* [PATCH 0/2] Move more code to softmmu/
@ 2020-10-06 9:19 Paolo Bonzini
2020-10-06 9:19 ` [PATCH 1/2] softmmu: move more files " Paolo Bonzini
2020-10-06 9:19 ` [PATCH 2/2] exec: split out non-softmmu-specific parts Paolo Bonzini
0 siblings, 2 replies; 14+ messages in thread
From: Paolo Bonzini @ 2020-10-06 9:19 UTC (permalink / raw)
To: qemu-devel
This small series moves more files to the softmmu/ directory.
Patch 1 covers files that can be moved completely, while
patch 2 splits exec.c into common and softmmu parts.
Based-on: <20201006072947.487729-1-pbonzini@redhat.com>
Paolo Bonzini (2):
softmmu: move more files to softmmu/
exec: split out non-softmmu-specific parts
MAINTAINERS | 4 +-
cpu.c | 452 ++++++++++++++++++++++
include/exec/cpu-common.h | 3 +
meson.build | 12 +-
bootdevice.c => softmmu/bootdevice.c | 0
device_tree.c => softmmu/device_tree.c | 0
dma-helpers.c => softmmu/dma-helpers.c | 0
softmmu/meson.build | 11 +
exec.c => softmmu/physmem.c | 454 +----------------------
qdev-monitor.c => softmmu/qdev-monitor.c | 0
qemu-seccomp.c => softmmu/qemu-seccomp.c | 0
tpm.c => softmmu/tpm.c | 0
12 files changed, 476 insertions(+), 460 deletions(-)
create mode 100644 cpu.c
rename bootdevice.c => softmmu/bootdevice.c (100%)
rename device_tree.c => softmmu/device_tree.c (100%)
rename dma-helpers.c => softmmu/dma-helpers.c (100%)
rename exec.c => softmmu/physmem.c (91%)
rename qdev-monitor.c => softmmu/qdev-monitor.c (100%)
rename qemu-seccomp.c => softmmu/qemu-seccomp.c (100%)
rename tpm.c => softmmu/tpm.c (100%)
--
2.26.2
^ permalink raw reply [flat|nested] 14+ messages in thread
* [PATCH 1/2] softmmu: move more files to softmmu/
2020-10-06 9:19 [PATCH 0/2] Move more code to softmmu/ Paolo Bonzini
@ 2020-10-06 9:19 ` Paolo Bonzini
2020-10-06 9:34 ` Philippe Mathieu-Daudé
2020-10-06 9:43 ` BALATON Zoltan via
2020-10-06 9:19 ` [PATCH 2/2] exec: split out non-softmmu-specific parts Paolo Bonzini
1 sibling, 2 replies; 14+ messages in thread
From: Paolo Bonzini @ 2020-10-06 9:19 UTC (permalink / raw)
To: qemu-devel
Keep most softmmu_ss files into the system-emulation-specific
directory.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
meson.build | 10 ----------
bootdevice.c => softmmu/bootdevice.c | 0
device_tree.c => softmmu/device_tree.c | 0
dma-helpers.c => softmmu/dma-helpers.c | 0
softmmu/meson.build | 10 ++++++++++
qdev-monitor.c => softmmu/qdev-monitor.c | 0
qemu-seccomp.c => softmmu/qemu-seccomp.c | 0
tpm.c => softmmu/tpm.c | 0
8 files changed, 10 insertions(+), 10 deletions(-)
rename bootdevice.c => softmmu/bootdevice.c (100%)
rename device_tree.c => softmmu/device_tree.c (100%)
rename dma-helpers.c => softmmu/dma-helpers.c (100%)
rename qdev-monitor.c => softmmu/qdev-monitor.c (100%)
rename qemu-seccomp.c => softmmu/qemu-seccomp.c (100%)
rename tpm.c => softmmu/tpm.c (100%)
diff --git a/meson.build b/meson.build
index 17c89c87c6..0e0577e81e 100644
--- a/meson.build
+++ b/meson.build
@@ -1360,17 +1360,7 @@ blockdev_ss.add(files(
# os-win32.c does not
blockdev_ss.add(when: 'CONFIG_POSIX', if_true: files('os-posix.c'))
softmmu_ss.add(when: 'CONFIG_WIN32', if_true: [files('os-win32.c')])
-
softmmu_ss.add_all(blockdev_ss)
-softmmu_ss.add(files(
- 'bootdevice.c',
- 'dma-helpers.c',
- 'qdev-monitor.c',
-), sdl)
-
-softmmu_ss.add(when: 'CONFIG_TPM', if_true: files('tpm.c'))
-softmmu_ss.add(when: 'CONFIG_SECCOMP', if_true: [files('qemu-seccomp.c'), seccomp])
-softmmu_ss.add(when: fdt, if_true: files('device_tree.c'))
common_ss.add(files('cpus-common.c'))
diff --git a/bootdevice.c b/softmmu/bootdevice.c
similarity index 100%
rename from bootdevice.c
rename to softmmu/bootdevice.c
diff --git a/device_tree.c b/softmmu/device_tree.c
similarity index 100%
rename from device_tree.c
rename to softmmu/device_tree.c
diff --git a/dma-helpers.c b/softmmu/dma-helpers.c
similarity index 100%
rename from dma-helpers.c
rename to softmmu/dma-helpers.c
diff --git a/softmmu/meson.build b/softmmu/meson.build
index 36c96e7b15..862ab24878 100644
--- a/softmmu/meson.build
+++ b/softmmu/meson.build
@@ -14,3 +14,13 @@ specific_ss.add(when: 'CONFIG_SOFTMMU', if_true: [files(
specific_ss.add(when: ['CONFIG_SOFTMMU', 'CONFIG_TCG'], if_true: [files(
'icount.c'
)])
+
+softmmu_ss.add(files(
+ 'bootdevice.c',
+ 'dma-helpers.c',
+ 'qdev-monitor.c',
+), sdl)
+
+softmmu_ss.add(when: 'CONFIG_TPM', if_true: files('tpm.c'))
+softmmu_ss.add(when: 'CONFIG_SECCOMP', if_true: [files('qemu-seccomp.c'), seccomp])
+softmmu_ss.add(when: fdt, if_true: files('device_tree.c'))
diff --git a/qdev-monitor.c b/softmmu/qdev-monitor.c
similarity index 100%
rename from qdev-monitor.c
rename to softmmu/qdev-monitor.c
diff --git a/qemu-seccomp.c b/softmmu/qemu-seccomp.c
similarity index 100%
rename from qemu-seccomp.c
rename to softmmu/qemu-seccomp.c
diff --git a/tpm.c b/softmmu/tpm.c
similarity index 100%
rename from tpm.c
rename to softmmu/tpm.c
--
2.26.2
^ permalink raw reply related [flat|nested] 14+ messages in thread
* [PATCH 2/2] exec: split out non-softmmu-specific parts
2020-10-06 9:19 [PATCH 0/2] Move more code to softmmu/ Paolo Bonzini
2020-10-06 9:19 ` [PATCH 1/2] softmmu: move more files " Paolo Bonzini
@ 2020-10-06 9:19 ` Paolo Bonzini
2020-10-06 9:44 ` Philippe Mathieu-Daudé
2020-10-08 7:47 ` Claudio Fontana
1 sibling, 2 replies; 14+ messages in thread
From: Paolo Bonzini @ 2020-10-06 9:19 UTC (permalink / raw)
To: qemu-devel
Over the years, most parts of exec.c that were not specific to softmmu
have been moved to accel/tcg; what's left is mostly the low-level part
of the memory API, which includes RAMBlock and AddressSpaceDispatch.
However exec.c also hosts 4-500 lines of code for the target specific
parts of the CPU QOM object, plus a few functions for user-mode
emulation that do not have a better place (they are not TCG-specific so
accel/tcg/user-exec.c is not a good place either).
Move these parts to a new file, so that exec.c can be moved to
softmmu/physmem.c.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
MAINTAINERS | 4 +-
cpu.c | 452 +++++++++++++++++++++++++++++++++++
include/exec/cpu-common.h | 3 +
meson.build | 2 +-
softmmu/meson.build | 3 +-
exec.c => softmmu/physmem.c | 454 +-----------------------------------
6 files changed, 467 insertions(+), 451 deletions(-)
create mode 100644 cpus.c
rename exec.c => softmmu/physmem.c (91%)
diff --git a/MAINTAINERS b/MAINTAINERS
index e9d85cc873..ddd1490767 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -117,7 +117,6 @@ R: Paolo Bonzini <pbonzini@redhat.com>
S: Maintained
F: softmmu/cpus.c
F: cpus-common.c
-F: exec.c
F: accel/tcg/
F: accel/stubs/tcg-stub.c
F: scripts/decodetree.py
@@ -1525,6 +1524,7 @@ Machine core
M: Eduardo Habkost <ehabkost@redhat.com>
M: Marcel Apfelbaum <marcel.apfelbaum@gmail.com>
S: Supported
+F: cpu.c
F: hw/core/cpu.c
F: hw/core/machine-qmp-cmds.c
F: hw/core/machine.c
@@ -2283,8 +2283,8 @@ F: include/exec/ram_addr.h
F: include/exec/ramblock.h
F: softmmu/ioport.c
F: softmmu/memory.c
+F: softmmu/physmem.c
F: include/exec/memory-internal.h
-F: exec.c
F: scripts/coccinelle/memory-region-housekeeping.cocci
SPICE
diff --git a/cpu.c b/cpu.c
new file mode 100644
index 0000000000..0be5dcb6f3
--- /dev/null
+++ b/cpu.c
@@ -0,0 +1,452 @@
+/*
+ * Target-specific parts of the CPU object
+ *
+ * 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 "qemu-common.h"
+#include "qapi/error.h"
+
+#include "exec/target_page.h"
+#include "hw/qdev-core.h"
+#include "hw/qdev-properties.h"
+#include "qemu/error-report.h"
+#include "migration/vmstate.h"
+#ifdef CONFIG_USER_ONLY
+#include "qemu.h"
+#else
+#include "exec/address-spaces.h"
+#endif
+#include "sysemu/tcg.h"
+#include "sysemu/kvm.h"
+#include "sysemu/replay.h"
+#include "translate-all.h"
+#include "exec/log.h"
+
+uintptr_t qemu_host_page_size;
+intptr_t qemu_host_page_mask;
+
+#ifndef CONFIG_USER_ONLY
+static int cpu_common_post_load(void *opaque, int version_id)
+{
+ CPUState *cpu = opaque;
+
+ /* 0x01 was CPU_INTERRUPT_EXIT. This line can be removed when the
+ version_id is increased. */
+ cpu->interrupt_request &= ~0x01;
+ tlb_flush(cpu);
+
+ /* loadvm has just updated the content of RAM, bypassing the
+ * usual mechanisms that ensure we flush TBs for writes to
+ * memory we've translated code from. So we must flush all TBs,
+ * which will now be stale.
+ */
+ tb_flush(cpu);
+
+ return 0;
+}
+
+static int cpu_common_pre_load(void *opaque)
+{
+ CPUState *cpu = opaque;
+
+ cpu->exception_index = -1;
+
+ return 0;
+}
+
+static bool cpu_common_exception_index_needed(void *opaque)
+{
+ CPUState *cpu = opaque;
+
+ return tcg_enabled() && cpu->exception_index != -1;
+}
+
+static const VMStateDescription vmstate_cpu_common_exception_index = {
+ .name = "cpu_common/exception_index",
+ .version_id = 1,
+ .minimum_version_id = 1,
+ .needed = cpu_common_exception_index_needed,
+ .fields = (VMStateField[]) {
+ VMSTATE_INT32(exception_index, CPUState),
+ VMSTATE_END_OF_LIST()
+ }
+};
+
+static bool cpu_common_crash_occurred_needed(void *opaque)
+{
+ CPUState *cpu = opaque;
+
+ return cpu->crash_occurred;
+}
+
+static const VMStateDescription vmstate_cpu_common_crash_occurred = {
+ .name = "cpu_common/crash_occurred",
+ .version_id = 1,
+ .minimum_version_id = 1,
+ .needed = cpu_common_crash_occurred_needed,
+ .fields = (VMStateField[]) {
+ VMSTATE_BOOL(crash_occurred, CPUState),
+ VMSTATE_END_OF_LIST()
+ }
+};
+
+const VMStateDescription vmstate_cpu_common = {
+ .name = "cpu_common",
+ .version_id = 1,
+ .minimum_version_id = 1,
+ .pre_load = cpu_common_pre_load,
+ .post_load = cpu_common_post_load,
+ .fields = (VMStateField[]) {
+ VMSTATE_UINT32(halted, CPUState),
+ VMSTATE_UINT32(interrupt_request, CPUState),
+ VMSTATE_END_OF_LIST()
+ },
+ .subsections = (const VMStateDescription*[]) {
+ &vmstate_cpu_common_exception_index,
+ &vmstate_cpu_common_crash_occurred,
+ NULL
+ }
+};
+#endif
+
+void cpu_exec_unrealizefn(CPUState *cpu)
+{
+ CPUClass *cc = CPU_GET_CLASS(cpu);
+
+ tlb_destroy(cpu);
+ cpu_list_remove(cpu);
+
+#ifdef CONFIG_USER_ONLY
+ assert(cc->vmsd == NULL);
+#else
+ if (cc->vmsd != NULL) {
+ vmstate_unregister(NULL, cc->vmsd, cpu);
+ }
+ if (qdev_get_vmsd(DEVICE(cpu)) == NULL) {
+ vmstate_unregister(NULL, &vmstate_cpu_common, cpu);
+ }
+ tcg_iommu_free_notifier_list(cpu);
+#endif
+}
+
+Property cpu_common_props[] = {
+#ifndef CONFIG_USER_ONLY
+ /* Create a memory property for softmmu CPU object,
+ * so users can wire up its memory. (This can't go in hw/core/cpu.c
+ * because that file is compiled only once for both user-mode
+ * and system builds.) The default if no link is set up is to use
+ * the system address space.
