On 7/2/2018 6:10 PM, Alberto Garcia wrote:
On Fri 19 Jan 2018 01:50:06 PM CET, Anton Nefedov wrote:This will help to identify how many of the user-issued discard operations (accounted on a device level) have actually suceeded down on the host file (even though the numbers will not be exactly the same if non-raw format driver is used (e.g. qcow2 sending metadata discards)). Signed-off-by: Anton Nefedov <anton.nefe...@virtuozzo.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsement...@virtuozzo.com> --- block/file-posix.c | 21 +++++++++++++++++++-- 1 file changed, 19 insertions(+), 2 deletions(-) diff --git a/block/file-posix.c b/block/file-posix.c index 36ee89e..544ae58 100644 --- a/block/file-posix.c +++ b/block/file-posix.c @@ -158,6 +158,11 @@ typedef struct BDRVRawState { bool page_cache_inconsistent:1; bool has_fallocate; bool needs_alignment; + struct { + int64_t discard_nb_ok; + int64_t discard_nb_failed; + int64_t discard_bytes_ok; + } stats;Shouldn't this new structure be defined in a header file so other drivers can use it? Or did you define it here because you don't see that happening soon?
I guess there's no reason to burden the common header files as long as it's not really used anywhere else.
