All of lore.kernel.org
 help / color / mirror / Atom feed
* nfs_lookup_revalidate BUG ?
@ 2012-08-14 10:00 Richard Ems
  2012-08-14 18:04 ` Myklebust, Trond
  0 siblings, 1 reply; 11+ messages in thread
From: Richard Ems @ 2012-08-14 10:00 UTC (permalink / raw)
  To: linux-nfs

Hi all !

We got the following BUG already 9 times in the last 6 days on 9 different nodes of our HPC cluster.
I searched but couldn't find this same BUG anywhere, but a similar one related to ecryptfs which we are not using.

The servers are running openSUSE 12.1 with kernel 3.3.6.
The nodes were updated from openSUSE 11.3 and kernel 2.6.34.7-0.5-default to openSUSE 12.1 and kernel 3.5.0.
Servers and nodes are running util-linux-2.20.1 and are all 64 bit systems.

The error appeared only after this update. Also after updating to kernel 3.5.1 we are getting the same error.


The BUG is triggered by a java application that runs as a batch job for hours on the nodes.
We are using autofs-5.0.7 on the HPC nodes and mounting only using NFS_V3, not using NFS_V4 at all.

On one of these nodes the mount options are:

c5n12:~ # mount | grep nfs
nfsd on /proc/fs/nfsd type nfsd (rw,relatime)
fs1:/data_4/ on /net/fs1/data_4 type nfs (rw,nosuid,nodev,relatime,vers=3,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=10.0.3.203,mountvers=3,mountport=58521,mountproto=udp,local_lock=none,addr=10.0.3.203)
fs1:/data_1/ on /net/fs1/data_1 type nfs (rw,nosuid,nodev,relatime,vers=3,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=10.0.3.203,mountvers=3,mountport=58521,mountproto=udp,local_lock=none,addr=10.0.3.203)
c3m:/opt/ on /net/c3m/opt type nfs (rw,nosuid,nodev,relatime,vers=3,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=10.0.3.201,mountvers=3,mountport=35374,mountproto=udp,local_lock=none,addr=10.0.3.201)


exportfs -v on the servers show all lines like:
/data_1         *.c5.xxx.com(rw,wdelay,no_root_squash,no_subtree_check)