+ */
+ DEFINE_PROP_LINK("memory", CPUState, memory, TYPE_MEMORY_REGION,
+ MemoryRegion *),
+#endif
+ DEFINE_PROP_BOOL("start-powered-off", CPUState, start_powered_off, false),
+ DEFINE_PROP_END_OF_LIST(),
+};
+
+void cpu_exec_initfn(CPUState *cpu)
+{
+ cpu->as = NULL;
+ cpu->num_ases = 0;
+
+#ifndef CONFIG_USER_ONLY
+ cpu->thread_id = qemu_get_thread_id();
+ cpu->memory = get_system_memory();
+ object_ref(OBJECT(cpu->memory));
+#endif
+}
+
+void cpu_exec_realizefn(CPUState *cpu, Error **errp)
+{
+ CPUClass *cc = CPU_GET_CLASS(cpu);
+ static bool tcg_target_initialized;
+
+ cpu_list_add(cpu);
+
+ if (tcg_enabled() && !tcg_target_initialized) {
+ tcg_target_initialized = true;
+ cc->tcg_initialize();
+ }
+ tlb_init(cpu);
+
+ qemu_plugin_vcpu_init_hook(cpu);
+
+#ifdef CONFIG_USER_ONLY
+ assert(cc->vmsd == NULL);
+#else /* !CONFIG_USER_ONLY */
+ if (qdev_get_vmsd(DEVICE(cpu)) == NULL) {
+ vmstate_register(NULL, cpu->cpu_index, &vmstate_cpu_common, cpu);
+ }
+ if (cc->vmsd != NULL) {
+ vmstate_register(NULL, cpu->cpu_index, cc->vmsd, cpu);
+ }
+
+ tcg_iommu_init_notifier_list(cpu);
+#endif
+}
+
+const char *parse_cpu_option(const char *cpu_option)
+{
+ ObjectClass *oc;
+ CPUClass *cc;
+ gchar **model_pieces;
+ const char *cpu_type;
+
+ model_pieces = g_strsplit(cpu_option, ",", 2);
+ if (!model_pieces[0]) {
+ error_report("-cpu option cannot be empty");
+ exit(1);
+ }
+
+ oc = cpu_class_by_name(CPU_RESOLVING_TYPE, model_pieces[0]);
+ if (oc == NULL) {
+ error_report("unable to find CPU model '%s'", model_pieces[0]);
+ g_strfreev(model_pieces);
+ exit(EXIT_FAILURE);
+ }
+
+ cpu_type = object_class_get_name(oc);
+ cc = CPU_CLASS(oc);
+ cc->parse_features(cpu_type, model_pieces[1], &error_fatal);
+ g_strfreev(model_pieces);
+ return cpu_type;
+}
+
+#if defined(CONFIG_USER_ONLY)
+void tb_invalidate_phys_addr(target_ulong addr)
+{
+ mmap_lock();
+ tb_invalidate_phys_page_range(addr, addr + 1);
+ mmap_unlock();
+}
+
+static void breakpoint_invalidate(CPUState *cpu, target_ulong pc)
+{
+ tb_invalidate_phys_addr(pc);
+}
+#else
+void tb_invalidate_phys_addr(AddressSpace *as, hwaddr addr, MemTxAttrs attrs)
+{
+ ram_addr_t ram_addr;
+ MemoryRegion *mr;
+ hwaddr l = 1;
+
+ if (!tcg_enabled()) {
+ return;
+ }
+
+ RCU_READ_LOCK_GUARD();
+ mr = address_space_translate(as, addr, &addr, &l, false, attrs);
+ if (!(memory_region_is_ram(mr)
+ || memory_region_is_romd(mr))) {
+ return;
+ }
+ ram_addr = memory_region_get_ram_addr(mr) + addr;
+ tb_invalidate_phys_page_range(ram_addr, ram_addr + 1);
+}
+
+static void breakpoint_invalidate(CPUState *cpu, target_ulong pc)
+{
+ /*
+ * There may not be a virtual to physical translation for the pc
+ * right now, but there may exist cached TB for this pc.
+ * Flush the whole TB cache to force re-translation of such TBs.
+ * This is heavyweight, but we're debugging anyway.
+ */
+ tb_flush(cpu);
+}
+#endif
+
+/* Add a breakpoint. */
+int cpu_breakpoint_insert(CPUState *cpu, vaddr pc, int flags,
+ CPUBreakpoint **breakpoint)
+{
+ CPUBreakpoint *bp;
+
+ bp = g_malloc(sizeof(*bp));
+
+ bp->pc = pc;
+ bp->flags = flags;
+
+ /* keep all GDB-injected breakpoints in front */
+ if (flags & BP_GDB) {
+ QTAILQ_INSERT_HEAD(&cpu->breakpoints, bp, entry);
+ } else {
+ QTAILQ_INSERT_TAIL(&cpu->breakpoints, bp, entry);
+ }
+
+ breakpoint_invalidate(cpu, pc);
+
+ if (breakpoint) {
+ *breakpoint = bp;
+ }
+ return 0;
+}
+
+/* Remove a specific breakpoint. */
+int cpu_breakpoint_remove(CPUState *cpu, vaddr pc, int flags)
+{
+ CPUBreakpoint *bp;
+
+ QTAILQ_FOREACH(bp, &cpu->breakpoints, entry) {
+ if (bp->pc == pc && bp->flags == flags) {
+ cpu_breakpoint_remove_by_ref(cpu, bp);
+ return 0;
+ }
+ }
+ return -ENOENT;
+}
+
+/* Remove a specific breakpoint by reference. */
+void cpu_breakpoint_remove_by_ref(CPUState *cpu, CPUBreakpoint *breakpoint)
+{
+ QTAILQ_REMOVE(&cpu->breakpoints, breakpoint, entry);
+
+ breakpoint_invalidate(cpu, breakpoint->pc);
+
+ g_free(breakpoint);
+}
+
+/* Remove all matching breakpoints. */
+void cpu_breakpoint_remove_all(CPUState *cpu, int mask)
+{
+ CPUBreakpoint *bp, *next;
+
+ QTAILQ_FOREACH_SAFE(bp, &cpu->breakpoints, entry, next) {
+ if (bp->flags & mask) {
+ cpu_breakpoint_remove_by_ref(cpu, bp);
+ }
+ }
+}
+
+/* enable or disable single step mode. EXCP_DEBUG is returned by the
+ CPU loop after each instruction */
+void cpu_single_step(CPUState *cpu, int enabled)
+{
+ if (cpu->singlestep_enabled != enabled) {
+ cpu->singlestep_enabled = enabled;
+ if (kvm_enabled()) {
+ kvm_update_guest_debug(cpu, 0);
+ } else {
+ /* must flush all the translated code to avoid inconsistencies */
+ /* XXX: only flush what is necessary */
+ tb_flush(cpu);
+ }
+ }
+}
+
+void cpu_abort(CPUState *cpu, const char *fmt, ...)
+{
+ va_list ap;
+ va_list ap2;
+
+ va_start(ap, fmt);
+ va_copy(ap2, ap);
+ fprintf(stderr, "qemu: fatal: ");
+ vfprintf(stderr, fmt, ap);
+ fprintf(stderr, "\n");
+ cpu_dump_state(cpu, stderr, CPU_DUMP_FPU | CPU_DUMP_CCOP);
+ if (qemu_log_separate()) {
+ FILE *logfile = qemu_log_lock();
+ qemu_log("qemu: fatal: ");
+ qemu_log_vprintf(fmt, ap2);
+ qemu_log("\n");
+ log_cpu_state(cpu, CPU_DUMP_FPU | CPU_DUMP_CCOP);
+ qemu_log_flush();
+ qemu_log_unlock(logfile);
+ qemu_log_close();
+ }
+ va_end(ap2);
+ va_end(ap);
+ replay_finish();
+#if defined(CONFIG_USER_ONLY)
+ {
+ struct sigaction act;
+ sigfillset(&act.sa_mask);
+ act.sa_handler = SIG_DFL;
+ act.sa_flags = 0;
+ sigaction(SIGABRT, &act, NULL);
+ }
+#endif
+ abort();
+}
+
+/* physical memory access (slow version, mainly for debug) */
+#if defined(CONFIG_USER_ONLY)
+int cpu_memory_rw_debug(CPUState *cpu, target_ulong addr,
+ void *ptr, target_ulong len, bool is_write)
+{
+ int flags;
+ target_ulong l, page;
+ void * p;
+ uint8_t *buf = ptr;
+
+ while (len > 0) {
+ page = addr & TARGET_PAGE_MASK;
+ l = (page + TARGET_PAGE_SIZE) - addr;
+ if (l > len)
+ l = len;
+ flags = page_get_flags(page);
+ if (!(flags & PAGE_VALID))
+ return -1;
+ if (is_write) {
+ if (!(flags & PAGE_WRITE))
+ return -1;
+ /* XXX: this code should not depend on lock_user */
+ if (!(p = lock_user(VERIFY_WRITE, addr, l, 0)))
+ return -1;
+ memcpy(p, buf, l);
+ unlock_user(p, addr, l);
+ } else {
+ if (!(flags & PAGE_READ))
+ return -1;
+ /* XXX: this code should not depend on lock_user */
+ if (!(p = lock_user(VERIFY_READ, addr, l, 1)))
+ return -1;
+ memcpy(buf, p, l);
+ unlock_user(p, addr, 0);
+ }
+ len -= l;
+ buf += l;
+ addr += l;
+ }
+ return 0;
+}
+#endif
+
+bool target_words_bigendian(void)
+{
+#if defined(TARGET_WORDS_BIGENDIAN)
+ return true;
+#else
+ return false;
+#endif
+}
+
+void page_size_init(void)
+{
+ /* NOTE: we can always suppose that qemu_host_page_size >=
+ TARGET_PAGE_SIZE */
+ if (qemu_host_page_size == 0) {
+ qemu_host_page_size = qemu_real_host_page_size;
+ }
+ if (qemu_host_page_size < TARGET_PAGE_SIZE) {
+ qemu_host_page_size = TARGET_PAGE_SIZE;
+ }
+ qemu_host_page_mask = -(intptr_t)qemu_host_page_size;
+}
diff --git a/include/exec/cpu-common.h b/include/exec/cpu-common.h
index d5e285d2b5..19805ed6db 100644
--- a/include/exec/cpu-common.h
+++ b/include/exec/cpu-common.h
@@ -14,6 +14,9 @@ void cpu_list_unlock(void);
void tcg_flush_softmmu_tlb(CPUState *cs);
+void tcg_iommu_init_notifier_list(CPUState *cpu);
+void tcg_iommu_free_notifier_list(CPUState *cpu);
+
#if !defined(CONFIG_USER_ONLY)
enum device_endian {
diff --git a/meson.build b/meson.build
index 0e0577e81e..60df291b00 100644
--- a/meson.build
+++ b/meson.build
@@ -1367,7 +1367,7 @@ common_ss.add(files('cpus-common.c'))
subdir('softmmu')
common_ss.add(capstone)
-specific_ss.add(files('disas.c', 'exec.c', 'gdbstub.c'), capstone, libpmem, libdaxctl)
+specific_ss.add(files('cpu.c', 'disas.c', 'gdbstub.c'), capstone)
specific_ss.add(files('exec-vary.c'))
specific_ss.add(when: 'CONFIG_TCG', if_true: files(
'fpu/softfloat.c',
diff --git a/softmmu/meson.build b/softmmu/meson.build
index 862ab24878..8f7210b4f0 100644
--- a/softmmu/meson.build
+++ b/softmmu/meson.build
@@ -3,6 +3,7 @@ specific_ss.add(when: 'CONFIG_SOFTMMU', if_true: [files(
'balloon.c',
'cpus.c',
'cpu-throttle.c',
+ 'physmem.c',
'ioport.c',
'memory.c',
'memory_mapping.c',
@@ -19,7 +20,7 @@ softmmu_ss.add(files(
'bootdevice.c',
'dma-helpers.c',
'qdev-monitor.c',
-), sdl)
+), sdl, libpmem, libdaxctl)
softmmu_ss.add(when: 'CONFIG_TPM', if_true: files('tpm.c'))
softmmu_ss.add(when: 'CONFIG_SECCOMP', if_true: [files('qemu-seccomp.c'), seccomp])
diff --git a/exec.c b/softmmu/physmem.c
similarity index 91%
rename from exec.c
rename to softmmu/physmem.c
index ec68f4a9ca..6acd220f53 100644
--- a/exec.c
+++ b/softmmu/physmem.c
@@ -1,5 +1,5 @@
/*
- * Virtual page mapping
+ * RAM allocation and memory access
*
* Copyright (c) 2003 Fabrice Bellard
*
@@ -28,10 +28,8 @@
#include "tcg/tcg.h"
#include "hw/qdev-core.h"
#include "hw/qdev-properties.h"
-#if !defined(CONFIG_USER_ONLY)
#include "hw/boards.h"
#include "hw/xen/xen.h"
-#endif
#include "sysemu/kvm.h"
#include "sysemu/sysemu.h"
#include "sysemu/tcg.h"
@@ -40,9 +38,6 @@
#include "qemu/config-file.h"
#include "qemu/error-report.h"
#include "qemu/qemu-print.h"
-#if defined(CONFIG_USER_ONLY)
-#include "qemu.h"
-#else /* !CONFIG_USER_ONLY */
#include "exec/memory.h"
#include "exec/ioport.h"
#include "sysemu/dma.h"
@@ -56,7 +51,6 @@
#include <linux/falloc.h>
#endif
-#endif
#include "qemu/rcu_queue.h"
#include "qemu/main-loop.h"
#include "translate-all.h"
@@ -83,7 +77,6 @@
//#define DEBUG_SUBPAGE
-#if !defined(CONFIG_USER_ONLY)
/* ram_list is read under rcu_read_lock()/rcu_read_unlock(). Writes
* are protected by the ramlist lock.
