[EN] [SOLVED] kernel 3.5 towo produces filesystem errors

[clubex]

DU a few minutes ago produces filesystem errors and reports of many KDE application rcs not writeable eg. dolphinrc. I suspect the kernel upgrade is the problem.

That's all for now as I'm off to bed. More info tomorrow if required.

[devil]

I cannot confirm that (on a 64-bit razor-qt box). Bool is clean and nothing strange after boot. Will upgrade KDE in the morning.

greetz
devil

[der_bud]

clubex, all fine here on 32-bit KDE fresh DU this morning.

Code: [Select]

Kernel: 3.5-0.towo-siduction-686 i686 (32 bit)
Desktop: KDE 4.8.4
Distro: siduction 12.1 Desperado
Check the usual suspects, like filesystem full (df -h) or apt-mirror problem (DU again).

[Lanzi]

The new kernel uses ext4 CRC32-hases to verify the data, if I understood corectly. So maybe your system is producing wrong hashes?

[clubex]

der_bud: 2.3G remaining. Redid du to no avail. Latest du has linux-libc-dev and whois which shouldn't help.

Lanzi: Could be as my system is all ext3. What consequences would this have?

More info:
After the DU the system boots to the KDE login. I can login to KDE via login dialogue but it hangs during the "disk drive" symbol. Reverting to kernel 3.4.5 via CTRL+ALT+F1 reveals system errors on the /home directory partition: which were repaired. Rebooting into kernel 3.5 I could log into KDE OK but all kde applications reported that their rcs were "not writeable". Non KD apps reported that the app was already running when it wasn't

On reverting to 3.4.5 again there were more errors in /home partition. Curiously kernel 3.5 doesn't appear in kernel-remover.

inxi -v 2
System:    Host: westfield Kernel: 3.4-5.towo-siduction-amd64 x86_64 (64 bit)
           Desktop: KDE 4.8.4 Distro: sidux 2008-04 Πόντος - kde-lite - (200812222256)
Machine:   Mobo: ASUSTeK model: M2N68-AM Plus version: Rev X.0x Bios: American Megatrends version: 1002 date: 11/19/2009
CPU:       Dual core AMD Athlon II X2 240 (-MCP-) clocked at 800.00 MHz
Graphics:  Card: NVIDIA GF119 [GeForce GT 520] X.Org: 1.12.3 driver: nvidia Resolution: 1680x1050@59.9hz
           GLX Renderer: GeForce GT 520/PCIe/SSE2 GLX Version: 4.2.0 NVIDIA 302.17
Network:   Card: NVIDIA MCP61 Ethernet driver: forcedeth
Drives:    HDD Total Size: 250.1GB (34.3% used) 1: model: WDC_WD2500AAJB
Info:      Processes: 138 Uptime: 1:27 Memory: 1085.9/3959.9MB Client: Shell inxi: 1.8.4

Edit: I should have added that kernels 3.4.3, 3.4.4 and 3.4.5 do not create these error/problems

[devil]

 Curiously kernel 3.5 doesn't appear in kernel-remover.

The latest and the booted kernel never appear there.
Other than that, this all sounds very strange. Your guess, that ext3 is the issue might be true, but it still is far fetched. Lets see if towo has any insight on this. So far there was noone else hit by anything like this.

greetz
devil

[clubex]

Devil: Just to get his straight kernel 3.5 won't appear in kernel-remover even when the system is running under into kernel 3.4.5?

Yes it's all rather strange. Hopefully towo will have some ideas. In the meantime I'm going to boot back into kernel 3.5 later today and see what the logs throw up.

[devil]

kern.log or dmesg output would indeed be helpful.

greetz
devil

[clubex]

