Hi,
I've compiled a new version of the package with --without-gcc-arch added to
./configure in debian/rules and placed it temporarily at
http://pkg-flash.alioth.debian.org/gnash/
I don't know if the error would be repetible with a package compiled in my
compiler, as I guess the version in Debian i386 might have been compiled by my
sponsor or by the buildd system. Anyway, lets try. I have used the same orig
tarball that is currently in Debian SID not to hide the results under a change
of the code.
Thanks for the report. Please, tell me if that solves the problem. In any
case, as that bug might be all along the packages, wouldn't it make sense to
tell debhelper maintainers to add --without-gcc-arch to the skeleton of
autotools-based packages?
Greetings,
Miry
PS: I'm CC my sponsor for the package, in case he's interested in this too.
--- Timo Juhani Lindfors <[EMAIL PROTECTED]> escribió:
> Package: gnash
> Version: 0.7.1+cvs20060924.1330-1
> Severity: grave
> Justification: renders package unusable
>
> Steps to reproduce:
> 1) gnash foo.swf
>
> Expected results:
> 1) gnash opens a window and shows output of the interpreter flash program.
>
> Actual results:
> 1) gnash crashes with the following message:
> Illegal instruction (core dumped)
>
> More info:
> 1) GDB shows the following backtrace:
>
> [New Thread 32769 (LWP 26043)]
> [New Thread 16386 (LWP 26044)]
> [Thread 16386 (LWP 26044) exited]
>
> Program received signal SIGILL, Illegal instruction.
> [Switching to Thread 16384 (LWP 26042)]
> 0x408afaf6 in gnash::Player::run () from /usr/lib/libgnashplayer.so.0
> (gdb) bt
> #0 0x408afaf6 in gnash::Player::run () from /usr/lib/libgnashplayer.so.0
> #1 0x0804a058 in ?? ()
> #2 0x411d53be in __libc_start_main () from /lib/libc.so.6
> #3 0x08049791 in ?? ()
> (gdb) x/6i $eip
> 0x408afaf6 <_ZN5gnash6Player3runEiPPcPKcS4_+342>: movss
> 0xffffff94(%ebp),%xmm0
> 0x408afafb <_ZN5gnash6Player3runEiPPcPKcS4_+347>: mov (%edi),%eax
> 0x408afafd <_ZN5gnash6Player3runEiPPcPKcS4_+349>: mov %edi,(%esp)
> 0x408afb00 <_ZN5gnash6Player3runEiPPcPKcS4_+352>: cvttss2si %xmm0,%esi
> 0x408afb04 <_ZN5gnash6Player3runEiPPcPKcS4_+356>: call *0x38(%eax)
> 0x408afb07 <_ZN5gnash6Player3runEiPPcPKcS4_+359>: fstps
> 0xffffff94(%ebp)
>
> 2) It seems that gnash dies because it tries to use the "movss"
> instruction which is only available in the SSE instruction set that my
> system does not have.
>
> 3) /proc/cpuinfo:
>
> processor : 0
> vendor_id : AuthenticAMD
> cpu family : 6
> model : 4
> model name : AMD Athlon(tm) processor
> stepping : 2
> cpu MHz : 701.604
> cache size : 256 KB
> fdiv_bug : no
> hlt_bug : no
> f00f_bug : no
> coma_bug : no
> fpu : yes
> fpu_exception : yes
> cpuid level : 1
> wp : yes
> flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca
> cmov pat pse36 mmx fxsr syscall mmxext 3dnowext 3dnow
> bogomips : 1399.19
>
> 4) I suggest you pass the --without-gcc-arch option to configure in
> debian/rules so that the build script don't pass options like
> "-march=prescot" to gcc even if the CPU of the build system supports those.
>
> Please let me know if you are unable to reproduce this bug on hardware
> without SSE support. If you have trouble getting access to such
> hardware I can also do further testing.
______________________________________________
LLama Gratis a cualquier PC del Mundo.
Llamadas a fijos y móviles desde 1 céntimo por minuto.
http://es.voice.yahoo.com
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