*/
@@ -96,12 +89,6 @@ AddressSpace address_space_io;
AddressSpace address_space_memory;
static MemoryRegion io_mem_unassigned;
-#endif
-
-uintptr_t qemu_host_page_size;
-intptr_t qemu_host_page_mask;
-
-#if !defined(CONFIG_USER_ONLY)
typedef struct PhysPageEntry PhysPageEntry;
@@ -179,10 +166,6 @@ struct DirtyBitmapSnapshot {
unsigned long dirty[];
};
-#endif
-
-#if !defined(CONFIG_USER_ONLY)
-
static void phys_map_node_reserve(PhysPageMap *map, unsigned nodes)
{
static unsigned alloc_hint = 16;
@@ -666,7 +649,7 @@ static void tcg_register_iommu_notifier(CPUState *cpu,
}
}
-static void tcg_iommu_free_notifier_list(CPUState *cpu)
+void tcg_iommu_free_notifier_list(CPUState *cpu)
{
/* Destroy the CPU's notifier list */
int i;
@@ -680,6 +663,11 @@ static void tcg_iommu_free_notifier_list(CPUState *cpu)
g_array_free(cpu->iommu_notifiers, true);
}
+void tcg_iommu_init_notifier_list(CPUState *cpu)
+{
+ cpu->iommu_notifiers = g_array_new(false, true, sizeof(TCGIOMMUNotifier *));
+}
+
/* Called from RCU critical section */
MemoryRegionSection *
address_space_translate_for_iotlb(CPUState *cpu, int asidx, hwaddr addr,
@@ -737,91 +725,6 @@ address_space_translate_for_iotlb(CPUState *cpu, int asidx, hwaddr addr,
translate_fail:
return &d->map.sections[PHYS_SECTION_UNASSIGNED];
}
-#endif
-
-#if !defined(CONFIG_USER_ONLY)
-
-static int cpu_common_post_load(void *opaque, int version_id)
-{
- CPUState *cpu = opaque;
-
- /* 0x01 was CPU_INTERRUPT_EXIT. This line can be removed when the
- version_id is increased. */
- cpu->interrupt_request &= ~0x01;
- tlb_flush(cpu);
-
- /* loadvm has just updated the content of RAM, bypassing the
- * usual mechanisms that ensure we flush TBs for writes to
- * memory we've translated code from. So we must flush all TBs,
- * which will now be stale.
- */
- tb_flush(cpu);
-
- return 0;
-}
-
-static int cpu_common_pre_load(void *opaque)
-{
- CPUState *cpu = opaque;
-
- cpu->exception_index = -1;
-
- return 0;
-}
-
-static bool cpu_common_exception_index_needed(void *opaque)
-{
- CPUState *cpu = opaque;
-
- return tcg_enabled() && cpu->exception_index != -1;
-}
-
-static const VMStateDescription vmstate_cpu_common_exception_index = {
- .name = "cpu_common/exception_index",
- .version_id = 1,
- .minimum_version_id = 1,
- .needed = cpu_common_exception_index_needed,
- .fields = (VMStateField[]) {
- VMSTATE_INT32(exception_index, CPUState),
- VMSTATE_END_OF_LIST()
- }
-};
-
-static bool cpu_common_crash_occurred_needed(void *opaque)
-{
- CPUState *cpu = opaque;
-
- return cpu->crash_occurred;
-}
-
-static const VMStateDescription vmstate_cpu_common_crash_occurred = {
- .name = "cpu_common/crash_occurred",
- .version_id = 1,
- .minimum_version_id = 1,
- .needed = cpu_common_crash_occurred_needed,
- .fields = (VMStateField[]) {
- VMSTATE_BOOL(crash_occurred, CPUState),
- VMSTATE_END_OF_LIST()
- }
-};
-
-const VMStateDescription vmstate_cpu_common = {
- .name = "cpu_common",
- .version_id = 1,
- .minimum_version_id = 1,
- .pre_load = cpu_common_pre_load,
- .post_load = cpu_common_post_load,
- .fields = (VMStateField[]) {
- VMSTATE_UINT32(halted, CPUState),
- VMSTATE_UINT32(interrupt_request, CPUState),
- VMSTATE_END_OF_LIST()
- },
- .subsections = (const VMStateDescription*[]) {
- &vmstate_cpu_common_exception_index,
- &vmstate_cpu_common_crash_occurred,
- NULL
- }
-};
void cpu_address_space_init(CPUState *cpu, int asidx,
const char *prefix, MemoryRegion *mr)
@@ -865,155 +768,7 @@ AddressSpace *cpu_get_address_space(CPUState *cpu, int asidx)
/* Return the AddressSpace corresponding to the specified index */
return cpu->cpu_ases[asidx].as;
}
-#endif
-
-void cpu_exec_unrealizefn(CPUState *cpu)
-{
- CPUClass *cc = CPU_GET_CLASS(cpu);
-
- tlb_destroy(cpu);
- cpu_list_remove(cpu);
-
- if (cc->vmsd != NULL) {
- vmstate_unregister(NULL, cc->vmsd, cpu);
- }
- if (qdev_get_vmsd(DEVICE(cpu)) == NULL) {
- vmstate_unregister(NULL, &vmstate_cpu_common, cpu);
- }
-#ifndef CONFIG_USER_ONLY
- tcg_iommu_free_notifier_list(cpu);
-#endif
-}
-
-Property cpu_common_props[] = {
-#ifndef CONFIG_USER_ONLY
- /* Create a memory property for softmmu CPU object,
- * so users can wire up its memory. (This can't go in hw/core/cpu.c
- * because that file is compiled only once for both user-mode
- * and system builds.) The default if no link is set up is to use
- * the system address space.
- */
- DEFINE_PROP_LINK("memory", CPUState, memory, TYPE_MEMORY_REGION,
- MemoryRegion *),
-#endif
- DEFINE_PROP_BOOL("start-powered-off", CPUState, start_powered_off, false),
- DEFINE_PROP_END_OF_LIST(),
-};
-
-void cpu_exec_initfn(CPUState *cpu)
-{
- cpu->as = NULL;
- cpu->num_ases = 0;
-
-#ifndef CONFIG_USER_ONLY
- cpu->thread_id = qemu_get_thread_id();
- cpu->memory = system_memory;
- object_ref(OBJECT(cpu->memory));
-#endif
-}
-
-void cpu_exec_realizefn(CPUState *cpu, Error **errp)
-{
- CPUClass *cc = CPU_GET_CLASS(cpu);
- static bool tcg_target_initialized;
-
- cpu_list_add(cpu);
-
- if (tcg_enabled() && !tcg_target_initialized) {
- tcg_target_initialized = true;
- cc->tcg_initialize();
- }
- tlb_init(cpu);
-
- qemu_plugin_vcpu_init_hook(cpu);
-
-#ifdef CONFIG_USER_ONLY
- assert(cc->vmsd == NULL);
-#else /* !CONFIG_USER_ONLY */
- if (qdev_get_vmsd(DEVICE(cpu)) == NULL) {
- vmstate_register(NULL, cpu->cpu_index, &vmstate_cpu_common, cpu);
- }
- if (cc->vmsd != NULL) {
- vmstate_register(NULL, cpu->cpu_index, cc->vmsd, cpu);
- }
-
- cpu->iommu_notifiers = g_array_new(false, true, sizeof(TCGIOMMUNotifier *));
-#endif
-}
-
-const char *parse_cpu_option(const char *cpu_option)
-{
- ObjectClass *oc;
- CPUClass *cc;
- gchar **model_pieces;
- const char *cpu_type;
-
- model_pieces = g_strsplit(cpu_option, ",", 2);
- if (!model_pieces[0]) {
- error_report("-cpu option cannot be empty");
- exit(1);
- }
-
- oc = cpu_class_by_name(CPU_RESOLVING_TYPE, model_pieces[0]);
- if (oc == NULL) {
- error_report("unable to find CPU model '%s'", model_pieces[0]);
- g_strfreev(model_pieces);
- exit(EXIT_FAILURE);
- }
-
- cpu_type = object_class_get_name(oc);
- cc = CPU_CLASS(oc);
- cc->parse_features(cpu_type, model_pieces[1], &error_fatal);
- g_strfreev(model_pieces);
- return cpu_type;
-}
-
-#if defined(CONFIG_USER_ONLY)
-void tb_invalidate_phys_addr(target_ulong addr)
-{
- mmap_lock();
- tb_invalidate_phys_page_range(addr, addr + 1);
- mmap_unlock();
-}
-
-static void breakpoint_invalidate(CPUState *cpu, target_ulong pc)
-{
- tb_invalidate_phys_addr(pc);
-}
-#else
-void tb_invalidate_phys_addr(AddressSpace *as, hwaddr addr, MemTxAttrs attrs)
-{
- ram_addr_t ram_addr;
- MemoryRegion *mr;
- hwaddr l = 1;
-
- if (!tcg_enabled()) {
- return;
- }
-
- RCU_READ_LOCK_GUARD();
- mr = address_space_translate(as, addr, &addr, &l, false, attrs);
- if (!(memory_region_is_ram(mr)
- || memory_region_is_romd(mr))) {
- return;
- }
- ram_addr = memory_region_get_ram_addr(mr) + addr;
- tb_invalidate_phys_page_range(ram_addr, ram_addr + 1);
-}
-static void breakpoint_invalidate(CPUState *cpu, target_ulong pc)
-{
- /*
- * There may not be a virtual to physical translation for the pc
- * right now, but there may exist cached TB for this pc.
- * Flush the whole TB cache to force re-translation of such TBs.
- * This is heavyweight, but we're debugging anyway.
- */
- tb_flush(cpu);
-}
-#endif
-
-#ifndef CONFIG_USER_ONLY
/* Add a watchpoint. */
int cpu_watchpoint_insert(CPUState *cpu, vaddr addr, vaddr len,
int flags, CPUWatchpoint **watchpoint)
@@ -1122,123 +877,7 @@ int cpu_watchpoint_address_matches(CPUState *cpu, vaddr addr, vaddr len)
}
return ret;
}
-#endif /* !CONFIG_USER_ONLY */
-
-/* Add a breakpoint. */
-int cpu_breakpoint_insert(CPUState *cpu, vaddr pc, int flags,
- CPUBreakpoint **breakpoint)
-{
- CPUBreakpoint *bp;
-
- bp = g_malloc(sizeof(*bp));
-
- bp->pc = pc;
- bp->flags = flags;
-
- /* keep all GDB-injected breakpoints in front */
- if (flags & BP_GDB) {
- QTAILQ_INSERT_HEAD(&cpu->breakpoints, bp, entry);
- } else {
- QTAILQ_INSERT_TAIL(&cpu->breakpoints, bp, entry);
- }
-
- breakpoint_invalidate(cpu, pc);
-
- if (breakpoint) {
- *breakpoint = bp;
- }
- return 0;
-}
-
-/* Remove a specific breakpoint. */
-int cpu_breakpoint_remove(CPUState *cpu, vaddr pc, int flags)
-{
- CPUBreakpoint *bp;
-
- QTAILQ_FOREACH(bp, &cpu->breakpoints, entry) {
- if (bp->pc == pc && bp->flags == flags) {
- cpu_breakpoint_remove_by_ref(cpu, bp);
- return 0;
- }
- }
- return -ENOENT;
-}
-/* Remove a specific breakpoint by reference. */
-void cpu_breakpoint_remove_by_ref(CPUState *cpu, CPUBreakpoint *breakpoint)
-{
- QTAILQ_REMOVE(&cpu->breakpoints, breakpoint, entry);
-
- breakpoint_invalidate(cpu, breakpoint->pc);
-
- g_free(breakpoint);
-}
-
-/* Remove all matching breakpoints. */
-void cpu_breakpoint_remove_all(CPUState *cpu, int mask)
-{
- CPUBreakpoint *bp, *next;
-
- QTAILQ_FOREACH_SAFE(bp, &cpu->breakpoints, entry, next) {
- if (bp->flags & mask) {
- cpu_breakpoint_remove_by_ref(cpu, bp);
- }
- }
-}
-
-/* enable or disable single step mode. EXCP_DEBUG is returned by the
- CPU loop after each instruction */
-void cpu_single_step(CPUState *cpu, int enabled)
-{
- if (cpu->singlestep_enabled != enabled) {
- cpu->singlestep_enabled = enabled;
- if (kvm_enabled()) {
- kvm_update_guest_debug(cpu, 0);
- } else {
- /* must flush all the translated code to avoid inconsistencies */
- /* XXX: only flush what is necessary */
- tb_flush(cpu);
- }
- }
-}
-
-void cpu_abort(CPUState *cpu, const char *fmt, ...)
