Could it be because I changed the DEFAULT_EAGER_BUF_SIZE in ib.h? I did a fresh cvs checkout today, and installed the patch against that. Its still failing though after further testing, but without any logging info this time, server processes just disappear w/o a segfault
I'll reboot my servers and see if it still happens. On Feb 6, 2008 2:42 PM, Pete Wyckoff <[EMAIL PROTECTED]> wrote: > [EMAIL PROTECTED] wrote on Wed, 06 Feb 2008 14:14 -0600: > > I applied your patch, and got the following immediately on some io: > > > > > > [E 02/06 14:12] Error: openib_check_cq: unknown opcode 11171. > > [E 02/06 14:12] [bt] bin/sbin/pvfs2-server(error+0xca) [0x4293ba] > > [E 02/06 14:12] [bt] bin/sbin/pvfs2-server [0x429f9b] > > [E 02/06 14:12] [bt] bin/sbin/pvfs2-server [0x425dc3] > > [E 02/06 14:12] [bt] bin/sbin/pvfs2-server [0x428129] > > [E 02/06 14:12] [bt] > > bin/sbin/pvfs2-server(BMI_testunexpected+0x19f) [0x > > 424d2f] > > [E 02/06 14:12] [bt] bin/sbin/pvfs2-server [0x4397a7] > > [E 02/06 14:12] [bt] /lib/libpthread.so.0 [0x2ba396d2df1a] > > [E 02/06 14:12] [bt] /lib/libc.so.6(__clone+0x72) [0x2ba39711c602] > > That "can't happen". My teensy patch didn't get anywhere near > there. It just changes some printfs and adds an extra test in the > RTS checking. Are you sure everything in the hardware is still > working? And you recompiled okay? Maybe yank out the printfs in > case there is some memory corruption going on somewhere. > > -- Pete > -- Kyle Schochenmaier _______________________________________________ Pvfs2-developers mailing list [email protected] http://www.beowulf-underground.org/mailman/listinfo/pvfs2-developers
