This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [patch v4 03/24] btrace: change branch trace data structure
- From: Jan Kratochvil <jan dot kratochvil at redhat dot com>
- To: "Metzger, Markus T" <markus dot t dot metzger at intel dot com>
- Cc: "gdb-patches at sourceware dot org" <gdb-patches at sourceware dot org>, "Himpel, Christian" <christian dot himpel at intel dot com>
- Date: Mon, 23 Sep 2013 09:14:54 +0200
- Subject: Re: [patch v4 03/24] btrace: change branch trace data structure
- Authentication-results: sourceware.org; auth=none
- References: <1372842874-28951-1-git-send-email-markus dot t dot metzger at intel dot com> <1372842874-28951-4-git-send-email-markus dot t dot metzger at intel dot com> <20130818190426 dot GC24153 at host2 dot jankratochvil dot net> <A78C989F6D9628469189715575E55B230A9CA7A2 at IRSMSX104 dot ger dot corp dot intel dot com> <20130912200927 dot GA29475 at host2 dot jankratochvil dot net> <A78C989F6D9628469189715575E55B230A9CE95A at IRSMSX104 dot ger dot corp dot intel dot com> <20130921194413 dot GA20539 at host2 dot jankratochvil dot net> <A78C989F6D9628469189715575E55B230A9D4934 at IRSMSX104 dot ger dot corp dot intel dot com>
On Mon, 23 Sep 2013 08:54:01 +0200, Metzger, Markus T wrote:
> The trace is collected in a cyclic buffer by the h/w. When the inferior
> stops, GDB reads that buffer which corresponds to the tail of the
> inferior's execution trace.
I somehow expected kernel SIGTRAPs the process when the buffer overflows so
that the buffer could be read in.
OK, GDB code now makes more sense to me, just it should be more described in
the manual, I have made such comment in [patch v6 21/21] after I reply it.
Thanks,
Jan