPU | BP_MEM_WRITE,
-                                        &env->cpu_watchpoint[index]);
-        }
-        break;
-
-    case DR7_TYPE_DATA_RW:
-        if (hw_breakpoint_enabled(dr7, index)) {
-            err = cpu_watchpoint_insert(cs, drN,
-                                        hw_breakpoint_len(dr7, index),
-                                        BP_CPU | BP_MEM_ACCESS,
-                                        &env->cpu_watchpoint[index]);
-        }
-        break;
-    }
-    if (err) {
-        env->cpu_breakpoint[index] = NULL;
-    }
-    return 0;
-}
-
-static void hw_breakpoint_remove(CPUX86State *env, int index)
-{
-    CPUState *cs = env_cpu(env);
-
-    switch (hw_breakpoint_type(env->dr[7], index)) {
-    case DR7_TYPE_BP_INST:
-        if (env->cpu_breakpoint[index]) {
-            cpu_breakpoint_remove_by_ref(cs, env->cpu_breakpoint[index]);
-            env->cpu_breakpoint[index] = NULL;
-        }
-        break;
-
-    case DR7_TYPE_DATA_WR:
-    case DR7_TYPE_DATA_RW:
-        if (env->cpu_breakpoint[index]) {
-            cpu_watchpoint_remove_by_ref(cs, env->cpu_watchpoint[index]);
-            env->cpu_breakpoint[index] = NULL;
-        }
-        break;
-
-    case DR7_TYPE_IO_RW:
-        /* HF_IOBPT_MASK cleared elsewhere.  */
-        break;
-    }
-}
-
-void cpu_x86_update_dr7(CPUX86State *env, uint32_t new_dr7)
-{
-    target_ulong old_dr7 = env->dr[7];
-    int iobpt = 0;
-    int i;
-
-    new_dr7 |= DR7_FIXED_1;
-
-    /* If nothing is changing except the global/local enable bits,
-       then we can make the change more efficient.  */
-    if (((old_dr7 ^ new_dr7) & ~0xff) == 0) {
-        /* Fold the global and local enable bits together into the
-           global fields, then xor to show which registers have
-           changed collective enable state.  */
-        int mod = ((old_dr7 | old_dr7 * 2) ^ (new_dr7 | new_dr7 * 2)) & 0xff;
-
-        for (i = 0; i < DR7_MAX_BP; i++) {
-            if ((mod & (2 << i * 2)) && !hw_breakpoint_enabled(new_dr7, i)) {
-                hw_breakpoint_remove(env, i);
-            }
-        }
-        env->dr[7] = new_dr7;
-        for (i = 0; i < DR7_MAX_BP; i++) {
-            if (mod & (2 << i * 2) && hw_breakpoint_enabled(new_dr7, i)) {
-                iobpt |= hw_breakpoint_insert(env, i);
-            } else if (hw_breakpoint_type(new_dr7, i) == DR7_TYPE_IO_RW
-                       && hw_breakpoint_enabled(new_dr7, i)) {
-                iobpt |= HF_IOBPT_MASK;
-            }
-        }
-    } else {
-        for (i = 0; i < DR7_MAX_BP; i++) {
-            hw_breakpoint_remove(env, i);
-        }
-        env->dr[7] = new_dr7;
-        for (i = 0; i < DR7_MAX_BP; i++) {
-            iobpt |= hw_breakpoint_insert(env, i);
-        }
-    }
-
-    env->hflags = (env->hflags & ~HF_IOBPT_MASK) | iobpt;
-}
-
-static bool check_hw_breakpoints(CPUX86State *env, bool force_dr6_update)
-{
-    target_ulong dr6;
-    int reg;
-    bool hit_enabled = false;
-
-    dr6 = env->dr[6] & ~0xf;
-    for (reg = 0; reg < DR7_MAX_BP; reg++) {
-        bool bp_match = false;
-        bool wp_match = false;
-
-        switch (hw_breakpoint_type(env->dr[7], reg)) {
-        case DR7_TYPE_BP_INST:
-            if (env->dr[reg] == env->eip) {
-                bp_match = true;
-            }
-            break;
-        case DR7_TYPE_DATA_WR:
-        case DR7_TYPE_DATA_RW:
-            if (env->cpu_watchpoint[reg] &&
-                env->cpu_watchpoint[reg]->flags & BP_WATCHPOINT_HIT) {
-                wp_match = true;
-            }
-            break;
-        case DR7_TYPE_IO_RW:
-            break;
-        }
-        if (bp_match || wp_match) {
-            dr6 |= 1 << reg;
-            if (hw_breakpoint_enabled(env->dr[7], reg)) {
-                hit_enabled = true;
-            }
-        }
-    }
-
-    if (hit_enabled || force_dr6_update) {
-        env->dr[6] = dr6;
-    }
-
-    return hit_enabled;
-}
-
-void breakpoint_handler(CPUState *cs)
-{
-    X86CPU *cpu = X86_CPU(cs);
-    CPUX86State *env = &cpu->env;
-    CPUBreakpoint *bp;
-
-    if (cs->watchpoint_hit) {
-        if (cs->watchpoint_hit->flags & BP_CPU) {
-            cs->watchpoint_hit = NULL;
-            if (check_hw_breakpoints(env, false)) {
-                raise_exception(env, EXCP01_DB);
-            } else {
-                cpu_loop_exit_noexc(cs);
-            }
-        }
-    } else {
-        QTAILQ_FOREACH(bp, &cs->breakpoints, entry) {
-            if (bp->pc == env->eip) {
-                if (bp->flags & BP_CPU) {
-                    check_hw_breakpoints(env, true);
-                    raise_exception(env, EXCP01_DB);
-                }
-                break;
-            }
-        }
-    }
-}
-#endif
-
 void helper_single_step(CPUX86State *env)
 {
 #ifndef CONFIG_USER_ONLY
@@ -252,41 +38,6 @@ void helper_rechecking_single_step(CPUX86State *env)
     }
 }
 
-void helper_set_dr(CPUX86State *env, int reg, target_ulong t0)
-{
-#ifndef CONFIG_USER_ONLY
-    switch (reg) {
-    case 0: case 1: case 2: case 3:
-        if (hw_breakpoint_enabled(env->dr[7], reg)
-            && hw_breakpoint_type(env->dr[7], reg) != DR7_TYPE_IO_RW) {
-            hw_breakpoint_remove(env, reg);
-            env->dr[reg] = t0;
-            hw_breakpoint_insert(env, reg);
-        } else {
-            env->dr[reg] = t0;
-        }
-        return;
-    case 4:
-        if (env->cr[4] & CR4_DE_MASK) {
-            break;
-        }
-        /* fallthru */
-    case 6:
-        env->dr[6] = t0 | DR6_FIXED_1;
-        return;
-    case 5:
-        if (env->cr[4] & CR4_DE_MASK) {
-            break;
-        }
-        /* fallthru */
-    case 7:
-        cpu_x86_update_dr7(env, t0);
-        return;
-    }
-    raise_exception_err_ra(env, EXCP06_ILLOP, 0, GETPC());
-#endif
-}
-
 target_ulong helper_get_dr(CPUX86State *env, int reg)
 {
     switch (reg) {
@@ -307,30 +58,3 @@ target_ulong helper_get_dr(CPUX86State *env, int reg)
     }
     raise_exception_err_ra(env, EXCP06_ILLOP, 0, GETPC());
 }
-
-/* Check if Port I/O is trapped by a breakpoint.  */
-void helper_bpt_io(CPUX86State *env, uint32_t port,
-                   uint32_t size, target_ulong next_eip)
-{
-#ifndef CONFIG_USER_ONLY
-    target_ulong dr7 = env->dr[7];
-    int i, hit = 0;
-
-    for (i = 0; i < DR7_MAX_BP; ++i) {
-        if (hw_breakpoint_type(dr7, i) == DR7_TYPE_IO_RW
-            && hw_breakpoint_enabled(dr7, i)) {
-            int bpt_len = hw_breakpoint_len(dr7, i);
-            if (port + size - 1 >= env->dr[i]
-                && port <= env->dr[i] + bpt_len - 1) {
-                hit |= 1 << i;
-            }
-        }
-    }
-
-    if (hit) {
-        env->dr[6] = (env->dr[6] & ~0xf) | hit;
-        env->eip = next_eip;
-        raise_exception(env, EXCP01_DB);
-    }
-#endif
-}
diff --git a/target/i386/tcg/sysemu/bpt_helper.c b/target/i386/tcg/sysemu/bpt_helper.c
new file mode 100644
index 0000000000..9bdf7e170b
--- /dev/null
+++ b/target/i386/tcg/sysemu/bpt_helper.c
@@ -0,0 +1,293 @@
+/*
+ *  i386 breakpoint helpers - sysemu code
+ *
+ *  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 2.1 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 "exec/exec-all.h"
+#include "exec/helper-proto.h"
+#include "tcg/helper-tcg.h"
+
+
+static inline bool hw_local_breakpoint_enabled(unsigned long dr7, int index)
+{
+    return (dr7 >> (index * 2)) & 1;
+}
+
+static inline bool hw_global_breakpoint_enabled(unsigned long dr7, int index)
+{
+    return (dr7 >> (index * 2)) & 2;
+
+}
+static inline bool hw_breakpoint_enabled(unsigned long dr7, int index)
+{
+    return hw_global_breakpoint_enabled(dr7, index) ||
+           hw_local_breakpoint_enabled(dr7, index);
+}
+
+static inline int hw_breakpoint_type(unsigned long dr7, int index)
+{
+    return (dr7 >> (DR7_TYPE_SHIFT + (index * 4))) & 3;
+}
+
+static inline int hw_breakpoint_len(unsigned long dr7, int index)
+{
+    int len = ((dr7 >> (DR7_LEN_SHIFT + (index * 4))) & 3);
+    return (len == 2) ? 8 : len + 1;
+}
+
+static int hw_breakpoint_insert(CPUX86State *env, int index)
+{
+    CPUState *cs = env_cpu(env);
+    target_ulong dr7 = env->dr[7];
+    target_ulong drN = env->dr[index];
+    int err = 0;
+
+    switch (hw_breakpoint_type(dr7, index)) {
+    case DR7_TYPE_BP_INST:
+        if (hw_breakpoint_enabled(dr7, index)) {
+            err = cpu_breakpoint_insert(cs, drN, BP_CPU,
+                                        &env->cpu_breakpoint[index]);
+        }
+        break;
+
+    case DR7_TYPE_IO_RW:
+        /* Notice when we should enable calls to bpt_io.  */
+        return hw_breakpoint_enabled(env->dr[7], index)
+               ? HF_IOBPT_MASK : 0;
+
+    case DR7_TYPE_DATA_WR:
+        if (hw_breakpoint_enabled(dr7, index)) {
+            err = cpu_watchpoint_insert(cs, drN,
+                                        hw_breakpoint_len(dr7, index),
+                                        BP_CPU | BP_MEM_WRITE,
+                                        &env->cpu_watchpoint[index]);
+        }
+        break;
+
+    case DR7_TYPE_DATA_RW:
+        if (hw_breakpoint_enabled(dr7, index)) {
+            err = cpu_watchpoint_insert(cs, drN,
+                                        hw_breakpoint_len(dr7, index),
+                                        BP_CPU | BP_MEM_ACCESS,
+                                        &env->cpu_watchpoint[index]);
+        }
+        break;
+    }
+    if (err) {
+        env->cpu_breakpoint[index] = NULL;
+    }
+    return 0;
+}
+
+static void hw_breakpoint_remove(CPUX86State *env, int index)
+{
+    CPUState *cs = env_cpu(env);
+
+    switch (hw_breakpoint_type(env->dr[7], index)) {
+    case DR7_TYPE_BP_INST:
+        if (env->cpu_breakpoint[index]) {
+            cpu_breakpoint_remove_by_ref(cs, env->cpu_breakpoint[index]);
+            env->cpu_breakpoint[index] = NULL;
+        }
+        break;
+
+    case DR7_TYPE_DATA_WR:
+    case DR7_TYPE_DATA_RW:
+        if (env->cpu_breakpoint[index]) {
+            cpu_watchpoint_remove_by_ref(cs, env->cpu_watchpoint[index]);
+            env->cpu_breakpoint[index] = NULL;
+        }
+        break;
+
+    case DR7_TYPE_IO_RW:
+        /* HF_IOBPT_MASK cleared elsewhere.  */
+        break;
+    }
+}
+
+void cpu_x86_update_dr7(CPUX86State *env, uint32_t new_dr7)
+{
+    target_ulong old_dr7 = env->dr[7];
+    int iobpt = 0;
+    int i;
+
+    new_dr7 |= DR7_FIXED_1;
+
+    /* If nothing is changing except the global/local enable bits,
+       then we can make the change more efficient.  */
+    if (((old_dr7 ^ new_dr7) & ~0xff) == 0) {
+        /* Fold the global and local enable bits together into the
+           global fields, then xor to show which registers have
+           changed collective enable state.  */
+        int mod = ((old_dr7 | old_dr7 * 2) ^ (new_dr7 | new_dr7 * 2)) & 0xff;
+
+        for (i = 0; i < DR7_MAX_BP; i++) {
+            if ((mod & (2 << i * 2)) && !hw_breakpoint_enabled(new_dr7, i)) {
+                hw_breakpoint_remove(env, i);
+            }
+        }
+        env->dr[7] = new_dr7;
+        for (i = 0; i < DR7_MAX_BP; i++) {
+            if (mod & (2 << i * 2) && hw_breakpoint_enabled(new_dr7, i)) {
+                iobpt |= hw_breakpoint_insert(env, i);
+            } else if (hw_breakpoint_type(new_dr7, i) == DR7_TYPE_IO_RW
+                       && hw_breakpoint_enabled(new_dr7, i)) {
+                iobpt |= HF_IOBPT_MASK;
+            }
+        }
+    } else {
+        for (i = 0; i < DR7_MAX_BP; i++) {
+            hw_breakpoint_remove(env, i);
+        }
+        env->dr[7] = new_dr7;
+        for (i = 0; i < DR7_MAX_BP; i++) {
+            iobpt |= hw_breakpoint_insert(env, i);
+        }
+    }
+
+    env->hflags = (env->hflags & ~HF_IOBPT_MASK) | iobpt;
+}
+
+bool check_hw_breakpoints(CPUX86State *env, bool force_dr6_update)
+{
+    target_ulong dr6;
+    int reg;
+    bool hit_enabled = false;
+
+    dr6 = env->dr[6] & ~0xf;
+    for (reg = 0; reg < DR7_MAX_BP; reg++) {
+        bool bp_match = false;
+        bool wp_match = false;
+
+        switch (hw_breakpoint_type(env->dr[7], reg)) {
+        case DR7_TYPE_BP_INST:
+            if (env->dr[reg] == env->eip) {
+                bp_match = true;
+            }
+            break;
+        case DR7_TYPE_DATA_WR:
+        case DR7_TYPE_DATA_RW:
+            if (env->cpu_watchpoint[reg] &&
+                env->cpu_watchpoint[reg]->flags & BP_WATCHPOINT_HIT) {
+                wp_match = true;
+            }
+            break;
+        case DR7_TYPE_IO_RW:
+            break;
+        }
+        if (bp_match || wp_match) {
+            dr6 |= 1 << reg;
+            if (hw_breakpoint_enabled(env->dr[7], reg)) {
+                hit_enabled = true;
+            }
+        }
+    }
+
+    if (hit_enabled || force_dr6_update) {
+        env->dr[6] = dr6;
+    }
+
+    return hit_enabled;
+}
+
+void breakpoint_handler(CPUState *cs)
+{
+    X86CPU *cpu = X86_CPU(cs);
+    CPUX86State *env = &cpu->env;
+    CPUBreakpoint *bp;
+
+    if (cs->watchpoint_hit) {
+        if (cs->watchpoint_hit->flags & BP_CPU) {
+            cs->watchpoint_hit = NULL;
+            if (check_hw_breakpoints(env, false)) {
+                raise_exception(env, EXCP01_DB);
+            } else {
+                cpu_loop_exit_noexc(cs);
+            }
+        }
+    } else {
+        QTAILQ_FOREACH(bp, &cs->breakpoints, entry) {
+            if (bp->pc == env->eip) {
+                if (bp->flags & BP_CPU) {
+                    check_hw_breakpoints(env, true);
+                    raise_exception(env, EXCP01_DB);
+                }
+                break;
+            }
+        }
+    }
+}
+
+void helper_set_dr(CPUX86State *env, int reg, target_ulong t0)
+{
+    switch (reg) {
+    case 0: case 1: case 2: case 3:
+        if (hw_breakpoint_enabled(env->dr[7], reg)
+            && hw_breakpoint_type(env->dr[7], reg) != DR7_TYPE_IO_RW) {
+            hw_breakpoint_remove(env, reg);
+            env->dr[reg] = t0;
+            hw_breakpoint_insert(env, reg);
+        } else {
+            env->dr[reg] = t0;
+        }
+        return;
+    case 4:
+        if (env->cr[4] & CR4_DE_MASK) {
+            break;
+        }
+        /* fallthru */
+    case 6:
+        env->dr[6] = t0 | DR6_FIXED_1;
+        return;
+    case 5:
+        if (env->cr[4] & CR4_DE_MASK) {
+            break;
+        }
+        /* fallthru */
+    case 7:
+        cpu_x86_update_dr7(env, t0);
+        return;
+    }
+    raise_exception_err_ra(env, EXCP06_ILLOP, 0, GETPC());
+}
+
+/* Check if Port I/O is trapped by a breakpoint.  */
+void helper_bpt_io(CPUX86State *env, uint32_t port,
+                   uint32_t size, target_ulong next_eip)
+{
+    target_ulong dr7 = env->dr[7];
+    int i, hit = 0;
+
+    for (i = 0; i < DR7_MAX_BP; ++i) {
+        if (hw_breakpoint_type(dr7, i) == DR7_TYPE_IO_RW
+            && hw_breakpoint_enabled(dr7, i)) {
+            int bpt_len = hw_breakpoint_len(dr7, i);
+            if (port + size - 1 >= env->dr[i]
+                && port <= env->dr[i] + bpt_len - 1) {
+                hit |= 1 << i;
+            }
+        }
+    }
+
+    if (hit) {
+        env->dr[6] = (env->dr[6] & ~0xf) | hit;
+        env->eip = next_eip;
+        raise_exception(env, EXCP01_DB);
+    }
+}
diff --git a/target/i386/tcg/translate.c b/target/i386/tcg/translate.c
index b882041ef0..6ecbbfa6c1 100644
--- a/target/i386/tcg/translate.c
+++ b/target/i386/tcg/translate.c
@@ -1117,16 +1117,20 @@ static inline void gen_cmps(DisasContext *s, MemOp ot)
 static void gen_bpt_io(DisasContext *s, TCGv_i32 t_port, int ot)
 {
     if (s->flags & HF_IOBPT_MASK) {
+#ifdef CONFIG_USER_ONLY
+        /* user-mode cpu should not be in IOBPT mode */
+        g_assert_not_reached();
+#else
         TCGv_i32 t_size = tcg_const_i32(1 << ot);
         TCGv t_next = tcg_const_tl(s->pc - s->cs_base);
 
         gen_helper_bpt_io(cpu_env, t_port, t_size, t_next);
         tcg_temp_free_i32(t_size);
         tcg_temp_free(t_next);
+#endif /* CONFIG_USER_ONLY */
     }
 }
 
-
 static inline void gen_ins(DisasContext *s, MemOp ot)
 {
     gen_string_movl_A0_EDI(s);
@@ -8055,6 +8059,7 @@ static target_ulong disas_insn(DisasContext *s, CPUState *cpu)
         if (s->cpl != 0) {
             gen_exception(s, EXCP0D_GPF, pc_start - s->cs_base);
         } else {
+#ifndef CONFIG_USER_ONLY
             modrm = x86_ldub_code(env, s);
             /* Ignore the mod bits (assume (modrm&0xc0)==0xc0).
              * AMD documentation (24594.pdf) and testing of
@@ -8083,6 +8088,7 @@ static target_ulong disas_insn(DisasContext *s, CPUState *cpu)
                 gen_helper_get_dr(s->T0, cpu_env, s->tmp2_i32);
                 gen_op_mov_reg_v(s, ot, rm, s->T0);
             }
+#endif /* !CONFIG_USER_ONLY */
         }
         break;
     case 0x106: /* clts */
diff --git a/target/i386/tcg/sysemu/meson.build b/target/i386/tcg/sysemu/meson.build
index 6d0a0a0fee..1580950141 100644
--- a/target/i386/tcg/sysemu/meson.build
+++ b/target/i386/tcg/sysemu/meson.build
@@ -2,4 +2,5 @@ i386_softmmu_ss.add(when: ['CONFIG_TCG', 'CONFIG_SOFTMMU'], if_true: files(
   'tcg-cpu.c',
   'smm_helper.c',
   'excp_helper.c',
+  'bpt_helper.c',
 ))
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 13/20] i386: split misc helper user stubs and sysemu part
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (11 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 12/20] i386: move TCG bpt_helper into sysemu/ Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 14/20] i386: separate fpu_helper sysemu-only parts Claudio Fontana
                   ` (8 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/tcg/misc_helper.c        | 463 ---------------------------
 target/i386/tcg/sysemu/misc_helper.c | 438 +++++++++++++++++++++++++
 target/i386/tcg/user/misc_stubs.c    |  75 +++++
 target/i386/tcg/sysemu/meson.build   |   1 +
 target/i386/tcg/user/meson.build     |   1 +
 5 files changed, 515 insertions(+), 463 deletions(-)
 create mode 100644 target/i386/tcg/sysemu/misc_helper.c
 create mode 100644 target/i386/tcg/user/misc_stubs.c

diff --git a/target/i386/tcg/misc_helper.c b/target/i386/tcg/misc_helper.c
index f02e4fd400..82fb7037ac 100644
--- a/target/i386/tcg/misc_helper.c
+++ b/target/i386/tcg/misc_helper.c
@@ -18,12 +18,9 @@
  */
 
 #include "qemu/osdep.h"
-#include "qemu/main-loop.h"
 #include "cpu.h"
 #include "exec/helper-proto.h"
 #include "exec/exec-all.h"
-#include "exec/cpu_ldst.h"
-#include "exec/address-spaces.h"
 #include "helper-tcg.h"
 
 /*
@@ -39,69 +36,6 @@ void cpu_load_eflags(CPUX86State *env, int eflags, int update_mask)
         (eflags & update_mask) | 0x2;
 }
 
-void helper_outb(CPUX86State *env, uint32_t port, uint32_t data)
-{
-#ifdef CONFIG_USER_ONLY
-    fprintf(stderr, "outb: port=0x%04x, data=%02x\n", port, data);
-#else
-    address_space_stb(&address_space_io, port, data,
-                      cpu_get_mem_attrs(env), NULL);
-#endif
-}
-
-target_ulong helper_inb(CPUX86State *env, uint32_t port)
-{
-#ifdef CONFIG_USER_ONLY
-    fprintf(stderr, "inb: port=0x%04x\n", port);
-    return 0;
-#else
-    return address_space_ldub(&address_space_io, port,
-                              cpu_get_mem_attrs(env), NULL);
-#endif
-}
-
-void helper_outw(CPUX86State *env, uint32_t port, uint32_t data)
-{
-#ifdef CONFIG_USER_ONLY
-    fprintf(stderr, "outw: port=0x%04x, data=%04x\n", port, data);
-#else
-    address_space_stw(&address_space_io, port, data,
-                      cpu_get_mem_attrs(env), NULL);
-#endif
-}
-
-target_ulong helper_inw(CPUX86State *env, uint32_t port)
-{
-#ifdef CONFIG_USER_ONLY
-    fprintf(stderr, "inw: port=0x%04x\n", port);
-    return 0;
-#else
-    return address_space_lduw(&address_space_io, port,
-                              cpu_get_mem_attrs(env), NULL);
-#endif
-}
-
-void helper_outl(CPUX86State *env, uint32_t port, uint32_t data)
-{
-#ifdef CONFIG_USER_ONLY
-    fprintf(stderr, "outl: port=0x%04x, data=%08x\n", port, data);
-#else
-    address_space_stl(&address_space_io, port, data,
-                      cpu_get_mem_attrs(env), NULL);
-#endif
-}
-
-target_ulong helper_inl(CPUX86State *env, uint32_t port)
-{
-#ifdef CONFIG_USER_ONLY
-    fprintf(stderr, "inl: port=0x%04x\n", port);
-    return 0;
-#else
-    return address_space_ldl(&address_space_io, port,
-                             cpu_get_mem_attrs(env), NULL);
-#endif
-}
-
 void helper_into(CPUX86State *env, int next_eip_addend)
 {
     int eflags;
@@ -126,64 +60,6 @@ void helper_cpuid(CPUX86State *env)
     env->regs[R_EDX] = edx;
 }
 
-#if defined(CONFIG_USER_ONLY)
-target_ulong helper_read_crN(CPUX86State *env, int reg)
-{
-    return 0;
-}
-
-void helper_write_crN(CPUX86State *env, int reg, target_ulong t0)
-{
-}
-#else
-target_ulong helper_read_crN(CPUX86State *env, int reg)
-{
-    target_ulong val;
-
-    cpu_svm_check_intercept_param(env, SVM_EXIT_READ_CR0 + reg, 0, GETPC());
-    switch (reg) {
-    default:
-        val = env->cr[reg];
-        break;
-    case 8:
-        if (!(env->hflags2 & HF2_VINTR_MASK)) {
-            val = cpu_get_apic_tpr(env_archcpu(env)->apic_state);
-        } else {
-            val = env->v_tpr;
-        }
-        break;
-    }
-    return val;
-}
-
-void helper_write_crN(CPUX86State *env, int reg, target_ulong t0)
-{
-    cpu_svm_check_intercept_param(env, SVM_EXIT_WRITE_CR0 + reg, 0, GETPC());
-    switch (reg) {
-    case 0:
-        cpu_x86_update_cr0(env, t0);
-        break;
-    case 3:
-        cpu_x86_update_cr3(env, t0);
-        break;
-    case 4:
-        cpu_x86_update_cr4(env, t0);
-        break;
-    case 8:
-        if (!(env->hflags2 & HF2_VINTR_MASK)) {
-            qemu_mutex_lock_iothread();
-            cpu_set_apic_tpr(env_archcpu(env)->apic_state, t0);
-            qemu_mutex_unlock_iothread();
-        }
-        env->v_tpr = t0 & 0x0f;
-        break;
-    default:
-        env->cr[reg] = t0;
-        break;
-    }
-}
-#endif
-
 void helper_lmsw(CPUX86State *env, target_ulong t0)
 {
     /* only 4 lower bits of CR0 are modified. PE cannot be set to zero
@@ -232,345 +108,6 @@ void helper_rdpmc(CPUX86State *env)
     raise_exception_err(env, EXCP06_ILLOP, 0);
 }
 
