[PATCH for-4.2 4/4] block/file-posix: Let post-EOF fallocate serialize

[Max Reitz <mreitz@redhat.com>]

The XFS kernel driver has a bug that may cause data loss when using
fallocate() in an I/O path, i.e. writing zeroes.  We can work around it
by treating post-EOF fallocates as serializing up until infinity
(INT64_MAX in practice).

Signed-off-by: Max Reitz <mreitz@redhat.com>
---
 block/file-posix.c | 38 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 38 insertions(+)

diff --git a/block/file-posix.c b/block/file-posix.c
index 0b7e904d48..d5460f3e45 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -2721,6 +2721,44 @@ raw_do_pwrite_zeroes(BlockDriverState *bs, int64_t offset, int bytes,
     RawPosixAIOData acb;
     ThreadPoolFunc *handler;
 
+#ifdef CONFIG_FALLOCATE
+    if (offset + bytes > bs->total_sectors * BDRV_SECTOR_SIZE) {
+        BdrvTrackedRequest *req;
+        uint64_t end;
+
+        /*
+         * This is a workaround for a bug in the Linux XFS driver,
+         * where writes submitted through the AIO interface will be
+         * discarded if they happen beyond a concurrently running
+         * fallocate() that increases the file length (i.e., both the
+         * write and the fallocate() happen beyond the EOF).
+         *
+         * To work around it, we extend the tracked request for this
+         * zero write until INT64_MAX (effectively infinity), and mark
+         * it as serializing.
+         *
+         * We have to enable this workaround for all filesystems and
+         * AIO modes (not just XFS with aio=native), because for
+         * remote filesystems we do not know the host configuration.
+         * However, this should have no effect as long as no block
+         * driver submits zero writes beyond the EOF.
+         */
+
+        req = bdrv_co_get_self_request(bs);
+        assert(req);
+        assert(req->type == BDRV_TRACKED_WRITE);
+        assert(req->offset <= offset);
+        assert(req->offset + req->bytes >= offset + bytes);
+
+        end = INT64_MAX & -(uint64_t)bs->bl.request_alignment;
+        req->bytes = end - req->offset;
+        req->overlap_bytes = req->bytes;
+
+        bdrv_mark_request_serialising(req, bs->bl.request_alignment);
+        bdrv_wait_serialising_requests(req);
+    }
+#endif
+
     acb = (RawPosixAIOData) {
         .bs             = bs,
         .aio_fildes     = s->fd,
-- 
2.21.0