Hi David, > > I have been testing a few boards that use the VT8235 and there have > > been some improvements. ... > > Thanks for the testing results. > > I know of one (elusive) bug that's not VIA-specific ... if when that's > finally found/clobbered then we still see such misbehavior on VIA,
Is it likely that this impacts more on the VIA than other implementations? Come to think of it, have you tested Intel's EHCI implementation found on newer versions of their chipsets? Is it's reliability more like the VIA or the NEC? >I > suspect I know some blunt instruments that could be applied when the > PCI vendor ID says "VIA". I considered blacklisting the VIA and allowing NEC and other working implementations to use the driver, but this isn't an ideal solution (it would allow USB 2.0 devices to run in USB 1.x mode without hanging). > > The 2.5. storage driver appears to have many advantages over the 2.4. > > Kernel's storage driver, but 2.5. series kernels are not viable for > > most solutions at this stage (there are unrelated problems with other > > modules etc). Are the 2.5 kernels simply too different from 2.4 for a > > backport to be possible? > > Matt may have more to say, but I for one wouldn't care to try to put > the usb_sg_*() calls on 2.4 ... they're more demanding of host > controller driver fault handling, and such changes would likely > destabilize 2.4 more than most folk would want. And by now I think the > SCSI (and block) layer code has diverged enough to make backports > troublesome. That's what I thought. The throughput speed is substaincially better on 2.5, but 2.4 is generally quite reliable. > > There must be other issues not related to the EHCI driver, but > > relating to the VT8235 in general on 2.4. ... > > Interesting. I know that VIA has a bit of history with PCI problems, > maybe it's (still) a factor here. It very well may be. After disabling the VIA USB controllers on this computer and used the NEC, the system would hang after copying 240Mb of a 1.5Gb file to a USB 2.0 HDD. On the KT133 (and an AMD 756 chipset on an older Athlon), I wouldn't think twice, the transfer would always work without ever hanging. ------------------------------------------------------- This SF.NET email is sponsored by: FREE SSL Guide from Thawte are you planning your Web Server Security? Click here to get a FREE Thawte SSL guide and find the answers to all your SSL security issues. http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0026en _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