-#if defined(CONFIG_USER_ONLY)
-void helper_wrmsr(CPUX86State *env)
-{
-}
-
-void helper_rdmsr(CPUX86State *env)
-{
-}
-#else
-void helper_wrmsr(CPUX86State *env)
-{
-    uint64_t val;
-    CPUState *cs = env_cpu(env);
-
-    cpu_svm_check_intercept_param(env, SVM_EXIT_MSR, 1, GETPC());
-
-    val = ((uint32_t)env->regs[R_EAX]) |
-        ((uint64_t)((uint32_t)env->regs[R_EDX]) << 32);
-
-    switch ((uint32_t)env->regs[R_ECX]) {
-    case MSR_IA32_SYSENTER_CS:
-        env->sysenter_cs = val & 0xffff;
-        break;
-    case MSR_IA32_SYSENTER_ESP:
-        env->sysenter_esp = val;
-        break;
-    case MSR_IA32_SYSENTER_EIP:
-        env->sysenter_eip = val;
-        break;
-    case MSR_IA32_APICBASE:
-        cpu_set_apic_base(env_archcpu(env)->apic_state, val);
-        break;
-    case MSR_EFER:
-        {
-            uint64_t update_mask;
-
-            update_mask = 0;
-            if (env->features[FEAT_8000_0001_EDX] & CPUID_EXT2_SYSCALL) {
-                update_mask |= MSR_EFER_SCE;
-            }
-            if (env->features[FEAT_8000_0001_EDX] & CPUID_EXT2_LM) {
-                update_mask |= MSR_EFER_LME;
-            }
-            if (env->features[FEAT_8000_0001_EDX] & CPUID_EXT2_FFXSR) {
-                update_mask |= MSR_EFER_FFXSR;
-            }
-            if (env->features[FEAT_8000_0001_EDX] & CPUID_EXT2_NX) {
-                update_mask |= MSR_EFER_NXE;
-            }
-            if (env->features[FEAT_8000_0001_ECX] & CPUID_EXT3_SVM) {
-                update_mask |= MSR_EFER_SVME;
-            }
-            if (env->features[FEAT_8000_0001_EDX] & CPUID_EXT2_FFXSR) {
-                update_mask |= MSR_EFER_FFXSR;
-            }
-            cpu_load_efer(env, (env->efer & ~update_mask) |
-                          (val & update_mask));
-        }
-        break;
-    case MSR_STAR:
-        env->star = val;
-        break;
-    case MSR_PAT:
-        env->pat = val;
-        break;
-    case MSR_IA32_PKRS:
-        if (val & 0xFFFFFFFF00000000ull) {
-            goto error;
-        }
-        env->pkrs = val;
-        tlb_flush(cs);
-        break;
-    case MSR_VM_HSAVE_PA:
-        env->vm_hsave = val;
-        break;
-#ifdef TARGET_X86_64
-    case MSR_LSTAR:
-        env->lstar = val;
-        break;
-    case MSR_CSTAR:
-        env->cstar = val;
-        break;
-    case MSR_FMASK:
-        env->fmask = val;
-        break;
-    case MSR_FSBASE:
-        env->segs[R_FS].base = val;
-        break;
-    case MSR_GSBASE:
-        env->segs[R_GS].base = val;
-        break;
-    case MSR_KERNELGSBASE:
-        env->kernelgsbase = val;
-        break;
-#endif
-    case MSR_MTRRphysBase(0):
-    case MSR_MTRRphysBase(1):
-    case MSR_MTRRphysBase(2):
-    case MSR_MTRRphysBase(3):
-    case MSR_MTRRphysBase(4):
-    case MSR_MTRRphysBase(5):
-    case MSR_MTRRphysBase(6):
-    case MSR_MTRRphysBase(7):
-        env->mtrr_var[((uint32_t)env->regs[R_ECX] -
-                       MSR_MTRRphysBase(0)) / 2].base = val;
-        break;
-    case MSR_MTRRphysMask(0):
-    case MSR_MTRRphysMask(1):
-    case MSR_MTRRphysMask(2):
-    case MSR_MTRRphysMask(3):
-    case MSR_MTRRphysMask(4):
-    case MSR_MTRRphysMask(5):
-    case MSR_MTRRphysMask(6):
-    case MSR_MTRRphysMask(7):
-        env->mtrr_var[((uint32_t)env->regs[R_ECX] -
-                       MSR_MTRRphysMask(0)) / 2].mask = val;
-        break;
-    case MSR_MTRRfix64K_00000:
-        env->mtrr_fixed[(uint32_t)env->regs[R_ECX] -
-                        MSR_MTRRfix64K_00000] = val;
-        break;
-    case MSR_MTRRfix16K_80000:
-    case MSR_MTRRfix16K_A0000:
-        env->mtrr_fixed[(uint32_t)env->regs[R_ECX] -
-                        MSR_MTRRfix16K_80000 + 1] = val;
-        break;
-    case MSR_MTRRfix4K_C0000:
-    case MSR_MTRRfix4K_C8000:
-    case MSR_MTRRfix4K_D0000:
-    case MSR_MTRRfix4K_D8000:
-    case MSR_MTRRfix4K_E0000:
-    case MSR_MTRRfix4K_E8000:
-    case MSR_MTRRfix4K_F0000:
-    case MSR_MTRRfix4K_F8000:
-        env->mtrr_fixed[(uint32_t)env->regs[R_ECX] -
-                        MSR_MTRRfix4K_C0000 + 3] = val;
-        break;
-    case MSR_MTRRdefType:
-        env->mtrr_deftype = val;
-        break;
-    case MSR_MCG_STATUS:
-        env->mcg_status = val;
-        break;
-    case MSR_MCG_CTL:
-        if ((env->mcg_cap & MCG_CTL_P)
-            && (val == 0 || val == ~(uint64_t)0)) {
-            env->mcg_ctl = val;
-        }
-        break;
-    case MSR_TSC_AUX:
-        env->tsc_aux = val;
-        break;
-    case MSR_IA32_MISC_ENABLE:
-        env->msr_ia32_misc_enable = val;
-        break;
-    case MSR_IA32_BNDCFGS:
-        /* FIXME: #GP if reserved bits are set.  */
-        /* FIXME: Extend highest implemented bit of linear address.  */
-        env->msr_bndcfgs = val;
-        cpu_sync_bndcs_hflags(env);
-        break;
-    default:
-        if ((uint32_t)env->regs[R_ECX] >= MSR_MC0_CTL
-            && (uint32_t)env->regs[R_ECX] < MSR_MC0_CTL +
-            (4 * env->mcg_cap & 0xff)) {
-            uint32_t offset = (uint32_t)env->regs[R_ECX] - MSR_MC0_CTL;
-            if ((offset & 0x3) != 0
-                || (val == 0 || val == ~(uint64_t)0)) {
-                env->mce_banks[offset] = val;
-            }
-            break;
-        }
-        /* XXX: exception? */
-        break;
-    }
-    return;
-error:
-    raise_exception_err_ra(env, EXCP0D_GPF, 0, GETPC());
-}
-
-void helper_rdmsr(CPUX86State *env)
-{
-    X86CPU *x86_cpu = env_archcpu(env);
-    uint64_t val;
-
-    cpu_svm_check_intercept_param(env, SVM_EXIT_MSR, 0, GETPC());
-
-    switch ((uint32_t)env->regs[R_ECX]) {
-    case MSR_IA32_SYSENTER_CS:
-        val = env->sysenter_cs;
-        break;
-    case MSR_IA32_SYSENTER_ESP:
-        val = env->sysenter_esp;
-        break;
-    case MSR_IA32_SYSENTER_EIP:
-        val = env->sysenter_eip;
-        break;
-    case MSR_IA32_APICBASE:
-        val = cpu_get_apic_base(env_archcpu(env)->apic_state);
-        break;
-    case MSR_EFER:
-        val = env->efer;
-        break;
-    case MSR_STAR:
-        val = env->star;
-        break;
-    case MSR_PAT:
-        val = env->pat;
-        break;
-    case MSR_IA32_PKRS:
-        val = env->pkrs;
-        break;
-    case MSR_VM_HSAVE_PA:
-        val = env->vm_hsave;
-        break;
-    case MSR_IA32_PERF_STATUS:
-        /* tsc_increment_by_tick */
-        val = 1000ULL;
-        /* CPU multiplier */
-        val |= (((uint64_t)4ULL) << 40);
-        break;
-#ifdef TARGET_X86_64
-    case MSR_LSTAR:
-        val = env->lstar;
-        break;
-    case MSR_CSTAR:
-        val = env->cstar;
-        break;
-    case MSR_FMASK:
-        val = env->fmask;
-        break;
-    case MSR_FSBASE:
-        val = env->segs[R_FS].base;
-        break;
-    case MSR_GSBASE:
-        val = env->segs[R_GS].base;
-        break;
-    case MSR_KERNELGSBASE:
-        val = env->kernelgsbase;
-        break;
-    case MSR_TSC_AUX:
-        val = env->tsc_aux;
-        break;
-#endif
-    case MSR_SMI_COUNT:
-        val = env->msr_smi_count;
-        break;
-    case MSR_MTRRphysBase(0):
-    case MSR_MTRRphysBase(1):
-    case MSR_MTRRphysBase(2):
-    case MSR_MTRRphysBase(3):
-    case MSR_MTRRphysBase(4):
-    case MSR_MTRRphysBase(5):
-    case MSR_MTRRphysBase(6):
-    case MSR_MTRRphysBase(7):
-        val = env->mtrr_var[((uint32_t)env->regs[R_ECX] -
-                             MSR_MTRRphysBase(0)) / 2].base;
-        break;
-    case MSR_MTRRphysMask(0):
-    case MSR_MTRRphysMask(1):
-    case MSR_MTRRphysMask(2):
-    case MSR_MTRRphysMask(3):
-    case MSR_MTRRphysMask(4):
-    case MSR_MTRRphysMask(5):
-    case MSR_MTRRphysMask(6):
-    case MSR_MTRRphysMask(7):
-        val = env->mtrr_var[((uint32_t)env->regs[R_ECX] -
-                             MSR_MTRRphysMask(0)) / 2].mask;
-        break;
-    case MSR_MTRRfix64K_00000:
-        val = env->mtrr_fixed[0];
-        break;
-    case MSR_MTRRfix16K_80000:
-    case MSR_MTRRfix16K_A0000:
-        val = env->mtrr_fixed[(uint32_t)env->regs[R_ECX] -
-                              MSR_MTRRfix16K_80000 + 1];
-        break;
-    case MSR_MTRRfix4K_C0000:
-    case MSR_MTRRfix4K_C8000:
-    case MSR_MTRRfix4K_D0000:
-    case MSR_MTRRfix4K_D8000:
-    case MSR_MTRRfix4K_E0000:
-    case MSR_MTRRfix4K_E8000:
-    case MSR_MTRRfix4K_F0000:
-    case MSR_MTRRfix4K_F8000:
-        val = env->mtrr_fixed[(uint32_t)env->regs[R_ECX] -
-                              MSR_MTRRfix4K_C0000 + 3];
-        break;
-    case MSR_MTRRdefType:
-        val = env->mtrr_deftype;
-        break;
-    case MSR_MTRRcap:
-        if (env->features[FEAT_1_EDX] & CPUID_MTRR) {
-            val = MSR_MTRRcap_VCNT | MSR_MTRRcap_FIXRANGE_SUPPORT |
-                MSR_MTRRcap_WC_SUPPORTED;
-        } else {
-            /* XXX: exception? */
-            val = 0;
-        }
-        break;
-    case MSR_MCG_CAP:
-        val = env->mcg_cap;
-        break;
-    case MSR_MCG_CTL:
-        if (env->mcg_cap & MCG_CTL_P) {
-            val = env->mcg_ctl;
-        } else {
-            val = 0;
-        }
-        break;
-    case MSR_MCG_STATUS:
-        val = env->mcg_status;
-        break;
-    case MSR_IA32_MISC_ENABLE:
-        val = env->msr_ia32_misc_enable;
-        break;
-    case MSR_IA32_BNDCFGS:
-        val = env->msr_bndcfgs;
-        break;
-     case MSR_IA32_UCODE_REV:
-        val = x86_cpu->ucode_rev;
-        break;
-    default:
-        if ((uint32_t)env->regs[R_ECX] >= MSR_MC0_CTL
-            && (uint32_t)env->regs[R_ECX] < MSR_MC0_CTL +
-            (4 * env->mcg_cap & 0xff)) {
-            uint32_t offset = (uint32_t)env->regs[R_ECX] - MSR_MC0_CTL;
-            val = env->mce_banks[offset];
-            break;
-        }
-        /* XXX: exception? */
-        val = 0;
-        break;
-    }
-    env->regs[R_EAX] = (uint32_t)(val);
-    env->regs[R_EDX] = (uint32_t)(val >> 32);
-}
-#endif
-
 static void do_pause(X86CPU *cpu)
 {
     CPUState *cs = CPU(cpu);
diff --git a/target/i386/tcg/sysemu/misc_helper.c b/target/i386/tcg/sysemu/misc_helper.c
new file mode 100644
index 0000000000..6a88840cf2
--- /dev/null
+++ b/target/i386/tcg/sysemu/misc_helper.c
@@ -0,0 +1,438 @@
+/*
+ *  x86 misc helpers - sysemu code
+ *
+ *  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 2.1 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 "qemu/main-loop.h"
+#include "cpu.h"
+#include "exec/helper-proto.h"
+#include "exec/cpu_ldst.h"
+#include "exec/address-spaces.h"
+#include "tcg/helper-tcg.h"
+
+void helper_outb(CPUX86State *env, uint32_t port, uint32_t data)
+{
+    address_space_stb(&address_space_io, port, data,
+                      cpu_get_mem_attrs(env), NULL);
+}
+
+target_ulong helper_inb(CPUX86State *env, uint32_t port)
+{
+    return address_space_ldub(&address_space_io, port,
+                              cpu_get_mem_attrs(env), NULL);
+}
+
+void helper_outw(CPUX86State *env, uint32_t port, uint32_t data)
+{
+    address_space_stw(&address_space_io, port, data,
+                      cpu_get_mem_attrs(env), NULL);
+}
+
+target_ulong helper_inw(CPUX86State *env, uint32_t port)
+{
+    return address_space_lduw(&address_space_io, port,
+                              cpu_get_mem_attrs(env), NULL);
+}
+
+void helper_outl(CPUX86State *env, uint32_t port, uint32_t data)
+{
+    address_space_stl(&address_space_io, port, data,
+                      cpu_get_mem_attrs(env), NULL);
+}
+
+target_ulong helper_inl(CPUX86State *env, uint32_t port)
+{
+    return address_space_ldl(&address_space_io, port,
+                             cpu_get_mem_attrs(env), NULL);
+}
+
+target_ulong helper_read_crN(CPUX86State *env, int reg)
+{
+    target_ulong val;
+
+    cpu_svm_check_intercept_param(env, SVM_EXIT_READ_CR0 + reg, 0, GETPC());
+    switch (reg) {
+    default:
+        val = env->cr[reg];
+        break;
+    case 8:
+        if (!(env->hflags2 & HF2_VINTR_MASK)) {
+            val = cpu_get_apic_tpr(env_archcpu(env)->apic_state);
+        } else {
+            val = env->v_tpr;
+        }
+        break;
+    }
+    return val;
+}
+
+void helper_write_crN(CPUX86State *env, int reg, target_ulong t0)
+{
+    cpu_svm_check_intercept_param(env, SVM_EXIT_WRITE_CR0 + reg, 0, GETPC());
+    switch (reg) {
+    case 0:
+        cpu_x86_update_cr0(env, t0);
+        break;
+    case 3:
+        cpu_x86_update_cr3(env, t0);
+        break;
+    case 4:
+        cpu_x86_update_cr4(env, t0);
+        break;
+    case 8:
+        if (!(env->hflags2 & HF2_VINTR_MASK)) {
+            qemu_mutex_lock_iothread();
+            cpu_set_apic_tpr(env_archcpu(env)->apic_state, t0);
+            qemu_mutex_unlock_iothread();
+        }
+        env->v_tpr = t0 & 0x0f;
+        break;
+    default:
+        env->cr[reg] = t0;
+        break;
+    }
+}
+
+void helper_wrmsr(CPUX86State *env)
+{
+    uint64_t val;
+    CPUState *cs = env_cpu(env);
+
+    cpu_svm_check_intercept_param(env, SVM_EXIT_MSR, 1, GETPC());
+
+    val = ((uint32_t)env->regs[R_EAX]) |
+        ((uint64_t)((uint32_t)env->regs[R_EDX]) << 32);
+
+    switch ((uint32_t)env->regs[R_ECX]) {
+    case MSR_IA32_SYSENTER_CS:
+        env->sysenter_cs = val & 0xffff;
+        break;
+    case MSR_IA32_SYSENTER_ESP:
+        env->sysenter_esp = val;
+        break;
+    case MSR_IA32_SYSENTER_EIP:
+        env->sysenter_eip = val;
+        break;
+    case MSR_IA32_APICBASE:
+        cpu_set_apic_base(env_archcpu(env)->apic_state, val);
+        break;
+    case MSR_EFER:
+        {
+            uint64_t update_mask;
+
+            update_mask = 0;
+            if (env->features[FEAT_8000_0001_EDX] & CPUID_EXT2_SYSCALL) {
+                update_mask |= MSR_EFER_SCE;
+            }
+            if (env->features[FEAT_8000_0001_EDX] & CPUID_EXT2_LM) {
+                update_mask |= MSR_EFER_LME;
+            }
+            if (env->features[FEAT_8000_0001_EDX] & CPUID_EXT2_FFXSR) {
+                update_mask |= MSR_EFER_FFXSR;
+            }
+            if (env->features[FEAT_8000_0001_EDX] & CPUID_EXT2_NX) {
+                update_mask |= MSR_EFER_NXE;
+            }
+            if (env->features[FEAT_8000_0001_ECX] & CPUID_EXT3_SVM) {
+                update_mask |= MSR_EFER_SVME;
+            }
+            if (env->features[FEAT_8000_0001_EDX] & CPUID_EXT2_FFXSR) {
+                update_mask |= MSR_EFER_FFXSR;
+            }
+            cpu_load_efer(env, (env->efer & ~update_mask) |
+                          (val & update_mask));
+        }
+        break;
+    case MSR_STAR:
+        env->star = val;
+        break;
+    case MSR_PAT:
+        env->pat = val;
+        break;
+    case MSR_IA32_PKRS:
+        if (val & 0xFFFFFFFF00000000ull) {
+            goto error;
+        }
+        env->pkrs = val;
+        tlb_flush(cs);
+        break;
+    case MSR_VM_HSAVE_PA:
+        env->vm_hsave = val;
+        break;
+#ifdef TARGET_X86_64
+    case MSR_LSTAR:
+        env->lstar = val;
+        break;
+    case MSR_CSTAR:
+        env->cstar = val;
+        break;
+    case MSR_FMASK:
+        env->fmask = val;
+        break;
+    case MSR_FSBASE:
+        env->segs[R_FS].base = val;
+        break;
+    case MSR_GSBASE:
+        env->segs[R_GS].base = val;
+        break;
+    case MSR_KERNELGSBASE:
+        env->kernelgsbase = val;
+        break;
+#endif
+    case MSR_MTRRphysBase(0):
+    case MSR_MTRRphysBase(1):
+    case MSR_MTRRphysBase(2):
+    case MSR_MTRRphysBase(3):
+    case MSR_MTRRphysBase(4):
+    case MSR_MTRRphysBase(5):
+    case MSR_MTRRphysBase(6):
+    case MSR_MTRRphysBase(7):
+        env->mtrr_var[((uint32_t)env->regs[R_ECX] -
+                       MSR_MTRRphysBase(0)) / 2].base = val;
+        break;
+    case MSR_MTRRphysMask(0):
+    case MSR_MTRRphysMask(1):
+    case MSR_MTRRphysMask(2):
+    case MSR_MTRRphysMask(3):
+    case MSR_MTRRphysMask(4):
+    case MSR_MTRRphysMask(5):
+    case MSR_MTRRphysMask(6):
+    case MSR_MTRRphysMask(7):
+        env->mtrr_var[((uint32_t)env->regs[R_ECX] -
+                       MSR_MTRRphysMask(0)) / 2].mask = val;
+        break;
+    case MSR_MTRRfix64K_00000:
+        env->mtrr_fixed[(uint32_t)env->regs[R_ECX] -
+                        MSR_MTRRfix64K_00000] = val;
+        break;
+    case MSR_MTRRfix16K_80000:
+    case MSR_MTRRfix16K_A0000:
+        env->mtrr_fixed[(uint32_t)env->regs[R_ECX] -
+                        MSR_MTRRfix16K_80000 + 1] = val;
+        break;
+    case MSR_MTRRfix4K_C0000:
+    case MSR_MTRRfix4K_C8000:
+    case MSR_MTRRfix4K_D0000:
+    case MSR_MTRRfix4K_D8000:
+    case MSR_MTRRfix4K_E0000:
+    case MSR_MTRRfix4K_E8000:
+    case MSR_MTRRfix4K_F0000:
+    case MSR_MTRRfix4K_F8000:
+        env->mtrr_fixed[(uint32_t)env->regs[R_ECX] -
+                        MSR_MTRRfix4K_C0000 + 3] = val;
+        break;
+    case MSR_MTRRdefType:
+        env->mtrr_deftype = val;
+        break;
+    case MSR_MCG_STATUS:
+        env->mcg_status = val;
+        break;
+    case MSR_MCG_CTL:
+        if ((env->mcg_cap & MCG_CTL_P)
+            && (val == 0 || val == ~(uint64_t)0)) {
+            env->mcg_ctl = val;
+        }
+        break;
+    case MSR_TSC_AUX:
+        env->tsc_aux = val;
+        break;
+    case MSR_IA32_MISC_ENABLE:
+        env->msr_ia32_misc_enable = val;
+        break;
+    case MSR_IA32_BNDCFGS:
+        /* FIXME: #GP if reserved bits are set.  */
+        /* FIXME: Extend highest implemented bit of linear address.  */
+        env->msr_bndcfgs = val;
+        cpu_sync_bndcs_hflags(env);
+        break;
+    default:
+        if ((uint32_t)env->regs[R_ECX] >= MSR_MC0_CTL
+            && (uint32_t)env->regs[R_ECX] < MSR_MC0_CTL +
+            (4 * env->mcg_cap & 0xff)) {
+            uint32_t offset = (uint32_t)env->regs[R_ECX] - MSR_MC0_CTL;
+            if ((offset & 0x3) != 0
+                || (val == 0 || val == ~(uint64_t)0)) {
+                env->mce_banks[offset] = val;
+            }
+            break;
+        }
+        /* XXX: exception? */
+        break;
+    }
+    return;
+error:
+    raise_exception_err_ra(env, EXCP0D_GPF, 0, GETPC());
+}
+
+void helper_rdmsr(CPUX86State *env)
+{
+    X86CPU *x86_cpu = env_archcpu(env);
+    uint64_t val;
+
+    cpu_svm_check_intercept_param(env, SVM_EXIT_MSR, 0, GETPC());
+
+    switch ((uint32_t)env->regs[R_ECX]) {
+    case MSR_IA32_SYSENTER_CS:
+        val = env->sysenter_cs;
+        break;
+    case MSR_IA32_SYSENTER_ESP:
+        val = env->sysenter_esp;
+        break;
+    case MSR_IA32_SYSENTER_EIP:
+        val = env->sysenter_eip;
+        break;
+    case MSR_IA32_APICBASE:
+        val = cpu_get_apic_base(env_archcpu(env)->apic_state);
+        break;
+    case MSR_EFER:
+        val = env->efer;
+        break;
+    case MSR_STAR:
+        val = env->star;
+        break;
+    case MSR_PAT:
+        val = env->pat;
+        break;
+    case MSR_IA32_PKRS:
+        val = env->pkrs;
+        break;
+    case MSR_VM_HSAVE_PA:
+        val = env->vm_hsave;
+        break;
+    case MSR_IA32_PERF_STATUS:
+        /* tsc_increment_by_tick */
+        val = 1000ULL;
+        /* CPU multiplier */
+        val |= (((uint64_t)4ULL) << 40);
+        break;
+#ifdef TARGET_X86_64
+    case MSR_LSTAR:
+        val = env->lstar;
+        break;
+    case MSR_CSTAR:
+        val = env->cstar;
+        break;
+    case MSR_FMASK:
+        val = env->fmask;
+        break;
+    case MSR_FSBASE:
+        val = env->segs[R_FS].base;
+        break;
+    case MSR_GSBASE:
+        val = env->segs[R_GS].base;
+        break;
+    case MSR_KERNELGSBASE:
+        val = env->kernelgsbase;
+        break;
+    case MSR_TSC_AUX:
+        val = env->tsc_aux;
+        break;
+#endif
+    case MSR_SMI_COUNT:
+        val = env->msr_smi_count;
+        break;
+    case MSR_MTRRphysBase(0):
+    case MSR_MTRRphysBase(1):
+    case MSR_MTRRphysBase(2):
+    case MSR_MTRRphysBase(3):
+    case MSR_MTRRphysBase(4):
+    case MSR_MTRRphysBase(5):
+    case MSR_MTRRphysBase(6):
+    case MSR_MTRRphysBase(7):
+        val = env->mtrr_var[((uint32_t)env->regs[R_ECX] -
+                             MSR_MTRRphysBase(0)) / 2].base;
+        break;
+    case MSR_MTRRphysMask(0):
+    case MSR_MTRRphysMask(1):
+    case MSR_MTRRphysMask(2):
+    case MSR_MTRRphysMask(3):
+    case MSR_MTRRphysMask(4):
+    case MSR_MTRRphysMask(5):
+    case MSR_MTRRphysMask(6):
+    case MSR_MTRRphysMask(7):
+        val = env->mtrr_var[((uint32_t)env->regs[R_ECX] -
+                             MSR_MTRRphysMask(0)) / 2].mask;
+        break;
+    case MSR_MTRRfix64K_00000:
+        val = env->mtrr_fixed[0];
+        break;
+    case MSR_MTRRfix16K_80000:
+    case MSR_MTRRfix16K_A0000:
+        val = env->mtrr_fixed[(uint32_t)env->regs[R_ECX] -
+                              MSR_MTRRfix16K_80000 + 1];
+        break;
+    case MSR_MTRRfix4K_C0000:
+    case MSR_MTRRfix4K_C8000:
+    case MSR_MTRRfix4K_D0000:
+    case MSR_MTRRfix4K_D8000:
+    case MSR_MTRRfix4K_E0000:
+    case MSR_MTRRfix4K_E8000:
+    case MSR_MTRRfix4K_F0000:
+    case MSR_MTRRfix4K_F8000:
+        val = env->mtrr_fixed[(uint32_t)env->regs[R_ECX] -
+                              MSR_MTRRfix4K_C0000 + 3];
+        break;
+    case MSR_MTRRdefType:
+        val = env->mtrr_deftype;
+        break;
+    case MSR_MTRRcap:
+        if (env->features[FEAT_1_EDX] & CPUID_MTRR) {
+            val = MSR_MTRRcap_VCNT | MSR_MTRRcap_FIXRANGE_SUPPORT |
+                MSR_MTRRcap_WC_SUPPORTED;
+        } else {
+            /* XXX: exception? */
+            val = 0;
+        }
+        break;
+    case MSR_MCG_CAP:
+        val = env->mcg_cap;
+        break;
+    case MSR_MCG_CTL:
+        if (env->mcg_cap & MCG_CTL_P) {
+            val = env->mcg_ctl;
+        } else {
+            val = 0;
+        }
+        break;
+    case MSR_MCG_STATUS:
+        val = env->mcg_status;
+        break;
+    case MSR_IA32_MISC_ENABLE:
+        val = env->msr_ia32_misc_enable;
+        break;
+    case MSR_IA32_BNDCFGS:
+        val = env->msr_bndcfgs;
+        break;
+     case MSR_IA32_UCODE_REV:
+        val = x86_cpu->ucode_rev;
+        break;
+    default:
+        if ((uint32_t)env->regs[R_ECX] >= MSR_MC0_CTL
+            && (uint32_t)env->regs[R_ECX] < MSR_MC0_CTL +
+            (4 * env->mcg_cap & 0xff)) {
+            uint32_t offset = (uint32_t)env->regs[R_ECX] - MSR_MC0_CTL;
+            val = env->mce_banks[offset];
+            break;
+        }
+        /* XXX: exception? */
+        val = 0;
+        break;
+    }
+    env->regs[R_EAX] = (uint32_t)(val);
+    env->regs[R_EDX] = (uint32_t)(val >> 32);
+}
diff --git a/target/i386/tcg/user/misc_stubs.c b/target/i386/tcg/user/misc_stubs.c
new file mode 100644
index 0000000000..84df4e65ff
--- /dev/null
+++ b/target/i386/tcg/user/misc_stubs.c
@@ -0,0 +1,75 @@
+/*
+ *  x86 misc 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 2.1 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 "exec/helper-proto.h"
+
+void helper_outb(CPUX86State *env, uint32_t port, uint32_t data)
+{
+    g_assert_not_reached();
+}
+
+target_ulong helper_inb(CPUX86State *env, uint32_t port)
+{
+    g_assert_not_reached();
+    return 0;
+}
+
+void helper_outw(CPUX86State *env, uint32_t port, uint32_t data)
+{
+    g_assert_not_reached();
+}
+
+target_ulong helper_inw(CPUX86State *env, uint32_t port)
+{
+    g_assert_not_reached();
+    return 0;
+}
+
+void helper_outl(CPUX86State *env, uint32_t port, uint32_t data)
+{
+    g_assert_not_reached();
+}
+
+target_ulong helper_inl(CPUX86State *env, uint32_t port)
+{
+    g_assert_not_reached();
+    return 0;
+}
+
+target_ulong helper_read_crN(CPUX86State *env, int reg)
+{
+    g_assert_not_reached();
+}
+
+void helper_write_crN(CPUX86State *env, int reg, target_ulong t0)
+{
+    g_assert_not_reached();
+}
+
+void helper_wrmsr(CPUX86State *env)
+{
+    g_assert_not_reached();
+}
+
+void helper_rdmsr(CPUX86State *env)
+{
+    g_assert_not_reached();
+}
diff --git a/target/i386/tcg/sysemu/meson.build b/target/i386/tcg/sysemu/meson.build
index 1580950141..b2aaab6eef 100644
--- a/target/i386/tcg/sysemu/meson.build
+++ b/target/i386/tcg/sysemu/meson.build
@@ -3,4 +3,5 @@ i386_softmmu_ss.add(when: ['CONFIG_TCG', 'CONFIG_SOFTMMU'], if_true: files(
   'smm_helper.c',
   'excp_helper.c',
   'bpt_helper.c',
+  'misc_helper.c',
 ))
diff --git a/target/i386/tcg/user/meson.build b/target/i386/tcg/user/meson.build
index e0ef0f02e2..2ab8bd903c 100644
--- a/target/i386/tcg/user/meson.build
+++ b/target/i386/tcg/user/meson.build
@@ -1,3 +1,4 @@
 i386_user_ss.add(when: ['CONFIG_TCG', 'CONFIG_USER_ONLY'], if_true: files(
   'excp_helper.c',
+  'misc_stubs.c',
 ))
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 14/20] i386: separate fpu_helper sysemu-only parts
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (12 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 13/20] i386: split misc helper user stubs and sysemu part Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 15/20] i386: split svm_helper into sysemu and stub-only user Claudio Fontana
                   ` (7 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

create a separate tcg/sysemu/fpu_helper.c for the sysemu-only parts.

For user mode, some small #ifdefs remain in tcg/fpu_helper.c
which do not seem worth splitting into their own user-mode module.

