Re: [patch+7.5.1] Work around PR libc/13097 "linux-vdso.so.1" #3

[Pedro Alves <palves@redhat.com>]

On 11/27/2012 09:59 PM, Mike Frysinger wrote:
> On Friday 23 November 2012 11:01:03 Pedro Alves wrote:
>> On 11/23/2012 12:39 PM, Jan Kratochvil wrote:
>>> Besides that there is no standard for any such rule and I consider this
>>> patch only as a workaround of a WONTFIXed glibc PR.
>>
>> Given the new order of glibc maintenance, if we believe glibc's WONTFIX was
>> wrong, then we should reopen the PR, and retry discussing the issue.
> 
> i think the glibc behavior is actually nice and we don't want to revert it.  
> if i'm single stepping through code that enters the vdso, if gdb doesn't know 
> about it, it looks like i just stepped off into the weeds and the program is 
> doing something wrong.  if, instead, it told me i was in the vdso (and we had 
> some way of communicating symbol information), that's a lot more useful.

Indeed.  GDB already loads the vDSO 's elf (symfile-mem.c).  GDB finds the vdso address
in the auxv.  I think in Fedora, the idea is to find the debug info file through
build id matching (I haven't checked if that's actually implemented, but in previous
online searches I found references to that).  In
<http://sourceware.org/ml/gdb-patches/2012-11/msg00628.html> I mentioned that this change
could make it possible to find the separate debug file by name.  I think it'd be
nice to turn things around, and list the vDSO in the shared libraries list instead of
hiding it.

BTW, before that change, when the vDSO was nameless, what made sure it didn't appear
in the DSO list?  Where was it skipped?

-- 
Pedro Alves