Patchwork dma-mapping: work around clang bug

login
register
mail settings
Submitter Arnd Bergmann
Date March 7, 2019, 8 a.m.
Message ID <20190307080038.3727426-1-arnd@arndb.de>
Download mbox | patch
Permalink /patch/743057/
State New
Headers show

Comments

Arnd Bergmann - March 7, 2019, 8 a.m.
Clang has a rather annoying behavior of checking for integer
arithmetic problems in code paths that are discarded by gcc
before that perfoms the same checks.

For DMA_BIT_MASK(64), this leads to a warning despite the
result of the macro being completely sensible:

arch/arm/plat-iop/adma.c:146:24: error: shift count >= width of type [-Werror,-Wshift-count-overflow]
                .coherent_dma_mask = DMA_BIT_MASK(64),

The best workaround I could come up with is to shift the
value twice, which makes the macro way less readable but
always has the same result.

Link: https://bugs.llvm.org/show_bug.cgi?id=38789
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
---
 include/linux/dma-mapping.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
Geert Uytterhoeven - March 7, 2019, 8:28 a.m.
Hi Arnd,

On Thu, Mar 7, 2019 at 9:01 AM Arnd Bergmann <arnd@arndb.de> wrote:
> Clang has a rather annoying behavior of checking for integer
> arithmetic problems in code paths that are discarded by gcc
> before that perfoms the same checks.
>
> For DMA_BIT_MASK(64), this leads to a warning despite the
> result of the macro being completely sensible:
>
> arch/arm/plat-iop/adma.c:146:24: error: shift count >= width of type [-Werror,-Wshift-count-overflow]
>                 .coherent_dma_mask = DMA_BIT_MASK(64),
>
> The best workaround I could come up with is to shift the
> value twice, which makes the macro way less readable but
> always has the same result.
>
> Link: https://bugs.llvm.org/show_bug.cgi?id=38789
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> ---
>  include/linux/dma-mapping.h | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h
> index 75e60be91e5f..380d3a95d02e 100644
> --- a/include/linux/dma-mapping.h
> +++ b/include/linux/dma-mapping.h
> @@ -138,7 +138,8 @@ struct dma_map_ops {
>  extern const struct dma_map_ops dma_virt_ops;
>  extern const struct dma_map_ops dma_dummy_ops;
>
> -#define DMA_BIT_MASK(n)        (((n) == 64) ? ~0ULL : ((1ULL<<(n))-1))
> +/* double shift to work around https://bugs.llvm.org/show_bug.cgi?id=38789 */
> +#define DMA_BIT_MASK(n)        (((n) == 64) ? ~0ULL : (((1ULL<<((n)-1))-1) << 1))

The second "-1" should be done on the final result, not on the
intermediate value.

Gr{oetje,eeting}s,

                        Geert
Arnd Bergmann - March 7, 2019, 8:50 a.m.
On Thu, Mar 7, 2019 at 9:28 AM Geert Uytterhoeven <geert@linux-m68k.org> wrote:
> On Thu, Mar 7, 2019 at 9:01 AM Arnd Bergmann <arnd@arndb.de> wrote:
> > Clang has a rather annoying behavior of checking for integer
> > arithmetic problems in code paths that are discarded by gcc
> > before that perfoms the same checks.
> >
> > For DMA_BIT_MASK(64), this leads to a warning despite the
> > result of the macro being completely sensible:
> >
> > arch/arm/plat-iop/adma.c:146:24: error: shift count >= width of type [-Werror,-Wshift-count-overflow]
> >                 .coherent_dma_mask = DMA_BIT_MASK(64),
> >
> > The best workaround I could come up with is to shift the
> > value twice, which makes the macro way less readable but
> > always has the same result.
> >
> > Link: https://bugs.llvm.org/show_bug.cgi?id=38789
> > Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> > ---
> >  include/linux/dma-mapping.h | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h
> > index 75e60be91e5f..380d3a95d02e 100644
> > --- a/include/linux/dma-mapping.h
> > +++ b/include/linux/dma-mapping.h
> > @@ -138,7 +138,8 @@ struct dma_map_ops {
> >  extern const struct dma_map_ops dma_virt_ops;
> >  extern const struct dma_map_ops dma_dummy_ops;
> >
> > -#define DMA_BIT_MASK(n)        (((n) == 64) ? ~0ULL : ((1ULL<<(n))-1))
> > +/* double shift to work around https://bugs.llvm.org/show_bug.cgi?id=38789 */
> > +#define DMA_BIT_MASK(n)        (((n) == 64) ? ~0ULL : (((1ULL<<((n)-1))-1) << 1))
>
> The second "-1" should be done on the final result, not on the
> intermediate value.

Ah, of course. I'll send an update patch in a bit, sorry about this.

      Arnd

Patch

diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h
index 75e60be91e5f..380d3a95d02e 100644
--- a/include/linux/dma-mapping.h
+++ b/include/linux/dma-mapping.h
@@ -138,7 +138,8 @@  struct dma_map_ops {
 extern const struct dma_map_ops dma_virt_ops;
 extern const struct dma_map_ops dma_dummy_ops;
 
-#define DMA_BIT_MASK(n)	(((n) == 64) ? ~0ULL : ((1ULL<<(n))-1))
+/* double shift to work around https://bugs.llvm.org/show_bug.cgi?id=38789 */
+#define DMA_BIT_MASK(n)	(((n) == 64) ? ~0ULL : (((1ULL<<((n)-1))-1) << 1))
 
 #define DMA_MASK_NONE	0x0ULL