Aug 14 06:25:00 c5n12 kernel: [53043.599388] BUG: unable to handle kernel NULL pointer dereference at 0000000000000038
Aug 14 06:25:00 c5n12 kernel: [53043.599523] IP: [<ffffffffa03789cd>] nfs_lookup_revalidate+0x2d/0x480 [nfs]
Aug 14 06:25:00 c5n12 kernel: [53043.599604] PGD 337c63067 PUD 0 
Aug 14 06:25:00 c5n12 kernel: [53043.599668] Oops: 0000 [#1] SMP 
Aug 14 06:25:00 c5n12 kernel: [53043.599732] CPU 5 
Aug 14 06:25:00 c5n12 kernel: [53043.599737] Modules linked in: nfs fscache nfsd lockd nfs_acl auth_rpcgss sunrpc af_packet binfmt_misc cpufreq_conservative cpufreq_userspace cpufreq_powersave dm_mod acpi_cpufreq mperf coretemp gpio_ich kvm_intel joydev kvm ioatdma hid_generic igb lpc_ich i7core_edac edac_core ptp serio_raw dca pcspkr i2c_i801 mfd_core sg pps_core usbhid crc32c_intel microcode button autofs4 uhci_hcd ttm drm_kms_helper drm i2c_algo_bit sysimgblt sysfillrect syscopyarea ehci_hcd usbcore usb_common scsi_dh_rdac scsi_dh_emc scsi_dh_hp_sw scsi_dh_alua scsi_dh edd fan ata_piix thermal processor thermal_sys
Aug 14 06:25:00 c5n12 kernel: [53043.600411] 
Aug 14 06:25:00 c5n12 kernel: [53043.600466] Pid: 30431, comm: java Not tainted 3.5.1-2-default #1 Supermicro X8DTT/X8DTT
Aug 14 06:25:00 c5n12 kernel: [53043.600594] RIP: 0010:[<ffffffffa03789cd>]  [<ffffffffa03789cd>] nfs_lookup_revalidate+0x2d/0x480 [nfs]
Aug 14 06:25:00 c5n12 kernel: [53043.600723] RSP: 0018:ffff8801b418bd38  EFLAGS: 00010292
Aug 14 06:25:00 c5n12 kernel: [53043.600787] RAX: 00000000fffffff6 RBX: ffff88032016d800 RCX: 0000000000000020
Aug 14 06:25:00 c5n12 kernel: [53043.600854] RDX: ffffffff00000000 RSI: 0000000000000000 RDI: ffff8801824a7b00
Aug 14 06:25:00 c5n12 kernel: [53043.600921] RBP: ffff8801b418bdf8 R08: 7fffff0034323030 R09: fffffffff04c03ed
Aug 14 06:25:00 c5n12 kernel: [53043.600989] R10: ffff8801824a7b00 R11: 0000000000000002 R12: ffff8801824a7b00
Aug 14 06:25:00 c5n12 kernel: [53043.601055] R13: ffff8801824a7b00 R14: 0000000000000000 R15: ffff8803201725d0
Aug 14 06:25:00 c5n12 kernel: [53043.601122] FS:  00002b53a46cb700(0000) GS:ffff88033fc20000(0000) knlGS:0000000000000000
Aug 14 06:25:00 c5n12 kernel: [53043.601241] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Aug 14 06:25:00 c5n12 kernel: [53043.601335] CR2: 0000000000000038 CR3: 000000020a426000 CR4: 00000000000007e0
Aug 14 06:25:00 c5n12 kernel: [53043.601401] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
Aug 14 06:25:00 c5n12 kernel: [53043.601466] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Aug 14 06:25:00 c5n12 kernel: [53043.601532] Process java (pid: 30431, threadinfo ffff8801b418a000, task ffff8801b5d20600)
Aug 14 06:25:00 c5n12 kernel: [53043.601650] Stack:
Aug 14 06:25:00 c5n12 kernel: [53043.601706]  ffff8801b418be44 ffff88032016d800 ffff8801b418bdf8 0000000000000000
Aug 14 06:25:00 c5n12 kernel: [53043.601829]  ffff8801824a7b00 ffff8801b418bdd7 ffff8803201725d0 ffffffff8116a9c0
Aug 14 06:25:00 c5n12 kernel: [53043.601952]  ffff8801b5c38dc0 0000000000000007 ffff88032016d800 0000000000000000
Aug 14 06:25:00 c5n12 kernel: [53043.602076] Call Trace:
Aug 14 06:25:00 c5n12 kernel: [53043.602153]  [<ffffffff8116a9c0>] lookup_dcache+0x80/0xe0
Aug 14 06:25:00 c5n12 kernel: [53043.602220]  [<ffffffff8116aa43>] __lookup_hash+0x23/0x90
Aug 14 06:25:00 c5n12 kernel: [53043.602284]  [<ffffffff8116b4a5>] lookup_one_len+0xc5/0x100
Aug 14 06:25:00 c5n12 kernel: [53043.602355]  [<ffffffffa03869a3>] nfs_sillyrename+0xe3/0x210 [nfs]
Aug 14 06:25:00 c5n12 kernel: [53043.602439]  [<ffffffff8116cadf>] vfs_unlink.part.25+0x7f/0xe0
Aug 14 06:25:00 c5n12 kernel: [53043.602504]  [<ffffffff8116f22c>] do_unlinkat+0x1ac/0x1d0
Aug 14 06:25:00 c5n12 kernel: [53043.602570]  [<ffffffff815717b9>] system_call_fastpath+0x16/0x1b
Aug 14 06:25:00 c5n12 kernel: [53043.602637]  [<00002b5348b5f527>] 0x2b5348b5f526
Aug 14 06:25:00 c5n12 kernel: [53043.602699] Code: ec 38 b8 f6 ff ff ff 4c 89 64 24 18 4c 89 74 24 28 49 89 fc 48 89 5c 24 08 48 89 6c 24 10 49 89 f6 4c 89 6c 24 20 4c 89 7c 24 30 <f6> 46 38 40 0f 85 d1 00 00 00 e8 c4 c4 df e0 48 8b 58 30 49 89 
Aug 14 06:25:00 c5n12 kernel: [53043.603008] RIP  [<ffffffffa03789cd>] nfs_lookup_revalidate+0x2d/0x480 [nfs]
Aug 14 06:25:00 c5n12 kernel: [53043.603080]  RSP <ffff8801b418bd38>
Aug 14 06:25:00 c5n12 kernel: [53043.603140] CR2: 0000000000000038
Aug 14 06:25:00 c5n12 kernel: [53043.603517] ---[ end trace 845113ed191985dd ]---


