[LTP] [PATCH v2] syscalls/signal06: loop being clobbered by syscall on clang
Viktor Martensson
vmartensson@google.com
Sat Aug 20 00:34:25 CEST 2022
Hello!
The syscall will get the return value in %rax I believe, so it may be wise
to indicate that it gets modified no? I would assume something like
*int foo; *
asm volatile ("syscall" : *"=a" (foo)*: "a"(__NR_tkill), "D"(pid),
"S"(SIGHUP) : "rcx", "r11");
would instruct the compiler to expect that %rax has been modified, and it
can't rely on it staying the same. An optimization pass may otherwise take
away the repeated setting of __NR_tkill in %rax.
/V
On Fri, Aug 19, 2022 at 2:31 PM 'Edward Liaw' via kernel-team <
kernel-team@android.com> wrote:
> Indicate to the compiler that the syscall will modify registers rcx
> and r11 to prevent loop from getting clobbered.
>
> Signed-off-by: Edward Liaw <edliaw@google.com>
>
> ---
> When I combined the asm instruction into one line, it threw an error
> that the "Asm-specifier for input or output variable conflicts with asm
> clobber list" for rax. I omitted it for now, but I'm not sure if that
> is correct.
>
> Also, is it ok to change the subject line like I did?
> ---
> testcases/kernel/syscalls/signal/signal06.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/testcases/kernel/syscalls/signal/signal06.c
> b/testcases/kernel/syscalls/signal/signal06.c
> index 64f886ee3..fba380610 100644
> --- a/testcases/kernel/syscalls/signal/signal06.c
> +++ b/testcases/kernel/syscalls/signal/signal06.c
> @@ -72,8 +72,7 @@ void test(void)
> while (D == VALUE && loop < LOOPS) {
> /* sys_tkill(pid, SIGHUP); asm to avoid save/reload
> * fp regs around c call */
> - asm ("" : : "a"(__NR_tkill), "D"(pid), "S"(SIGHUP));
> - asm ("syscall" : : : "ax");
> + asm volatile ("syscall" : : "a"(__NR_tkill), "D"(pid),
> "S"(SIGHUP) : "rcx", "r11");
>
> loop++;
> }
> --
> 2.37.2.609.g9ff673ca1a-goog
>
> --
> To unsubscribe from this group and stop receiving emails from it, send an
> email to kernel-team+unsubscribe@android.com.
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20220819/66bf43e0/attachment.htm>
More information about the ltp
mailing list