On Tue, 11 Mar 2003, Ken Marx wrote: > Hi. > > Thanks for the suggestion, but after playing with > various placements (just in case your 'iirc' wasn't) > of novbe and noddc, I still get the sporadic > 1-2 minute hangup described below.
> Anyway, here's more or less what I found, for > what it's worth (doing egrep -i 'ddc' of the logfile): > > WithOUT noddc or novbe set, or if > noddc is in ServerFlags section: > ------------------------------------ > > (II) Loading sub module "ddc" > (II) LoadModule: "ddc" > (II) Loading /usr/X11R6/lib/modules/libddc.a > (II) Module ddc: vendor="The XFree86 Project" > (II) MGA(0): I2C bus "DDC" initialized. > (II) MGA(0): I2C device "DDC:ddc2" registered. > (II) MGA(0): I2C device "DDC:ddc2" removed. > (II) MGA(0): I2C device "DDC:ddc2" registered. > (II) MGA(0): I2C device "DDC:ddc2" removed. > > With noddc in Monitor or card's Device section: > ----------------------------------------------- > (II) Loading sub module "ddc" > (II) LoadModule: "ddc" > (II) Loading /usr/X11R6/lib/modules/libddc.a > (II) Module ddc: vendor="The XFree86 Project" > (II) MGA(0): I2C bus "DDC" initialized. > (**) MGA(0): Option "NoDDC" > (II) MGA(0): DDC Monitor info: 0x0 > (II) MGA(0): end of DDC Monitor info So noDDC works in monitor or device sections. Good. (The code prints the DDC info even if there is none, hence the two "DDC Monitor info" lines). > novbe: > ------- > In Monitor, card's Device section: > > (WW) MGA(0): Option "novbe" is not used > > In ServerFlags section, no warning, but no > diagnostic is given at all. IIRC novbe is broken in 4.2.x I may or may not have implemented a novbeddc at some point. > I'm happy to try other ideas if anyone has > any. Does nobody else see this in 4.2.1 > with dvi cards/monitor combinations? (esp., mga) In 4.2.x novbe should make no difference, since the mga driver doesn't use VBE for DDC at run time (it does use it with X -configure). In 4.3 the mga driver tries vbe if ddc2 and ddc1 fail. In DVI mode, my MGA G550 was slow to start, until I added noDDC1 and noDDC2, but I don't remember trying noDDC. However you say that even with noDDC you get the delay. > >>My current headache is with a Matrox g400 and Iiyama AU4831D > >>19" lcd (max 1600x1200) [BTW, has anyone seen this work > >>in digital full res? I'd love to see a config for it]. A colleague has an Iiyama AU4831D working fine on the DVI port of a Matrox G550, with the config at http://www.dpmms.cam.ac.uk/~werdna/xfree86-devel/xf86config.iiyama19TFT It isn't on my desk, so I haven't noticed whether it has the delay. > >> .... > >> (==) MGA(0): Write-combining range (0xe4000000,0x4000) was already clear > >> (==) MGA(0): Write-combining range (0xe8000000,0x2000000) > >> (==) MGA(0): Write-combining range (0xe5000000,0x800000) was already clear > >> (II) MGA(0): vgaHWGetIOBase: hwp->IOBase is 0x03d0, hwp->PIOOffset is 0x0000 > >> (==) MGA(0): Write-combining range (0xa0000,0x10000) was already clear > >> > >>! ...>>>90-120+ second wait<<<<< > >> > >> (--) MGA(0): 16 DWORD fifo > >> (==) MGA(0): Default visual is TrueColor > >> (II) MGA(0): Using 4042 lines for offscreen memory. > >> (II) MGA(0): Using XFree86 Acceleration Architecture (XAA) > >> Screen to screen bit blits > >> Solid filled rectangles > >> ....etc. My logfiles have two vgaHWGetIOBase lines, but if there is nothing missing in the 90-120 second wait that is after DDC happens. (The two DVI monitors we have on G550 cards behave very differently, the iiyama AU4831D works with the XFree86 shipped with Red Hat 8.0, but my SGI 1600SW+MLA requires the matrox hal). -- Dr. Andrew C. Aitchison Computer Officer, DPMMS, Cambridge [EMAIL PROTECTED] http://www.dpmms.cam.ac.uk/~werdna _______________________________________________ XFree86 mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xfree86
