From: Bart Van Assche <bvanassche@acm.org>
To: Danil Kipnis <danil.kipnis@cloud.ionos.com>,
Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jack Wang <jinpuwang@gmail.com>,
linux-block@vger.kernel.org, linux-rdma@vger.kernel.org,
Jens Axboe <axboe@kernel.dk>,
Christoph Hellwig <hch@infradead.org>,
Sagi Grimberg <sagi@grimberg.me>,
Leon Romanovsky <leon@kernel.org>,
Doug Ledford <dledford@redhat.com>,
Jack Wang <jinpu.wang@cloud.ionos.com>,
rpenyaev@suse.de, pankaj.gupta@cloud.ionos.com
Subject: Re: [PATCH v9 06/25] RDMA/rtrs: client: main functionality
Date: Tue, 3 Mar 2020 08:04:09 -0800 [thread overview]
Message-ID: <597777e7-ac5a-5fc4-c1f7-3ffa5876a6f2@acm.org> (raw)
In-Reply-To: <CAHg0HuxJgjQca3Vy7nbcVKs0bUoj=-Uqoa5DOzYoqRHQ6Vph7A@mail.gmail.com>
On 3/2/20 5:20 AM, Danil Kipnis wrote:
> On Sun, Mar 1, 2020 at 2:33 AM Bart Van Assche <bvanassche@acm.org> wrote:
>> On 2020-02-21 02:47, Jack Wang wrote:
>>> +static struct rtrs_permit *
>>> +__rtrs_get_permit(struct rtrs_clt *clt, enum rtrs_clt_con_type con_type)
>>> +{
>>> + size_t max_depth = clt->queue_depth;
>>> + struct rtrs_permit *permit;
>>> + int cpu, bit;
>>> +
>>> + /* Combined with cq_vector, we pin the IO to the the cpu it comes */
>>
>> This comment is confusing. Please clarify this comment. All I see below
>> is that preemption is disabled. I don't see pinning of I/O to the CPU of
>> the caller.
> The comment is addressing a use-case of the driver: The user can
> assign (under /proc/irq/) the irqs of the HCA cq_vectors "one-to-one"
> to each cpu. This will "force" the driver to process io response on
> the same cpu the io has been submitted on.
> In the code below only preemption is disabled. This can lead to the
> situation that callers from different cpus will grab the same bit,
> since find_first_zero_bit is not atomic. But then the
> test_and_set_bit_lock will fail for all the callers but one, so that
> they will loop again. This way an explicit spinlock is not required.
> Will extend the comment.
If the purpose of get_cpu() and put_cpu() calls is to serialize code
against other threads, please use locking instead of disabling
preemption. This will help tools that verify locking like lockdep and
the kernel thread sanitizer (https://github.com/google/ktsan/wiki).
>>> +static int rtrs_post_send_rdma(struct rtrs_clt_con *con,
>>> + struct rtrs_clt_io_req *req,
>>> + struct rtrs_rbuf *rbuf, u32 off,
>>> + u32 imm, struct ib_send_wr *wr)
>>> +{
>>> + struct rtrs_clt_sess *sess = to_clt_sess(con->c.sess);
>>> + enum ib_send_flags flags;
>>> + struct ib_sge sge;
>>> +
>>> + if (unlikely(!req->sg_size)) {
>>> + rtrs_wrn(con->c.sess,
>>> + "Doing RDMA Write failed, no data supplied\n");
>>> + return -EINVAL;
>>> + }
>>> +
>>> + /* user data and user message in the first list element */
>>> + sge.addr = req->iu->dma_addr;
>>> + sge.length = req->sg_size;
>>> + sge.lkey = sess->s.dev->ib_pd->local_dma_lkey;
>>> +
>>> + /*
>>> + * From time to time we have to post signalled sends,
>>> + * or send queue will fill up and only QP reset can help.
>>> + */
>>> + flags = atomic_inc_return(&con->io_cnt) % sess->queue_depth ?
>>> + 0 : IB_SEND_SIGNALED;
>>> +
>>> + ib_dma_sync_single_for_device(sess->s.dev->ib_dev, req->iu->dma_addr,
>>> + req->sg_size, DMA_TO_DEVICE);
>>> +
>>> + return rtrs_iu_post_rdma_write_imm(&con->c, req->iu, &sge, 1,
>>> + rbuf->rkey, rbuf->addr + off,
>>> + imm, flags, wr);
>>> +}
>>
>> I don't think that posting a signalled send from time to time is
>> sufficient to prevent send queue overflow. Please address Jason's
>> comment from January 7th: "Not quite. If the SQ depth is 16 and you post
>> 16 things and then signal the last one, you *cannot* post new work until
>> you see the completion. More SQ space *ONLY* becomes available upon
>> receipt of a completion. This is why you can't have an unsignaled SQ."
