BBB USB Transfer Rate

Untill now, the maximum transfer rate I could achieve is ~11 MB/s. I’ve tried this with different kernels with and without DMA enabled.

Could anyone get more than this? If so, how can I achieve that?

Thank you!

TC

I am having transfer rate issues as well. My post over on BeagleBone group has gotten zero attention.
https://groups.google.com/forum/#!topic/beaglebone/KxYwZewmRLY
In my case, the BBB cannot handle the video throughput of my camera, meanwhile a Rpi can.
Jeff

I can also confirm ~10MB/sec transfer rates with some testing I've
done. This was some time ago, maybe 6 months, and with Angstrom.

Chris

Just to let you all, and anyone else know my own findings . . . I setup a rootfs to load from an external USB hard drive( last year-ish ). I found that writes were as high as 20MB/s, but reads were around like the both of you seem to say. 10MB/s. So my own numbers seem to jibe with both of yours as well. At least for reads.

Also as far as I know there was no DMA involved, and I could not say for sure which block size I used . . . maybe I posted that in an old post, but I do not remember. The methodology I used for testing was dd( /dev/zero to disk ) and time, together. So nothing fancy.

Just as a reference on my test result

I use usb gadget network and run iperf as benchtools

uname -a

Linux buildroot 3.14.33 #10 SMP PREEMPT Sun Feb 15 13:57:32 CST 2015 armv7l GNU/Linux

cat /proc/cpuinfo

processor : 0
model name : ARMv7 Processor rev 2 (v7l)
BogoMIPS : 996.14
Features : swp half thumb fastmult vfp edsp thumbee neon vfpv3 tls vfpd32
CPU implementer : 0x41
CPU architecture: 7
CPU variant : 0x3
CPU part : 0xc08
CPU revision : 2

Hardware : Generic AM33XX (Flattened Device Tree)
Revision : 0000
Serial : 0000000000000000

my PC ip 10.10.0.1
board ip 10.10.0.2

Hi,

I’m not sure if this is relevant or not … maybe Robert or someone with a better understand of these things would know … ???

AM335x SDK6/7 kernel: USB isochronous with DMA patches, known issues

AMSDK8.0 (3.14 kernel) has been released last week. Can you please check if the issue still exists in kernel 3.14?

http://e2e.ti.com/support/embedded/linux/f/354/p/332530/1423886#1423886

our 3.14.x-ti kernel is based on ^^^^

You can test via:

sudo apt-get update ; sudo apt-get install linux-image-3.14.33-ti-r50
; sudo reboot

Regards,

Merci Robert and thanks for all your great work - it is appreciated !
////
I’d say it is not fixed

root@beaglebone:~# uname -a
Linux beaglebone 3.14.33-ti-r50 #1 SMP PREEMPT Wed Feb 11 18:16:43 UTC 2015 armv7l GNU/Linux

root@beaglebone:~# hdparm -t /dev/sda

/dev/sda:
Timing buffered disk reads: 28 MB in 3.16 seconds = 8.87 MB/sec

based on the release notes and known issues - there are still open points
http://processors.wiki.ti.com/index.php/Sitara_Linux_SDK_Kernel_Release_Notes

Merci Robert and thanks for all your great work - it is appreciated !
////
I'd say it is not fixed

root@beaglebone:~# uname -a
Linux beaglebone 3.14.33-ti-r50 #1 SMP PREEMPT Wed Feb 11 18:16:43 UTC
2015 armv7l GNU/Linux

root@beaglebone:~# hdparm -t /dev/sda

/dev/sda:
Timing buffered disk reads: 28 MB in 3.16 seconds = 8.87 MB/sec

So you have a crappy usb-flash drive... How do we know that's the usb
bus speed and not just a slow drive. :wink:

based on the release notes and known issues - there are still open points
http://processors.wiki.ti.com/index.php/Sitara_Linux_SDK_Kernel_Release_Notes

There will always be open points...

Regards,

fair point … I should have the same drive on my desktop with

https://github.com/beagleboard/linux/blob/3.14/arch/arm/configs/bb.org_defconfig#L3728