[Qemu-devel] [PATCH 0/7] scsi and megasas update

[Qemu-devel] [PATCH 0/7] scsi and megasas update

[Hannes Reinecke]

Hi all,

I've run an I/O exerciser against the megasas emulation and
found quite some issues, both with the HBA and the SCSI
emulation.

- SG_IO returns -EDOM if the internal queue is exhausted;
  we should be mapping this to QUEUE_FULL.
- Codingstyle fixes for megasas as suggested by blueswirl
- We should be using the status codes as defined by SAM,
  not the linux ones (which are shifted by one)
- Each driver has it's own mechanism of storing the
  sense data, so we should be implementing an accessor
  scsi_req_sense() to retrieve it.
- And, of course, a rather largish megasas update

Patches are relative to nab's qemu-kvm git tree at
git://git.kernel.org/pub/scm/virt/kvm/nab/qemu-kvm.git

Nab, I think this patchset should make Windows happy.

Cheers,

Hannes
-- 
Dr. Hannes Reinecke		      zSeries & Storage
hare@suse.de			      +49 911 74053 688
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Markus Rex, HRB 16746 (AG Nürnberg)

Re: [Qemu-devel] [PATCH 0/7] scsi and megasas update

[Alexander Graf]

On 15.06.2010, at 11:15, Hannes Reinecke wrote:

> Hi all,
> 
> I've run an I/O exerciser against the megasas emulation and
> found quite some issues, both with the HBA and the SCSI
> emulation.
> 
> - SG_IO returns -EDOM if the internal queue is exhausted;
>  we should be mapping this to QUEUE_FULL.
> - Codingstyle fixes for megasas as suggested by blueswirl
> - We should be using the status codes as defined by SAM,
>  not the linux ones (which are shifted by one)
> - Each driver has it's own mechanism of storing the
>  sense data, so we should be implementing an accessor
>  scsi_req_sense() to retrieve it.
> - And, of course, a rather largish megasas update
> 
> Patches are relative to nab's qemu-kvm git tree at
> git://git.kernel.org/pub/scm/virt/kvm/nab/qemu-kvm.git
> 
> Nab, I think this patchset should make Windows happy.

Got an updated version? :)
I'd also love to see some performance numbers on this. How well does it perform against virtio?


Alex

Re: [Qemu-devel] [PATCH 0/7] scsi and megasas update

[Hannes Reinecke]

On 11/04/2010 09:49 PM, Alexander Graf wrote:
> 
> On 15.06.2010, at 11:15, Hannes Reinecke wrote:
> 
>> Hi all,
>>
>> I've run an I/O exerciser against the megasas emulation and
>> found quite some issues, both with the HBA and the SCSI
>> emulation.
>>
>> - SG_IO returns -EDOM if the internal queue is exhausted;
>>  we should be mapping this to QUEUE_FULL.
>> - Codingstyle fixes for megasas as suggested by blueswirl
>> - We should be using the status codes as defined by SAM,
>>  not the linux ones (which are shifted by one)
>> - Each driver has it's own mechanism of storing the
>>  sense data, so we should be implementing an accessor
>>  scsi_req_sense() to retrieve it.
>> - And, of course, a rather largish megasas update
>>
>> Patches are relative to nab's qemu-kvm git tree at
>> git://git.kernel.org/pub/scm/virt/kvm/nab/qemu-kvm.git
>>
>> Nab, I think this patchset should make Windows happy.
> 
> Got an updated version? :)

Sadly, no. I've got a git repository setup at kernel.org,
but so far failed to do anything with it.
Real life intercepting as per normal.

But on the good side I have the patchset cleaned up
to avoid the contested SCSI update.
IE the megasas emulation is now using the 'standard'
SCSI callback mechanism.

I see to have the patchset pushed to kernel.org
and will send a pointer then.

Cheers,

Hannes
-- 
Dr. Hannes Reinecke		      zSeries & Storage
hare@suse.de			      +49 911 74053 688
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Markus Rex, HRB 16746 (AG Nürnberg)

Re: [Qemu-devel] [PATCH 0/7] scsi and megasas update

[Stefan Hajnoczi]

On Fri, Nov 5, 2010 at 7:21 AM, Hannes Reinecke <hare@suse.de> wrote:
> On 11/04/2010 09:49 PM, Alexander Graf wrote:
>>
>> Got an updated version? :)
>
> I see to have the patchset pushed to kernel.org
> and will send a pointer then.

I'm also interested in seeing the latest and greatest megasas :).

Stefan