On Sat, 19 May 2012, Ming Lei wrote: > IMO, it is not a good idea. > > Firstly, after writing RESET on command register, the default value of > CMD register > should be restored to '00080000h (00080B00h if Asynchronous Schedule Park > Capability is a one)'[1], and both 'Interrupt Threshold Control' and > 'Park mode' has > been set in ehci_init(), so using ehci->command after ehci_reset is correct. > > Secondly, remaining other bits from HW is not good since HW may be buggy > and the default value after reset may be uncertain, and it is better > to not depend > on them. For some newer hardware, if the reserved fields are to be > used, just do > it in ehci_init like setting lpm bits. > > Finally, before the commit 3d9545cc375d117554a9b35dfddadf9189c62775( > EHCI: maintain the ehci->command value properly), the ehci->command is used > just directly in ehci_run and the other HW bits are not kept from HW, > so it is better > to be consistent as before to avoid introducing possible regression. > > [1], 4.1 Host Controller Initialization of EHCI Specification 1.0 > > > Would anyone like to code that up? > > Looks removing the line of 'ehci->command = ehci_readl(ehci, > &ehci->regs->command);' in ehci_reset is OK. You have convinced me. In addition, my earlier reasoning didn't make much sense. Nobody is developing new versions of EHCI these days; they're all working on USB-3. I'll send in a new patch simply to remove that line of code. Or you can, if you like. Alan Stern -- 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