Is this a known BUG ?
Which other information or test can I do to contribute searching/resolving this issue?

Many thanks,
Richard



-- 
Richard Ems       mail: Richard.Ems@Cape-Horn-Eng.com

Cape Horn Engineering S.L.
C/ Dr. J.J. Dómine 1, 5º piso
46011 Valencia
Tel : +34 96 3242923 / Fax 924
http://www.cape-horn-eng.com

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: nfs_lookup_revalidate BUG ?
  2012-08-14 10:00 nfs_lookup_revalidate BUG ? Richard Ems
@ 2012-08-14 18:04 ` Myklebust, Trond
  2012-08-14 18:35   ` Al Viro
  0 siblings, 1 reply; 11+ messages in thread
From: Myklebust, Trond @ 2012-08-14 18:04 UTC (permalink / raw)
  To: Richard Ems; +Cc: linux-nfs@vger.kernel.org
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^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: nfs_lookup_revalidate BUG ?
  2012-08-14 18:04 ` Myklebust, Trond
@ 2012-08-14 18:35   ` Al Viro
  2012-08-14 18:48     ` Myklebust, Trond
  0 siblings, 1 reply; 11+ messages in thread
From: Al Viro @ 2012-08-14 18:35 UTC (permalink / raw)
  To: Myklebust, Trond; +Cc: Richard Ems, linux-nfs@vger.kernel.org

On Tue, Aug 14, 2012 at 06:04:08PM +0000, Myklebust, Trond wrote:
> > Is this a known BUG ?
> > Which other information or test can I do to contribute searching/resolving this issue?
> 
> I'm not 100% certain, but it looks to me as if the call to dget_parent()
> in nfs_lookup_revalidate is returning NULL.
> 
> Could you please apply the following patch, and see if that triggers the
> WARN_ON instead of the above Oops?

Unlikely as hell.  dget_parent() has only one return in it, and it's
        BUG_ON(!ret->d_count);
        ret->d_count++;
        spin_unlock(&ret->d_lock);
        return ret;
so it's extremely unlikely that it would managed to return NULL.  Not
without oopsing...

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: nfs_lookup_revalidate BUG ?
  2012-08-14 18:35   ` Al Viro
@ 2012-08-14 18:48     ` Myklebust, Trond
  2012-08-14 19:11       ` Jeff Layton
  0 siblings, 1 reply; 11+ messages in thread
From: Myklebust, Trond @ 2012-08-14 18:48 UTC (permalink / raw)
  To: Al Viro; +Cc: Richard Ems, linux-nfs@vger.kernel.org

