This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: gdbserver regression [Re: [COMMIT PATCH][GDBserver] Multi-process + multi-arch]
- From: Pedro Alves <palves at redhat dot com>
- To: Jan Kratochvil <jan dot kratochvil at redhat dot com>
- Cc: gdb-patches at sourceware dot org
- Date: Tue, 11 Jun 2013 16:44:53 +0100
- Subject: Re: gdbserver regression [Re: [COMMIT PATCH][GDBserver] Multi-process + multi-arch]
- References: <20130530121335 dot 15047 dot 12654 dot stgit at brno dot lan> <51B1BB04 dot 9020304 at redhat dot com> <20130608015339 dot GA10329 at host2 dot jankratochvil dot net> <51B72C8E dot 8060905 at redhat dot com> <20130611143849 dot GA30462 at host2 dot jankratochvil dot net>
On 06/11/2013 03:38 PM, Jan Kratochvil wrote:
> On Tue, 11 Jun 2013 15:56:30 +0200, Pedro Alves wrote:
>> Unfortunately I can't reproduce this. :-/
>
> I have checked now that I get PASS->FAIL when I link gdbserver with -lmcheck.
> BTW the regression is specific to gdbserver (not gdb).
>
> I get PASS->FAIL for any tests running gdbserver, such as:
> runtest gdb.server/server-run.exp
>
> For some reason regsets_info->disabled_regsets gets set which causes:
>
> -Sending packet: $?#3f...Packet received: T0506:0000000000000000;07:20dbffffff7f0000;10:20d4ddf7ff7f0000;thread:p48bf.48bf;core:6;^M
> +Sending packet: $?#3f...Packet received: T0506:0000000000000000;07:0000000000000000;10:0000000000000000;thread:p48fd.48fd;core:6;^M
>
> I guess you can fix it easily now, I did not try valgrind on gdbserver but it
> should catch it I think.
Thanks! I'll try that out.
--
Pedro Alves