This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH v3 3/3] Dynamic core regset sections support
- From: Mark Kettenis <mark dot kettenis at xs4all dot nl>
- To: arnez at linux dot vnet dot ibm dot com
- Cc: gdb-patches at sourceware dot org, Ulrich dot Weigand at de dot ibm dot com
- Date: Wed, 12 Jun 2013 17:21:57 +0200 (CEST)
- Subject: Re: [PATCH v3 3/3] Dynamic core regset sections support
- References: <877ghzmkmj dot fsf at br87z6lw dot de dot ibm dot com> <8738sngy5e dot fsf at br87z6lw dot de dot ibm dot com>
> From: Andreas Arnez <arnez@linux.vnet.ibm.com>
> Date: Wed, 12 Jun 2013 17:13:33 +0200
>
> Rework the handling of core file register note sections. So far the
> target-dependent code provided a static list of supported register
> note sections via set_gdbarch_core_regset_sections. This patch
> replaces the static list by a target-dependent iterator function which
> can be set with set_gdbarch_iterate_over_regset_sections. Only those
> architectures that have been using set_gdbarch_core_regset_sections
> are affected by this change.
>
> The patch is needed on newer z/Architecture systems for correct gcore
> handling. With the static approach, gcore always writes the TDB
> regset into the core file, even if the TDB registers are currently
> invalid. The new iterator omits the TDB in this case.
Sorry, but I really don't like the obfuscation that this diff brings
to the amd64 & i186 Linux targets. Is there really no other way to
this? Is it really that bad to write out the invalid TDB registers?
If GDB recognizes them as invalid, this shouldn't be a big issue
should it?