T24gVHVlLCAyMDEyLTA4LTE0IGF0IDE5OjM1ICswMTAwLCBBbCBWaXJvIHdyb3RlOg0KPiBPbiBU
dWUsIEF1ZyAxNCwgMjAxMiBhdCAwNjowNDowOFBNICswMDAwLCBNeWtsZWJ1c3QsIFRyb25kIHdy
b3RlOg0KPiA+ID4gSXMgdGhpcyBhIGtub3duIEJVRyA/DQo+ID4gPiBXaGljaCBvdGhlciBpbmZv
cm1hdGlvbiBvciB0ZXN0IGNhbiBJIGRvIHRvIGNvbnRyaWJ1dGUgc2VhcmNoaW5nL3Jlc29sdmlu
ZyB0aGlzIGlzc3VlPw0KPiA+IA0KPiA+IEknbSBub3QgMTAwJSBjZXJ0YWluLCBidXQgaXQgbG9v
a3MgdG8gbWUgYXMgaWYgdGhlIGNhbGwgdG8gZGdldF9wYXJlbnQoKQ0KPiA+IGluIG5mc19sb29r
dXBfcmV2YWxpZGF0ZSBpcyByZXR1cm5pbmcgTlVMTC4NCj4gPiANCj4gPiBDb3VsZCB5b3UgcGxl
YXNlIGFwcGx5IHRoZSBmb2xsb3dpbmcgcGF0Y2gsIGFuZCBzZWUgaWYgdGhhdCB0cmlnZ2VycyB0
aGUNCj4gPiBXQVJOX09OIGluc3RlYWQgb2YgdGhlIGFib3ZlIE9vcHM/DQo+IA0KPiBVbmxpa2Vs
eSBhcyBoZWxsLiAgZGdldF9wYXJlbnQoKSBoYXMgb25seSBvbmUgcmV0dXJuIGluIGl0LCBhbmQg
aXQncw0KPiAgICAgICAgIEJVR19PTighcmV0LT5kX2NvdW50KTsNCj4gICAgICAgICByZXQtPmRf
Y291bnQrKzsNCj4gICAgICAgICBzcGluX3VubG9jaygmcmV0LT5kX2xvY2spOw0KPiAgICAgICAg
IHJldHVybiByZXQ7DQo+IHNvIGl0J3MgZXh0cmVtZWx5IHVubGlrZWx5IHRoYXQgaXQgd291bGQg
bWFuYWdlZCB0byByZXR1cm4gTlVMTC4gIE5vdA0KPiB3aXRob3V0IG9vcHNpbmcuLi4NCg0KQWdy
ZWVkLCBidXQgdGhlIE9vcHMgaXMgYSBOVUxMIHBvaW50ZXIgZGVyZWZlcmVuY2Ugb2Ygb2Zmc2V0
IDAwMDM4LA0Kd2hpY2ggbG9va3Mgc3VzcGljaW91c2x5IGxpa2UgYSBkZWZlcmVuY2Ugb2YgKHN0
cnVjdCBkZW50cnkpLT5kX2lub2RlLg0KVGhlcmUgYXJlIG9ubHkgMiBvZiB0aG9zZSBpbiB0aGUg
dmljaW5pdHkgb2YgbmZzX2xvb2t1cF9yZXZhbGlkYXRlKzB4MmQ6DQpvbmUgaXMgYSBkZXJlZmVy
ZW5jZSBvZiBwYXJlbnQtPmRfaW5vZGUsIHRoZSBvdGhlciBpcyBkZW50cnktPmRfaW5vZGUuLi4N
Cg0KLS0gDQpUcm9uZCBNeWtsZWJ1c3QNCkxpbnV4IE5GUyBjbGllbnQgbWFpbnRhaW5lcg0KDQpO
ZXRBcHANClRyb25kLk15a2xlYnVzdEBuZXRhcHAuY29tDQp3d3cubmV0YXBwLmNvbQ0KDQo=