>
>> See also https://lore.kernel.org/linux-rdma/20200107182528.GB26174@ziepe.ca/
> In our case we set the send queue of each QP belonging to one
> "session" to the one supported by the hardware (max_qp_wr) which is
> around 5K on our hardware. The queue depth of our "session" is 512.
> Those 512 are "shared" by all the QPs (number of CPUs on client side)
> belonging to that session. So we have at most 512 and 512/num_cpus on
> average inflights on each QP. We never experienced send queue full
> event in any of our performance tests or production usage. The
> alternative would be to count submitted requests and completed
> requests, check the difference before submission and wait if the
> difference multiplied by the queue depth of "session" exceeds the max
> supported by the hardware. The check will require quite some code and
> will most probably affect performance. I do not think it is worth it
> to introduce a code path which is triggered only on a condition which
> is known to never become true.
> Jason, do you think it's necessary to implement such tracking?
Please either make sure that send queues do not overflow by providing
enough space for 512 in-flight requests fit or implement tracking for
the number of in-flight requests.
Thanks,
Bart.
next prev parent reply other threads:[~2020-03-03 16:04 UTC|newest]
Thread overview: 89+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-02-21 10:46 [PATCH v9 00/25] RTRS (former IBTRS) RDMA Transport Library and RNBD (former IBNBD) RDMA Network Block Device Jack Wang
2020-02-21 10:46 ` [PATCH v9 01/25] sysfs: export sysfs_remove_file_self() Jack Wang
2020-03-01 0:24 ` Bart Van Assche
2020-03-02 14:37 ` Jinpu Wang
2020-02-21 10:46 ` [PATCH v9 02/25] RDMA/rtrs: public interface header to establish RDMA connections Jack Wang
2020-03-01 0:31 ` Bart Van Assche
2020-03-02 8:39 ` Jinpu Wang
2020-03-03 9:40 ` Leon Romanovsky
2020-03-03 14:05 ` Jinpu Wang
2020-03-03 14:16 ` Leon Romanovsky
2020-03-03 14:23 ` Jinpu Wang
2020-02-21 10:46 ` [PATCH v9 03/25] RDMA/rtrs: private headers with rtrs protocol structs and helpers Jack Wang
2020-03-01 0:37 ` Bart Van Assche
2020-03-02 9:21 ` Jinpu Wang
2020-03-03 9:45 ` Leon Romanovsky
2020-03-03 13:52 ` Jinpu Wang
2020-03-03 14:05 ` Jason Gunthorpe
2020-03-03 16:13 ` Jinpu Wang
2020-02-21 10:47 ` [PATCH v9 04/25] RDMA/rtrs: core: lib functions shared between client and server modules Jack Wang
2020-03-01 0:47 ` Bart Van Assche
2020-03-02 8:40 ` Jinpu Wang
2020-03-03 9:57 ` Leon Romanovsky
2020-03-04 11:21 ` Jinpu Wang
2020-02-21 10:47 ` [PATCH v9 05/25] RDMA/rtrs: client: private header with client structs and functions Jack Wang
2020-03-01 0:51 ` Bart Van Assche
2020-03-02 13:49 ` Jinpu Wang
2020-03-02 16:13 ` Bart Van Assche
2020-03-02 16:18 ` Jinpu Wang
2020-02-21 10:47 ` [PATCH v9 06/25] RDMA/rtrs: client: main functionality Jack Wang
2020-03-01 1:33 ` Bart Van Assche
2020-03-02 13:20 ` Danil Kipnis
2020-03-03 16:04 ` Bart Van Assche [this message]
2020-03-04 16:43 ` Jinpu Wang
2020-03-04 16:49 ` Jason Gunthorpe
2020-03-05 11:26 ` Jinpu Wang
2020-03-05 13:27 ` Jason Gunthorpe
2020-03-05 13:37 ` Jinpu Wang
2020-03-05 13:54 ` Jason Gunthorpe
2020-03-06 13:33 ` Jinpu Wang
2020-02-21 10:47 ` [PATCH v9 07/25] RDMA/rtrs: client: statistics functions Jack Wang
2020-03-03 11:28 ` Leon Romanovsky
2020-03-03 11:46 ` Jinpu Wang
2020-02-21 10:47 ` [PATCH v9 08/25] RDMA/rtrs: client: sysfs interface functions Jack Wang
