Bug#760450: libc6-prof is unusable

2014-10-16 Thread Aurelien Jarno
control: found -1 760450 libc6-prof/2.11.3-4

On Thu, Sep 04, 2014 at 12:41:06PM +0400, Dmitry Eremin-Solenikov wrote:
> Package: libc6-prof
> Version: 2.19-10
> Severity: grave
> 
> The library from libc6-prof package is unusable.
> 
> I'm compiling the simple example program:
> 
> int main(void)
> {
>   return sleep(3);
> }
> 
> If I follow directly the suggestion from the README.Debian, I get the
> following error:
> 
> $ gcc -g -pg test.c  -static-libgcc -lc_p
> /usr/bin/ld: dynamic STT_GNU_IFUNC symbol `strcmp' with pointer equality
> in
> `/usr/lib/gcc/x86_64-linux-gnu/4.9/../../../x86_64-linux-gnu/libc_p.a(strcmp.op)'
> can not be used when making an executable; recompile with -fPIE and
> relink with -pie
> collect2: error: ld returned 1 exit status

The problem seems to be related to the STT_GNU_IFUNC support in glibc.
It's something that has been added in Squeeze, and indeed the problem is
reproducible there.

There might be some way to get it working again, but given there are now
better way to do profiling, even without the need to recompile and using
the normal debugging symbols, I think it's time to just get rid of of
this package. It doesn't seems to be really used anyway [1].

Any thoughts?

[1] https://qa.debian.org/popcon.php?package=eglibc
-- 
Aurelien Jarno  GPG: 4096R/1DDD8C9B
aurel...@aurel32.net http://www.aurel32.net


-- 
To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org



Bug#760450: libc6-prof is unusable

2014-09-04 Thread Dmitry Eremin-Solenikov
Package: libc6-prof
Version: 2.19-10
Severity: grave

The library from libc6-prof package is unusable.

I'm compiling the simple example program:

int main(void)
{
return sleep(3);
}

If I follow directly the suggestion from the README.Debian, I get the
following error:

$ gcc -g -pg test.c  -static-libgcc -lc_p
/usr/bin/ld: dynamic STT_GNU_IFUNC symbol `strcmp' with pointer equality
in
`/usr/lib/gcc/x86_64-linux-gnu/4.9/../../../x86_64-linux-gnu/libc_p.a(strcmp.op)'
can not be used when making an executable; recompile with -fPIE and
relink with -pie
collect2: error: ld returned 1 exit status


I was able to make executable using one of the following compile lines:

$ gcc -g -pg test.c -nodefaultlibs  -lc_p -lgcc_eh -lc_p

or

$ gcc -g -pg test.c -nodefaultlibs -Wl,--start-group -lc_p -lgcc_eh
-Wl,--end-group

However the resulting executable segfaults:

$ ./a.out 
Segmentation fault
$ valgrind --quiet ./a.out
==1440== Invalid read of size 4
==1440==at 0x459DBB: __mcount_internal (in /tmp/tt/t/a.out)
==1440==by 0x4347D3: mcount (in /tmp/tt/t/a.out)
==1440==by 0x400F48: ??? (in /tmp/tt/t/a.out)
==1440==by 0xFFF000387: ???
==1440==  Address 0x18 is not stack'd, malloc'd or (recently) free'd
==1440== 
==1440== 
==1440== Process terminating with default action of signal 11 (SIGSEGV)
==1440==  Access not within mapped region at address 0x18
==1440==at 0x459DBB: __mcount_internal (in /tmp/tt/t/a.out)
==1440==by 0x4347D3: mcount (in /tmp/tt/t/a.out)
==1440==by 0x400F48: ??? (in /tmp/tt/t/a.out)
==1440==by 0xFFF000387: ???
==1440==  If you believe this happened as a result of a stack
==1440==  overflow in your program's main thread (unlikely but
==1440==  possible), you can try to increase the size of the
==1440==  main thread stack using the --main-stacksize= flag.
==1440==  The main thread stack size used in this run was 8388608.
Segmentation fault


-- System Information:
Debian Release: jessie/sid
  APT prefers testing
  APT policy: (500, 'testing')
Architecture: amd64 (x86_64)
Foreign Architectures: i386

Kernel: Linux 3.14-1-amd64 (SMP w/8 CPU cores)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages libc6-prof depends on:
ii  libc6  2.19-10

libc6-prof recommends no packages.

libc6-prof suggests no packages.

-- no debconf information


-- 
To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org