[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH v2 1/3] icount: preserve cflags when custom tb is about to ex
Re: [PATCH v2 1/3] icount: preserve cflags when custom tb is about to execute
Tue, 16 Nov 2021 10:40:50 +0300
Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.13.0
On 11.11.2021 15:20, Alex Bennée wrote:
Pavel Dovgalyuk <email@example.com> writes:
When debugging with the watchpoints, qemu may need to create
TB with single instruction. This is achieved by setting cpu->cflags_next_tb.
But when this block is about to execute, it may be interrupted by another
thread. In this case cflags will be lost and next executed TB will not
be the special one.
This patch checks TB exit reason and restores cflags_next_tb to allow
finding the interrupted block.
Signed-off-by: Pavel Dovgalyuk <Pavel.Dovgalyuk@ispras.ru>
accel/tcg/cpu-exec.c | 10 ++++++++++
1 file changed, 10 insertions(+)
diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
index 2d14d02f6c..df12452b8f 100644
@@ -846,6 +846,16 @@ static inline void cpu_loop_exec_tb(CPUState *cpu,
* cpu_handle_interrupt. cpu_handle_interrupt will also
* clear cpu->icount_decr.u16.high.
+ if (cpu->cflags_next_tb == -1
+ && (!use_icount || !(tb->cflags & CF_USE_ICOUNT)
Why check use_icount here? The cflags should always have CF_USE_ICOUNT
set when icount is enabled. Lets not over complicate the inverted ||
tests we have here.
Not really. Sometimes we use non-icount blocks in icount mode.
But AFAIR they are used only for triggering the exeptions, but not for
+ || cpu_neg(cpu)->icount_decr.u16.low >= tb->icount))
Is u16.low ever set when icount isn't enabled?
This condition is checked for icount mode only.
u16.low is not used without icount.
+ * icount is disabled or there are enough instructions
+ * in the budget, do not retranslate this block with
+ * different parameters.
+ cpu->cflags_next_tb = tb->cflags;
[PATCH v2 2/3] softmmu: fix watchpoint-interrupt races, Pavel Dovgalyuk, 2021/11/11
[PATCH v2 3/3] softmmu: fix watchpoints on memory used by vCPU internals, Pavel Dovgalyuk, 2021/11/11