[LAD] Possible snd_pcm_hw_params_set_buffer_time_near() gotcha

[Victor Lazzarini]

I guess the problem is that none of the right side variables/constant  
is unsigned in:

>  unsigned buffer_time = 1000000*period*nperiods/rate;

so regardless of storing it in an unsigned, it overflows before you  
try to store it, and so it messes up everything.
Note that you could also prevent this by using the correct constant  
10000000U

Victor


On 7 Dec 2009, at 19:22, Mario Lang wrote:

> Hi.
>
> I just happened to stumble about a (what I call) pretty subtle
> bug in typical ALSA PCM code:
>
> root at fzidpc73:/tmp# cat test.c
> #include <stdio.h>
> int main() {
>  int period = 1024;
>  int nperiods = 3;
>  int rate = 96000;
>  unsigned buffer_time = 1000000*period*nperiods/rate;
>  printf("%d\n", buffer_time);
> }
> root at fzidpc73:/tmp# ./test
> -12739
> root at fzidpc73:/tmp# sed -i -e 's/int rate/unsigned int rate/' test.c
> root at fzidpc73:/tmp# gcc -o test test.c
> root at fzidpc73:/tmp# ./test
> 32000
>
> I.e., don't trust this boilerplate formula blindly.
>
> -- 
> CYa,
>  ⡍⠁⠗⠊⠕
> _______________________________________________
> Linux-audio-dev mailing list
> Linux-audio-dev at lists.linuxaudio.org
> http://lists.linuxaudio.org/mailman/listinfo/linux-audio-dev