diff mbox series

[v2,3/9] block: Use uint64_t for BdrvTrackedRequest byte fields

Message ID 20180705073701.10558-4-famz@redhat.com
State New
Headers show
Series block: Fix dst reading after tail copy offloading | expand

Commit Message

Fam Zheng July 5, 2018, 7:36 a.m. UTC
This matches the types used for bytes in the rest parts of block layer.
In the case of bdrv_co_truncate, new_bytes can be the image size which
probably doesn't fit in a 32 bit int.

Signed-off-by: Fam Zheng <famz@redhat.com>
---
 block/io.c                | 2 +-
 include/block/block_int.h | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

Comments

Kevin Wolf July 5, 2018, 11:16 a.m. UTC | #1
Am 05.07.2018 um 09:36 hat Fam Zheng geschrieben:
> This matches the types used for bytes in the rest parts of block layer.
> In the case of bdrv_co_truncate, new_bytes can be the image size which
> probably doesn't fit in a 32 bit int.
> 
> Signed-off-by: Fam Zheng <famz@redhat.com>

mark_request_serialising() has this:

    unsigned int overlap_bytes = ROUND_UP(req->offset + req->bytes, align)
                               - overlap_offset;


There is also:

static bool tracked_request_overlaps(BdrvTrackedRequest *req,
                                     int64_t offset, unsigned int bytes)

Don't these need to be uint64_t now as well?

>  block/io.c                | 2 +-
>  include/block/block_int.h | 4 ++--
>  2 files changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/block/io.c b/block/io.c
> index 3e00667a2a..443a8584c4 100644
> --- a/block/io.c
> +++ b/block/io.c
> @@ -587,7 +587,7 @@ static void tracked_request_end(BdrvTrackedRequest *req)
>  static void tracked_request_begin(BdrvTrackedRequest *req,
>                                    BlockDriverState *bs,
>                                    int64_t offset,
> -                                  unsigned int bytes,
> +                                  uint64_t bytes,
>                                    enum BdrvTrackedRequestType type)
>  {
>      *req = (BdrvTrackedRequest){

Should we assert that offset + bytes <= INT64_MAX? We make this
assumption in basically all of the calculations.

Kevin
diff mbox series

Patch

diff --git a/block/io.c b/block/io.c
index 3e00667a2a..443a8584c4 100644
--- a/block/io.c
+++ b/block/io.c
@@ -587,7 +587,7 @@  static void tracked_request_end(BdrvTrackedRequest *req)
 static void tracked_request_begin(BdrvTrackedRequest *req,
                                   BlockDriverState *bs,
                                   int64_t offset,
-                                  unsigned int bytes,
+                                  uint64_t bytes,
                                   enum BdrvTrackedRequestType type)
 {
     *req = (BdrvTrackedRequest){
diff --git a/include/block/block_int.h b/include/block/block_int.h
index af71b414be..66c0e50d82 100644
--- a/include/block/block_int.h
+++ b/include/block/block_int.h
@@ -69,12 +69,12 @@  enum BdrvTrackedRequestType {
 typedef struct BdrvTrackedRequest {
     BlockDriverState *bs;
     int64_t offset;
-    unsigned int bytes;
+    uint64_t bytes;
     enum BdrvTrackedRequestType type;
 
     bool serialising;
     int64_t overlap_offset;
-    unsigned int overlap_bytes;
+    uint64_t overlap_bytes;
 
     QLIST_ENTRY(BdrvTrackedRequest) list;
     Coroutine *co; /* owner, used for deadlock detection */