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/cpu.h                   |  3 ++
 target/i386/tcg/fpu_helper.c        | 41 +--------------------
 target/i386/tcg/sysemu/fpu_helper.c | 57 +++++++++++++++++++++++++++++
 target/i386/tcg/sysemu/meson.build  |  1 +
 4 files changed, 63 insertions(+), 39 deletions(-)
 create mode 100644 target/i386/tcg/sysemu/fpu_helper.c

diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index c8a84a9033..3797789dc2 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -1816,7 +1816,10 @@ int cpu_x86_support_mca_broadcast(CPUX86State *env);
 int cpu_get_pic_interrupt(CPUX86State *s);
 /* MSDOS compatibility mode FPU exception support */
 void x86_register_ferr_irq(qemu_irq irq);
+void fpu_check_raise_ferr_irq(CPUX86State *s);
 void cpu_set_ignne(void);
+void cpu_clear_ignne(void);
+
 /* mpx_helper.c */
 void cpu_sync_bndcs_hflags(CPUX86State *env);
 
diff --git a/target/i386/tcg/fpu_helper.c b/target/i386/tcg/fpu_helper.c
index 20e4d2e715..1b30f1bb73 100644
--- a/target/i386/tcg/fpu_helper.c
+++ b/target/i386/tcg/fpu_helper.c
@@ -21,17 +21,10 @@
 #include <math.h>
 #include "cpu.h"
 #include "exec/helper-proto.h"
-#include "qemu/host-utils.h"
-#include "exec/exec-all.h"
-#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"
-#endif
-
 /* float macros */
 #define FT0    (env->ft0)
 #define ST0    (env->fpregs[env->fpstt].d)
@@ -75,36 +68,6 @@
 #define floatx80_ln2_d make_floatx80(0x3ffe, 0xb17217f7d1cf79abLL)
 #define floatx80_pi_d make_floatx80(0x4000, 0xc90fdaa22168c234LL)
 
-#if !defined(CONFIG_USER_ONLY)
-static qemu_irq ferr_irq;
-
-void x86_register_ferr_irq(qemu_irq irq)
-{
-    ferr_irq = irq;
-}
-
-static void cpu_clear_ignne(void)
-{
-    CPUX86State *env = &X86_CPU(first_cpu)->env;
-    env->hflags2 &= ~HF2_IGNNE_MASK;
-}
-
-void cpu_set_ignne(void)
-{
-    CPUX86State *env = &X86_CPU(first_cpu)->env;
-    env->hflags2 |= HF2_IGNNE_MASK;
-    /*
-     * We get here in response to a write to port F0h.  The chipset should
-     * deassert FP_IRQ and FERR# instead should stay signaled until FPSW_SE is
-     * cleared, because FERR# and FP_IRQ are two separate pins on real
-     * hardware.  However, we don't model FERR# as a qemu_irq, so we just
-     * do directly what the chipset would do, i.e. deassert FP_IRQ.
-     */
-    qemu_irq_lower(ferr_irq);
-}
-#endif
-
-
 static inline void fpush(CPUX86State *env)
 {
     env->fpstt = (env->fpstt - 1) & 7;
@@ -202,8 +165,8 @@ static void fpu_raise_exception(CPUX86State *env, uintptr_t retaddr)
         raise_exception_ra(env, EXCP10_COPR, retaddr);
     }
 #if !defined(CONFIG_USER_ONLY)
-    else if (ferr_irq && !(env->hflags2 & HF2_IGNNE_MASK)) {
-        qemu_irq_raise(ferr_irq);
+    else {
+        fpu_check_raise_ferr_irq(env);
     }
 #endif
 }
diff --git a/target/i386/tcg/sysemu/fpu_helper.c b/target/i386/tcg/sysemu/fpu_helper.c
new file mode 100644
index 0000000000..1c3610da3b
--- /dev/null
+++ b/target/i386/tcg/sysemu/fpu_helper.c
@@ -0,0 +1,57 @@
+/*
+ *  x86 FPU, MMX/3DNow!/SSE/SSE2/SSE3/SSSE3/SSE4/PNI helpers (sysemu code)
+ *
+ *  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 2.1 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 "hw/irq.h"
+
+static qemu_irq ferr_irq;
+
+void x86_register_ferr_irq(qemu_irq irq)
+{
+    ferr_irq = irq;
+}
+
+void fpu_check_raise_ferr_irq(CPUX86State *env)
+{
+    if (ferr_irq && !(env->hflags2 & HF2_IGNNE_MASK)) {
+        qemu_irq_raise(ferr_irq);
+        return;
+    }
+}
+
+void cpu_clear_ignne(void)
+{
+    CPUX86State *env = &X86_CPU(first_cpu)->env;
+    env->hflags2 &= ~HF2_IGNNE_MASK;
+}
+
+void cpu_set_ignne(void)
+{
+    CPUX86State *env = &X86_CPU(first_cpu)->env;
+    env->hflags2 |= HF2_IGNNE_MASK;
+    /*
+     * We get here in response to a write to port F0h.  The chipset should
+     * deassert FP_IRQ and FERR# instead should stay signaled until FPSW_SE is
+     * cleared, because FERR# and FP_IRQ are two separate pins on real
+     * hardware.  However, we don't model FERR# as a qemu_irq, so we just
+     * do directly what the chipset would do, i.e. deassert FP_IRQ.
+     */
+    qemu_irq_lower(ferr_irq);
+}
diff --git a/target/i386/tcg/sysemu/meson.build b/target/i386/tcg/sysemu/meson.build
index b2aaab6eef..f84519a213 100644
--- a/target/i386/tcg/sysemu/meson.build
+++ b/target/i386/tcg/sysemu/meson.build
@@ -4,4 +4,5 @@ i386_softmmu_ss.add(when: ['CONFIG_TCG', 'CONFIG_SOFTMMU'], if_true: files(
   'excp_helper.c',
   'bpt_helper.c',
   'misc_helper.c',
+  'fpu_helper.c',
 ))
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 15/20] i386: split svm_helper into sysemu and stub-only user
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (13 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 14/20] i386: separate fpu_helper sysemu-only parts Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 16/20] i386: split seg_helper into user-only and sysemu parts Claudio Fontana
                   ` (6 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

For now we just copy over the previous user stubs, but really,

everything that requires s->cpl == 0 should be impossible
to trigger from user-mode emulation.

Later on we should add a check that asserts this easily f.e.:

static bool check_cpl0(DisasContext *s)
{
     int cpl = s->cpl;
 #ifdef CONFIG_USER_ONLY
     assert(cpl == 3);
 #endif
     if (cpl != 0) {
         gen_exception(s, EXCP0D_GPF, s->pc_start - s->cs_base);
         return false;
     }
     return true;
}

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/tcg/{ => sysemu}/svm_helper.c | 62 +-----------------
 target/i386/tcg/user/svm_stubs.c          | 76 +++++++++++++++++++++++
 target/i386/tcg/meson.build               |  1 -
 target/i386/tcg/sysemu/meson.build        |  1 +
 target/i386/tcg/user/meson.build          |  1 +
 5 files changed, 80 insertions(+), 61 deletions(-)
 rename target/i386/tcg/{ => sysemu}/svm_helper.c (96%)
 create mode 100644 target/i386/tcg/user/svm_stubs.c

diff --git a/target/i386/tcg/svm_helper.c b/target/i386/tcg/sysemu/svm_helper.c
similarity index 96%
rename from target/i386/tcg/svm_helper.c
rename to target/i386/tcg/sysemu/svm_helper.c
index 097bb9b83d..5b9c6f18be 100644
--- a/target/i386/tcg/svm_helper.c
+++ b/target/i386/tcg/sysemu/svm_helper.c
@@ -1,5 +1,5 @@
 /*
- *  x86 SVM helpers
+ *  x86 SVM helpers (sysemu only)
  *
  *  Copyright (c) 2003 Fabrice Bellard
  *
@@ -22,66 +22,10 @@
 #include "exec/helper-proto.h"
 #include "exec/exec-all.h"
 #include "exec/cpu_ldst.h"
-#include "helper-tcg.h"
+#include "tcg/helper-tcg.h"
 
 /* Secure Virtual Machine helpers */
 
-#if defined(CONFIG_USER_ONLY)
-
-void helper_vmrun(CPUX86State *env, int aflag, int next_eip_addend)
-{
-}
-
-void helper_vmmcall(CPUX86State *env)
-{
-}
-
-void helper_vmload(CPUX86State *env, int aflag)
-{
-}
-
-void helper_vmsave(CPUX86State *env, int aflag)
-{
-}
-
-void helper_stgi(CPUX86State *env)
-{
-}
-
-void helper_clgi(CPUX86State *env)
-{
-}
-
-void helper_skinit(CPUX86State *env)
-{
-}
-
-void helper_invlpga(CPUX86State *env, int aflag)
-{
-}
-
-void cpu_vmexit(CPUX86State *nenv, uint32_t exit_code, uint64_t exit_info_1,
-                uintptr_t retaddr)
-{
-    assert(0);
-}
-
-void helper_svm_check_intercept_param(CPUX86State *env, uint32_t type,
-                                      uint64_t param)
-{
-}
-
-void cpu_svm_check_intercept_param(CPUX86State *env, uint32_t type,
-                                   uint64_t param, uintptr_t retaddr)
-{
-}
-
-void helper_svm_check_io(CPUX86State *env, uint32_t port, uint32_t param,
-                         uint32_t next_eip_addend)
-{
-}
-#else
-
 static inline void svm_save_seg(CPUX86State *env, hwaddr addr,
                                 const SegmentCache *sc)
 {
@@ -797,5 +741,3 @@ void do_vmexit(CPUX86State *env, uint32_t exit_code, uint64_t exit_info_1)
        host's code segment or non-canonical (in the case of long mode), a
        #GP fault is delivered inside the host. */
 }
-
-#endif
diff --git a/target/i386/tcg/user/svm_stubs.c b/target/i386/tcg/user/svm_stubs.c
new file mode 100644
index 0000000000..97528b56ad
--- /dev/null
+++ b/target/i386/tcg/user/svm_stubs.c
@@ -0,0 +1,76 @@
+/*
+ *  x86 SVM helpers (user-mode)
+ *
+ *  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 2.1 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 "exec/helper-proto.h"
+#include "tcg/helper-tcg.h"
+
+void helper_vmrun(CPUX86State *env, int aflag, int next_eip_addend)
+{
+}
+
+void helper_vmmcall(CPUX86State *env)
+{
+}
+
+void helper_vmload(CPUX86State *env, int aflag)
+{
+}
+
+void helper_vmsave(CPUX86State *env, int aflag)
+{
+}
+
+void helper_stgi(CPUX86State *env)
+{
+}
+
+void helper_clgi(CPUX86State *env)
+{
+}
+
+void helper_skinit(CPUX86State *env)
+{
+}
+
+void helper_invlpga(CPUX86State *env, int aflag)
+{
+}
+
+void cpu_vmexit(CPUX86State *nenv, uint32_t exit_code, uint64_t exit_info_1,
+                uintptr_t retaddr)
+{
+    assert(0);
+}
+
+void helper_svm_check_intercept_param(CPUX86State *env, uint32_t type,
+                                      uint64_t param)
+{
+}
+
+void cpu_svm_check_intercept_param(CPUX86State *env, uint32_t type,
+                                   uint64_t param, uintptr_t retaddr)
+{
+}
+
+void helper_svm_check_io(CPUX86State *env, uint32_t port, uint32_t param,
+                         uint32_t next_eip_addend)
+{
+}
diff --git a/target/i386/tcg/meson.build b/target/i386/tcg/meson.build
index 449d9719ef..f9110e890c 100644
--- a/target/i386/tcg/meson.build
+++ b/target/i386/tcg/meson.build
@@ -8,7 +8,6 @@ i386_ss.add(when: 'CONFIG_TCG', if_true: files(
   'misc_helper.c',
   'mpx_helper.c',
   'seg_helper.c',
-  'svm_helper.c',
   'tcg-cpu.c',
   'translate.c'), if_false: files('tcg-stub.c'))
 
diff --git a/target/i386/tcg/sysemu/meson.build b/target/i386/tcg/sysemu/meson.build
index f84519a213..126528d0c9 100644
--- a/target/i386/tcg/sysemu/meson.build
+++ b/target/i386/tcg/sysemu/meson.build
@@ -5,4 +5,5 @@ i386_softmmu_ss.add(when: ['CONFIG_TCG', 'CONFIG_SOFTMMU'], if_true: files(
   'bpt_helper.c',
   'misc_helper.c',
   'fpu_helper.c',
+  'svm_helper.c',
 ))
diff --git a/target/i386/tcg/user/meson.build b/target/i386/tcg/user/meson.build
index 2ab8bd903c..3edaee7402 100644
--- a/target/i386/tcg/user/meson.build
+++ b/target/i386/tcg/user/meson.build
@@ -1,4 +1,5 @@
 i386_user_ss.add(when: ['CONFIG_TCG', 'CONFIG_USER_ONLY'], if_true: files(
   'excp_helper.c',
   'misc_stubs.c',
+  'svm_stubs.c',
 ))
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 16/20] i386: split seg_helper into user-only and sysemu parts
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (14 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 15/20] i386: split svm_helper into sysemu and stub-only user Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 17/20] i386: split off sysemu part of cpu.c Claudio Fontana
                   ` (5 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/tcg/helper-tcg.h        |   5 +
 target/i386/tcg/seg_helper.h        |  66 ++++++++
 target/i386/tcg/seg_helper.c        | 233 +---------------------------
 target/i386/tcg/sysemu/seg_helper.c | 125 +++++++++++++++
 target/i386/tcg/user/seg_helper.c   | 109 +++++++++++++
 target/i386/tcg/sysemu/meson.build  |   1 +
 target/i386/tcg/user/meson.build    |   1 +
 7 files changed, 311 insertions(+), 229 deletions(-)
 create mode 100644 target/i386/tcg/seg_helper.h
 create mode 100644 target/i386/tcg/sysemu/seg_helper.c
 create mode 100644 target/i386/tcg/user/seg_helper.c

diff --git a/target/i386/tcg/helper-tcg.h b/target/i386/tcg/helper-tcg.h
index b420b3356d..30eacdbbc9 100644
--- a/target/i386/tcg/helper-tcg.h
+++ b/target/i386/tcg/helper-tcg.h
@@ -88,6 +88,11 @@ 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);
+void do_interrupt_all(X86CPU *cpu, int intno, int is_int,
+                      int error_code, target_ulong next_eip, int is_hw);
+void handle_even_inj(CPUX86State *env, int intno, int is_int,
+                     int error_code, int is_hw, int rm);
+int exception_has_error_code(int intno);
 
 /* smm_helper.c */
 void do_smm_enter(X86CPU *cpu);
diff --git a/target/i386/tcg/seg_helper.h b/target/i386/tcg/seg_helper.h
new file mode 100644
index 0000000000..ebf1035277
--- /dev/null
+++ b/target/i386/tcg/seg_helper.h
@@ -0,0 +1,66 @@
+/*
+ *  x86 segmentation related helpers macros
+ *
+ *  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 2.1 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 SEG_HELPER_H
+#define SEG_HELPER_H
+
+//#define DEBUG_PCALL
+
+#ifdef DEBUG_PCALL
+# define LOG_PCALL(...) qemu_log_mask(CPU_LOG_PCALL, ## __VA_ARGS__)
+# define LOG_PCALL_STATE(cpu)                                  \
+    log_cpu_state_mask(CPU_LOG_PCALL, (cpu), CPU_DUMP_CCOP)
+#else
+# define LOG_PCALL(...) do { } while (0)
+# define LOG_PCALL_STATE(cpu) do { } while (0)
+#endif
+
+/*
+ * TODO: Convert callers to compute cpu_mmu_index_kernel once
+ * and use *_mmuidx_ra directly.
+ */
+#define cpu_ldub_kernel_ra(e, p, r) \
+    cpu_ldub_mmuidx_ra(e, p, cpu_mmu_index_kernel(e), r)
+#define cpu_lduw_kernel_ra(e, p, r) \
+    cpu_lduw_mmuidx_ra(e, p, cpu_mmu_index_kernel(e), r)
+#define cpu_ldl_kernel_ra(e, p, r) \
+    cpu_ldl_mmuidx_ra(e, p, cpu_mmu_index_kernel(e), r)
+#define cpu_ldq_kernel_ra(e, p, r) \
+    cpu_ldq_mmuidx_ra(e, p, cpu_mmu_index_kernel(e), r)
+
+#define cpu_stb_kernel_ra(e, p, v, r) \
+    cpu_stb_mmuidx_ra(e, p, v, cpu_mmu_index_kernel(e), r)
+#define cpu_stw_kernel_ra(e, p, v, r) \
+    cpu_stw_mmuidx_ra(e, p, v, cpu_mmu_index_kernel(e), r)
+#define cpu_stl_kernel_ra(e, p, v, r) \
+    cpu_stl_mmuidx_ra(e, p, v, cpu_mmu_index_kernel(e), r)
+#define cpu_stq_kernel_ra(e, p, v, r) \
+    cpu_stq_mmuidx_ra(e, p, v, cpu_mmu_index_kernel(e), r)
+
+#define cpu_ldub_kernel(e, p)    cpu_ldub_kernel_ra(e, p, 0)
+#define cpu_lduw_kernel(e, p)    cpu_lduw_kernel_ra(e, p, 0)
+#define cpu_ldl_kernel(e, p)     cpu_ldl_kernel_ra(e, p, 0)
+#define cpu_ldq_kernel(e, p)     cpu_ldq_kernel_ra(e, p, 0)
+
+#define cpu_stb_kernel(e, p, v)  cpu_stb_kernel_ra(e, p, v, 0)
+#define cpu_stw_kernel(e, p, v)  cpu_stw_kernel_ra(e, p, v, 0)
+#define cpu_stl_kernel(e, p, v)  cpu_stl_kernel_ra(e, p, v, 0)
+#define cpu_stq_kernel(e, p, v)  cpu_stq_kernel_ra(e, p, v, 0)
+
+#endif /* SEG_HELPER_H */
diff --git a/target/i386/tcg/seg_helper.c b/target/i386/tcg/seg_helper.c
index d04fbdd7cd..cf3f051524 100644
--- a/target/i386/tcg/seg_helper.c
+++ b/target/i386/tcg/seg_helper.c
@@ -26,49 +26,7 @@
 #include "exec/cpu_ldst.h"
 #include "exec/log.h"
 #include "helper-tcg.h"
-
-//#define DEBUG_PCALL
-
-#ifdef DEBUG_PCALL
-# define LOG_PCALL(...) qemu_log_mask(CPU_LOG_PCALL, ## __VA_ARGS__)
-# define LOG_PCALL_STATE(cpu)                                  \
-    log_cpu_state_mask(CPU_LOG_PCALL, (cpu), CPU_DUMP_CCOP)
-#else
-# define LOG_PCALL(...) do { } while (0)
-# define LOG_PCALL_STATE(cpu) do { } while (0)
-#endif
-
-/*
- * TODO: Convert callers to compute cpu_mmu_index_kernel once
- * and use *_mmuidx_ra directly.
- */
-#define cpu_ldub_kernel_ra(e, p, r) \
-    cpu_ldub_mmuidx_ra(e, p, cpu_mmu_index_kernel(e), r)
-#define cpu_lduw_kernel_ra(e, p, r) \
-    cpu_lduw_mmuidx_ra(e, p, cpu_mmu_index_kernel(e), r)
-#define cpu_ldl_kernel_ra(e, p, r) \
-    cpu_ldl_mmuidx_ra(e, p, cpu_mmu_index_kernel(e), r)
-#define cpu_ldq_kernel_ra(e, p, r) \
-    cpu_ldq_mmuidx_ra(e, p, cpu_mmu_index_kernel(e), r)
-
-#define cpu_stb_kernel_ra(e, p, v, r) \
-    cpu_stb_mmuidx_ra(e, p, v, cpu_mmu_index_kernel(e), r)
-#define cpu_stw_kernel_ra(e, p, v, r) \
-    cpu_stw_mmuidx_ra(e, p, v, cpu_mmu_index_kernel(e), r)
-#define cpu_stl_kernel_ra(e, p, v, r) \
-    cpu_stl_mmuidx_ra(e, p, v, cpu_mmu_index_kernel(e), r)
-#define cpu_stq_kernel_ra(e, p, v, r) \
-    cpu_stq_mmuidx_ra(e, p, v, cpu_mmu_index_kernel(e), r)
-
-#define cpu_ldub_kernel(e, p)    cpu_ldub_kernel_ra(e, p, 0)
-#define cpu_lduw_kernel(e, p)    cpu_lduw_kernel_ra(e, p, 0)
-#define cpu_ldl_kernel(e, p)     cpu_ldl_kernel_ra(e, p, 0)
-#define cpu_ldq_kernel(e, p)     cpu_ldq_kernel_ra(e, p, 0)
-
-#define cpu_stb_kernel(e, p, v)  cpu_stb_kernel_ra(e, p, v, 0)
-#define cpu_stw_kernel(e, p, v)  cpu_stw_kernel_ra(e, p, v, 0)
-#define cpu_stl_kernel(e, p, v)  cpu_stl_kernel_ra(e, p, v, 0)
-#define cpu_stq_kernel(e, p, v)  cpu_stq_kernel_ra(e, p, v, 0)
+#include "seg_helper.h"
 
 /* return non zero if error */
 static inline int load_segment_ra(CPUX86State *env, uint32_t *e1_ptr,
@@ -531,7 +489,7 @@ static inline unsigned int get_sp_mask(unsigned int e2)
     }
 }
 
-static int exception_has_error_code(int intno)
+int exception_has_error_code(int intno)
 {
     switch (intno) {
     case 8:
@@ -976,72 +934,6 @@ static void do_interrupt64(CPUX86State *env, int intno, int is_int,
 }
 #endif
 
-#ifdef TARGET_X86_64
-#if defined(CONFIG_USER_ONLY)
-void helper_syscall(CPUX86State *env, int next_eip_addend)
-{
-    CPUState *cs = env_cpu(env);
-
-    cs->exception_index = EXCP_SYSCALL;
-    env->exception_is_int = 0;
-    env->exception_next_eip = env->eip + next_eip_addend;
-    cpu_loop_exit(cs);
-}
-#else
-void helper_syscall(CPUX86State *env, int next_eip_addend)
-{
-    int selector;
-
-    if (!(env->efer & MSR_EFER_SCE)) {
-        raise_exception_err_ra(env, EXCP06_ILLOP, 0, GETPC());
-    }
-    selector = (env->star >> 32) & 0xffff;
-    if (env->hflags & HF_LMA_MASK) {
-        int code64;
-
-        env->regs[R_ECX] = env->eip + next_eip_addend;
-        env->regs[11] = cpu_compute_eflags(env) & ~RF_MASK;
-
-        code64 = env->hflags & HF_CS64_MASK;
-
-        env->eflags &= ~(env->fmask | RF_MASK);
-        cpu_load_eflags(env, env->eflags, 0);
-        cpu_x86_load_seg_cache(env, R_CS, selector & 0xfffc,
-                           0, 0xffffffff,
-                               DESC_G_MASK | DESC_P_MASK |
-                               DESC_S_MASK |
-                               DESC_CS_MASK | DESC_R_MASK | DESC_A_MASK |
-                               DESC_L_MASK);
-        cpu_x86_load_seg_cache(env, R_SS, (selector + 8) & 0xfffc,
-                               0, 0xffffffff,
-                               DESC_G_MASK | DESC_B_MASK | DESC_P_MASK |
-                               DESC_S_MASK |
-                               DESC_W_MASK | DESC_A_MASK);
-        if (code64) {
-            env->eip = env->lstar;
-        } else {
-            env->eip = env->cstar;
-        }
-    } else {
-        env->regs[R_ECX] = (uint32_t)(env->eip + next_eip_addend);
-
-        env->eflags &= ~(IF_MASK | RF_MASK | VM_MASK);
-        cpu_x86_load_seg_cache(env, R_CS, selector & 0xfffc,
-                           0, 0xffffffff,
-                               DESC_G_MASK | DESC_B_MASK | DESC_P_MASK |
-                               DESC_S_MASK |
-                               DESC_CS_MASK | DESC_R_MASK | DESC_A_MASK);
-        cpu_x86_load_seg_cache(env, R_SS, (selector + 8) & 0xfffc,
-                               0, 0xffffffff,
-                               DESC_G_MASK | DESC_B_MASK | DESC_P_MASK |
-                               DESC_S_MASK |
-                               DESC_W_MASK | DESC_A_MASK);
-        env->eip = (uint32_t)env->star;
-    }
-}
-#endif
-#endif
-
 #ifdef TARGET_X86_64
 void helper_sysret(CPUX86State *env, int dflag)
 {
@@ -1136,84 +1028,13 @@ static void do_interrupt_real(CPUX86State *env, int intno, int is_int,
     env->eflags &= ~(IF_MASK | TF_MASK | AC_MASK | RF_MASK);
 }
 
-#if defined(CONFIG_USER_ONLY)
-/* fake user mode interrupt. is_int is TRUE if coming from the int
- * instruction. next_eip is the env->eip value AFTER the interrupt
- * instruction. It is only relevant if is_int is TRUE or if intno
- * is EXCP_SYSCALL.
- */
-static void do_interrupt_user(CPUX86State *env, int intno, int is_int,
-                              int error_code, target_ulong next_eip)
-{
-    if (is_int) {
-        SegmentCache *dt;
-        target_ulong ptr;
-        int dpl, cpl, shift;
-        uint32_t e2;
-
-        dt = &env->idt;
-        if (env->hflags & HF_LMA_MASK) {
-            shift = 4;
-        } else {
-            shift = 3;
-        }
-        ptr = dt->base + (intno << shift);
-        e2 = cpu_ldl_kernel(env, ptr + 4);
-
-        dpl = (e2 >> DESC_DPL_SHIFT) & 3;
-        cpl = env->hflags & HF_CPL_MASK;
-        /* check privilege if software int */
-        if (dpl < cpl) {
-            raise_exception_err(env, EXCP0D_GPF, (intno << shift) + 2);
-        }
-    }
-
-    /* Since we emulate only user space, we cannot do more than
-       exiting the emulation with the suitable exception and error
-       code. So update EIP for INT 0x80 and EXCP_SYSCALL. */
-    if (is_int || intno == EXCP_SYSCALL) {
-        env->eip = next_eip;
-    }
-}
-
-#else
-
-static void handle_even_inj(CPUX86State *env, int intno, int is_int,
-                            int error_code, int is_hw, int rm)
-{
-    CPUState *cs = env_cpu(env);
-    uint32_t event_inj = x86_ldl_phys(cs, env->vm_vmcb + offsetof(struct vmcb,
-                                                          control.event_inj));
-
-    if (!(event_inj & SVM_EVTINJ_VALID)) {
-        int type;
-
-        if (is_int) {
-            type = SVM_EVTINJ_TYPE_SOFT;
-        } else {
-            type = SVM_EVTINJ_TYPE_EXEPT;
-        }
-        event_inj = intno | type | SVM_EVTINJ_VALID;
-        if (!rm && exception_has_error_code(intno)) {
-            event_inj |= SVM_EVTINJ_VALID_ERR;
-            x86_stl_phys(cs, env->vm_vmcb + offsetof(struct vmcb,
-                                             control.event_inj_err),
-                     error_code);
-        }
-        x86_stl_phys(cs,
-                 env->vm_vmcb + offsetof(struct vmcb, control.event_inj),
-                 event_inj);
-    }
-}
-#endif
-
 /*
  * Begin execution of an interruption. is_int is TRUE if coming from
  * the int instruction. next_eip is the env->eip value AFTER the interrupt
  * instruction. It is only relevant if is_int is TRUE.
  */
-static void do_interrupt_all(X86CPU *cpu, int intno, int is_int,
-                             int error_code, target_ulong next_eip, int is_hw)
+void do_interrupt_all(X86CPU *cpu, int intno, int is_int,
+                      int error_code, target_ulong next_eip, int is_hw)
 {
     CPUX86State *env = &cpu->env;
 
@@ -1289,36 +1110,6 @@ static void do_interrupt_all(X86CPU *cpu, int intno, int is_int,
 #endif
 }
 
-void x86_cpu_do_interrupt(CPUState *cs)
-{
-    X86CPU *cpu = X86_CPU(cs);
-    CPUX86State *env = &cpu->env;
-
-#if defined(CONFIG_USER_ONLY)
-    /* if user mode only, we simulate a fake exception
-       which will be handled outside the cpu execution
-       loop */
-    do_interrupt_user(env, cs->exception_index,
-                      env->exception_is_int,
-                      env->error_code,
-                      env->exception_next_eip);
-    /* successfully delivered */
-    env->old_exception = -1;
-#else
-    if (cs->exception_index >= EXCP_VMEXIT) {
-        assert(env->old_exception == -1);
-        do_vmexit(env, cs->exception_index - EXCP_VMEXIT, env->error_code);
-    } else {
-        do_interrupt_all(cpu, cs->exception_index,
-                         env->exception_is_int,
-                         env->error_code,
-                         env->exception_next_eip, 0);
-        /* successfully delivered */
-        env->old_exception = -1;
-    }
-#endif
-}
-
 void do_interrupt_x86_hardirq(CPUX86State *env, int intno, int is_hw)
 {
     do_interrupt_all(env_archcpu(env), intno, 0, 0, 0, is_hw);
@@ -2626,22 +2417,6 @@ void helper_verw(CPUX86State *env, target_ulong selector1)
     CC_SRC = eflags | CC_Z;
 }
 
