From: Keith Busch <kbusch@kernel.org>
To: Tokunori Ikegami <ikegami.t@gmail.com>
Cc: Chaitanya Kulkarni <Chaitanya.Kulkarni@wdc.com>,
"linux-block@vger.kernel.org" <linux-block@vger.kernel.org>,
"linux-nvme@lists.infradead.org" <linux-nvme@lists.infradead.org>
Subject: Re: [PATCH] block, nvme: Increase max segments parameter setting value
Date: Tue, 24 Mar 2020 09:02:37 +0900 [thread overview]
Message-ID: <20200324000237.GB15091@redsun51.ssa.fujisawa.hgst.com> (raw)
In-Reply-To: <cff52955-e55c-068a-44a6-8ed4edc0696f@gmail.com>
On Tue, Mar 24, 2020 at 08:09:19AM +0900, Tokunori Ikegami wrote:
> Hi,
> > The change looks okay, but why do we need such a large data length ?
> >
> > Do you have a use-case or performance numbers ?
>
> We use the large data length to get log page by the NVMe admin command.
> In the past it was able to get with the same length but failed currently
> with it.
>
> So it seems that depended on the kernel version as caused by the version up.
We didn't have 32-bit max segments before, though. Why was 16-bits
enough in older kernels? Which kernel did this stop working?
> Also I have confirmed that currently failed with the length 0x10000000
> 256MB.
If your hitting max segment limits before any other limit, you should be
able to do larger transfers with more physically contiguous memory. Huge
pages can get the same data length in fewer segments, if you want to
try that.
But wouldn't it be better if your application splits the transfer into
smaller chunks across multiple commands? NVMe log page command supports
offsets for this reason.
WARNING: multiple messages have this Message-ID (diff)
From: Keith Busch <kbusch@kernel.org>
To: Tokunori Ikegami <ikegami.t@gmail.com>
Cc: "linux-block@vger.kernel.org" <linux-block@vger.kernel.org>,
"linux-nvme@lists.infradead.org" <linux-nvme@lists.infradead.org>,
Chaitanya Kulkarni <Chaitanya.Kulkarni@wdc.com>
Subject: Re: [PATCH] block, nvme: Increase max segments parameter setting value
Date: Tue, 24 Mar 2020 09:02:37 +0900 [thread overview]
Message-ID: <20200324000237.GB15091@redsun51.ssa.fujisawa.hgst.com> (raw)
In-Reply-To: <cff52955-e55c-068a-44a6-8ed4edc0696f@gmail.com>
On Tue, Mar 24, 2020 at 08:09:19AM +0900, Tokunori Ikegami wrote:
> Hi,
> > The change looks okay, but why do we need such a large data length ?
> >
> > Do you have a use-case or performance numbers ?
>
> We use the large data length to get log page by the NVMe admin command.
> In the past it was able to get with the same length but failed currently
> with it.
>
> So it seems that depended on the kernel version as caused by the version up.
We didn't have 32-bit max segments before, though. Why was 16-bits
enough in older kernels? Which kernel did this stop working?
> Also I have confirmed that currently failed with the length 0x10000000
> 256MB.
If your hitting max segment limits before any other limit, you should be
able to do larger transfers with more physically contiguous memory. Huge
pages can get the same data length in fewer segments, if you want to
try that.
But wouldn't it be better if your application splits the transfer into
smaller chunks across multiple commands? NVMe log page command supports
offsets for this reason.
_______________________________________________
linux-nvme mailing list
linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme
next prev parent reply other threads:[~2020-03-24 0:02 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-03-23 18:23 [PATCH] block, nvme: Increase max segments parameter setting value Tokunori Ikegami
2020-03-23 18:23 ` Tokunori Ikegami
2020-03-23 19:14 ` Chaitanya Kulkarni
2020-03-23 19:14 ` Chaitanya Kulkarni
2020-03-23 23:09 ` Tokunori Ikegami
2020-03-23 23:09 ` Tokunori Ikegami
2020-03-24 0:02 ` Keith Busch [this message]
2020-03-24 0:02 ` Keith Busch
2020-03-24 16:51 ` Tokunori Ikegami
2020-03-24 16:51 ` Tokunori Ikegami
2020-03-27 17:50 ` Tokunori Ikegami
2020-03-27 17:50 ` Tokunori Ikegami
2020-03-27 18:18 ` Keith Busch
2020-03-27 18:18 ` Keith Busch
2020-03-28 2:11 ` Ming Lei
2020-03-28 2:11 ` Ming Lei
2020-03-28 3:13 ` Keith Busch
2020-03-28 3:13 ` Keith Busch
2020-03-28 8:28 ` Ming Lei
2020-03-28 8:28 ` Ming Lei
2020-03-28 12:57 ` Tokunori Ikegami
2020-03-28 12:57 ` Tokunori Ikegami
2020-03-29 3:01 ` Ming Lei
2020-03-29 3:01 ` Ming Lei
2020-03-30 9:15 ` Tokunori Ikegami
2020-03-30 9:15 ` Tokunori Ikegami
2020-03-30 13:53 ` Keith Busch
2020-03-30 13:53 ` Keith Busch
2020-03-31 15:24 ` Tokunori Ikegami
2020-03-31 15:24 ` Tokunori Ikegami
2020-03-31 14:13 ` Joshi
2020-03-31 14:13 ` Joshi
2020-03-31 15:37 ` Tokunori Ikegami
2020-03-31 15:37 ` Tokunori Ikegami
2020-03-24 7:16 ` Hannes Reinecke
2020-03-24 7:16 ` Hannes Reinecke
2020-03-24 17:17 ` Tokunori Ikegami
2020-03-24 17:17 ` Tokunori Ikegami
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=20200324000237.GB15091@redsun51.ssa.fujisawa.hgst.com \
--to=kbusch@kernel.org \
--cc=Chaitanya.Kulkarni@wdc.com \
--cc=ikegami.t@gmail.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-nvme@lists.infradead.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.