On Sat, Nov 21, 2020 at 12:10:50PM +0100, Lukas Wunner wrote: > > > > > + /* > > > > + * If ATTN is present and MRL is triggered > > > > + * ignore the Presence Change Event. > > > > + */ > > > > + if (ATTN_BUTTN(ctrl) && (events & PCI_EXP_SLTSTA_MRLSC)) > > > > + events &= ~PCI_EXP_SLTSTA_PDC; > > > > > > An Attention Button press results in a synthesized PDC event after > > > 5 seconds, which may get lost due to the above if-statement. > > > > When its synthesized you don't get the MRLSC? So we won't nuke the PDC then > > correct? > > I just meant to say, pciehp_queue_pushbutton_work() will synthesize > a PDC event after 5 seconds and with the above code snippet, if an > MRL event happens simultaneously, that synthesized PDC event would > be lost. So I'd just drop the above code snippet. I think you > just need to subscribe to MRL events and propagate them to > pciehp_handle_presence_or_link_change(). There, you'd bring down > the slot if an MRL event has occurred (same as DLLSC or PDC). > Then, check whether MRL is closed. If so, and if presence or link > is up, try to bring up the slot. > > If the MRL is open when slot or presence has gone up, the slot is not > brought up. But once MRL is closed, there'll be another MRL event and > *then* the slot is brought up. > Sounds good.. I'll send the update patch.
