Message ID | 87y420l1ak.fsf@fimbulvetr.bsc.es |
---|---|
State | New |
Headers | show |
On 7 October 2016 at 14:46, Lluís Vilanova <vilanova@ac.upc.edu> wrote: > diff --git a/qom/cpu.c b/qom/cpu.c > index 484c493..40f2eb1 100644 > --- a/qom/cpu.c > +++ b/qom/cpu.c > @@ -356,12 +356,15 @@ static void cpu_common_initfn(Object *obj) > qemu_mutex_init(&cpu->work_mutex); > QTAILQ_INIT(&cpu->breakpoints); > QTAILQ_INIT(&cpu->watchpoints); > - bitmap_zero(cpu->trace_dstate, TRACE_VCPU_EVENT_COUNT); > + > + cpu->trace_dstate = bitmap_new(min(trace_get_vcpu_event_count(), 1); > } > > static void cpu_common_finalize(Object *obj) > { > - cpu_exec_exit(CPU(obj)); > + CPUState *cpu = CPU(obj); > + cpu_exec_exit(cpu); > + g_free(cpu->trace_dstate); > } > > static int64_t cpu_common_get_arch_id(CPUState *cpu) > > Beware that min() needs to be defined somewhere on that file first (should it be > available on some generic header?). osdep.h defines a MIN() macro. thanks -- PMM
On Fri, Oct 07, 2016 at 03:46:27PM +0200, Lluís Vilanova wrote: > Peter Maydell writes: > > > On 7 October 2016 at 11:09, Stefan Hajnoczi <stefanha@redhat.com> wrote: > >> The following changes since commit e902754e3d0890945ddcc1b33748ed73762ddb8d: > >> > >> Merge remote-tracking branch 'remotes/dgibson/tags/ppc-for-2.8-20161006' into staging (2016-10-06 13:34:00 +0100) > >> > >> are available in the git repository at: > >> > >> git://github.com/stefanha/qemu.git tags/tracing-pull-request > >> > >> for you to fetch changes up to bcdc1363dcb9adb5f7ea016a4d0b41f0357c8182: > >> > >> trace: Add missing execution mode of guest events (2016-10-07 09:17:49 +0100) > >> > >> ---------------------------------------------------------------- > >> > >> ---------------------------------------------------------------- > > > This makes all the linux-user binaries dump core on startup: > > > $ gdb --args /home/petmay01/linaro/qemu-for-merges/build/all-linux-static/x86_64-linux-user/qemu-x86_64 > > /bin/ls > > (gdb) r > > Starting program: > > /home/petmay01/linaro/qemu-for-merges/build/all-linux-static/x86_64-linux-user/qemu-x86_64 > > /bin/ls > > [Thread debugging using libthread_db enabled] > > Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". > > [New Thread 0x7ffff7ffa700 (LWP 14965)] > > > Thread 1 "qemu-x86_64" received signal SIGABRT, Aborted. > > 0x000000006019fbe8 in raise () > > (gdb) bt > > #0 0x000000006019fbe8 in raise () > > #1 0x00000000601a030a in abort () > > #2 0x00000000600bdd02 in bitmap_new (nbits=0) > > at /home/petmay01/linaro/qemu-for-merges/include/qemu/bitmap.h:99 > > #3 0x00000000600be99f in cpu_common_initfn (obj=0x625f6340) > > at /home/petmay01/linaro/qemu-for-merges/qom/cpu.c:360 > > #4 0x00000000600bfe97 in object_init_with_type (obj=0x625f6340, ti=0x625eb9a0) > > at /home/petmay01/linaro/qemu-for-merges/qom/object.c:339 > > #5 0x00000000600bfe79 in object_init_with_type (obj=0x625f6340, ti=0x625e2d80) > > at /home/petmay01/linaro/qemu-for-merges/qom/object.c:335 > > #6 0x00000000600bfe79 in object_init_with_type (obj=0x625f6340, ti=0x625e78d0) > > at /home/petmay01/linaro/qemu-for-merges/qom/object.c:335 > > #7 0x00000000600c00aa in object_initialize_with_type > > (data=0x625f6340, size=38416, type=0x625e78d0) > > at /home/petmay01/linaro/qemu-for-merges/qom/object.c:370 > > #8 