From 1cdc3239f1bb8c8f18954defe3cb813edc9df4a0 Mon Sep 17 00:00:00 2001 From: "Denis V. Lunev" Date: Fri, 30 Jan 2015 11:42:16 +0300 Subject: block: use fallocate(FALLOC_FL_PUNCH_HOLE) & fallocate(0) to write zeroes This sequence works efficiently if FALLOC_FL_ZERO_RANGE is not supported. Unfortunately, FALLOC_FL_ZERO_RANGE is supported on really modern systems and only for a couple of filesystems. FALLOC_FL_PUNCH_HOLE is much more mature. The sequence of 2 operations FALLOC_FL_PUNCH_HOLE and 0 is necessary due to the following reasons: - FALLOC_FL_PUNCH_HOLE creates a hole in the file, the file becomes sparse. In order to retain original functionality we must allocate disk space afterwards. This is done using fallocate(0) call - fallocate(0) without preceeding FALLOC_FL_PUNCH_HOLE will do nothing if called above already allocated areas of the file, i.e. the content will not be zeroed This should increase the performance a bit for not-so-modern kernels. CC: Max Reitz CC: Kevin Wolf CC: Stefan Hajnoczi CC: Peter Lieven CC: Fam Zheng Signed-off-by: Denis V. Lunev Reviewed-by: Max Reitz Signed-off-by: Kevin Wolf --- block/raw-posix.c | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/block/raw-posix.c b/block/raw-posix.c index 1c88ad8..7b42f37 100644 --- a/block/raw-posix.c +++ b/block/raw-posix.c @@ -967,6 +967,25 @@ static ssize_t handle_aiocb_write_zeroes(RawPosixAIOData *aiocb) } #endif +#ifdef CONFIG_FALLOCATE_PUNCH_HOLE + if (s->has_discard && s->has_fallocate) { + int ret = do_fallocate(s->fd, + FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE, + aiocb->aio_offset, aiocb->aio_nbytes); + if (ret == 0) { + ret = do_fallocate(s->fd, 0, aiocb->aio_offset, aiocb->aio_nbytes); + if (ret == 0 || ret != -ENOTSUP) { + return ret; + } + s->has_fallocate = false; + } else if (ret != -ENOTSUP) { + return ret; + } else { + s->has_discard = false; + } + } +#endif + #ifdef CONFIG_FALLOCATE if (s->has_fallocate && aiocb->aio_offset >= bdrv_getlength(aiocb->bs)) { int ret = do_fallocate(s->fd, 0, aiocb->aio_offset, aiocb->aio_nbytes); -- cgit v1.1