Skip to contentSkip to conversation

New: From version 2.6.27.8 smbd lock ups kernel after 2-10 hours uptime. Keyboard, sshd, smbd, acpi button not work. Syslog and hdd not activity. but Nmap port scan and pinged normally on all ifaces.

Bugme-new Andrew Morton 2 participants 2 messages Dec 14, 2008 – Dec 15, 2008 1 message not in the archive VFS

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]/

#1 lore raw

(switched to email. Please respond via emailed reply-to-all, not via the bugzilla web interface).

Quoted 20 lines · On Sun, 14 Dec 2008 12:51:44 -0800 (PST) [email protected] wrote:

http://bugzilla.kernel.org/show_bug.cgi?id=12225

Summary: From version 2.6.27.8 smbd lock ups kernel after 2-10
hours uptime. Keyboard, sshd, smbd, acpi button not work. Syslog and hdd not activity. but Nmap port scan and pinged normally on all ifaces.
Product: Platform Specific/Hardware
Version: 2.5
KernelVersion: 2.6.27.8
Platform: All
OS/Version: Linux
Tree: Mainline
Status: NEW
Severity: blocking
Priority: P1
Component: i386
AssignedTo: [email protected]
ReportedBy: [email protected]

Quoted 2 lines

Latest working kernel version: 2.6.27.7
Earliest failing kernel version: 2.6.27.8

A pretty small number of commits there.

Quoted 67 lines

Distribution:
Hardware Environment: First server - Athlon64 3000+ , ASUS K8N-E motherboard (NForce3, CK8S Ethernet), video S3 Trio 64V2 PCI, Comport PCI 9835 (2 COM), 3 network PCI-s (3c905B, DGE-528T, DL10050), soft md RAID on 2 IDE HDD. Second server: AthlonXP+ 2600, A7N8X-E Deluxe motherboard (NForce2), 3 DVB SkyStar3 PCI, 2 IDE HDD and 2 SATA HDD, video S3 Inc. ViRGE/DX. Three server: Intel Celeron 1100, ECS P6IPAT motherboard (i815e), 2 IDE HDD, 2 DVB Skystar2 PCI, 1 DVB SkyStar1 PCI, network RTL8139D and DL2000. all motherboard BIOS-s is latest version

Software Environment: linux debian unstable, all pkg is up-to-date

Problem Description:

Quoted 1 line

From version 2.6.27.8 smbd lock ups kernel after 1-10 hours uptime. Keyboard,

sshd, smbd, acpi button not work. Syslog and hdd not activity. but Nmap port scan and pinged normally on all ifaces.

tested configuration:
Athlon64 2.6.27.7 - work
Athlon64 2.6.27.8 - lock (gcc 3.4.6, 4.2.3) AthlonXP 2.6.27.7 - work
AthlonXP 2.6.27.8 and .9 - lock (gcc 3.4.6 4.2.3) IntelCel1100 2.6.27.7 - work
IntelCel1100 2.6.27.9 - lock (gcc 4.1.3)
for each server - kernel compiled with own processor option.

In lock state - keyboard not work, acpi power button to shutdown not work, sshd not work, apache not work, smb not work, but network scan nmap show this ports and ping is work. Hard drive led indicator not signal. /var/log/syslog not contain log string on this lock. Only powerbutton 5 sec or hard reset to reboot. Softdog daemon and module is runned and not work in lock state.

Lock up's server five times - usually after work day in 18-19:00. Personal see server stops in 9:00 morning and reboot it three servers to work... Only one time after disable screen saver console I foto lock screen AthlonXP+ server and typed manually from foto:

tda10086 snd_intel8x0 snd_mpu401 snd_ac97_codec snd_cs4232 snd_opl3_lib snd_hwdep snd_cs4231_1ib ac97_bus snd_pcm_oss snd_pcm snd_mpu401_uart snd_mixer_oss rtc_cmos snd_rawmidi snd_seq_oss rtc_core rtc_lib sata_sil floppy snd_seq_midi_event budget budget_core saa7146 libata snd_seq ttpci_eeprom scsi_mod dvb_core snd_timer snd_seq_device dock 8250_pnp skge snd soundcore forcedeth snd_page_alloc ehci_hcd ohci_hcd usbcore i2c_nforce2 i2c_core 8250 serial_core parport_pc parport thermal button processor nvidia_agp agpgart evdev unix

Pid:7896, comm: smbd Not tainted (2.6.27.9 #15)
EIP: 0060:[<c02d4dd5>l EFLAGS: 00000246 CPU: 0 EIP is at __mutex_lock_slowpath+0x143/0x154
EAX: ef7931ec EBX: c8ae1f28 ECX: c8ae0000 EDX: c8ae0000
ESI: ef7931e8 EDI: 00000246 EBP: ef875550 ESP: c8ae1f24
DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 0068
CR0: 8005003b CR2: b786be90 CR3: 23451000 CR4: 000006f0
DR0: 00000000 DR1: 00000000 DR2: 00000000 DR3: 00000000
DR6: ffff0ff0 DR7: 00000400
[<c017f7cb>] ? inotify_destroy+0x67/0xb1 [<c017f918>] ? inotify_release+0x17/0xb5 [<c015edfd>] ? __fput+0x8a/0x138
[<c015c93f>l ? filp_close+0x4d/0x53 [<c015c991>l ? sys_close+0x4c/0x7a [<c010369a>] 7 syscall_call+0x7/0xb
=============================

On Athlon64 NX-bit is work, test_nx is syslog.
In this time this 3 servers work normally (4 days uptime) on 2.6.27.7 kernel.

Quoted 7 lines

Steps to reproduce:
compile 2.6.27.8 or .9 kernel, run samba 3.2.5 daemon to share file on network... and server lock up's and not answer in any way, only pinged on all network interfaces.
config file from AthlonXP+ 2600
http://rapidshare.com/files/173376113/config.rar

I'd say that we regressed due to

commit 839bbb396bf8e6e966d36daeae89a7787e451909

Author: Al Viro <[email protected]>
Date:   Sat Nov 15 01:15:43 2008 +0000

Fix inotify watch removal/umount races

commit 8f7b0ba1c853919b85b54774775f567f30006107 upstream.

#2 lore raw
Quoted 10 lines · Andrew Morton
Quoted 7 lines

Steps to reproduce:
compile 2.6.27.8 or .9 kernel, run samba 3.2.5 daemon to share file on network... and server lock up's and not answer in any way, only pinged on all network interfaces.
config file from AthlonXP+ 2600
http://rapidshare.com/files/173376113/config.rar

I'd say that we regressed due to

Quoted 7 lines

commit 839bbb396bf8e6e966d36daeae89a7787e451909

Author: Al Viro <[email protected]>
Date:   Sat Nov 15 01:15:43 2008 +0000

Fix inotify watch removal/umount races

commit 8f7b0ba1c853919b85b54774775f567f30006107 upstream.

Hm, I have reports that by adding the following git commit:

	http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=711a49a07f84f914aac26a52143f6e7526571143

it fixes the samba issues for them.

Can you also verify this?