Bug 795195

Summary: fs/dcache.c:2458 prepend_path+0x15a/0x170() Root dentry has weird name <>
Product: [Fedora] Fedora Reporter: Wendell Baker <wendellcraigbaker>
Component: kernelAssignee: Eric Sandeen <esandeen>
Status: CLOSED DUPLICATE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 16CC: gansalmon, itamar, jonathan, kernel-maint, madhu.chinakonda
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2012-03-07 19:44:20 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Attachments:
Description Flags
/var/log/messages for the time period a few minutes prior to the incident until stability is recovered none

Description Wendell Baker 2012-02-20 00:36:34 UTC
Description of problem:

Kernel complains ...

Feb 19 16:15:01 wrinklie kernel: [4264330.074159] ------------[ cut here ]------------
Feb 19 16:15:01 wrinklie kernel: [4264330.074186] WARNING: at fs/dcache.c:2458 prepend_path+0x15a/0x170()
Feb 19 16:15:01 wrinklie kernel: [4264330.074194] Hardware name: Stylistic ST5020D
Feb 19 16:15:01 wrinklie kernel: [4264330.074201] Root dentry has weird name <>
Feb 19 16:15:01 wrinklie kernel: [4264330.074207] Modules linked in: nfs fscache auth_rpcgss nfs_acl lockd ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 nf_conntrack_ipv4 nf_defrag_ipv4 ip6table_filter xt_state nf_conntrack ip6_tables snd_intel8x0m snd_intel8x0 snd_ac97_codec ac97_bus snd_seq snd_seq_device apanel snd_pcm input_polldev arc4 ath5k ath mac80211 snd_timer microcode tg3 snd cfg80211 irda rfkill crc_ccitt fujitsu_laptop i2c_i801 iTCO_wdt iTCO_vendor_support soundcore snd_page_alloc serio_raw uinput sunrpc xts gf128mul dm_crypt firewire_ohci firewire_core crc_itu_t yenta_socket i915 drm_kms_helper drm i2c_algo_bit i2c_core video [last unloaded: scsi_wait_scan]
Feb 19 16:15:01 wrinklie kernel: [4264330.074350] Pid: 28242, comm: sudo Not tainted 3.1.6-1.fc16.i686 #1
Feb 19 16:15:01 wrinklie kernel: [4264330.074358] Call Trace:
Feb 19 16:15:01 wrinklie kernel: [4264330.074376]  [<c044c122>] warn_slowpath_common+0x72/0xa0
Feb 19 16:15:01 wrinklie kernel: [4264330.074388]  [<c053ab0a>] ? prepend_path+0x15a/0x170
Feb 19 16:15:01 wrinklie kernel: [4264330.074399]  [<c053ab0a>] ? prepend_path+0x15a/0x170
Feb 19 16:15:01 wrinklie kernel: [4264330.074410]  [<c044c1f3>] warn_slowpath_fmt+0x33/0x40
Feb 19 16:15:01 wrinklie kernel: [4264330.074421]  [<c053ab0a>] prepend_path+0x15a/0x170
Feb 19 16:15:01 wrinklie kernel: [4264330.074433]  [<c053c7ff>] sys_getcwd+0xbf/0x170
Feb 19 16:15:01 wrinklie kernel: [4264330.074447]  [<c090d4f4>] syscall_call+0x7/0xb
Feb 19 16:15:01 wrinklie kernel: [4264330.074456] ---[ end trace 56cb2ff4cada9e8b ]---


More generally ...

$ sudo grep WARNING /var/log/messages*
/var/log/messages:Feb 19 16:15:01 wrinklie kernel: [4264330.074186] WARNING: at fs/dcache.c:2458 prepend_path+0x15a/0x170()
/var/log/messages:Feb 19 16:15:01 wrinklie kernel: [4264330.109741] WARNING: at fs/dcache.c:2458 prepend_path+0x15a/0x170()
/var/log/messages:Feb 19 16:19:08 wrinklie kernel: [4264577.311191] WARNING: at fs/dcache.c:2458 prepend_path+0x15a/0x170()
/var/log/messages:Feb 19 16:19:59 wrinklie kernel: [4264628.539660] WARNING: at fs/dcache.c:2458 prepend_path+0x15a/0x170()
/var/log/messages:Feb 19 16:19:59 wrinklie kernel: [4264628.573399] WARNING: at fs/dcache.c:2458 prepend_path+0x15a/0x170()
/var/log/messages:Feb 19 16:20:04 wrinklie kernel: [4264633.012196] WARNING: at fs/dcache.c:2458 prepend_path+0x15a/0x170()
/var/log/messages:Feb 19 16:20:04 wrinklie kernel: [4264633.045625] WARNING: at fs/dcache.c:2458 prepend_path+0x15a/0x170()


It's fairly common across my site (from the central syslog collector box).
It occurs across a range of hardware (though nearly all Fujitsu Stylistic equipment with various network and wireless peripherals).

$ sudo grep -c 'kernel.*WARNING.*prepend_path' /var/log/messages*
/var/log/messages:6
/var/log/messages-20111211:0
/var/log/messages-20111218:0
/var/log/messages-20111225:7
/var/log/messages-20120101:2
/var/log/messages-20120108:0
/var/log/messages-20120115:114
/var/log/messages-20120122:5
/var/log/messages-20120129:24
/var/log/messages-20120205:0
/var/log/messages-20120212:35
/var/log/messages-20120219:0


Version-Release number of selected component (if applicable):

Kernel versions range across 3.1 and 3.2 in the Fedora 16 era.
This instance occurred on

$ uname -a
Linux wrinklie 3.1.6-1.fc16.i686 #1 SMP Wed Dec 21 23:18:01 UTC 2011 i686 i686 i386 GNU/Linux


How reproducible:

The problem is there ... it reproduces itself over time.
It is not excitable by a deterministic pattern (that I know of yet).