-{
- va_list ap;
- va_list ap2;
-
- va_start(ap, fmt);
- va_copy(ap2, ap);
- fprintf(stderr, "qemu: fatal: ");
- vfprintf(stderr, fmt, ap);
- fprintf(stderr, "\n");
- cpu_dump_state(cpu, stderr, CPU_DUMP_FPU | CPU_DUMP_CCOP);
- if (qemu_log_separate()) {
- FILE *logfile = qemu_log_lock();
- qemu_log("qemu: fatal: ");
- qemu_log_vprintf(fmt, ap2);
- qemu_log("\n");
- log_cpu_state(cpu, CPU_DUMP_FPU | CPU_DUMP_CCOP);
- qemu_log_flush();
- qemu_log_unlock(logfile);
- qemu_log_close();
- }
- va_end(ap2);
- va_end(ap);
- replay_finish();
-#if defined(CONFIG_USER_ONLY)
- {
- struct sigaction act;
- sigfillset(&act.sa_mask);
- act.sa_handler = SIG_DFL;
- act.sa_flags = 0;
- sigaction(SIGABRT, &act, NULL);
- }
-#endif
- abort();
-}
-
-#if !defined(CONFIG_USER_ONLY)
/* Called from RCU critical section */
static RAMBlock *qemu_get_ram_block(ram_addr_t addr)
{
@@ -1425,9 +1064,6 @@ hwaddr memory_region_section_get_iotlb(CPUState *cpu,
AddressSpaceDispatch *d = flatview_to_dispatch(section->fv);
return section - d->map.sections;
}
-#endif /* defined(CONFIG_USER_ONLY) */
-
-#if !defined(CONFIG_USER_ONLY)
static int subpage_register(subpage_t *mmio, uint32_t start, uint32_t end,
uint16_t section);
@@ -3028,52 +2664,6 @@ MemoryRegion *get_system_io(void)
return system_io;
}
-#endif /* !defined(CONFIG_USER_ONLY) */
-
-/* physical memory access (slow version, mainly for debug) */
-#if defined(CONFIG_USER_ONLY)
-int cpu_memory_rw_debug(CPUState *cpu, target_ulong addr,
- void *ptr, target_ulong len, bool is_write)
-{
- int flags;
- target_ulong l, page;
- void * p;
- uint8_t *buf = ptr;
-
- while (len > 0) {
- page = addr & TARGET_PAGE_MASK;
- l = (page + TARGET_PAGE_SIZE) - addr;
- if (l > len)
- l = len;
- flags = page_get_flags(page);
- if (!(flags & PAGE_VALID))
- return -1;
- if (is_write) {
- if (!(flags & PAGE_WRITE))
- return -1;
- /* XXX: this code should not depend on lock_user */
- if (!(p = lock_user(VERIFY_WRITE, addr, l, 0)))
- return -1;
- memcpy(p, buf, l);
- unlock_user(p, addr, l);
- } else {
- if (!(flags & PAGE_READ))
- return -1;
- /* XXX: this code should not depend on lock_user */
- if (!(p = lock_user(VERIFY_READ, addr, l, 1)))
- return -1;
- memcpy(buf, p, l);
- unlock_user(p, addr, 0);
- }
- len -= l;
- buf += l;
- addr += l;
- }
- return 0;
-}
-
-#else
-
static void invalidate_and_set_dirty(MemoryRegion *mr, hwaddr addr,
hwaddr length)
{
@@ -3862,18 +3452,7 @@ int qemu_target_page_bits_min(void)
{
return TARGET_PAGE_BITS_MIN;
}
-#endif
-
-bool target_words_bigendian(void)
-{
-#if defined(TARGET_WORDS_BIGENDIAN)
- return true;
-#else
- return false;
-#endif
-}
-#ifndef CONFIG_USER_ONLY
bool cpu_physical_memory_is_io(hwaddr phys_addr)
{
MemoryRegion*mr;
@@ -4003,23 +3582,6 @@ bool ramblock_is_pmem(RAMBlock *rb)
return rb->flags & RAM_PMEM;
}
-#endif
-
-void page_size_init(void)
-{
- /* NOTE: we can always suppose that qemu_host_page_size >=
- TARGET_PAGE_SIZE */
- if (qemu_host_page_size == 0) {
- qemu_host_page_size = qemu_real_host_page_size;
- }
- if (qemu_host_page_size < TARGET_PAGE_SIZE) {
- qemu_host_page_size = TARGET_PAGE_SIZE;
- }
- qemu_host_page_mask = -(intptr_t)qemu_host_page_size;
-}
-
-#if !defined(CONFIG_USER_ONLY)
-
static void mtree_print_phys_entries(int start, int end, int skip, int ptr)
{
if (start == end - 1) {
@@ -4152,5 +3714,3 @@ bool ram_block_discard_is_required(void)
{
return qatomic_read(&ram_block_discard_disabled) < 0;
}
-
-#endif
--
2.26.2
^ permalink raw reply related [flat|nested] 14+ messages in thread
* Re: [PATCH 1/2] softmmu: move more files to softmmu/
2020-10-06 9:19 ` [PATCH 1/2] softmmu: move more files " Paolo Bonzini
@ 2020-10-06 9:34 ` Philippe Mathieu-Daudé
2020-10-06 9:38 ` Philippe Mathieu-Daudé
2020-10-06 9:43 ` BALATON Zoltan via
1 sibling, 1 reply; 14+ messages in thread
From: Philippe Mathieu-Daudé @ 2020-10-06 9:34 UTC (permalink / raw)
To: Paolo Bonzini, qemu-devel
On 10/6/20 11:19 AM, Paolo Bonzini wrote:
> Keep most softmmu_ss files into the system-emulation-specific
> directory.
>
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
> meson.build | 10 ----------
> bootdevice.c => softmmu/bootdevice.c | 0
> device_tree.c => softmmu/device_tree.c | 0
> dma-helpers.c => softmmu/dma-helpers.c | 0
> softmmu/meson.build | 10 ++++++++++
> qdev-monitor.c => softmmu/qdev-monitor.c | 0
> qemu-seccomp.c => softmmu/qemu-seccomp.c | 0
> tpm.c => softmmu/tpm.c | 0
> 8 files changed, 10 insertions(+), 10 deletions(-)
> rename bootdevice.c => softmmu/bootdevice.c (100%)
> rename device_tree.c => softmmu/device_tree.c (100%)
> rename dma-helpers.c => softmmu/dma-helpers.c (100%)
> rename qdev-monitor.c => softmmu/qdev-monitor.c (100%)
TIL qdev-monitor.c :)
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> rename qemu-seccomp.c => softmmu/qemu-seccomp.c (100%)
> rename tpm.c => softmmu/tpm.c (100%)
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH 1/2] softmmu: move more files to softmmu/
2020-10-06 9:34 ` Philippe Mathieu-Daudé
@ 2020-10-06 9:38 ` Philippe Mathieu-Daudé
0 siblings, 0 replies; 14+ messages in thread
From: Philippe Mathieu-Daudé @ 2020-10-06 9:38 UTC (permalink / raw)
To: Paolo Bonzini, qemu-devel
On 10/6/20 11:34 AM, Philippe Mathieu-Daudé wrote:
> On 10/6/20 11:19 AM, Paolo Bonzini wrote:
>> Keep most softmmu_ss files into the system-emulation-specific
>> directory.
Oops, you forgot to update MAINTAINERS.
>>
>> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
>> ---
>> meson.build | 10 ----------
>> bootdevice.c => softmmu/bootdevice.c | 0
>> device_tree.c => softmmu/device_tree.c | 0
>> dma-helpers.c => softmmu/dma-helpers.c | 0
>> softmmu/meson.build | 10 ++++++++++
>> qdev-monitor.c => softmmu/qdev-monitor.c | 0
>> qemu-seccomp.c => softmmu/qemu-seccomp.c | 0
>> tpm.c => softmmu/tpm.c | 0
>> 8 files changed, 10 insertions(+), 10 deletions(-)
>> rename bootdevice.c => softmmu/bootdevice.c (100%)
"Gonglei (Arei)" <arei.gonglei@huawei.com> (maintainer:Bootdevice)
>> rename device_tree.c => softmmu/device_tree.c (100%)
Alistair Francis <alistair.francis@wdc.com> (maintainer:Device Tree)
David Gibson <david@gibson.dropbear.id.au> (reviewer:Device Tree)
>> rename dma-helpers.c => softmmu/dma-helpers.c (100%)
No maintainers.
>> rename qdev-monitor.c => softmmu/qdev-monitor.c (100%)
Paolo Bonzini <pbonzini@redhat.com> (supporter:QOM)
"Daniel P. Berrangé" <berrange@redhat.com> (reviewer:QOM)
Eduardo Habkost <ehabkost@redhat.com> (reviewer:QOM)
>
> TIL qdev-monitor.c :)
>
> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
>
>> rename qemu-seccomp.c => softmmu/qemu-seccomp.c (100%)
Eduardo Otubo <otubo@redhat.com> (supporter:Seccomp)
>> rename tpm.c => softmmu/tpm.c (100%)
Stefan Berger <stefanb@linux.ibm.com> (maintainer:TPM)
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH 1/2] softmmu: move more files to softmmu/
2020-10-06 9:19 ` [PATCH 1/2] softmmu: move more files " Paolo Bonzini
2020-10-06 9:34 ` Philippe Mathieu-Daudé
@ 2020-10-06 9:43 ` BALATON Zoltan via
2020-10-06 9:54 ` Paolo Bonzini
1 sibling, 1 reply; 14+ messages in thread
From: BALATON Zoltan via @ 2020-10-06 9:43 UTC (permalink / raw)
To: Paolo Bonzini; +Cc: qemu-devel
On Tue, 6 Oct 2020, Paolo Bonzini wrote:
> Keep most softmmu_ss files into the system-emulation-specific
> directory.
The name of this dir may be misleading. I think it originally stood for
the actual MMU emulation but now it seems everything related to system
emulation is dumped here. Is it better to keep MMU emulation separate and
put other files in a "sysemu" dir or rename this dir if it keeps mixing
MMU emulation and system emulation parts? (I think the MMU emulation is a
weak part of QEMU regarding performance so it would be better to keep that
cleanly separated so that it's easier to analyse and optimise it in the
future, which is more difficult if it's mixed with other parts where it's
not even clear what constitutes the actual MMU emulation. So I vote for
keeping it separate.)
Regards,
BALATON Zoltan
>
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
> meson.build | 10 ----------
> bootdevice.c => softmmu/bootdevice.c | 0
> device_tree.c => softmmu/device_tree.c | 0
> dma-helpers.c => softmmu/dma-helpers.c | 0
> softmmu/meson.build | 10 ++++++++++
> qdev-monitor.c => softmmu/qdev-monitor.c | 0
> qemu-seccomp.c => softmmu/qemu-seccomp.c | 0
> tpm.c => softmmu/tpm.c | 0
> 8 files changed, 10 insertions(+), 10 deletions(-)
> rename bootdevice.c => softmmu/bootdevice.c (100%)
> rename device_tree.c => softmmu/device_tree.c (100%)
> rename dma-helpers.c => softmmu/dma-helpers.c (100%)
> rename qdev-monitor.c => softmmu/qdev-monitor.c (100%)
> rename qemu-seccomp.c => softmmu/qemu-seccomp.c (100%)
> rename tpm.c => softmmu/tpm.c (100%)
>
> diff --git a/meson.build b/meson.build
> index 17c89c87c6..0e0577e81e 100644
> --- a/meson.build
> +++ b/meson.build
> @@ -1360,17 +1360,7 @@ blockdev_ss.add(files(
> # os-win32.c does not
> blockdev_ss.add(when: 'CONFIG_POSIX', if_true: files('os-posix.c'))
> softmmu_ss.add(when: 'CONFIG_WIN32', if_true: [files('os-win32.c')])
> -
> softmmu_ss.add_all(blockdev_ss)
> -softmmu_ss.add(files(
> - 'bootdevice.c',
> - 'dma-helpers.c',
> - 'qdev-monitor.c',
> -), sdl)
> -
> -softmmu_ss.add(when: 'CONFIG_TPM', if_true: files('tpm.c'))
> -softmmu_ss.add(when: 'CONFIG_SECCOMP', if_true: [files('qemu-seccomp.c'), seccomp])
> -softmmu_ss.add(when: fdt, if_true: files('device_tree.c'))
>
> common_ss.add(files('cpus-common.c'))
>
> diff --git a/bootdevice.c b/softmmu/bootdevice.c
> similarity index 100%
> rename from bootdevice.c
> rename to softmmu/bootdevice.c
> diff --git a/device_tree.c b/softmmu/device_tree.c
> similarity index 100%
> rename from device_tree.c
> rename to softmmu/device_tree.c
> diff --git a/dma-helpers.c b/softmmu/dma-helpers.c
> similarity index 100%
> rename from dma-helpers.c
> rename to softmmu/dma-helpers.c
> diff --git a/softmmu/meson.build b/softmmu/meson.build
> index 36c96e7b15..862ab24878 100644
> --- a/softmmu/meson.build
> +++ b/softmmu/meson.build
> @@ -14,3 +14,13 @@ specific_ss.add(when: 'CONFIG_SOFTMMU', if_true: [files(
> specific_ss.add(when: ['CONFIG_SOFTMMU', 'CONFIG_TCG'], if_true: [files(
> 'icount.c'
> )])
> +
> +softmmu_ss.add(files(
> + 'bootdevice.c',
> + 'dma-helpers.c',
> + 'qdev-monitor.c',
> +), sdl)
> +
> +softmmu_ss.add(when: 'CONFIG_TPM', if_true: files('tpm.c'))
> +softmmu_ss.add(when: 'CONFIG_SECCOMP', if_true: [files('qemu-seccomp.c'), seccomp])
> +softmmu_ss.add(when: fdt, if_true: files('device_tree.c'))
> diff --git a/qdev-monitor.c b/softmmu/qdev-monitor.c
> similarity index 100%
> rename from qdev-monitor.c
> rename to softmmu/qdev-monitor.c
> diff --git a/qemu-seccomp.c b/softmmu/qemu-seccomp.c
> similarity index 100%
> rename from qemu-seccomp.c
> rename to softmmu/qemu-seccomp.c
> diff --git a/tpm.c b/softmmu/tpm.c
> similarity index 100%
> rename from tpm.c
> rename to softmmu/tpm.c
>
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH 2/2] exec: split out non-softmmu-specific parts
2020-10-06 9:19 ` [PATCH 2/2] exec: split out non-softmmu-specific parts Paolo Bonzini
@ 2020-10-06 9:44 ` Philippe Mathieu-Daudé
2020-10-08 7:47 ` Claudio Fontana
1 sibling, 0 replies; 14+ messages in thread
From: Philippe Mathieu-Daudé @ 2020-10-06 9:44 UTC (permalink / raw)
To: Paolo Bonzini, qemu-devel
On 10/6/20 11:19 AM, Paolo Bonzini wrote:
> Over the years, most parts of exec.c that were not specific to softmmu
> have been moved to accel/tcg; what's left is mostly the low-level part
> of the memory API, which includes RAMBlock and AddressSpaceDispatch.
> However exec.c also hosts 4-500 lines of code for the target specific
> parts of the CPU QOM object, plus a few functions for user-mode
> emulation that do not have a better place (they are not TCG-specific so
> accel/tcg/user-exec.c is not a good place either).
>
> Move these parts to a new file, so that exec.c can be moved to
> softmmu/physmem.c.
>
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
> MAINTAINERS | 4 +-
> cpu.c | 452 +++++++++++++++++++++++++++++++++++
> include/exec/cpu-common.h | 3 +
> meson.build | 2 +-
> softmmu/meson.build | 3 +-
> exec.c => softmmu/physmem.c | 454 +-----------------------------------
> 6 files changed, 467 insertions(+), 451 deletions(-)
> create mode 100644 cpus.c
> rename exec.c => softmmu/physmem.c (91%)
I suppose it would be easier to review split in 2 patches,
first extract cpu.c, then move to physmem.c, but it is
now reviewed.
