thanks steve chen a lot! but in my dm355 evm have test again and the output are below: bash-3.00# mount /dev/mmcblk0p1 /mnt -o sync bash-3.00# ls ColorRuler.dat cmemk.ko loadmodules.sh test_camerad areg core mapdmaq areg.ko dm350mmap.ko ramdisk.img.gz bash-3.00# time cp ramdisk.img.gz /mnt/ real 0m17.967s user 0m0.000s sys 0m1.900s bash-3.00# dd if=/dev/zero of=/mnt/test.img bs=1M count=5 5+0 records in 5+0 records out bash-3.00# rm /mnt/test.img bash-3.00# time dd if=/dev/zero of=/mnt/test.img bs=1M count=5 5+0 records in 5+0 records out real 0m17.874s user 0m0.000s sys 0m0.410s bash-3.00# time dd if=/dev/zero of=/mnt/test.img bs=1M count=100 100+0 records in 100+0 records out real 5m54.738s user 0m0.000s sys 0m6.830s
bash-3.00# uname -a Linux 10.0.0.17 2.6.18_pro500-davinci_evm-arm_v5t_le so the speed is about less than 300k!my sd is scandisk ultra-ii,why so slow? 2009-08-04 zuowenping 发件人： Steve Chen 发送时间： 2009-08-03 22:39:02 收件人： Vipin 抄送： 'zuowenping'; 'davinci-linux-open-source' 主题： RE: Re: why is the sd/mmc so slowly in writing about dm355? On Mon, 2009-08-03 at 18:09 +0530, Vipin wrote: > Is this read or write speed? Can you isolate and find different > speed? Can you also check the read/write command code coming to the > driver? I remember of an issue where single block commands being sent > from MMC core resulting in low performance. > > Thanks and regards, > ~Vipin > > From: davinci-linux-open-source-boun...@linux.davincidsp.com > [mailto:davinci-linux-open-source-boun...@linux.davincidsp.com] On > Behalf Of zuowenping > Sent: Monday, August 03, 2009 5:02 PM > To: Steve Chen > Cc: davinci-linux-open-source > Subject: Re: Re: why is the sd/mmc so slowly in writing about dm355? > > thansk Steve Chen a lot! I am testing with a sd card(ScanDisk 1.0G)! > IIRC single block command affects MMC but not SD. I did a few quick tests with a HP 2G SD card on the my DM355 EVM, and the console output are below. r...@192.168.40.102:~# dd if=/dev/zero of=/mnt/disk/write-test1 bs=1M count=20 20+0 records in 20+0 records out 20971520 bytes (21 MB) copied, 1.07127 seconds, 19.6 MB/s r...@192.168.40.102:~# r...@192.168.40.102:~# dd if=/dev/zero of=/tmp/write-test2 bs=1M count=20 20+0 records in 20+0 records out 20971520 bytes (21 MB) copied, 0.531279 seconds, 39.5 MB/s r...@192.168.40.102:~# r...@192.168.40.102:~# dd if=/tmp/write-test2 of=/mnt/disk/write-test2 40960+0 records in 40960+0 records out 20971520 bytes (21 MB) copied, 8.96917 seconds, 2.3 MB/s r...@192.168.40.102:~# dd if=/dev/zero of=/mnt/disk/write-test3 bs=1M count=100 100+0 records in 100+0 records out 104857600 bytes (105 MB) copied, 44.9918 seconds, 2.3 MB/s The first dd (write-test1) show unusually high data rate due to caching. dd returned before data is actually written to the SD card. I'm seeing about 2.3MB/s with an idle system on MV 2.6.18 kernel. The other MMC/SD slot has similar performance numbers. I'm not sure what caused the drastic difference in write performance between our systems. Regards, Steve
_______________________________________________ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
