Does anyone know if there are any known factors that could cause higher/lower CPU load while playing the same mp3 songs using gplay? The OS is Linux 2.6.35, and with the same amount of DDR2 memory.
Where is this mp3 file stored, on an SD card? NAND flash? Different SD cards may have different reading speed, that may impact CPU loading.
The songs are stored in emmc on sd0, the same condition for both platforms.
Could it be the different DDR2 used that cause the cpu load differences?
The cpu load is around 30% steady while playing mp3 on the MX287 board, but
varies from 30% to 65% on MX280 board playing the same song.
On Nov 5, 2013 4:34 PM, "Yongcai Huang" <firstname.lastname@example.org>
To identify the root cause, I think you should isolate these modules one by one, below are my idea:
1. Write a simple script to let cpu counting in a loop with ++ and usleep, then check the cpu loading, this is to see whether it is caused by different CPU type;
2. Do memory copy test, check the DDR speed and CPU loading,, this is to check whether it is caused by different DDR2;
3. If available, using a ramfs, then playing mp3 to see whether it is caused by emmc.
This is my idea, not sure whether it works, but I did NOT have mx28 baords here, so you may need to try them on your side if you think they are making sense.
Thanks for the quick responses. We will do tests accordingly.
On Nov 5, 2013 5:02 PM, "Yongcai Huang" <email@example.com>
Retrieving data ...