[...]
> diff --git a/exec.c b/softmmu/physmem.c
> similarity index 91%
> rename from exec.c
> rename to softmmu/physmem.c
> index ec68f4a9ca..6acd220f53 100644
> --- a/exec.c
> +++ b/softmmu/physmem.c
> @@ -1,5 +1,5 @@
> /*
> - * Virtual page mapping
> + * RAM allocation and memory access
[...]
This comment and the one from softmmu/memory.c are confusing:
/*
* Physical memory management
*
...
Otherwise:
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH 1/2] softmmu: move more files to softmmu/
2020-10-06 9:43 ` BALATON Zoltan via
@ 2020-10-06 9:54 ` Paolo Bonzini
2020-10-06 10:02 ` Philippe Mathieu-Daudé
0 siblings, 1 reply; 14+ messages in thread
From: Paolo Bonzini @ 2020-10-06 9:54 UTC (permalink / raw)
To: BALATON Zoltan; +Cc: qemu-devel
On 06/10/20 11:43, BALATON Zoltan wrote:
>
> The name of this dir may be misleading. I think it originally stood
> for the actual MMU emulation but now it seems everything related to
> system emulation is dumped here. Is it better to keep MMU emulation
> separate and put other files in a "sysemu" dir or rename this dir if
> it keeps mixing MMU emulation and system emulation parts? (I think
> the MMU emulation is a weak part of QEMU regarding performance so it
> would be better to keep that cleanly separated so that it's easier to
> analyse and optimise it in the future, which is more difficult if
> it's mixed with other parts where it's not even clear what
> constitutes the actual MMU emulation. So I vote for keeping it
> separate.)
I agree that softmmu/ should have been called sysemu/. But I think it's
unnecessary churn to do that now.
Paolo
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH 1/2] softmmu: move more files to softmmu/
2020-10-06 9:54 ` Paolo Bonzini
@ 2020-10-06 10:02 ` Philippe Mathieu-Daudé
0 siblings, 0 replies; 14+ messages in thread
From: Philippe Mathieu-Daudé @ 2020-10-06 10:02 UTC (permalink / raw)
To: Paolo Bonzini, BALATON Zoltan; +Cc: qemu-devel
On 10/6/20 11:54 AM, Paolo Bonzini wrote:
> On 06/10/20 11:43, BALATON Zoltan wrote:
>>
>> The name of this dir may be misleading. I think it originally stood
>> for the actual MMU emulation but now it seems everything related to
>> system emulation is dumped here. Is it better to keep MMU emulation
>> separate and put other files in a "sysemu" dir or rename this dir if
>> it keeps mixing MMU emulation and system emulation parts? (I think
>> the MMU emulation is a weak part of QEMU regarding performance so it
>> would be better to keep that cleanly separated so that it's easier to
>> analyse and optimise it in the future, which is more difficult if
>> it's mixed with other parts where it's not even clear what
>> constitutes the actual MMU emulation. So I vote for keeping it
>> separate.)
>
> I agree that softmmu/ should have been called sysemu/. But I think it's
> unnecessary churn to do that now.
It is not necessary, but there are so many files moved around recently
so it is not a bad time to clarify it neither :)
OTOH yes many files have been moved around recently, so we might wait
for the next release before attacking this cleanup.
>
> Paolo
>
>
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH 2/2] exec: split out non-softmmu-specific parts
2020-10-06 9:19 ` [PATCH 2/2] exec: split out non-softmmu-specific parts Paolo Bonzini
2020-10-06 9:44 ` Philippe Mathieu-Daudé
@ 2020-10-08 7:47 ` Claudio Fontana
2020-10-08 7:56 ` Paolo Bonzini
1 sibling, 1 reply; 14+ messages in thread
From: Claudio Fontana @ 2020-10-08 7:47 UTC (permalink / raw)
To: Paolo Bonzini, qemu-devel
On 10/6/20 11:19 AM, Paolo Bonzini wrote:
> Over the years, most parts of exec.c that were not specific to softmmu
> have been moved to accel/tcg; what's left is mostly the low-level part
> of the memory API, which includes RAMBlock and AddressSpaceDispatch.
> However exec.c also hosts 4-500 lines of code for the target specific
> parts of the CPU QOM object, plus a few functions for user-mode
> emulation that do not have a better place (they are not TCG-specific so
> accel/tcg/user-exec.c is not a good place either).
>
> Move these parts to a new file, so that exec.c can be moved to
> softmmu/physmem.c.
>
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Hi Paolo,
the comment does not talk about cpu.c, which is now created in the top source directory.
What is the role of this new module?
Also, could we find a more descriptive file name than cpu.c?
Do you plan further renaming of this new module functions?
Or its this basically a "leftovers" file for which we did not find a proper role yet?
Ciao,
CLaudio
> ---
> MAINTAINERS | 4 +-
> cpu.c | 452 +++++++++++++++++++++++++++++++++++
> include/exec/cpu-common.h | 3 +
> meson.build | 2 +-
> softmmu/meson.build | 3 +-
> exec.c => softmmu/physmem.c | 454 +-----------------------------------
> 6 files changed, 467 insertions(+), 451 deletions(-)
> create mode 100644 cpus.c
> rename exec.c => softmmu/physmem.c (91%)
>
> diff --git a/MAINTAINERS b/MAINTAINERS
> index e9d85cc873..ddd1490767 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -117,7 +117,6 @@ R: Paolo Bonzini <pbonzini@redhat.com>
> S: Maintained
> F: softmmu/cpus.c
> F: cpus-common.c
> -F: exec.c
> F: accel/tcg/
> F: accel/stubs/tcg-stub.c
> F: scripts/decodetree.py
> @@ -1525,6 +1524,7 @@ Machine core
> M: Eduardo Habkost <ehabkost@redhat.com>
> M: Marcel Apfelbaum <marcel.apfelbaum@gmail.com>
> S: Supported
> +F: cpu.c
> F: hw/core/cpu.c
> F: hw/core/machine-qmp-cmds.c
> F: hw/core/machine.c
> @@ -2283,8 +2283,8 @@ F: include/exec/ram_addr.h
> F: include/exec/ramblock.h
> F: softmmu/ioport.c
> F: softmmu/memory.c
> +F: softmmu/physmem.c
> F: include/exec/memory-internal.h
> -F: exec.c
> F: scripts/coccinelle/memory-region-housekeeping.cocci
>
> SPICE
> diff --git a/cpu.c b/cpu.c
> new file mode 100644
> index 0000000000..0be5dcb6f3
> --- /dev/null
> +++ b/cpu.c
> @@ -0,0 +1,452 @@
> +/*
> + * Target-specific parts of the CPU object
> + *
> + * 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 "qemu-common.h"
> +#include "qapi/error.h"
> +
> +#include "exec/target_page.h"
> +#include "hw/qdev-core.h"
> +#include "hw/qdev-properties.h"
> +#include "qemu/error-report.h"
> +#include "migration/vmstate.h"
> +#ifdef CONFIG_USER_ONLY
> +#include "qemu.h"
> +#else
> +#include "exec/address-spaces.h"
> +#endif
> +#include "sysemu/tcg.h"
> +#include "sysemu/kvm.h"
> +#include "sysemu/replay.h"
> +#include "translate-all.h"
> +#include "exec/log.h"
> +
> +uintptr_t qemu_host_page_size;
> +intptr_t qemu_host_page_mask;
> +
> +#ifndef CONFIG_USER_ONLY
> +static int cpu_common_post_load(void *opaque, int version_id)
> +{
> + CPUState *cpu = opaque;
> +
> + /* 0x01 was CPU_INTERRUPT_EXIT. This line can be removed when the
> + version_id is increased. */
> + cpu->interrupt_request &= ~0x01;
> + tlb_flush(cpu);
> +
> + /* loadvm has just updated the content of RAM, bypassing the
> + * usual mechanisms that ensure we flush TBs for writes to
> + * memory we've translated code from. So we must flush all TBs,
> + * which will now be stale.
> + */
> + tb_flush(cpu);
> +
> + return 0;
> +}
> +
> +static int cpu_common_pre_load(void *opaque)
> +{
> + CPUState *cpu = opaque;
> +
> + cpu->exception_index = -1;
> +
> + return 0;
> +}
> +
> +static bool cpu_common_exception_index_needed(void *opaque)
> +{
> + CPUState *cpu = opaque;
> +
> + return tcg_enabled() && cpu->exception_index != -1;
> +}
> +
> +static const VMStateDescription vmstate_cpu_common_exception_index = {
> + .name = "cpu_common/exception_index",
> + .version_id = 1,
> + .minimum_version_id = 1,
> + .needed = cpu_common_exception_index_needed,
> + .fields = (VMStateField[]) {
> + VMSTATE_INT32(exception_index, CPUState),
> + VMSTATE_END_OF_LIST()
> + }
> +};
> +
> +static bool cpu_common_crash_occurred_needed(void *opaque)
> +{
> + CPUState *cpu = opaque;
> +
> + return cpu->crash_occurred;
> +}
> +
> +static const VMStateDescription vmstate_cpu_common_crash_occurred = {
> + .name = "cpu_common/crash_occurred",
> + .version_id = 1,
> + .minimum_version_id = 1,
> + .needed = cpu_common_crash_occurred_needed,
> + .fields = (VMStateField[]) {
> + VMSTATE_BOOL(crash_occurred, CPUState),
> + VMSTATE_END_OF_LIST()
> + }
> +};
> +
> +const VMStateDescription vmstate_cpu_common = {
> + .name = "cpu_common",
> + .version_id = 1,
> + .minimum_version_id = 1,
> + .pre_load = cpu_common_pre_load,
> + .post_load = cpu_common_post_load,
> + .fields = (VMStateField[]) {
> + VMSTATE_UINT32(halted, CPUState),
> + VMSTATE_UINT32(interrupt_request, CPUState),
> + VMSTATE_END_OF_LIST()
> + },
> + .subsections = (const VMStateDescription*[]) {
> + &vmstate_cpu_common_exception_index,
> + &vmstate_cpu_common_crash_occurred,
> + NULL
> + }
> +};
> +#endif
> +
> +void cpu_exec_unrealizefn(CPUState *cpu)
> +{
> + CPUClass *cc = CPU_GET_CLASS(cpu);
> +
> + tlb_destroy(cpu);
> + cpu_list_remove(cpu);
> +
> +#ifdef CONFIG_USER_ONLY
> + assert(cc->vmsd == NULL);
> +#else
> + if (cc->vmsd != NULL) {
> + vmstate_unregister(NULL, cc->vmsd, cpu);
> + }
> + if (qdev_get_vmsd(DEVICE(cpu)) == NULL) {
> + vmstate_unregister(NULL, &vmstate_cpu_common, cpu);
> + }
> + tcg_iommu_free_notifier_list(cpu);
> +#endif
> +}
> +
> +Property cpu_common_props[] = {
> +#ifndef CONFIG_USER_ONLY
> + /* Create a memory property for softmmu CPU object,
> + * so users can wire up its memory. (This can't go in hw/core/cpu.c
> + * because that file is compiled only once for both user-mode
> + * and system builds.) The default if no link is set up is to use
> + * the system address space.
> + */
> + DEFINE_PROP_LINK("memory", CPUState, memory, TYPE_MEMORY_REGION,
> + MemoryRegion *),
> +#endif
> + DEFINE_PROP_BOOL("start-powered-off", CPUState, start_powered_off, false),
> + DEFINE_PROP_END_OF_LIST(),
> +};
> +
> +void cpu_exec_initfn(CPUState *cpu)
> +{
> + cpu->as = NULL;
> + cpu->num_ases = 0;
> +
> +#ifndef CONFIG_USER_ONLY
> + cpu->thread_id = qemu_get_thread_id();
> + cpu->memory = get_system_memory();
> + object_ref(OBJECT(cpu->memory));
> +#endif
> +}
> +
> +void cpu_exec_realizefn(CPUState *cpu, Error **errp)
> +{
> + CPUClass *cc = CPU_GET_CLASS(cpu);
> + static bool tcg_target_initialized;
> +
> + cpu_list_add(cpu);
> +
> + if (tcg_enabled() && !tcg_target_initialized) {
> + tcg_target_initialized = true;
> + cc->tcg_initialize();
> + }
> + tlb_init(cpu);
> +
> + qemu_plugin_vcpu_init_hook(cpu);
> +
> +#ifdef CONFIG_USER_ONLY
> + assert(cc->vmsd == NULL);
> +#else /* !CONFIG_USER_ONLY */
> + if (qdev_get_vmsd(DEVICE(cpu)) == NULL) {
> + vmstate_register(NULL, cpu->cpu_index, &vmstate_cpu_common, cpu);
> + }
> + if (cc->vmsd != NULL) {
> + vmstate_register(NULL, cpu->cpu_index, cc->vmsd, cpu);
> + }
> +
> + tcg_iommu_init_notifier_list(cpu);
> +#endif
> +}
> +
> +const char *parse_cpu_option(const char *cpu_option)
> +{
> + ObjectClass *oc;
> + CPUClass *cc;
> + gchar **model_pieces;
> + const char *cpu_type;
> +
> + model_pieces = g_strsplit(cpu_option, ",", 2);
> + if (!model_pieces[0]) {
> + error_report("-cpu option cannot be empty");
> + exit(1);
> + }
> +
> + oc = cpu_class_by_name(CPU_RESOLVING_TYPE, model_pieces[0]);
> + if (oc == NULL) {
> + error_report("unable to find CPU model '%s'", model_pieces[0]);
> + g_strfreev(model_pieces);
> + exit(EXIT_FAILURE);
> + }
> +
> + cpu_type = object_class_get_name(oc);
> + cc = CPU_CLASS(oc);
> + cc->parse_features(cpu_type, model_pieces[1], &error_fatal);
> + g_strfreev(model_pieces);
> + return cpu_type;
> +}
> +
> +#if defined(CONFIG_USER_ONLY)
> +void tb_invalidate_phys_addr(target_ulong addr)
> +{
> + mmap_lock();
> + tb_invalidate_phys_page_range(addr, addr + 1);
> + mmap_unlock();
> +}
> +
> +static void breakpoint_invalidate(CPUState *cpu, target_ulong pc)
> +{
> + tb_invalidate_phys_addr(pc);
> +}
> +#else
> +void tb_invalidate_phys_addr(AddressSpace *as, hwaddr addr, MemTxAttrs attrs)
> +{
> + ram_addr_t ram_addr;
> + MemoryRegion *mr;
> + hwaddr l = 1;
> +
> + if (!tcg_enabled()) {
> + return;
> + }
> +
> + RCU_READ_LOCK_GUARD();
> + mr = address_space_translate(as, addr, &addr, &l, false, attrs);
> + if (!(memory_region_is_ram(mr)
> + || memory_region_is_romd(mr))) {
> + return;
> + }
> + ram_addr = memory_region_get_ram_addr(mr) + addr;
> + tb_invalidate_phys_page_range(ram_addr, ram_addr + 1);
> +}
> +
> +static void breakpoint_invalidate(CPUState *cpu, target_ulong pc)
> +{
> + /*
> + * There may not be a virtual to physical translation for the pc
> + * right now, but there may exist cached TB for this pc.