0x00000000600c0508 in object_new_with_type (type=0x625e78d0) > > at /home/petmay01/linaro/qemu-for-merges/qom/object.c:478 > > #9 0x00000000600c0542 in object_new (typename=0x625e2fe0 "qemu64-x86_64-cpu") > > at /home/petmay01/linaro/qemu-for-merges/qom/object.c:488 > > #10 0x00000000600be03f in cpu_generic_init (typename=0x6025ad54 > > "x86_64-cpu", cpu_model=0x60247ca5 "qemu64") > > at /home/petmay01/linaro/qemu-for-merges/qom/cpu.c:76 > > #11 0x0000000060089438 in cpu_x86_init (cpu_model=0x60247ca5 "qemu64") > > at /home/petmay01/linaro/qemu-for-merges/target-i386/cpu.c:2266 > > #12 0x0000000060031031 in main (argc=2, argv=0x7fffffffe4b8, > > envp=0x7fffffffe4d0) > > at /home/petmay01/linaro/qemu-for-merges/linux-user/main.c:4253 > > Aha! I didn't take that abort into consideration during the review. The fix > should be on patch 13 of the series (trace: dynamically allocate trace_dstate in > CPUState): > > diff --git a/qom/cpu.c b/qom/cpu.c > index 484c493..40f2eb1 100644 > --- a/qom/cpu.c > +++ b/qom/cpu.c > @@ -356,12 +356,15 @@ static void cpu_common_initfn(Object *obj) > qemu_mutex_init(&cpu->work_mutex); > QTAILQ_INIT(&cpu->breakpoints); > QTAILQ_INIT(&cpu->watchpoints); > - bitmap_zero(cpu->trace_dstate, TRACE_VCPU_EVENT_COUNT); > + > + cpu->trace_dstate = bitmap_new(min(trace_get_vcpu_event_count(), 1); > } > > static void cpu_common_finalize(Object *obj) > { > - cpu_exec_exit(CPU(obj)); > + CPUState *cpu = CPU(obj); > + cpu_exec_exit(cpu); > + g_free(cpu->trace_dstate); > } > > static int64_t cpu_common_get_arch_id(CPUState *cpu) No, this isn't what we want todo. The real problem is that we should have been registering the trace events in the linux-user/bsd-user programs. I sent a fix for that. Regards, Daniel
Daniel P Berrange writes: > On Fri, Oct 07, 2016 at 03:46:27PM +0200, Lluís Vilanova wrote: >> Peter Maydell writes: >> >> > On 7 October 2016 at 11:09, Stefan Hajnoczi <stefanha@redhat.com> wrote: >> >> The following changes since commit e902754e3d0890945ddcc1b33748ed73762ddb8d: >> >> >> >> Merge remote-tracking branch 'remotes/dgibson/tags/ppc-for-2.8-20161006' into staging (2016-10-06 13:34:00 +0100) >> >> >> >> are available in the git repository at: >> >> >> >> git://github.com/stefanha/qemu.git tags/tracing-pull-request >> >> >> >> for you to fetch changes up to bcdc1363dcb9adb5f7ea016a4d0b41f0357c8182: >> >> >> >> trace: Add missing execution mode of guest events (2016-10-07 09:17:49 +0100) >> >> >> >> ---------------------------------------------------------------- >> >> >> >> ---------------------------------------------------------------- >> >> > This makes all the linux-user binaries dump core on startup: >> >> > $ gdb --args /home/petmay01/linaro/qemu-for-merges/build/all-linux-static/x86_64-linux-user/qemu-x86_64 >> > /bin/ls >> > (gdb) r >> > Starting program: >> > /home/petmay01/linaro/qemu-for-merges/build/all-linux-static/x86_64-linux-user/qemu-x86_64 >> > /bin/ls >> > [Thread debugging using libthread_db enabled] >> > Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". >> > [New Thread 0x7ffff7ffa700 (LWP 14965)] >> >> > Thread 1 "qemu-x86_64" received signal SIGABRT, Aborted. >> > 0x000000006019fbe8 in raise () >> > (gdb) bt >> > #0 0x000000006019fbe8 in raise () >> > #1 0x00000000601a030a in abort () >> > #2 