All of lore.kernel.org
 help / color / mirror / Atom feed
From: Vladimir Sementsov-Ogievskiy <vladimir.sementsov-ogievskiy@openvz.org>
To: qemu-block@nongnu.org
Cc: kwolf@redhat.com, v.sementsov-og@mail.ru, jsnow@redhat.com,
	qemu-devel@nongnu.org, armbru@redhat.com, hreitz@redhat.com,
	vsementsov@openvz.org, stefanha@redhat.com, eblake@redhat.com
Subject: [PATCH v3 6/7] block/copy-before-write: implement cbw-timeout option
Date: Wed,  6 Apr 2022 21:08:00 +0300	[thread overview]
Message-ID: <20220406180801.374844-7-vsementsov@openvz.org> (raw)
In-Reply-To: <20220406180801.374844-1-vsementsov@openvz.org>

In some scenarios, when copy-before-write operations lasts too long
time, it's better to cancel it.

Most useful would be to use the new option together with
on-cbw-error=break-snapshot: this way if cbw operation takes too long
time we'll just cancel backup process but do not disturb the guest too
much.

Note the tricky point of realization: we keep additional point in
bs->in_fligth during block_copy operation even if it's timed-out.
Background "cancelled" block_copy operations will finish at some point
and will want to access state. We should care to not free the state in
.bdrv_close() earlier.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@openvz.org>
---
 qapi/block-core.json      |  8 +++++++-
 block/copy-before-write.c | 23 ++++++++++++++++++++++-
 2 files changed, 29 insertions(+), 2 deletions(-)

diff --git a/qapi/block-core.json b/qapi/block-core.json
index 085f1666af..e9cd7e88f6 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -4206,12 +4206,18 @@
 # @on-cbw-error: Behavior on failure of copy-before-write operation.
 #                Default is @break-guest-write. (Since 7.0)
 #
+# @cbw-timeout: Zero means no limit. Non-zero sets the timeout in seconds
+#               for copy-before-write operation. When a timeout occurs,
+#               the respective copy-before-write operation will fail, and
+#               the @on-cbw-error parameter will decide how this failure
+#               is handled. Default 0. (Since 7.1)
+#
 # Since: 6.2
 ##
 { 'struct': 'BlockdevOptionsCbw',
   'base': 'BlockdevOptionsGenericFormat',
   'data': { 'target': 'BlockdevRef', '*bitmap': 'BlockDirtyBitmap',
-            '*on-cbw-error': 'OnCbwError' } }
+            '*on-cbw-error': 'OnCbwError', '*cbw-timeout': 'uint32' } }
 
 ##
 # @BlockdevOptions:
diff --git a/block/copy-before-write.c b/block/copy-before-write.c
index b7487e4ffe..13992d28c2 100644
--- a/block/copy-before-write.c
+++ b/block/copy-before-write.c
@@ -42,6 +42,7 @@ typedef struct BDRVCopyBeforeWriteState {
     BlockCopyState *bcs;
     BdrvChild *target;
     OnCbwError on_cbw_error;
+    uint32_t cbw_timeout_ns;
 
     /*
      * @lock: protects access to @access_bitmap, @done_bitmap and
@@ -83,6 +84,14 @@ static coroutine_fn int cbw_co_preadv(
     return bdrv_co_preadv(bs->file, offset, bytes, qiov, flags);
 }
 
+static void block_copy_cb(void *opaque)
+{
+    BlockDriverState *bs = opaque;
+
+    bs->in_flight--;
+    aio_wait_kick();
+}
+
 /*
  * Do copy-before-write operation.
  *
@@ -111,7 +120,16 @@ static coroutine_fn int cbw_do_copy_before_write(BlockDriverState *bs,
     off = QEMU_ALIGN_DOWN(offset, cluster_size);
     end = QEMU_ALIGN_UP(offset + bytes, cluster_size);
 
-    ret = block_copy(s->bcs, off, end - off, true, 0, NULL, NULL);
+    /*
+     * Increase in_flight, so that in case of timed-out block-copy, the
+     * remaining background block_copy() request (which can't be immediately
+     * cancelled by timeout) is presented in bs->in_flight. This way we are
+     * sure that on bs close() we'll previously wait for all timed-out but yet
+     * running block_copy calls.
+     */
+    bs->in_flight++;
+    ret = block_copy(s->bcs, off, end - off, true, s->cbw_timeout_ns,
+                     block_copy_cb, bs);
     if (ret < 0 && s->on_cbw_error == ON_CBW_ERROR_BREAK_GUEST_WRITE) {
         return ret;
     }
@@ -377,6 +395,7 @@ static BlockdevOptions *cbw_parse_options(QDict *options, Error **errp)
      */
     qdict_extract_subqdict(options, NULL, "bitmap");
     qdict_del(options, "on-cbw-error");
+    qdict_del(options, "cbw-timeout");
 
 out:
     visit_free(v);
@@ -422,6 +441,8 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags,
     }
     s->on_cbw_error = opts->has_on_cbw_error ? opts->on_cbw_error :
             ON_CBW_ERROR_BREAK_GUEST_WRITE;
+    s->cbw_timeout_ns = opts->has_cbw_timeout ?
+        opts->cbw_timeout * NANOSECONDS_PER_SECOND : 0;
 
     bs->total_sectors = bs->file->bs->total_sectors;
     bs->supported_write_flags = BDRV_REQ_WRITE_UNCHANGED |
-- 
2.35.1



  parent reply	other threads:[~2022-04-06 18:14 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-04-06 18:07 [PATCH v3 0/7] copy-before-write: on-cbw-error and cbw-timeout Vladimir Sementsov-Ogievskiy
2022-04-06 18:07 ` [PATCH v3 1/7] block/copy-before-write: refactor option parsing Vladimir Sementsov-Ogievskiy
2022-04-07  7:08   ` Hanna Reitz
2022-04-06 18:07 ` [PATCH v3 2/7] block/copy-before-write: add on-cbw-error open parameter Vladimir Sementsov-Ogievskiy
2022-04-07  8:22   ` Hanna Reitz
2022-04-06 18:07 ` [PATCH v3 3/7] iotests: add copy-before-write: on-cbw-error tests Vladimir Sementsov-Ogievskiy
2022-04-06 18:07 ` [PATCH v3 4/7] util: add qemu-co-timeout Vladimir Sementsov-Ogievskiy
2022-04-07  9:20   ` Hanna Reitz
2022-04-06 18:07 ` [PATCH v3 5/7] block/block-copy: block_copy(): add timeout_ns parameter Vladimir Sementsov-Ogievskiy
2022-04-07  9:20   ` Hanna Reitz
2022-04-06 18:08 ` Vladimir Sementsov-Ogievskiy [this message]
2022-04-07  9:20   ` [PATCH v3 6/7] block/copy-before-write: implement cbw-timeout option Hanna Reitz
2022-04-06 18:08 ` [PATCH v3 7/7] iotests: copy-before-write: add cases for " Vladimir Sementsov-Ogievskiy
2022-04-07  9:19   ` Hanna Reitz
2022-04-07 10:54     ` 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=20220406180801.374844-7-vsementsov@openvz.org \
    --to=vladimir.sementsov-ogievskiy@openvz.org \
    --cc=armbru@redhat.com \
    --cc=eblake@redhat.com \
    --cc=hreitz@redhat.com \
    --cc=jsnow@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=stefanha@redhat.com \
    --cc=v.sementsov-og@mail.ru \
    --cc=vsementsov@openvz.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.