This is the bug/patch tracker for Rockbox. Click here for more information.
Quick links: Bugs · Patches · Rockbox frontpage
FS#9498 - AAC IRAM improvements
Attached to Project:
Rockbox
Opened by MichaelGiacomelli (saratoga) - Saturday, 18 October 2008, 05:17 GMT+2
Last edited by Andree Buschmann (Buschel) - Wednesday, 08 June 2011, 22:00 GMT+2
Opened by MichaelGiacomelli (saratoga) - Saturday, 18 October 2008, 05:17 GMT+2
Last edited by Andree Buschmann (Buschel) - Wednesday, 08 June 2011, 22:00 GMT+2
|
DetailsThis patch reorganizes variables by removing a seldom used array while putting a more frequently used one in IRAM.
This gives a ~2% speedup on PP5024, but I would be interested in any Coldfire results, where the difference will hopefully be larger. |
This task depends upon
Closed by Andree Buschmann (Buschel)
Wednesday, 08 June 2011, 22:00 GMT+2
Reason for closing: Out of Date
Additional comments about closing: The patch (and even the idea behind) does not apply to the reworked libfaad.
Wednesday, 08 June 2011, 22:00 GMT+2
Reason for closing: Out of Date
Additional comments about closing: The patch (and even the idea behind) does not apply to the reworked libfaad.
Results on H300 (tested with a VBR AAC track, ~190kbps): SVN 188% realtime, +patch 192% realtime. So about the same speedup as on PP502x.
Same test on iPod G2: SVN 155.5% realtime, +patch 157.0% realtime, i.e. about 1% speedup on PP5002.
In that case I will probably save this extra memory and try to use it for AAC-HE and/or dual core optimizations later on.