Re: [PULL 12/15] migration/block: rename BLOCK_SIZE macro

[Peter Maydell]

On Fri, 28 Feb 2020 at 09:30, Juan Quintela <quintela@redhat.com> wrote:
>
> From: Stefan Hajnoczi <stefanha@redhat.com>
>
> Both <linux/fs.h> and <sys/mount.h> define BLOCK_SIZE macros.  Avoiding
> using that name in block/migration.c.
>
> I noticed this when including <liburing.h> (Linux io_uring) from
> "block/aio.h" and compilation failed.  Although patches adding that
> include haven't been sent yet, it makes sense to rename the macro now in
> case someone else stumbles on it in the meantime.

A rather old change, and it didn't even introduce the code that
Coverity is complaining about, but this seems as good a point as
any to hang the email off of...

BLK_MIG_BLOCK_SIZE doesn't have a ULL suffix, so it's 32 bits,
and so Coverity complains about places where we multiply some
block count by it and then use that in a 64-bit result, eg here:

> @@ -770,7 +771,7 @@ static int block_save_iterate(QEMUFile *f, void *opaque)
>
>      /* control the rate of transfer */
>      blk_mig_lock();
> -    while (block_mig_state.read_done * BLOCK_SIZE <
> +    while (block_mig_state.read_done * BLK_MIG_BLOCK_SIZE <
>             qemu_file_get_rate_limit(f) &&
>             block_mig_state.submitted < MAX_PARALLEL_IO &&
>             (block_mig_state.submitted + block_mig_state.read_done) <

and here:

> @@ -874,13 +875,13 @@ static void block_save_pending(QEMUFile *f, void 
> *opaque, uint64_t max_size,
>      qemu_mutex_unlock_iothread();
>
>      blk_mig_lock();
> -    pending += block_mig_state.submitted * BLOCK_SIZE +
> -               block_mig_state.read_done * BLOCK_SIZE;
> +    pending += block_mig_state.submitted * BLK_MIG_BLOCK_SIZE +
> +               block_mig_state.read_done * BLK_MIG_BLOCK_SIZE;
>      blk_mig_unlock();

Putting a suitable cast to ensure the multiply is done at
64 bits would satisfy Coverity.

This is CID 1487136, 1487175.

thanks
-- PMM