Pasi Kärkkäinen
2011-Oct-31 18:22 UTC
[Xen-devel] Xen 4.1.1 HVM guest cdrom trouble, lost interrupts, ata failed commands (frozen)
Hello,
I was testing Fedora 16 (rc2) Xen host, with included Xen 4.1.1 rpms
and Linux 3.1.0 dom0 kernel. Fedora 16 PV domUs seem to work nicely.
I noticed a problem with Fedora 16 Xen HVM guests though.
The F16 guest kernel (Linux 3.1.0) fails with the qemu-dm emulated DVD-ROM
drive..
Full HVM guest kernel dmesg attached to this email.
See the end of this email for Xen cfgfile for the domain and some workarounds..
Important parts of the guest dmesg:
[ 0.000000] Linux version 3.1.0-5.fc16.x86_64
(mockbuild@x86-10.phx2.fedoraproject.org) (gcc version 4.6.1 20111003 (Red Hat
4.6.1-10) (GCC) ) #1 SMP Thu Oct 27 03:46:50 UTC 2011
[ 0.000000] Command line: initrd=initrd.img console=ttyS0,38400 console=tty0
BOOT_IMAGE=vmlinuz
<snip>
[ 4.310149] ata2.00: ATAPI: QEMU DVD-ROM, 0.10.2, max UDMA/100
[ 4.325107] ata1.00: ATA-7: QEMU HARDDISK, 0.10.2, max UDMA/100
[ 4.339397] ata1.00: 62914560 sectors, multi 16: LBA48
[ 4.362818] ata1.00: configured for MWDMA2
[ 4.373968] scsi 0:0:0:0: Direct-Access ATA QEMU HARDDISK 0.10
PQ: 0 ANSI: 5
[ 4.495976] ata2.00: configured for MWDMA2
[ 4.540361] sd 0:0:0:0: Attached scsi generic sg0 type 0
[ 4.556311] sd 0:0:0:0: [sda] 62914560 512-byte logical blocks: (32.2 GB/30.0
GiB)
[ 4.575201] sd 0:0:0:0: [sda] Write Protect is off
[ 4.586948] sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled,
doesn''t support DPO or FUA
[ 9.511381] ata2.00: qc timeout (cmd 0xa0)
[ 9.514923] ata2.00: TEST_UNIT_READY failed (err_mask=0x4)
[ 9.675770] ata2.00: configured for MWDMA2
[ 14.680417] ata2.00: qc timeout (cmd 0xa0)
[ 14.685010] ata2.00: TEST_UNIT_READY failed (err_mask=0x4)
[ 14.690798] ata2.00: limiting speed to MWDMA2:PIO3
[ 14.850811] ata2.00: configured for MWDMA2
[ 19.855357] ata2.00: qc timeout (cmd 0xa0)
[ 19.859931] ata2.00: TEST_UNIT_READY failed (err_mask=0x4)
[ 19.865765] ata2.00: disabled
[ 19.869676] ata2: soft resetting link
[ 20.056857] ata2: EH complete
[ 35.712153] ata1: lost interrupt (Status 0x50)
[ 35.722381] ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
[ 35.735437] ata1.00: failed command: READ DMA
[ 35.745912] ata1.00: cmd c8/00:08:00:00:00/00:00:00:00:00/e0 tag 0 dma 4096
in
[ 35.745913] res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4
(timeout)
[ 35.777373] ata1.00: status: { DRDY }
[ 35.785709] ata1: soft resetting link
[ 35.946815] ata1.00: configured for MWDMA2
[ 35.955695] ata1.00: device reported invalid CHS sector 0
[ 35.967512] ata1: EH complete
[ 66.784177] ata1: lost interrupt (Status 0x50)
[ 66.788709] ata1.00: limiting speed to MWDMA1:PIO2
[ 66.793395] ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
[ 66.799857] ata1.00: failed command: READ DMA
[ 66.804193] ata1.00: cmd c8/00:08:00:00:00/00:00:00:00:00/e0 tag 0 dma 4096
in
[ 66.804194] res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4
(timeout)
[ 66.818098] ata1.00: status: { DRDY }
[ 66.822310] ata1: soft resetting link
[ 66.981440] ata1.00: configured for MWDMA1
[ 66.985550] ata1.00: device reported invalid CHS sector 0
[ 66.990704] ata1: EH complete
[ 97.760172] ata1: lost interrupt (Status 0x50)
[ 97.765465] ata1.00: limiting speed to PIO2
[ 97.770093] ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
[ 97.777360] ata1.00: failed command: READ DMA
[ 97.781823] ata1.00: cmd c8/00:08:00:00:00/00:00:00:00:00/e0 tag 0 dma 4096
in
[ 97.781824] res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4
(timeout)
[ 97.795560] ata1.00: status: { DRDY }
[ 97.799269] ata1: soft resetting link
[ 97.957830] ata1.00: configured for PIO2
[ 97.961338] ata1.00: device reported invalid CHS sector 0
[ 97.965874] ata1: EH complete
[ 128.736132] ata1: lost interrupt (Status 0x58)
[ 128.830108] ata1.00: limiting speed to PIO1