Here's the tail end of kern.log using kernel 3.5

 nf_conntrack: automatic helper assignment is deprecated and it will be removed soon. Use the iptables CT target to attach helpers instead.
Jul 23 13:36:32 westfield kernel: [  169.264501] NVRM: Your system is not currently configured to drive a VGA console
Jul 23 13:36:32 westfield kernel: [  169.264512] NVRM: on the primary VGA device. The NVIDIA Linux graphics driver
Jul 23 13:36:32 westfield kernel: [  169.264519] NVRM: requires the use of a text-mode VGA console. Use of other console
Jul 23 13:36:32 westfield kernel: [  169.264525] NVRM: drivers including, but not limited to, vesafb, may result in
Jul 23 13:36:32 westfield kernel: [  169.264530] NVRM: corruption and stability problems, and is not supported.
Jul 23 13:36:55 westfield kernel: [  192.010352] ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6
Jul 23 13:36:55 westfield kernel: [  192.010363] ata1.00: BMDMA stat 0x64
Jul 23 13:36:55 westfield kernel: [  192.010372] ata1.00: failed command: WRITE DMA
Jul 23 13:36:55 westfield kernel: [  192.010388] ata1.00: cmd ca/00:58:c8:e1:d4/00:00:00:00:00/e1 tag 0 dma 45056 out
Jul 23 13:36:55 westfield kernel: [  192.010388]          res 51/84:00:1f:e2:d4/00:00:00:00:00/e1 Emask 0x10 (ATA bus error)
Jul 23 13:36:55 westfield kernel: [  192.010397] ata1.00: status: { DRDY ERR }
Jul 23 13:36:55 westfield kernel: [  192.010402] ata1.00: error: { ICRC ABRT }
Jul 23 13:36:55 westfield kernel: [  192.010439] ata1: soft resetting link
Jul 23 13:36:55 westfield kernel: [  192.167619] ata1: nv_mode_filter: 0x7f39f&0x7f39f->0x7f39f, BIOS=0x7f000 (0xc7c50000) ACPI=0x7f01f (15:30:0x15)
Jul 23 13:36:55 westfield kernel: [  192.167635] ata1: nv_mode_filter: 0x1f39f&0x1f39f->0x1f39f, BIOS=0x1f000 (0xc7c50000) ACPI=0x1f01f (15:30:0x15)
Jul 23 13:36:55 westfield kernel: [  192.174168] ata1.00: configured for UDMA/133
Jul 23 13:36:55 westfield kernel: [  192.179941] ata1.01: configured for UDMA/66
Jul 23 13:36:55 westfield kernel: [  192.181089] sd 0:0:0:0: [sda]  
Jul 23 13:36:55 westfield kernel: [  192.181098] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
Jul 23 13:36:55 westfield kernel: [  192.181105] sd 0:0:0:0: [sda]  
Jul 23 13:36:55 westfield kernel: [  192.181109] Sense Key : Aborted Command [current] [descriptor]
Jul 23 13:36:55 westfield kernel: [  192.181116] Descriptor sense data with sense descriptors (in hex):
Jul 23 13:36:55 westfield kernel: [  192.181121]         72 0b 47 00 00 00 00 0c 00 0a 80 00 00 00 00 00
Jul 23 13:36:55 westfield kernel: [  192.181142]         01 d4 e2 1f
Jul 23 13:36:55 westfield kernel: [  192.181152] sd 0:0:0:0: [sda]  
Jul 23 13:36:55 westfield kernel: [  192.181159] Add. Sense: Scsi parity error
Jul 23 13:36:55 westfield kernel: [  192.181165] sd 0:0:0:0: [sda] CDB:
Jul 23 13:36:55 westfield kernel: [  192.181168] Write(10): 2a 00 01 d4 e1 c8 00 00 58 00
Jul 23 13:36:55 westfield kernel: [  192.181186] end_request: I/O error, dev sda, sector 30728648
Jul 23 13:36:55 westfield kernel: [  192.181195] Buffer I/O error on device sda2, logical block 1546
Jul 23 13:36:55 westfield kernel: [  192.181200] lost page write due to I/O error on sda2
Jul 23 13:36:55 westfield kernel: [  192.181236] ata1: EH complete
Jul 23 13:36:55 westfield kernel: [  192.181878] ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6
Jul 23 13:36:55 westfield kernel: [  192.181887] ata1.00: BMDMA stat 0x64
Jul 23 13:36:55 westfield kernel: [  192.181894] ata1.00: failed command: WRITE DMA
Jul 23 13:36:55 westfield kernel: [  192.181909] ata1.00: cmd ca/00:58:28:e2:d4/00:00:00:00:00/e1 tag 0 dma 45056 out
Jul 23 13:36:55 westfield kernel: [  192.181909]          res 51/84:00:7f:e2:d4/00:00:00:00:00/e1 Emask 0x10 (ATA bus error)
Jul 23 13:36:55 westfield kernel: [  192.181917] ata1.00: status: { DRDY ERR }
Jul 23 13:36:55 westfield kernel: [  192.181922] ata1.00: error: { ICRC ABRT }
Jul 23 13:36:55 westfield kernel: [  192.181959] ata1: soft resetting link
Jul 23 13:36:55 westfield kernel: [  192.339758] ata1: nv_mode_filter: 0x7f39f&0x7f39f->0x7f39f, BIOS=0x7f000 (0xc7c50000) ACPI=0x7f01f (15:30:0x15)
Jul 23 13:36:55 westfield kernel: [  192.339773] ata1: nv_mode_filter: 0x1f39f&0x1f39f->0x1f39f, BIOS=0x1f000 (0xc7c50000) ACPI=0x1f01f (15:30:0x15)
Jul 23 13:36:55 westfield kernel: [  192.346051] ata1.00: configured for UDMA/133
Jul 23 13:36:55 westfield kernel: [  192.351830] ata1.01: configured for UDMA/66
Jul 23 13:36:55 westfield kernel: [  192.352994] sd 0:0:0:0: [sda]  
Jul 23 13:36:55 westfield kernel: [  192.353004] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
Jul 23 13:36:55 westfield kernel: [  192.353009] sd 0:0:0:0: [sda]  
Jul 23 13:36:55 westfield kernel: [  192.353013] Sense Key : Aborted Command [current] [descriptor]
Jul 23 13:36:55 westfield kernel: [  192.353020] Descriptor sense data with sense descriptors (in hex):
Jul 23 13:36:55 westfield kernel: [  192.353024]         72 0b 47 00 00 00 00 0c 00 0a 80 00 00 00 00 00
Jul 23 13:36:55 westfield kernel: [  192.353044]         01 d4 e2 7f
Jul 23 13:36:55 westfield kernel: [  192.353054] sd 0:0:0:0: [sda]  
Jul 23 13:36:55 westfield kernel: [  192.353060] Add. Sense: Scsi parity error
Jul 23 13:36:55 westfield kernel: [  192.353066] sd 0:0:0:0: [sda] CDB:
Jul 23 13:36:55 westfield kernel: [  192.353069] Write(10): 2a 00 01 d4 e2 28 00 00 58 00
Jul 23 13:36:55 westfield kernel: [  192.353086] end_request: I/O error, dev sda, sector 30728744
Jul 23 13:36:55 westfield kernel: [  192.353159] ata1: EH complete
Jul 23 13:36:55 westfield kernel: [  192.353220] Aborting journal on device sda2.
Jul 23 13:36:55 westfield kernel: [  192.353311] EXT3-fs (sda2): error in ext3_reserve_inode_write: Journal has aborted
Jul 23 13:36:55 westfield kernel: [  192.375501] EXT3-fs (sda2): error in ext3_dirty_inode: Journal has aborted
Jul 23 13:36:55 westfield kernel: [  192.376012] EXT3-fs (sda2): error: ext3_journal_start_sb: Detected aborted journal
Jul 23 13:36:55 westfield kernel: [  192.376025] EXT3-fs (sda2): error: remounting filesystem read-only

