> Am 05.06.2019 07:58 schrieb Martin Kepplinger: >> I can test later but I think I can see what happens. It fixes the issue, >> but still prints "ERROR: Keyboard set scancode failed!".
There's also a chance that it'd prematurely "return;" slightly later - in this case my tiny patch wouldn't fix the issue and commenting out the another return would be also required On Thu, Jun 6, 2019 at 9:50 AM Martin Kepplinger <mart...@posteo.de> wrote: > > On 06.06.19 08:20, Paul Menzel wrote: > > Dear Martin, > > > > > > First, please try to use interleaved style for quoting. > > > > > > On 05.06.19 13:47, Martin Kepplinger wrote: > > > > > >> Am 05.06.2019 07:58 schrieb Martin Kepplinger: > >>> I can test later but I think I can see what happens. It fixes the issue, > >>> but still prints "ERROR: Keyboard set scancode failed!". > >>> > >>> If we know that it's no error on at least one platform, we shouldn't > >>> print "ERROR" IMO, but as long as the bug gets fixes, I guess I'm fine. > > > > It’s how it was done before, and I think it is useful to have error > > messages especially in case it aborts. Just to be sure, is the message > > > > ERROR: Keyboard set scancode failed! > > > > also new for you, or did you see it in the past before the change > > resetting the keyboard, but the keyboard kept working. (Maybe with no > > reset it was still working because of the initialization by SeaBIOS or > > GRUB.) > > it's new. Before the change that adds the reset, all was fine and no > errors printed. > > > > >> Paul, what do you think about removing the return > >> and error message from the RESET cmd? Other thoughts? > > > > Your error is not about the failing reset command anymore. > > > > You should be ablet to work around it by selecting > > LP_PC_KEYBOARD_IGNORE_INIT_FAILURE in the libpayload configuration. > > > how do I select that? > > > > > Maybe that should be selected by default or we should indeed remove all > > the returns from the error paths in the hope the keyboard will be > > functional despite the non-working commands before. > > > We must add a fix. If you can select said config by default for > libpayload users, that's fine. Otherwise, remove the return I guess... > > thanks, > martin _______________________________________________ coreboot mailing list -- coreboot@coreboot.org To unsubscribe send an email to coreboot-le...@coreboot.org