[PATCH] nvme-pci : redundancies in nvme_alloc_queue and nvme_init_queue

[PATCH] nvme-pci : redundancies in nvme_alloc_queue and nvme_init_queue

[Irvin Cote]

cq_head cq_phase and q_db are getting set in nvme_alloc_queue
and in nvme_init_queue, make it so that they are only set once in
nvme_init_queue.

Signed-off-by: Irvin Cote <irvincoteg@gmail.com>
---
 drivers/nvme/host/pci.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index 32244582fdb0..49ae26980a0f 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -1533,9 +1533,6 @@ static int nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth)
 	nvmeq->dev = dev;
 	spin_lock_init(&nvmeq->sq_lock);
 	spin_lock_init(&nvmeq->cq_poll_lock);
-	nvmeq->cq_head = 0;
-	nvmeq->cq_phase = 1;
-	nvmeq->q_db = &dev->dbs[qid * 2 * dev->db_stride];
 	nvmeq->qid = qid;
 	dev->ctrl.queue_count++;
 
-- 
2.39.2

[PATCH] nvme-pci: redundancies in nvme_alloc_queue and nvme_init_queue

[Irvin Cote]

cq_head cq_phase and q_db are getting set in nvme_alloc_queue
and in nvme_init_queue, make it so that they are only set once in
nvme_init_queue.

Signed-off-by: Irvin Cote <irvincoteg@gmail.com>
---
 drivers/nvme/host/pci.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index 32244582fdb0..49ae26980a0f 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -1533,9 +1533,6 @@ static int nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth)
 	nvmeq->dev = dev;
 	spin_lock_init(&nvmeq->sq_lock);
 	spin_lock_init(&nvmeq->cq_poll_lock);
-	nvmeq->cq_head = 0;
-	nvmeq->cq_phase = 1;
-	nvmeq->q_db = &dev->dbs[qid * 2 * dev->db_stride];
 	nvmeq->qid = qid;
 	dev->ctrl.queue_count++;
 
-- 
2.39.2

[PATCH] nvme-pci : redundancies in nvme_alloc_queue and nvme_init_queue

[Irvin Cote]

cq_head cq_phase and q_db are getting set in nvme_alloc_queue
and in nvme_init_queue, make it so that they are only set once in
nvme_init_queue.

Signed-off-by: Irvin Cote <irvincoteg@gmail.com>
---
 drivers/nvme/host/pci.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index 32244582fdb0..49ae26980a0f 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -1533,9 +1533,6 @@ static int nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth)
 	nvmeq->dev = dev;
 	spin_lock_init(&nvmeq->sq_lock);
 	spin_lock_init(&nvmeq->cq_poll_lock);
-	nvmeq->cq_head = 0;
-	nvmeq->cq_phase = 1;
-	nvmeq->q_db = &dev->dbs[qid * 2 * dev->db_stride];
 	nvmeq->qid = qid;
 	dev->ctrl.queue_count++;
 
-- 
2.39.2

Re: [PATCH] nvme-pci : redundancies in nvme_alloc_queue and nvme_init_queue

[irvin cote]

I'm trying to resend the patch but somehow all the new attempts seem
to bind to this thread...

On Sat, 20 May 2023 at 23:46, Irvin Cote <irvincoteg@gmail.com> wrote:
>
> cq_head cq_phase and q_db are getting set in nvme_alloc_queue
> and in nvme_init_queue, make it so that they are only set once in
> nvme_init_queue.
>
> Signed-off-by: Irvin Cote <irvincoteg@gmail.com>
> ---
>  drivers/nvme/host/pci.c | 3 ---
>  1 file changed, 3 deletions(-)
>
> diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
> index 32244582fdb0..49ae26980a0f 100644
> --- a/drivers/nvme/host/pci.c
> +++ b/drivers/nvme/host/pci.c
> @@ -1533,9 +1533,6 @@ static int nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth)
>         nvmeq->dev = dev;
>         spin_lock_init(&nvmeq->sq_lock);
>         spin_lock_init(&nvmeq->cq_poll_lock);
> -       nvmeq->cq_head = 0;
> -       nvmeq->cq_phase = 1;
> -       nvmeq->q_db = &dev->dbs[qid * 2 * dev->db_stride];
>         nvmeq->qid = qid;
>         dev->ctrl.queue_count++;
>
> --
> 2.39.2
>

Re: [PATCH] nvme-pci : redundancies in nvme_alloc_queue and nvme_init_queue

[Keith Busch]

On Sat, May 20, 2023 at 11:47:37PM -0300, irvin cote wrote:
> I'm trying to resend the patch but somehow all the new attempts seem
> to bind to this thread...

It's not really a big deal to me that it's continuing the previous
thread, but how are you sending your patches? If you just use 'git
send-email' without specifying any reply identifier, that should start a
new thread.

