Skip to contentSkip to conversation

Bug on barriers for virtio_blk device - end_request: I/O error, dev vda, sector 0

DISCUSSION Massimo Cetra 3 participants 8 messages Oct 14, 2009 – Oct 15, 2009

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

Hello all,

i ended up with some thiusands of errors after upgrading to 2.6.31.2 like this one:

end_request: I/O error, dev vda, sector 0
end_request: I/O error, dev vda, sector 0
end_request: I/O error, dev vda, sector 0
end_request: I/O error, dev vda, sector 0

I am using ocfs2 on top of drbd on top of LVM in a KVM guest that uses virtio_blk.

The bug is already known, see here:
https://bugzilla.redhat.com/show_bug.cgi?id=514901

A patch exists and the problem has already been submitted here in the list. See here: http://lkml.org/lkml/2009/8/6/153

That fix is not in 2.6.31 nor in 2.6.32-RCX

Could you please consider pulling it and adding this fix for inclusion in 2.6.31.$next and also to mainline?

Thanks.

#1 lore raw

Massimo Cetra ha scritto:

Quoted 14 lines

Hello all,

i ended up with some thiusands of errors after upgrading to 2.6.31.2 like this one:

end_request: I/O error, dev vda, sector 0
end_request: I/O error, dev vda, sector 0
end_request: I/O error, dev vda, sector 0
end_request: I/O error, dev vda, sector 0

I am using ocfs2 on top of drbd on top of LVM in a KVM guest that uses virtio_blk.

The bug is already known, see here:
https://bugzilla.redhat.com/show_bug.cgi?id=514901

Quoted 8 lines

A patch exists and the problem has already been submitted here in the list. See here: http://lkml.org/lkml/2009/8/6/153

That fix is not in 2.6.31 nor in 2.6.32-RCX

Could you please consider pulling it and adding this fix for inclusion in 2.6.31.$next and also to mainline?

Thanks.

No one listening here ?
This makes virtio_blk on LVM unusable.

Max

#2 lore raw

Added CCs for the author of and people CCed on the patch.

The patch has been submitted twice (6 Aug, 3 Sept), the first time there was one concern voiced which seems to have been addressed in replies. The second time there were no reactions.

The patch seems to work for Massimo...

Quoted 21 lines · Massimo Cetra

Massimo Cetra ha scritto:

Quoted 20 lines

Hello all,

i ended up with some thiusands of errors after upgrading to 2.6.31.2 like this one:

end_request: I/O error, dev vda, sector 0
end_request: I/O error, dev vda, sector 0
end_request: I/O error, dev vda, sector 0
end_request: I/O error, dev vda, sector 0

I am using ocfs2 on top of drbd on top of LVM in a KVM guest that uses virtio_blk.

The bug is already known, see here:
https://bugzilla.redhat.com/show_bug.cgi?id=514901

A patch exists and the problem has already been submitted here in the list.
See here: http://lkml.org/lkml/2009/8/6/153

Quoted 8 lines
Quoted 6 lines

That fix is not in 2.6.31 nor in 2.6.32-RCX

Could you please consider pulling it and adding this fix for inclusion in 2.6.31.$next and also to mainline?

Thanks.

No one listening here ?

It helps if you CC the people who maintain the relevant code.

Quoted 1 line

This makes virtio_blk on LVM unusable.

Cheers,
FJP

#3 lore raw

Frans Pop ha scritto:

Quoted 5 lines

Added CCs for the author of and people CCed on the patch.

The patch has been submitted twice (6 Aug, 3 Sept), the first time there was one concern voiced which seems to have been addressed in replies. The second time there were no reactions.

The patch seems to work for Massimo...

[CUT]

Quoted 3 lines
Quoted 1 line

No one listening here ?

It helps if you CC the people who maintain the relevant code.

Thanks, Frans.

By the way, where can i find a list of the mantainers of the various areas ? I found it some years ago, but it seems i can't find it anymore. It's probably under my eyes... but i'm blind.