[ 128.834274] ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
[ 128.840650] ata1.00: failed command: READ MULTIPLE
[ 128.845289] ata1.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 pio 4096
in
[ 128.845291] res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4
(timeout)
[ 128.859072] ata1.00: status: { DRDY }
[ 128.863253] ata1: soft resetting link
[ 129.022319] ata1.00: configured for PIO1
[ 129.026279] ata1.00: device reported invalid CHS sector 0
[ 129.031398] ata1: EH complete
[ 159.712158] ata1: lost interrupt (Status 0x58)
[ 159.802324] ata1.00: limiting speed to PIO0
[ 159.806496] ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
[ 159.812888] ata1.00: failed command: READ MULTIPLE
[ 159.817503] ata1.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 pio 4096
in
[ 159.817505] res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4
(timeout)
[ 159.831339] ata1.00: status: { DRDY }
[ 159.835482] ata1: soft resetting link
[ 159.994295] ata1.00: configured for PIO0
[ 159.998279] ata1.00: device reported invalid CHS sector 0
[ 160.003401] ata1: EH complete
[ 190.816215] ata1: lost interrupt (Status 0x58)
[ 190.898787] ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
[ 190.904509] ata1.00: failed command: READ MULTIPLE
[ 190.908616] ata1.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 pio 4096
in
[ 190.908617] res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4
(timeout)
[ 190.920911] ata1.00: status: { DRDY }
[ 190.924608] ata1: soft resetting link
[ 191.082728] ata1.00: configured for PIO0
[ 191.086575] ata1.00: device reported invalid CHS sector 0
[ 191.091675] sd 0:0:0:0: [sda] Result: hostbyte=DID_OK
driverbyte=DRIVER_SENSE
[ 191.099098] sd 0:0:0:0: [sda] Sense Key : Aborted Command [current]
[descriptor]
[ 191.108415] Descriptor sense data with sense descriptors (in hex):
[ 191.114713] 72 0b 00 00 00 00 00 0c 00 0a 80 00 00 00 00 00
[ 191.130002] 00 00 00 00
[ 191.136199] sd 0:0:0:0: [sda] Add. Sense: No additional sense information
[ 191.143574] sd 0:0:0:0: [sda] CDB: Read(10): 28 00 00 00 00 00 00 00 08 00
[ 191.157045] end_request: I/O error, dev sda, sector 0
[ 191.161675] Buffer I/O error on device sda, logical block 0
[ 191.166372] ata1: EH complete
[ 221.792133] ata1: lost interrupt (Status 0x58)
[ 221.875310] ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
[ 221.882176] ata1.00: failed command: READ MULTIPLE
[ 221.886834] ata1.00: cmd c4/00:08:00:00:00/00:00:00:00:00/e0 tag 0 pio 4096
in
[ 221.886836] res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4
(timeout)
[ 221.901648] ata1.00: status: { DRDY }
[ 221.906550] ata1: soft resetting link
[ 222.124783] ata1.00: configured for PIO0
[ 222.127744] ata1.00: device reported invalid CHS sector 0
[ 222.132141] ata1: EH complete
/etc/xen/<vm> cfgfile:
kernel = "hvmloader"
builder=''hvm''
device_model = ''qemu-dm''
name = "f16test64hvm"
memory = 1024
vcpus=1
pae=1
acpi=1
apic=1
vif = [ ''type=ioemu, mac=00:16:3e:04:01:14, bridge=virbr0'' ]
disk = [ ''phy:/dev/vg_f16/f16test64hvm,hda,w'',
''file:/root/iso/Fedora-16-Final-RC2-x86_64-DVD.iso,hdc:cdrom,r''
]
boot=''cd''
xen_platform_pci=0
on_poweroff = ''destroy''
on_reboot = ''restart''
on_crash = ''restart''
sdl=0
vnc=1
vncpasswd=''''
stdvga=0
serial=''pty''
tsc_mode=0
usb=1
usbdevice=''tablet''
keymap=''fi''
I''m using "xm create" to start the HVM guest.
I noticed I can workaround the problem with the following methods:
- Specify "xen_platform_pci=1" in the domain cfgfile. The emulated
dvdrom drive works then,
but there are other problems (nic doesn''t work - the mac address is
all empty and the installer fails).
Guest dmesg from this case also attached to this email.
- Upgrade Xen hypervisor to 4.1.2 (rpms available in f16 updates-testing).
Emulated dvdrom works then with the original hvm domain cfgfile above.
guest kernel dmesg for this working method attached aswell.
