qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] fedora 7, cdrom and GPCMD_GET_EVENT_STATUS_NOTIFICATIO


From: dmc
Subject: Re: [Qemu-devel] fedora 7, cdrom and GPCMD_GET_EVENT_STATUS_NOTIFICATION
Date: Mon, 30 Jul 2007 00:21:27 -0500
User-agent: Thunderbird 2.0.0.4 (X11/20070615)

Matthew Kent wrote:
Been digging around for a solution or confirmation of this for a while
now without much luck. Forgive me if there's a solution to this I'm just
not seeing.

Getting some ugly libata related errors in my guest like

ata2: DRQ=1 with device error, dev_stat 0x49
ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x2 frozen
ata2.00: cmd a0/00:00:00:00:20/00:00:00:00:00/a0 tag 0 cdb 0x4a data 8
in
         res 41/50:03:00:00:20/00:00:00:00:00/a0 Emask 0x3 (HSM
violation)
ata2: soft resetting port
ata2.00: configured for MWDMA2
ata2: EH complete

when running a stock install of Fedora 7 (x86_64) in qemu (the stock
fedora package of qemu 0.9.0 or kvm-33). Tried upgrading to the latest
updated kernel 2.6.22.1-33.fc7, then the latest development one with no
change.

Tracked it down to hald-addon-storage which is polling the cdrom with
SG_IO and GPCMD_GET_EVENT_STATUS_NOTIFICATION which is trying to "check
if eject button was pressed".
In my search I found a brief example program in this ticket
https://bugzilla.novell.com/show_bug.cgi?id=145147 which replicates the
issue reliably here without involving hal.

The workaround I found was to run

hal-disable-polling --device /dev/scd0

but given the lack of success I had finding other people reporting this
exact issue in libata proper, I'm guessing it might be something with
the cd-rom emulation of which fixing is way over my head :)

Guest dmesg and test case attached.

Anyone else seeing this?


Yeah. I mentioned it on address@hidden I was forwarded to a link which mentioned the workaround you mentioned, as well as prefixing it with a 'killall hald-addon-storage'.

I also ran across this little patch, which I haven't bothered to try yet.

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=431829

If you try it, let me know if it works.

-dmc





reply via email to

[Prev in Thread] Current Thread [Next in Thread]