On Thu, 2018-06-07 at 00:10 +0200, Kevin Brace wrote: > I was thinking that it is not really desirable to let VBE start > tinkering with I2C bus related registers this early in the PreInit > callback,
That's probably a reasonable intuition, these days. I was just trying to describe the context behind why it does what it does. In particular, "just ask VBE" is the kind of broad brush stroke you can do across a bunch of drivers and have a reasonable expection (in 2000) that it would work, without needing to know all the details of how to get initialized far enough to have working I2C. In this sense ddxDesign.xml is out of date, and it should describe how RANDRful drivers should defer mode setup until ScreenInit. > and besides that, if r128 is being used on PowerPC (or Alpha or > SPARC) platform, there is no INT 10H VBE. Kiiiinda. There's not an x86 system firmware and there's not a vm86 syscall, but the int10 (x86emu really) code knows how to emulate that. So if you plugged in an r128 with an x86 firmware on it, we really could POST it, or anyway try to. Now if that r128 had a PPC firmware with it because it's the one built into an iMac, then yes vbe wouldn't do anything, but the device would also already be initialized, and you'd probably be using fbdev. - ajax _______________________________________________ xorg-devel@lists.x.org: X.Org development Archives: http://lists.x.org/archives/xorg-devel Info: https://lists.x.org/mailman/listinfo/xorg-devel