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: nfs_lookup_revalidate BUG ?
  2012-08-14 18:48     ` Myklebust, Trond
@ 2012-08-14 19:11       ` Jeff Layton
  2012-08-15 11:57         ` Richard Ems
  0 siblings, 1 reply; 11+ messages in thread
From: Jeff Layton @ 2012-08-14 19:11 UTC (permalink / raw)
  To: Myklebust, Trond; +Cc: Al Viro, Richard Ems, linux-nfs@vger.kernel.org

On Tue, 14 Aug 2012 18:48:35 +0000
"Myklebust, Trond" <Trond.Myklebust@netapp.com> wrote:

> On Tue, 2012-08-14 at 19:35 +0100, Al Viro wrote:
> > On Tue, Aug 14, 2012 at 06:04:08PM +0000, Myklebust, Trond wrote:
> > > > Is this a known BUG ?
> > > > Which other information or test can I do to contribute searching/resolving this issue?
> > > 
> > > I'm not 100% certain, but it looks to me as if the call to dget_parent()
> > > in nfs_lookup_revalidate is returning NULL.
> > > 
> > > Could you please apply the following patch, and see if that triggers the
> > > WARN_ON instead of the above Oops?
> > 
> > Unlikely as hell.  dget_parent() has only one return in it, and it's
> >         BUG_ON(!ret->d_count);
> >         ret->d_count++;
> >         spin_unlock(&ret->d_lock);
> >         return ret;
> > so it's extremely unlikely that it would managed to return NULL.  Not
> > without oopsing...
> 
> Agreed, but the Oops is a NULL pointer dereference of offset 00038,
> which looks suspiciously like a deference of (struct dentry)->d_inode.
> There are only 2 of those in the vicinity of nfs_lookup_revalidate+0x2d:
> one is a dereference of parent->d_inode, the other is dentry->d_inode...
> 

One way to determine this would be to follow the instructions here,
substituting nfs.ko for cifs.ko:

    http://wiki.samba.org/index.php/LinuxCIFS_troubleshooting#Oopses

...and see if you can get a listing of the spot where it oopsed. You
may need kernel debuginfo installed if you're using a distro kernel
though.

[1]: yes, I know that page is about CIFS troubleshooting, but that
method should work for nfs.ko too.

-- 
Jeff Layton <jlayton@redhat.com>

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: nfs_lookup_revalidate BUG ?
  2012-08-14 19:11       ` Jeff Layton
@ 2012-08-15 11:57         ` Richard Ems
  2012-08-15 13:20           ` Myklebust, Trond
  0 siblings, 1 reply; 11+ messages in thread
From: Richard Ems @ 2012-08-15 11:57 UTC (permalink / raw)
  To: Jeff Layton; +Cc: Myklebust, Trond, Al Viro, linux-nfs@vger.kernel.org

On 08/14/2012 09:11 PM, Jeff Layton wrote:
> One way to determine this would be to follow the instructions here,
> substituting nfs.ko for cifs.ko:
> 
>     http://wiki.samba.org/index.php/LinuxCIFS_troubleshooting#Oopses

Hi all, thanks for your responses. I first tried Jeff's proposal, here are the results:

c5n12:~ # gdb /lib/modules/3.5.1-2-default/kernel/fs/nfs/nfs.ko

GNU gdb (GDB) SUSE (7.3-41.1.2)
Copyright (C) 2011 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-suse-linux".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /lib/modules/3.5.1-2-default/kernel/fs/nfs/nfs.ko...Reading symbols from /usr/lib/debug/lib/modules/3.5.1-2-default/kernel/fs/nfs/nfs.ko.debug...done.
done.