2020-02-21 10:47 ` [PATCH v9 09/25] RDMA/rtrs: server: private header with server structs and functions Jack Wang
2020-02-21 10:47 ` [PATCH v9 10/25] RDMA/rtrs: server: main functionality Jack Wang
2020-03-01 1:42 ` Bart Van Assche
2020-03-02 14:39 ` Jinpu Wang
2020-03-03 11:37 ` Leon Romanovsky
2020-03-03 16:41 ` Jinpu Wang
2020-03-03 16:59 ` Leon Romanovsky
2020-03-04 11:03 ` Jinpu Wang
2020-03-05 8:00 ` Leon Romanovsky
[not found] ` <CAHg0Huyc=pn1=WSKGLjm+c8AcchyQ8q7JS-0ToQyiBRgpGG=jA@mail.gmail.com>
2020-03-05 12:16 ` Leon Romanovsky
2020-03-05 12:28 ` Jinpu Wang
2020-03-05 12:35 ` Leon Romanovsky
2020-03-05 13:02 ` Jinpu Wang
2020-02-21 10:47 ` [PATCH v9 11/25] RDMA/rtrs: server: statistics functions Jack Wang
2020-02-21 10:47 ` [PATCH v9 12/25] RDMA/rtrs: server: sysfs interface functions Jack Wang
2020-02-21 10:47 ` [PATCH v9 13/25] RDMA/rtrs: include client and server modules into kernel compilation Jack Wang
2020-02-21 10:47 ` [PATCH v9 14/25] RDMA/rtrs: a bit of documentation Jack Wang
2020-02-21 10:47 ` [PATCH v9 15/25] block/rnbd: private headers with rnbd protocol structs and helpers Jack Wang
2020-03-01 2:12 ` Bart Van Assche
2020-03-02 16:37 ` Jinpu Wang
2020-02-21 10:47 ` [PATCH v9 16/25] block/rnbd: client: private header with client structs and functions Jack Wang
2020-03-01 2:26 ` Bart Van Assche
2020-03-02 14:59 ` Jinpu Wang
2020-02-21 10:47 ` [PATCH v9 17/25] block/rnbd: client: main functionality Jack Wang
2020-03-01 2:46 ` Bart Van Assche
2020-03-02 14:58 ` Jinpu Wang
2020-02-21 10:47 ` [PATCH v9 18/25] block/rnbd: client: sysfs interface functions Jack Wang
2020-02-21 10:47 ` [PATCH v9 19/25] block/rnbd: server: private header with server structs and functions Jack Wang
2020-03-01 2:47 ` Bart Van Assche
2020-03-02 10:07 ` Danil Kipnis
2020-02-21 10:47 ` [PATCH v9 20/25] block/rnbd: server: main functionality Jack Wang
2020-03-01 2:58 ` Bart Van Assche
2020-03-02 9:58 ` Danil Kipnis
2020-03-03 5:57 ` Bart Van Assche
2020-02-21 10:47 ` [PATCH v9 21/25] block/rnbd: server: functionality for IO submission to file or block dev Jack Wang
2020-03-01 3:09 ` Bart Van Assche
2020-03-02 10:06 ` Danil Kipnis
2020-03-03 16:20 ` Jinpu Wang
2020-03-03 16:28 ` Bart Van Assche
2020-03-03 16:43 ` Jinpu Wang
2020-02-21 10:47 ` [PATCH v9 22/25] block/rnbd: server: sysfs interface functions Jack Wang
2020-02-21 10:47 ` [PATCH v9 23/25] block/rnbd: include client and server modules into kernel compilation Jack Wang
2020-02-21 10:47 ` [PATCH v9 24/25] block/rnbd: a bit of documentation Jack Wang
2020-02-21 10:47 ` [PATCH v9 25/25] MAINTAINERS: Add maintainers for RNBD/RTRS modules Jack Wang
2020-03-03 9:28 ` [PATCH v9 00/25] RTRS (former IBTRS) RDMA Transport Library and RNBD (former IBNBD) RDMA Network Block Device Leon Romanovsky
2020-03-04 14:06 ` Jinpu Wang
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=597777e7-ac5a-5fc4-c1f7-3ffa5876a6f2@acm.org \
--to=bvanassche@acm.org \
--cc=axboe@kernel.dk \
--cc=danil.kipnis@cloud.ionos.com \
--cc=dledford@redhat.com \
--cc=hch@infradead.org \
--cc=jgg@ziepe.ca \
--cc=jinpu.wang@cloud.ionos.com \
--cc=jinpuwang@gmail.com \
--cc=leon@kernel.org \
--cc=linux-block@vger.kernel.org \
--cc=linux-rdma@vger.kernel.org \
--cc=pankaj.gupta@cloud.ionos.com \
--cc=rpenyaev@suse.de \
--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;
as well as URLs for NNTP newsgroup(s).