Bug #1494
closed
page fault in nfs_readdir/nfs_bioread
0%
Description
while running a cpdup to a nfs server (linux)
#15 0xc03527a7 in calltrap () at
/usr/src/sys/platform/pc32/i386/exception.s:785
#16 0xe048de99 in nfs_bioread (vp=0xec44a5e8, uio=0xe816fc58, ioflag=0)
at /usr/src/sys/vfs/nfs/nfs_bio.c:680
#17 0xe04a483f in nfs_readdir (ap=0xe816fbf4) at
/usr/src/sys/vfs/nfs/nfs_vnops.c:2246
#18 0xc02402bc in vop_readdir (ops=0xdff7d730, vp=0xec44a5e8,
uio=0xe816fc58, cred=0xe062e618,
eofflag=0xe816fc4c, ncookies=0x0, cookies=0x0) at
/usr/src/sys/kern/vfs_vopops.c:569
(kgdb) fra 16
#16 0xe048de99 in nfs_bioread (vp=0xec44a5e8, uio=0xe816fc58, ioflag=0)
at /usr/src/sys/vfs/nfs/nfs_bio.c:680
680 dp = (struct nfs_dirent *)cpos;
(kgdb) p cpos
$1 = 0xce3c1000 <Address 0xce3c1000 out of bounds>
(kgdb) p epos
$2 = 0xce3c1f44 <Address 0xce3c1f44 out of bounds>
(kgdb) p bp->b_data
$3 = 0xce3c0000 "�205�a"
(kgdb) p bp->b_data + boff
$4 = 0xce3c0f44 "�205�a"
kernel core on leaf in ~corecode/crash
Updated by dillon over 15 years ago
:while running a cpdup to a nfs server (linux)
:
:#15 0xc03527a7 in calltrap () at=20
:/usr/src/sys/platform/pc32/i386/exception.s:785
Ok, should be fixed now. It was an unsigned underflow which caused
a szmin() check to fail.-Matt
Updated by alexh over 15 years ago
Seems to be fixed in c0b6e0f5ae8831878d11646c6bd9dc4fec703029.
Cheers,
Alex Hornung