As Maarten Baert recently reported, the current dmix code applies the
semaphore unnecessarily around mixing streams even when the lockless
mix operation is used on x86. This was rather introduced mistakenly
at the commit 267d7c7281 ("Add support of little-endian on
i386/x86_64 dmix") where the generic dmix code was included on x86,
too.
For achieving the original performance back, this patch changes the
semaphore handling to be checked at run time instead of statically at
compile time.
Reviewed-by: Jaroslav Kysela <perex@perex.cz>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In the case of the sum buffer being uninitialized, the source sample
must be negated not only when writing to the sum but also when writing
to the destination.
i386/x86_64 alsa-lib may need to handle big-endian formats, e.g.
when running via qemu on PPC. The generic dmix code already has
both endian support, so let's use it as fallback.
Here's a patch for generic dmix which fixes S16 byte swapping.
Tested on powerpc with snd-usb-audio. (Without the patch I get crackling.)
Signed-off-by: Juergen Kreileder <jk@blackdown.de>
Add to the dmix plugin support for the S24_3LE sample format which is
used by 24-bit USB devices.
The optimized assembler version uses only 23 bits for sample data so
that the lowest bit can be used for synchronization because there is no
24-bit cmpxchg instruction.
Support dmix on generic architectures without atomic operations but
using a semaphore to avoid concurrent accesses. This is less effective
than atomic operations but should work on every system.
