Re: [Dri-devel] Radeon on KT7A
On Wed, 30 May 2001, Alan Hourihane wrote: On Wed, May 30, 2001 at 04:59:44PM +0300, Pasi Kärkkäinen wrote: On Wed, 30 May 2001, Alan Hourihane wrote: When I do that, the whole box crashes immediately after I press enter..:( even sysrq wont work anymore.. O.k. Can you check your BIOS for something called AGP Driving Value ? usually in Chipset Features Setup. Set this to 0xFF. Also check for another option called AGP Driving Control and make sure it's set to MANUAL (the other option is AUTO). Then it allows you to enter a hex digit. Use the following... If yours is a KT133 board try 0xE7, if it's a Apollo Pro 133A try 0xDA. See if that helps. Tried 0xFF, 0xE7 and 0xDA. Crashes with all of them. MB has KT133A-chipset (Abit KT7A). Can you try adding this (after you've done them changes) to your XF86Config file in the Device section. Option AGPMode 4 Didn't help. Still crashes. I've got Fail Safe Defaults in BIOS, plus the changes above, and 4x agpmode in XF86Config. Any other ideas? (Reading older posts, if the Rate parameter is set to zero, people noticed crashes.. I'm trying to change it). - Pasi Kärkkäinen ^ . . Linux /-\ Choice.of.the .Next.Generation. ___ Dri-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] Radeon on KT7A
On Wed, 30 May 2001, Alan Hourihane wrote: When I do that, the whole box crashes immediately after I press enter..:( even sysrq wont work anymore.. O.k. Can you check your BIOS for something called AGP Driving Value ? usually in Chipset Features Setup. Set this to 0xFF. Also check for another option called AGP Driving Control and make sure it's set to MANUAL (the other option is AUTO). Then it allows you to enter a hex digit. Use the following... If yours is a KT133 board try 0xE7, if it's a Apollo Pro 133A try 0xDA. See if that helps. Tried 0xFF, 0xE7 and 0xDA. Crashes with all of them. MB has KT133A-chipset (Abit KT7A). - Pasi Kärkkäinen ^ . . Linux /-\ Choice.of.the .Next.Generation. ___ Dri-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] Radeon on KT7A
On Wed, 30 May 2001, Alan Hourihane wrote: I replaced -s 0:0.0 with the address of Radeon, and tried both of those.. didn't help. Still crashes in the same way as without that setpci-setting.. Any other ideas? No. No. That's exactly why I asked for the lspci -vv output to check where things where. Please use -s 0:0.0 OK. When I do that, the whole box crashes immediately after I press enter..:( even sysrq wont work anymore.. - Pasi Kärkkäinen ^ . . Linux /-\ Choice.of.the .Next.Generation. ___ Dri-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] Glide warning/error and bug
On Mon, 28 May 2001, Marc Aurele La France wrote:
Right. makedepend doesn't (yet) deal with parameterised macros. If
things still build, makedepend's complaints should be ignored (for now).
I happened to have a /usr/include/glide3 directory populated by
older Glide3 headers, which made me believe that the build
process was ignoring my explicit settings in the spec file:
%if %{BuildTdfx}
#define HasGlide3 YES
#define Glide3LibName glide3
#define Glide3IncDir$RPM_BUILD_ROOT/usr/include/glide3
%endif
Examining the host.def that gets written out on a build at home
shows:
#define HasGlide3 YES
#define Glide3LibName glide3
#define Glide3IncDir
/home/mharris/rpmbuild/tmp/XFree86-4.0.99.900-root/usr/include/glide3
Should I be setting Glide3IncDir using relative symlinks instead?
Or is Glide3IncDir used in multiple places that would prevent
the using of relative links? The above should work nonetheless.
It appears however that it only looks in Glide3IncDir if the
headers aren't found somewhere else first, like /usr/include/glide3
Bug?
In makedepend's search order? Possibly. I'll have a look.
Well, I've checked this out and I don't any way /usr/include/glide3/glide.h
could have been included without a -I/usr/include/glide3 flag to
makedepend. Check the makedepend command line that was invoked.
Marc.
+--+---+
| Marc Aurele La France | work: 1-780-492-9310 |
| Computing and Network Services | fax:1-780-492-1729 |
| 352 General Services Building | email: [EMAIL PROTECTED] |
| University of Alberta +---+
| Edmonton, Alberta | |
| T6G 2H1 | Standard disclaimers apply|
| CANADA | |
+--+---+
XFree86 Core Team member. ATI driver and X server internals.
___
Dri-devel mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/lists/listinfo/dri-devel
