Re: [PATCH v1.1 2/16 update] arcmsr: Adding code to support MSI-X interrupt

2014-05-06 Thread ching
Thanks for Tomas' suggestion. I will have a new update for this patch. On Mon, 2014-05-05 at 14:53 +0200, Tomas Henzl wrote: > On 05/05/2014 01:15 PM, ching wrote: > > From: Ching > > > > Adding code for supporting MSI-X interrupt. > > > > This update is for fixing bug of previous [PATCH v1.1 2/16

[PATCH v1.1 2/16 update 2] arcmsr: Adding code to support MSI-X interrupt

2014-05-06 Thread ching
From: Ching Adding code to support MSI-X interrupt. This update has modification by Tomas' suggestion. And I add a msix_vector_count for free_irq. Signed-off-by: Ching --- diff -uprN a/drivers/scsi/arcmsr/arcmsr.h b/drivers/scsi/arcmsr/arcmsr.h --- a/drivers/scsi/arcmsr/arcmsr.h 2014-04-28

virtio-scsi: two questions related with picking up queue

2014-05-06 Thread Ming Lei
Hi Paolo and All, One question is about ACCESS_ONCE() in virtscsi_pick_vq(), looks it needn't since both reading and writing tgt->req_vq holds tgt->tgt_lock. Another one is about the comment in virtscsi_req_done(), which said smp_read_barrier_depends() is needed for avoiding out of order between

RE: How to get more sequential IO merged at elevator

2014-05-06 Thread Desai, Kashyap
I got some clue on what was going on while doing 4K sequential read using fio. If I use ioengine in fio script as "libaio", I see " do_io_submit" call plug/unplug the queue before submitting the IO. It means after every IO, we expect to send IO immediately to the next layer. If at all there are

Re: [PATCH v1.1 2/16 update 2] arcmsr: Adding code to support MSI-X interrupt

2014-05-06 Thread Tomas Henzl
On 05/06/2014 10:17 AM, ching wrote: > From: Ching > > Adding code to support MSI-X interrupt. > > This update has modification by Tomas' suggestion. > And I add a msix_vector_count for free_irq. > > Signed-off-by: Ching > --- > > diff -uprN a/drivers/scsi/arcmsr/arcmsr.h b/drivers/scsi/arcmsr/arcm

Re: [PATCH v1.1 2/16 update 2] arcmsr: Adding code to support MSI-X interrupt

2014-05-06 Thread Alexander Gordeev
On Tue, May 06, 2014 at 02:54:45PM +0200, Tomas Henzl wrote: > On 05/06/2014 10:17 AM, ching wrote: > > From: Ching > > > > Adding code to support MSI-X interrupt. > > > > This update has modification by Tomas' suggestion. > > And I add a msix_vector_count for free_irq. > > > > Signed-off-by: Ching

Re: virtio-scsi: two questions related with picking up queue

2014-05-06 Thread Paolo Bonzini
Il 06/05/2014 11:26, Ming Lei ha scritto: Hi Paolo and All, One question is about ACCESS_ONCE() in virtscsi_pick_vq(), looks it needn't since both reading and writing tgt->req_vq holds tgt->tgt_lock. You're right. It should be possible to avoid the lock in virtscsi_pick_vq like this:

Re: How to get more sequential IO merged at elevator

2014-05-06 Thread Jens Axboe
On 05/06/2014 04:06 AM, Desai, Kashyap wrote: > I got some clue on what was going on while doing 4K sequential read using fio. > > If I use ioengine in fio script as "libaio", I see " do_io_submit" call > plug/unplug the queue before submitting the IO. > It means after every IO, we expect to send

Re: [PATCH] MAINTAINERS: add an entry for all the NCR5380 drivers

2014-05-06 Thread Michael Schmitz
Acked-by: Michael Schmitz On Mon, May 5, 2014 at 5:35 PM, Finn Thain wrote: > > Signed-off-by: Finn Thain > Cc: Michael Schmitz > > --- > > As requested: > http://marc.info/?l=linux-arm-kernel&m=139853302724112&w=2 > > diff --git a/MAINTAINERS b/MAINTAINERS > index e67ea24..60ea600 100644 > --

Re: virtio-scsi: two questions related with picking up queue

2014-05-06 Thread Ming Lei
Hi Paolo, On Tue, May 6, 2014 at 9:17 PM, Paolo Bonzini wrote: > Il 06/05/2014 11:26, Ming Lei ha scritto: > >> Hi Paolo and All, >> >> One question is about ACCESS_ONCE() in virtscsi_pick_vq(), >> looks it needn't since both reading and writing tgt->req_vq holds >> tgt->tgt_lock. > > > You're ri