Re: Question about trap
- From: Peter Flass <Peter_Flass@xxxxxxxxx>
- Date: Fri, 26 Feb 2010 19:07:21 -0500
David Schwartz wrote:
On Feb 26, 5:08 am, Pete <Peter_Fl...@xxxxxxxxx> wrote:
Thanks. I guess that's the answer than. I usually look for Linux-
specific documentation, forgetting about the Sys V ABI spec.
It's also in the ELF x86 ABI spec. The SysV ABI spec seems to be the
first x86 ABI to specify this, at least as far as I could tell.
I was
noticing that there wasn't a signal specific to this. I guess I
thought it would generate a flavor of SIGFPE, although that doesn't
really apply either.
No signal seems to perfectly apply. You can make arguments for
SIGSEGV, SIGBUS, SIGFPE, and SIGILL, but mostly the arguments consist
of pointing out that nothing is better. They picked SIGSEGV.
Not really a problem, but I would have liked to see a documented value of 'si_code' to indicate what's going on.
BTW, my last post seems to have gotten truncated. I was saying I was looking at the siginfo struct, and getting si_signo=11 (SIGSEGV), si_error=0, and si_code=128 (0x00000080).
Thanks for the help.
.
- Follow-Ups:
- Re: Question about trap
- From: Alan Curry
- Re: Question about trap
- References:
- Question about trap
- From: Pete
- Re: Question about trap
- From: David Schwartz
- Re: Question about trap
- From: Pete
- Re: Question about trap
- From: David Schwartz
- Question about trap
- Prev by Date: Re: Question about trap
- Next by Date: Re: Question about trap
- Previous by thread: Re: Question about trap
- Next by thread: Re: Question about trap
- Index(es):