(gdb) list *(nfs_lookup_revalidate+0x2d)
0x59cd is in nfs_lookup_revalidate (/usr/src/debug/kernel-default-3.5.1/linux-3.5/fs/nfs/dir.c:1129).
1124            struct dentry *parent;
1125            struct nfs_fh *fhandle = NULL;
1126            struct nfs_fattr *fattr = NULL;
1127            int error;
1128
1129            if (nd->flags & LOOKUP_RCU)
1130                    return -ECHILD;
1131
1132            parent = dget_parent(dentry);
1133            dir = parent->d_inode;


(gdb) list *(nfs_lookup_revalidate+0x2d/0x480)
0x59a0 is in nfs_lookup_revalidate (/usr/src/debug/kernel-default-3.5.1/linux-3.5/fs/nfs/dir.c:1121).
1116     *
1117     * If the parent directory is seen to have changed, we throw out the
1118     * cached dentry and do a new lookup.
1119     */
1120    static int nfs_lookup_revalidate(struct dentry *dentry, struct nameidata *nd)
1121    {
1122            struct inode *dir;
1123            struct inode *inode;
1124            struct dentry *parent;
1125            struct nfs_fh *fhandle = NULL;


Does this info help?
Trond, should I go ahead and apply that patch to fs/nfs/dir.c ? It's long time ago that I last applied patches to the kernel and recompiled modules ... 8)
And I will have to wait for the bug to trigger, since it just happens only sometimes, it could be days waiting for it.

Many thanks,
Richard


-- 
Richard Ems       mail: Richard.Ems@Cape-Horn-Eng.com

Cape Horn Engineering S.L.
C/ Dr. J.J. Dómine 1, 5º piso
46011 Valencia
Tel : +34 96 3242923 / Fax 924
http://www.cape-horn-eng.com

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: nfs_lookup_revalidate BUG ?
  2012-08-15 11:57         ` Richard Ems
@ 2012-08-15 13:20           ` Myklebust, Trond
  2012-08-15 13:22             ` Richard Ems
  0 siblings, 1 reply; 11+ messages in thread
From: Myklebust, Trond @ 2012-08-15 13:20 UTC (permalink / raw)
  To: Richard Ems; +Cc: Jeff Layton, Al Viro, linux-nfs@vger.kernel.org
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^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: nfs_lookup_revalidate BUG ?
  2012-08-15 13:20           ` Myklebust, Trond
@ 2012-08-15 13:22             ` Richard Ems
  2012-08-15 13:38               ` Myklebust, Trond
  0 siblings, 1 reply; 11+ messages in thread
From: Richard Ems @ 2012-08-15 13:22 UTC (permalink / raw)
  To: Myklebust, Trond; +Cc: Jeff Layton, Al Viro, linux-nfs@vger.kernel.org

On 08/15/2012 03:20 PM, Myklebust, Trond wrote:
>> (gdb) list *(nfs_lookup_revalidate+0x2d)
>> 0x59cd is in nfs_lookup_revalidate (/usr/src/debug/kernel-default-3.5.1/linux-3.5/fs/nfs/dir.c:1129).
>> 1124            struct dentry *parent;
>> 1125            struct nfs_fh *fhandle = NULL;
>> 1126            struct nfs_fattr *fattr = NULL;
>> 1127            int error;
>> 1128
>> 1129            if (nd->flags & LOOKUP_RCU)
> 
> Bummer... We check if 'nd' is NULL everywhere except here...
> 
> OK. Changing the above line to
> 
> 	if (nd && (nd->flags & LOOKUP_RCU))
> 
> will fix the problem.

Ok, great !
Do you think this fix will go into 3.5.2 ?