> + * Flush the whole TB cache to force re-translation of such TBs.
> + * This is heavyweight, but we're debugging anyway.
> + */
> + tb_flush(cpu);
> +}
> +#endif
> +
> +/* Add a breakpoint. */
> +int cpu_breakpoint_insert(CPUState *cpu, vaddr pc, int flags,
> + CPUBreakpoint **breakpoint)
> +{
> + CPUBreakpoint *bp;
> +
> + bp = g_malloc(sizeof(*bp));
> +
> + bp->pc = pc;
> + bp->flags = flags;
> +
> + /* keep all GDB-injected breakpoints in front */
> + if (flags & BP_GDB) {
> + QTAILQ_INSERT_HEAD(&cpu->breakpoints, bp, entry);
> + } else {
> + QTAILQ_INSERT_TAIL(&cpu->breakpoints, bp, entry);
> + }
> +
> + breakpoint_invalidate(cpu, pc);
> +
> + if (breakpoint) {
> + *breakpoint = bp;
> + }
> + return 0;
> +}
> +
> +/* Remove a specific breakpoint. */
> +int cpu_breakpoint_remove(CPUState *cpu, vaddr pc, int flags)
> +{
> + CPUBreakpoint *bp;
> +
> + QTAILQ_FOREACH(bp, &cpu->breakpoints, entry) {
> + if (bp->pc == pc && bp->flags == flags) {
> + cpu_breakpoint_remove_by_ref(cpu, bp);
> + return 0;
> + }
> + }
> + return -ENOENT;
> +}
> +
> +/* Remove a specific breakpoint by reference. */
> +void cpu_breakpoint_remove_by_ref(CPUState *cpu, CPUBreakpoint *breakpoint)
> +{
> + QTAILQ_REMOVE(&cpu->breakpoints, breakpoint, entry);
> +
> + breakpoint_invalidate(cpu, breakpoint->pc);
> +
> + g_free(breakpoint);
> +}
> +
> +/* Remove all matching breakpoints. */
> +void cpu_breakpoint_remove_all(CPUState *cpu, int mask)
> +{
> + CPUBreakpoint *bp, *next;
> +
> + QTAILQ_FOREACH_SAFE(bp, &cpu->breakpoints, entry, next) {
> + if (bp->flags & mask) {
> + cpu_breakpoint_remove_by_ref(cpu, bp);
> + }
> + }
> +}
> +
> +/* enable or disable single step mode. EXCP_DEBUG is returned by the
> + CPU loop after each instruction */
> +void cpu_single_step(CPUState *cpu, int enabled)
> +{
> + if (cpu->singlestep_enabled != enabled) {
> + cpu->singlestep_enabled = enabled;
> + if (kvm_enabled()) {
> + kvm_update_guest_debug(cpu, 0);
> + } else {
> + /* must flush all the translated code to avoid inconsistencies */
> + /* XXX: only flush what is necessary */
> + tb_flush(cpu);
> + }
> + }
> +}
> +
> +void cpu_abort(CPUState *cpu, const char *fmt, ...)
> +{
> + va_list ap;
> + va_list ap2;
> +
> + va_start(ap, fmt);
> + va_copy(ap2, ap);
> + fprintf(stderr, "qemu: fatal: ");
> + vfprintf(stderr, fmt, ap);
> + fprintf(stderr, "\n");
> + cpu_dump_state(cpu, stderr, CPU_DUMP_FPU | CPU_DUMP_CCOP);
> + if (qemu_log_separate()) {
> + FILE *logfile = qemu_log_lock();
> + qemu_log("qemu: fatal: ");
> + qemu_log_vprintf(fmt, ap2);
> + qemu_log("\n");
> + log_cpu_state(cpu, CPU_DUMP_FPU | CPU_DUMP_CCOP);
> + qemu_log_flush();
> + qemu_log_unlock(logfile);
> + qemu_log_close();
> + }
> + va_end(ap2);
> + va_end(ap);
> + replay_finish();
> +#if defined(CONFIG_USER_ONLY)
> + {
> + struct sigaction act;
> + sigfillset(&act.sa_mask);
> + act.sa_handler = SIG_DFL;
> + act.sa_flags = 0;
> + sigaction(SIGABRT, &act, NULL);
> + }
> +#endif
> + abort();
> +}
> +
> +/* physical memory access (slow version, mainly for debug) */
> +#if defined(CONFIG_USER_ONLY)
> +int cpu_memory_rw_debug(CPUState *cpu, target_ulong addr,
> + void *ptr, target_ulong len, bool is_write)
> +{
> + int flags;
> + target_ulong l, page;
> + void * p;
> + uint8_t *buf = ptr;
> +
> + while (len > 0) {
> + page = addr & TARGET_PAGE_MASK;
> + l = (page + TARGET_PAGE_SIZE) - addr;
> + if (l > len)
> + l = len;
> + flags = page_get_flags(page);
> + if (!(flags & PAGE_VALID))
> + return -1;
> + if (is_write) {
> + if (!(flags & PAGE_WRITE))
> + return -1;
> + /* XXX: this code should not depend on lock_user */
> + if (!(p = lock_user(VERIFY_WRITE, addr, l, 0)))
> + return -1;
> + memcpy(p, buf, l);
> + unlock_user(p, addr, l);
> + } else {
> + if (!(flags & PAGE_READ))
> + return -1;
> + /* XXX: this code should not depend on lock_user */
> + if (!(p = lock_user(VERIFY_READ, addr, l, 1)))
> + return -1;
> + memcpy(buf, p, l);
> + unlock_user(p, addr, 0);
> + }
> + len -= l;
> + buf += l;
> + addr += l;
> + }
> + return 0;
> +}
> +#endif
> +
> +bool target_words_bigendian(void)
> +{
> +#if defined(TARGET_WORDS_BIGENDIAN)
> + return true;
> +#else
> + return false;
> +#endif
> +}
> +
> +void page_size_init(void)
> +{
> + /* NOTE: we can always suppose that qemu_host_page_size >=
> + TARGET_PAGE_SIZE */
> + if (qemu_host_page_size == 0) {
> + qemu_host_page_size = qemu_real_host_page_size;
> + }
> + if (qemu_host_page_size < TARGET_PAGE_SIZE) {
> + qemu_host_page_size = TARGET_PAGE_SIZE;
> + }
> + qemu_host_page_mask = -(intptr_t)qemu_host_page_size;
> +}
> diff --git a/include/exec/cpu-common.h b/include/exec/cpu-common.h
> index d5e285d2b5..19805ed6db 100644
> --- a/include/exec/cpu-common.h
> +++ b/include/exec/cpu-common.h
> @@ -14,6 +14,9 @@ void cpu_list_unlock(void);
>
> void tcg_flush_softmmu_tlb(CPUState *cs);
>
> +void tcg_iommu_init_notifier_list(CPUState *cpu);
> +void tcg_iommu_free_notifier_list(CPUState *cpu);
> +
> #if !defined(CONFIG_USER_ONLY)
>
> enum device_endian {
> diff --git a/meson.build b/meson.build
> index 0e0577e81e..60df291b00 100644
> --- a/meson.build
> +++ b/meson.build
> @@ -1367,7 +1367,7 @@ common_ss.add(files('cpus-common.c'))
> subdir('softmmu')
>
> common_ss.add(capstone)
> -specific_ss.add(files('disas.c', 'exec.c', 'gdbstub.c'), capstone, libpmem, libdaxctl)
> +specific_ss.add(files('cpu.c', 'disas.c', 'gdbstub.c'), capstone)
> specific_ss.add(files('exec-vary.c'))
> specific_ss.add(when: 'CONFIG_TCG', if_true: files(
> 'fpu/softfloat.c',
> diff --git a/softmmu/meson.build b/softmmu/meson.build
> index 862ab24878..8f7210b4f0 100644
> --- a/softmmu/meson.build
> +++ b/softmmu/meson.build
> @@ -3,6 +3,7 @@ specific_ss.add(when: 'CONFIG_SOFTMMU', if_true: [files(
> 'balloon.c',
> 'cpus.c',
> 'cpu-throttle.c',
> + 'physmem.c',
> 'ioport.c',
> 'memory.c',
> 'memory_mapping.c',
> @@ -19,7 +20,7 @@ softmmu_ss.add(files(
> 'bootdevice.c',
> 'dma-helpers.c',
> 'qdev-monitor.c',
> -), sdl)
> +), sdl, libpmem, libdaxctl)
>
> softmmu_ss.add(when: 'CONFIG_TPM', if_true: files('tpm.c'))
> softmmu_ss.add(when: 'CONFIG_SECCOMP', if_true: [files('qemu-seccomp.c'), seccomp])
> diff --git a/exec.c b/softmmu/physmem.c
> similarity index 91%
> rename from exec.c
> rename to softmmu/physmem.c
> index ec68f4a9ca..6acd220f53 100644
> --- a/exec.c
> +++ b/softmmu/physmem.c
> @@ -1,5 +1,5 @@
> /*
> - * Virtual page mapping
> + * RAM allocation and memory access
> *
> * Copyright (c) 2003 Fabrice Bellard
> *
> @@ -28,10 +28,8 @@
> #include "tcg/tcg.h"
> #include "hw/qdev-core.h"
> #include "hw/qdev-properties.h"
> -#if !defined(CONFIG_USER_ONLY)
> #include "hw/boards.h"
> #include "hw/xen/xen.h"
> -#endif
> #include "sysemu/kvm.h"
> #include "sysemu/sysemu.h"
> #include "sysemu/tcg.h"
> @@ -40,9 +38,6 @@
> #include "qemu/config-file.h"
> #include "qemu/error-report.h"
> #include "qemu/qemu-print.h"
> -#if defined(CONFIG_USER_ONLY)
> -#include "qemu.h"
> -#else /* !CONFIG_USER_ONLY */
> #include "exec/memory.h"
> #include "exec/ioport.h"
> #include "sysemu/dma.h"
> @@ -56,7 +51,6 @@
> #include <linux/falloc.h>
> #endif
>
> -#endif
> #include "qemu/rcu_queue.h"
> #include "qemu/main-loop.h"
> #include "translate-all.h"
> @@ -83,7 +77,6 @@
>
> //#define DEBUG_SUBPAGE
>
> -#if !defined(CONFIG_USER_ONLY)
> /* ram_list is read under rcu_read_lock()/rcu_read_unlock(). Writes
> * are protected by the ramlist lock.
> */
> @@ -96,12 +89,6 @@ AddressSpace address_space_io;
> AddressSpace address_space_memory;
>
> static MemoryRegion io_mem_unassigned;
> -#endif
> -
> -uintptr_t qemu_host_page_size;
> -intptr_t qemu_host_page_mask;
> -
> -#if !defined(CONFIG_USER_ONLY)
>
> typedef struct PhysPageEntry PhysPageEntry;
>
> @@ -179,10 +166,6 @@ struct DirtyBitmapSnapshot {
> unsigned long dirty[];
> };
>
> -#endif
> -
> -#if !defined(CONFIG_USER_ONLY)
> -
> static void phys_map_node_reserve(PhysPageMap *map, unsigned nodes)
> {
> static unsigned alloc_hint = 16;
> @@ -666,7 +649,7 @@ static void tcg_register_iommu_notifier(CPUState *cpu,
> }
> }
>
> -static void tcg_iommu_free_notifier_list(CPUState *cpu)
> +void tcg_iommu_free_notifier_list(CPUState *cpu)
> {
> /* Destroy the CPU's notifier list */
> int i;
> @@ -680,6 +663,11 @@ static void tcg_iommu_free_notifier_list(CPUState *cpu)
> g_array_free(cpu->iommu_notifiers, true);
> }
>
> +void tcg_iommu_init_notifier_list(CPUState *cpu)
> +{
> + cpu->iommu_notifiers = g_array_new(false, true, sizeof(TCGIOMMUNotifier *));
> +}
> +
> /* Called from RCU critical section */
> MemoryRegionSection *
> address_space_translate_for_iotlb(CPUState *cpu, int asidx, hwaddr addr,
> @@ -737,91 +725,6 @@ address_space_translate_for_iotlb(CPUState *cpu, int asidx, hwaddr addr,
> translate_fail:
> return &d->map.sections[PHYS_SECTION_UNASSIGNED];
> }
> -#endif
> -
> -#if !defined(CONFIG_USER_ONLY)
> -
> -static int cpu_common_post_load(void *opaque, int version_id)
> -{
> - CPUState *cpu = opaque;
> -
> - /* 0x01 was CPU_INTERRUPT_EXIT. This line can be removed when the
> - version_id is increased. */
> - cpu->interrupt_request &= ~0x01;
> - tlb_flush(cpu);
> -
> - /* loadvm has just updated the content of RAM, bypassing the
> - * usual mechanisms that ensure we flush TBs for writes to
> - * memory we've translated code from. So we must flush all TBs,
> - * which will now be stale.
