qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH v2 10/23] linux-user/i386: Implement setup_sigtramp


From: Peter Maydell
Subject: Re: [PATCH v2 10/23] linux-user/i386: Implement setup_sigtramp
Date: Tue, 29 Jun 2021 15:40:40 +0100

On Fri, 18 Jun 2021 at 20:38, Richard Henderson
<richard.henderson@linaro.org> wrote:
>
> Create and record the two signal trampolines.
> Use them when the guest does not use SA_RESTORER.
> Note that x86_64 does not use this code.
>
> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
> ---
>  linux-user/i386/target_signal.h   |  2 ++
>  linux-user/x86_64/target_signal.h |  3 +++
>  linux-user/i386/signal.c          | 42 ++++++++++++++++++-------------
>  3 files changed, 29 insertions(+), 18 deletions(-)
>
> diff --git a/linux-user/i386/target_signal.h b/linux-user/i386/target_signal.h
> index 50361af874..64d09f2e75 100644
> --- a/linux-user/i386/target_signal.h
> +++ b/linux-user/i386/target_signal.h
> @@ -22,4 +22,6 @@ typedef struct target_sigaltstack {
>  #include "../generic/signal.h"
>
>  #define TARGET_ARCH_HAS_SETUP_FRAME
> +#define TARGET_ARCH_HAS_SIGTRAMP_PAGE 1
> +
>  #endif /* I386_TARGET_SIGNAL_H */
> diff --git a/linux-user/x86_64/target_signal.h 
> b/linux-user/x86_64/target_signal.h
> index 4ea74f20dd..4673c5a886 100644
> --- a/linux-user/x86_64/target_signal.h
> +++ b/linux-user/x86_64/target_signal.h
> @@ -21,4 +21,7 @@ typedef struct target_sigaltstack {
>
>  #include "../generic/signal.h"
>
> +/* For x86_64, use of SA_RESTORER is mandatory. */
> +#define TARGET_ARCH_HAS_SIGTRAMP_PAGE 0
> +
>  #endif /* X86_64_TARGET_SIGNAL_H */
> diff --git a/linux-user/i386/signal.c b/linux-user/i386/signal.c
> index 8701774e37..a83ecba54f 100644
> --- a/linux-user/i386/signal.c
> +++ b/linux-user/i386/signal.c
> @@ -337,16 +337,7 @@ void setup_frame(int sig, struct target_sigaction *ka,
>      if (ka->sa_flags & TARGET_SA_RESTORER) {
>          __put_user(ka->sa_restorer, &frame->pretcode);
>      } else {
> -        uint16_t val16;
> -        abi_ulong retcode_addr;
> -        retcode_addr = frame_addr + offsetof(struct sigframe, retcode);
> -        __put_user(retcode_addr, &frame->pretcode);
> -        /* This is popl %eax ; movl $,%eax ; int $0x80 */
> -        val16 = 0xb858;
> -        __put_user(val16, (uint16_t *)(frame->retcode+0));
> -        __put_user(TARGET_NR_sigreturn, (int *)(frame->retcode+2));
> -        val16 = 0x80cd;
> -        __put_user(val16, (uint16_t *)(frame->retcode+6));
> +        __put_user(default_sigreturn, &frame->pretcode);
>

In the kernel in arch/x86/kernel/signal.c there is a comment:

        /*
         * This is popl %eax ; movl $__NR_sigreturn, %eax ; int $0x80
         *
         * WE DO NOT USE IT ANY MORE! It's only left here for historical
         * reasons and because gdb uses it as a signature to notice
         * signal handler stack frames.
         */

which suggests that we also should continue to fill in the
retcode bytes in the signal frame for gdb's benefit even though
we don't actually execute them any more.

thanks
-- PMM



reply via email to

[Prev in Thread] Current Thread [Next in Thread]