qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Hanna Reitz <hreitz@redhat.com>
To: qemu-block@nongnu.org
Cc: Kevin Wolf <kwolf@redhat.com>, Hanna Reitz <hreitz@redhat.com>,
	Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>,
	Eric Blake <eblake@redhat.com>,
	qemu-devel@nongnu.org
Subject: [PATCH v4 01/12] job: Context changes in job_completed_txn_abort()
Date: Tue,  7 Sep 2021 14:42:34 +0200	[thread overview]
Message-ID: <20210907124245.143492-2-hreitz@redhat.com> (raw)
In-Reply-To: <20210907124245.143492-1-hreitz@redhat.com>

Finalizing the job may cause its AioContext to change.  This is noted by
job_exit(), which points at job_txn_apply() to take this fact into
account.

However, job_completed() does not necessarily invoke job_txn_apply()
(through job_completed_txn_success()), but potentially also
job_completed_txn_abort().  The latter stores the context in a local
variable, and so always acquires the same context at its end that it has
released in the beginning -- which may be a different context from the
one that job_exit() releases at its end.  If it is different, qemu
aborts ("qemu_mutex_unlock_impl: Operation not permitted").

Drop the local @outer_ctx variable from job_completed_txn_abort(), and
instead re-acquire the actual job's context at the end of the function,
so job_exit() will release the same.

Signed-off-by: Hanna Reitz <hreitz@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
---
 job.c | 22 +++++++++++++++++-----
 1 file changed, 17 insertions(+), 5 deletions(-)

diff --git a/job.c b/job.c
index e7a5d28854..810e6a2065 100644
--- a/job.c
+++ b/job.c
@@ -737,7 +737,6 @@ static void job_cancel_async(Job *job, bool force)
 
 static void job_completed_txn_abort(Job *job)
 {
-    AioContext *outer_ctx = job->aio_context;
     AioContext *ctx;
     JobTxn *txn = job->txn;
     Job *other_job;
@@ -751,10 +750,14 @@ static void job_completed_txn_abort(Job *job)
     txn->aborting = true;
     job_txn_ref(txn);
 
-    /* We can only hold the single job's AioContext lock while calling
+    /*
+     * We can only hold the single job's AioContext lock while calling
      * job_finalize_single() because the finalization callbacks can involve
-     * calls of AIO_WAIT_WHILE(), which could deadlock otherwise. */
-    aio_context_release(outer_ctx);
+     * calls of AIO_WAIT_WHILE(), which could deadlock otherwise.
+     * Note that the job's AioContext may change when it is finalized.
+     */
+    job_ref(job);
+    aio_context_release(job->aio_context);
 
     /* Other jobs are effectively cancelled by us, set the status for
      * them; this job, however, may or may not be cancelled, depending
@@ -769,6 +772,10 @@ static void job_completed_txn_abort(Job *job)
     }
     while (!QLIST_EMPTY(&txn->jobs)) {
         other_job = QLIST_FIRST(&txn->jobs);
+        /*
+         * The job's AioContext may change, so store it in @ctx so we
+         * release the same context that we have acquired before.
+         */
         ctx = other_job->aio_context;
         aio_context_acquire(ctx);
         if (!job_is_completed(other_job)) {
@@ -779,7 +786,12 @@ static void job_completed_txn_abort(Job *job)
         aio_context_release(ctx);
     }
 
-    aio_context_acquire(outer_ctx);
+    /*
+     * Use job_ref()/job_unref() so we can read the AioContext here
+     * even if the job went away during job_finalize_single().
+     */
+    aio_context_acquire(job->aio_context);
+    job_unref(job);
 
     job_txn_unref(txn);
 }
-- 
2.31.1



  reply	other threads:[~2021-09-07 13:02 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-07 12:42 [PATCH v4 00/12] mirror: Handle errors after READY cancel Hanna Reitz
2021-09-07 12:42 ` Hanna Reitz [this message]
2021-09-07 12:42 ` [PATCH v4 02/12] mirror: Keep s->synced on error Hanna Reitz
2021-09-07 12:42 ` [PATCH v4 03/12] mirror: Drop s->synced Hanna Reitz
2021-09-07 12:42 ` [PATCH v4 04/12] job: Force-cancel jobs in a failed transaction Hanna Reitz
2021-09-07 12:42 ` [PATCH v4 05/12] job: @force parameter for job_cancel_sync() Hanna Reitz
2021-09-08 16:20   ` Eric Blake
2021-09-08 16:33   ` Vladimir Sementsov-Ogievskiy
2021-09-14 17:20     ` Hanna Reitz
2021-09-15  7:07       ` Vladimir Sementsov-Ogievskiy
2021-09-07 12:42 ` [PATCH v4 06/12] jobs: Give Job.force_cancel more meaning Hanna Reitz
2021-09-07 12:42 ` [PATCH v4 07/12] job: Add job_cancel_requested() Hanna Reitz
2021-09-07 12:42 ` [PATCH v4 08/12] mirror: Use job_is_cancelled() Hanna Reitz
2021-09-07 12:42 ` [PATCH v4 09/12] mirror: Check job_is_cancelled() earlier Hanna Reitz
2021-09-07 12:42 ` [PATCH v4 10/12] mirror: Stop active mirroring after force-cancel Hanna Reitz
2021-09-07 12:42 ` [PATCH v4 11/12] mirror: Do not clear .cancelled Hanna Reitz
2021-09-07 12:42 ` [PATCH v4 12/12] iotests: Add mirror-ready-cancel-error test Hanna Reitz
2021-09-15  7:45 ` [PATCH v4 00/12] mirror: Handle errors after READY cancel Vladimir Sementsov-Ogievskiy

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=20210907124245.143492-2-hreitz@redhat.com \
    --to=hreitz@redhat.com \
    --cc=eblake@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=vsementsov@virtuozzo.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).