0x00000000600bdd02 in bitmap_new (nbits=0) >> > at /home/petmay01/linaro/qemu-for-merges/include/qemu/bitmap.h:99 >> > #3 0x00000000600be99f in cpu_common_initfn (obj=0x625f6340) >> > at /home/petmay01/linaro/qemu-for-merges/qom/cpu.c:360 >> > #4 0x00000000600bfe97 in object_init_with_type (obj=0x625f6340, ti=0x625eb9a0) >> > at /home/petmay01/linaro/qemu-for-merges/qom/object.c:339 >> > #5 0x00000000600bfe79 in object_init_with_type (obj=0x625f6340, ti=0x625e2d80) >> > at /home/petmay01/linaro/qemu-for-merges/qom/object.c:335 >> > #6 0x00000000600bfe79 in object_init_with_type (obj=0x625f6340, ti=0x625e78d0) >> > at /home/petmay01/linaro/qemu-for-merges/qom/object.c:335 >> > #7 0x00000000600c00aa in object_initialize_with_type >> > (data=0x625f6340, size=38416, type=0x625e78d0) >> > at /home/petmay01/linaro/qemu-for-merges/qom/object.c:370 >> > #8 0x00000000600c0508 in object_new_with_type (type=0x625e78d0) >> > at /home/petmay01/linaro/qemu-for-merges/qom/object.c:478 >> > #9 0x00000000600c0542 in object_new (typename=0x625e2fe0 "qemu64-x86_64-cpu") >> > at /home/petmay01/linaro/qemu-for-merges/qom/object.c:488 >> > #10 0x00000000600be03f in cpu_generic_init (typename=0x6025ad54 >> > "x86_64-cpu", cpu_model=0x60247ca5 "qemu64") >> > at /home/petmay01/linaro/qemu-for-merges/qom/cpu.c:76 >> > #11 0x0000000060089438 in cpu_x86_init (cpu_model=0x60247ca5 "qemu64") >> > at /home/petmay01/linaro/qemu-for-merges/target-i386/cpu.c:2266 >> > #12 0x0000000060031031 in main (argc=2, argv=0x7fffffffe4b8, >> > envp=0x7fffffffe4d0) >> > at /home/petmay01/linaro/qemu-for-merges/linux-user/main.c:4253 >> >> Aha! I didn't take that abort into consideration during the review. The fix >> should be on patch 13 of the series (trace: dynamically allocate trace_dstate in >> CPUState): >> >> diff --git a/qom/cpu.c b/qom/cpu.c >> index 484c493..40f2eb1 100644 >> --- a/qom/cpu.c >> +++ b/qom/cpu.c >> @@ -356,12 +356,15 @@ static void cpu_common_initfn(Object *obj) >> qemu_mutex_init(&cpu->work_mutex); >> QTAILQ_INIT(&cpu->breakpoints); >> QTAILQ_INIT(&cpu->watchpoints); >> - bitmap_zero(cpu->trace_dstate, TRACE_VCPU_EVENT_COUNT); >> + >> + cpu->trace_dstate = bitmap_new(min(trace_get_vcpu_event_count(), 1); >> } >> >> static void cpu_common_finalize(Object *obj) >> { >> - cpu_exec_exit(CPU(obj)); >> + CPUState *cpu = CPU(obj); >> + cpu_exec_exit(cpu); >> + g_free(cpu->trace_dstate); >> } >> >> static int64_t cpu_common_get_arch_id(CPUState *cpu) > No, this isn't what we want todo. The real problem is that we should > have been registering the trace events in the linux-user/bsd-user > programs. I sent a fix for that. Well, that'll work as long as we keep at least one vcpu event :) Cheers, Lluis
diff --git a/qom/cpu.c b/qom/cpu.c index 484c493..40f2eb1 100644 --- a/qom/cpu.c +++ b/qom/cpu.c @@ -356,12 +356,15 @@ static void cpu_common_initfn(Object *obj) qemu_mutex_init(&cpu->work_mutex); QTAILQ_INIT(&cpu->breakpoints); QTAILQ_INIT(&cpu->watchpoints); - bitmap_zero(cpu->trace_dstate, TRACE_VCPU_EVENT_COUNT); + + cpu->trace_dstate = bitmap_new(min(trace_get_vcpu_event_count(), 1); } static void cpu_common_finalize(Object *obj) { - cpu_exec_exit(CPU(obj)); + CPUState *cpu = CPU(obj); + cpu_exec_exit(cpu); + g_free(cpu->trace_dstate); } static int64_t cpu_common_get_arch_id(CPUState *cpu)