This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH] printcmd.c (ui_printf): make internalvar string can be printf and eval when inferior cannot alloc memory
- From: Jan Kratochvil <jan dot kratochvil at redhat dot com>
- To: Hui Zhu <teawater at gmail dot com>
- Cc: Tom Tromey <tromey at redhat dot com>, gdb-patches ml <gdb-patches at sourceware dot org>
- Date: Tue, 6 Sep 2011 10:50:07 +0200
- Subject: Re: [PATCH] printcmd.c (ui_printf): make internalvar string can be printf and eval when inferior cannot alloc memory
- References: <CANFwon0RttecHT1BrVx886NC6yCsojHm-MWfr7ttS8=CjoNEzg@mail.gmail.com> <m339hz1i9v.fsf@fleche.redhat.com> <CANFwon0wqsOirAz078-y+ti-WFNHGJ45F_-HfS89ViYM02UDFA@mail.gmail.com> <m3fwl2ebq1.fsf@fleche.redhat.com> <CANFwon1h3fytbmBB5DiXjLt=pa7nt_JTuFhuGDxbVjCS=7FwtQ@mail.gmail.com> <m3k4acaysh.fsf@fleche.redhat.com> <CANFwon3kjh4HBH59BXyM=R451iAozjJjJ1kACg=Jbv607PV6aA@mail.gmail.com> <m3k4a9lc8i.fsf@fleche.redhat.com> <CANFwon0=Trk5i-j4MRPOAXx0AeJaqsZk6Tg052n9_EUEiQ=LrQ@mail.gmail.com>
On Tue, 06 Sep 2011 10:21:53 +0200, Hui Zhu wrote:
> + if (TYPE_CODE (check_typedef (value_type (val_args[i])))
> + == TYPE_CODE_ARRAY)
> + fprintf_filtered (stream, current_substring,
> + (char *) value_contents (val_args[i]));
(gdb) set $a={'a','b'}
(gdb) printf "%s\n",$a
==23392== Invalid read of size 1
==23392== at: vfprintf (vfprintf.c:1568)
==23392== by: vasprintf (vasprintf.c:64)
==23392== by: xstrvprintf (common-utils.c:131)
==23392== by: vfprintf_maybe_filtered (utils.c:2379)
==23392== by: vfprintf_filtered (utils.c:2389)
==23392== by: fprintf_filtered (utils.c:2441)
==23392== by: ui_printf (printcmd.c:2348)
[...]
==23392== Address 0xd24bb82 is 0 bytes after a block of size 2 alloc'd
==23392== at: calloc (vg_replace_malloc.c:467)
==23392== by: xcalloc (common-utils.c:92)
==23392== by: xzalloc (common-utils.c:102)
==23392== by: allocate_value_contents (value.c:690)
==23392== by: allocate_value (value.c:700)
==23392== by: value_copy (value.c:1299)
==23392== by: value_of_internalvar (value.c:1725)
==23392== by: evaluate_subexp_standard (eval.c:903)
==23392== by: evaluate_subexp_c (c-lang.c:720)
==23392== by: evaluate_subexp (eval.c:76)
==23392== by: evaluate_expression (eval.c:151)
==23392== by: parse_to_comma_and_eval (eval.c:136)
==23392== by: ui_printf (printcmd.c:2328)
[...]
The original code was not completely correct in such cases but GDB could not
crash, now it can, I find it as a regression.
I would welcome a testcase.
Thanks,
Jan