> - */
> - tb_flush(cpu);
> -
> - return 0;
> -}
> -
> -static int cpu_common_pre_load(void *opaque)
> -{
> - CPUState *cpu = opaque;
> -
> - cpu->exception_index = -1;
> -
> - return 0;
> -}
> -
> -static bool cpu_common_exception_index_needed(void *opaque)
> -{
> - CPUState *cpu = opaque;
> -
> - return tcg_enabled() && cpu->exception_index != -1;
> -}
> -
> -static const VMStateDescription vmstate_cpu_common_exception_index = {
> - .name = "cpu_common/exception_index",
> - .version_id = 1,
> - .minimum_version_id = 1,
> - .needed = cpu_common_exception_index_needed,
> - .fields = (VMStateField[]) {
> - VMSTATE_INT32(exception_index, CPUState),
> - VMSTATE_END_OF_LIST()
> - }
> -};
> -
> -static bool cpu_common_crash_occurred_needed(void *opaque)
> -{
> - CPUState *cpu = opaque;
> -
> - return cpu->crash_occurred;
> -}
> -
> -static const VMStateDescription vmstate_cpu_common_crash_occurred = {
> - .name = "cpu_common/crash_occurred",
> - .version_id = 1,
> - .minimum_version_id = 1,
> - .needed = cpu_common_crash_occurred_needed,
> - .fields = (VMStateField[]) {
> - VMSTATE_BOOL(crash_occurred, CPUState),
> - VMSTATE_END_OF_LIST()
> - }
> -};
> -
> -const VMStateDescription vmstate_cpu_common = {
> - .name = "cpu_common",
> - .version_id = 1,
> - .minimum_version_id = 1,
> - .pre_load = cpu_common_pre_load,
> - .post_load = cpu_common_post_load,
> - .fields = (VMStateField[]) {
> - VMSTATE_UINT32(halted, CPUState),
> - VMSTATE_UINT32(interrupt_request, CPUState),
> - VMSTATE_END_OF_LIST()
> - },
> - .subsections = (const VMStateDescription*[]) {
> - &vmstate_cpu_common_exception_index,
> - &vmstate_cpu_common_crash_occurred,
> - NULL
> - }
> -};
>
> void cpu_address_space_init(CPUState *cpu, int asidx,
> const char *prefix, MemoryRegion *mr)
> @@ -865,155 +768,7 @@ AddressSpace *cpu_get_address_space(CPUState *cpu, int asidx)
> /* Return the AddressSpace corresponding to the specified index */
> return cpu->cpu_ases[asidx].as;
> }
> -#endif
> -
> -void cpu_exec_unrealizefn(CPUState *cpu)
> -{
> - CPUClass *cc = CPU_GET_CLASS(cpu);
> -
> - tlb_destroy(cpu);
> - cpu_list_remove(cpu);
> -
> - if (cc->vmsd != NULL) {
> - vmstate_unregister(NULL, cc->vmsd, cpu);
> - }
> - if (qdev_get_vmsd(DEVICE(cpu)) == NULL) {
> - vmstate_unregister(NULL, &vmstate_cpu_common, cpu);
> - }
> -#ifndef CONFIG_USER_ONLY
> - tcg_iommu_free_notifier_list(cpu);
> -#endif
> -}
> -
> -Property cpu_common_props[] = {
> -#ifndef CONFIG_USER_ONLY
> - /* Create a memory property for softmmu CPU object,
> - * so users can wire up its memory. (This can't go in hw/core/cpu.c
> - * because that file is compiled only once for both user-mode
> - * and system builds.) The default if no link is set up is to use
> - * the system address space.
> - */
> - DEFINE_PROP_LINK("memory", CPUState, memory, TYPE_MEMORY_REGION,
> - MemoryRegion *),
> -#endif
> - DEFINE_PROP_BOOL("start-powered-off", CPUState, start_powered_off, false),
> - DEFINE_PROP_END_OF_LIST(),
> -};
> -
> -void cpu_exec_initfn(CPUState *cpu)
> -{
> - cpu->as = NULL;
> - cpu->num_ases = 0;
> -
> -#ifndef CONFIG_USER_ONLY
> - cpu->thread_id = qemu_get_thread_id();
> - cpu->memory = system_memory;
> - object_ref(OBJECT(cpu->memory));
> -#endif
> -}
> -
> -void cpu_exec_realizefn(CPUState *cpu, Error **errp)
> -{
> - CPUClass *cc = CPU_GET_CLASS(cpu);
> - static bool tcg_target_initialized;
> -
> - cpu_list_add(cpu);
> -
> - if (tcg_enabled() && !tcg_target_initialized) {
> - tcg_target_initialized = true;
> - cc->tcg_initialize();
> - }
> - tlb_init(cpu);
> -
> - qemu_plugin_vcpu_init_hook(cpu);
> -
> -#ifdef CONFIG_USER_ONLY
> - assert(cc->vmsd == NULL);
> -#else /* !CONFIG_USER_ONLY */
> - if (qdev_get_vmsd(DEVICE(cpu)) == NULL) {
> - vmstate_register(NULL, cpu->cpu_index, &vmstate_cpu_common, cpu);
> - }
> - if (cc->vmsd != NULL) {
> - vmstate_register(NULL, cpu->cpu_index, cc->vmsd, cpu);
> - }
> -
> - cpu->iommu_notifiers = g_array_new(false, true, sizeof(TCGIOMMUNotifier *));
> -#endif
> -}
> -
> -const char *parse_cpu_option(const char *cpu_option)
> -{
> - ObjectClass *oc;
> - CPUClass *cc;
> - gchar **model_pieces;
> - const char *cpu_type;
> -
> - model_pieces = g_strsplit(cpu_option, ",", 2);
> - if (!model_pieces[0]) {
> - error_report("-cpu option cannot be empty");
> - exit(1);
> - }
> -
> - oc = cpu_class_by_name(CPU_RESOLVING_TYPE, model_pieces[0]);
> - if (oc == NULL) {
> - error_report("unable to find CPU model '%s'", model_pieces[0]);
> - g_strfreev(model_pieces);
> - exit(EXIT_FAILURE);
> - }
> -
> - cpu_type = object_class_get_name(oc);
> - cc = CPU_CLASS(oc);
> - cc->parse_features(cpu_type, model_pieces[1], &error_fatal);
> - g_strfreev(model_pieces);
> - return cpu_type;
> -}
> -
> -#if defined(CONFIG_USER_ONLY)
> -void tb_invalidate_phys_addr(target_ulong addr)
> -{
> - mmap_lock();
> - tb_invalidate_phys_page_range(addr, addr + 1);
> - mmap_unlock();
> -}
> -
> -static void breakpoint_invalidate(CPUState *cpu, target_ulong pc)
> -{
> - tb_invalidate_phys_addr(pc);
> -}
> -#else
> -void tb_invalidate_phys_addr(AddressSpace *as, hwaddr addr, MemTxAttrs attrs)
> -{
> - ram_addr_t ram_addr;
> - MemoryRegion *mr;
> - hwaddr l = 1;
> -
> - if (!tcg_enabled()) {
> - return;
> - }
> -
> - RCU_READ_LOCK_GUARD();
> - mr = address_space_translate(as, addr, &addr, &l, false, attrs);
> - if (!(memory_region_is_ram(mr)
> - || memory_region_is_romd(mr))) {
> - return;
> - }
> - ram_addr = memory_region_get_ram_addr(mr) + addr;
> - tb_invalidate_phys_page_range(ram_addr, ram_addr + 1);
> -}
>
> -static void breakpoint_invalidate(CPUState *cpu, target_ulong pc)
> -{
> - /*
> - * There may not be a virtual to physical translation for the pc
> - * right now, but there may exist cached TB for this pc.
> - * Flush the whole TB cache to force re-translation of such TBs.
> - * This is heavyweight, but we're debugging anyway.
> - */
> - tb_flush(cpu);
> -}
> -#endif
> -
> -#ifndef CONFIG_USER_ONLY
> /* Add a watchpoint. */
> int cpu_watchpoint_insert(CPUState *cpu, vaddr addr, vaddr len,
> int flags, CPUWatchpoint **watchpoint)
> @@ -1122,123 +877,7 @@ int cpu_watchpoint_address_matches(CPUState *cpu, vaddr addr, vaddr len)
> }
> return ret;
> }
> -#endif /* !CONFIG_USER_ONLY */
> -
> -/* Add a breakpoint. */
> -int cpu_breakpoint_insert(CPUState *cpu, vaddr pc, int flags,
> - CPUBreakpoint **breakpoint)
> -{
> - CPUBreakpoint *bp;
> -
> - bp = g_malloc(sizeof(*bp));
> -
> - bp->pc = pc;
> - bp->flags = flags;
> -
> - /* keep all GDB-injected breakpoints in front */
> - if (flags & BP_GDB) {
> - QTAILQ_INSERT_HEAD(&cpu->breakpoints, bp, entry);
> - } else {
> - QTAILQ_INSERT_TAIL(&cpu->breakpoints, bp, entry);
> - }
> -
> - breakpoint_invalidate(cpu, pc);
> -
> - if (breakpoint) {
> - *breakpoint = bp;
> - }
> - return 0;
> -}
> -
> -/* Remove a specific breakpoint. */
> -int cpu_breakpoint_remove(CPUState *cpu, vaddr pc, int flags)
> -{
> - CPUBreakpoint *bp;
> -
> - QTAILQ_FOREACH(bp, &cpu->breakpoints, entry) {
> - if (bp->pc == pc && bp->flags == flags) {
> - cpu_breakpoint_remove_by_ref(cpu, bp);
> - return 0;
> - }
> - }
> - return -ENOENT;
> -}
>
> -/* Remove a specific breakpoint by reference. */
> -void cpu_breakpoint_remove_by_ref(CPUState *cpu, CPUBreakpoint *breakpoint)
> -{
> - QTAILQ_REMOVE(&cpu->breakpoints, breakpoint, entry);
> -
> - breakpoint_invalidate(cpu, breakpoint->pc);
> -
> - g_free(breakpoint);
> -}
> -
> -/* Remove all matching breakpoints. */
> -void cpu_breakpoint_remove_all(CPUState *cpu, int mask)
> -{
> - CPUBreakpoint *bp, *next;
> -
> - QTAILQ_FOREACH_SAFE(bp, &cpu->breakpoints, entry, next) {
> - if (bp->flags & mask) {
> - cpu_breakpoint_remove_by_ref(cpu, bp);
> - }
> - }
> -}
> -
> -/* enable or disable single step mode. EXCP_DEBUG is returned by the
> - CPU loop after each instruction */
> -void cpu_single_step(CPUState *cpu, int enabled)
> -{
> - if (cpu->singlestep_enabled != enabled) {
> - cpu->singlestep_enabled = enabled;
> - if (kvm_enabled()) {
> - kvm_update_guest_debug(cpu, 0);
> - } else {
> - /* must flush all the translated code to avoid inconsistencies */
> - /* XXX: only flush what is necessary */
> - tb_flush(cpu);
> - }
> - }
> -}
> -
> -void cpu_abort(CPUState *cpu, const char *fmt, ...)
> -{
> - va_list ap;
> - va_list ap2;
> -
> - va_start(ap, fmt);
> - va_copy(ap2, ap);
> - fprintf(stderr, "qemu: fatal: ");
> - vfprintf(stderr, fmt, ap);
> - fprintf(stderr, "\n");
> - cpu_dump_state(cpu, stderr, CPU_DUMP_FPU | CPU_DUMP_CCOP);
> - if (qemu_log_separate()) {
> - FILE *logfile = qemu_log_lock();
> - qemu_log("qemu: fatal: ");
> - qemu_log_vprintf(fmt, ap2);
> - qemu_log("\n");
> - log_cpu_state(cpu, CPU_DUMP_FPU | CPU_DUMP_CCOP);
> - qemu_log_flush();
> - qemu_log_unlock(logfile);
> - qemu_log_close();
> - }
> - va_end(ap2);
> - va_end(ap);
> - replay_finish();
> -#if defined(CONFIG_USER_ONLY)
> - {
> - struct sigaction act;
> - sigfillset(&act.sa_mask);
> - act.sa_handler = SIG_DFL;
> - act.sa_flags = 0;
> - sigaction(SIGABRT, &act, NULL);
> - }
> -#endif
> - abort();
> -}
> -
> -#if !defined(CONFIG_USER_ONLY)
> /* Called from RCU critical section */
> static RAMBlock *qemu_get_ram_block(ram_addr_t addr)
> {
> @@ -1425,9 +1064,6 @@ hwaddr memory_region_section_get_iotlb(CPUState *cpu,
> AddressSpaceDispatch *d = flatview_to_dispatch(section->fv);
> return section - d->map.sections;
> }
> -#endif /* defined(CONFIG_USER_ONLY) */
> -
> -#if !defined(CONFIG_USER_ONLY)
>
> static int subpage_register(subpage_t *mmio, uint32_t start, uint32_t end,
> uint16_t section);
> @@ -3028,52 +2664,6 @@ MemoryRegion *get_system_io(void)
> return system_io;
> }
>
> -#endif /* !defined(CONFIG_USER_ONLY) */
> -
> -/* physical memory access (slow version, mainly for debug) */
> -#if defined(CONFIG_USER_ONLY)
> -int cpu_memory_rw_debug(CPUState *cpu, target_ulong addr,
> - void *ptr, target_ulong len, bool is_write)
> -{
> - int flags;
> - target_ulong l, page;
> - void * p;
> - uint8_t *buf = ptr;
> -
> - while (len > 0) {
> - page = addr & TARGET_PAGE_MASK;
> - l = (page + TARGET_PAGE_SIZE) - addr;
> - if (l > len)
> - l = len;
> - flags = page_get_flags(page);
> - if (!(flags & PAGE_VALID))
> - return -1;
> - if (is_write) {
> - if (!(flags & PAGE_WRITE))
> - return -1;
> - /* XXX: this code should not depend on lock_user */
> - if (!(p = lock_user(VERIFY_WRITE, addr, l, 0)))
> - return -1;
> - memcpy(p, buf, l);
> - unlock_user(p, addr, l);
> - } else {
> - if (!(flags & PAGE_READ))
> - return -1;
> - /* XXX: this code should not depend on lock_user */
> - if (!(p = lock_user(VERIFY_READ, addr, l, 1)))
> - return -1;
> - memcpy(buf, p, l);
> - unlock_user(p, addr, 0);
> - }
> - len -= l;
> - buf += l;
> - addr += l;
> - }
> - return 0;
> -}
> -
> -#else
> -
> static void invalidate_and_set_dirty(MemoryRegion *mr, hwaddr addr,
> hwaddr length)
> {
> @@ -3862,18 +3452,7 @@ int qemu_target_page_bits_min(void)
> {
> return TARGET_PAGE_BITS_MIN;
> }
> -#endif
> -
> -bool target_words_bigendian(void)
> -{
> -#if defined(TARGET_WORDS_BIGENDIAN)
> - return true;
> -#else
> - return false;
> -#endif
> -}
>
> -#ifndef CONFIG_USER_ONLY
> bool cpu_physical_memory_is_io(hwaddr phys_addr)
> {
> MemoryRegion*mr;
> @@ -4003,23 +3582,6 @@ bool ramblock_is_pmem(RAMBlock *rb)
> return rb->flags & RAM_PMEM;
> }
>
> -#endif
> -
> -void page_size_init(void)
> -{
> - /* NOTE: we can always suppose that qemu_host_page_size >=
> - TARGET_PAGE_SIZE */
> - if (qemu_host_page_size == 0) {
> - qemu_host_page_size = qemu_real_host_page_size;
> - }
> - if (qemu_host_page_size < TARGET_PAGE_SIZE) {
> - qemu_host_page_size = TARGET_PAGE_SIZE;
> - }
> - qemu_host_page_mask = -(intptr_t)qemu_host_page_size;
> -}
> -
> -#if !defined(CONFIG_USER_ONLY)
> -
> static void mtree_print_phys_entries(int start, int end, int skip, int ptr)
> {
> if (start == end - 1) {
> @@ -4152,5 +3714,3 @@ bool ram_block_discard_is_required(void)
> {
> return qatomic_read(&ram_block_discard_disabled) < 0;
> }
> -
> -#endif
>
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH 2/2] exec: split out non-softmmu-specific parts
2020-10-08 7:47 ` Claudio Fontana
@ 2020-10-08 7:56 ` Paolo Bonzini
2020-10-08 11:02 ` Claudio Fontana
0 siblings, 1 reply; 14+ messages in thread
From: Paolo Bonzini @ 2020-10-08 7:56 UTC (permalink / raw)
To: Claudio Fontana, qemu-devel
On 08/10/20 09:47, Claudio Fontana wrote:
> On 10/6/20 11:19 AM, Paolo Bonzini wrote:
>> Over the years, most parts of exec.c that were not specific to softmmu
>> have been moved to accel/tcg; what's left is mostly the low-level part
>> of the memory API, which includes RAMBlock and AddressSpaceDispatch.
