Intel Extreme Driver Does Not Work With 945G - GMA950

[tonestone57]

Haiku freezes when loading the Intel video driver and does not boot. I checked the vendor id and my video card should be supported by the driver.

Seems to mention an error with PLL. Required 25.17 : Found 25.1429 Mhz. Will attach screenshot & syslog.

From listdev:

device Display controller (VGA compatible controller, VGA controller) [3|0|0]
  vendor 8086: Intel Corporation
  device 2772: 82945G/GZ Integrated Graphics Controller

[tonestone57]

PLL-limits

[tonestone57]

syslog

[siarzhuk]

Just a bit of additional debug info acquired on the same hardware:

PLL limits, min: p 5 (p1 1, p2 10), n 5, m 70 (m1 12, m2 7) PLL limits, max: p 80 (p1 8, p2 5), n 10, m 120 (m1 22, m2 11) required MHz: 108 found: 107.755 MHz, p = 14 (p1 = 1, p2 = 14), n = 7, m = 110 (m1 = 20, m2 = 10) PLL limits, min: p 5 (p1 1, p2 10), n 5, m 70 (m1 12, m2 7) PLL limits, max: p 80 (p1 8, p2 5), n 10, m 120 (m1 22, m2 11) required MHz: 108 found: 107.52 MHz, p = 20 (p1 = 2, p2 = 10), n = 5, m = 112 (m1 = 21, m2 = 7)

Complete syslog.15 is attached.

[siarzhuk]

Same problem on different PC

[axeld]

(In #2540) Okay, thanks for the note!

[zaunpfahl]

Confirmed with 945GT, Abit iL-90mv. Save mode works.

[tonestone57]

EXTRA INFO:

I noticed two things of interest while looking over David's similar issue on mailing list  http://www.freelists.org/post/haiku-development/Asus-Eee-PC-B202,20

Refer to my screenshot. 1) You will notice that p2 is out of range on the first attempt. X Org driver has updated p2 values to 14 and 7. 2nd attempt uses changed PLL limits which correspond with driver but fails to boot ( black screen ). 2) My frequency is ~25 Mhz whereas siarzhuk & David both get ~108 Mhz. Is my frequency being divided by 4? ie: should it not be 100 Mhz instead?

[tonestone57]

(In #3464) Fairly sure this bug is a duplicate of (#3149)

[axeld]

(In #5530) Nice to look up the old duplicates - but it certainly doesn't make sense to add another one.

[axeld]

I am not able/willing to fix this bug without having access to the hardware.

[tonestone57]

I have not checked any changes you may have done to intel_extreme driver for very long time. I will try to test again within next 3 weeks and update ticket information. Please wait for my update first!

Germany is too far to send you my hardware. :-)

This is from Intel (manufacturer & chipset) ATOM 330 desktop board. Since lots of people are having issue with this driver & on-board video (I think from this particular mobo - with later version of i945 graphics card), maybe you can get Haiku Inc. to pay to get you this motherboard ($100-$120)? Or maybe there is someone you know with ATOM 330 board (or similar mobo with this issue) that could lend it to you?

[Michael S.]

Replying to axeld:

I am not able/willing to fix this bug without having access to the hardware.

I buy on ebay MOBO with the necessary integrated videocard and I specify you quality of the buyer. Such variant will approach?

P.S. The bug #5530 has been created to pay attention to this small group of bugs.

[axeld]

It seems like there are two different issues with this hardware; there was a possible hang if the hardware detected an enabled head B without a LVDS panel. This one issue has been fixed in hrev36941. This should allow some of the affected system/monitor combinations to work now.

[Michael S.]

Many thanks! With D-Sub-VGA TFT-matrix driver works now.

[tonestone57]

Thank you Axel. D-Sub VGA on LCD monitor works with Intel Extreme driver now. Tested with hrev36941.

~> listimage | grep accelerant
  793        /boot/system/add-ons/accelerants/intel_extreme.accelerant 0x8d8000 0x8e4000    0          0
TEAM 3291 (/bin/grep accelerant):

[Michael S.]

hrev37003 KDL on booting

[mr.Noisy]

GCC4 build hrev37641 doesn't work. GCC2 hrev37739 just fine.

[umccullough]

What combination exactly is still broken here? LVDS/DVI output?

I seem to recall that my particular Atom 330 box with this chipset does work now with analog VGA (it didn't have DVI output).

[umccullough]

Please retest - I suspect this issue may have since been resolved. I'll give it a few more months, then probably close it since a lot of changes have been made to the intel_extreme driver in 3 years.