Steps to Reproduce:
1.
2.
3.
  
Actual results:

should never occur...

Expected results:


Additional info:


Other patterns (this is a separate occurrence from the instance reported above)

Feb  7 10:44:14 hippie kernel: [73230.531218] WARNING: at fs/dcache.c:2485 prepend_path+0x15a/0x170()
Feb  7 10:44:14 hippie kernel: [73230.531225] Hardware name: Stylistic ST5112
Feb  7 10:44:14 hippie kernel: [73230.531230] Root dentry has weird name <>
Feb  7 10:44:14 hippie kernel: [73230.531234] Modules linked in: 8139too 8139cp mii nfs fscache rfcomm bnep ip6t_REJECT nf_conntrack_ipv4 nf_conntrack_ipv6 n
f_defrag_ipv4 nf_defrag_ipv6 xt_state ip6table_filter nf_conntrack ip6_tables snd_hda_codec_idt snd_hda_intel snd_hda_codec snd_hwdep snd_seq snd_seq_device 
snd_pcm btusb arc4 iwl3945 iwlegacy nfsd snd_timer mac80211 lockd apanel sky2 input_polldev snd iTCO_wdt iTCO_vendor_support irda soundcore i2c_i801 fujitsu_
laptop bluetooth nfs_acl cfg80211 snd_page_alloc auth_rpcgss crc_ccitt uinput microcode serio_raw rfkill sunrpc sdhci_pci sdhci firewire_ohci mmc_core firewi
re_core crc_itu_t yenta_socket i915 drm_kms_helper drm i2c_algo_bit i2c_core video [last unloaded: scsi_wait_scan]
Feb  7 10:44:14 hippie kernel: [73230.531366] Pid: 26180, comm: sh Tainted: G        W    3.2.3-2.fc16.i686.PAE #1
Feb  7 10:44:14 hippie kernel: [73230.531372] Call Trace:
Feb  7 10:44:14 hippie kernel: [73230.531384]  [<c045e352>] warn_slowpath_common+0x72/0xa0
Feb  7 10:44:14 hippie kernel: [73230.531394]  [<c05536ba>] ? prepend_path+0x15a/0x170
Feb  7 10:44:14 hippie kernel: [73230.531402]  [<c05536ba>] ? prepend_path+0x15a/0x170
Feb  7 10:44:14 hippie kernel: [73230.531411]  [<c045e423>] warn_slowpath_fmt+0x33/0x40
Feb  7 10:44:14 hippie kernel: [73230.531420]  [<c05536ba>] prepend_path+0x15a/0x170
Feb  7 10:44:14 hippie kernel: [73230.531429]  [<c05553cf>] sys_getcwd+0xbf/0x170
Feb  7 10:44:14 hippie kernel: [73230.531439]  [<c094349f>] sysenter_do_call+0x12/0x28
Feb  7 10:44:14 hippie kernel: [73230.531446] ---[ end trace ecaafbbea8f577c4 ]---

Comment 1 Dave Jones 2012-02-20 00:59:42 UTC
was this after a suspend/hibernate ?

Comment 2 Wendell Baker 2012-02-20 02:34:54 UTC
Created attachment 564273 [details]
/var/log/messages for the time period a few minutes prior to the incident until stability is recovered

No.  This is not after a suspend/hibernate.  None of the events that I'm describing are after a suspend or hibernate. Although these units are tablet form factor they are basically sitting in a machine room closet at runlevel 3 on AC all the time.  I use them as lowrange compute servers (servers, routers, etc.)

This behavior did not occur in Fedora 13 or 14.  This behavior started with Fedora 16.  I skipped Fedora 15.

My daily driver tablet is on Fedora 14, it does not exhibit this (though it is different hardware, being an x86_64 not an i686 as is the case here).  It is hibernated and suspended regularly and this issue is not exhibited on that machine.

The user-level event just prior to the occurrence with 'wrinklie' was a 'make install' activity that threw up into /etc/sysconfig/network-scripts two files that NetworkManager had to respond to.

1. a modification of /etc/sysconfig/network-scripts/ifcfg-em1
2. a new installation of /etc/sysconfig/network-scripts/ifcfg-wlan0

This triggered a reenablement of the wireline and an attempt to bring up wlan0.

This 'make install' activity was performed within an nfs-mounted directory /srv, which is shown as causing 'umount.nfs: /srv: device is busy'  This is expected; there is no problem to be hunted in that part of the trace.

Stability was recovered; wrinklie not need rebooting  There are no further issues.

Unrelated (probably), wlan0 won't associate.  It has behavior smilar to
https://bugzilla.redhat.com/show_bug.cgi?id=795189
I reported that against an intel wifi chip on a different machine.  But 'wrinklie' has atheros wifi gear on it.  So maybe this is a general behavior in the 3-series kernels.   A wide range of Atheros and Intel network interfaces seem to be happy on the 2.6-series kernels of Fedora 14.

Comment 3 Dave Jones 2012-03-07 19:44:20 UTC

*** This bug has been marked as a duplicate of bug 787171 ***