Skip to contentSkip to conversation

helping debugging

DISCUSSION Thierry Vignaud 2 participants 2 messages Jun 28, 2007 – Jun 29, 2007

lore
Share & tools
lore
#0 lore raw

Hi

Here's a suggestion from Dave Airlie:

http://airlied.livejournal.com/46087.html

"If you are going to have lots of unsigned int sized free running pointers called things like hw_ptr or hw.ptr or anything, for debugging it would be really cool if you started those counters at INT_MAX-100000 or something instead of at zero,

24 hours waiting for a bug to trip in the pcm rate conversion is not fun and I'm guessing it'll be some stupid interger overflow or dumb boundary condition hitting..."

#1 lore raw

At Thu, 28 Jun 2007 15:08:57 +0200,

Quoted 7 lines · Thierry Vignaud

Hi

Here's a suggestion from Dave Airlie:

http://airlied.livejournal.com/46087.html

Quoted 8 lines

"If you are going to have lots of unsigned int sized free running pointers called things like hw_ptr or hw.ptr or anything, for debugging it would be really cool if you started those counters at INT_MAX-100000 or something instead of at zero,

24 hours waiting for a bug to trip in the pcm rate conversion is not fun and I'm guessing it'll be some stupid interger overflow or dumb boundary condition hitting..."

Yeah, that's what I wanted to do, too.
I was busy in this week for a company event and had no time for ALSA work, so hopefully will be able to work on this in the next week...

Takashi