On Mon, May 09, 2016 at 10:46:08AM -0400, Andre Smagin wrote: > Hello. > > cat audio.wav > /dev/audio > results in a cryptic "audio0: play interrupt but not playing" > on the console, but the audio file plays ok.
this one is harmless > If I use aucat or mpg123 or flac123 to play something, I get > "pxa2x0_dma_to_fifo: bogus length 17640" on the console and the player just > hangs there with no sound. now this one is fixed; the problem was that the audio(4) driver used to allow block sizes larger than the maximum supported by the zaudio(4) driver. still audio stutters on my machine; the way zaudio(4) does dma doesn't seem correct, which probably causes the stuttering. > On a related note, while browsing src/sys/arch/arm/xscale I noticed that > some error printf's have '\n', but some don't, resulting in the errors as > at the end of the dmesg below. What is the preferred way to handle new > lines in the kernel? printf with '\n' or some standard error macros/functions? you're right; messages are supposed to be terminated with '\n'. The following is not right: > > pxa2x0_dma_to_fifo: bogus length 17640pxa2x0_dma_to_fifo: bogus length > 17640pxa2x0_dma_to_fifo: bogus length 17640 >
