On Thu, 22 Dec 2005, Pete Zaitcev wrote: > On Thu, 22 Dec 2005 14:33:11 -0500, "Brown, Len" <[EMAIL PROTECTED]> wrote: > > > 2.6.12 worked, but neither ehci-hcd, nor ohci-hcd loaded and > > instead the keyboard interrupts came in over IRQ1 i8042 > > so I guess there is some sort of emulation at play there? > > Unclear why it isn't at play in other releases... > > Alan, do you remember when the default usb-handoff went in effect? > This smells like the one.
Default usb-handoff wasn't present in 2.6.14. > Yes, i8042 is mostly emulated these days. Laptops were emulated for > a very long time. Servers still ship with real i8042, until Intel > stops making southbridges with it. With kernels before 2.6.15-rc1, if you didn't load any HCD modules and you didn't use the usb-handoff kernel parameter, there would not be any BIOS handoff. The BIOS would drive the USB keyboard and emulate a serial keyboard. > I am afraid that Len is hitting an SMM BIOS bug here. I saw such things > with RHEL. Basically, even if we take controllers away from BIOS with > the handoff procedure, the BIOS still manages to get involved into > the bus sharing between EHCI and its companion OHCI, even though it > shouldn't. Dell shipped updates for BIOS on the PE 380 for that reason, > to name one example. However Len's problem doesn't involve bus sharing. Even without ohci-hcd, he still sees that ehci-hcd is unable to drive the docking station's high-speed hub. Alan Stern ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
