[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-devel] [PATCH v3 08/15] cpu: atomically modify cpu->exit_request
From: |
Alex Bennée |
Subject: |
[Qemu-devel] [PATCH v3 08/15] cpu: atomically modify cpu->exit_request |
Date: |
Fri, 30 Sep 2016 22:30:59 +0100 |
ThreadSanitizer picks up potential races although we already use
barriers to ensure things are in the correct order when processing exit
requests. For true C11 defined behaviour across threads we need to use
relaxed atomic_set/atomic_read semantics to reassure tsan.
Signed-off-by: Alex Bennée <address@hidden>
---
cpu-exec.c | 8 ++++----
qom/cpu.c | 4 ++--
2 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/cpu-exec.c b/cpu-exec.c
index 8823d23..e114fcd 100644
--- a/cpu-exec.c
+++ b/cpu-exec.c
@@ -192,7 +192,7 @@ static inline tcg_target_ulong cpu_tb_exec(CPUState *cpu,
TranslationBlock *itb)
/* We were asked to stop executing TBs (probably a pending
* interrupt. We've now stopped, so clear the flag.
*/
- cpu->tcg_exit_req = 0;
+ atomic_set(&cpu->tcg_exit_req, 0);
}
return ret;
}
@@ -490,8 +490,8 @@ static inline void cpu_handle_interrupt(CPUState *cpu,
*last_tb = NULL;
}
}
- if (unlikely(cpu->exit_request || replay_has_interrupt())) {
- cpu->exit_request = 0;
+ if (unlikely(atomic_read(&cpu->exit_request) || replay_has_interrupt())) {
+ atomic_set(&cpu->exit_request, 0);
cpu->exception_index = EXCP_INTERRUPT;
cpu_loop_exit(cpu);
}
@@ -503,7 +503,7 @@ static inline void cpu_loop_exec_tb(CPUState *cpu,
TranslationBlock *tb,
{
uintptr_t ret;
- if (unlikely(cpu->exit_request)) {
+ if (unlikely(atomic_read(&cpu->exit_request))) {
return;
}
diff --git a/qom/cpu.c b/qom/cpu.c
index ef905da..e765bc0 100644
--- a/qom/cpu.c
+++ b/qom/cpu.c
@@ -120,10 +120,10 @@ void cpu_reset_interrupt(CPUState *cpu, int mask)
void cpu_exit(CPUState *cpu)
{
- cpu->exit_request = 1;
+ atomic_set(&cpu->exit_request, 1);
/* Ensure cpu_exec will see the exit request after TCG has exited. */
smp_wmb();
- cpu->tcg_exit_req = 1;
+ atomic_set(&cpu->tcg_exit_req, 1);
}
int cpu_write_elf32_qemunote(WriteCoreDumpFunction f, CPUState *cpu,
--
2.9.3
- [Qemu-devel] [PATCH v3 01/15] atomic.h: fix __SANITIZE_THREAD__ build, (continued)
- [Qemu-devel] [PATCH v3 01/15] atomic.h: fix __SANITIZE_THREAD__ build, Alex Bennée, 2016/09/30
- [Qemu-devel] [PATCH v3 03/15] exec-all.h: revert tb_page_addr_t to target_ulong, Alex Bennée, 2016/09/30
- [Qemu-devel] [PATCH v3 06/15] qom/object: update class cache atomically, Alex Bennée, 2016/09/30
- [Qemu-devel] [PATCH v3 02/15] atomic.h: comment on use of atomic_read/set, Alex Bennée, 2016/09/30
- [Qemu-devel] [PATCH v3 04/15] tcg/optimize: move default return out of if statement, Alex Bennée, 2016/09/30
- [Qemu-devel] [PATCH v3 10/15] linux-user/syscall: extend lock around cpu-list, Alex Bennée, 2016/09/30
- [Qemu-devel] [PATCH v3 05/15] seqlock: use atomic writes for the sequence, Alex Bennée, 2016/09/30
- [Qemu-devel] [PATCH v3 08/15] cpu: atomically modify cpu->exit_request,
Alex Bennée <=
- [Qemu-devel] [PATCH v3 07/15] qom/cpu: atomically clear the tb_jmp_cache, Alex Bennée, 2016/09/30
- [Qemu-devel] [PATCH v3 11/15] qga/command: use QEMU atomic primitives, Alex Bennée, 2016/09/30
- [Qemu-devel] [PATCH v3 12/15] .travis.yml: add gcc sanitizer build, Alex Bennée, 2016/09/30
- [Qemu-devel] [PATCH v3 13/15] tcg: ensure cpu_tb_exec/tb_gen_code use atomic_read/write, Alex Bennée, 2016/09/30
- [Qemu-devel] [PATCH v3 15/15] translate-all: mark updates to PageDesc as atomic, Alex Bennée, 2016/09/30
- [Qemu-devel] [PATCH v3 09/15] util/qht: atomically set b->hashes, Alex Bennée, 2016/09/30
- [Qemu-devel] [PATCH v3 14/15] tcg: update remaining TranslationBuffer fields atomically, Alex Bennée, 2016/09/30
- Re: [Qemu-devel] [PATCH v3 00/15] A number of fixes for ThreadSanitizer, no-reply, 2016/09/30
- Re: [Qemu-devel] [PATCH v3 00/15] A number of fixes for ThreadSanitizer, no-reply, 2016/09/30