From: Christoph Hellwig <hch@lst.de>
To: Keith Busch <kbusch@kernel.org>
Cc: bijan.mottahedeh@oracle.com, hch@lst.de,
linux-nvme@lists.infradead.org, sagi@grimberg.me
Subject: Re: [PATCH 2/2] nvme-pci: Remove two-pass completions
Date: Wed, 4 Mar 2020 17:08:59 +0100 [thread overview]
Message-ID: <20200304160859.GB11029@lst.de> (raw)
In-Reply-To: <20200302194616.2432-3-kbusch@kernel.org>
> +static inline int nvme_process_cq(struct nvme_queue *nvmeq)
> {
> int found = 0;
>
> while (nvme_cqe_pending(nvmeq)) {
> ++found;
> + nvme_handle_cqe(nvmeq, nvmeq->cq_head);
> nvme_update_cq_head(nvmeq);
> }
>
> + if (found)
> nvme_ring_cq_doorbell(nvmeq);
> return found;
> }
> @@ -1011,21 +1000,16 @@ static irqreturn_t nvme_irq(int irq, void *data)
> {
> struct nvme_queue *nvmeq = data;
> irqreturn_t ret = IRQ_NONE;
> - u16 start, end;
>
> /*
> * The rmb/wmb pair ensures we see all updates from a previous run of
> * the irq handler, even if that was on another CPU.
> */
> rmb();
> - nvme_process_cq(nvmeq, &start, &end);
> + if (nvme_process_cq(nvmeq))
> + ret = IRQ_HANDLED;
> wmb();
>
> - if (start != end) {
> - nvme_complete_cqes(nvmeq, start, end);
> - return IRQ_HANDLED;
> - }
> -
> return ret;
> }
>
> @@ -1044,7 +1028,6 @@ static irqreturn_t nvme_irq_check(int irq, void *data)
> static int nvme_poll_irqdisable(struct nvme_queue *nvmeq)
> {
> struct pci_dev *pdev = to_pci_dev(nvmeq->dev->dev);
> - u16 start, end;
> int found;
>
> /*
> @@ -1054,32 +1037,29 @@ static int nvme_poll_irqdisable(struct nvme_queue *nvmeq)
> */
> if (test_bit(NVMEQ_POLLED, &nvmeq->flags)) {
> spin_lock(&nvmeq->cq_poll_lock);
> - found = nvme_process_cq(nvmeq, &start, &end);
> + found = nvme_process_cq(nvmeq);
> spin_unlock(&nvmeq->cq_poll_lock);
> } else {
> disable_irq(pci_irq_vector(pdev, nvmeq->cq_vector));
> - found = nvme_process_cq(nvmeq, &start, &end);
> + found = nvme_process_cq(nvmeq);
> enable_irq(pci_irq_vector(pdev, nvmeq->cq_vector));
> }
>
> - nvme_complete_cqes(nvmeq, start, end);
> return found;
I think we can just make nvme_timeout call nvme_poll for the
polled case, and just handle the irq driven case in this function
(should probably be another patch on top of this one).
_______________________________________________
linux-nvme mailing list
linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme
next prev parent reply other threads:[~2020-03-04 16:09 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-03-02 19:46 [PATCH 0/2] nvme-pci: Simpler completions Keith Busch
2020-03-02 19:46 ` [PATCH 1/2] nvme-pci: Remove tag from process cq Keith Busch
2020-03-04 16:05 ` Christoph Hellwig
2020-03-04 16:45 ` Keith Busch
2020-03-02 19:46 ` [PATCH 2/2] nvme-pci: Remove two-pass completions Keith Busch
2020-03-04 16:08 ` Christoph Hellwig [this message]
2020-03-05 20:52 ` Sagi Grimberg
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200304160859.GB11029@lst.de \
--to=hch@lst.de \
--cc=bijan.mottahedeh@oracle.com \
--cc=kbusch@kernel.org \
--cc=linux-nvme@lists.infradead.org \
--cc=sagi@grimberg.me \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox