Bugzilla Bug 198041

please read the troubleshooting on the forums
http://forums.gentoo.org/viewtopic-t-604538-highlight-.html

Reproducible: Sometimes

Steps to Reproduce:
1. use a sata dvd
2. build kernel with ahci drivers inside kernel, not module
3. try to reboot 10 times

Actual Results:  
with ahci in kernel the system hangs up occasionaly at freeing kernel memory.
when it passes trough this the dvd works fine. with ahci as Module the dvd does
not work (see dmesg in forums). with nvidia drivers dvd does not work...

Expected Results:  
i expect linux not to hang at "freeing unused kernel memory"

and let me know what else i could check...

(In reply to comment #0)
> please read the troubleshooting on the forums
> http://forums.gentoo.org/viewtopic-t-604538-highlight-.html

Nope, please post all relevant info here, referring to forums and other
third-party site for bug descriptions makes search useless for anyone having
the same issue.

in general i have 2 different issues one when ahci is in kernel and one when
its  as Module.in kernel occasionaly the whole pc hang up at freeing the memory
and in module i get a delay. afterwards the system boots but w/o dvd support.
please check the dmesg of the 'ahci in module' where you can find a "qc
timeout".

the dmesg of the "ahci in kernel" kernel version is allway ok as when the
system sporadicaly hangs i am not able to make a dmesg for you.

finaly when i unplug my dvd i have 100% success in booting in any of 2 upper
mentioned cases.



hardware:
motherboard: MSI K9N Neo-F V3 - nForce560, SATA II
dvd: PLEXTOR PX-810SA SATA (pluged into 1st sata port)

kernel: 2.6.22-gentoo-r8

.config (with ahci in kernel): http://pastebin.com/m4ee2c8a3 

dmesg with 3 kernel configurations: http://pastebin.com/f805a585
1. dmesg: with Nvidia Sata support in kernel (dvd does not work)
2. dmesg: ahci as module (booting is delayed - "qc timeout", dvd does not work)
3. dmesg: with ahci in kernel (boot sporadicaly hangs when freeing memory, dvd
works)



i did try to fix the "ahci as module" issue with irqpoll in grub boot line but
it did not fix the "qc timeout" as stated here:
http://fak3r.com/2007/06/22/failed-to-set-xfermode-solved/

And *please* no pastebins; the links expire in days. Once again, please
post/attach everything relevant to this bug, do not refer to any third-party
links for bug description.

Created an attachment (id=135335) [details]
.config (with ahci in kernel)

Created an attachment (id=135336) [details]
dmesg with 3 kernel configurations

1. dmesg: with Nvidia Sata support in kernel (dvd does not work)
2. dmesg: ahci as module (booting is delayed - "qc timeout", dvd does not work)
3. dmesg: with ahci in kernel (boot sporadicaly hangs when freeing memory, dvd
works)

now i get it ;) you want me to use attachments.

all relevant links from comment #2 are in attachments

the 'qc timeout' issue with AHCI as Module was fixed as described below but
does not work for me:

There’s a known bug in Ubuntu 7.04 (Feisty) with some ata detection routine
that causes the system to take over 2 minutes to boot. Since this has happened
to me more than once I’m documenting it here for me, and for other desperate
souls that may find their way here. If your system is very slow to boot, and
you see error messages in your dmesg (`dmesg | grep ata`) such as this:

[ 34.122465] ata1.00: qc timeout (cmd 0xef)
[ 34.122519] ata1.00: failed to set xfermode (err_mask=0x4)
[ 34.122565] ata1: failed to recover some devices, retrying in 5 secs
[ 46.260055] ata1: port is slow to respond, please be patient (Status 0x90)
[ 69.218482] ata1: port failed to respond (30 secs, Status 0x90)

You just need to ad `irqpoll` to your grub line. So in so in
/boot/grub/menu.lst I added irqpoll to the kernel line:

kernel /boot/vmlinuz-2.6.20-15-generic
root=UUID=48c5a348-eb39-4171-8531-671a49fdb75b ro quiet splash irqpoll

and it fixes the issue. Probably a work around, but since this resets every
time you install a new kernel you’ll realize when it’s broken and when
it’s fixed. Oh, and my system boots in 21 seconds now…is it geeky that I
know that, and I tweaked the system to make it boot faster than the 27 seconds
it was booting in? I guess we’ll never know! ;)

Is this still an issue with later kernels. I see a numerous number of commits
in ahci. Can you test with vanilla-sources-2.6.24_rc5 and post the results.

Hi Mike, i did build a kernel from here
ftp://ftp.eu.kernel.org/pub/linux/kernel/v2.6/testing/linux-2.6.24-rc6.tar.bz2
hope this is vanilla, and will test it next weeks. 

after the kernel was build i had to reboot several times due to some other
issue, so until now i do not confirm the 'freeing mem. issue'. as soon as i get
it during boot ups i will report it.

ps: is the link now ok or what shell i pastebin

Created an attachment (id=139261) [details]
linux-2.6.24-rc6.tar.gz.sign - signature of the kernel i have installed

with the vanilla 2.6.24-rc6 kernel and AHCI as Module the 'qc timeout' issue
disapeared.

i will keep on testing with AHCI in kernel, until now i have more than 20
sucessfull boots and no memory freeing hangup.

Thanks for keeping us updated. It would be interesting to know if this was
fixed in 2.6.23 or in the 2.6.24-rc series. Testing 2.6.23 would clarify that,
although may not be worth the effort given that nobody else seems to have run
into this bug.

no problem, but i will go on vacation now. will return on 30dec2007. you want
to check vanilla again or kernel provided by gentoo?

Either one will do, thanks! enjoy your vacation :)

with vanilla 2.6.23.9 i got the same issues as with the 2.6.22... kernel.
tested with ahci as module and in kernel.

i will stick now with the 2.6.24-rc6 for a while.

and thanks the vacation was short but great, looking forward the next one :D

Created an attachment (id=142437) [details]
dmesg with qc timeout in 2.6.24

now with vanilla 2.6.24 kernel and ahci in kernel:

i could not reproduce the "Booting hangs at "Freeing unused kernel memory""
issue, which is good :)

but i noticed that i get sporadicaly the "qc timeout" folowed by a retry in 5
seconds and the 2nd time was allways a success.. 

so in general i have now none of this bug's issues. just wanted to let you
know.

I am not making any progress on figuring out which post-2.6.23 change fixed
this problem, and given that it only seems to affect 1 person plus we are
hoping for 2.6.24 stable quite soon, I'm closing this as a "fixed in 2.6.24"
bug, with no backports to 2.6.23.

Thanks for all the testing.