Izumi Tsutsui wrote Sun, 26 Jun 2022 15:53:49 +0900: > Just FYI, ancient XFree86 driver seems to have "iplan2p2", "iplan2p4", > and "iplan2p8" drivers (but I have not investigated them yet): > http://cvsweb.netbsd.org/bsdweb.cgi/xsrc/xfree/xc/programs/Xserver/iplan2p4/Attic/ I suppose those are the "native" planar mode drivers. > Random comments: > > - as rin@ said, it's safer to leave WSDISPLAYIO_GTYPE not mandatory, > (i.e. not treat errors as fatal) especially if WSDISPLAYIO_GET_FBINFO > is available for newer drivers > - we don't have well-defined ioctl specifications so we should not rely > on implicit expectations Ok then, I've updated the patch to just log an warning if WSDISPLAYIO_GTYPE fails. > - I wonder if we should consider about "len" for mmap in WsfbScreenInit()? > fPtr->fbi.fbi_fbsize is enough? What about fPtr->fbmem_len? In case of Amiga fbi_fbsize is correct, and fbmem_len is derived from it. > - As commented in WsfbScreenInit(), shadowFB is not supported > on depth < 8, so I wonder we should bother to check bitsPerPixel > (I doubt there could be >8bpp planar) For length calculation? Yes, seems safe to assume exactly 8bpp there. > - wonder what "Afb" stands for, even in Xorg > (eventually we could have Afb4, for luna68k and atari etc.?) I don't think it has any particular meaning beyond being the identifier for Amiga planar mode routines in Xorg. Atari seems to have separate routines: https://lists.x.org/archives/xorg-devel/2013-March/035831.html --jukka
Attachment:
wsfb2.diff
Description: Binary data