-#if defined(CONFIG_USER_ONLY)
-void cpu_x86_load_seg(CPUX86State *env, X86Seg seg_reg, int selector)
-{
-    if (!(env->cr[0] & CR0_PE_MASK) || (env->eflags & VM_MASK)) {
-        int dpl = (env->eflags & VM_MASK) ? 3 : 0;
-        selector &= 0xffff;
-        cpu_x86_load_seg_cache(env, seg_reg, selector,
-                               (selector << 4), 0xffff,
-                               DESC_P_MASK | DESC_S_MASK | DESC_W_MASK |
-                               DESC_A_MASK | (dpl << DESC_DPL_SHIFT));
-    } else {
-        helper_load_seg(env, seg_reg, selector);
-    }
-}
-#endif
-
 /* check if Port I/O is allowed in TSS */
 static inline void check_io(CPUX86State *env, int addr, int size,
                             uintptr_t retaddr)
diff --git a/target/i386/tcg/sysemu/seg_helper.c b/target/i386/tcg/sysemu/seg_helper.c
new file mode 100644
index 0000000000..10076d1341
--- /dev/null
+++ b/target/i386/tcg/sysemu/seg_helper.c
@@ -0,0 +1,125 @@
+/*
+ *  x86 segmentation related helpers: (sysemu-only code)
+ *  TSS, interrupts, system calls, jumps and call/task gates, descriptors
+ *
+ *  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 2.1 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 "exec/helper-proto.h"
+#include "exec/cpu_ldst.h"
+#include "tcg/helper-tcg.h"
+
+#ifdef TARGET_X86_64
+void helper_syscall(CPUX86State *env, int next_eip_addend)
+{
+    int selector;
+
+    if (!(env->efer & MSR_EFER_SCE)) {
+        raise_exception_err_ra(env, EXCP06_ILLOP, 0, GETPC());
+    }
+    selector = (env->star >> 32) & 0xffff;
+    if (env->hflags & HF_LMA_MASK) {
+        int code64;
+
+        env->regs[R_ECX] = env->eip + next_eip_addend;
+        env->regs[11] = cpu_compute_eflags(env) & ~RF_MASK;
+
+        code64 = env->hflags & HF_CS64_MASK;
+
+        env->eflags &= ~(env->fmask | RF_MASK);
+        cpu_load_eflags(env, env->eflags, 0);
+        cpu_x86_load_seg_cache(env, R_CS, selector & 0xfffc,
+                           0, 0xffffffff,
+                               DESC_G_MASK | DESC_P_MASK |
+                               DESC_S_MASK |
+                               DESC_CS_MASK | DESC_R_MASK | DESC_A_MASK |
+                               DESC_L_MASK);
+        cpu_x86_load_seg_cache(env, R_SS, (selector + 8) & 0xfffc,
+                               0, 0xffffffff,
+                               DESC_G_MASK | DESC_B_MASK | DESC_P_MASK |
+                               DESC_S_MASK |
+                               DESC_W_MASK | DESC_A_MASK);
+        if (code64) {
+            env->eip = env->lstar;
+        } else {
+            env->eip = env->cstar;
+        }
+    } else {
+        env->regs[R_ECX] = (uint32_t)(env->eip + next_eip_addend);
+
+        env->eflags &= ~(IF_MASK | RF_MASK | VM_MASK);
+        cpu_x86_load_seg_cache(env, R_CS, selector & 0xfffc,
+                           0, 0xffffffff,
+                               DESC_G_MASK | DESC_B_MASK | DESC_P_MASK |
+                               DESC_S_MASK |
+                               DESC_CS_MASK | DESC_R_MASK | DESC_A_MASK);
+        cpu_x86_load_seg_cache(env, R_SS, (selector + 8) & 0xfffc,
+                               0, 0xffffffff,
+                               DESC_G_MASK | DESC_B_MASK | DESC_P_MASK |
+                               DESC_S_MASK |
+                               DESC_W_MASK | DESC_A_MASK);
+        env->eip = (uint32_t)env->star;
+    }
+}
+#endif /* TARGET_X86_64 */
+
+void handle_even_inj(CPUX86State *env, int intno, int is_int,
+                     int error_code, int is_hw, int rm)
+{
+    CPUState *cs = env_cpu(env);
+    uint32_t event_inj = x86_ldl_phys(cs, env->vm_vmcb + offsetof(struct vmcb,
+                                                          control.event_inj));
+
+    if (!(event_inj & SVM_EVTINJ_VALID)) {
+        int type;
+
+        if (is_int) {
+            type = SVM_EVTINJ_TYPE_SOFT;
+        } else {
+            type = SVM_EVTINJ_TYPE_EXEPT;
+        }
+        event_inj = intno | type | SVM_EVTINJ_VALID;
+        if (!rm && exception_has_error_code(intno)) {
+            event_inj |= SVM_EVTINJ_VALID_ERR;
+            x86_stl_phys(cs, env->vm_vmcb + offsetof(struct vmcb,
+                                             control.event_inj_err),
+                     error_code);
+        }
+        x86_stl_phys(cs,
+                 env->vm_vmcb + offsetof(struct vmcb, control.event_inj),
+                 event_inj);
+    }
+}
+
+void x86_cpu_do_interrupt(CPUState *cs)
+{
+    X86CPU *cpu = X86_CPU(cs);
+    CPUX86State *env = &cpu->env;
+
+    if (cs->exception_index >= EXCP_VMEXIT) {
+        assert(env->old_exception == -1);
+        do_vmexit(env, cs->exception_index - EXCP_VMEXIT, env->error_code);
+    } else {
+        do_interrupt_all(cpu, cs->exception_index,
+                         env->exception_is_int,
+                         env->error_code,
+                         env->exception_next_eip, 0);
+        /* successfully delivered */
+        env->old_exception = -1;
+    }
+}
diff --git a/target/i386/tcg/user/seg_helper.c b/target/i386/tcg/user/seg_helper.c
new file mode 100644
index 0000000000..67481b0aa8
--- /dev/null
+++ b/target/i386/tcg/user/seg_helper.c
@@ -0,0 +1,109 @@
+/*
+ *  x86 segmentation related helpers (user-mode code):
+ *  TSS, interrupts, system calls, jumps and call/task gates, descriptors
+ *
+ *  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 2.1 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 "exec/helper-proto.h"
+#include "exec/exec-all.h"
+#include "exec/cpu_ldst.h"
+#include "tcg/helper-tcg.h"
+#include "tcg/seg_helper.h"
+
+#ifdef TARGET_X86_64
+void helper_syscall(CPUX86State *env, int next_eip_addend)
+{
+    CPUState *cs = env_cpu(env);
+
+    cs->exception_index = EXCP_SYSCALL;
+    env->exception_is_int = 0;
+    env->exception_next_eip = env->eip + next_eip_addend;
+    cpu_loop_exit(cs);
+}
+#endif /* TARGET_X86_64 */
+
+/*
+ * fake user mode interrupt. is_int is TRUE if coming from the int
+ * instruction. next_eip is the env->eip value AFTER the interrupt
+ * instruction. It is only relevant if is_int is TRUE or if intno
+ * is EXCP_SYSCALL.
+ */
+static void do_interrupt_user(CPUX86State *env, int intno, int is_int,
+                              int error_code, target_ulong next_eip)
+{
+    if (is_int) {
+        SegmentCache *dt;
+        target_ulong ptr;
+        int dpl, cpl, shift;
+        uint32_t e2;
+
+        dt = &env->idt;
+        if (env->hflags & HF_LMA_MASK) {
+            shift = 4;
+        } else {
+            shift = 3;
+        }
+        ptr = dt->base + (intno << shift);
+        e2 = cpu_ldl_kernel(env, ptr + 4);
+
+        dpl = (e2 >> DESC_DPL_SHIFT) & 3;
+        cpl = env->hflags & HF_CPL_MASK;
+        /* check privilege if software int */
+        if (dpl < cpl) {
+            raise_exception_err(env, EXCP0D_GPF, (intno << shift) + 2);
+        }
+    }
+
+    /* Since we emulate only user space, we cannot do more than
+       exiting the emulation with the suitable exception and error
+       code. So update EIP for INT 0x80 and EXCP_SYSCALL. */
+    if (is_int || intno == EXCP_SYSCALL) {
+        env->eip = next_eip;
+    }
+}
+
+void x86_cpu_do_interrupt(CPUState *cs)
+{
+    X86CPU *cpu = X86_CPU(cs);
+    CPUX86State *env = &cpu->env;
+
+    /* if user mode only, we simulate a fake exception
+       which will be handled outside the cpu execution
+       loop */
+    do_interrupt_user(env, cs->exception_index,
+                      env->exception_is_int,
+                      env->error_code,
+                      env->exception_next_eip);
+    /* successfully delivered */
+    env->old_exception = -1;
+}
+
+void cpu_x86_load_seg(CPUX86State *env, X86Seg seg_reg, int selector)
+{
+    if (!(env->cr[0] & CR0_PE_MASK) || (env->eflags & VM_MASK)) {
+        int dpl = (env->eflags & VM_MASK) ? 3 : 0;
+        selector &= 0xffff;
+        cpu_x86_load_seg_cache(env, seg_reg, selector,
+                               (selector << 4), 0xffff,
+                               DESC_P_MASK | DESC_S_MASK | DESC_W_MASK |
+                               DESC_A_MASK | (dpl << DESC_DPL_SHIFT));
+    } else {
+        helper_load_seg(env, seg_reg, selector);
+    }
+}
diff --git a/target/i386/tcg/sysemu/meson.build b/target/i386/tcg/sysemu/meson.build
index 126528d0c9..2e444e766a 100644
--- a/target/i386/tcg/sysemu/meson.build
+++ b/target/i386/tcg/sysemu/meson.build
@@ -6,4 +6,5 @@ i386_softmmu_ss.add(when: ['CONFIG_TCG', 'CONFIG_SOFTMMU'], if_true: files(
   'misc_helper.c',
   'fpu_helper.c',
   'svm_helper.c',
+  'seg_helper.c',
 ))
diff --git a/target/i386/tcg/user/meson.build b/target/i386/tcg/user/meson.build
index 3edaee7402..9eac0e69ca 100644
--- a/target/i386/tcg/user/meson.build
+++ b/target/i386/tcg/user/meson.build
@@ -2,4 +2,5 @@ i386_user_ss.add(when: ['CONFIG_TCG', 'CONFIG_USER_ONLY'], if_true: files(
   'excp_helper.c',
   'misc_stubs.c',
   'svm_stubs.c',
+  'seg_helper.c',
 ))
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 17/20] i386: split off sysemu part of cpu.c
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (15 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 16/20] i386: split seg_helper into user-only and sysemu parts Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 18/20] target/i386: gdbstub: introduce aux functions to read/write CS64 regs Claudio Fontana
                   ` (4 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/cpu-internal.h |  70 +++++++
 target/i386/cpu-sysemu.c   | 352 +++++++++++++++++++++++++++++++++
 target/i386/cpu.c          | 385 +------------------------------------
 target/i386/meson.build    |   1 +
 4 files changed, 429 insertions(+), 379 deletions(-)
 create mode 100644 target/i386/cpu-internal.h
 create mode 100644 target/i386/cpu-sysemu.c

diff --git a/target/i386/cpu-internal.h b/target/i386/cpu-internal.h
new file mode 100644
index 0000000000..9baac5c0b4
--- /dev/null
+++ b/target/i386/cpu-internal.h
@@ -0,0 +1,70 @@
+/*
+ * i386 CPU internal definitions to be shared between cpu.c and cpu-sysemu.c
+ *
+ *  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 2.1 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_CPU_INTERNAL_H
+#define I386_CPU_INTERNAL_H
+
+typedef enum FeatureWordType {
+   CPUID_FEATURE_WORD,
+   MSR_FEATURE_WORD,
+} FeatureWordType;
+
+typedef struct FeatureWordInfo {
+    FeatureWordType type;
+    /* feature flags names are taken from "Intel Processor Identification and
+     * the CPUID Instruction" and AMD's "CPUID Specification".
+     * In cases of disagreement between feature naming conventions,
+     * aliases may be added.
+     */
+    const char *feat_names[64];
+    union {
+        /* If type==CPUID_FEATURE_WORD */
+        struct {
+            uint32_t eax;   /* Input EAX for CPUID */
+            bool needs_ecx; /* CPUID instruction uses ECX as input */
+            uint32_t ecx;   /* Input ECX value for CPUID */
+            int reg;        /* output register (R_* constant) */
+        } cpuid;
+        /* If type==MSR_FEATURE_WORD */
+        struct {
+            uint32_t index;
+        } msr;
+    };
+    uint64_t tcg_features; /* Feature flags supported by TCG */
+    uint64_t unmigratable_flags; /* Feature flags known to be unmigratable */
+    uint64_t migratable_flags; /* Feature flags known to be migratable */
+    /* Features that shouldn't be auto-enabled by "-cpu host" */
+    uint64_t no_autoenable_flags;
+} FeatureWordInfo;
+
+extern FeatureWordInfo feature_word_info[];
+
+void x86_cpu_expand_features(X86CPU *cpu, Error **errp);
+
+#ifndef CONFIG_USER_ONLY
+GuestPanicInformation *x86_cpu_get_crash_info(CPUState *cs);
+void x86_cpu_get_crash_info_qom(Object *obj, Visitor *v,
+                                const char *name, void *opaque, Error **errp);
+
+void x86_cpu_apic_create(X86CPU *cpu, Error **errp);
+void x86_cpu_apic_realize(X86CPU *cpu, Error **errp);
+void x86_cpu_machine_reset_cb(void *opaque);
+#endif /* !CONFIG_USER_ONLY */
+
+#endif /* I386_CPU_INTERNAL_H */
diff --git a/target/i386/cpu-sysemu.c b/target/i386/cpu-sysemu.c
new file mode 100644
index 0000000000..6477584313
--- /dev/null
+++ b/target/i386/cpu-sysemu.c
@@ -0,0 +1,352 @@
+/*
+ *  i386 CPUID, CPU class, definitions, models: sysemu-only code
+ *
+ *  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 2.1 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 "sysemu/xen.h"
+#include "sysemu/whpx.h"
+#include "kvm/kvm_i386.h"
+#include "qapi/error.h"
+#include "qapi/qapi-visit-run-state.h"
+#include "qapi/qmp/qdict.h"
+#include "qom/qom-qobject.h"
+#include "qapi/qapi-commands-machine-target.h"
+#include "hw/qdev-properties.h"
+
+#include "exec/address-spaces.h"
+#include "hw/i386/apic_internal.h"
+
+#include "cpu-internal.h"
+
+/* Return a QDict containing keys for all properties that can be included
+ * in static expansion of CPU models. All properties set by x86_cpu_load_model()
+ * must be included in the dictionary.
+ */
+static QDict *x86_cpu_static_props(void)
+{
+    FeatureWord w;
+    int i;
+    static const char *props[] = {
+        "min-level",
+        "min-xlevel",
+        "family",
+        "model",
+        "stepping",
+        "model-id",
+        "vendor",
+        "lmce",
+        NULL,
+    };
+    static QDict *d;
+
+    if (d) {
+        return d;
+    }
+
+    d = qdict_new();
+    for (i = 0; props[i]; i++) {
+        qdict_put_null(d, props[i]);
+    }
+
+    for (w = 0; w < FEATURE_WORDS; w++) {
+        FeatureWordInfo *fi = &feature_word_info[w];
+        int bit;
+        for (bit = 0; bit < 64; bit++) {
+            if (!fi->feat_names[bit]) {
+                continue;
+            }
+            qdict_put_null(d, fi->feat_names[bit]);
+        }
+    }
+
+    return d;
+}
+
+/* Add an entry to @props dict, with the value for property. */
+static void x86_cpu_expand_prop(X86CPU *cpu, QDict *props, const char *prop)
+{
+    QObject *value = object_property_get_qobject(OBJECT(cpu), prop,
+                                                 &error_abort);
+
+    qdict_put_obj(props, prop, value);
+}
+
+/* Convert CPU model data from X86CPU object to a property dictionary
+ * that can recreate exactly the same CPU model.
+ */
+static void x86_cpu_to_dict(X86CPU *cpu, QDict *props)
+{
+    QDict *sprops = x86_cpu_static_props();
+    const QDictEntry *e;
+
+    for (e = qdict_first(sprops); e; e = qdict_next(sprops, e)) {
+        const char *prop = qdict_entry_key(e);
+        x86_cpu_expand_prop(cpu, props, prop);
+    }
+}
+
+/* Convert CPU model data from X86CPU object to a property dictionary
+ * that can recreate exactly the same CPU model, including every
+ * writeable QOM property.
+ */
+static void x86_cpu_to_dict_full(X86CPU *cpu, QDict *props)
+{
+    ObjectPropertyIterator iter;
+    ObjectProperty *prop;
+
+    object_property_iter_init(&iter, OBJECT(cpu));
+    while ((prop = object_property_iter_next(&iter))) {
+        /* skip read-only or write-only properties */
+        if (!prop->get || !prop->set) {
+            continue;
+        }
+
+        /* "hotplugged" is the only property that is configurable
+         * on the command-line but will be set differently on CPUs
+         * created using "-cpu ... -smp ..." and by CPUs created
+         * on the fly by x86_cpu_from_model() for querying. Skip it.
+         */
+        if (!strcmp(prop->name, "hotplugged")) {
+            continue;
+        }
+        x86_cpu_expand_prop(cpu, props, prop->name);
+    }
+}
+
+static void object_apply_props(Object *obj, QDict *props, Error **errp)
+{
+    const QDictEntry *prop;
+
+    for (prop = qdict_first(props); prop; prop = qdict_next(props, prop)) {
+        if (!object_property_set_qobject(obj, qdict_entry_key(prop),
+                                         qdict_entry_value(prop), errp)) {
+            break;
+        }
+    }
+}
+
+/* Create X86CPU object according to model+props specification */
+static X86CPU *x86_cpu_from_model(const char *model, QDict *props, Error **errp)
+{
+    X86CPU *xc = NULL;
+    X86CPUClass *xcc;
+    Error *err = NULL;
+
+    xcc = X86_CPU_CLASS(cpu_class_by_name(TYPE_X86_CPU, model));
+    if (xcc == NULL) {
+        error_setg(&err, "CPU model '%s' not found", model);
+        goto out;
+    }
+
+    xc = X86_CPU(object_new_with_class(OBJECT_CLASS(xcc)));
+    if (props) {
+        object_apply_props(OBJECT(xc), props, &err);
+        if (err) {
+            goto out;
+        }
+    }
+
+    x86_cpu_expand_features(xc, &err);
+    if (err) {
+        goto out;
+    }
+
+out:
+    if (err) {
+        error_propagate(errp, err);
+        object_unref(OBJECT(xc));
+        xc = NULL;
+    }
+    return xc;
+}
+
+CpuModelExpansionInfo *
+qmp_query_cpu_model_expansion(CpuModelExpansionType type,
+                                                      CpuModelInfo *model,
+                                                      Error **errp)
+{
+    X86CPU *xc = NULL;
+    Error *err = NULL;
+    CpuModelExpansionInfo *ret = g_new0(CpuModelExpansionInfo, 1);
+    QDict *props = NULL;
+    const char *base_name;
+
+    xc = x86_cpu_from_model(model->name,
+                            model->has_props ?
+                                qobject_to(QDict, model->props) :
+                                NULL, &err);
+    if (err) {
+        goto out;
+    }
+
+    props = qdict_new();
+    ret->model = g_new0(CpuModelInfo, 1);
+    ret->model->props = QOBJECT(props);
+    ret->model->has_props = true;
+
+    switch (type) {
+    case CPU_MODEL_EXPANSION_TYPE_STATIC:
+        /* Static expansion will be based on "base" only */
+        base_name = "base";
+        x86_cpu_to_dict(xc, props);
+    break;
+    case CPU_MODEL_EXPANSION_TYPE_FULL:
+        /* As we don't return every single property, full expansion needs
+         * to keep the original model name+props, and add extra
+         * properties on top of that.
+         */
+        base_name = model->name;
+        x86_cpu_to_dict_full(xc, props);
+    break;
+    default:
+        error_setg(&err, "Unsupported expansion type");
+        goto out;
+    }
+
+    x86_cpu_to_dict(xc, props);
+
+    ret->model->name = g_strdup(base_name);
+
+out:
+    object_unref(OBJECT(xc));
+    if (err) {
+        error_propagate(errp, err);
+        qapi_free_CpuModelExpansionInfo(ret);
+        ret = NULL;
+    }
+    return ret;
+}
+
+void cpu_clear_apic_feature(CPUX86State *env)
+{
+    env->features[FEAT_1_EDX] &= ~CPUID_APIC;
+}
+
+bool cpu_is_bsp(X86CPU *cpu)
+{
+    return cpu_get_apic_base(cpu->apic_state) & MSR_IA32_APICBASE_BSP;
+}
+
+/* TODO: remove me, when reset over QOM tree is implemented */
+void x86_cpu_machine_reset_cb(void *opaque)
+{
+    X86CPU *cpu = opaque;
+    cpu_reset(CPU(cpu));
+}
+
+APICCommonClass *apic_get_class(void)
+{
+    const char *apic_type = "apic";
+
+    /* TODO: in-kernel irqchip for hvf */
+    if (kvm_apic_in_kernel()) {
+        apic_type = "kvm-apic";
+    } else if (xen_enabled()) {
+        apic_type = "xen-apic";
+    } else if (whpx_apic_in_platform()) {
+        apic_type = "whpx-apic";
+    }
+
+    return APIC_COMMON_CLASS(object_class_by_name(apic_type));
+}
+
+void x86_cpu_apic_create(X86CPU *cpu, Error **errp)
+{
+    APICCommonState *apic;
+    ObjectClass *apic_class = OBJECT_CLASS(apic_get_class());
+
+    cpu->apic_state = DEVICE(object_new_with_class(apic_class));
+
+    object_property_add_child(OBJECT(cpu), "lapic",
+                              OBJECT(cpu->apic_state));
+    object_unref(OBJECT(cpu->apic_state));
+
+    qdev_prop_set_uint32(cpu->apic_state, "id", cpu->apic_id);
+    /* TODO: convert to link<> */
+    apic = APIC_COMMON(cpu->apic_state);
+    apic->cpu = cpu;
+    apic->apicbase = APIC_DEFAULT_ADDRESS | MSR_IA32_APICBASE_ENABLE;
+}
+
+void x86_cpu_apic_realize(X86CPU *cpu, Error **errp)
+{
+    APICCommonState *apic;
+    static bool apic_mmio_map_once;
+
+    if (cpu->apic_state == NULL) {
+        return;
+    }
+    qdev_realize(DEVICE(cpu->apic_state), NULL, errp);
+
+    /* Map APIC MMIO area */
+    apic = APIC_COMMON(cpu->apic_state);
+    if (!apic_mmio_map_once) {
+        memory_region_add_subregion_overlap(get_system_memory(),
+                                            apic->apicbase &
+                                            MSR_IA32_APICBASE_BASE,
+                                            &apic->io_memory,
+                                            0x1000);
+        apic_mmio_map_once = true;
+     }
+}
+
+GuestPanicInformation *x86_cpu_get_crash_info(CPUState *cs)
+{
+    X86CPU *cpu = X86_CPU(cs);
+    CPUX86State *env = &cpu->env;
+    GuestPanicInformation *panic_info = NULL;
+
+    if (env->features[FEAT_HYPERV_EDX] & HV_GUEST_CRASH_MSR_AVAILABLE) {
+        panic_info = g_malloc0(sizeof(GuestPanicInformation));
+
+        panic_info->type = GUEST_PANIC_INFORMATION_TYPE_HYPER_V;
+
+        assert(HV_CRASH_PARAMS >= 5);
+        panic_info->u.hyper_v.arg1 = env->msr_hv_crash_params[0];
+        panic_info->u.hyper_v.arg2 = env->msr_hv_crash_params[1];
+        panic_info->u.hyper_v.arg3 = env->msr_hv_crash_params[2];
+        panic_info->u.hyper_v.arg4 = env->msr_hv_crash_params[3];
+        panic_info->u.hyper_v.arg5 = env->msr_hv_crash_params[4];
+    }
+
+    return panic_info;
+}
+void x86_cpu_get_crash_info_qom(Object *obj, Visitor *v,
+                                const char *name, void *opaque,
+                                Error **errp)
+{
+    CPUState *cs = CPU(obj);
+    GuestPanicInformation *panic_info;
+
+    if (!cs->crash_occurred) {
+        error_setg(errp, "No crash occured");
+        return;
+    }
+
+    panic_info = x86_cpu_get_crash_info(cs);
+    if (panic_info == NULL) {
+        error_setg(errp, "No crash information");
+        return;
+    }
+
+    visit_type_GuestPanicInformation(v, "crash-information", &panic_info,
+                                     errp);
+    qapi_free_GuestPanicInformation(panic_info);
+}
+
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 14e2a60ee5..4da7b2f7a9 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -1,5 +1,5 @@
 /*
- *  i386 CPUID helper functions
+ *  i386 CPUID, CPU class, definitions, models
  *
  *  Copyright (c) 2003 Fabrice Bellard
  *
@@ -20,35 +20,26 @@
 #include "qemu/osdep.h"
 #include "qemu/units.h"
 #include "qemu/cutils.h"
-#include "qemu/bitops.h"
 #include "qemu/qemu-print.h"
 #include "cpu.h"
 #include "tcg/helper-tcg.h"
-#include "exec/exec-all.h"
-#include "sysemu/kvm.h"
 #include "sysemu/reset.h"
 #include "sysemu/hvf.h"
-#include "sysemu/xen.h"
-#include "sysemu/whpx.h"
 #include "kvm/kvm_i386.h"
 #include "sev_i386.h"
-#include "qemu/module.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 "qom/qom-qobject.h"
 #include "qapi/qapi-commands-machine-target.h"
 #include "standard-headers/asm-x86/kvm_para.h"
 #include "hw/qdev-properties.h"
 #include "hw/i386/topology.h"
 #ifndef CONFIG_USER_ONLY
 #include "exec/address-spaces.h"
-#include "hw/i386/apic_internal.h"
 #include "hw/boards.h"
 #endif
 
 #include "disas/capstone.h"
+#include "cpu-internal.h"
 
 /* Helpers for building CPUID[2] descriptors: */
 
@@ -663,40 +654,7 @@ void x86_cpu_vendor_words2str(char *dst, uint32_t vendor1,
           CPUID_XSAVE_XSAVEC, CPUID_XSAVE_XSAVES */
 #define TCG_14_0_ECX_FEATURES 0
 
-typedef enum FeatureWordType {
-   CPUID_FEATURE_WORD,
-   MSR_FEATURE_WORD,
-} FeatureWordType;
-
-typedef struct FeatureWordInfo {
-    FeatureWordType type;
-    /* feature flags names are taken from "Intel Processor Identification and
-     * the CPUID Instruction" and AMD's "CPUID Specification".
-     * In cases of disagreement between feature naming conventions,
-     * aliases may be added.
-     */
-    const char *feat_names[64];
-    union {
-        /* If type==CPUID_FEATURE_WORD */
-        struct {
-            uint32_t eax;   /* Input EAX for CPUID */
-            bool needs_ecx; /* CPUID instruction uses ECX as input */
-            uint32_t ecx;   /* Input ECX value for CPUID */
-            int reg;        /* output register (R_* constant) */
-        } cpuid;
-        /* If type==MSR_FEATURE_WORD */
-        struct {
-            uint32_t index;
-        } msr;
-    };
-    uint64_t tcg_features; /* Feature flags supported by TCG */
-    uint64_t unmigratable_flags; /* Feature flags known to be unmigratable */
-    uint64_t migratable_flags; /* Feature flags known to be migratable */
-    /* Features that shouldn't be auto-enabled by "-cpu host" */
-    uint64_t no_autoenable_flags;
-} FeatureWordInfo;
-
-static FeatureWordInfo feature_word_info[FEATURE_WORDS] = {
+FeatureWordInfo feature_word_info[FEATURE_WORDS] = {
     [FEAT_1_EDX] = {
         .type = CPUID_FEATURE_WORD,
         .feat_names = {
@@ -4738,7 +4696,6 @@ static void x86_cpu_parse_featurestr(const char *typename, char *features,
     }
 }
 
-static void x86_cpu_expand_features(X86CPU *cpu, Error **errp);
 static void x86_cpu_filter_features(X86CPU *cpu, bool verbose);
 
 /* Build a list with the name of all features on a feature word array */
@@ -5108,207 +5065,6 @@ static void x86_cpu_load_model(X86CPU *cpu, X86CPUModel *model)
     memset(&env->user_features, 0, sizeof(env->user_features));
 }
 
-#ifndef CONFIG_USER_ONLY
-/* Return a QDict containing keys for all properties that can be included
- * in static expansion of CPU models. All properties set by x86_cpu_load_model()
- * must be included in the dictionary.
- */
-static QDict *x86_cpu_static_props(void)
-{
-    FeatureWord w;
-    int i;
-    static const char *props[] = {
-        "min-level",
-        "min-xlevel",
-        "family",
-        "model",
-        "stepping",
-        "model-id",
-        "vendor",
-        "lmce",
-        NULL,
-    };
-    static QDict *d;
-
-    if (d) {
-        return d;
-    }
-
-    d = qdict_new();
-    for (i = 0; props[i]; i++) {
-        qdict_put_null(d, props[i]);
-    }
-
-    for (w = 0; w < FEATURE_WORDS; w++) {
-        FeatureWordInfo *fi = &feature_word_info[w];
-        int bit;
-        for (bit = 0; bit < 64; bit++) {
-            if (!fi->feat_names[bit]) {
-                continue;
-            }
-            qdict_put_null(d, fi->feat_names[bit]);
-        }
-    }
-
-    return d;
-}
-
-/* Add an entry to @props dict, with the value for property. */
-static void x86_cpu_expand_prop(X86CPU *cpu, QDict *props, const char *prop)
-{
-    QObject *value = object_property_get_qobject(OBJECT(cpu), prop,
-                                                 &error_abort);
-
-    qdict_put_obj(props, prop, value);
-}
-
-/* Convert CPU model data from X86CPU object to a property dictionary
- * that can recreate exactly the same CPU model.
- */
-static void x86_cpu_to_dict(X86CPU *cpu, QDict *props)
-{
-    QDict *sprops = x86_cpu_static_props();
-    const QDictEntry *e;
-
-    for (e = qdict_first(sprops); e; e = qdict_next(sprops, e)) {
-        const char *prop = qdict_entry_key(e);
-        x86_cpu_expand_prop(cpu, props, prop);
-    }
-}
-
-/* Convert CPU model data from X86CPU object to a property dictionary
- * that can recreate exactly the same CPU model, including every
- * writeable QOM property.
- */
-static void x86_cpu_to_dict_full(X86CPU *cpu, QDict *props)
-{
-    ObjectPropertyIterator iter;
-    ObjectProperty *prop;
-
-    object_property_iter_init(&iter, OBJECT(cpu));
-    while ((prop = object_property_iter_next(&iter))) {
-        /* skip read-only or write-only properties */
-        if (!prop->get || !prop->set) {
-            continue;
-        }
-
-        /* "hotplugged" is the only property that is configurable
-         * on the command-line but will be set differently on CPUs
-         * created using "-cpu ... -smp ..." and by CPUs created
-         * on the fly by x86_cpu_from_model() for querying. Skip it.
-         */
-        if (!strcmp(prop->name, "hotplugged")) {
-            continue;
-        }
-        x86_cpu_expand_prop(cpu, props, prop->name);
-    }
-}
-
-static void object_apply_props(Object *obj, QDict *props, Error **errp)
-{
-    const QDictEntry *prop;
-
-    for (prop = qdict_first(props); prop; prop = qdict_next(props, prop)) {
-        if (!object_property_set_qobject(obj, qdict_entry_key(prop),
-                                         qdict_entry_value(prop), errp)) {
-            break;
-        }
-    }
-}
-
-/* Create X86CPU object according to model+props specification */
-static X86CPU *x86_cpu_from_model(const char *model, QDict *props, Error **errp)
-{
-    X86CPU *xc = NULL;
-    X86CPUClass *xcc;
-    Error *err = NULL;
-
-    xcc = X86_CPU_CLASS(cpu_class_by_name(TYPE_X86_CPU, model));
-    if (xcc == NULL) {
-        error_setg(&err, "CPU model '%s' not found", model);
-        goto out;
-    }
-
-    xc = X86_CPU(object_new_with_class(OBJECT_CLASS(xcc)));
-    if (props) {
-        object_apply_props(OBJECT(xc), props, &err);
-        if (err) {
-            goto out;
-        }
-    }
-
-    x86_cpu_expand_features(xc, &err);
-    if (err) {
-        goto out;
-    }
-
-out:
-    if (err) {
-        error_propagate(errp, err);
-        object_unref(OBJECT(xc));
-        xc = NULL;
-    }
-    return xc;
-}
-
-CpuModelExpansionInfo *
-qmp_query_cpu_model_expansion(CpuModelExpansionType type,
-                                                      CpuModelInfo *model,
-                                                      Error **errp)
-{
-    X86CPU *xc = NULL;
-    Error *err = NULL;
-    CpuModelExpansionInfo *ret = g_new0(CpuModelExpansionInfo, 1);
-    QDict *props = NULL;
-    const char *base_name;
-
-    xc = x86_cpu_from_model(model->name,
-                            model->has_props ?
-                                qobject_to(QDict, model->props) :
-                                NULL, &err);
-    if (err) {
-        goto out;
-    }
-
-    props = qdict_new();
-    ret->model = g_new0(CpuModelInfo, 1);
-    ret->model->props = QOBJECT(props);
-    ret->model->has_props = true;
-
-    switch (type) {
-    case CPU_MODEL_EXPANSION_TYPE_STATIC:
-        /* Static expansion will be based on "base" only */
-        base_name = "base";
-        x86_cpu_to_dict(xc, props);
-    break;
-    case CPU_MODEL_EXPANSION_TYPE_FULL:
-        /* As we don't return every single property, full expansion needs
-         * to keep the original model name+props, and add extra
-         * properties on top of that.
-         */
-        base_name = model->name;
-        x86_cpu_to_dict_full(xc, props);
-    break;
-    default:
-        error_setg(&err, "Unsupported expansion type");
-        goto out;
-    }
-
-    x86_cpu_to_dict(xc, props);
-
-    ret->model->name = g_strdup(base_name);
-
-out:
-    object_unref(OBJECT(xc));
-    if (err) {
-        error_propagate(errp, err);
-        qapi_free_CpuModelExpansionInfo(ret);
-        ret = NULL;
-    }
-    return ret;
-}
-#endif  /* !CONFIG_USER_ONLY */
-
 static gchar *x86_gdb_arch_name(CPUState *cs)
 {
 #ifdef TARGET_X86_64
@@ -5383,15 +5139,6 @@ static void x86_register_cpudef_types(X86CPUDefinition *def)
 
