[media] ti-vpe: get rid of some smatch warnings
authorMauro Carvalho Chehab <mchehab@s-opensource.com>
Tue, 22 Nov 2016 11:05:59 +0000 (09:05 -0200)
committerMauro Carvalho Chehab <mchehab@s-opensource.com>
Tue, 22 Nov 2016 11:08:36 +0000 (09:08 -0200)
When compiled on i386, it produces several warnings:

./arch/x86/include/asm/bitops.h:457:22: warning: asm output is not an lvalue
./arch/x86/include/asm/bitops.h:457:22: warning: asm output is not an lvalue
./arch/x86/include/asm/bitops.h:457:22: warning: asm output is not an lvalue
./arch/x86/include/asm/bitops.h:457:22: warning: asm output is not an lvalue
./arch/x86/include/asm/bitops.h:457:22: warning: asm output is not an lvalue
./arch/x86/include/asm/bitops.h:457:22: warning: asm output is not an lvalue

I suspect that some gcc optimization could be causing the asm code to be
incorrectly generated. Splitting it into two macro calls fix the issues
and gets us rid of 6 smatch warnings, with is a good thing. As it should
not cause any troubles, as we're basically doing the same thing, let's
apply such change to vpe.c.

Cc: Benoit Parrot <bparrot@ti.com>
Cc: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
drivers/media/platform/ti-vpe/vpe.c

index 0626593a8b229d5e56389cc176383932a6a0b8c3..f0156b7759e9263d303abdf5e145de6e54f3ac33 100644 (file)
@@ -1615,20 +1615,32 @@ static int __vpe_try_fmt(struct vpe_ctx *ctx, struct v4l2_format *f,
         */
        depth_bytes = depth >> 3;
 
-       if (depth_bytes == 3)
+       if (depth_bytes == 3) {
                /*
                 * if bpp is 3(as in some RGB formats), the pixel width doesn't
                 * really help in ensuring line stride is 16 byte aligned
                 */
                w_align = 4;
-       else
+       } else {
                /*
                 * for the remainder bpp(4, 2 and 1), the pixel width alignment
                 * can ensure a line stride alignment of 16 bytes. For example,
                 * if bpp is 2, then the line stride can be 16 byte aligned if
                 * the width is 8 byte aligned
                 */
-               w_align = order_base_2(VPDMA_DESC_ALIGN / depth_bytes);
+
+               /*
+                * HACK: using order_base_2() here causes lots of asm output
+                * errors with smatch, on i386:
+                * ./arch/x86/include/asm/bitops.h:457:22:
+                *               warning: asm output is not an lvalue
+                * Perhaps some gcc optimization is doing the wrong thing
+                * there.
+                * Let's get rid of them by doing the calculus on two steps
+                */
+               w_align = roundup_pow_of_two(VPDMA_DESC_ALIGN / depth_bytes);
+               w_align = ilog2(w_align);
+       }
 
        v4l_bound_align_image(&pix->width, MIN_W, MAX_W, w_align,
                              &pix->height, MIN_H, MAX_H, H_ALIGN,