From: Paul Gofman Subject: Re: [PATCH v2 1/3] winex11.drv: Add GUID_DISPLAY_DEVICE_ARRIVAL interface to display device. Message-Id: <9315ee00-239f-4ab1-ad88-23f68175a7a5@codeweavers.com> Date: Mon, 6 Sep 2021 18:14:52 +0300 In-Reply-To: <86897d56-ec06-76a7-5a2a-cfd71399c52b@codeweavers.com> References: <20210906124239.59283-1-pgofman@codeweavers.com> <9f582756-19f6-b26b-c6ab-737c03e691c3@codeweavers.com> <12a53a10-11bc-34f5-9478-fe921247ef2f@codeweavers.com> <86897d56-ec06-76a7-5a2a-cfd71399c52b@codeweavers.com> On 9/6/21 17:45, Zhiyi Zhang wrote: > > On 9/6/21 10:40 PM, Paul Gofman wrote: >> On 9/6/21 17:26, Zhiyi Zhang wrote: >>> iface_data is not used. You can pass NULL instead. >> Yeah, thanks, I will resend. >> >> Not strictly related, but is there any reason why your Proton patch >> "winex11.drv: Link display device interfaces" cannot go upstream? Yes, I >> know the functionality is managed in PnP driver on Windows, but since we >> don't have one and unlikely to have it in the near future is there >> anything wrong in making DIGCF_PRESENT flag work this way with display >> devices? Some games depend on that, and my other patch for >> D3DKMTOpenAdapterFromDeviceName() implementation is hard to get right >> without that one working. >> > It's because the way DIGCF_PRESENT that is implemented in Proton is incorrect. > So "winex11.drv: Link display device interfaces" is also not proper. To summarize, we had a brief conversation with Zhiyi in the chat and DIGCF_PRESENT is actually implemented upstream for device interfaces, so the current version of the "Link display device interfaces" patch makes some sense upstream too. So Zhiyi is going to look into upstreaming that patch. While my patch for GUID_DISPLAY_DEVICE_ARRIVAL is not actually much useful with SetupDiGetClassDevsW() not returning any interfaces for display devices when DIGCF_PRESENT is specified (the only game which I know depends on that specifies that flag), so I will resend the patch once that part is upstream.