From: Pavel Begunkov <[email protected]>
To: Jens Axboe <[email protected]>, io-uring <[email protected]>
Subject: Re: [PATCH v2] io_uring/net: ensure async prep handlers always initialize ->done_io
Date: Fri, 15 Mar 2024 23:19:54 +0000 [thread overview]
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>
On 3/15/24 23:13, Pavel Begunkov wrote:
> On 3/15/24 23:09, Pavel Begunkov wrote:
>> On 3/15/24 22:48, Jens Axboe wrote:
>>> If we get a request with IOSQE_ASYNC set, then we first run the prep
>>> async handlers. But if we then fail setting it up and want to post
>>> a CQE with -EINVAL, we use ->done_io. This was previously guarded with
>>> REQ_F_PARTIAL_IO, and the normal setup handlers do set it up before any
>>> potential errors, but we need to cover the async setup too.
>>
>> You can hit io_req_defer_failed() { opdef->fail(); }
>> off of an early submission failure path where def->prep has
>> not yet been called, I don't think the patch will fix the
>> problem.
>>
>> ->fail() handlers are fragile, maybe we should skip them
>> if def->prep() wasn't called. Not even compile tested:
>>
>>
>> diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
>> index 846d67a9c72e..56eed1490571 100644
>> --- a/io_uring/io_uring.c
>> +++ b/io_uring/io_uring.c
[...]
>> def->fail(req);
>> io_req_complete_defer(req);
>> }
>> @@ -2201,8 +2201,7 @@ static int io_init_req(struct io_ring_ctx *ctx, struct io_kiocb *req,
>> }
>> req->flags |= REQ_F_CREDS;
>> }
>> -
>> - return def->prep(req, sqe);
>> + return 0;
>> }
>>
>> static __cold int io_submit_fail_init(const struct io_uring_sqe *sqe,
>> @@ -2250,8 +2249,15 @@ static inline int io_submit_sqe(struct io_ring_ctx *ctx, struct io_kiocb *req,
>> int ret;
>>
>> ret = io_init_req(ctx, req, sqe);
>> - if (unlikely(ret))
>> + if (unlikely(ret)) {
>> +fail:
Obvious the diff is crap, but still bugging me enough to write
that the label should've been one line below, otherwise we'd
flag after ->prep as well.
>> + req->flags |= REQ_F_EARLY_FAIL;
>> return io_submit_fail_init(sqe, req, ret);
>> + }
>> +
>> + ret = def->prep(req, sqe);
>> + if (unlikely(ret))
>> + goto fail;
>>
>> trace_io_uring_submit_req(req);
>>
>
--
Pavel Begunkov
next prev parent reply other threads:[~2024-03-15 23:21 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-03-15 22:48 [PATCH v2] io_uring/net: ensure async prep handlers always initialize ->done_io Jens Axboe
2024-03-15 23:09 ` Pavel Begunkov
2024-03-15 23:13 ` Pavel Begunkov
2024-03-15 23:19 ` Pavel Begunkov [this message]
2024-03-15 23:25 ` Jens Axboe
2024-03-15 23:28 ` Pavel Begunkov
2024-03-15 23:53 ` Jens Axboe
2024-03-16 16:14 ` Jens Axboe
2024-03-16 16:28 ` Pavel Begunkov
2024-03-16 16:31 ` Jens Axboe
2024-03-16 16:32 ` Pavel Begunkov
2024-03-16 16:34 ` Pavel Begunkov
2024-03-16 16:36 ` Jens Axboe
2024-03-16 16:36 ` Pavel Begunkov
2024-03-16 16:40 ` Pavel Begunkov
2024-03-16 16:42 ` Jens Axboe
2024-03-16 16:46 ` Pavel Begunkov
2024-03-16 16:51 ` Jens Axboe
2024-03-16 16:57 ` Pavel Begunkov
2024-03-16 17:01 ` Jens Axboe
2024-03-16 17:42 ` Pavel Begunkov
2024-03-16 23:58 ` Jens Axboe
2024-03-17 20:45 ` Pavel Begunkov
2024-03-15 23:13 ` Jens Axboe
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 \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
/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