 }
 
-#if !defined(CONFIG_USER_ONLY)
-
-void cpu_clear_apic_feature(CPUX86State *env)
-{
-    env->features[FEAT_1_EDX] &= ~CPUID_APIC;
-}
-
-#endif /* !CONFIG_USER_ONLY */
-
 void cpu_x86_cpuid(CPUX86State *env, uint32_t index, uint32_t count,
                    uint32_t *eax, uint32_t *ebx,
                    uint32_t *ecx, uint32_t *edx)
@@ -6040,20 +5787,6 @@ static void x86_cpu_reset(DeviceState *dev)
 #endif
 }
 
-#ifndef CONFIG_USER_ONLY
-bool cpu_is_bsp(X86CPU *cpu)
-{
-    return cpu_get_apic_base(cpu->apic_state) & MSR_IA32_APICBASE_BSP;
-}
-
-/* TODO: remove me, when reset over QOM tree is implemented */
-static void x86_cpu_machine_reset_cb(void *opaque)
-{
-    X86CPU *cpu = opaque;
-    cpu_reset(CPU(cpu));
-}
-#endif
-
 static void mce_init(X86CPU *cpu)
 {
     CPUX86State *cenv = &cpu->env;
@@ -6071,68 +5804,6 @@ static void mce_init(X86CPU *cpu)
     }
 }
 
-#ifndef CONFIG_USER_ONLY
-APICCommonClass *apic_get_class(void)
-{
-    const char *apic_type = "apic";
-
-    /* TODO: in-kernel irqchip for hvf */
-    if (kvm_apic_in_kernel()) {
-        apic_type = "kvm-apic";
-    } else if (xen_enabled()) {
-        apic_type = "xen-apic";
-    } else if (whpx_apic_in_platform()) {
-        apic_type = "whpx-apic";
-    }
-
-    return APIC_COMMON_CLASS(object_class_by_name(apic_type));
-}
-
-static void x86_cpu_apic_create(X86CPU *cpu, Error **errp)
-{
-    APICCommonState *apic;
-    ObjectClass *apic_class = OBJECT_CLASS(apic_get_class());
-
-    cpu->apic_state = DEVICE(object_new_with_class(apic_class));
-
-    object_property_add_child(OBJECT(cpu), "lapic",
-                              OBJECT(cpu->apic_state));
-    object_unref(OBJECT(cpu->apic_state));
-
-    qdev_prop_set_uint32(cpu->apic_state, "id", cpu->apic_id);
-    /* TODO: convert to link<> */
-    apic = APIC_COMMON(cpu->apic_state);
-    apic->cpu = cpu;
-    apic->apicbase = APIC_DEFAULT_ADDRESS | MSR_IA32_APICBASE_ENABLE;
-}
-
-static void x86_cpu_apic_realize(X86CPU *cpu, Error **errp)
-{
-    APICCommonState *apic;
-    static bool apic_mmio_map_once;
-
-    if (cpu->apic_state == NULL) {
-        return;
-    }
-    qdev_realize(DEVICE(cpu->apic_state), NULL, errp);
-
-    /* Map APIC MMIO area */
-    apic = APIC_COMMON(cpu->apic_state);
-    if (!apic_mmio_map_once) {
-        memory_region_add_subregion_overlap(get_system_memory(),
-                                            apic->apicbase &
-                                            MSR_IA32_APICBASE_BASE,
-                                            &apic->io_memory,
-                                            0x1000);
-        apic_mmio_map_once = true;
-     }
-}
-#else
-static void x86_cpu_apic_realize(X86CPU *cpu, Error **errp)
-{
-}
-#endif
-
 static void x86_cpu_adjust_level(X86CPU *cpu, uint32_t *min, uint32_t value)
 {
     if (*min < value) {
@@ -6236,7 +5907,7 @@ static void x86_cpu_enable_xsave_components(X86CPU *cpu)
 /* Expand CPU configuration data, based on configured features
  * and host/accelerator capabilities when appropriate.
  */
-static void x86_cpu_expand_features(X86CPU *cpu, Error **errp)
+void x86_cpu_expand_features(X86CPU *cpu, Error **errp)
 {
     CPUX86State *env = &cpu->env;
     FeatureWord w;
@@ -6608,10 +6279,12 @@ static void x86_cpu_realizefn(DeviceState *dev, Error **errp)
             ht_warned = true;
     }
 
+#ifndef CONFIG_USER_ONLY
     x86_cpu_apic_realize(cpu, &local_err);
     if (local_err != NULL) {
         goto out;
     }
+#endif /* !CONFIG_USER_ONLY */
     cpu_reset(cs);
 
     xcc->parent_realize(dev, &local_err);
@@ -6735,52 +6408,6 @@ static void x86_cpu_register_feature_bit_props(X86CPUClass *xcc,
     x86_cpu_register_bit_prop(xcc, name, w, bitnr);
 }
 
-#if !defined(CONFIG_USER_ONLY)
-static GuestPanicInformation *x86_cpu_get_crash_info(CPUState *cs)
-{
-    X86CPU *cpu = X86_CPU(cs);
-    CPUX86State *env = &cpu->env;
-    GuestPanicInformation *panic_info = NULL;
-
-    if (env->features[FEAT_HYPERV_EDX] & HV_GUEST_CRASH_MSR_AVAILABLE) {
-        panic_info = g_malloc0(sizeof(GuestPanicInformation));
-
-        panic_info->type = GUEST_PANIC_INFORMATION_TYPE_HYPER_V;
-
-        assert(HV_CRASH_PARAMS >= 5);
-        panic_info->u.hyper_v.arg1 = env->msr_hv_crash_params[0];
-        panic_info->u.hyper_v.arg2 = env->msr_hv_crash_params[1];
-        panic_info->u.hyper_v.arg3 = env->msr_hv_crash_params[2];
-        panic_info->u.hyper_v.arg4 = env->msr_hv_crash_params[3];
-        panic_info->u.hyper_v.arg5 = env->msr_hv_crash_params[4];
-    }
-
-    return panic_info;
-}
-static void x86_cpu_get_crash_info_qom(Object *obj, Visitor *v,
-                                       const char *name, void *opaque,
-                                       Error **errp)
-{
-    CPUState *cs = CPU(obj);
-    GuestPanicInformation *panic_info;
-
-    if (!cs->crash_occurred) {
-        error_setg(errp, "No crash occured");
-        return;
-    }
-
-    panic_info = x86_cpu_get_crash_info(cs);
-    if (panic_info == NULL) {
-        error_setg(errp, "No crash information");
-        return;
-    }
-
-    visit_type_GuestPanicInformation(v, "crash-information", &panic_info,
-                                     errp);
-    qapi_free_GuestPanicInformation(panic_info);
-}
-#endif /* !CONFIG_USER_ONLY */
-
 static void x86_cpu_initfn(Object *obj)
 {
     X86CPU *cpu = X86_CPU(obj);
diff --git a/target/i386/meson.build b/target/i386/meson.build
index cac26a4581..0f50f32b33 100644
--- a/target/i386/meson.build
+++ b/target/i386/meson.build
@@ -18,6 +18,7 @@ i386_softmmu_ss.add(files(
   'arch_memory_mapping.c',
   'machine.c',
   'monitor.c',
+  'cpu-sysemu.c',
 ))
 i386_user_ss = ss.source_set()
 
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 18/20] target/i386: gdbstub: introduce aux functions to read/write CS64 regs
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (16 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 17/20] i386: split off sysemu part of cpu.c Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 19/20] target/i386: gdbstub: only write CR0/CR2/CR3/EFER for sysemu Claudio Fontana
                   ` (3 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

a number of registers are read as 64bit under the condition that
(hflags & HF_CS64_MASK) || TARGET_X86_64)

and a number of registers are written as 64bit under the condition that
(hflags & HF_CS64_MASK).

Provide some auxiliary functions that do that.

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/gdbstub.c | 155 ++++++++++++++----------------------------
 1 file changed, 51 insertions(+), 104 deletions(-)

diff --git a/target/i386/gdbstub.c b/target/i386/gdbstub.c
index 41e265fc67..4ad1295425 100644
--- a/target/i386/gdbstub.c
+++ b/target/i386/gdbstub.c
@@ -78,6 +78,23 @@ static const int gpr_map32[8] = { 0, 1, 2, 3, 4, 5, 6, 7 };
 #define GDB_FORCE_64 0
 #endif
 
+static int gdb_read_reg_cs64(uint32_t hflags, GByteArray *buf, target_ulong val)
+{
+    if ((hflags & HF_CS64_MASK) || GDB_FORCE_64) {
+        return gdb_get_reg64(buf, val);
+    }
+    return gdb_get_reg32(buf, val);
+}
+
+static int gdb_write_reg_cs64(uint32_t hflags, uint8_t *buf, target_ulong *val)
+{
+    if (hflags & HF_CS64_MASK) {
+        *val = ldq_p(buf);
+        return 8;
+    }
+    *val = ldl_p(buf);
+    return 4;
+}
 
 int x86_cpu_gdb_read_register(CPUState *cs, GByteArray *mem_buf, int n)
 {
@@ -142,25 +159,14 @@ int x86_cpu_gdb_read_register(CPUState *cs, GByteArray *mem_buf, int n)
             return gdb_get_reg32(mem_buf, env->segs[R_FS].selector);
         case IDX_SEG_REGS + 5:
             return gdb_get_reg32(mem_buf, env->segs[R_GS].selector);
-
         case IDX_SEG_REGS + 6:
-            if ((env->hflags & HF_CS64_MASK) || GDB_FORCE_64) {
-                return gdb_get_reg64(mem_buf, env->segs[R_FS].base);
-            }
-            return gdb_get_reg32(mem_buf, env->segs[R_FS].base);
-
+            return gdb_read_reg_cs64(env->hflags, mem_buf, env->segs[R_FS].base);
         case IDX_SEG_REGS + 7:
-            if ((env->hflags & HF_CS64_MASK) || GDB_FORCE_64) {
-                return gdb_get_reg64(mem_buf, env->segs[R_GS].base);
-            }
-            return gdb_get_reg32(mem_buf, env->segs[R_GS].base);
+            return gdb_read_reg_cs64(env->hflags, mem_buf, env->segs[R_GS].base);
 
         case IDX_SEG_REGS + 8:
 #ifdef TARGET_X86_64
-            if ((env->hflags & HF_CS64_MASK) || GDB_FORCE_64) {
-                return gdb_get_reg64(mem_buf, env->kernelgsbase);
-            }
-            return gdb_get_reg32(mem_buf, env->kernelgsbase);
+            return gdb_read_reg_cs64(env->hflags, mem_buf, env->kernelgsbase);
 #else
             return gdb_get_reg32(mem_buf, 0);
 #endif
@@ -188,45 +194,23 @@ int x86_cpu_gdb_read_register(CPUState *cs, GByteArray *mem_buf, int n)
             return gdb_get_reg32(mem_buf, env->mxcsr);
 
         case IDX_CTL_CR0_REG:
-            if ((env->hflags & HF_CS64_MASK) || GDB_FORCE_64) {
-                return gdb_get_reg64(mem_buf, env->cr[0]);
-            }
-            return gdb_get_reg32(mem_buf, env->cr[0]);
-
+            return gdb_read_reg_cs64(env->hflags, mem_buf, env->cr[0]);
         case IDX_CTL_CR2_REG:
-            if ((env->hflags & HF_CS64_MASK) || GDB_FORCE_64) {
-                return gdb_get_reg64(mem_buf, env->cr[2]);
-            }
-            return gdb_get_reg32(mem_buf, env->cr[2]);
-
+            return gdb_read_reg_cs64(env->hflags, mem_buf, env->cr[2]);
         case IDX_CTL_CR3_REG:
-            if ((env->hflags & HF_CS64_MASK) || GDB_FORCE_64) {
-                return gdb_get_reg64(mem_buf, env->cr[3]);
-            }
-            return gdb_get_reg32(mem_buf, env->cr[3]);
-
+            return gdb_read_reg_cs64(env->hflags, mem_buf, env->cr[3]);
         case IDX_CTL_CR4_REG:
-            if ((env->hflags & HF_CS64_MASK) || GDB_FORCE_64) {
-                return gdb_get_reg64(mem_buf, env->cr[4]);
-            }
-            return gdb_get_reg32(mem_buf, env->cr[4]);
-
+            return gdb_read_reg_cs64(env->hflags, mem_buf, env->cr[4]);
         case IDX_CTL_CR8_REG:
-#ifdef CONFIG_SOFTMMU
+#ifndef CONFIG_USER_ONLY
             tpr = cpu_get_apic_tpr(cpu->apic_state);
 #else
             tpr = 0;
 #endif
-            if ((env->hflags & HF_CS64_MASK) || GDB_FORCE_64) {
-                return gdb_get_reg64(mem_buf, tpr);
-            }
-            return gdb_get_reg32(mem_buf, tpr);
+            return gdb_read_reg_cs64(env->hflags, mem_buf, tpr);
 
         case IDX_CTL_EFER_REG:
-            if ((env->hflags & HF_CS64_MASK) || GDB_FORCE_64) {
-                return gdb_get_reg64(mem_buf, env->efer);
-            }
-            return gdb_get_reg32(mem_buf, env->efer);
+            return gdb_read_reg_cs64(env->hflags, mem_buf, env->efer);
         }
     }
     return 0;
@@ -266,7 +250,8 @@ int x86_cpu_gdb_write_register(CPUState *cs, uint8_t *mem_buf, int n)
 {
     X86CPU *cpu = X86_CPU(cs);
     CPUX86State *env = &cpu->env;
-    uint32_t tmp;
+    target_ulong tmp;
+    int len;
 
     /* N.B. GDB can't deal with changes in registers or sizes in the middle
        of a session. So if we're in 32-bit mode on a 64-bit cpu, still act
@@ -329,30 +314,13 @@ int x86_cpu_gdb_write_register(CPUState *cs, uint8_t *mem_buf, int n)
             return x86_cpu_gdb_load_seg(cpu, R_FS, mem_buf);
         case IDX_SEG_REGS + 5:
             return x86_cpu_gdb_load_seg(cpu, R_GS, mem_buf);
-
         case IDX_SEG_REGS + 6:
-            if (env->hflags & HF_CS64_MASK) {
-                env->segs[R_FS].base = ldq_p(mem_buf);
-                return 8;
-            }
-            env->segs[R_FS].base = ldl_p(mem_buf);
-            return 4;
-
+            return gdb_write_reg_cs64(env->hflags, mem_buf, &env->segs[R_FS].base);
         case IDX_SEG_REGS + 7:
-            if (env->hflags & HF_CS64_MASK) {
-                env->segs[R_GS].base = ldq_p(mem_buf);
-                return 8;
-            }
-            env->segs[R_GS].base = ldl_p(mem_buf);
-            return 4;
-
+            return gdb_write_reg_cs64(env->hflags, mem_buf, &env->segs[R_GS].base);
         case IDX_SEG_REGS + 8:
 #ifdef TARGET_X86_64
-            if (env->hflags & HF_CS64_MASK) {
-                env->kernelgsbase = ldq_p(mem_buf);
-                return 8;
-            }
-            env->kernelgsbase = ldl_p(mem_buf);
+            return gdb_write_reg_cs64(env->hflags, mem_buf, &env->kernelgsbase);
 #endif
             return 4;
 
@@ -382,57 +350,36 @@ int x86_cpu_gdb_write_register(CPUState *cs, uint8_t *mem_buf, int n)
             return 4;
 
         case IDX_CTL_CR0_REG:
-            if (env->hflags & HF_CS64_MASK) {
-                cpu_x86_update_cr0(env, ldq_p(mem_buf));
-                return 8;
-            }
-            cpu_x86_update_cr0(env, ldl_p(mem_buf));
-            return 4;
+            len = gdb_write_reg_cs64(env->hflags, mem_buf, &tmp);
+            cpu_x86_update_cr0(env, tmp);
+            return len;
 
         case IDX_CTL_CR2_REG:
-            if (env->hflags & HF_CS64_MASK) {
-                env->cr[2] = ldq_p(mem_buf);
-                return 8;
-            }
-            env->cr[2] = ldl_p(mem_buf);
-            return 4;
+            len = gdb_write_reg_cs64(env->hflags, mem_buf, &tmp);
+            env->cr[2] = tmp;
+            return len;
 
         case IDX_CTL_CR3_REG:
-            if (env->hflags & HF_CS64_MASK) {
-                cpu_x86_update_cr3(env, ldq_p(mem_buf));
-                return 8;
-            }
-            cpu_x86_update_cr3(env, ldl_p(mem_buf));
-            return 4;
+            len = gdb_write_reg_cs64(env->hflags, mem_buf, &tmp);
+            cpu_x86_update_cr3(env, tmp);
+            return len;
 
         case IDX_CTL_CR4_REG:
-            if (env->hflags & HF_CS64_MASK) {
-                cpu_x86_update_cr4(env, ldq_p(mem_buf));
-                return 8;
-            }
-            cpu_x86_update_cr4(env, ldl_p(mem_buf));
-            return 4;
+            len = gdb_write_reg_cs64(env->hflags, mem_buf, &tmp);
+            cpu_x86_update_cr4(env, tmp);
+            return len;
 
         case IDX_CTL_CR8_REG:
-            if (env->hflags & HF_CS64_MASK) {
-#ifdef CONFIG_SOFTMMU
-                cpu_set_apic_tpr(cpu->apic_state, ldq_p(mem_buf));
+            len = gdb_write_reg_cs64(env->hflags, mem_buf, &tmp);
+#ifndef CONFIG_USER_ONLY
+            cpu_set_apic_tpr(cpu->apic_state, tmp);
 #endif
-                return 8;
-            }
-#ifdef CONFIG_SOFTMMU
-            cpu_set_apic_tpr(cpu->apic_state, ldl_p(mem_buf));
-#endif
-            return 4;
+            return len;
 
         case IDX_CTL_EFER_REG:
-            if (env->hflags & HF_CS64_MASK) {
-                cpu_load_efer(env, ldq_p(mem_buf));
-                return 8;
-            }
-            cpu_load_efer(env, ldl_p(mem_buf));
-            return 4;
-
+            len = gdb_write_reg_cs64(env->hflags, mem_buf, &tmp);
+            cpu_load_efer(env, tmp);
+            return len;
         }
     }
     /* Unrecognised register.  */
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 19/20] target/i386: gdbstub: only write CR0/CR2/CR3/EFER for sysemu
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (17 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 18/20] target/i386: gdbstub: introduce aux functions to read/write CS64 regs Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  8:54 ` [PATCH v26 20/20] i386: make cpu_load_efer sysemu-only Claudio Fontana
                   ` (2 subsequent siblings)
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/gdbstub.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/target/i386/gdbstub.c b/target/i386/gdbstub.c
index 4ad1295425..098a2ad15a 100644
--- a/target/i386/gdbstub.c
+++ b/target/i386/gdbstub.c
@@ -351,22 +351,30 @@ int x86_cpu_gdb_write_register(CPUState *cs, uint8_t *mem_buf, int n)
 
         case IDX_CTL_CR0_REG:
             len = gdb_write_reg_cs64(env->hflags, mem_buf, &tmp);
+#ifndef CONFIG_USER_ONLY
             cpu_x86_update_cr0(env, tmp);
+#endif
             return len;
 
         case IDX_CTL_CR2_REG:
             len = gdb_write_reg_cs64(env->hflags, mem_buf, &tmp);
+#ifndef CONFIG_USER_ONLY
             env->cr[2] = tmp;
+#endif
             return len;
 
         case IDX_CTL_CR3_REG:
             len = gdb_write_reg_cs64(env->hflags, mem_buf, &tmp);
+#ifndef CONFIG_USER_ONLY
             cpu_x86_update_cr3(env, tmp);
+#endif
             return len;
 
         case IDX_CTL_CR4_REG:
             len = gdb_write_reg_cs64(env->hflags, mem_buf, &tmp);
+#ifndef CONFIG_USER_ONLY
             cpu_x86_update_cr4(env, tmp);
+#endif
             return len;
 
         case IDX_CTL_CR8_REG:
@@ -378,7 +386,9 @@ int x86_cpu_gdb_write_register(CPUState *cs, uint8_t *mem_buf, int n)
 
         case IDX_CTL_EFER_REG:
             len = gdb_write_reg_cs64(env->hflags, mem_buf, &tmp);
+#ifndef CONFIG_USER_ONLY
             cpu_load_efer(env, tmp);
+#endif
             return len;
         }
     }
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* [PATCH v26 20/20] i386: make cpu_load_efer sysemu-only
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (18 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 19/20] target/i386: gdbstub: only write CR0/CR2/CR3/EFER for sysemu Claudio Fontana
@ 2021-03-01  8:54 ` Claudio Fontana
  2021-03-01  9:18 ` [PATCH v26 00/20] i386 cleanup PART 2 no-reply
  2021-03-08 12:57 ` Claudio Fontana
  21 siblings, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-01  8:54 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, Claudio Fontana,
	qemu-devel

cpu_load_efer is now used only for sysemu code.

Therefore, move this function implementation to
sysemu-only section of helper.c

Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/i386/cpu.h    | 20 +++++---------------
 target/i386/helper.c | 13 +++++++++++++
 2 files changed, 18 insertions(+), 15 deletions(-)

diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index 3797789dc2..a1268abe9f 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -1957,6 +1957,11 @@ static inline AddressSpace *cpu_addressspace(CPUState *cs, MemTxAttrs attrs)
     return cpu_get_address_space(cs, cpu_asidx_from_attrs(cs, attrs));
 }
 
+/*
+ * load efer and update the corresponding hflags. XXX: do consistency
+ * checks with cpuid bits?
+ */
+void cpu_load_efer(CPUX86State *env, uint64_t val);
 uint8_t x86_ldub_phys(CPUState *cs, hwaddr addr);
 uint32_t x86_lduw_phys(CPUState *cs, hwaddr addr);
 uint32_t x86_ldl_phys(CPUState *cs, hwaddr addr);
@@ -2053,21 +2058,6 @@ static inline uint32_t cpu_compute_eflags(CPUX86State *env)
     return eflags;
 }
 
-
-/* load efer and update the corresponding hflags. XXX: do consistency
-   checks with cpuid bits? */
-static inline void cpu_load_efer(CPUX86State *env, uint64_t val)
-{
-    env->efer = val;
-    env->hflags &= ~(HF_LMA_MASK | HF_SVME_MASK);
-    if (env->efer & MSR_EFER_LMA) {
-        env->hflags |= HF_LMA_MASK;
-    }
-    if (env->efer & MSR_EFER_SVME) {
-        env->hflags |= HF_SVME_MASK;
-    }
-}
-
 static inline MemTxAttrs cpu_get_mem_attrs(CPUX86State *env)
 {
     return ((MemTxAttrs) { .secure = (env->hflags & HF_SMM_MASK) != 0 });
diff --git a/target/i386/helper.c b/target/i386/helper.c
index 618ad1c409..7304721a94 100644
--- a/target/i386/helper.c
+++ b/target/i386/helper.c
@@ -574,6 +574,19 @@ void do_cpu_sipi(X86CPU *cpu)
 #endif
 
 #ifndef CONFIG_USER_ONLY
+
+void cpu_load_efer(CPUX86State *env, uint64_t val)
+{
+    env->efer = val;
+    env->hflags &= ~(HF_LMA_MASK | HF_SVME_MASK);
+    if (env->efer & MSR_EFER_LMA) {
+        env->hflags |= HF_LMA_MASK;
+    }
+    if (env->efer & MSR_EFER_SVME) {
+        env->hflags |= HF_SVME_MASK;
+    }
+}
+
 uint8_t x86_ldub_phys(CPUState *cs, hwaddr addr)
 {
     X86CPU *cpu = X86_CPU(cs);
-- 
2.26.2



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (19 preceding siblings ...)
  2021-03-01  8:54 ` [PATCH v26 20/20] i386: make cpu_load_efer sysemu-only Claudio Fontana
@ 2021-03-01  9:18 ` no-reply
  2021-03-08 12:57 ` Claudio Fontana
  21 siblings, 0 replies; 40+ messages in thread
From: no-reply @ 2021-03-01  9:18 UTC (permalink / raw)
  To: cfontana
  Cc: lvivier, peter.maydell, thuth, ehabkost, philmd,
	richard.henderson, qemu-devel, r.bolshakov, cfontana, pbonzini,
	alex.bennee

Patchew URL: https://patchew.org/QEMU/20210301085450.1732-1-cfontana@suse.de/



Hi,

This series seems to have some coding style problems. See output below for
more information:

Type: series
Message-id: 20210301085450.1732-1-cfontana@suse.de
Subject: [PATCH v26 00/20] i386 cleanup PART 2

=== 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/20210301085450.1732-1-cfontana@suse.de -> patchew/20210301085450.1732-1-cfontana@suse.de
Switched to a new branch 'test'
00a5877 i386: make cpu_load_efer sysemu-only
bc37797 target/i386: gdbstub: only write CR0/CR2/CR3/EFER for sysemu
2b17816 target/i386: gdbstub: introduce aux functions to read/write CS64 regs
3a208cf i386: split off sysemu part of cpu.c
2c7e080 i386: split seg_helper into user-only and sysemu parts
d83014b i386: split svm_helper into sysemu and stub-only user
b6d69a4 i386: separate fpu_helper sysemu-only parts
93cb687 i386: split misc helper user stubs and sysemu part
d8352df i386: move TCG bpt_helper into sysemu/
8fbb05f i386: split tcg excp_helper into sysemu and user parts
0a91873 i386: split smm helper (sysemu)
86ad6ce i386: split off sysemu-only functionality in tcg-cpu
ecd60a2 meson: add target_user_arch
eec9c82 accel-cpu: make cpu_realizefn return a bool
007716a target/i386: fix host_cpu_adjust_phys_bits error handling
48530bb accel: introduce new accessor functions
cedc6f6 cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn
ffbf00d i386: split cpu accelerators from cpu.c, using AccelCPUClass
d8033c6 target/i386: Split out do_fsave, do_frstor, do_fxsave, do_fxrstor
d516237 target/i386: Rename helper_fldt, helper_fstt

=== OUTPUT BEGIN ===
1/20 Checking commit d5162371be89 (target/i386: Rename helper_fldt, helper_fstt)
2/20 Checking commit d8033c640ed4 (target/i386: Split out do_fsave, do_frstor, do_fxsave, do_fxrstor)
3/20 Checking commit ffbf00db2eb3 (i386: split cpu accelerators from cpu.c, using AccelCPUClass)
WARNING: line over 80 characters
#1337: FILE: target/i386/tcg/tcg-cpu.c:125:
+    memory_region_add_subregion_overlap(cpu->cpu_as_root, 0, cpu->cpu_as_mem, 0);

total: 0 errors, 1 warnings, 1256 lines checked

Patch 3/20 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
4/20 Checking commit cedc6f6e4c18 (cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn)
5/20 Checking commit 48530bbeb4c2 (accel: introduce new accessor functions)
6/20 Checking commit 007716a87edf (target/i386: fix host_cpu_adjust_phys_bits error handling)
7/20 Checking commit eec9c82ab85c (accel-cpu: make cpu_realizefn return a bool)
8/20 Checking commit ecd60a2189f5 (meson: add target_user_arch)
9/20 Checking commit 86ad6ce0eb6a (i386: split off sysemu-only functionality in tcg-cpu)
Use of uninitialized value $acpi_testexpected in string eq at ./scripts/checkpatch.pl line 1529.
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#31: 
new file mode 100644

WARNING: line over 80 characters
#116: FILE: target/i386/tcg/sysemu/tcg-cpu.c:72:
+    memory_region_add_subregion_overlap(cpu->cpu_as_root, 0, cpu->cpu_as_mem, 0);

total: 0 errors, 2 warnings, 212 lines checked

Patch 9/20 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
10/20 Checking commit 0a9187311234 (i386: split smm helper (sysemu))
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#24: 
 target/i386/tcg/{ => sysemu}/smm_helper.c | 19 ++-----------------

total: 0 errors, 1 warnings, 84 lines checked

Patch 10/20 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
11/20 Checking commit 8fbb05f77818 (i386: split tcg excp_helper into sysemu and user parts)
Use of uninitialized value $acpi_testexpected in string eq at ./scripts/checkpatch.pl line 1529.
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#599: 
new file mode 100644

WARNING: Block comments use a leading /* on a separate line
#756: FILE: target/i386/tcg/sysemu/excp_helper.c:153:
+            /* Bits 20-13 provide bits 39-32 of the address, bit 21 is reserved.