Thanks again,
Richard


-- 
Richard Ems       mail: Richard.Ems@Cape-Horn-Eng.com

Cape Horn Engineering S.L.
C/ Dr. J.J. Dómine 1, 5º piso
46011 Valencia
Tel : +34 96 3242923 / Fax 924
http://www.cape-horn-eng.com

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: nfs_lookup_revalidate BUG ?
  2012-08-15 13:22             ` Richard Ems
@ 2012-08-15 13:38               ` Myklebust, Trond
  2012-08-15 13:40                 ` Richard Ems
  0 siblings, 1 reply; 11+ messages in thread
From: Myklebust, Trond @ 2012-08-15 13:38 UTC (permalink / raw)
  To: Richard Ems; +Cc: Jeff Layton, Al Viro, linux-nfs@vger.kernel.org
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^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: nfs_lookup_revalidate BUG ?
  2012-08-15 13:38               ` Myklebust, Trond
@ 2012-08-15 13:40                 ` Richard Ems
  2012-08-15 13:55                   ` Myklebust, Trond
  0 siblings, 1 reply; 11+ messages in thread
From: Richard Ems @ 2012-08-15 13:40 UTC (permalink / raw)
  To: Myklebust, Trond; +Cc: Jeff Layton, Al Viro, linux-nfs@vger.kernel.org

On 08/15/2012 03:38 PM, Myklebust, Trond wrote:
> On Wed, 2012-08-15 at 15:22 +0200, Richard Ems wrote:
>> On 08/15/2012 03:20 PM, Myklebust, Trond wrote:
>>>> (gdb) list *(nfs_lookup_revalidate+0x2d)
>>>> 0x59cd is in nfs_lookup_revalidate (/usr/src/debug/kernel-default-3.5.1/linux-3.5/fs/nfs/dir.c:1129).
>>>> 1124            struct dentry *parent;
>>>> 1125            struct nfs_fh *fhandle = NULL;
>>>> 1126            struct nfs_fattr *fattr = NULL;
>>>> 1127            int error;
>>>> 1128
>>>> 1129            if (nd->flags & LOOKUP_RCU)
>>>
>>> Bummer... We check if 'nd' is NULL everywhere except here...
>>>
>>> OK. Changing the above line to
>>>
>>> 	if (nd && (nd->flags & LOOKUP_RCU))
>>>
>>> will fix the problem.
>>
>> Ok, great !
>> Do you think this fix will go into 3.5.2 ?
> 
> I'll try to get it into stable. 3.6-rcX is unaffected by the bug due to
> recent VFS changes that have removed the "struct nameidata *" argument
> to nfs_lookup_revalidate.
> 

Ok, many thanks. I have found a couple of other accesses to nd with no
check for NULL at *nfs_atomic_lookup and nfs4_lookup_revalidate, but
don't know if those are relevant.

Thanks again,
Richard

-- 
Richard Ems       mail: Richard.Ems@Cape-Horn-Eng.com

Cape Horn Engineering S.L.
C/ Dr. J.J. Dómine 1, 5º piso
46011 Valencia
Tel : +34 96 3242923 / Fax 924
http://www.cape-horn-eng.com

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: nfs_lookup_revalidate BUG ?
  2012-08-15 13:40                 ` Richard Ems
@ 2012-08-15 13:55                   ` Myklebust, Trond
  0 siblings, 0 replies; 11+ messages in thread
From: Myklebust, Trond @ 2012-08-15 13:55 UTC (permalink / raw)
  To: Richard Ems; +Cc: Jeff Layton, Al Viro, linux-nfs@vger.kernel.org
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^ permalink raw reply	[flat|nested] 11+ messages in thread

end of thread, other threads:[~2012-08-15 13:56 UTC | newest]

Thread overview: 11+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-08-14 10:00 nfs_lookup_revalidate BUG ? Richard Ems
2012-08-14 18:04 ` Myklebust, Trond
2012-08-14 18:35   ` Al Viro
2012-08-14 18:48     ` Myklebust, Trond
2012-08-14 19:11       ` Jeff Layton
2012-08-15 11:57         ` Richard Ems
2012-08-15 13:20           ` Myklebust, Trond
2012-08-15 13:22             ` Richard Ems
2012-08-15 13:38               ` Myklebust, Trond
2012-08-15 13:40                 ` Richard Ems
2012-08-15 13:55                   ` Myklebust, Trond

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.