Hi,

we got a gcc crash during Sparc package build on buildd servers.

could be same kind of issue.




/usr/bin/g++ -c -Wall -Wno-format-y2k -pthread -O2 -fstack-protector --param=ssp-buffer-size=4 -Wformat -Werror=format-security -fPIC -DNDEBUG -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -D_FORTIFY_SOURCE=2 -D_MT -D_REENTRANT -D_THREAD_SAFE -I/build/buildd-ncbi-blast+_2.2.27-1-sparc-CRxRu2/ncbi-blast+-2.2.27/c++/BUILD/inc -I/build/buildd-ncbi-blast+_2.2.27-1-sparc-CRxRu2/ncbi-blast+-2.2.27/c++/include /build/buildd-ncbi-blast+_2.2.27-1-sparc-CRxRu2/ncbi-blast+-2.2.27/c++/src/objects/genomecoll/genome_collection__.cpp -o genome_collection__.o
/usr/bin/g++ -c -Wall -Wno-format-y2k -pthread -O2 -fstack-protector --param=ssp-buffer-size=4 -Wformat -Werror=format-security -fPIC -DNDEBUG -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -D_FORTIFY_SOURCE=2 -D_MT -D_REENTRANT -D_THREAD_SAFE -I/build/buildd-ncbi-blast+_2.2.27-1-sparc-CRxRu2/ncbi-blast+-2.2.27/c++/BUILD/inc -I/build/buildd-ncbi-blast+_2.2.27-1-sparc-CRxRu2/ncbi-blast+-2.2.27/c++/include /build/buildd-ncbi-blast+_2.2.27-1-sparc-CRxRu2/ncbi-blast+-2.2.27/c++/src/objects/genomecoll/genome_collection___.cpp -o genome_collection___.o
In file included from /build/buildd-ncbi-blast+_2.2.27-1-sparc-CRxRu2/ncbi-blast+-2.2.27/c++/src/objects/genomecoll/genome_collection___.cpp:21:0:
/build/buildd-ncbi-blast+_2.2.27-1-sparc-CRxRu2/ncbi-blast+-2.2.27/c++/src/objects/genomecoll/NCBI_GenomeCollection_module.cpp:87:1: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.
See <file:///usr/share/doc/gcc-4.6/README.Bugs> for instructions.
The bug is not reproducible, so it is likely a hardware or OS problem.
make[5]: *** [genome_collection___.o] Error 1
Full log is available [0]



If this is not the same issue, please tell me so that I create a new
bug.



Thanks



Olivier



[0]
https://buildd.debian.org/status/fetch.php?pkg=ncbi-blast%2B&arch=sparc&ver=2.2.27-1&stamp=1348478235

--


gpg key id: 4096R/326D8438 (keyring.debian.org)
Key fingerprint = 5FB4 6F83 D3B9 5204 6335 D26D 78DC 68DB 326D 8438

On 27.09.2012 09:38, Olivier Sallou wrote:
> The bug is not reproducible, so it is likely a hardware or OS problem.

so what didn't you understand?


--
To UNSUBSCRIBE, email to debian-gcc-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 50645680.7040200@debian.org">http://lists.debian.org/50645680.7040200@debian.org

Le 9/27/12 3:37 PM, Matthias Klose a
écrit*:



On 27.09.2012 09:38, Olivier Sallou wrote:


The bug is not reproducible, so it is likely a hardware or OS problem.



so what didn't you understand?

This occurs on buildd servers, on Sparc only.* I saw this bug faced
the same error, so I wonder if those are related.

compiler creates a seg fault, not only an error. And log specifies
to submit a bug:


/build/buildd-ncbi-blast+_2.2.27-1-sparc-CRxRu2/ncbi-blast+-2.2.27/c++/src/objects/genomecoll/NCBI_GenomeCollection_module.cpp:87:1: internal compiler error: Segmentation fault
Please submit a full bug report,


Now, as it impacts build on Debian servers, I can't say if it is OS
or hardware issue.



Olivier













--


gpg key id: 4096R/326D8438 (keyring.debian.org)
Key fingerprint = 5FB4 6F83 D3B9 5204 6335 D26D 78DC 68DB 326D 8438

Matthias Klose <doko@debian.org> writes:

> On 27.09.2012 09:38, Olivier Sallou wrote:
>> The bug is not reproducible, so it is likely a hardware or OS problem.
>
> so what didn't you understand?

How GCC is so sure of that. I now see that it does literally try to
reproduce the error by re-running cc1plus. ncbi-blast+ will need a new
upload soon anyway to address other bugs, so there's probably no point
in requesting a give-back.

Sorry to bother you!

--
Aaron M. Ucko, KB1CJC (amu at alum.mit.edu, ucko at debian.org)
http://www.mit.edu/~amu/ | http://stuff.mit.edu/cgi/finger/?amu@monk.mit.edu


--
To UNSUBSCRIBE, email to debian-gcc-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: udl3923mqhz.fsf@dr-wily.mit.edu">http://lists.debian.org/udl3923mqhz.fsf@dr-wily.mit.edu