On Wed, Jul 31, 2024 at 2:47 PM Mathias Nyman <mathias.nyman@xxxxxxxxxxxxxxx> wrote: > > On 31.7.2024 15.28, Łukasz Bartosik wrote: > > On Tue, Jul 30, 2024 at 5:45 PM Mathias Nyman > > <mathias.nyman@xxxxxxxxxxxxxxx> wrote: > >> > >> On 30.7.2024 3.17, Łukasz Bartosik wrote: > >>> On Mon, Jul 29, 2024 at 4:11 PM Mathias Nyman > >>> <mathias.nyman@xxxxxxxxxxxxxxx> wrote: > >>>> > >>>> Hi > >>>> > >>>> On 25.7.2024 10.48, Łukasz Bartosik wrote: > >>>>> DbC driver does not handle ClearFeature Halt requests correctly > >>>>> which in turn may lead to dropping packets on the receive path. > >>>> > >>>> Nice catch. > >>>> Looks like a halted endpoint is treated almost as a disconnect. > >>>> > >> > >> > >> I pushed my thoughts to a "fix_dbc_halted_ep" branch, it compiles but is not complete. > >> It mostly focuses on getting the STALL case for bulk-in working which this report was > >> about. > >> > >> I think the code itself best describes what I'm trying to do. > >> > >> git://git.kernel.org/pub/scm/linux/kernel/git/mnyman/xhci.git fix_dbc_halted_ep > >> https://git.kernel.org/pub/scm/linux/kernel/git/mnyman/xhci.git/commit/?h=fix_dbc_halted_ep&id=8532b621314e93336535528d5d45e41974c75e01 > >> > >> If you can try it out it would be great. > >> > > > > Sure I will test your patch and get back with the result. > > Thanks, that patch was missing a "ctrl = readl(&dbc->regs->control);" line > > It's now fixed here: > https://git.kernel.org/pub/scm/linux/kernel/git/mnyman/xhci.git/commit/?h=fix_dbc_halted_ep&id=cf99b473a1477c1b3510af0021877197a039c43f > > Can you try that instead > I will. Thanks for the update. Łukasz > Thanks > Mathias >
