On Sun, May 31, 2015 at 10:12:07AM +0200, Takashi Iwai wrote:
At Sun, 31 May 2015 09:18:57 +0200, Takashi Iwai wrote:
At Sat, 30 May 2015 17:38:10 +0200, Mikko Rapeli wrote:
Fixes userspace compilation error:
error: expected specifier-qualifier-list before ‘DECLARE_BITMAP’ DECLARE_BITMAP(gpr_valid, 0x200); /* bitmask of valid initializers */
DECLARE_BITMAP etc macros are not meant for userspace headers and thus added here as private copies for emu10k.h.
Fix was suggested by Arnd Bergmann arnd@arndb.de in message 2168807.4Yxh5gl11Q@wuerfel on lkml.
Signed-off-by: Mikko Rapeli mikko.rapeli@iki.fi
include/uapi/sound/emu10k1.h | 23 ++++++++++++++++++++--- 1 file changed, 20 insertions(+), 3 deletions(-)
diff --git a/include/uapi/sound/emu10k1.h b/include/uapi/sound/emu10k1.h index ec1535b..7575f0f 100644 --- a/include/uapi/sound/emu10k1.h +++ b/include/uapi/sound/emu10k1.h @@ -34,6 +34,23 @@
#define EMU10K1_FX8010_PCM_COUNT 8
+/*
- Following definitions are copies from kernel headers to support compiling
- this header file in userspace. The definitions are not generally available
- in uapi headers so the needed things are copied here wtih __EMU10k1 prefix.
- */
+/* From include/linux/bitops.h */ +#define __EMU10K1_BITS_PER_BYTE 8 +/* From include/linux/kernel.h */ +#define __EMU10K1_DIV_ROUND_UP(n,d) (((n) + (d) - 1) / (d)) +/* From include/linux/bitops.h */ +#define __EMU10K1_BITS_TO_LONGS(nr) \
__EMU10K1_DIV_ROUND_UP(nr, __EMU10K1_BITS_PER_BYTE * sizeof(long))+/* From include/linux/types.h */ +#define __EMU10K1_DECLARE_BITMAP(name,bits) \
unsigned long name[__EMU10K1_BITS_TO_LONGS(bits)]This is way too complicated just for a few expansion of the aligned size arrays, IMO. Rather simplify it like below with a comment:
/* the array must be aligned to unsigned long (i.e. 4 or 8 bytes) */ #define __EMU10K1_DECLARE_BITMAP(name) \ unsigned long name[(bits) / sizeof(unsigned long)]
Crap, of course, it must be multiplied with 8.
/* the array must be aligned to unsigned long (i.e. 32 or 64) */ #define __EMU10K1_DECLARE_BITMAP(name) \ unsigned long name[(bits) / (sizeof(unsigned long) * 8)]
Thanks, I'll use this definition instead.
-Mikko