From: hch@lst.de (Christoph Hellwig)
Subject: [PATCH v2 1/3] nvme-rdma: don't suppress send completions
Date: Thu, 9 Nov 2017 10:18:58 +0100 [thread overview]
Message-ID: <20171109091858.GA16966@lst.de> (raw)
In-Reply-To: <20171108100616.26605-2-sagi@grimberg.me>
On Wed, Nov 08, 2017@12:06:14PM +0200, Sagi Grimberg wrote:
> The entire completions suppress mechanism is currently
> broken because the HCA might retry a send operation
> (due to dropped ack) after the nvme transaction has completed.
>
> In order to handle this, we signal all send completions (besides
> async event which is not racing anything).
Oh well, so much for all these unsignalled completion optimizations..
So in which cases do unsignalled completions work at all? Seems like
we need to fix up a lot of other ULPs as well.
Looks fine:
Reviewed-by: Christoph Hellwig <hch at lst.de>
> - */
> - if (nvme_rdma_queue_sig_limit(queue) || flush)
> - wr.send_flags |= IB_SEND_SIGNALED;
> + wr.send_flags = IB_SEND_SIGNALED;
But.. Is there any benefit in just setting IB_SIGNAL_ALL_WR on the QP?
WARNING: multiple messages have this Message-ID (diff)
From: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
To: Sagi Grimberg <sagi-NQWnxTmZq1alnMjI0IkVqw@public.gmane.org>
Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-nvme-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org,
Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>,
Max Gurtuvoy <maxg-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Subject: Re: [PATCH v2 1/3] nvme-rdma: don't suppress send completions
Date: Thu, 9 Nov 2017 10:18:58 +0100 [thread overview]
Message-ID: <20171109091858.GA16966@lst.de> (raw)
In-Reply-To: <20171108100616.26605-2-sagi-NQWnxTmZq1alnMjI0IkVqw@public.gmane.org>
On Wed, Nov 08, 2017 at 12:06:14PM +0200, Sagi Grimberg wrote:
> The entire completions suppress mechanism is currently
> broken because the HCA might retry a send operation
> (due to dropped ack) after the nvme transaction has completed.
>
> In order to handle this, we signal all send completions (besides
> async event which is not racing anything).
Oh well, so much for all these unsignalled completion optimizations..
So in which cases do unsignalled completions work at all? Seems like
we need to fix up a lot of other ULPs as well.
Looks fine:
Reviewed-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
> - */
> - if (nvme_rdma_queue_sig_limit(queue) || flush)
> - wr.send_flags |= IB_SEND_SIGNALED;
> + wr.send_flags = IB_SEND_SIGNALED;
But.. Is there any benefit in just setting IB_SIGNAL_ALL_WR on the QP?
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
next prev parent reply other threads:[~2017-11-09 9:18 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-11-08 10:06 [PATCH v2 0/3] Fix request completion holes Sagi Grimberg
2017-11-08 10:06 ` Sagi Grimberg
2017-11-08 10:06 ` [PATCH v2 1/3] nvme-rdma: don't suppress send completions Sagi Grimberg
2017-11-08 10:06 ` Sagi Grimberg
2017-11-09 9:18 ` Christoph Hellwig [this message]
2017-11-09 9:18 ` Christoph Hellwig
2017-11-09 11:08 ` Sagi Grimberg
2017-11-09 11:08 ` Sagi Grimberg
2017-11-20 8:18 ` Christoph Hellwig
2017-11-20 8:18 ` Christoph Hellwig
2017-11-20 8:33 ` Sagi Grimberg
2017-11-20 8:33 ` Sagi Grimberg
2017-11-20 9:32 ` Christoph Hellwig
2017-11-20 9:32 ` Christoph Hellwig
2017-11-08 10:06 ` [PATCH v2 2/3] nvme-rdma: don't complete requests before a send work request has completed Sagi Grimberg
2017-11-08 10:06 ` Sagi Grimberg
2017-11-09 9:21 ` Christoph Hellwig
2017-11-09 9:21 ` Christoph Hellwig
2017-11-09 11:14 ` Sagi Grimberg
2017-11-09 11:14 ` Sagi Grimberg
2017-11-20 8:31 ` Christoph Hellwig
2017-11-20 8:31 ` Christoph Hellwig
2017-11-20 8:37 ` Sagi Grimberg
2017-11-20 8:37 ` Sagi Grimberg
2017-11-20 8:41 ` Christoph Hellwig
2017-11-20 8:41 ` Christoph Hellwig
2017-11-20 9:04 ` Sagi Grimberg
2017-11-20 9:04 ` Sagi Grimberg
2017-11-20 9:28 ` Sagi Grimberg
2017-11-20 9:28 ` Sagi Grimberg
2017-11-20 10:49 ` Christoph Hellwig
2017-11-20 10:49 ` Christoph Hellwig
2017-11-20 11:12 ` Sagi Grimberg
2017-11-20 11:12 ` Sagi Grimberg
2017-11-20 11:16 ` Christoph Hellwig
2017-11-20 11:16 ` Christoph Hellwig
2017-11-08 10:06 ` [PATCH v2 3/3] nvme-rdma: wait for local invalidation before completing a request Sagi Grimberg
2017-11-08 10:06 ` Sagi Grimberg
2017-11-09 9:39 ` Christoph Hellwig
2017-11-09 9:39 ` Christoph Hellwig
2017-11-13 22:10 ` [PATCH v2 0/3] Fix request completion holes Doug Ledford
2017-11-13 22:10 ` Doug Ledford
2017-11-16 15:39 ` Sagi Grimberg
2017-11-16 15:39 ` Sagi Grimberg
2017-11-16 15:58 ` Doug Ledford
2017-11-16 15:58 ` Doug Ledford
2017-11-20 7:37 ` Christoph Hellwig
2017-11-20 7:37 ` Christoph Hellwig
2017-11-20 8:33 ` Sagi Grimberg
2017-11-20 8:33 ` 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=20171109091858.GA16966@lst.de \
--to=hch@lst.de \
/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.