ATA errors cause LONG boot time in Feisty

[David Warde-Farley]

After upgrading to feisty boots take a long time, it's stuck at the very first position of the progress bar, and this seems to be what's causing it:

[ 29.128243] ata2: PATA max UDMA/133 cmd 0x00010170 ctl 0x00010376 bmdma 0x0001ffa8 irq 15
[ 29.798376] ata2.00: ATAPI, max UDMA/33
[ 29.798380] ata2.01: ATAPI, max UDMA/33
[ 59.783373] ata2.00: qc timeout (cmd 0xef)
[ 59.783381] ata2.00: failed to set xfermode (err_mask=0x4)
[ 59.783385] ata2: failed to recover some devices, retrying in 5 secs
[ 95.253820] ata2.00: qc timeout (cmd 0xef)
[ 95.253827] ata2.00: failed to set xfermode (err_mask=0x4)
[ 95.253832] ata2.00: limiting speed to UDMA/33:PIO3
[ 95.253834] ata2: failed to recover some devices, retrying in 5 secs
[ 130.724260] ata2.00: qc timeout (cmd 0xef)
[ 130.724268] ata2.00: failed to set xfermode (err_mask=0x4)
[ 130.724271] ata2.00: disabled
[ 130.724274] ata2: failed to recover some devices, retrying in 5 secs
[ 135.725789] ata2.01: failed to set xfermode (err_mask=0x40)
[ 135.725793] ata2: failed to recover some devices, retrying in 5 secs
[ 141.211380] ata2.01: configured for UDMA/33

I should note I have two CD-ROM drives hooked up on ata2 (i.e. secondary IDE channel) but only one is visible in the device manager. it is an " _NEC DVD_RW ND-2510A 2.15 PQ: 0 ANSI: 5".

Bug 96104 might be related, but I'm not sure.

Googling has revealed that there might be patches available that fix this issue for some drives:

http://<email address hidden>/msg03088.html

[jfs (jfschmakeit)]

I have exactly the same problem...

Sounds like Bug 96826 (have a look at some comments further down the page)..
Somebody posted a bugfix there as well, but I haven't had time to check it out so far..

[David Warde-Farley (david-warde-farley)]

As described in Bug 96826 "irqpoll" added to grub's menu.lst works for me.

[beerfan (beerfan)]

I've had this problem since Fiesty release. Bootup takes about 4 minutes. I keep hoping the next kernel update will fix it but no dice. I've tried adding "irqpoll" to the menu.lst boot options but that has no effect for me. Bug 96826 seems to be related to a cdburner but my computer does not have a Lite-on burner, only a regular cdrom.

MB: Asus P4S8X
HD: Western Digital 320GB IDE (on primary IDE interface) (I recently upgraded but had same issue with last HD)
CDR: Aopen CD-952E (on second IDE interface)

In addition to trying the irqpoll boot option, I've tried disabling the cdrom and turning off the "Plug and Play OS" bios option (workarounds reported in other bugs). If there are other workarounds I haven't found them.

I was hoping yesterday's kernel update would be the one to fix this problem but instead it just broke my menu.lst. It completely removed my windows dual-boot menu option, changed the Ubuntu root back to hd0,0 but it's now on hd0,1 since the hard-drive upgrade, and the UUID is back to the pre-upgrade volume ID. It seems like it replaced some backup version of the menu.lst file to what it was after install but how can it do that? Anyway, that's all off-topic but it pisses me off just the same.

If any logs will help just let me know. My dmesg output is quite similar to the original poster's but I've attached it.

[yeti (utu)]

Fedora 7 has solved this problem for my hardware.

Its latest kernel, 2.6.21-1.3228.fc7, no longer needs 'irqpoll' (here, anyway) to get around the long boot-up delay and the lack of access to some ide resources.

This is a little bit ahead of the 2.6.22 'upstream' kernel patch rumored to address this problem.

The irqpoll work-around has been a satisfactory work-around for my ubuntu install. I can't say I've found any downside to it in my uses of linux, which are all pretty tame things like wireless, firefox, openoffice and such.

[DrSlump (sembee-slump)]

I got similar errors, and irqpoll and acpi=force does not resolve them for me.
This is my pc:

Asus P5W DH Deluxe
Intel E6600
2Gb ddr2 ram in dualchannell
Optiarc AD7173A dvd burner on primary intel pata port
Maxtor 320Gb hd on primary sata port, in ahci mode
Seagate 40Gb eide hd on primary pata port of the jmicron controller

i tryed various versions of ubuntu from 7.04 to the 7.10 beta. If i use the 7.10 alternate version, manually partitioning the seagate hd, i can install the system but i get error 21 from grub on reboot.
This are the errors i get (in this case i was get a try with wubi):
http://www.flickr.com/photos/14067309@N03/

[beerfan (beerfan)]

I am mystified as to how the priority of this bug got set to "undecided". A kernel bug like this should have been resolved long ago but I've been suffering with it for more than 5 months now. I just downloaded the Gutsy beta and I'm hoping that it doesn't contain this misfeature.

[Oleksij Rempel (olerem)]

Thank you for taking the time to report this bug and helping to make Ubuntu better. This bug did not have a package associated with it, which is important for ensuring that it gets looked at by the proper developers. You can learn more about finding the right package at [WWW] https://wiki.ubuntu.com/Bugs/FindRightPackage. I have classified this bug as a bug in linux-source-2.6.20.

[Oleksij Rempel (olerem)]

Please include the following additional information, if you have not already done.

1. Please include the output of the command "uname -a" in your next response. It should be one, long line of text which includes the exact kernel version you're running, as well as the CPU architecture.
2. Please run the command "dmesg > dmesg.log" after a fresh boot and attach the resulting file "dmesg.log" to this bug report.
3. Please run the command "sudo lspci -vvnn > lspci-vvnn.log" and attach the resulting file "lspci-vvnn.log" to this bug report.
4. Please run the command "hdparm -i /dev/cdrom" ( or your broken mas storage device ) and post result here.

Regards,
Alex

[Oleksij Rempel (olerem)]

Please test if it was fixed with Ubuntu Gutsy.

[Arthur Penn (arthur-penn)]

I had this issue and can confirm that it went away under Gutsy.

[Chris Coulson (chrisccoulson)]

Marking linux task as Fix Released due to the last comment regarding this being fixed in a later version of Ubuntu.

Thanks

[Leann Ogasawara (leannogasawara)]

The 18 month support period for Feisty Fawn 6.10 has reached it's end of life - http://www.ubuntu.com/news/ubuntu-7.04-end-of-life . As a result, I'm closing the linux-source-2.6.20 kernel task.