WARNING: Block comments use a leading /* on a separate line
#830: FILE: target/i386/tcg/sysemu/excp_helper.c:227:
+/* return value:

WARNING: line over 80 characters
#931: FILE: target/i386/tcg/sysemu/excp_helper.c:328:
+            pdpe_addr = ((pml4e & PG_ADDRESS_MASK) + (((addr >> 30) & 0x1ff) << 3)) &

WARNING: Block comments use a leading /* on a separate line
#1025: FILE: target/i386/tcg/sysemu/excp_helper.c:422:
+            /* Bits 20-13 provide bits 39-32 of the address, bit 21 is reserved.

WARNING: Block comments use a leading /* on a separate line
#1120: FILE: target/i386/tcg/sysemu/excp_helper.c:517:
+        /* only set write access if already dirty... otherwise wait

WARNING: Block comments use * on subsequent lines
#1121: FILE: target/i386/tcg/sysemu/excp_helper.c:518:
+        /* only set write access if already dirty... otherwise wait
+           for dirty access */

WARNING: Block comments use a trailing */ on a separate line
#1121: FILE: target/i386/tcg/sysemu/excp_helper.c:518:
+           for dirty access */

WARNING: Block comments use a leading /* on a separate line
#1134: FILE: target/i386/tcg/sysemu/excp_helper.c:531:
+    /* Even if 4MB pages, we map only one 4KB page in the cache to

WARNING: Block comments use * on subsequent lines
#1135: FILE: target/i386/tcg/sysemu/excp_helper.c:532:
+    /* Even if 4MB pages, we map only one 4KB page in the cache to
+       avoid filling it too fast */

WARNING: Block comments use a trailing */ on a separate line
#1135: FILE: target/i386/tcg/sysemu/excp_helper.c:532:
+       avoid filling it too fast */

ERROR: braces {} are necessary for all arms of this statement
#1149: FILE: target/i386/tcg/sysemu/excp_helper.c:546:
+    if (is_user)
[...]

total: 1 errors, 11 warnings, 631 lines checked

Patch 11/20 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.

12/20 Checking commit d8352df00ca5 (i386: move TCG bpt_helper into sysemu/)
Use of uninitialized value $acpi_testexpected in string eq at ./scripts/checkpatch.pl line 1529.
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#364: 
new file mode 100644

WARNING: Block comments use a leading /* on a separate line
#500: FILE: target/i386/tcg/sysemu/bpt_helper.c:132:
+    /* If nothing is changing except the global/local enable bits,

WARNING: Block comments use * on subsequent lines
#501: FILE: target/i386/tcg/sysemu/bpt_helper.c:133:
+    /* If nothing is changing except the global/local enable bits,
+       then we can make the change more efficient.  */

WARNING: Block comments use a trailing */ on a separate line
#501: FILE: target/i386/tcg/sysemu/bpt_helper.c:133:
+       then we can make the change more efficient.  */

WARNING: Block comments use a leading /* on a separate line
#503: FILE: target/i386/tcg/sysemu/bpt_helper.c:135:
+        /* Fold the global and local enable bits together into the

WARNING: Block comments use * on subsequent lines
#504: FILE: target/i386/tcg/sysemu/bpt_helper.c:136:
+        /* Fold the global and local enable bits together into the
+           global fields, then xor to show which registers have

WARNING: Block comments use a trailing */ on a separate line
#505: FILE: target/i386/tcg/sysemu/bpt_helper.c:137:
+           changed collective enable state.  */

total: 0 errors, 7 warnings, 628 lines checked

Patch 12/20 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
13/20 Checking commit 93cb687b7b8c (i386: split misc helper user stubs and sysemu part)
Use of uninitialized value $acpi_testexpected in string eq at ./scripts/checkpatch.pl line 1529.
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#527: 
new file mode 100644

ERROR: switch and case should be at the same indent
#820: FILE: target/i386/tcg/sysemu/misc_helper.c:289:
+    switch ((uint32_t)env->regs[R_ECX]) {
[...]
+     case MSR_IA32_UCODE_REV:

total: 1 errors, 1 warnings, 1012 lines checked

Patch 13/20 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.

14/20 Checking commit b6d69a4d70a7 (i386: separate fpu_helper sysemu-only parts)
Use of uninitialized value $acpi_testexpected in string eq at ./scripts/checkpatch.pl line 1529.
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#108: 
new file mode 100644

total: 0 errors, 1 warnings, 135 lines checked

Patch 14/20 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
15/20 Checking commit d83014bb1c38 (i386: split svm_helper into sysemu and stub-only user)
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#34: 
 target/i386/tcg/{ => sysemu}/svm_helper.c | 62 +------------------------

total: 0 errors, 1 warnings, 169 lines checked

Patch 15/20 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
16/20 Checking commit 2c7e080c0171 (i386: split seg_helper into user-only and sysemu parts)
Use of uninitialized value $acpi_testexpected in string eq at ./scripts/checkpatch.pl line 1529.
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#321: 
new file mode 100644

ERROR: do not use C99 // comments
#348: FILE: target/i386/tcg/seg_helper.h:23:
+//#define DEBUG_PCALL

WARNING: Block comments use a leading /* on a separate line
#621: FILE: target/i386/tcg/user/seg_helper.c:73:
+    /* Since we emulate only user space, we cannot do more than

WARNING: Block comments use * on subsequent lines
#622: FILE: target/i386/tcg/user/seg_helper.c:74:
+    /* Since we emulate only user space, we cannot do more than
+       exiting the emulation with the suitable exception and error

WARNING: Block comments use a trailing */ on a separate line
#623: FILE: target/i386/tcg/user/seg_helper.c:75:
+       code. So update EIP for INT 0x80 and EXCP_SYSCALL. */

WARNING: Block comments use a leading /* on a separate line
#634: FILE: target/i386/tcg/user/seg_helper.c:86:
+    /* if user mode only, we simulate a fake exception

WARNING: Block comments use * on subsequent lines
#635: FILE: target/i386/tcg/user/seg_helper.c:87:
+    /* if user mode only, we simulate a fake exception
+       which will be handled outside the cpu execution

WARNING: Block comments use a trailing */ on a separate line
#636: FILE: target/i386/tcg/user/seg_helper.c:88:
+       loop */

total: 1 errors, 7 warnings, 595 lines checked

Patch 16/20 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.

17/20 Checking commit 3a208cf87826 (i386: split off sysemu part of cpu.c)
Use of uninitialized value $acpi_testexpected in string eq at ./scripts/checkpatch.pl line 1529.
WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
#18: 
new file mode 100644

WARNING: Block comments use a leading /* on a separate line
#52: FILE: target/i386/cpu-internal.h:30:
+    /* feature flags names are taken from "Intel Processor Identification and

WARNING: Block comments use a leading /* on a separate line
#135: FILE: target/i386/cpu-sysemu.c:37:
+/* Return a QDict containing keys for all properties that can be included

WARNING: Block comments use a leading /* on a separate line
#188: FILE: target/i386/cpu-sysemu.c:90:
+/* Convert CPU model data from X86CPU object to a property dictionary

WARNING: Block comments use a leading /* on a separate line
#202: FILE: target/i386/cpu-sysemu.c:104:
+/* Convert CPU model data from X86CPU object to a property dictionary

WARNING: Block comments use a leading /* on a separate line
#218: FILE: target/i386/cpu-sysemu.c:120:
+        /* "hotplugged" is the only property that is configurable

WARNING: Block comments use a leading /* on a separate line
#308: FILE: target/i386/cpu-sysemu.c:210:
+        /* As we don't return every single property, full expansion needs

total: 0 errors, 7 warnings, 901 lines checked

Patch 17/20 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
18/20 Checking commit 2b1781677162 (target/i386: gdbstub: introduce aux functions to read/write CS64 regs)
WARNING: line over 80 characters
#62: FILE: target/i386/gdbstub.c:163:
+            return gdb_read_reg_cs64(env->hflags, mem_buf, env->segs[R_FS].base);

WARNING: line over 80 characters
#68: FILE: target/i386/gdbstub.c:165:
+            return gdb_read_reg_cs64(env->hflags, mem_buf, env->segs[R_GS].base);

WARNING: line over 80 characters
#156: FILE: target/i386/gdbstub.c:318:
+            return gdb_write_reg_cs64(env->hflags, mem_buf, &env->segs[R_FS].base);

WARNING: line over 80 characters
#165: FILE: target/i386/gdbstub.c:320:
+            return gdb_write_reg_cs64(env->hflags, mem_buf, &env->segs[R_GS].base);

total: 0 errors, 4 warnings, 221 lines checked

Patch 18/20 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
19/20 Checking commit bc377979cd9a (target/i386: gdbstub: only write CR0/CR2/CR3/EFER for sysemu)
20/20 Checking commit 00a5877fc1a3 (i386: make cpu_load_efer sysemu-only)
=== OUTPUT END ===

Test command exited with code: 1


The full log is available at
http://patchew.org/logs/20210301085450.1732-1-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] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
                   ` (20 preceding siblings ...)
  2021-03-01  9:18 ` [PATCH v26 00/20] i386 cleanup PART 2 no-reply
@ 2021-03-08 12:57 ` Claudio Fontana
  2021-03-08 13:27   ` Philippe Mathieu-Daudé
  2021-03-08 14:02   ` Alex Bennée
  21 siblings, 2 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-08 12:57 UTC (permalink / raw)
  To: Paolo Bonzini, Richard Henderson, Philippe Mathieu-Daudé,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, qemu-devel

Hi,

anything else for me to do here?

The latest rebased state of this series should be always available here:

https://gitlab.com/hw-claudio/qemu/-/tree/i386_cleanup_8

When it comes to the ARM cleanup series,
I would like to have the tests pass for ARM, before doing even more changes, could you help me there Philippe?

Maybe applying some of your changes on top would fix the failures? I tried, for example with the arm-cpu-features ones, but it didn't work for me..

Thanks,

Claudio


On 3/1/21 9:54 AM, Claudio Fontana wrote:
> v25 -> v26:
> * i386: separate fpu_helper into user and sysemu parts
>  - removed the splitting of the user mode into their own user/fpu_helper.c,
>    seems not worth it.
> 
> v24 -> v25:
> 
> * i386: separate fpu_helper into user and sysemu parts
> 
>  - added 2 preliminary patches to the series (from Richard)
>  - rebased on those
> 
> * i386: move TCG btp_helper into sysemu/
> 
>  - fixed title typo (Alex)
>  - nested #ifdef more easily (Richard)
> 
> v23 -> v24:
> 
> * i386: gdbstub: only write CR0/CR2/CR3/EFER for sysemu                                                                                     
>  - remove additional #ifdef TARGET_X86_64
>  - split in two patches, so it is easier to understand.
> 
> v22 -> v23:
> 
> * i386: move TCG btp_helper into sysemu/
>  - extended the #ifndef CONFIG_USER_ONLY to entire else of
>    if (cpl != 0).
> 
> * i386: split misc helper into user and sysemu parts
>  - added g_assert_not_reached() and changed user file name to -stubs.
> 
> * i386: separate fpu_helper into user and sysemu parts
>  - removed unused return value
>  - added comment abut issues with current cpu_x86_fsave.
> 
> * i386: split off sysemu part of cpu.c
>  - rename cpu-softmmu.c to cpu-sysemu.c
>  - fixed two mispelled comments, and add two comments
>    in the headers of cpu.c and cpu-sysemu.c to describe them
> 
> * i386: gdbstub: only write CR0/CR2/CR3/EFER for sysemu
>  - defined some aux functions to reduce repeated code
> 
> * i386: make cpu_load_efer sysemu-only
>  - move the function to helper.c, remove "inline"
> 
> v21 -> v22: replace "softmmu" with "sysemu"
> 
> v20 -> v21:
> 
> * meson: add target_user_arch
>   - add hexagon
> 
> v19 -> v20:
> 
> * add new patch to make gdbstub only write certain registers for softmmu.
>   In particular, CR0, CR2, CR3 and EFER should not be changed under
>   CONFIG_USER_ONLY. (Paolo)
> 
> * add new patch to make cpu_load_efer softmmu-only (Paolo)
> 
> * i386: split svm_helper into softmmu and stub-only user
> 
>   - fixed commit message spelling (Eric)
> 
>   - mention in commit message that this reproduces the existing stubs,
>     but really everything that requires s->cpl == 0 should be impossible
>     to trigger from user-mode, and we could find a way to assert that
>     more consistently.
> 
> v18 -> v19:
> 
> * i386: split smm helper (softmmu)
>   - add g_assert_not_reached and cpu_abort for invalid states in
>     CONFIG_USER_ONLY (Paolo)
> 
> * i386: move TCG btp_helper into softmmu/
>   - for CONFIG_USER_ONLY, assert that the hidden IOBPT flags are not set
>     while attempting to generate io_bpt helpers.
>     Theory to verify (Paolo)
> 
> * i386: slit svm_helper into softmmu and stub-only user
>   - added XXX in the commit message to highlight the question about
>     whether the same check should be done controlling access to
>     cpu_load_efer() and state of the hidden SVME flag. (Paolo)
> 
> v17 -> v18:
> 
> * meson: add target_user_arch
> 
>  - add target_user_arch to all targets which build user.
>    Otherwise meson complains about missing key for archs without it.
>    (Paolo)
> 
> * wrap a few gen_helper_ calls around ifndef CONFIG_USER_ONLY.
>   This would need a look from someone like Alex or Richard I think,
>   as potentially we could remove even more code I think around the
>   gen_helper_ calls for CONFIG_USER_ONLY.
> 
>   In the current master code, we have empty helpers for user mode,
>   but still we generate the preamble code, temporary variables etc,
>   just to then call a helper_() function that does nothing.
> 
>   In particular I am referring to patches:
> 
>   i386: split tcg btp_helper into softmmu and user parts
>         DEF_HELPER_FLAGS_3(set_dr, TCG_CALL_NO_WG, void, env, int, tl)
>         DEF_HELPER_FLAGS_4(bpt_io, TCG_CALL_NO_WG, void, env, i32, i32, tl)
>         gen_bpt_io
>         gen_helper_set_dr(cpu_env, s->tmp2_i32, s->T0);
> 
>   i386: split smm helper (softmmu)
>         DEF_HELPER_1(rsm, void, env)
> 	gen_helper_rsm(cpu_env);
> 
>   (Alex, Richard?)
> 
> * removed suffixes from user/ and softmmu/ modules
>   (Alex, Philippe).
>   Where possible, removed user stubs entirely.
>   Renamed the leftover svm_helper stubs to user/svm_stubs.c
> 
> * cleaned up lefover unnecessary header files and squashed them.
>  
> 
> v16 -> v17: changed to RFC
> 
> * tcg_ops are already in master, removed from the series
> 
> * i386: split cpu accelerators from cpu.c, using AccelCPUClass:
>   removed spurious ; and added spacing before/after functions (Richard)
> 
> * added new patches as RFC for the next steps, introducing target-specific
>   user-mode specific meson variables, and applied to i386/tcg as an
>   example, in order to gather feedback.
> 
> v15 -> v16:
> 
> * cpu: Move synchronize_from_tb() to tcg_ops:
>   - adjusted comments (Alex)
> 
> * cpu: tcg_ops: move to tcg-cpu-ops.h, keep a pointer in CPUClass:
>   - remove forward decl. of AccelCPUClass, should be in a later patch. (Alex)
>   - simplified comment about tcg_ops in struct CPUClass (Alex)
>   - remove obsolete comment about ARM blocking TCGCPUOps from being const.
>     (Alex)
> 
> * accel: replace struct CpusAccel with AccelOpsClass:
>   - reworded commit message to be clearer about the objective (Alex)
> 
> * accel: introduce AccelCPUClass extending CPUClass
>   - reworded commit message to be clearer about the objective (Alex)
> 
> * hw/core/cpu: call qemu_init_vcpu in cpu_common_realizefn:
>   - dropped this patch (Alex, Philippe)
> 
>   will try again later, also in the context of:
>   https://www.mail-archive.com/qemu-devel@nongnu.org/msg686480.html
> 
> * accel: introduce new accessor functions
>   - squashed comments in previous patch introducing accel-cpu.h. (Philippe)
> 
> * accel-cpu: make cpu_realizefn return a bool
>   - split in two patches, separating the change to the phys_bits check
>     (Philippe)
> 
> v14 -> v15:
> 
> * change the TcgCpuOperations so that all fields of the struct are
>   defined unconditionally, as per original patch by Eduardo,
>   before moving them to a separate struct referenced by a pointer
>   (Richard, Eduardo).
> 
> * changed (c) year to 2021
> 
> * added a patch to make accel_cpu->cpu_realizefn return bool, and
>   adapt host_cpu, kvm_cpu, hvf_cpu and tcg_cpu in i386 accordingly.
>   Ultimately, consistently moving to a pattern of realize functions
>   returning bool will require a rework of all devices.
> 
> v13 -> v14: rebased on latest master.
> v12 -> v13: rebased on latest master.
> 
> v11 -> v12: reordered patches and improved tcg_ops
> 
> * reordered all TcgCpuOperations stuff so it is at the beginning
> 
> * added patches for ARM-specific tcg ops
>   debug_check_watchpoint and adjust_watchpoint_address
> 
> * added a patch that puts a forward declared pointer in the struct,
>   so as to reduce the change of misuse between common_ss and specific_ss code,
>   and tidy up as a consequence all targets, by defining dedicated structs.
> 
> v10 -> v11: split off PART 2,
> 
> no further changes to PART 2 other than the split.
> 
> v9 -> v10: minor tweaks and fixes
> 
> * in "i386: split cpu accelerators from cpu.c",
> 
> use kvm/kvm-cpu.c, hvf/hvf-cpu.c, tcg/tcg-cpu.c.
> Easier to understand compared to editing multiple cpu.c files,
> and matches the header files if needed (kvm-cpu.h).
> 
> * in "accel: replace struct CpusAccel with AccelOpsClass",
> 
> make it a bit more consistent, by naming the files defining
> the AccelOpsClass types "...-accel-ops.c" instead of the old
> naming "...-cpus.c".
> 
> * in "cpu: move cc->transaction_failed to tcg_ops",
> 
> protect with CONFIG_TCG the use of tcg_ops for hw/misc/jazz.c,
> 
>  #include "exec/memattrs.h" (Philippe, Eduardo)
> 
> * in "cpu: Move synchronize_from_tb() to tcg_ops",
> 
>  #include "hw/core/cpu.h" (Philippe, Eduardo)
> 
> do not remove the comment about struct TcgCpuOperations (Philippe)
> 
> * in "accel/tcg: split TCG-only code from cpu_exec_realizefn",
> 
> invert tcg_target_initialized set order (Alex)
> 
> * in "i386: move TCG cpu class initialization out of helper.c",
> 
> extract helper-tcg.h, tcg-cpu.c, and tcg-cpu.h directly into
> tcg/, avoiding the extra move later to tcg/ (Alex)
> 
> 
> 
> 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 (18):
>   i386: split cpu accelerators from cpu.c, using AccelCPUClass
>   cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn
>   accel: introduce new accessor functions
>   target/i386: fix host_cpu_adjust_phys_bits error handling
>   accel-cpu: make cpu_realizefn return a bool
>   meson: add target_user_arch
>   i386: split off sysemu-only functionality in tcg-cpu
>   i386: split smm helper (sysemu)
>   i386: split tcg excp_helper into sysemu and user parts
>   i386: move TCG bpt_helper into sysemu/
>   i386: split misc helper user stubs and sysemu part
>   i386: separate fpu_helper sysemu-only parts
>   i386: split svm_helper into sysemu and stub-only user
>   i386: split seg_helper into user-only and sysemu parts
>   i386: split off sysemu part of cpu.c
>   target/i386: gdbstub: introduce aux functions to read/write CS64 regs
>   target/i386: gdbstub: only write CR0/CR2/CR3/EFER for sysemu
>   i386: make cpu_load_efer sysemu-only
> 
> Richard Henderson (2):
>   target/i386: Rename helper_fldt, helper_fstt
>   target/i386: Split out do_fsave, do_frstor, do_fxsave, do_fxrstor
> 
>  meson.build                               |   5 +
>  include/hw/core/accel-cpu.h               |   2 +-
>  include/qemu/accel.h                      |  13 +
>  target/i386/cpu-internal.h                |  70 ++
>  target/i386/cpu.h                         |  43 +-
>  target/i386/helper.h                      |  11 +
>  target/i386/host-cpu.h                    |  19 +
>  target/i386/kvm/kvm-cpu.h                 |  41 ++
>  target/i386/tcg/helper-tcg.h              |   8 +
>  target/i386/tcg/seg_helper.h              |  66 ++
>  target/i386/tcg/tcg-cpu.h                 |  21 +-
>  accel/accel-common.c                      |  19 +
>  cpu.c                                     |   5 +-
>  hw/i386/pc_piix.c                         |   1 +
>  target/i386/cpu-sysemu.c                  | 352 ++++++++++
>  target/i386/cpu.c                         | 782 ++--------------------
>  target/i386/gdbstub.c                     | 165 ++---
>  target/i386/helper.c                      |  13 +
>  target/i386/host-cpu.c                    | 204 ++++++
>  target/i386/hvf/hvf-cpu.c                 |  68 ++
>  target/i386/kvm/kvm-cpu.c                 | 151 +++++
>  target/i386/kvm/kvm.c                     |   3 +-
>  target/i386/tcg/bpt_helper.c              | 276 --------
>  target/i386/tcg/excp_helper.c             | 572 ----------------
>  target/i386/tcg/fpu_helper.c              | 122 ++--
>  target/i386/tcg/misc_helper.c             | 463 -------------
>  target/i386/tcg/seg_helper.c              | 237 +------
>  target/i386/tcg/sysemu/bpt_helper.c       | 293 ++++++++
>  target/i386/tcg/sysemu/excp_helper.c      | 581 ++++++++++++++++
>  target/i386/tcg/sysemu/fpu_helper.c       |  57 ++
>  target/i386/tcg/sysemu/misc_helper.c      | 438 ++++++++++++
>  target/i386/tcg/sysemu/seg_helper.c       | 125 ++++
>  target/i386/tcg/{ => sysemu}/smm_helper.c |  19 +-
>  target/i386/tcg/{ => sysemu}/svm_helper.c |  62 +-
>  target/i386/tcg/sysemu/tcg-cpu.c          |  83 +++
>  target/i386/tcg/tcg-cpu.c                 |  50 +-
>  target/i386/tcg/translate.c               |  13 +-
>  target/i386/tcg/user/excp_helper.c        |  39 ++
>  target/i386/tcg/user/misc_stubs.c         |  75 +++
>  target/i386/tcg/user/seg_helper.c         | 109 +++
>  target/i386/tcg/user/svm_stubs.c          |  76 +++
>  MAINTAINERS                               |   2 +-
>  target/alpha/meson.build                  |   3 +
>  target/arm/meson.build                    |   2 +
>  target/cris/meson.build                   |   3 +
>  target/hexagon/meson.build                |   3 +
>  target/hppa/meson.build                   |   3 +
>  target/i386/hvf/meson.build               |   1 +
>  target/i386/kvm/meson.build               |   7 +-
>  target/i386/meson.build                   |   9 +-
>  target/i386/tcg/meson.build               |   5 +-
>  target/i386/tcg/sysemu/meson.build        |  10 +
>  target/i386/tcg/user/meson.build          |   6 +
>  target/m68k/meson.build                   |   3 +
>  target/microblaze/meson.build             |   3 +
>  target/mips/meson.build                   |   3 +
>  target/nios2/meson.build                  |   3 +
>  target/openrisc/meson.build               |   3 +
>  target/ppc/meson.build                    |   3 +
>  target/riscv/meson.build                  |   3 +
>  target/s390x/meson.build                  |   3 +
>  target/sh4/meson.build                    |   3 +
>  target/sparc/meson.build                  |   3 +
>  target/tilegx/meson.build                 |   3 +
>  target/tricore/meson.build                |   3 +
>  target/xtensa/meson.build                 |   3 +
>  66 files changed, 3266 insertions(+), 2579 deletions(-)
>  create mode 100644 target/i386/cpu-internal.h
>  create mode 100644 target/i386/host-cpu.h
>  create mode 100644 target/i386/kvm/kvm-cpu.h
>  create mode 100644 target/i386/tcg/seg_helper.h
>  create mode 100644 target/i386/cpu-sysemu.c
>  create mode 100644 target/i386/host-cpu.c
>  create mode 100644 target/i386/hvf/hvf-cpu.c
>  create mode 100644 target/i386/kvm/kvm-cpu.c
>  create mode 100644 target/i386/tcg/sysemu/bpt_helper.c
>  create mode 100644 target/i386/tcg/sysemu/excp_helper.c
>  create mode 100644 target/i386/tcg/sysemu/fpu_helper.c
>  create mode 100644 target/i386/tcg/sysemu/misc_helper.c
>  create mode 100644 target/i386/tcg/sysemu/seg_helper.c
>  rename target/i386/tcg/{ => sysemu}/smm_helper.c (98%)
>  rename target/i386/tcg/{ => sysemu}/svm_helper.c (96%)
>  create mode 100644 target/i386/tcg/sysemu/tcg-cpu.c
>  create mode 100644 target/i386/tcg/user/excp_helper.c
>  create mode 100644 target/i386/tcg/user/misc_stubs.c
>  create mode 100644 target/i386/tcg/user/seg_helper.c
>  create mode 100644 target/i386/tcg/user/svm_stubs.c
>  create mode 100644 target/i386/tcg/sysemu/meson.build
>  create mode 100644 target/i386/tcg/user/meson.build
> 



^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-03-08 12:57 ` Claudio Fontana
@ 2021-03-08 13:27   ` Philippe Mathieu-Daudé
  2021-03-08 13:52     ` Claudio Fontana
  2021-03-08 14:02   ` Alex Bennée
  1 sibling, 1 reply; 40+ messages in thread
From: Philippe Mathieu-Daudé @ 2021-03-08 13:27 UTC (permalink / raw)
  To: Claudio Fontana, Paolo Bonzini, Richard Henderson,
	Eduardo Habkost, Peter Maydell, Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, qemu-devel

Hi Claudio,

On 3/8/21 1:57 PM, Claudio Fontana wrote:
> Hi,
> 
> anything else for me to do here?
> 
> The latest rebased state of this series should be always available here:
> 
> https://gitlab.com/hw-claudio/qemu/-/tree/i386_cleanup_8
> 
> When it comes to the ARM cleanup series,
> I would like to have the tests pass for ARM, before doing even more changes, could you help me there Philippe?
> 
> Maybe applying some of your changes on top would fix the failures? I tried, for example with the arm-cpu-features ones, but it didn't work for me..

TBH I wrote these patches during my personal spare time and this
became a real Pandora box that drained too much energy. I prefer
to step back and focus on finishing smaller tasks before burning
out. That said I appreciate your effort and am interested in
following / reviewing your work.

Regards,

Phil.


^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-03-08 13:27   ` Philippe Mathieu-Daudé
@ 2021-03-08 13:52     ` Claudio Fontana
  2021-03-08 14:57       ` Philippe Mathieu-Daudé
  0 siblings, 1 reply; 40+ messages in thread
From: Claudio Fontana @ 2021-03-08 13:52 UTC (permalink / raw)
  To: Philippe Mathieu-Daudé,
	Paolo Bonzini, Richard Henderson, Eduardo Habkost, Peter Maydell,
	Alex Bennée
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, qemu-devel

On 3/8/21 2:27 PM, Philippe Mathieu-Daudé wrote:
> Hi Claudio,
> 
> On 3/8/21 1:57 PM, Claudio Fontana wrote:
>> Hi,
>>
>> anything else for me to do here?
>>
>> The latest rebased state of this series should be always available here:
>>
>> https://gitlab.com/hw-claudio/qemu/-/tree/i386_cleanup_8
>>
>> When it comes to the ARM cleanup series,
>> I would like to have the tests pass for ARM, before doing even more changes, could you help me there Philippe?
>>
>> Maybe applying some of your changes on top would fix the failures? I tried, for example with the arm-cpu-features ones, but it didn't work for me..
> 
> TBH I wrote these patches during my personal spare time and this
> became a real Pandora box that drained too much energy. I prefer
> to step back and focus on finishing smaller tasks before burning
> out. That said I appreciate your effort and am interested in
> following / reviewing your work.
> 
> Regards,
> 
> Phil.
> 

Thanks Philippe for sharing this, and I agree completely, it is very draining.

The effort of making tests happy that run in artificial environments in particular often feels to me
as too disconnected from actually ensuring that there is no real run time regression.

qtest_enabled() (implicitly, or explicitly via open-ended else statements) is another painful variable to keep in mind in cpu and machine code, so it is not helpful in my view.

