media: platform: sh_veu: use 64-bit arithmetic instead of 32-bit

Cast left and top to dma_addr_t in order to give the compiler complete
information about the proper arithmetic to use. Notice that these
variables are being used in contexts that expect expressions of
type dma_addr_t (64 bit, unsigned).

Such expressions are currently being evaluated using 32-bit arithmetic.

Also, move the expression (((dma_addr_t)left * veu->vfmt_out.fmt->depth) >> 3)
at the end in order to avoid a line wrapping checkpatch.pl warning.

Addresses-Coverity-ID: 1056807 ("Unintentional integer overflow")
Addresses-Coverity-ID: 1056808 ("Unintentional integer overflow")

Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
This commit is contained in:
Gustavo A. R. Silva 2018-02-06 11:52:24 -05:00 committed by Mauro Carvalho Chehab
parent 8328ad0f82
commit eb64311ffc

View file

@ -520,8 +520,8 @@ static void sh_veu_colour_offset(struct sh_veu_dev *veu, struct sh_veu_vfmt *vfm
/* dst_left and dst_top validity will be verified in CROP / COMPOSE */
unsigned int left = vfmt->frame.left & ~0x03;
unsigned int top = vfmt->frame.top;
dma_addr_t offset = ((left * veu->vfmt_out.fmt->depth) >> 3) +
top * veu->vfmt_out.bytesperline;
dma_addr_t offset = (dma_addr_t)top * veu->vfmt_out.bytesperline +
(((dma_addr_t)left * veu->vfmt_out.fmt->depth) >> 3);
unsigned int y_line;
vfmt->offset_y = offset;