On Tue, Apr 04, 2006 at 04:13:59PM +0100, Bahadir Balban wrote: > Hi, > > I am trying to measure usb transfer throughput by timing file read > operations from a usb flashdisk into main memory. The results seems to > be imprecise and I suspect this is due to filesystem caching. How > could I prohibit filesystem caching in the kernel? > > Do you know of any GPL'ed usb benchmarks? Maybe I should measure this > in the kernel rather then application space? No, use usbfs to send raw data down to a device that you know can handle line speeds in order to measure the kernel throughput. Otherwise you are only testing the throughput of the USB device itself (which is almost always slower than what the kernel can provide.) Also, your host controller chip needs to be a good one, otherwise that is a big bottleneck. What do you really want to test this for? What are you expecting to do with the results? FYI, I have had reports of people who have done this kind of testing in the past and they say that Linux has the best USB throughput of any current operating system, we max out the hardware. thanks, greg k-h -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