The following guest kernel options didn''t help with the dvdrom problem:
- libata.dma=0
- pci=nocrs
Is this a known problem? Any proper workarounds for Xen 4.1.1,
or should everyone upgrade to Xen 4.1.2 ?
Info about the Fedora 16 dom0 host:
# uname -a
Linux f16dom0.localdomain 3.1.0-5.fc16.x86_64 #1 SMP Thu Oct 27 03:46:50 UTC
2011 x86_64 x86_64 x86_64 GNU/Linux
# rpm -qa|grep xen
xen-hypervisor-4.1.1-8.fc16.x86_64
xen-libs-4.1.1-8.fc16.x86_64
xen-runtime-4.1.1-8.fc16.x86_64
xen-4.1.1-8.fc16.x86_64
xen-licenses-4.1.1-8.fc16.x86_64
Thanks,
-- Pasi
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel
sandr8
2011-Nov-27 01:53 UTC
Re: Xen 4.1.1 HVM guest cdrom trouble, lost interrupts, ata failed commands (frozen)
Hi, I am seeing exactly the same behavior... did you manage to get around this issue? thank you! -alessandro- -- View this message in context: http://xen.1045712.n5.nabble.com/Xen-4-1-1-HVM-guest-cdrom-trouble-lost-interrupts-ata-failed-commands-frozen-tp4953147p5025774.html Sent from the Xen - Dev mailing list archive at Nabble.com.
Stefano Stabellini
2011-Nov-28 10:19 UTC
Re: Xen 4.1.1 HVM guest cdrom trouble, lost interrupts, ata failed commands (frozen)
On Sun, 27 Nov 2011, sandr8 wrote:> Hi, > > I am seeing exactly the same behavior... did you manage to get around this > issue?it is fixed in xen 4.1.2
Pasi Kärkkäinen
2011-Nov-28 12:44 UTC
Re: Xen 4.1.1 HVM guest cdrom trouble, lost interrupts, ata failed commands (frozen)
On Sat, Nov 26, 2011 at 05:53:34PM -0800, sandr8 wrote:> Hi, > > I am seeing exactly the same behavior... did you manage to get around this > issue? >Yep, as stated in my original email, upgrading to Xen 4.1.2 fixes the problem. I''m not aware of any workarounds if you''re still using Xen 4.1.1. -- Pasi
Pasi Kärkkäinen
2011-Nov-28 12:45 UTC
Re: Xen 4.1.1 HVM guest cdrom trouble, lost interrupts, ata failed commands (frozen)
On Mon, Nov 28, 2011 at 10:19:32AM +0000, Stefano Stabellini wrote:> On Sun, 27 Nov 2011, sandr8 wrote: > > Hi, > > > > I am seeing exactly the same behavior... did you manage to get around this > > issue? > > it is fixed in xen 4.1.2 >Btw are there any known workarounds for <= 4.1.1 ? -- Pasi
Stefano Stabellini
2011-Nov-28 13:50 UTC
Re: Xen 4.1.1 HVM guest cdrom trouble, lost interrupts, ata failed commands (frozen)
On Mon, 28 Nov 2011, Pasi Kärkkäinen wrote:> On Mon, Nov 28, 2011 at 10:19:32AM +0000, Stefano Stabellini wrote: > > On Sun, 27 Nov 2011, sandr8 wrote: > > > Hi, > > > > > > I am seeing exactly the same behavior... did you manage to get around this > > > issue? > > > > it is fixed in xen 4.1.2 > > > > Btw are there any known workarounds for <= 4.1.1 ?you can apply the following change to the guest kernel: diff --git a/arch/x86/pci/xen.c b/arch/x86/pci/xen.c index 8214724..6b57f90 100644 --- a/arch/x86/pci/xen.c +++ b/arch/x86/pci/xen.c @@ -308,8 +308,7 @@ int __init pci_xen_init(void) int __init pci_xen_hvm_init(void) { - if (!xen_feature(XENFEAT_hvm_pirqs)) - return 0; + return 0; #ifdef CONFIG_ACPI /* or you can apply the following change to xen: diff -r e73ada19a69d xen/common/kernel.c --- a/xen/common/kernel.c Thu Nov 17 09:13:25 2011 +0000 +++ b/xen/common/kernel.c Mon Nov 28 10:24:24 2011 +0000 @@ -294,8 +294,7 @@ DO(xen_version)(int cmd, XEN_GUEST_HANDL (1U << XENFEAT_gnttab_map_avail_bits); else fi.submap |= (1U << XENFEAT_hvm_safe_pvclock) | - (1U << XENFEAT_hvm_callback_vector) | - (1U << XENFEAT_hvm_pirqs); + (1U << XENFEAT_hvm_callback_vector); #endif break; default: --8323329-988113412-1322488272=:31179 Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel --8323329-988113412-1322488272=:31179--