From: Christoph Hellwig <hch@lst.de>
To: Max Gurtovoy <maxg@mellanox.com>
Cc: Christoph Hellwig <hch@lst.de>,
Sagi Grimberg <sagig@mellanox.com>,
linux-rdma@vger.kernel.org, target-devel@vger.kernel.org
Subject: Re: [PATCH 2/4] IB/isert: convert to new CQ API
Date: Wed, 17 Feb 2016 15:18:52 +0100 [thread overview]
Message-ID: <20160217141852.GA15646@lst.de> (raw)
In-Reply-To: <56C364B8.6010807@mellanox.com>
On Tue, Feb 16, 2016 at 08:04:40PM +0200, Max Gurtovoy wrote:
>> @@ -977,7 +959,10 @@ isert_post_recvm(struct isert_conn *isert_conn, u32 count)
>>
>> for (rx_wr = isert_conn->rx_wr, i = 0; i < count; i++, rx_wr++) {
>> rx_desc = &isert_conn->rx_descs[i];
>> - rx_wr->wr_id = (uintptr_t)rx_desc;
>> +
>> + rx_desc->rx_cqe.done = isert_recv_done;
>
> perhaps we can do it once in isert_alloc_rx_descriptors ?
Ok.
>> +static void
>> +isert_recv_done(struct ib_cq *cq, struct ib_wc *wc)
>> {
>> + struct isert_conn *isert_conn = wc->qp->qp_context;
>> + struct ib_device *ib_dev = isert_conn->cm_id->device;
>> + struct iser_rx_desc *rx_desc =
>> + container_of(wc->wr_cqe, struct iser_rx_desc, rx_cqe);
>
> maybe we can define it as:
> static inline struct iser_rx_desc *
> isert_rx(struct ib_cqe *rx_cqe)
> {
> return container_of(rx_cqe, struct iser_rx_desc, rx_cqe);
> }
> in the .h file ?
There is just a single use of this pattern, but if everyone prefers
the helper I can add it.
>> + struct iscsi_hdr *hdr = &rx_desc->iscsi_header;
>> struct iser_ctrl *iser_ctrl = &rx_desc->iser_header;
>> uint64_t read_va = 0, write_va = 0;
>
>
>
>> + if (unlikely(wc->status != IB_WC_SUCCESS)) {
>> + isert_print_wc(wc);
>> + if (!--isert_conn->post_recv_buf_count)
>
> maybe we can do isert_conn->post_recv_buf_count-- once at the beginning of
> the function?
I'd rather not move the assignment around in this patch. If you
think this can sensible be done please send a follow up patch that
also explains why it's safe.
> maybe we can do isert_conn->post_recv_buf_count-- once at the beginning of
> the function?
Same as above.
>> + struct iser_tx_desc *desc =
>> + container_of(wc->wr_cqe, struct iser_tx_desc, tx_cqe);
>
> maybe we can define it as:
> static inline struct iser_tx_desc *
> isert_tx(struct ib_cqe *tx_cqe)
> {
> return container_of(tx_cqe, struct iser_tx_desc, tx_cqe);
> }
We've got a couple more instances of this one, so this might be
worth it.
next prev parent reply other threads:[~2016-02-17 14:18 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-02-15 20:10 isert: convert to the new CQ API and a few random cleanups Christoph Hellwig
2016-02-15 20:10 ` [PATCH 1/4] IB/isert: properly type the login buffer Christoph Hellwig
[not found] ` <1455567060-18381-2-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2016-02-16 4:40 ` Leon Romanovsky
[not found] ` <20160216044058.GA27580-2ukJVAZIZ/Y@public.gmane.org>
2016-02-17 10:29 ` Christoph Hellwig
2016-02-16 17:19 ` Max Gurtovoy
2016-02-17 10:33 ` Christoph Hellwig
2016-02-16 6:47 ` Or Gerlitz
2016-02-15 20:10 ` [PATCH 2/4] IB/isert: convert to new CQ API Christoph Hellwig
2016-02-16 18:04 ` Max Gurtovoy
2016-02-17 14:18 ` Christoph Hellwig [this message]
2016-02-18 12:32 ` Max Gurtovoy
[not found] ` <1455567060-18381-1-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2016-02-15 20:10 ` [PATCH 3/4] IB/isert: kill struct isert_rdma_wr Christoph Hellwig
[not found] ` <1455567060-18381-4-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2016-02-17 9:03 ` Max Gurtovoy
2016-02-15 20:11 ` [PATCH 4/4] IB/isert: the kill ->isert_cmd back pointer in the struct iser_tx_desc Christoph Hellwig
2016-02-17 9:31 ` Max Gurtovoy
2016-02-17 14:19 ` Christoph Hellwig
2016-02-16 10:53 ` isert: convert to the new CQ API and a few random cleanups 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=20160217141852.GA15646@lst.de \
--to=hch@lst.de \
--cc=linux-rdma@vger.kernel.org \
--cc=maxg@mellanox.com \
--cc=sagig@mellanox.com \
--cc=target-devel@vger.kernel.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.