Rockbox

Tasklist

FS#7447 - CPU remains boosted during playback if voice is in use.

Attached to Project: Rockbox
Opened by Steve Bavin (pondlife) - Tuesday, 17 July 2007, 12:05 GMT
Last edited by Steve Bavin (pondlife) - Tuesday, 31 July 2007, 10:42 GMT
Task Type Bugs
Category Applications
Status Closed
Assigned To No-one
Operating System SW-codec
Severity Medium
Priority Normal
Reported Version Daily build (which?)
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

If a voice file is in use, the CPU is always boosted during playback.

When in the "unwanted boost" state, the CPU boost log shows the following:
B main.c:328
B common/dircache.c:554
U common/dircache.c:563
B tagcache.c:3835
U tagcache.c:3853
U main.c:542
B tagcache.c:3803
B thread.c:915
B playlist.c:1219
B talk.c:218
U talk.c:230
U tagcache.c:3816
U playlist.c:1244

If I read this correctly, it indicates that the unwanted boost was started in thread.c:915. This just indicates a call to trigger_cpu_boost() was used, very likely one of the calls in playback.c...
This task depends upon

Closed by  Steve Bavin (pondlife)
Tuesday, 31 July 2007, 10:42 GMT
Reason for closing:  Fixed
Additional comments about closing:  jhMikeS fixed in a better way.
Comment by Linus Nielsen Feltzing (linusnielsen) - Tuesday, 17 July 2007, 15:26 GMT
This sounds an awful lot like the bug I fixed in february. It must have resurfaced due to some other changes.

http://svn.rockbox.org/viewvc.cgi/trunk/apps/playback.c?r1=12304&r2=12314
Comment by Steve Bavin (pondlife) - Tuesday, 17 July 2007, 16:18 GMT
Could it be the new "timeout loop" added to voice_request_buffer_callback() in here:
http://svn.rockbox.org/viewvc.cgi/trunk/apps/playback.c?r1=12807&r2=12843
Comment by Linus Nielsen Feltzing (linusnielsen) - Tuesday, 17 July 2007, 16:43 GMT
The FIXME on line 929 looks suspicious. :-)
Comment by Michael Sevakis (MikeS) - Tuesday, 17 July 2007, 17:40 GMT
/me shall henceforth refrain from adding suspicious comments :P
Comment by Steve Bavin (pondlife) - Tuesday, 31 July 2007, 09:11 GMT
This seems to help... not sure if it results in the boost clearing too quickly, but it doesn't make my H340 slow down noticably.

Maybe MikeS could look at it, and an iPod Video user could test it?

Loading...