You are not logged in.

#1 2010-01-27 18:46:35

nogoma
Member
From: Cranston, RI
Registered: 2006-03-01
Posts: 217

[unsolved (spoke too soon)] CD burner hanging

Edit 2: Turned out not to be VMWare; once I started using the machine today (w/o running vmware), the drive went out again.

Edit: Well, I think this is actually a VMWare issue. I usually have VMs running for work, but rebooted at the end of the day yesterday and didn't fire up VMWare, and the drive seems to be behaving itself. So, marking as solved for now, and will follow up on the vmware side...

For the last couple of months, I've had an issue with my PATA CD drive where it randomly "hangs" sometime after booting. At first I thought it was a hardware issue, but replacing the drive yields the same results. I don't see any particular pattern to what triggers it, but it's definitely the case that the drive is fine on initial boot (am able to mount discs, burn via k3b, etc.); eventually, however, the following shows up in dmesg:

[2010-01-27 13:39:02 alfheim:/home/nogoma]
1042 ~ % dmesg|grep hdc
hdc: TSSTcorp CDDVDW SH-S222L, ATAPI CD/DVD-ROM drive
hdc: host max PIO4 wanted PIO255(auto-tune) selected PIO4
hdc: drive side 80-wire cable detection failed, limiting max speed to UDMA33
hdc: UDMA/33 mode selected
ide-cd: hdc: ATAPI 48X DVD-ROM DVD-R/RAM CD-R/RW drive, 2048kB Cache
hdc: request sense failure: status=0xd0 { Busy }
hdc: possibly failed opcode: 0xa0

Which will quickly be followed by many:

dmesg wrote:

INFO: task hald-addon-stor:4207 blocked for more than 120 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
hald-addon-st D f673facc     0  4207   4154 0x00000000
f5dccb60 00000086 f673fb0c f673facc fb2781a3 c30082e0 c1178e60 1773d780
00004d08 00000000 f6919dc0 c14472a0 c14472a0 17731af0 00004d08 f5dccd10
c1442764 f5dccd10 c14472a0 c14472a0 00000000 00000000 f649d7d0 f5deb800
Call Trace:
[<fb2781a3>] ? ide_cd_do_request+0x103/0x260 [ide_cd_mod]
[<c1178e60>] ? delay_tsc+0x90/0xb8
[<f92af8e1>] ? do_ide_request+0x361/0x590 [ide_core]
[<c12b785d>] ? schedule_timeout+0x1ad/0x2d0
[<fb2781a3>] ? ide_cd_do_request+0x103/0x260 [ide_cd_mod]
[<c104e8f7>] ? lock_timer_base+0x27/0x50
[<c104f195>] ? del_timer+0x65/0xe0
[<c12b68d5>] ? wait_for_common+0x95/0x120
[<c10379a0>] ? default_wake_function+0x0/0x10
[<c1165755>] ? blk_execute_rq+0x85/0xf0
[<c1165600>] ? blk_end_sync_rq+0x0/0x30
[<c11571bf>] ? elv_set_request+0xf/0x30
[<c11619fd>] ? get_request+0x32d/0x380
[<c11620cd>] ? get_request_wait+0x1d/0x1a0
[<c1069378>] ? tick_program_event+0x28/0x40
[<c105e75b>] ? hrtimer_interrupt+0x1ab/0x240
[<fb278490>] ? ide_cd_queue_pc+0x160/0x1e0 [ide_cd_mod]
[<c10220ee>] ? hpet_interrupt_handler+0xe/0x30

Which makes sense, as I assume the device can no longer be accessed. In fact, if anything tries to access /dev/hdc at this point, it will simply cause that process to hang and become unkillable. For example, mkinitcpio will fail to run when it's in this state; I have to actually remove the device node ('rm /dev/hdc') in order to get anything that scans devices to work.

I'm currently running kernel v2.6.32 (i686); but, again, this has actually been happening for a while. I've searched both the bbs and google for these sorts of errors, but haven't turned up too much. There's some talk of adding options to grub lines, but none of the suggestions seemed specific enough to simply try and pray. Other details:

1046 ~ % lspci|grep -i ide
00:11.0 SATA controller: ATI Technologies Inc SB700/SB800 SATA Controller [IDE mode]
00:14.1 IDE interface: ATI Technologies Inc SB700/SB800 IDE Controller
1047 ~ % lsmod|grep -i pata
pata_acpi               2264  0
pata_atiixp             2689  0
libata                135162  4 ata_generic,pata_acpi,pata_atiixp,ahci

Last edited by nogoma (2010-02-05 20:03:58)


-nogoma
---
Code Happy, Code Ruby!
http://www.last.fm/user/nogoma/

Offline

#2 2010-05-11 02:11:00

NevarMaor
Member
Registered: 2009-04-22
Posts: 29

Re: [unsolved (spoke too soon)] CD burner hanging

I get the same error messages with my pata cd/dvd burner. It is sr0 on my system though, not hd*. I can read / burn a cd / dvd fine immediately after bootup but if I wait or eject and try a second the device is totally inaccessible. I also have a pata hard drive which is sda and two sata hard drives which are sdb and sdc.

Note the drive works flawlessly under both Windows and Debian. Under Debian the pata hard drive is hda, not sda.

Is this a hal issue or a udev issue? I also cannot find anything definite in searches. It is very frustrating and I would prefer not having to go back to Debian but unless this is solved I may have no choice.

Offline

#3 2010-05-11 13:19:05

nogoma
Member
From: Cranston, RI
Registered: 2006-03-01
Posts: 217

Re: [unsolved (spoke too soon)] CD burner hanging

I had sort of decided this was some weird hardware issue, and stopped investigating it. I "get by" by blacklisting the ide_cd_mod kernel module; when I need to use the drive (relatively infrequently), I modprobe ide_cd_mod, do what needs doing, then rmmod it. This is about 90% effective, but every now and then the problems triggers while the module is loaded, and I have to reboot to get the device working properly. Anyway, from my experience it's really just the existence of the device node in /dev that causes the issue; if I allow the kernel module to load but manually rm the device node, I don't see the device hang. So, given that and the randomness of the timing, this does seem like some process or set of processes accessing the device node in an unexpected manner... it's a frustratingly hard bug to google for; I test it out every few kernel releases, but have had the issues continue. I guess once hal is fully deprecated and I can remove that, it'll be one fewer potential problem source.


-nogoma
---
Code Happy, Code Ruby!
http://www.last.fm/user/nogoma/

Offline

#4 2010-05-13 02:15:03

NevarMaor
Member
Registered: 2009-04-22
Posts: 29

Re: [unsolved (spoke too soon)] CD burner hanging

When I get some more free time I am going to have to investigate the differences between my Arch and Debian installs. Maybe I can narrow down what the problem is with Arch (or maybe I'll just give up and switch back to Debian for good). It just seems to me that this should not be an issue at all. Ide  cd drives should "just work" by now.

Offline

Board footer

Powered by FluxBB