From patchwork Tue Dec 5 10:41:34 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Stefan Hajnoczi X-Patchwork-Id: 844684 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (mailfrom) smtp.mailfrom=nongnu.org (client-ip=2001:4830:134:3::11; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=) Received: from lists.gnu.org (lists.gnu.org [IPv6:2001:4830:134:3::11]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 3yrdkB1sS3z9tB7 for ; Tue, 5 Dec 2017 21:47:46 +1100 (AEDT) Received: from localhost ([::1]:47787 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1eMAlM-0007je-Cp for incoming@patchwork.ozlabs.org; Tue, 05 Dec 2017 05:47:44 -0500 Received: from eggs.gnu.org ([2001:4830:134:3::10]:54514) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1eMAfp-00027d-VV for qemu-devel@nongnu.org; Tue, 05 Dec 2017 05:42:06 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1eMAfo-0002OW-JV for qemu-devel@nongnu.org; Tue, 05 Dec 2017 05:42:02 -0500 Received: from mx1.redhat.com ([209.132.183.28]:37614) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1eMAfl-0002M8-GA; Tue, 05 Dec 2017 05:41:57 -0500 Received: from smtp.corp.redhat.com (int-mx03.intmail.prod.int.phx2.redhat.com [10.5.11.13]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 984EF883AB; Tue, 5 Dec 2017 10:41:56 +0000 (UTC) Received: from localhost (ovpn-117-172.ams2.redhat.com [10.36.117.172]) by smtp.corp.redhat.com (Postfix) with ESMTP id 23F8B60841; Tue, 5 Dec 2017 10:41:55 +0000 (UTC) From: Stefan Hajnoczi To: Date: Tue, 5 Dec 2017 10:41:34 +0000 Message-Id: <20171205104141.28882-3-stefanha@redhat.com> In-Reply-To: <20171205104141.28882-1-stefanha@redhat.com> References: <20171205104141.28882-1-stefanha@redhat.com> X-Scanned-By: MIMEDefang 2.79 on 10.5.11.13 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.26]); Tue, 05 Dec 2017 10:41:56 +0000 (UTC) X-detected-operating-system: by eggs.gnu.org: GNU/Linux 2.2.x-3.x [generic] [fuzzy] X-Received-From: 209.132.183.28 Subject: [Qemu-devel] [PATCH 2/9] block: don't keep AioContext acquired after external_snapshot_prepare() X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.21 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Kevin Wolf , John Snow , Stefan Hajnoczi , qemu-block@nongnu.org Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: "Qemu-devel" It is not necessary to hold AioContext across transactions anymore since bdrv_drained_begin/end() is used to keep the nodes quiesced. In fact, using the AioContext lock for this purpose was always buggy. This patch reduces the scope of AioContext locked regions. This is not just a cleanup but also fixes hangs that occur in BDRV_POLL_WHILE() because it is unware of recursive locking and does not release the AioContext the necessary number of times to allow progress to be made. Signed-off-by: Stefan Hajnoczi --- blockdev.c | 71 ++++++++++++++++++++++++++++++++++++++++++-------------------- 1 file changed, 48 insertions(+), 23 deletions(-) diff --git a/blockdev.c b/blockdev.c index 3c8d994ced..3b598f8f0e 100644 --- a/blockdev.c +++ b/blockdev.c @@ -1606,7 +1606,6 @@ typedef struct ExternalSnapshotState { BlkActionState common; BlockDriverState *old_bs; BlockDriverState *new_bs; - AioContext *aio_context; bool overlay_appended; } ExternalSnapshotState; @@ -1626,6 +1625,7 @@ static void external_snapshot_prepare(BlkActionState *common, ExternalSnapshotState *state = DO_UPCAST(ExternalSnapshotState, common, common); TransactionAction *action = common->action; + AioContext *aio_context; /* 'blockdev-snapshot' and 'blockdev-snapshot-sync' have similar * purpose but a different set of parameters */ @@ -1662,31 +1662,32 @@ static void external_snapshot_prepare(BlkActionState *common, return; } - /* Acquire AioContext now so any threads operating on old_bs stop */ - state->aio_context = bdrv_get_aio_context(state->old_bs); - aio_context_acquire(state->aio_context); + aio_context = bdrv_get_aio_context(state->old_bs); + aio_context_acquire(aio_context); + + /* Paired with .clean() */ bdrv_drained_begin(state->old_bs); if (!bdrv_is_inserted(state->old_bs)) { error_setg(errp, QERR_DEVICE_HAS_NO_MEDIUM, device); - return; + goto out; } if (bdrv_op_is_blocked(state->old_bs, BLOCK_OP_TYPE_EXTERNAL_SNAPSHOT, errp)) { - return; + goto out; } if (!bdrv_is_read_only(state->old_bs)) { if (bdrv_flush(state->old_bs)) { error_setg(errp, QERR_IO_ERROR); - return; + goto out; } } if (!bdrv_is_first_non_filter(state->old_bs)) { error_setg(errp, QERR_FEATURE_DISABLED, "snapshot"); - return; + goto out; } if (action->type == TRANSACTION_ACTION_KIND_BLOCKDEV_SNAPSHOT_SYNC) { @@ -1698,13 +1699,13 @@ static void external_snapshot_prepare(BlkActionState *common, if (node_name && !snapshot_node_name) { error_setg(errp, "New snapshot node name missing"); - return; + goto out; } if (snapshot_node_name && bdrv_lookup_bs(snapshot_node_name, snapshot_node_name, NULL)) { error_setg(errp, "New snapshot node name already in use"); - return; + goto out; } flags = state->old_bs->open_flags; @@ -1717,7 +1718,7 @@ static void external_snapshot_prepare(BlkActionState *common, int64_t size = bdrv_getlength(state->old_bs); if (size < 0) { error_setg_errno(errp, -size, "bdrv_getlength failed"); - return; + goto out; } bdrv_img_create(new_image_file, format, state->old_bs->filename, @@ -1725,7 +1726,7 @@ static void external_snapshot_prepare(BlkActionState *common, NULL, size, flags, false, &local_err); if (local_err) { error_propagate(errp, local_err); - return; + goto out; } } @@ -1740,30 +1741,30 @@ static void external_snapshot_prepare(BlkActionState *common, errp); /* We will manually add the backing_hd field to the bs later */ if (!state->new_bs) { - return; + goto out; } if (bdrv_has_blk(state->new_bs)) { error_setg(errp, "The snapshot is already in use"); - return; + goto out; } if (bdrv_op_is_blocked(state->new_bs, BLOCK_OP_TYPE_EXTERNAL_SNAPSHOT, errp)) { - return; + goto out; } if (state->new_bs->backing != NULL) { error_setg(errp, "The snapshot already has a backing image"); - return; + goto out; } if (!state->new_bs->drv->supports_backing) { error_setg(errp, "The snapshot does not support backing images"); - return; + goto out; } - bdrv_set_aio_context(state->new_bs, state->aio_context); + bdrv_set_aio_context(state->new_bs, aio_context); /* This removes our old bs and adds the new bs. This is an operation that * can fail, so we need to do it in .prepare; undoing it for abort is @@ -1772,15 +1773,22 @@ static void external_snapshot_prepare(BlkActionState *common, bdrv_append(state->new_bs, state->old_bs, &local_err); if (local_err) { error_propagate(errp, local_err); - return; + goto out; } state->overlay_appended = true; + +out: + aio_context_release(aio_context); } static void external_snapshot_commit(BlkActionState *common) { ExternalSnapshotState *state = DO_UPCAST(ExternalSnapshotState, common, common); + AioContext *aio_context; + + aio_context = bdrv_get_aio_context(state->old_bs); + aio_context_acquire(aio_context); /* We don't need (or want) to use the transactional * bdrv_reopen_multiple() across all the entries at once, because we @@ -1789,6 +1797,8 @@ static void external_snapshot_commit(BlkActionState *common) bdrv_reopen(state->old_bs, state->old_bs->open_flags & ~BDRV_O_RDWR, NULL); } + + aio_context_release(aio_context); } static void external_snapshot_abort(BlkActionState *common) @@ -1797,11 +1807,18 @@ static void external_snapshot_abort(BlkActionState *common) DO_UPCAST(ExternalSnapshotState, common, common); if (state->new_bs) { if (state->overlay_appended) { + AioContext *aio_context; + + aio_context = bdrv_get_aio_context(state->old_bs); + aio_context_acquire(aio_context); + bdrv_ref(state->old_bs); /* we can't let bdrv_set_backind_hd() close state->old_bs; we need it */ bdrv_set_backing_hd(state->new_bs, NULL, &error_abort); bdrv_replace_node(state->new_bs, state->old_bs, &error_abort); bdrv_unref(state->old_bs); /* bdrv_replace_node() ref'ed old_bs */ + + aio_context_release(aio_context); } } } @@ -1810,11 +1827,19 @@ static void external_snapshot_clean(BlkActionState *common) { ExternalSnapshotState *state = DO_UPCAST(ExternalSnapshotState, common, common); - if (state->aio_context) { - bdrv_drained_end(state->old_bs); - bdrv_unref(state->new_bs); - aio_context_release(state->aio_context); + AioContext *aio_context; + + if (!state->old_bs) { + return; } + + aio_context = bdrv_get_aio_context(state->old_bs); + aio_context_acquire(aio_context); + + bdrv_drained_end(state->old_bs); + bdrv_unref(state->new_bs); + + aio_context_release(aio_context); } typedef struct DriveBackupState {