From: Damien Le Moal <damien.lemoal@opensource.wdc.com>
To: Gabriel Krisman Bertazi <krisman@collabora.com>,
lsf-pc@lists.linux-foundation.org, linux-block@vger.kernel.org
Subject: Re: [LSF/MM/BPF TOPIC] block drivers in user space
Date: Thu, 24 Feb 2022 07:40:47 +0900 [thread overview]
Message-ID: <3702afe7-2918-42e7-110b-efa75c0b58e8@opensource.wdc.com> (raw)
In-Reply-To: <YhXsQdkOpBY2nmFG@B-P7TQMD6M-0146.local>
On 2/23/22 17:11, Gao Xiang wrote:
> On Wed, Feb 23, 2022 at 04:46:41PM +0900, Damien Le Moal wrote:
>> On 2/23/22 14:57, Gao Xiang wrote:
>>> On Mon, Feb 21, 2022 at 02:59:48PM -0500, Gabriel Krisman Bertazi wrote:
>>>> I'd like to discuss an interface to implement user space block devices,
>>>> while avoiding local network NBD solutions. There has been reiterated
>>>> interest in the topic, both from researchers [1] and from the community,
>>>> including a proposed session in LSFMM2018 [2] (though I don't think it
>>>> happened).
>>>>
>>>> I've been working on top of the Google iblock implementation to find
>>>> something upstreamable and would like to present my design and gather
>>>> feedback on some points, in particular zero-copy and overall user space
>>>> interface.
>>>>
>>>> The design I'm pending towards uses special fds opened by the driver to
>>>> transfer data to/from the block driver, preferably through direct
>>>> splicing as much as possible, to keep data only in kernel space. This
>>>> is because, in my use case, the driver usually only manipulates
>>>> metadata, while data is forwarded directly through the network, or
>>>> similar. It would be neat if we can leverage the existing
>>>> splice/copy_file_range syscalls such that we don't ever need to bring
>>>> disk data to user space, if we can avoid it. I've also experimented
>>>> with regular pipes, But I found no way around keeping a lot of pipes
>>>> opened, one for each possible command 'slot'.
>>>>
>>>> [1] https://dl.acm.org/doi/10.1145/3456727.3463768
>>>> [2] https://www.spinics.net/lists/linux-fsdevel/msg120674.html
>>>
>>> I'm interested in this general topic too. One of our use cases is
>>> that we need to process network data in some degree since many
>>> protocols are application layer protocols so it seems more reasonable
>>> to process such protocols in userspace. And another difference is that
>>> we may have thousands of devices in a machine since we'd better to run
>>> containers as many as possible so the block device solution seems
>>> suboptimal to us. Yet I'm still interested in this topic to get more
>>> ideas.
>>>
>>> Btw, As for general userspace block device solutions, IMHO, there could
>>> be some deadlock issues out of direct reclaim, writeback, and userspace
>>> implementation due to writeback user requests can be tripped back to
>>> the kernel side (even the dependency crosses threads). I think they are
>>> somewhat hard to fix with user block device solutions. For example,
>>> https://lore.kernel.org/r/CAM1OiDPxh0B1sXkyGCSTEpdgDd196-ftzLE-ocnM8Jd2F9w7AA@mail.gmail.com
>>
>> This is already fixed with prctl() support. See:
>>
>> https://lore.kernel.org/linux-fsdevel/20191112001900.9206-1-mchristi@redhat.com/
>
> As I mentioned above, IMHO, we could add some per-task state to avoid
> the majority of such deadlock cases (also what I mentioned above), but
> there may still some potential dependency could happen between threads,
> such as using another kernel workqueue and waiting on it (in principle
> at least) since userspace program can call any syscall in principle (
> which doesn't like in-kernel drivers). So I think it can cause some
> risk due to generic userspace block device restriction, please kindly
> correct me if I'm wrong.
Not sure what you mean with all this. prctl() works per process/thread
and a context that has PR_SET_IO_FLUSHER set will have PF_MEMALLOC_NOIO
set. So for the case of a user block device driver, setting this means
that it cannot reenter itself during a memory allocation, regardless of
the system call it executes (FS etc): all memory allocations in any
syscall executed by the context will have GFP_NOIO.