>> However exec.c also hosts 4-500 lines of code for the target specific
>> parts of the CPU QOM object, plus a few functions for user-mode
>> emulation that do not have a better place (they are not TCG-specific so
>> accel/tcg/user-exec.c is not a good place either).
>>
>> Move these parts to a new file, so that exec.c can be moved to
>> softmmu/physmem.c.
>>
>> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
>
> Hi Paolo,
>
> the comment does not talk about cpu.c, which is now created in the top source directory.
> What is the role of this new module?
It's actually in the commit message: "4-500 lines of code for the target
specific parts of the CPU QOM object, plus a few functions for user-mode
emulation that do not have a better place".
It's basically sitting between hw/core/cpu.c and target/*/cpu.c. Hence
the non-descriptive name. :)
> Also, could we find a more descriptive file name than cpu.c?
> Do you plan further renaming of this new module functions?
>
> Or its this basically a "leftovers" file for which we did not find a proper role yet?
The user-mode parts are, but most of it is implementing the QOM CPU
object. We can move those functions to hw/core/cpu.c and make that file
target-dependent, I wouldn't object to that. But since there are some
opportunities for simplification, I'd rather do that in a separate patch
and keep the pure code-movement in this one.
Paolo
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH 2/2] exec: split out non-softmmu-specific parts
2020-10-08 7:56 ` Paolo Bonzini
@ 2020-10-08 11:02 ` Claudio Fontana
2020-10-08 11:36 ` Paolo Bonzini
0 siblings, 1 reply; 14+ messages in thread
From: Claudio Fontana @ 2020-10-08 11:02 UTC (permalink / raw)
To: Paolo Bonzini; +Cc: qemu-devel
On 10/8/20 9:56 AM, Paolo Bonzini wrote:
> On 08/10/20 09:47, Claudio Fontana wrote:
>> On 10/6/20 11:19 AM, Paolo Bonzini wrote:
>>> Over the years, most parts of exec.c that were not specific to softmmu
>>> have been moved to accel/tcg; what's left is mostly the low-level part
>>> of the memory API, which includes RAMBlock and AddressSpaceDispatch.
>>> However exec.c also hosts 4-500 lines of code for the target specific
>>> parts of the CPU QOM object, plus a few functions for user-mode
>>> emulation that do not have a better place (they are not TCG-specific so
>>> accel/tcg/user-exec.c is not a good place either).
>>>
>>> Move these parts to a new file, so that exec.c can be moved to
>>> softmmu/physmem.c.
>>>
>>> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
>>
>> Hi Paolo,
>>
>> the comment does not talk about cpu.c, which is now created in the top source directory.
>> What is the role of this new module?
>
> It's actually in the commit message: "4-500 lines of code for the target
> specific parts of the CPU QOM object, plus a few functions for user-mode
> emulation that do not have a better place".
>
> It's basically sitting between hw/core/cpu.c and target/*/cpu.c. Hence
> the non-descriptive name. :)
>
>> Also, could we find a more descriptive file name than cpu.c?
>> Do you plan further renaming of this new module functions?
>>
>> Or its this basically a "leftovers" file for which we did not find a proper role yet?
>
> The user-mode parts are, but most of it is implementing the QOM CPU
> object. We can move those functions to hw/core/cpu.c and make that file
> target-dependent, I wouldn't object to that. But since there are some
> opportunities for simplification, I'd rather do that in a separate patch
> and keep the pure code-movement in this one.
>
> Paolo
>
Ciao Paolo,
this gives me an idea, we already basically have a target-specific part of a cpu QEMU object.
I basically was looking for a place to graft accelerator-specific code in order to refactor target/i386/cpu...,
to split between tcg stuff and non-tcg stuff, and thus refactor even more code.
In the past I thought to put them here for example:
diff --git a/target/i386/cpu-qom.h b/target/i386/cpu-qom.h
index 3e96f8d668..3716c3e949 100644
--- a/target/i386/cpu-qom.h
+++ b/target/i386/cpu-qom.h
@@ -72,6 +72,12 @@ typedef struct X86CPUClass {
DeviceRealize parent_realize;
DeviceUnrealize parent_unrealize;
DeviceReset parent_reset;
+
+ /* methods operating on CPUX86State */
+ uint32_t (*cpu_compute_eflags)(CPUX86State *env);
+ void (*cpu_set_mxcsr)(CPUX86State *env, uint32_t mxcsr);
+ void (*cpu_set_fpuc)(CPUX86State *env, uint16_t fpuc);
+ void (*cpu_report_tpr_access)(CPUX86State *env, TPRAccess access);
} X86CPUClass;
typedef struct X86CPU X86CPU;
But maybe that is the right component?
Just thinking out loud here.
Ciao,
Claudio
^ permalink raw reply related [flat|nested] 14+ messages in thread
* Re: [PATCH 2/2] exec: split out non-softmmu-specific parts
2020-10-08 11:02 ` Claudio Fontana
@ 2020-10-08 11:36 ` Paolo Bonzini
2020-10-08 12:42 ` Claudio Fontana
0 siblings, 1 reply; 14+ messages in thread
From: Paolo Bonzini @ 2020-10-08 11:36 UTC (permalink / raw)
To: Claudio Fontana; +Cc: qemu-devel
On 08/10/20 13:02, Claudio Fontana wrote:
>>> What is the role of this new module?
>>
>> It's [...] "4-500 lines of code for the target
>> specific parts of the CPU QOM object, plus a few functions for user-mode
>> emulation that do not have a better place".
>>
>> It's basically sitting between hw/core/cpu.c and target/*/cpu.c. Hence
>> the non-descriptive name. :)
>>
>>> Or its this basically a "leftovers" file for which we did not find a proper role yet?
>>
>> The user-mode parts are, but most of it is implementing the QOM CPU
>> object. We can move those functions to hw/core/cpu.c and make that file
>> target-dependent, I wouldn't object to that.
>
> this gives me an idea, we already basically have a target-specific part of a cpu QEMU object.
Which is? :) Sorry I don't follow. We have one that depends on the
target architecture (methods in the CPU class), but not one that depends
on the target kind. We could add more methods in the CPU class for
that, but I'm not sure it would be useful because (unlike CPUs of which
in theory there could be >1 class in the system) the whole emulation
_has_ to be either user-level or system.
> I basically was looking for a place to graft accelerator-specific code in order to refactor target/i386/cpu...,
> to split between tcg stuff and non-tcg stuff, and thus refactor even more code.
>
> In the past I thought to put them here for example:
>
> diff --git a/target/i386/cpu-qom.h b/target/i386/cpu-qom.h
> index 3e96f8d668..3716c3e949 100644
> --- a/target/i386/cpu-qom.h
> +++ b/target/i386/cpu-qom.h
> @@ -72,6 +72,12 @@ typedef struct X86CPUClass {
> DeviceRealize parent_realize;
> DeviceUnrealize parent_unrealize;
> DeviceReset parent_reset;
> +
> + /* methods operating on CPUX86State */
> + uint32_t (*cpu_compute_eflags)(CPUX86State *env);
> + void (*cpu_set_mxcsr)(CPUX86State *env, uint32_t mxcsr);
> + void (*cpu_set_fpuc)(CPUX86State *env, uint16_t fpuc);
> + void (*cpu_report_tpr_access)(CPUX86State *env, TPRAccess access);
> } X86CPUClass;
>
> typedef struct X86CPU X86CPU;
I think in this case you would have an X86AccelOps struct and a global
variable pointing to it.
Paolo
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH 2/2] exec: split out non-softmmu-specific parts
2020-10-08 11:36 ` Paolo Bonzini
@ 2020-10-08 12:42 ` Claudio Fontana
0 siblings, 0 replies; 14+ messages in thread
From: Claudio Fontana @ 2020-10-08 12:42 UTC (permalink / raw)
To: Paolo Bonzini; +Cc: qemu-devel
On 10/8/20 1:36 PM, Paolo Bonzini wrote:
> On 08/10/20 13:02, Claudio Fontana wrote:
>>>> What is the role of this new module?
>>>
>>> It's [...] "4-500 lines of code for the target
>>> specific parts of the CPU QOM object, plus a few functions for user-mode
>>> emulation that do not have a better place".
>>>
>>> It's basically sitting between hw/core/cpu.c and target/*/cpu.c. Hence
>>> the non-descriptive name. :)
>>>
>>>> Or its this basically a "leftovers" file for which we did not find a proper role yet?
>>>
>>> The user-mode parts are, but most of it is implementing the QOM CPU
>>> object. We can move those functions to hw/core/cpu.c and make that file
>>> target-dependent, I wouldn't object to that.
>>
>> this gives me an idea, we already basically have a target-specific part of a cpu QEMU object.
>
> Which is? :) Sorry I don't follow. We have one that depends on the
> target architecture (methods in the CPU class), but not one that depends
> on the target kind. We could add more methods in the CPU class for
> that, but I'm not sure it would be useful because (unlike CPUs of which
> in theory there could be >1 class in the system) the whole emulation
> _has_ to be either user-level or system.
>
>> I basically was looking for a place to graft accelerator-specific code in order to refactor target/i386/cpu...,
>> to split between tcg stuff and non-tcg stuff, and thus refactor even more code.
>>
>> In the past I thought to put them here for example:
>>
>> diff --git a/target/i386/cpu-qom.h b/target/i386/cpu-qom.h
>> index 3e96f8d668..3716c3e949 100644
>> --- a/target/i386/cpu-qom.h
>> +++ b/target/i386/cpu-qom.h
>> @@ -72,6 +72,12 @@ typedef struct X86CPUClass {
>> DeviceRealize parent_realize;
>> DeviceUnrealize parent_unrealize;
>> DeviceReset parent_reset;
>> +
>> + /* methods operating on CPUX86State */
>> + uint32_t (*cpu_compute_eflags)(CPUX86State *env);
>> + void (*cpu_set_mxcsr)(CPUX86State *env, uint32_t mxcsr);
>> + void (*cpu_set_fpuc)(CPUX86State *env, uint16_t fpuc);
>> + void (*cpu_report_tpr_access)(CPUX86State *env, TPRAccess access);
>> } X86CPUClass;
>>
>> typedef struct X86CPU X86CPU;
>
> I think in this case you would have an X86AccelOps struct and a global
> variable pointing to it.
>
> Paolo
>
>
I was hoping to make use of some of the object model ... but lets get to this when we get to this later on.
Ciao, thanks!
Claudio
^ permalink raw reply [flat|nested] 14+ messages in thread
end of thread, other threads:[~2020-10-08 12:43 UTC | newest]
Thread overview: 14+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-10-06 9:19 [PATCH 0/2] Move more code to softmmu/ Paolo Bonzini
2020-10-06 9:19 ` [PATCH 1/2] softmmu: move more files " Paolo Bonzini
2020-10-06 9:34 ` Philippe Mathieu-Daudé
2020-10-06 9:38 ` Philippe Mathieu-Daudé
2020-10-06 9:43 ` BALATON Zoltan via
2020-10-06 9:54 ` Paolo Bonzini
2020-10-06 10:02 ` Philippe Mathieu-Daudé
2020-10-06 9:19 ` [PATCH 2/2] exec: split out non-softmmu-specific parts Paolo Bonzini
2020-10-06 9:44 ` Philippe Mathieu-Daudé
2020-10-08 7:47 ` Claudio Fontana
2020-10-08 7:56 ` Paolo Bonzini
2020-10-08 11:02 ` Claudio Fontana
2020-10-08 11:36 ` Paolo Bonzini
2020-10-08 12:42 ` Claudio Fontana
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.