Re: [PATCH] nvme-pci : redundancies in nvme_alloc_queue and nvme_init_queue

[irvin cote]

I was just re-executing the same initial git send-mail command.
In the end I just changed the subject and it worked, maybe the subject
was used as a reply identifier

On Thu, 25 May 2023 at 12:36, Keith Busch <kbusch@kernel.org> wrote:
>
> On Sat, May 20, 2023 at 11:47:37PM -0300, irvin cote wrote:
> > I'm trying to resend the patch but somehow all the new attempts seem
> > to bind to this thread...
>
> It's not really a big deal to me that it's continuing the previous
> thread, but how are you sending your patches? If you just use 'git
> send-email' without specifying any reply identifier, that should start a
> new thread.

Re: [PATCH] nvme-pci : redundancies in nvme_alloc_queue and nvme_init_queue

[Keith Busch]

Thanks, applied for nvme-6.5.

[PATCH] nvme-pci : redundancies in nvme_alloc_queue and nvme_init_queue

[Irvin Cote]

Signed-off-by: Irvin Cote <irvincoteg@gmail.com>
---
 drivers/nvme/host/pci.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index 32244582fdb0..49ae26980a0f 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -1533,9 +1533,6 @@ static int nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth)
 	nvmeq->dev = dev;
 	spin_lock_init(&nvmeq->sq_lock);
 	spin_lock_init(&nvmeq->cq_poll_lock);
-	nvmeq->cq_head = 0;
-	nvmeq->cq_phase = 1;
-	nvmeq->q_db = &dev->dbs[qid * 2 * dev->db_stride];
 	nvmeq->qid = qid;
 	dev->ctrl.queue_count++;
 
-- 
2.39.2

Re: [PATCH] nvme-pci : redundancies in nvme_alloc_queue and nvme_init_queue

[Jens Axboe]

On 5/20/23 6:02 PM, Irvin Cote wrote:
> Signed-off-by: Irvin Cote <irvincoteg@gmail.com>
> ---
>  drivers/nvme/host/pci.c | 3 ---
>  1 file changed, 3 deletions(-)
> 
> diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
> index 32244582fdb0..49ae26980a0f 100644
> --- a/drivers/nvme/host/pci.c
> +++ b/drivers/nvme/host/pci.c
> @@ -1533,9 +1533,6 @@ static int nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth)
>  	nvmeq->dev = dev;
>  	spin_lock_init(&nvmeq->sq_lock);
>  	spin_lock_init(&nvmeq->cq_poll_lock);
> -	nvmeq->cq_head = 0;
> -	nvmeq->cq_phase = 1;
> -	nvmeq->q_db = &dev->dbs[qid * 2 * dev->db_stride];
>  	nvmeq->qid = qid;
>  	dev->ctrl.queue_count++;

Please put an actual commit message in here, any commit should have
an explanation of WHY the change is being made. Your subject line
is NOT enough, detail exactly why this is redundant.

From a quick look the patch does look fine, but I'm not going to
ack a patch that doesn't have a good commit message.

-- 
Jens Axboe

Re: [PATCH] nvme-pci : redundancies in nvme_alloc_queue and nvme_init_queue

[irvin cote]

Hi, I am really confused I had written one and apparently it didn't go
through, my apologies.
I will resend it

On Sat, 20 May 2023 at 22:48, Jens Axboe <axboe@kernel.dk> wrote:
>
> On 5/20/23 6:02 PM, Irvin Cote wrote:
> > Signed-off-by: Irvin Cote <irvincoteg@gmail.com>
> > ---
> >  drivers/nvme/host/pci.c | 3 ---
> >  1 file changed, 3 deletions(-)
> >
> > diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
> > index 32244582fdb0..49ae26980a0f 100644
> > --- a/drivers/nvme/host/pci.c
> > +++ b/drivers/nvme/host/pci.c
> > @@ -1533,9 +1533,6 @@ static int nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth)
> >       nvmeq->dev = dev;
> >       spin_lock_init(&nvmeq->sq_lock);
> >       spin_lock_init(&nvmeq->cq_poll_lock);
> > -     nvmeq->cq_head = 0;
> > -     nvmeq->cq_phase = 1;
> > -     nvmeq->q_db = &dev->dbs[qid * 2 * dev->db_stride];
> >       nvmeq->qid = qid;
> >       dev->ctrl.queue_count++;
>
> Please put an actual commit message in here, any commit should have
> an explanation of WHY the change is being made. Your subject line
> is NOT enough, detail exactly why this is redundant.
>
> From a quick look the patch does look fine, but I'm not going to
> ack a patch that doesn't have a good commit message.
>
> --
> Jens Axboe
>
>