If the kernel-side driver for the user block device driver does any
allocation that does not have GFP_NOIO, or cause any such allocation
(e.g. within a workqueue it is waiting for), then that is a kernel bug.
Block device drivers are not supposed to ever do a memory allocation in
the IO hot path without GFP_NOIO.
>
> Thanks,
> Gao Xiang
>
>>
>>
>> --
>> Damien Le Moal
>> Western Digital Research
--
Damien Le Moal
Western Digital Research
next prev parent reply other threads:[~2022-02-23 22:40 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-02-21 19:59 [LSF/MM/BPF TOPIC] block drivers in user space Gabriel Krisman Bertazi
2022-02-21 23:16 ` Damien Le Moal
2022-02-21 23:30 ` Gabriel Krisman Bertazi
2022-02-22 6:57 ` Hannes Reinecke
2022-02-22 14:46 ` Sagi Grimberg
2022-02-22 17:46 ` Hannes Reinecke
2022-02-22 18:05 ` Gabriel Krisman Bertazi
2022-02-24 9:37 ` Xiaoguang Wang
2022-02-24 10:12 ` Sagi Grimberg
2022-03-01 23:24 ` Khazhy Kumykov
2022-03-02 16:16 ` Mike Christie
2022-03-13 21:15 ` Sagi Grimberg
2022-03-14 17:12 ` Mike Christie
2022-03-15 8:03 ` Sagi Grimberg
2022-03-14 19:21 ` Bart Van Assche
2022-03-15 6:52 ` Hannes Reinecke
2022-03-15 8:08 ` Sagi Grimberg
2022-03-15 8:12 ` Christoph Hellwig
2022-03-15 8:38 ` Sagi Grimberg
2022-03-15 8:42 ` Christoph Hellwig
2022-03-23 19:42 ` Gabriel Krisman Bertazi
2022-03-24 17:05 ` Sagi Grimberg
2022-03-15 8:04 ` Sagi Grimberg
2022-02-22 18:05 ` Bart Van Assche
2022-03-02 23:04 ` Gabriel Krisman Bertazi
2022-03-03 7:17 ` Hannes Reinecke
2022-03-27 16:35 ` Ming Lei
2022-03-28 5:47 ` Kanchan Joshi
2022-03-28 5:48 ` Hannes Reinecke
2022-03-28 20:20 ` Gabriel Krisman Bertazi
2022-03-29 0:30 ` Ming Lei
2022-03-29 17:20 ` Gabriel Krisman Bertazi
2022-03-30 1:55 ` Ming Lei
2022-03-30 18:22 ` Gabriel Krisman Bertazi
2022-03-31 1:38 ` Ming Lei
2022-03-31 3:49 ` Bart Van Assche
2022-04-08 6:52 ` Xiaoguang Wang
2022-04-08 7:44 ` Ming Lei
2022-02-23 5:57 ` Gao Xiang
2022-02-23 7:46 ` Damien Le Moal
2022-02-23 8:11 ` Gao Xiang
2022-02-23 22:40 ` Damien Le Moal [this message]
2022-02-24 0:58 ` Gao Xiang
2022-06-09 2:01 ` Ming Lei
2022-06-09 2:28 ` Gao Xiang
2022-06-09 4:06 ` Ming Lei
2022-06-09 4:55 ` Gao Xiang
2022-06-10 1:52 ` Ming Lei
2022-07-28 8:23 ` Pavel Machek
2022-03-02 16:52 ` Mike Christie
2022-03-03 7:09 ` Hannes Reinecke
2022-03-14 17:04 ` Mike Christie
2022-03-15 6:45 ` Hannes Reinecke
2022-03-05 7:29 ` Dongsheng Yang
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=3702afe7-2918-42e7-110b-efa75c0b58e8@opensource.wdc.com \
--to=damien.lemoal@opensource.wdc.com \
--cc=krisman@collabora.com \
--cc=linux-block@vger.kernel.org \
--cc=lsf-pc@lists.linux-foundation.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 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).