Am 22.05.2012 18:17, schrieb Paolo Bonzini:
> Formats are entirely in charge of flushes for metadata writes. For
> guest-initiated writes, a writethrough cache is faked in the block layer.
>
> Signed-off-by: Paolo Bonzini <[email protected]>
> ---
> block.c | 2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/block.c b/block.c
> index 7add33c..43b3de6 100644
> --- a/block.c
> +++ b/block.c
> @@ -661,7 +661,7 @@ static int bdrv_open_common(BlockDriverState *bs, const
> char *filename,
> if (drv->bdrv_file_open) {
> ret = drv->bdrv_file_open(bs, filename, open_flags);
Why don't you add the writeback flag here? Would the code is raw-posix
that adds O_DSYNC be unused then. and can it still be used in other ways
than format=file/host_* with this patch?
> } else {
> - ret = bdrv_file_open(&bs->file, filename, open_flags);
> + ret = bdrv_file_open(&bs->file, filename, open_flags |
> BDRV_O_CACHE_WB);
> if (ret >= 0) {
> ret = drv->bdrv_open(bs, open_flags);
> }
Kevin