I'll check on dmesg.log later today.

[dibl]

I updated and rebooted 4 systems this morning.  Three are KDE, one is LXDE, all use ext4.  Running 3.5-0, I find no issues in any of them, FYI.

[clubex]

Seems pretty clear that towo's 3.5 kernel has compatibility issues with ext3 but not with ext4

[towo]

You should have a further look at your paste, the problem seems your controler since there are tons of errors. I don't think it's a problem with ext3.

[OppaErich]

With a bit of luck just a bad SATA cable.

[clubex]

towo: And I don't think it's an error with ext3 but with the kernel + ext3 file systems. Your kernels 3.4-5 or 3.4-4 and 3.4-3 work fine with my ext3 system just not with 3.5. (I'm using 3.4-5 while I'm writing this.)

I'm going to try reinstalling kernel 3.5 as maybe a glitch occurred during installation. As kernel-remover can't remove the latest kernel would using apt-get remove --purge the 3.5 headers and image from within kernel 3.4-5 be OK?

[towo]

And I don't think it's an error with ext3 but with the kernel + ext3 file systems.

Code: [Select]

towo:Defiant> uname -a
Linux Defiant 3.5-0.towo-siduction-amd64 #1 SMP PREEMPT Sun Jul 22 12:10:39 UTC 2012 x86_64 GNU/Linux

/media/video
towo:Defiant> mount | grep ext3
/dev/sdb2 on /media/video type ext3 (rw,relatime,errors=continue,barrier=1,data=writeback)
No problems at all.