Re: 4.3.0 i810_drv i845G lockup when switching VTs while Server is busy

2003-04-05 Thread Enrik Berkhan
On Fri, Apr 04, 2003 at 01:00:00PM +, Martin Spott wrote:
 Enrik Berkhan [EMAIL PROTECTED] wrote:
  we have some compact PCs with integrated i845G graphics.
 Is it possible to get details on the hardware you use ? We're currently
 rolling out some 'compact' PCs from MSI with the same chipset and I would
 love to know what I have to fear _before_ it happens 

Actually, ours are MSIs. Something from the Hermes barebone series. I
will post the exact model when back at work on monday.

Enrik


smime.p7s
Description: S/MIME cryptographic signature


4.3.0 i810_drv i845G lockup when switching VTs while Server is busy

2003-04-04 Thread Enrik Berkhan
Hi,

we have some compact PCs with integrated i845G graphics. When running
the 4.3.0 server, it might lockup when switching VTs. A reliable way to
reproduce the behaviour is

# X :0 
# xterm -display :0 
# switch to X VT
in xterm:
# startkde
# switch back to inital vt while startkde is whacking the server

I have built i810_drv with some debugging enabled. A log is attached to
this mail. (I used the source for pre-liminary 4.3.0 debian packages
from [EMAIL PROTECTED] and have verified that the i810
directory doesn't differ from xfree86 cvs current.)

This has been tested on Linux only. Please tell me if you need more
specific information that I haven't provided so far.

Any ideas?

Thanks,
Enrik


smime.p7s
Description: S/MIME cryptographic signature


Re: 4.3.0 i810_drv i845G lockup when switching VTs while Server is busy

2003-04-04 Thread Enrik Berkhan
Hi,

On Fri, Apr 04, 2003 at 12:09:10PM +0200, Enrik Berkhan wrote:
A log is attached to
 this mail.

Grmpf, never never never write this before actually attaching.

Sorry,
Enrik


200304041108.log.gz
Description: Binary data


smime.p7s
Description: S/MIME cryptographic signature


nasty little bug in radeon driver

2003-03-07 Thread Enrik Berkhan
Hi,

after having some problems with an R300 based ATI card (dotclocks seemed
to be too slowed by a factor of about 1.17, text mode restoral didn't
work), I've found the following:

Index: radeon_reg.h
===
RCS file: /cvs/xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_reg.h,v
retrieving revision 1.25
diff -u -r1.25 radeon_reg.h
--- radeon_reg.h2003/02/07 18:08:59 1.25
+++ radeon_reg.h2003/03/07 08:34:09
@@ -879,7 +879,7 @@
 #   define RADEON_P2PLL_REF_DIV_MASK0x03ff
 #   define RADEON_P2PLL_ATOMIC_UPDATE_R (1  15) /* same as _W */
 #   define RADEON_P2PLL_ATOMIC_UPDATE_W (1  15) /* same as _R */
-#   define R300_PPLL_REF_DIV_ACC_MASK   (0x3ff  18)
+#   define R300_PPLL_REF_DIV_ACC_MASK   (0x3ff  18)
 #   define R300_PPLL_REF_DIV_ACC_SHIFT  18
 #define RADEON_PALETTE_DATA 0x00b4
 #define RADEON_PALETTE_30_DATA  0x00b8

Now it works for me.

Cheers,
Enrik
___
Devel mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/devel