This is the mail archive of the
gdb@sourceware.org
mailing list for the GDB project.
Re: Fwd: Time when the Heap snapshot is taken when dumping Core
- From: Abhishek Karoliya <abhikaro at gmail dot com>
- To: Pedro Alves <palves at redhat dot com>
- Cc: gdb at sourceware dot org
- Date: Thu, 31 Jul 2014 20:16:02 +0530
- Subject: Re: Fwd: Time when the Heap snapshot is taken when dumping Core
- Authentication-results: sourceware.org; auth=none
- References: <CAEEVgBVDzEf3LH1_SxYyvMKy4K5v__Qvc1HmcOmkdyrTNt96LA at mail dot gmail dot com> <CAEEVgBVZa5s-gOJhj3Ps4nCBp3jKL4a38_Ai-HWz6RT5B6Vtdg at mail dot gmail dot com> <53D906A8 dot 4080308 at redhat dot com>
Thanks!
I am now sure that the core won't be generated up till the point
abort() is called, which happens after we have printed out our
diagnostics and heap parameters. To narrow down the race window you
are talking about, we will have to sacrifice most of these.
On 30 July 2014 20:22, Pedro Alves <palves@redhat.com> wrote:
> On 07/25/2014 04:23 PM, Abhishek Karoliya wrote:
>
>> My question is:
>> A) Is my theory correct?
>
> It's plausible at least. Though I'd suspect a simpler
> explanation is more likely.
>
>> B) Am I right in presuming that the heap memory snapshot is not taken
>> at the time crash (signal being raised) but at in the final moments of
>> the process?
>> C) Address/location that caused a SIGSEGV can still be used (by other threads)?
>
> Both correct, though if you don't have a SIGSEGV handler, the race
> window is very narrow.
>
> --
> Thanks,
> Pedro Alves
>