Hi Rafael, On Wed, Jun 12, 2024 at 05:26:46PM +0200, Rafael J. Wysocki wrote: > Hi, > > On Wed, Jun 12, 2024 at 4:30 PM Hans de Goede <hdegoede@xxxxxxxxxx> wrote: > > > > Hi, > > > > On 6/12/24 3:06 PM, Rafael J. Wysocki wrote: > > > Hi Sakari, > > > > > > On Wed, Jun 12, 2024 at 2:47 PM Sakari Ailus > > > <sakari.ailus@xxxxxxxxxxxxxxx> wrote: > > >> > > >> Hi Rafael, > > >> > > >> On Wed, Jun 12, 2024 at 02:32:26PM +0200, Rafael J. Wysocki wrote: > > >>>>>>> I just hit the same problem on another Dell laptop. It seems that > > >>>>>>> all Dell laptops with IPU6 camera from the Tiger Lake, Alder Lake > > >>>>>>> and Raptor Lake generations suffer from this problem. > > >>>>>>> > > >>>>>>> So instead of playing whack a mole with DMI matches we should > > >>>>>>> simply disable ACPI MIPI DISCO support on all Dell laptops > > >>>>>>> with those CPUs. I'm preparing a fix for this to replace > > >>>>>>> the DMI matching now. > > >>>>>> > > >>>>>> DisCo for Imaging support shouldn't be dropped on these systems, and this > > >>>>>> isn't what your patch does either. Instead the ACPI graph port nodes (as > > >>>>>> per Linux specific definitions) are simply dropped, i.e. this isn't related > > >>>>>> to DisCo for Imaging at all. > > >>>>> > > >>>>> So it looks like the changelog of that patch could be improved, right? > > >>>> > > >>>> Well, yes. The reason the function is in the file is that nearly all camera > > >>>> related parsing is located there, not that it would be related to DisCo for > > >>>> Imaging as such. > > >>> > > >>> So IIUC the camera graph port nodes are created by default with the > > >>> help of the firmware-supplied information, but if that is defective a > > >>> quirk can be added to skip the creation of those ports in which case > > >>> they will be created elsewhere. > > >>> > > >>> Is this correct? > > >> > > >> Yes. > > > > > > So it would be good to add a comment to this effect to > > > acpi_nondev_subnode_extract() where acpi_graph_ignore_port() is > > > called. > > > > > > And there is a somewhat tangential question that occurred to me: If > > > the nodes are created elsewhere when acpi_graph_ignore_port() is true, > > > why is it necessary to consult the platform firmware for the > > > information on them at all? Wouldn't it be better to simply always > > > create them elsewhere? > > > > That is a good question. The ACPI MIPI DISCO specification is an > > attempt standardize how MIPI cameras and their sensors are described > > in ACPI. > > > > But this is not actually being used by any Windows drivers atm. The windows > > drivers rely on their own custom ACPI data which gets translated into > > standard Linux device-properties by: drivers/media/pci/intel/ipu-bridge.c > > > > and so far AFAIK there are 0 laptops where there actually is 100% functional > > ACPI MIPI information. I believe that some work is in place to get correct > > usable ACPI MIPI information in place in the ACPI tables of some Meteor Lake > > laptops. But I believe that there too it does not work yet with the BIOS > > version with which current Windows models are shipping. It is being fixed > > for systems which have Linux support from the vendor but I suspect that > > on other models if ACPI MIPI DISCO information is there it will not > > necessarily be reliable because AFAICT Windows does not actually use it. > > > > And TBH this has me worried about camera support for Meteor Lake devices > > going forward. We really need to have 1 reliable source of truth here and > > using information which is ignored by Windows does not seem like the best > > source to use. > > > > Sakari I know you have been pushing for MIPI camera descriptions under > > ACPI to move to a standardized format and I can see how that is a good > > thing, but atm it seems to mainly cause things to break and before > > the ACPI MIPI DISCO support landed in 6.8 we did not have these issues, > > since the information used by the ipu-bridge code does seem to be correct. > > Well, if Windows doesn't use this information, it is almost guaranteed > to be garbage. No ACPI DSDT in production systems uses DisCo for Imaging as of now at least to my knowledge. > > So maybe it would be better to make acpi_graph_ignore_port() return > true by default and false only when the information is known to be > valid. IOW, whitelist things instead of adding blacklist entries in > perpetuum. What could be gained from this? > > And hopefully we'll eventually get to the point at which we are able > to say "whitelist everything from now on". -- Kind regards, Sakari Ailus
