Hi Marek, I'm Ccing Mathias Nyman, who is the new xHCI driver maintainer. He can help you look into the issue with the AsMedia host controller. Is there any chance these are desktop systems? If so, you could get a Rensas PCIe add-in card that will better than the AsMedia host under Linux. The good news is that Dan Williams has been working on a patchset to address the issue with the ASIX ethernet chipset. It addresses the technical debt in the xHCI driver that makes it not compliant with the xHCI 1.0 specification. I believe Dan was putting the finishing touches on a 10-patch RFC. Marek, perhaps he can work with you and see if the patches address your issues? I don't hold out much hope that they will fix the issues with the AsMedia chipset, but they should make the ASIX chipset work under either a Rensas host or an Intel host. David, in the future, would you please direct people with xHCI issues to Mathias Nyman, so that he's aware of any problems with the xHCI driver? Thanks, Sarah Sharp On Tue, Jul 22, 2014 at 04:08:01PM +0200, Marek Uher wrote: > > > Hi David, > > that's not good news. I have three same computers with ASMedia USB > Host XHCI Controller. I have also three same D-Link DUB-1312 adapters. > Now I can throw them out of the window. I am afraid that these bugs anyone > will fixed soon. It looks like another never ending story. In any case thank > you for your help and time. > > Regards, > > Marek > > ---------- Původní zpráva ---------- > Od: David Laight > Komu: 'Marek Uher' > Datum: 22. 7. 2014 15:52:27 > Předmět: RE: Problem with D-Link USB 3.0 to Gigabit Ethernet Adapter DUB-1312 > > From: Marek Uher > > Hi David, > > > > thank you very much for your explanation. My answers to your questions > > follow: > > > > * Yes, I have ASMedia ASM1042 SuperSpeed USB Host (XHCI) Controller > > in my computer. > > > > * Yes, D-Link DUB-1312 is using asix88179_178a linux kernel driver. > > > > Based on information from you it seems that my hardware is completely > > unusable under Linux until your patches will be adapted back to kernel and > > side effects will be fixed. Is it right? > > Basically yes. > > I didn't get around to trying a fix for the problem with the ASMedia > hot controller. Adding enough trace slows the driver down enough so > that it works. > Every time it rings the 'doorbell' only one of the many queued requests > is processed - and you can see it is the oldest one. > > With the Intel Panther point (Ivy bridge) chipset is sort of works > for a while, until the ring end occurs at the wrong alignment. > > I've no longer got the USB adapters - they were on loan for evaluation. > We decided not to use them for other reasons. > > David -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