I'll try to push more to get the tests running again, if you have any comment or idea, feel free to just point me in the right direction,
that is very valuable to me, even without working code.

Currently I am struggling with arm-cpu-features and other tests (device-introspect-test, qom-test, test-hmp).

I'll publish arm_cleanup_v5 soon,

Thanks a lot,

Claudio


^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-03-08 12:57 ` Claudio Fontana
  2021-03-08 13:27   ` Philippe Mathieu-Daudé
@ 2021-03-08 14:02   ` Alex Bennée
  2021-05-05  8:27     ` Claudio Fontana
  1 sibling, 1 reply; 40+ messages in thread
From: Alex Bennée @ 2021-03-08 14:02 UTC (permalink / raw)
  To: Claudio Fontana
  Cc: Laurent Vivier, Peter Maydell, Thomas Huth, Eduardo Habkost,
	Richard Henderson, qemu-devel, Roman Bolshakov, Paolo Bonzini,
	Philippe Mathieu-Daudé


Claudio Fontana <cfontana@suse.de> writes:

> Hi,
>
> anything else for me to do here?

It looks to me that this series is looking pretty good. Every patch has
at least one review so I think it's just waiting on the maintainers to
pick it up.

Paolo/Richard - are you intending to take the series as is or are you
waiting for something else? I'd like to see the patch delta reduced for
the ARM cleanup work which is still ongoing.

>
> The latest rebased state of this series should be always available here:
>
> https://gitlab.com/hw-claudio/qemu/-/tree/i386_cleanup_8
>
> When it comes to the ARM cleanup series,
> I would like to have the tests pass for ARM, before doing even more changes, could you help me there Philippe?
>
> Maybe applying some of your changes on top would fix the failures? I tried, for example with the arm-cpu-features ones, but it didn't work for me..
>
<snip>

-- 
Alex Bennée


^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-03-08 13:52     ` Claudio Fontana
@ 2021-03-08 14:57       ` Philippe Mathieu-Daudé
  2021-03-08 15:03         ` Claudio Fontana
  2021-03-08 15:07         ` Daniel P. Berrangé
  0 siblings, 2 replies; 40+ messages in thread
From: Philippe Mathieu-Daudé @ 2021-03-08 14:57 UTC (permalink / raw)
  To: Claudio Fontana, Paolo Bonzini, Richard Henderson,
	Eduardo Habkost, Peter Maydell, Alex Bennée,
	Daniel P . Berrange
  Cc: Laurent Vivier, Thomas Huth, Roman Bolshakov, qemu-devel

On 3/8/21 2:52 PM, Claudio Fontana wrote:
> On 3/8/21 2:27 PM, Philippe Mathieu-Daudé wrote:
>> Hi Claudio,
>>
>> On 3/8/21 1:57 PM, Claudio Fontana wrote:
>>> Hi,
>>>
>>> anything else for me to do here?
>>>
>>> The latest rebased state of this series should be always available here:
>>>
>>> https://gitlab.com/hw-claudio/qemu/-/tree/i386_cleanup_8
>>>
>>> When it comes to the ARM cleanup series,
>>> I would like to have the tests pass for ARM, before doing even more changes, could you help me there Philippe?
>>>
>>> Maybe applying some of your changes on top would fix the failures? I tried, for example with the arm-cpu-features ones, but it didn't work for me..
>>
>> TBH I wrote these patches during my personal spare time and this
>> became a real Pandora box that drained too much energy. I prefer
>> to step back and focus on finishing smaller tasks before burning
>> out. That said I appreciate your effort and am interested in
>> following / reviewing your work.
>>
>> Regards,
>>
>> Phil.
>>
> 
> Thanks Philippe for sharing this, and I agree completely, it is very draining.
> 
> The effort of making tests happy that run in artificial environments in particular often feels to me
> as too disconnected from actually ensuring that there is no real run time regression.
> 
> qtest_enabled() (implicitly, or explicitly via open-ended else statements) is another painful variable to keep in mind in cpu and machine code, so it is not helpful in my view.
> 
> I'll try to push more to get the tests running again, if you have any comment or idea, feel free to just point me in the right direction,
> that is very valuable to me, even without working code.

Basically I gave up after realizing from Daniel reviews that we need
QMP commands to query QEMU at runtime its built-in features, so we
have build-agnostic tests easier to maintain. I agree this is the
best way to resolve this particular case, but also scale for various
other cases.


^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-03-08 14:57       ` Philippe Mathieu-Daudé
@ 2021-03-08 15:03         ` Claudio Fontana
  2021-03-08 15:07         ` Daniel P. Berrangé
  1 sibling, 0 replies; 40+ messages in thread
From: Claudio Fontana @ 2021-03-08 15:03 UTC (permalink / raw)
  To: Philippe Mathieu-Daudé,
	Paolo Bonzini, Richard Henderson, Eduardo Habkost, Peter Maydell,
	Alex Bennée, Daniel P . Berrange
  Cc: Laurent Vivier, Thomas Huth, Jim Fehlig, Roman Bolshakov, qemu-devel

On 3/8/21 3:57 PM, Philippe Mathieu-Daudé wrote:
> On 3/8/21 2:52 PM, Claudio Fontana wrote:
>> On 3/8/21 2:27 PM, Philippe Mathieu-Daudé wrote:
>>> Hi Claudio,
>>>
>>> On 3/8/21 1:57 PM, Claudio Fontana wrote:
>>>> Hi,
>>>>
>>>> anything else for me to do here?
>>>>
>>>> The latest rebased state of this series should be always available here:
>>>>
>>>> https://gitlab.com/hw-claudio/qemu/-/tree/i386_cleanup_8
>>>>
>>>> When it comes to the ARM cleanup series,
>>>> I would like to have the tests pass for ARM, before doing even more changes, could you help me there Philippe?
>>>>
>>>> Maybe applying some of your changes on top would fix the failures? I tried, for example with the arm-cpu-features ones, but it didn't work for me..
>>>
>>> TBH I wrote these patches during my personal spare time and this
>>> became a real Pandora box that drained too much energy. I prefer
>>> to step back and focus on finishing smaller tasks before burning
>>> out. That said I appreciate your effort and am interested in
>>> following / reviewing your work.
>>>
>>> Regards,
>>>
>>> Phil.
>>>
>>
>> Thanks Philippe for sharing this, and I agree completely, it is very draining.
>>
>> The effort of making tests happy that run in artificial environments in particular often feels to me
>> as too disconnected from actually ensuring that there is no real run time regression.
>>
>> qtest_enabled() (implicitly, or explicitly via open-ended else statements) is another painful variable to keep in mind in cpu and machine code, so it is not helpful in my view.
>>
>> I'll try to push more to get the tests running again, if you have any comment or idea, feel free to just point me in the right direction,
>> that is very valuable to me, even without working code.
> 
> Basically I gave up after realizing from Daniel reviews that we need
> QMP commands to query QEMU at runtime its built-in features, so we
> have build-agnostic tests easier to maintain. I agree this is the
> best way to resolve this particular case, but also scale for various
> other cases.
> 

Well, yes, but in order to get things to work, even just a kvm-build fix would be good until we have the perfect solution, no?

We also fixed the tcg tests when doing this for i386, so I think we can fix these issues for arm too.

But this doesn't mean that we don't need the QMP commands to query QEMU at runtime for its "built-in"/module-loaded features.
We need that too, and I suspect this will be more and more needed by libvirt, as QEMU modularizes.

I just think the two things could proceed in parallel..

Ciao,

Claudio






^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-03-08 14:57       ` Philippe Mathieu-Daudé
  2021-03-08 15:03         ` Claudio Fontana
@ 2021-03-08 15:07         ` Daniel P. Berrangé
  1 sibling, 0 replies; 40+ messages in thread
From: Daniel P. Berrangé @ 2021-03-08 15:07 UTC (permalink / raw)
  To: Philippe Mathieu-Daudé
  Cc: Laurent Vivier, Peter Maydell, Thomas Huth, Eduardo Habkost,
	Richard Henderson, qemu-devel, Roman Bolshakov, Claudio Fontana,
	Paolo Bonzini, Alex Bennée

On Mon, Mar 08, 2021 at 03:57:29PM +0100, Philippe Mathieu-Daudé wrote:
> On 3/8/21 2:52 PM, Claudio Fontana wrote:
> > On 3/8/21 2:27 PM, Philippe Mathieu-Daudé wrote:
> >> Hi Claudio,
> >>
> >> On 3/8/21 1:57 PM, Claudio Fontana wrote:
> >>> Hi,
> >>>
> >>> anything else for me to do here?
> >>>
> >>> The latest rebased state of this series should be always available here:
> >>>
> >>> https://gitlab.com/hw-claudio/qemu/-/tree/i386_cleanup_8
> >>>
> >>> When it comes to the ARM cleanup series,
> >>> I would like to have the tests pass for ARM, before doing even more changes, could you help me there Philippe?
> >>>
> >>> Maybe applying some of your changes on top would fix the failures? I tried, for example with the arm-cpu-features ones, but it didn't work for me..
> >>
> >> TBH I wrote these patches during my personal spare time and this
> >> became a real Pandora box that drained too much energy. I prefer
> >> to step back and focus on finishing smaller tasks before burning
> >> out. That said I appreciate your effort and am interested in
> >> following / reviewing your work.
> >>
> >> Regards,
> >>
> >> Phil.
> >>
> > 
> > Thanks Philippe for sharing this, and I agree completely, it is very draining.
> > 
> > The effort of making tests happy that run in artificial environments in particular often feels to me
> > as too disconnected from actually ensuring that there is no real run time regression.
> > 
> > qtest_enabled() (implicitly, or explicitly via open-ended else statements) is another painful variable to keep in mind in cpu and machine code, so it is not helpful in my view.
> > 
> > I'll try to push more to get the tests running again, if you have any comment or idea, feel free to just point me in the right direction,
> > that is very valuable to me, even without working code.
> 
> Basically I gave up after realizing from Daniel reviews that we need
> QMP commands to query QEMU at runtime its built-in features, so we
> have build-agnostic tests easier to maintain. I agree this is the
> best way to resolve this particular case, but also scale for various
> other cases.

Which patch / review are you referring to here ?


Regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|



^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-03-08 14:02   ` Alex Bennée
@ 2021-05-05  8:27     ` Claudio Fontana
  2021-05-05 10:04       ` Alex Bennée
  0 siblings, 1 reply; 40+ messages in thread
From: Claudio Fontana @ 2021-05-05  8:27 UTC (permalink / raw)
  To: Alex Bennée, Richard Henderson, Paolo Bonzini
  Cc: Laurent Vivier, Peter Maydell, Thomas Huth, Eduardo Habkost,
	qemu-devel, Philippe Mathieu-Daudé,
	Roman Bolshakov

On 3/8/21 3:02 PM, Alex Bennée wrote:
> 
> Claudio Fontana <cfontana@suse.de> writes:
> 
>> Hi,
>>
>> anything else for me to do here?
> 
> It looks to me that this series is looking pretty good. Every patch has
> at least one review so I think it's just waiting on the maintainers to
> pick it up.
> 
> Paolo/Richard - are you intending to take the series as is or are you
> waiting for something else? I'd like to see the patch delta reduced for
> the ARM cleanup work which is still ongoing.

I am a bit at a loss here, as this has been reviewed for a while, but nothing is happening.
Rebasing is starting to become more and more draining;

I am seeing some changes from Phil that redo some of the patches here (like target arch user),
maybe you would like to drive this?

> 
>>
>> The latest rebased state of this series should be always available here:
>>
>> https://gitlab.com/hw-claudio/qemu/-/tree/i386_cleanup_8

https://gitlab.com/hw-claudio/qemu/-/tree/i386_cleanup_9

is the more recent rebase, but now it is out of date again.

Phil could you help rebase this again?
Otherwise I could only get at this in about two weeks.

Thanks,

Claudio

>>
>> When it comes to the ARM cleanup series,
>> I would like to have the tests pass for ARM, before doing even more changes, could you help me there Philippe?
>>
>> Maybe applying some of your changes on top would fix the failures? I tried, for example with the arm-cpu-features ones, but it didn't work for me..
>>
> <snip>
> 



^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-05-05  8:27     ` Claudio Fontana
@ 2021-05-05 10:04       ` Alex Bennée
  2021-05-05 12:15         ` Philippe Mathieu-Daudé
  0 siblings, 1 reply; 40+ messages in thread
From: Alex Bennée @ 2021-05-05 10:04 UTC (permalink / raw)
  To: Claudio Fontana
  Cc: Laurent Vivier, Peter Maydell, Thomas Huth, Eduardo Habkost,
	Richard Henderson, qemu-devel, Philippe Mathieu-Daudé,
	Roman Bolshakov, Paolo Bonzini


Claudio Fontana <cfontana@suse.de> writes:

> On 3/8/21 3:02 PM, Alex Bennée wrote:
>> 
>> Claudio Fontana <cfontana@suse.de> writes:
>> 
>>> Hi,
>>>
>>> anything else for me to do here?
>> 
>> It looks to me that this series is looking pretty good. Every patch has
>> at least one review so I think it's just waiting on the maintainers to
>> pick it up.
>> 
>> Paolo/Richard - are you intending to take the series as is or are you
>> waiting for something else? I'd like to see the patch delta reduced for
>> the ARM cleanup work which is still ongoing.
>
> I am a bit at a loss here, as this has been reviewed for a while, but nothing is happening.
> Rebasing is starting to become more and more draining;

This is still the latest re-factor right?

  Subject: [PATCH v28 00/23] i386 cleanup PART 2
  Date: Mon, 22 Mar 2021 14:27:36 +0100
  Message-Id: <20210322132800.7470-1-cfontana@suse.de>

> I am seeing some changes from Phil that redo some of the patches here (like target arch user),
> maybe you would like to drive this?

AIUI his changes where to get qtest passing. I've just been chatting to
Paolo on IRC so I think we are almost ready to go.

  > bonzini_: what's currently holding up getting Claudio's re-factoring
      work merged?
  > f4bug: also ^ - I'm a little worried we have splintering
      re-factoring forks
  *** First activity: f4bug joined 2 hours 8 minutes 6 seconds ago.
  <f4bug> stsquad: the qtests series is pending imammedo review then
      could go in
  <f4bug> stsquad: which would simplify a bit Claudio's series (on
      respin he could drop a pair of patches)
  <f4bug> stsquad: my understanding was bonzini_ would merge the x86
      series, then pm215 could merge the arm one on top
  *** First activity: bonzini_ joined 1 hour 17 minutes 25 seconds ago.
  <bonzini_> ok i can queue it in my next PR
  <f4bug> the only blocking part is qtest not passing, but Claudio's
      refactor series is not the culprit
  <bonzini_> ok

-- 
Alex Bennée


^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-05-05 10:04       ` Alex Bennée
@ 2021-05-05 12:15         ` Philippe Mathieu-Daudé
  2021-05-05 12:24           ` Philippe Mathieu-Daudé
  2021-05-05 19:31           ` Eduardo Habkost
  0 siblings, 2 replies; 40+ messages in thread
From: Philippe Mathieu-Daudé @ 2021-05-05 12:15 UTC (permalink / raw)
  To: Alex Bennée, Claudio Fontana, Igor Mammedov, Thomas Huth
  Cc: Laurent Vivier, Peter Maydell, Eduardo Habkost,
	Richard Henderson, qemu-devel, Markus Armbruster,
	Roman Bolshakov, Paolo Bonzini

On 5/5/21 12:04 PM, Alex Bennée wrote:
> Claudio Fontana <cfontana@suse.de> writes:
>> On 3/8/21 3:02 PM, Alex Bennée wrote:
>>> Claudio Fontana <cfontana@suse.de> writes:
>>>
>>>> Hi,
>>>>
>>>> anything else for me to do here?
>>>
>>> It looks to me that this series is looking pretty good. Every patch has
>>> at least one review so I think it's just waiting on the maintainers to
>>> pick it up.
>>>
>>> Paolo/Richard - are you intending to take the series as is or are you
>>> waiting for something else? I'd like to see the patch delta reduced for
>>> the ARM cleanup work which is still ongoing.
>>
>> I am a bit at a loss here, as this has been reviewed for a while, but nothing is happening.
>> Rebasing is starting to become more and more draining;
> 
> This is still the latest re-factor right?
> 
>   Subject: [PATCH v28 00/23] i386 cleanup PART 2
>   Date: Mon, 22 Mar 2021 14:27:36 +0100
>   Message-Id: <20210322132800.7470-1-cfontana@suse.de>
> 
>> I am seeing some changes from Phil that redo some of the patches here (like target arch user),
>> maybe you would like to drive this?
> 
> AIUI his changes where to get qtest passing.

I hadn't read Claudio's mail, I think he's mentioning commit 46369b50ee3

    meson: Introduce meson_user_arch source set for arch-specific user-mode

    Similarly to the 'target_softmmu_arch' source set which allows
    to restrict target-specific sources to system emulation, add
    the equivalent 'target_user_arch' set for user emulation.

The patch only contains 6 lines in 2 hunks, if it introduced a conflict
it should be trivial to resolve (I wasn't expecting it to conflict with
your work when I merged it TBH).

> I've just been chatting to
> Paolo on IRC so I think we are almost ready to go.
> 
>   > bonzini_: what's currently holding up getting Claudio's re-factoring
>       work merged?
>   > f4bug: also ^ - I'm a little worried we have splintering
>       re-factoring forks
>   *** First activity: f4bug joined 2 hours 8 minutes 6 seconds ago.
>   <f4bug> stsquad: the qtests series is pending imammedo review then
>       could go in
>   <f4bug> stsquad: which would simplify a bit Claudio's series (on
>       respin he could drop a pair of patches)
>   <f4bug> stsquad: my understanding was bonzini_ would merge the x86
>       series, then pm215 could merge the arm one on top
>   *** First activity: bonzini_ joined 1 hour 17 minutes 25 seconds ago.
>   <bonzini_> ok i can queue it in my next PR
>   <f4bug> the only blocking part is qtest not passing, but Claudio's
>       refactor series is not the culprit
>   <bonzini_> ok

I was referring to:
https://www.mail-archive.com/qemu-devel@nongnu.org/msg804015.html

Then these aren't required:
- tests: restrict TCG-only arm-cpu-features tests to TCG builds
- tests: device-introspect-test: cope with ARM TCG-only devices

These could be reworked to use qtest_has_accel() instead of the
/* XXX */ comments:
- tests: do not run test-hmp on all machines for ARM KVM-only
- tests: do not run qom-test on all machines for ARM KVM-only

I didn't noticed the following patch had its content changed:
Revert "target/arm: Restrict v8M IDAU to TCG"
So now this is not a full revert, only the TYPE_IDAU_INTERFACE
type is moved back.

While this fixes the build, we still have a QOM design problem.
QOM interfaces can't be Kconfig-selected out. <- Eduardo, Markus?


More generally I think more code should be automatically stripped
out by Kconfig instead. In [1,2] I suggested to tie accel-specific
Kconfig selectors:

  config ARM_V7R
    bool
    depends on TCG && ARM

  config ARM_V7M
    bool
    depends on TCG && ARM
    select PTIMER

  config XLNX_ZYNQMP_ARM
    bool
    default y if TCG && AARCH64

But this can certainly be done on top of Claudio's work.

[1] https://www.mail-archive.com/qemu-devel@nongnu.org/msg777710.html
[2] https://www.mail-archive.com/qemu-devel@nongnu.org/msg777719.html



^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-05-05 12:15         ` Philippe Mathieu-Daudé
@ 2021-05-05 12:24           ` Philippe Mathieu-Daudé
  2021-05-05 19:31           ` Eduardo Habkost
  1 sibling, 0 replies; 40+ messages in thread
From: Philippe Mathieu-Daudé @ 2021-05-05 12:24 UTC (permalink / raw)
  To: Alex Bennée, Claudio Fontana, Igor Mammedov, Thomas Huth
  Cc: Laurent Vivier, Peter Maydell, Eduardo Habkost,
	Richard Henderson, qemu-devel, Markus Armbruster,
	Roman Bolshakov, Paolo Bonzini

On 5/5/21 2:15 PM, Philippe Mathieu-Daudé wrote:
> On 5/5/21 12:04 PM, Alex Bennée wrote:
>> Claudio Fontana <cfontana@suse.de> writes:
>>> On 3/8/21 3:02 PM, Alex Bennée wrote:
>>>> Claudio Fontana <cfontana@suse.de> writes:
>>>>
>>>>> Hi,
>>>>>
>>>>> anything else for me to do here?
>>>>
>>>> It looks to me that this series is looking pretty good. Every patch has
>>>> at least one review so I think it's just waiting on the maintainers to
>>>> pick it up.
>>>>
>>>> Paolo/Richard - are you intending to take the series as is or are you
>>>> waiting for something else? I'd like to see the patch delta reduced for
>>>> the ARM cleanup work which is still ongoing.
>>>
>>> I am a bit at a loss here, as this has been reviewed for a while, but nothing is happening.
>>> Rebasing is starting to become more and more draining;

> I didn't noticed the following patch had its content changed:
> Revert "target/arm: Restrict v8M IDAU to TCG"
> So now this is not a full revert, only the TYPE_IDAU_INTERFACE
> type is moved back.

Correcting myself. This is a plain revert, and I disagree with it
because if TCG is not available, TCG-specific devices/machines
shouldn't be loaded into QEMU. Cortex-M and v8M IDAU *are* TCG
specific and no such object should be built/registered, so the
interface shouldn't be there neither.

Below is a possible way to get there. Reverting the commit and
build Cortex-M devices in a non-TCG build is a kludge / shortcut
which allows Claudio to remove a big chunk of non-TCG code, so
is probably acceptable, and we can keep cleaning later.

> More generally I think more code should be automatically stripped
> out by Kconfig instead. In [1,2] I suggested to tie accel-specific
> Kconfig selectors:
> 
>   config ARM_V7R
>     bool
>     depends on TCG && ARM
> 
>   config ARM_V7M
>     bool
>     depends on TCG && ARM
>     select PTIMER
> 
>   config XLNX_ZYNQMP_ARM
>     bool
>     default y if TCG && AARCH64
> 
> But this can certainly be done on top of Claudio's work.
> 
> [1] https://www.mail-archive.com/qemu-devel@nongnu.org/msg777710.html
> [2] https://www.mail-archive.com/qemu-devel@nongnu.org/msg777719.html
> 



^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-05-05 12:15         ` Philippe Mathieu-Daudé
  2021-05-05 12:24           ` Philippe Mathieu-Daudé
@ 2021-05-05 19:31           ` Eduardo Habkost
  2021-05-06  7:55             ` Claudio Fontana
  1 sibling, 1 reply; 40+ messages in thread
From: Eduardo Habkost @ 2021-05-05 19:31 UTC (permalink / raw)
  To: Philippe Mathieu-Daudé
  Cc: Laurent Vivier, Peter Maydell, Thomas Huth, Richard Henderson,
	qemu-devel, Markus Armbruster, Roman Bolshakov, Claudio Fontana,
	Paolo Bonzini, Igor Mammedov, Alex Bennée

On Wed, May 05, 2021 at 02:15:29PM +0200, Philippe Mathieu-Daudé wrote:
> On 5/5/21 12:04 PM, Alex Bennée wrote:
> > Claudio Fontana <cfontana@suse.de> writes:
> >> On 3/8/21 3:02 PM, Alex Bennée wrote:
> >>> Claudio Fontana <cfontana@suse.de> writes:
> >>>
> >>>> Hi,
> >>>>
> >>>> anything else for me to do here?
> >>>
> >>> It looks to me that this series is looking pretty good. Every patch has
> >>> at least one review so I think it's just waiting on the maintainers to
> >>> pick it up.
> >>>
> >>> Paolo/Richard - are you intending to take the series as is or are you
> >>> waiting for something else? I'd like to see the patch delta reduced for
> >>> the ARM cleanup work which is still ongoing.
> >>
> >> I am a bit at a loss here, as this has been reviewed for a while, but nothing is happening.
> >> Rebasing is starting to become more and more draining;
> > 
> > This is still the latest re-factor right?
> > 
> >   Subject: [PATCH v28 00/23] i386 cleanup PART 2
> >   Date: Mon, 22 Mar 2021 14:27:36 +0100
> >   Message-Id: <20210322132800.7470-1-cfontana@suse.de>
> > 
> >> I am seeing some changes from Phil that redo some of the patches here (like target arch user),
> >> maybe you would like to drive this?
> > 
> > AIUI his changes where to get qtest passing.
> 
> I hadn't read Claudio's mail, I think he's mentioning commit 46369b50ee3
> 
>     meson: Introduce meson_user_arch source set for arch-specific user-mode
> 
>     Similarly to the 'target_softmmu_arch' source set which allows
>     to restrict target-specific sources to system emulation, add
>     the equivalent 'target_user_arch' set for user emulation.
> 
> The patch only contains 6 lines in 2 hunks, if it introduced a conflict
> it should be trivial to resolve (I wasn't expecting it to conflict with
> your work when I merged it TBH).
> 
> > I've just been chatting to
> > Paolo on IRC so I think we are almost ready to go.
> > 
> >   > bonzini_: what's currently holding up getting Claudio's re-factoring
> >       work merged?
> >   > f4bug: also ^ - I'm a little worried we have splintering
> >       re-factoring forks
> >   *** First activity: f4bug joined 2 hours 8 minutes 6 seconds ago.
> >   <f4bug> stsquad: the qtests series is pending imammedo review then
> >       could go in
> >   <f4bug> stsquad: which would simplify a bit Claudio's series (on
> >       respin he could drop a pair of patches)
> >   <f4bug> stsquad: my understanding was bonzini_ would merge the x86
> >       series, then pm215 could merge the arm one on top
> >   *** First activity: bonzini_ joined 1 hour 17 minutes 25 seconds ago.
> >   <bonzini_> ok i can queue it in my next PR
> >   <f4bug> the only blocking part is qtest not passing, but Claudio's
> >       refactor series is not the culprit
> >   <bonzini_> ok
> 
> I was referring to:
> https://www.mail-archive.com/qemu-devel@nongnu.org/msg804015.html
> 
> Then these aren't required:
> - tests: restrict TCG-only arm-cpu-features tests to TCG builds
> - tests: device-introspect-test: cope with ARM TCG-only devices
> 
> These could be reworked to use qtest_has_accel() instead of the
> /* XXX */ comments:
> - tests: do not run test-hmp on all machines for ARM KVM-only
> - tests: do not run qom-test on all machines for ARM KVM-only
> 
> I didn't noticed the following patch had its content changed:
> Revert "target/arm: Restrict v8M IDAU to TCG"
> So now this is not a full revert, only the TYPE_IDAU_INTERFACE
> type is moved back.
> 
> While this fixes the build, we still have a QOM design problem.
> QOM interfaces can't be Kconfig-selected out. <- Eduardo, Markus?

I don't get it.  Why exactly QOM interfaces can't be
Kconfig-selected out?

Do you want to be able to compile out an interface while not
compiling out types that implement that interface?  Why?


> 
> 
> More generally I think more code should be automatically stripped
> out by Kconfig instead. In [1,2] I suggested to tie accel-specific
> Kconfig selectors:
> 
>   config ARM_V7R
>     bool
>     depends on TCG && ARM
> 
>   config ARM_V7M
>     bool
>     depends on TCG && ARM
>     select PTIMER
> 
>   config XLNX_ZYNQMP_ARM
>     bool
>     default y if TCG && AARCH64
> 
> But this can certainly be done on top of Claudio's work.
> 
> [1] https://www.mail-archive.com/qemu-devel@nongnu.org/msg777710.html
> [2] https://www.mail-archive.com/qemu-devel@nongnu.org/msg777719.html
> 

-- 
Eduardo



^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-05-05 19:31           ` Eduardo Habkost
@ 2021-05-06  7:55             ` Claudio Fontana
  2021-05-06  9:10               ` Philippe Mathieu-Daudé
  0 siblings, 1 reply; 40+ messages in thread
From: Claudio Fontana @ 2021-05-06  7:55 UTC (permalink / raw)
  To: Eduardo Habkost, Philippe Mathieu-Daudé
  Cc: Laurent Vivier, Peter Maydell, Thomas Huth, Richard Henderson,
	qemu-devel, Markus Armbruster, Roman Bolshakov, Paolo Bonzini,
	Igor Mammedov, Alex Bennée

On 5/5/21 9:31 PM, Eduardo Habkost wrote:
> On Wed, May 05, 2021 at 02:15:29PM +0200, Philippe Mathieu-Daudé wrote:
>> On 5/5/21 12:04 PM, Alex Bennée wrote:
>>> Claudio Fontana <cfontana@suse.de> writes:
>>>> On 3/8/21 3:02 PM, Alex Bennée wrote:
>>>>> Claudio Fontana <cfontana@suse.de> writes:
>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> anything else for me to do here?
>>>>>
>>>>> It looks to me that this series is looking pretty good. Every patch has
>>>>> at least one review so I think it's just waiting on the maintainers to
>>>>> pick it up.
>>>>>
>>>>> Paolo/Richard - are you intending to take the series as is or are you
>>>>> waiting for something else? I'd like to see the patch delta reduced for
>>>>> the ARM cleanup work which is still ongoing.
>>>>
>>>> I am a bit at a loss here, as this has been reviewed for a while, but nothing is happening.
>>>> Rebasing is starting to become more and more draining;
>>>
>>> This is still the latest re-factor right?
>>>
>>>   Subject: [PATCH v28 00/23] i386 cleanup PART 2
>>>   Date: Mon, 22 Mar 2021 14:27:36 +0100
>>>   Message-Id: <20210322132800.7470-1-cfontana@suse.de>
>>>
>>>> I am seeing some changes from Phil that redo some of the patches here (like target arch user),
>>>> maybe you would like to drive this?
>>>
>>> AIUI his changes where to get qtest passing.
>>
>> I hadn't read Claudio's mail, I think he's mentioning commit 46369b50ee3
>>
>>     meson: Introduce meson_user_arch source set for arch-specific user-mode
>>
>>     Similarly to the 'target_softmmu_arch' source set which allows
>>     to restrict target-specific sources to system emulation, add
>>     the equivalent 'target_user_arch' set for user emulation.
>>
>> The patch only contains 6 lines in 2 hunks, if it introduced a conflict
>> it should be trivial to resolve (I wasn't expecting it to conflict with
>> your work when I merged it TBH).
>>
>>> I've just been chatting to
>>> Paolo on IRC so I think we are almost ready to go.
>>>
>>>   > bonzini_: what's currently holding up getting Claudio's re-factoring
>>>       work merged?
>>>   > f4bug: also ^ - I'm a little worried we have splintering
>>>       re-factoring forks
>>>   *** First activity: f4bug joined 2 hours 8 minutes 6 seconds ago.
>>>   <f4bug> stsquad: the qtests series is pending imammedo review then
>>>       could go in
>>>   <f4bug> stsquad: which would simplify a bit Claudio's series (on
>>>       respin he could drop a pair of patches)
>>>   <f4bug> stsquad: my understanding was bonzini_ would merge the x86
>>>       series, then pm215 could merge the arm one on top
>>>   *** First activity: bonzini_ joined 1 hour 17 minutes 25 seconds ago.
>>>   <bonzini_> ok i can queue it in my next PR
>>>   <f4bug> the only blocking part is qtest not passing, but Claudio's
>>>       refactor series is not the culprit
>>>   <bonzini_> ok
>>
>> I was referring to:
>> https://www.mail-archive.com/qemu-devel@nongnu.org/msg804015.html
>>
>> Then these aren't required:
>> - tests: restrict TCG-only arm-cpu-features tests to TCG builds
>> - tests: device-introspect-test: cope with ARM TCG-only devices
>>
>> These could be reworked to use qtest_has_accel() instead of the
>> /* XXX */ comments:
>> - tests: do not run test-hmp on all machines for ARM KVM-only
>> - tests: do not run qom-test on all machines for ARM KVM-only
>>
>> I didn't noticed the following patch had its content changed:
>> Revert "target/arm: Restrict v8M IDAU to TCG"
>> So now this is not a full revert, only the TYPE_IDAU_INTERFACE
>> type is moved back.
>>
>> While this fixes the build, we still have a QOM design problem.
>> QOM interfaces can't be Kconfig-selected out. <- Eduardo, Markus?
> 
> I don't get it.  Why exactly QOM interfaces can't be
> Kconfig-selected out?
> 
> Do you want to be able to compile out an interface while not
> compiling out types that implement that interface?  Why?


