From: Yu Kuai <yukuai3@huawei.com>
To: <axboe@kernel.dk>, <josef@toxicpanda.com>, <ming.lei@redhat.com>,
<bvanassche@acm.org>
Cc: <linux-block@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
<nbd@other.debian.org>, <yukuai3@huawei.com>,
<yi.zhang@huawei.com>
Subject: [PATCH v3 4/5] nbd: make sure request completion won't concurrent
Date: Tue, 24 Aug 2021 22:12:26 +0800 [thread overview]
Message-ID: <20210824141227.808340-5-yukuai3@huawei.com> (raw)
In-Reply-To: <20210824141227.808340-1-yukuai3@huawei.com>
commit cddce0116058 ("nbd: Aovid double completion of a request")
try to fix that nbd_clear_que() and recv_work() can complete a
request concurrently. However, the problem still exists:
t1 t2 t3
nbd_disconnect_and_put
flush_workqueue
recv_work
blk_mq_complete_request
blk_mq_complete_request_remote -> this is true
WRITE_ONCE(rq->state, MQ_RQ_COMPLETE)
blk_mq_raise_softirq
blk_done_softirq
blk_complete_reqs
nbd_complete_rq
blk_mq_end_request
blk_mq_free_request
WRITE_ONCE(rq->state, MQ_RQ_IDLE)
nbd_clear_que
blk_mq_tagset_busy_iter
nbd_clear_req
__blk_mq_free_request
blk_mq_put_tag
blk_mq_complete_request
There are three places where request can be completed in nbd:
recv_work(), nbd_clear_que() and nbd_xmit_timeout(). Since they
all hold cmd->lock before completing the request, it's easy to
avoid the problem by setting and checking a cmd flag.
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
---
drivers/block/nbd.c | 12 ++++++++++--
1 file changed, 10 insertions(+), 2 deletions(-)
diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index 7b9e19675224..4d5098d01758 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -416,12 +416,15 @@ static enum blk_eh_timer_return nbd_xmit_timeout(struct request *req,
struct nbd_cmd *cmd = blk_mq_rq_to_pdu(req);
struct nbd_device *nbd = cmd->nbd;
struct nbd_config *config;
+ bool need_complete;
if (!mutex_trylock(&cmd->lock))
return BLK_EH_RESET_TIMER;
if (!refcount_inc_not_zero(&nbd->config_refs)) {
cmd->status = BLK_STS_TIMEOUT;
+ need_complete =
+ test_and_clear_bit(NBD_CMD_INFLIGHT, &cmd->flags);
mutex_unlock(&cmd->lock);
goto done;
}
@@ -490,11 +493,13 @@ static enum blk_eh_timer_return nbd_xmit_timeout(struct request *req,
dev_err_ratelimited(nbd_to_dev(nbd), "Connection timed out\n");
set_bit(NBD_RT_TIMEDOUT, &config->runtime_flags);
cmd->status = BLK_STS_IOERR;
+ need_complete = test_and_clear_bit(NBD_CMD_INFLIGHT, &cmd->flags);
mutex_unlock(&cmd->lock);
sock_shutdown(nbd);
nbd_config_put(nbd);
done:
- blk_mq_complete_request(req);
+ if (need_complete)
+ blk_mq_complete_request(req);
return BLK_EH_DONE;
}
@@ -849,6 +854,7 @@ static void recv_work(struct work_struct *work)
static bool nbd_clear_req(struct request *req, void *data, bool reserved)
{
struct nbd_cmd *cmd = blk_mq_rq_to_pdu(req);
+ bool need_complete;
/* don't abort one completed request */
if (blk_mq_request_completed(req))
@@ -856,9 +862,11 @@ static bool nbd_clear_req(struct request *req, void *data, bool reserved)
mutex_lock(&cmd->lock);
cmd->status = BLK_STS_IOERR;
+ need_complete = test_and_clear_bit(NBD_CMD_INFLIGHT, &cmd->flags);
mutex_unlock(&cmd->lock);
- blk_mq_complete_request(req);
+ if (need_complete)
+ blk_mq_complete_request(req);
return true;
}
--
2.31.1
next prev parent reply other threads:[~2021-08-24 14:02 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-08-24 14:12 [PATCH v3 0/5] fix request uaf in nbd_read_stat() Yu Kuai
2021-08-24 14:12 ` [PATCH v3 1/5] blk-mq: add a new interface to get request by tag Yu Kuai
2021-08-24 14:12 ` [PATCH v3 2/5] nbd: convert to use blk_mq_get_rq_by_tag() Yu Kuai
2021-08-24 14:12 ` [PATCH v3 3/5] nbd: don't handle response without a corresponding request message Yu Kuai
2021-08-24 14:12 ` Yu Kuai [this message]
2021-08-24 14:12 ` [PATCH v3 5/5] nbd: don't start request if nbd_queue_rq() failed Yu Kuai
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=20210824141227.808340-5-yukuai3@huawei.com \
--to=yukuai3@huawei.com \
--cc=axboe@kernel.dk \
--cc=bvanassche@acm.org \
--cc=josef@toxicpanda.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=ming.lei@redhat.com \
--cc=nbd@other.debian.org \
--cc=yi.zhang@huawei.com \
/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).