block: ignore flush requests when storage is clean

Some guests (win2008 server for example) do a lot of unnecessary
flushing when underlying media has not changed. This adds additional
overhead on host when calling fsync/fdatasync.

This change introduces a write generation scheme in BlockDriverState.
Current write generation is checked against last flushed generation to
avoid unnessesary flushes.

The problem with excessive flushing was found by a performance test
which does parallel directory tree creation (from 2 processes).
Results improved from 0.424 loops/sec to 0.432 loops/sec.
Each loop creates 10^3 directories with 10 files in each.

This affected some blkdebug testcases that were expecting error logs from
failure-injected flushes which are now skipped entirely
(tests 026 071 089).

This also affects the performance of block jobs and thus BLOCK_JOB_READY
events for driver-mirror and active block-commit commands now arrives
faster, before QMP send successfully returns to caller (tests 141 144).

Signed-off-by: Evgeny Yakovlev <eyakovlev@virtuozzo.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1468870792-7411-5-git-send-email-den@openvz.org
CC: Kevin Wolf <kwolf@redhat.com>
CC: Max Reitz <mreitz@redhat.com>
CC: Stefan Hajnoczi <stefanha@redhat.com>
CC: Fam Zheng <famz@redhat.com>
CC: John Snow <jsnow@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>

1 files changed, 21 insertions, 0 deletions

diff --git a/block/io.c b/block/io.c
index 2887394..cfda714 100644
--- a/block/io.c
+++ b/block/io.c
@@ -1303,6 +1303,7 @@ static int coroutine_fn bdrv_aligned_pwritev(BlockDriverState *bs,
     }
     bdrv_debug_event(bs, BLKDBG_PWRITEV_DONE);
 
+    ++bs->write_gen;
     bdrv_set_dirty(bs, start_sector, end_sector - start_sector);
 
     if (bs->wr_highest_offset < offset + bytes) {
@@ -2236,6 +2237,15 @@ int coroutine_fn bdrv_co_flush(BlockDriverState *bs)
 
     tracked_request_begin(&req, bs, 0, 0, BDRV_TRACKED_FLUSH);
 
+    int current_gen = bs->write_gen;
+
+    /* Wait until any previous flushes are completed */
+    while (bs->flush_started_gen != bs->flushed_gen) {
+        qemu_co_queue_wait(&bs->flush_queue);
+    }
+
+    bs->flush_started_gen = current_gen;
+
     /* Write back all layers by calling one driver function */
     if (bs->drv->bdrv_co_flush) {
         ret = bs->drv->bdrv_co_flush(bs);
@@ -2256,6 +2266,11 @@ int coroutine_fn bdrv_co_flush(BlockDriverState *bs)
         goto flush_parent;
     }
 
+    /* Check if we really need to flush anything */
+    if (bs->flushed_gen == current_gen) {
+        goto flush_parent;
+    }
+
     BLKDBG_EVENT(bs->file, BLKDBG_FLUSH_TO_DISK);
     if (bs->drv->bdrv_co_flush_to_disk) {
         ret = bs->drv->bdrv_co_flush_to_disk(bs);
@@ -2286,6 +2301,7 @@ int coroutine_fn bdrv_co_flush(BlockDriverState *bs)
          */
         ret = 0;
     }
+
     if (ret < 0) {
         goto out;
     }
@@ -2296,6 +2312,10 @@ int coroutine_fn bdrv_co_flush(BlockDriverState *bs)
 flush_parent:
     ret = bs->file ? bdrv_co_flush(bs->file->bs) : 0;
 out:
+    /* Notify any pending flushes that we have completed */
+    bs->flushed_gen = current_gen;
+    qemu_co_queue_restart_all(&bs->flush_queue);
+
     tracked_request_end(&req);
     return ret;
 }
@@ -2421,6 +2441,7 @@ int coroutine_fn bdrv_co_discard(BlockDriverState *bs, int64_t sector_num,
     }
     ret = 0;
 out:
+    ++bs->write_gen;
     bdrv_set_dirty(bs, req.offset >> BDRV_SECTOR_BITS,
                    req.bytes >> BDRV_SECTOR_BITS);
     tracked_request_end(&req);