>> The problem with quake probably is caused by the mmap handling. The sb
>> code continue to use the full-duplex hack even if mmap is used.
>
>But this is not correct. If output is requested to be 16-bit and mmap is
^^^^^^^^^^^^^^^^^^^
Sure, it' s a bug. I didn' t want to tell otherwise. It' s been my first
objection 1 msec after I read the sb-dublex patch before it' s been
included in ac and then in Linus's tree.
>used for output, then 16-bit dma should be used for output. Same for input,
>and mmap of input and output both in 16-bit mode or both in 8-bit mode
>should give an error.
It's more simple, if mmap is used sb->fullduplex must be set to 0.
>Is it possible to switch dma's on mmap?
I think no. If it's possible, sure it isn' t trivial and safe as
unsetting a bit flag ;-).
Andrea Arcangeli
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/