I'd suggest to move the discussions about the ARM series to the arm series thread.

I am concerned here about the groundwork and x86 part.

Thanks,

Claudio

> 
> 
>>
>>
>> More generally I think more code should be automatically stripped
>> out by Kconfig instead. In [1,2] I suggested to tie accel-specific
>> Kconfig selectors:
>>
>>   config ARM_V7R
>>     bool
>>     depends on TCG && ARM
>>
>>   config ARM_V7M
>>     bool
>>     depends on TCG && ARM
>>     select PTIMER
>>
>>   config XLNX_ZYNQMP_ARM
>>     bool
>>     default y if TCG && AARCH64
>>
>> But this can certainly be done on top of Claudio's work.
>>
>> [1] https://www.mail-archive.com/qemu-devel@nongnu.org/msg777710.html
>> [2] https://www.mail-archive.com/qemu-devel@nongnu.org/msg777719.html
>>
> 



^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-05-06  7:55             ` Claudio Fontana
@ 2021-05-06  9:10               ` Philippe Mathieu-Daudé
  2021-05-06  9:33                 ` Philippe Mathieu-Daudé
  0 siblings, 1 reply; 40+ messages in thread
From: Philippe Mathieu-Daudé @ 2021-05-06  9:10 UTC (permalink / raw)
  To: Claudio Fontana, Eduardo Habkost
  Cc: Laurent Vivier, Peter Maydell, Thomas Huth, Richard Henderson,
	qemu-devel, Markus Armbruster, Roman Bolshakov, Igor Mammedov,
	Paolo Bonzini, Alex Bennée

On 5/6/21 9:55 AM, Claudio Fontana wrote:
> On 5/5/21 9:31 PM, Eduardo Habkost wrote:
>> On Wed, May 05, 2021 at 02:15:29PM +0200, Philippe Mathieu-Daudé wrote:
>>> On 5/5/21 12:04 PM, Alex Bennée wrote:
>>>> Claudio Fontana <cfontana@suse.de> writes:
>>>>> On 3/8/21 3:02 PM, Alex Bennée wrote:
>>>>>> Claudio Fontana <cfontana@suse.de> writes:
>>>>>>
>>>>>>> Hi,
>>>>>>>
>>>>>>> anything else for me to do here?
>>>>>>
>>>>>> It looks to me that this series is looking pretty good. Every patch has
>>>>>> at least one review so I think it's just waiting on the maintainers to
>>>>>> pick it up.
>>>>>>
>>>>>> Paolo/Richard - are you intending to take the series as is or are you
>>>>>> waiting for something else? I'd like to see the patch delta reduced for
>>>>>> the ARM cleanup work which is still ongoing.
>>>>>
>>>>> I am a bit at a loss here, as this has been reviewed for a while, but nothing is happening.
>>>>> Rebasing is starting to become more and more draining;
>>>>
>>>> This is still the latest re-factor right?
>>>>
>>>>   Subject: [PATCH v28 00/23] i386 cleanup PART 2
>>>>   Date: Mon, 22 Mar 2021 14:27:36 +0100
>>>>   Message-Id: <20210322132800.7470-1-cfontana@suse.de>
>>>>
>>>>> I am seeing some changes from Phil that redo some of the patches here (like target arch user),
>>>>> maybe you would like to drive this?
>>>>
>>>> AIUI his changes where to get qtest passing.
>>>
>>> I hadn't read Claudio's mail, I think he's mentioning commit 46369b50ee3
>>>
>>>     meson: Introduce meson_user_arch source set for arch-specific user-mode
>>>
>>>     Similarly to the 'target_softmmu_arch' source set which allows
>>>     to restrict target-specific sources to system emulation, add
>>>     the equivalent 'target_user_arch' set for user emulation.
>>>
>>> The patch only contains 6 lines in 2 hunks, if it introduced a conflict
>>> it should be trivial to resolve (I wasn't expecting it to conflict with
>>> your work when I merged it TBH).

> I'd suggest to move the discussions about the ARM series to the arm series thread.
> 
> I am concerned here about the groundwork and x86 part.

OK sorry, I was explaining the IRC chat.

Is there any issue rebasing the groundwork on top of commit 46369b50ee3?

Maybe my qtest/accel series is irrelevant to your x86 part, TBH I don't
remember.

Regards,

Phil.



^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-05-06  9:10               ` Philippe Mathieu-Daudé
@ 2021-05-06  9:33                 ` Philippe Mathieu-Daudé
  2021-05-12  7:17                   ` Philippe Mathieu-Daudé
  0 siblings, 1 reply; 40+ messages in thread
From: Philippe Mathieu-Daudé @ 2021-05-06  9:33 UTC (permalink / raw)
  To: Claudio Fontana, Eduardo Habkost
  Cc: Laurent Vivier, Peter Maydell, Thomas Huth, Richard Henderson,
	qemu-devel, Markus Armbruster, Roman Bolshakov, Igor Mammedov,
	Paolo Bonzini, Alex Bennée

On 5/6/21 11:10 AM, Philippe Mathieu-Daudé wrote:
> On 5/6/21 9:55 AM, Claudio Fontana wrote:
>> On 5/5/21 9:31 PM, Eduardo Habkost wrote:
>>> On Wed, May 05, 2021 at 02:15:29PM +0200, Philippe Mathieu-Daudé wrote:
>>>> On 5/5/21 12:04 PM, Alex Bennée wrote:
>>>>> Claudio Fontana <cfontana@suse.de> writes:
>>>>>> On 3/8/21 3:02 PM, Alex Bennée wrote:
>>>>>>> Claudio Fontana <cfontana@suse.de> writes:
>>>>>>>
>>>>>>>> Hi,
>>>>>>>>
>>>>>>>> anything else for me to do here?
>>>>>>>
>>>>>>> It looks to me that this series is looking pretty good. Every patch has
>>>>>>> at least one review so I think it's just waiting on the maintainers to
>>>>>>> pick it up.
>>>>>>>
>>>>>>> Paolo/Richard - are you intending to take the series as is or are you
>>>>>>> waiting for something else? I'd like to see the patch delta reduced for
>>>>>>> the ARM cleanup work which is still ongoing.
>>>>>>
>>>>>> I am a bit at a loss here, as this has been reviewed for a while, but nothing is happening.
>>>>>> Rebasing is starting to become more and more draining;
>>>>>
>>>>> This is still the latest re-factor right?
>>>>>
>>>>>   Subject: [PATCH v28 00/23] i386 cleanup PART 2
>>>>>   Date: Mon, 22 Mar 2021 14:27:36 +0100
>>>>>   Message-Id: <20210322132800.7470-1-cfontana@suse.de>
>>>>>
>>>>>> I am seeing some changes from Phil that redo some of the patches here (like target arch user),
>>>>>> maybe you would like to drive this?
>>>>>
>>>>> AIUI his changes where to get qtest passing.
>>>>
>>>> I hadn't read Claudio's mail, I think he's mentioning commit 46369b50ee3
>>>>
>>>>     meson: Introduce meson_user_arch source set for arch-specific user-mode
>>>>
>>>>     Similarly to the 'target_softmmu_arch' source set which allows
>>>>     to restrict target-specific sources to system emulation, add
>>>>     the equivalent 'target_user_arch' set for user emulation.
>>>>
>>>> The patch only contains 6 lines in 2 hunks, if it introduced a conflict
>>>> it should be trivial to resolve (I wasn't expecting it to conflict with
>>>> your work when I merged it TBH).
> 
>> I'd suggest to move the discussions about the ARM series to the arm series thread.
>>
>> I am concerned here about the groundwork and x86 part.
> 
> OK sorry, I was explaining the IRC chat.
> 
> Is there any issue rebasing the groundwork on top of commit 46369b50ee3?
> 
> Maybe my qtest/accel series is irrelevant to your x86 part, TBH I don't
> remember.

So far I could rebase this series on top of commit 229a834518b
(Mon Mar 8 15:45:48 2021).

First trivial conflict with 2cc1a90166b ("Remove deprecated
target tilegx"), removing target/tilegx/meson.build resolves it.

Then apparently commit b8184135835 ("target/i386: allow modifying
TCG phys-addr-bits") which is not trivial (to me).

Finally "meson: add target_user_arch" indeed clashes with commit
46369b50ee3 ("meson: Introduce meson_user_arch source set for
arch-specific user-mode"). I am sorry I didn't notice your patch,
but am glad we both sent a similar patch :) Mine allows optional
sourceset, so you can simplify your patch, only keeping the x86
part:

-- >8 --
diff --git a/target/i386/meson.build b/target/i386/meson.build
index fd24479590..cac26a4581 100644
--- a/target/i386/meson.build
+++ b/target/i386/meson.build
@@ -19,6 +19,7 @@ i386_softmmu_ss.add(files(
   'machine.c',
   'monitor.c',
 ))
+i386_user_ss = ss.source_set()

 subdir('kvm')
 subdir('hax')
@@ -28,3 +29,4 @@ subdir('tcg')

 target_arch += {'i386': i386_ss}
 target_softmmu_arch += {'i386': i386_softmmu_ss}
+target_user_arch += {'i386': i386_user_ss}
---

Regards,

Phil.



^ permalink raw reply related	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-05-06  9:33                 ` Philippe Mathieu-Daudé
@ 2021-05-12  7:17                   ` Philippe Mathieu-Daudé
  2021-05-17  6:17                     ` Claudio Fontana
  0 siblings, 1 reply; 40+ messages in thread
From: Philippe Mathieu-Daudé @ 2021-05-12  7:17 UTC (permalink / raw)
  To: Claudio Fontana, Eduardo Habkost
  Cc: Laurent Vivier, Peter Maydell, Thomas Huth, Richard Henderson,
	qemu-devel, Markus Armbruster, Roman Bolshakov, Igor Mammedov,
	Paolo Bonzini, Alex Bennée

On 5/6/21 11:33 AM, Philippe Mathieu-Daudé wrote:
> On 5/6/21 11:10 AM, Philippe Mathieu-Daudé wrote:
>> On 5/6/21 9:55 AM, Claudio Fontana wrote:
>>> On 5/5/21 9:31 PM, Eduardo Habkost wrote:
>>>> On Wed, May 05, 2021 at 02:15:29PM +0200, Philippe Mathieu-Daudé wrote:
>>>>> On 5/5/21 12:04 PM, Alex Bennée wrote:
>>>>>> Claudio Fontana <cfontana@suse.de> writes:
>>>>>>> On 3/8/21 3:02 PM, Alex Bennée wrote:
>>>>>>>> Claudio Fontana <cfontana@suse.de> writes:
>>>>>>>>
>>>>>>>>> Hi,
>>>>>>>>>
>>>>>>>>> anything else for me to do here?
>>>>>>>>
>>>>>>>> It looks to me that this series is looking pretty good. Every patch has
>>>>>>>> at least one review so I think it's just waiting on the maintainers to
>>>>>>>> pick it up.
>>>>>>>>
>>>>>>>> Paolo/Richard - are you intending to take the series as is or are you
>>>>>>>> waiting for something else? I'd like to see the patch delta reduced for
>>>>>>>> the ARM cleanup work which is still ongoing.
>>>>>>>
>>>>>>> I am a bit at a loss here, as this has been reviewed for a while, but nothing is happening.
>>>>>>> Rebasing is starting to become more and more draining;
>>>>>>
>>>>>> This is still the latest re-factor right?
>>>>>>
>>>>>>   Subject: [PATCH v28 00/23] i386 cleanup PART 2
>>>>>>   Date: Mon, 22 Mar 2021 14:27:36 +0100
>>>>>>   Message-Id: <20210322132800.7470-1-cfontana@suse.de>
>>>>>>
>>>>>>> I am seeing some changes from Phil that redo some of the patches here (like target arch user),
>>>>>>> maybe you would like to drive this?
>>>>>>
>>>>>> AIUI his changes where to get qtest passing.
>>>>>
>>>>> I hadn't read Claudio's mail, I think he's mentioning commit 46369b50ee3
>>>>>
>>>>>     meson: Introduce meson_user_arch source set for arch-specific user-mode
>>>>>
>>>>>     Similarly to the 'target_softmmu_arch' source set which allows
>>>>>     to restrict target-specific sources to system emulation, add
>>>>>     the equivalent 'target_user_arch' set for user emulation.
>>>>>
>>>>> The patch only contains 6 lines in 2 hunks, if it introduced a conflict
>>>>> it should be trivial to resolve (I wasn't expecting it to conflict with
>>>>> your work when I merged it TBH).
>>
>>> I'd suggest to move the discussions about the ARM series to the arm series thread.
>>>
>>> I am concerned here about the groundwork and x86 part.
>>
>> OK sorry, I was explaining the IRC chat.
>>
>> Is there any issue rebasing the groundwork on top of commit 46369b50ee3?
>>
>> Maybe my qtest/accel series is irrelevant to your x86 part, TBH I don't
>> remember.
> 
> So far I could rebase this series on top of commit 229a834518b
> (Mon Mar 8 15:45:48 2021).

Pfff I realized late you asked on the v26 series, but v28 is
available... Let's try again.

> First trivial conflict with 2cc1a90166b ("Remove deprecated
> target tilegx"), removing target/tilegx/meson.build resolves it.
> 
> Then apparently commit b8184135835 ("target/i386: allow modifying
> TCG phys-addr-bits") which is not trivial (to me).
> 
> Finally "meson: add target_user_arch" indeed clashes with commit
> 46369b50ee3 ("meson: Introduce meson_user_arch source set for
> arch-specific user-mode"). I am sorry I didn't notice your patch,
> but am glad we both sent a similar patch :) Mine allows optional
> sourceset, so you can simplify your patch, only keeping the x86
> part:
> 
> -- >8 --
> diff --git a/target/i386/meson.build b/target/i386/meson.build
> index fd24479590..cac26a4581 100644
> --- a/target/i386/meson.build
> +++ b/target/i386/meson.build
> @@ -19,6 +19,7 @@ i386_softmmu_ss.add(files(
>    'machine.c',
>    'monitor.c',
>  ))
> +i386_user_ss = ss.source_set()
> 
>  subdir('kvm')
>  subdir('hax')
> @@ -28,3 +29,4 @@ subdir('tcg')
> 
>  target_arch += {'i386': i386_ss}
>  target_softmmu_arch += {'i386': i386_softmmu_ss}
> +target_user_arch += {'i386': i386_user_ss}
> ---
> 
> Regards,
> 
> Phil.
> 



^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-05-12  7:17                   ` Philippe Mathieu-Daudé
@ 2021-05-17  6:17                     ` Claudio Fontana
  2021-05-19 18:17                       ` Philippe Mathieu-Daudé
  0 siblings, 1 reply; 40+ messages in thread
From: Claudio Fontana @ 2021-05-17  6:17 UTC (permalink / raw)
  To: Philippe Mathieu-Daudé, Eduardo Habkost
  Cc: Laurent Vivier, Peter Maydell, Thomas Huth, Richard Henderson,
	qemu-devel, Markus Armbruster, Roman Bolshakov, Paolo Bonzini,
	Igor Mammedov, Alex Bennée

On 5/12/21 9:17 AM, Philippe Mathieu-Daudé wrote:
> On 5/6/21 11:33 AM, Philippe Mathieu-Daudé wrote:
>> On 5/6/21 11:10 AM, Philippe Mathieu-Daudé wrote:
>>> On 5/6/21 9:55 AM, Claudio Fontana wrote:
>>>> On 5/5/21 9:31 PM, Eduardo Habkost wrote:
>>>>> On Wed, May 05, 2021 at 02:15:29PM +0200, Philippe Mathieu-Daudé wrote:
>>>>>> On 5/5/21 12:04 PM, Alex Bennée wrote:
>>>>>>> Claudio Fontana <cfontana@suse.de> writes:
>>>>>>>> On 3/8/21 3:02 PM, Alex Bennée wrote:
>>>>>>>>> Claudio Fontana <cfontana@suse.de> writes:
>>>>>>>>>
>>>>>>>>>> Hi,
>>>>>>>>>>
>>>>>>>>>> anything else for me to do here?
>>>>>>>>>
>>>>>>>>> It looks to me that this series is looking pretty good. Every patch has
>>>>>>>>> at least one review so I think it's just waiting on the maintainers to
>>>>>>>>> pick it up.
>>>>>>>>>
>>>>>>>>> Paolo/Richard - are you intending to take the series as is or are you
>>>>>>>>> waiting for something else? I'd like to see the patch delta reduced for
>>>>>>>>> the ARM cleanup work which is still ongoing.
>>>>>>>>
>>>>>>>> I am a bit at a loss here, as this has been reviewed for a while, but nothing is happening.
>>>>>>>> Rebasing is starting to become more and more draining;
>>>>>>>
>>>>>>> This is still the latest re-factor right?
>>>>>>>
>>>>>>>   Subject: [PATCH v28 00/23] i386 cleanup PART 2
>>>>>>>   Date: Mon, 22 Mar 2021 14:27:36 +0100
>>>>>>>   Message-Id: <20210322132800.7470-1-cfontana@suse.de>
>>>>>>>
>>>>>>>> I am seeing some changes from Phil that redo some of the patches here (like target arch user),
>>>>>>>> maybe you would like to drive this?
>>>>>>>
>>>>>>> AIUI his changes where to get qtest passing.
>>>>>>
>>>>>> I hadn't read Claudio's mail, I think he's mentioning commit 46369b50ee3
>>>>>>
>>>>>>     meson: Introduce meson_user_arch source set for arch-specific user-mode
>>>>>>
>>>>>>     Similarly to the 'target_softmmu_arch' source set which allows
>>>>>>     to restrict target-specific sources to system emulation, add
>>>>>>     the equivalent 'target_user_arch' set for user emulation.
>>>>>>
>>>>>> The patch only contains 6 lines in 2 hunks, if it introduced a conflict
>>>>>> it should be trivial to resolve (I wasn't expecting it to conflict with
>>>>>> your work when I merged it TBH).
>>>
>>>> I'd suggest to move the discussions about the ARM series to the arm series thread.
>>>>
>>>> I am concerned here about the groundwork and x86 part.
>>>
>>> OK sorry, I was explaining the IRC chat.
>>>
>>> Is there any issue rebasing the groundwork on top of commit 46369b50ee3?
>>>
>>> Maybe my qtest/accel series is irrelevant to your x86 part, TBH I don't
>>> remember.
>>
>> So far I could rebase this series on top of commit 229a834518b
>> (Mon Mar 8 15:45:48 2021).
> 
> Pfff I realized late you asked on the v26 series, but v28 is
> available... Let's try again.

Hi Philippe,

I think that Paolo picked it up in the end?

Thanks,

Claudio

> 
>> First trivial conflict with 2cc1a90166b ("Remove deprecated
>> target tilegx"), removing target/tilegx/meson.build resolves it.
>>
>> Then apparently commit b8184135835 ("target/i386: allow modifying
>> TCG phys-addr-bits") which is not trivial (to me).
>>
>> Finally "meson: add target_user_arch" indeed clashes with commit
>> 46369b50ee3 ("meson: Introduce meson_user_arch source set for
>> arch-specific user-mode"). I am sorry I didn't notice your patch,
>> but am glad we both sent a similar patch :) Mine allows optional
>> sourceset, so you can simplify your patch, only keeping the x86
>> part:
>>
>> -- >8 --
>> diff --git a/target/i386/meson.build b/target/i386/meson.build
>> index fd24479590..cac26a4581 100644
>> --- a/target/i386/meson.build
>> +++ b/target/i386/meson.build
>> @@ -19,6 +19,7 @@ i386_softmmu_ss.add(files(
>>    'machine.c',
>>    'monitor.c',
>>  ))
>> +i386_user_ss = ss.source_set()
>>
>>  subdir('kvm')
>>  subdir('hax')
>> @@ -28,3 +29,4 @@ subdir('tcg')
>>
>>  target_arch += {'i386': i386_ss}
>>  target_softmmu_arch += {'i386': i386_softmmu_ss}
>> +target_user_arch += {'i386': i386_user_ss}
>> ---
>>
>> Regards,
>>
>> Phil.
>>
> 
> 



^ permalink raw reply	[flat|nested] 40+ messages in thread

* Re: [PATCH v26 00/20] i386 cleanup PART 2
  2021-05-17  6:17                     ` Claudio Fontana
@ 2021-05-19 18:17                       ` Philippe Mathieu-Daudé
  0 siblings, 0 replies; 40+ messages in thread
From: Philippe Mathieu-Daudé @ 2021-05-19 18:17 UTC (permalink / raw)
  To: Claudio Fontana, Eduardo Habkost
  Cc: Laurent Vivier, Peter Maydell, Thomas Huth, Richard Henderson,
	qemu-devel, Markus Armbruster, Roman Bolshakov, Paolo Bonzini,
	Igor Mammedov, Alex Bennée

On 5/17/21 8:17 AM, Claudio Fontana wrote:
> On 5/12/21 9:17 AM, Philippe Mathieu-Daudé wrote:
>> On 5/6/21 11:33 AM, Philippe Mathieu-Daudé wrote:
>>> On 5/6/21 11:10 AM, Philippe Mathieu-Daudé wrote:
>>>> On 5/6/21 9:55 AM, Claudio Fontana wrote:
>>>>> On 5/5/21 9:31 PM, Eduardo Habkost wrote:
>>>>>> On Wed, May 05, 2021 at 02:15:29PM +0200, Philippe Mathieu-Daudé wrote:
>>>>>>> On 5/5/21 12:04 PM, Alex Bennée wrote:
>>>>>>>> Claudio Fontana <cfontana@suse.de> writes:
>>>>>>>>> On 3/8/21 3:02 PM, Alex Bennée wrote:
>>>>>>>>>> Claudio Fontana <cfontana@suse.de> writes:
>>>>>>>>>>
>>>>>>>>>>> Hi,
>>>>>>>>>>>
>>>>>>>>>>> anything else for me to do here?
>>>>>>>>>>
>>>>>>>>>> It looks to me that this series is looking pretty good. Every patch has
>>>>>>>>>> at least one review so I think it's just waiting on the maintainers to
>>>>>>>>>> pick it up.
>>>>>>>>>>
>>>>>>>>>> Paolo/Richard - are you intending to take the series as is or are you
>>>>>>>>>> waiting for something else? I'd like to see the patch delta reduced for
>>>>>>>>>> the ARM cleanup work which is still ongoing.
>>>>>>>>>
>>>>>>>>> I am a bit at a loss here, as this has been reviewed for a while, but nothing is happening.
>>>>>>>>> Rebasing is starting to become more and more draining;
>>>>>>>>
>>>>>>>> This is still the latest re-factor right?
>>>>>>>>
>>>>>>>>   Subject: [PATCH v28 00/23] i386 cleanup PART 2
>>>>>>>>   Date: Mon, 22 Mar 2021 14:27:36 +0100
>>>>>>>>   Message-Id: <20210322132800.7470-1-cfontana@suse.de>
>>>>>>>>
>>>>>>>>> I am seeing some changes from Phil that redo some of the patches here (like target arch user),
>>>>>>>>> maybe you would like to drive this?
>>>>>>>>
>>>>>>>> AIUI his changes where to get qtest passing.
>>>>>>>
>>>>>>> I hadn't read Claudio's mail, I think he's mentioning commit 46369b50ee3
>>>>>>>
>>>>>>>     meson: Introduce meson_user_arch source set for arch-specific user-mode
>>>>>>>
>>>>>>>     Similarly to the 'target_softmmu_arch' source set which allows
>>>>>>>     to restrict target-specific sources to system emulation, add
>>>>>>>     the equivalent 'target_user_arch' set for user emulation.
>>>>>>>
>>>>>>> The patch only contains 6 lines in 2 hunks, if it introduced a conflict
>>>>>>> it should be trivial to resolve (I wasn't expecting it to conflict with
>>>>>>> your work when I merged it TBH).
>>>>
>>>>> I'd suggest to move the discussions about the ARM series to the arm series thread.
>>>>>
>>>>> I am concerned here about the groundwork and x86 part.
>>>>
>>>> OK sorry, I was explaining the IRC chat.
>>>>
>>>> Is there any issue rebasing the groundwork on top of commit 46369b50ee3?
>>>>
>>>> Maybe my qtest/accel series is irrelevant to your x86 part, TBH I don't
>>>> remember.
>>>
>>> So far I could rebase this series on top of commit 229a834518b
>>> (Mon Mar 8 15:45:48 2021).
>>
>> Pfff I realized late you asked on the v26 series, but v28 is
>> available... Let's try again.
> 
> Hi Philippe,
> 
> I think that Paolo picked it up in the end?

Fortunately yes (for you; unfortunately for me... I thought you
pinged the latest series and tried to rebase it, and Paolo comment
on your v28 was "Looks good to me, thanks for all the effort!"
which was not obvious to me the series was already queued).

Anyway we are done with that part for good now, thanks.
Yes,



^ permalink raw reply	[flat|nested] 40+ messages in thread

end of thread, other threads:[~2021-05-19 18:20 UTC | newest]

Thread overview: 40+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-03-01  8:54 [PATCH v26 00/20] i386 cleanup PART 2 Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 01/20] target/i386: Rename helper_fldt, helper_fstt Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 02/20] target/i386: Split out do_fsave, do_frstor, do_fxsave, do_fxrstor Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 03/20] i386: split cpu accelerators from cpu.c, using AccelCPUClass Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 04/20] cpu: call AccelCPUClass::cpu_realizefn in cpu_exec_realizefn Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 05/20] accel: introduce new accessor functions Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 06/20] target/i386: fix host_cpu_adjust_phys_bits error handling Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 07/20] accel-cpu: make cpu_realizefn return a bool Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 08/20] meson: add target_user_arch Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 09/20] i386: split off sysemu-only functionality in tcg-cpu Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 10/20] i386: split smm helper (sysemu) Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 11/20] i386: split tcg excp_helper into sysemu and user parts Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 12/20] i386: move TCG bpt_helper into sysemu/ Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 13/20] i386: split misc helper user stubs and sysemu part Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 14/20] i386: separate fpu_helper sysemu-only parts Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 15/20] i386: split svm_helper into sysemu and stub-only user Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 16/20] i386: split seg_helper into user-only and sysemu parts Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 17/20] i386: split off sysemu part of cpu.c Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 18/20] target/i386: gdbstub: introduce aux functions to read/write CS64 regs Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 19/20] target/i386: gdbstub: only write CR0/CR2/CR3/EFER for sysemu Claudio Fontana
2021-03-01  8:54 ` [PATCH v26 20/20] i386: make cpu_load_efer sysemu-only Claudio Fontana
2021-03-01  9:18 ` [PATCH v26 00/20] i386 cleanup PART 2 no-reply
2021-03-08 12:57 ` Claudio Fontana
2021-03-08 13:27   ` Philippe Mathieu-Daudé
2021-03-08 13:52     ` Claudio Fontana
2021-03-08 14:57       ` Philippe Mathieu-Daudé
2021-03-08 15:03         ` Claudio Fontana
2021-03-08 15:07         ` Daniel P. Berrangé
2021-03-08 14:02   ` Alex Bennée
2021-05-05  8:27     ` Claudio Fontana
2021-05-05 10:04       ` Alex Bennée
2021-05-05 12:15         ` Philippe Mathieu-Daudé
2021-05-05 12:24           ` Philippe Mathieu-Daudé
2021-05-05 19:31           ` Eduardo Habkost
2021-05-06  7:55             ` Claudio Fontana
2021-05-06  9:10               ` Philippe Mathieu-Daudé
2021-05-06  9:33                 ` Philippe Mathieu-Daudé
2021-05-12  7:17                   ` Philippe Mathieu-Daudé
2021-05-17  6:17                     ` Claudio Fontana
2021-05-19 18:17                       ` Philippe Mathieu-Daudé

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).