This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
[+rfc] Re: [patch v6 00/21] record-btrace: reverse
- From: Jan Kratochvil <jan dot kratochvil at redhat dot com>
- To: Markus Metzger <markus dot t dot metzger at intel dot com>
- Cc: gdb-patches at sourceware dot org
- Date: Sun, 6 Oct 2013 21:59:13 +0200
- Subject: [+rfc] Re: [patch v6 00/21] record-btrace: reverse
- Authentication-results: sourceware.org; auth=none
- References: <1379676639-31802-1-git-send-email-markus dot t dot metzger at intel dot com>
On Fri, 20 Sep 2013 13:30:18 +0200, Markus Metzger wrote:
> This is a small update of v5 that fixes the test fail when using "finish"
> in a tail call frame. The fix only changes "record-btrace: extend unwinder"
> and resolves a resulting conflict in "record-btrace: add (reverse-)stepping
> support".
Besides some nitpicks there remains the larger request
Re: [patch v4 23/24] record-btrace: add (reverse-)stepping support
https://sourceware.org/ml/gdb-patches/2013-09/msg01004.html
Message-ID: <20130930102533.GA29665@host2.jankratochvil.net>
to always use to_resume + to_wait for any change of history position. This
also means the "hacks" like
+record_btrace_goto_end (void)
+ print_stack_frame (get_selected_frame (NULL), 1, SRC_AND_LOC, 1);
can be removed as normal GDB inferior stop will print the frame.
Other comments are welcome if it is not a too strict requirement from me but
I think we could face some forgotten resetting of future caches avoiding the
normal resume+wait path this way. Or ... we would not face them but it would
violate localized of cache resets in to_resume+to_wait code path.
Thanks,
Jan