qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Kevin Wolf <kwolf@redhat.com>
To: qemu-block@nongnu.org
Cc: kwolf@redhat.com, hreitz@redhat.com, eesposit@redhat.com,
	pbonzini@redhat.com, vsementsov@yandex-team.ru,
	qemu-devel@nongnu.org
Subject: [PATCH 09/18] test-bdrv-drain: Fix incorrrect drain assumptions
Date: Wed,  7 Dec 2022 14:18:29 +0100	[thread overview]
Message-ID: <20221207131838.239125-10-kwolf@redhat.com> (raw)
In-Reply-To: <20221207131838.239125-1-kwolf@redhat.com>

The test case assumes that a drain only happens in one specific place
where it drains explicitly. This assumption happened to hold true until
now, but block layer functions may drain interally (any graph
modifications are going to do that through bdrv_graph_wrlock()), so this
is incorrect. Make sure that the test code in .drained_begin only runs
where we actually want it to run.

When scheduling a BH from .drained_begin, we also need to increase the
in_flight counter to make sure that the operation is actually completed
in time before the node that it works on goes away.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 tests/unit/test-bdrv-drain.c | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/tests/unit/test-bdrv-drain.c b/tests/unit/test-bdrv-drain.c
index 2686a8acee..8cedea4959 100644
--- a/tests/unit/test-bdrv-drain.c
+++ b/tests/unit/test-bdrv-drain.c
@@ -1107,6 +1107,7 @@ struct detach_by_parent_data {
     BlockDriverState *c;
     BdrvChild *child_c;
     bool by_parent_cb;
+    bool detach_on_drain;
 };
 static struct detach_by_parent_data detach_by_parent_data;
 
@@ -1114,6 +1115,7 @@ static void detach_indirect_bh(void *opaque)
 {
     struct detach_by_parent_data *data = opaque;
 
+    bdrv_dec_in_flight(data->child_b->bs);
     bdrv_unref_child(data->parent_b, data->child_b);
 
     bdrv_ref(data->c);
@@ -1128,12 +1130,21 @@ static void detach_by_parent_aio_cb(void *opaque, int ret)
 
     g_assert_cmpint(ret, ==, 0);
     if (data->by_parent_cb) {
+        bdrv_inc_in_flight(data->child_b->bs);
         detach_indirect_bh(data);
     }
 }
 
 static void detach_by_driver_cb_drained_begin(BdrvChild *child)
 {
+    struct detach_by_parent_data *data = &detach_by_parent_data;
+
+    if (!data->detach_on_drain) {
+        return;
+    }
+    data->detach_on_drain = false;
+
+    bdrv_inc_in_flight(data->child_b->bs);
     aio_bh_schedule_oneshot(qemu_get_current_aio_context(),
                             detach_indirect_bh, &detach_by_parent_data);
     child_of_bds.drained_begin(child);
@@ -1174,8 +1185,14 @@ static void test_detach_indirect(bool by_parent_cb)
         detach_by_driver_cb_class = child_of_bds;
         detach_by_driver_cb_class.drained_begin =
             detach_by_driver_cb_drained_begin;
+        detach_by_driver_cb_class.drained_end = NULL;
+        detach_by_driver_cb_class.drained_poll = NULL;
     }
 
+    detach_by_parent_data = (struct detach_by_parent_data) {
+        .detach_on_drain = false,
+    };
+
     /* Create all involved nodes */
     parent_a = bdrv_new_open_driver(&bdrv_test, "parent-a", BDRV_O_RDWR,
                                     &error_abort);
@@ -1227,6 +1244,7 @@ static void test_detach_indirect(bool by_parent_cb)
         .child_b = child_b,
         .c = c,
         .by_parent_cb = by_parent_cb,
+        .detach_on_drain = true,
     };
     acb = blk_aio_preadv(blk, 0, &qiov, 0, detach_by_parent_aio_cb, NULL);
     g_assert(acb != NULL);
-- 
2.38.1



  parent reply	other threads:[~2022-12-07 13:26 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-12-07 13:18 [PATCH 00/18] block: Introduce a block graph rwlock Kevin Wolf
2022-12-07 13:18 ` [PATCH 01/18] block: Factor out bdrv_drain_all_begin_nopoll() Kevin Wolf
2022-12-07 13:18 ` [PATCH 02/18] graph-lock: Introduce a lock to protect block graph operations Kevin Wolf
2022-12-07 13:18 ` [PATCH 03/18] graph-lock: Implement guard macros Kevin Wolf
2022-12-07 13:18 ` [PATCH 04/18] async: Register/unregister aiocontext in graph lock list Kevin Wolf
2022-12-07 13:18 ` [PATCH 05/18] Import clang-tsa.h Kevin Wolf
2022-12-07 13:18 ` [PATCH 06/18] clang-tsa: Add TSA_ASSERT() macro Kevin Wolf
2022-12-07 13:18 ` [PATCH 07/18] clang-tsa: Add macros for shared locks Kevin Wolf
2022-12-07 13:18 ` [PATCH 08/18] configure: Enable -Wthread-safety if present Kevin Wolf
2022-12-07 13:18 ` Kevin Wolf [this message]
2022-12-07 13:18 ` [PATCH 10/18] block: Fix locking in external_snapshot_prepare() Kevin Wolf
2022-12-07 13:18 ` [PATCH 11/18] block: wrlock in bdrv_replace_child_noperm Kevin Wolf
2022-12-07 13:18 ` [PATCH 12/18] block: remove unnecessary assert_bdrv_graph_writable() Kevin Wolf
2022-12-07 13:18 ` [PATCH 13/18] block: assert that graph read and writes are performed correctly Kevin Wolf
2022-12-07 13:18 ` [PATCH 14/18] graph-lock: TSA annotations for lock/unlock functions Kevin Wolf
2022-12-07 13:18 ` [PATCH 15/18] Mark assert_bdrv_graph_readable/writable() GRAPH_RD/WRLOCK Kevin Wolf
2022-12-07 13:18 ` [PATCH 16/18] block-coroutine-wrapper.py: introduce annotations that take the graph rdlock Kevin Wolf
2022-12-07 13:18 ` [PATCH 17/18] block: use co_wrapper_mixed_bdrv_rdlock in functions taking the rdlock Kevin Wolf
2022-12-07 13:18 ` [PATCH 18/18] block: GRAPH_RDLOCK for functions only called by co_wrappers Kevin Wolf
2022-12-07 14:12 ` [PATCH 00/18] block: Introduce a block graph rwlock Emanuele Giuseppe Esposito
2022-12-07 16:08   ` Kevin Wolf
2022-12-12 17:14   ` Kevin Wolf

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=20221207131838.239125-10-kwolf@redhat.com \
    --to=kwolf@redhat.com \
    --cc=eesposit@redhat.com \
    --cc=hreitz@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=vsementsov@yandex-team.ru \
    /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).