On Tue, Nov 08, 2016 at 08:12:25AM +0100, Hannes Reinecke wrote:
> Use pci_alloc_irq_vectors and drop the hand-crafted
> interrupt affinity routines.
There are a couple more things we can do here. I actually have a patch
in my tree that goes a little further, I'll post it in a bit.
--
To unsubscr
On Tue, Nov 08, 2016 at 04:00:46PM +0100, Hannes Reinecke wrote:
> Right. I'll wait for it.
>
> If you also happen to have patches for megaraid and mpt3sas I'd be very
> much interested in looking into them; I'm currently working on converting
> them, too.
I've looked at the them but haven't sta
On 11/08/2016 04:01 PM, Christoph Hellwig wrote:
On Tue, Nov 08, 2016 at 04:00:46PM +0100, Hannes Reinecke wrote:
Right. I'll wait for it.
If you also happen to have patches for megaraid and mpt3sas I'd be very
much interested in looking into them; I'm currently working on converting
them, too.
On Tue, Nov 08, 2016 at 04:02:54PM +0100, Hannes Reinecke wrote:
>> I've also started playing with lpfc, something I'll need to send to
>> you and James for testing and feedback soon.
>>
> Bah. Another duplicate then.
> Let's see who is faster :-)
lpfc depends on the series adding the post_vetors
On 11/08/2016 03:58 PM, Christoph Hellwig wrote:
On Tue, Nov 08, 2016 at 08:12:25AM +0100, Hannes Reinecke wrote:
Use pci_alloc_irq_vectors and drop the hand-crafted
interrupt affinity routines.
There are a couple more things we can do here. I actually have a patch
in my tree that goes a litt
> -Original Message-
> From: Hannes Reinecke [mailto:h...@suse.de]
> Sent: Tuesday, November 08, 2016 1:12 AM
> To: Martin K. Petersen
> Cc: Christoph Hellwig; James Bottomley; linux-scsi@vger.kernel.org; Hannes
> Reinecke; Hannes Reinecke; Don Brace
> Subject: [PATCH] hpsa: switch to pci_a
On Wed, Nov 09, 2016 at 03:32:48PM +, Don Brace wrote:
> Do we need to add an entry for map_queues?
The existing driver only supports a single submission queue.
If some of the devices support more than one submission queue
we could enhance the driver to support it, but we'd need docs
and hardw
On 11/09/2016 04:36 PM, Christoph Hellwig wrote:
On Wed, Nov 09, 2016 at 03:32:48PM +, Don Brace wrote:
Do we need to add an entry for map_queues?
The existing driver only supports a single submission queue.
If some of the devices support more than one submission queue
we could enhance the
> -Original Message-
> From: Hannes Reinecke [mailto:h...@suse.de]
> Sent: Wednesday, November 09, 2016 9:45 AM
> To: Christoph Hellwig; Don Brace
> Cc: Martin K. Petersen; James Bottomley; linux-scsi@vger.kernel.org; Hannes
> Reinecke
> Subject: Re: [P
On Tue, Nov 08, 2016 at 04:02:54PM +0100, Hannes Reinecke wrote:
>> I've looked at the them but haven't started work, so feel free to go
>> ahead. The MSI-X handling in mpt3sas is pretty nasty, so be careful.
>>
>> I've also started playing with lpfc, something I'll need to send to
>> you and Jame
On 11/09/2016 10:39 PM, Christoph Hellwig wrote:
> On Tue, Nov 08, 2016 at 04:02:54PM +0100, Hannes Reinecke wrote:
>>> I've looked at the them but haven't started work, so feel free to go
>>> ahead. The MSI-X handling in mpt3sas is pretty nasty, so be careful.
>>>
>>> I've also started playing wi
On Thu, Nov 10, 2016 at 07:48:20AM +0100, Hannes Reinecke wrote:
> What I find quite irritating is that we still have to call
> irq_set_affinity_hint(irq, NULL) when freeing up interrupts.
> Can't we roll that into the call to free_irq() ?
If you do call it that's irritation, because you should no
On 11/10/2016 03:48 PM, Christoph Hellwig wrote:
> On Thu, Nov 10, 2016 at 07:48:20AM +0100, Hannes Reinecke wrote:
>> What I find quite irritating is that we still have to call
>> irq_set_affinity_hint(irq, NULL) when freeing up interrupts.
>> Can't we roll that into the call to free_irq() ?
>
>
13 matches
Mail list logo