Thanks again
Massimo

#4 lore raw
Quoted 14 lines · Massimo Cetra

Frans Pop ha scritto:

Quoted 8 lines

Added CCs for the author of and people CCed on the patch.

The patch has been submitted twice (6 Aug, 3 Sept), the first time there was one concern voiced which seems to have been addressed in replies. The
second time there were no reactions.

The patch seems to work for Massimo...

[CUT]

Quoted 2 lines
Quoted 1 line

No one listening here ?

Quoted 8 lines
Quoted 1 line

It helps if you CC the people who maintain the relevant code.

Thanks, Frans.

By the way, where can i find a list of the mantainers of the various areas ?
I found it some years ago, but it seems i can't find it anymore. It's probably under my eyes... but i'm blind.

See http://lwn.net/Articles/356890/
"Ask a kernel developer, part 2"
(you'll have to wait a week if you aren't a subscriber).

(or try ./scripts/get_maintainer.pl)

NeilBrown

Quoted 4 lines

Thanks again
Massimo

#5 lore raw
In reply to Massimo Cetra
Quoted 14 lines · Massimo Cetra

Frans Pop ha scritto:

Quoted 8 lines

Added CCs for the author of and people CCed on the patch.

The patch has been submitted twice (6 Aug, 3 Sept), the first time there was one concern voiced which seems to have been addressed in replies. The
second time there were no reactions.

The patch seems to work for Massimo...

[CUT]

Quoted 2 lines
Quoted 1 line

No one listening here ?

Quoted 8 lines
Quoted 1 line

It helps if you CC the people who maintain the relevant code.

Thanks, Frans.

By the way, where can i find a list of the mantainers of the various areas ?
I found it some years ago, but it seems i can't find it anymore. It's probably under my eyes... but i'm blind.

See http://lwn.net/Articles/356890/
"Ask a kernel developer, part 2"
(you'll have to wait a week if you aren't a subscriber).

(or try ./scripts/get_maintainer.pl)

NeilBrown

Quoted 4 lines

Thanks again
Massimo

#6 lore raw
In reply to Frans Pop

Frans Pop ha scritto:

Quoted 7 lines

Added CCs for the author of and people CCed on the patch.

The patch has been submitted twice (6 Aug, 3 Sept), the first time there was one concern voiced which seems to have been addressed in replies. The second time there were no reactions.

The patch seems to work for Massimo...

[CUT]

Quoted 3 lines
Quoted 1 line

No one listening here ?

It helps if you CC the people who maintain the relevant code.

Thanks, Frans.

By the way, where can i find a list of the mantainers of the various areas ? I found it some years ago, but it seems i can't find it anymore. It's probably under my eyes... but i'm blind.

Thanks again
Massimo

#7 lore raw
In reply to Massimo Cetra

Added CCs for the author of and people CCed on the patch.

The patch has been submitted twice (6 Aug, 3 Sept), the first time there was one concern voiced which seems to have been addressed in replies. The second time there were no reactions.

The patch seems to work for Massimo...

Quoted 21 lines · Massimo Cetra

Massimo Cetra ha scritto:

Quoted 20 lines

Hello all,

i ended up with some thiusands of errors after upgrading to 2.6.31.2 like this one:

end_request: I/O error, dev vda, sector 0
end_request: I/O error, dev vda, sector 0
end_request: I/O error, dev vda, sector 0
end_request: I/O error, dev vda, sector 0

I am using ocfs2 on top of drbd on top of LVM in a KVM guest that uses virtio_blk.

The bug is already known, see here:
https://bugzilla.redhat.com/show_bug.cgi?id=514901

A patch exists and the problem has already been submitted here in the list.
See here: http://lkml.org/lkml/2009/8/6/153

Quoted 8 lines
Quoted 6 lines

That fix is not in 2.6.31 nor in 2.6.32-RCX

Could you please consider pulling it and adding this fix for inclusion in 2.6.31.$next and also to mainline?

Thanks.

No one listening here ?

It helps if you CC the people who maintain the relevant code.

Quoted 1 line

This makes virtio_blk on LVM unusable.

Cheers,
FJP