[Bug libc/22380] Inconsistency detected by ld.so: dl-open.c: 717: _dl_open: Assertion `_dl_debug_initialize (0, args.nsid)->r_state == RT_CONSISTENT' failed!

[ragaprasanth at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=22380

--- Comment #4 from Prasanth <ragaprasanth at gmail dot com> ---
(In reply to Carlos O'Donell from comment #3)
> (In reply to Prasanth from comment #2)
> > Created attachment 10565 [details]
> > attachment-17785-0.html
> > 
> > No, it's not multi threading program, you could see in the man page it's
> > suit simple c program
> 
> Prasanth,
> 
> Thank you.
> 
> The loading of the shared library for analysis fails, and we are in RT_ADD
> state instead of a consistent state:
> 
> 706	  assert (_dl_debug_initialize (0, args.nsid)->r_state == RT_CONSISTENT);
> (gdb) p _r_debug
> $1 = {r_version = 1, r_map = 0x7ffff7ffe130, r_brk = 140737351944064,
> r_state = RT_ADD, 
>   r_ldbase = 140737351872512}
> 
> Then in _dl_find_dso_for_object:
> 
> 241   if (__glibc_unlikely (mode & __RTLD_SPROF))
> 242     /* This happens only if we load a DSO for 'sprof'.  */
> 243     return;
> 
> We stop doing any work before marking the object as consistent. Which seems
> odd, perhaps I've done something wrong in my analysis.
> 
> This looks like sprof has been broken for a long time if this is true, and
> we've not had any regression test to cover this.
> 
> Florian,
> 
> Given, that you're poking around the dynamic loader, would you be able to
> provide any input on what might be wrong?
> 
> With upstream master I see:
> "sprof: error while loading shared libraries: sprof: cannot open shared
> object file"
> 
> That doesn't make much sense though since the shared library is in the path.

Hi carlos,

 I too tried with the upstream patch, That end-up with the "Can't find the
object file error"

-- 
You are receiving this mail because:
You are on the CC list for the bug.