Bug 60987

Summary: intel driver not working with 915GM/GMA900: failed to get resources: Invalid argument
Product: xorg Reporter: Harald Rudell <harald>
Component: Driver/intelAssignee: Chris Wilson <chris>
Status: RESOLVED NOTABUG QA Contact: Intel GFX Bugs mailing list <intel-gfx-bugs>
Severity: blocker    
Priority: medium    
Version: unspecified   
Hardware: x86 (IA32)   
OS: All   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
dmesg
none
xorg.0.log none

Description Harald Rudell 2013-02-17 03:39:25 UTC 
1. Ever since kms, the X display driver "intel" has not worked with my 915GM/GMA900.

2. Symptom is a message "failed to get resources: Invalid argument"
I remember the first time I saw this when I was forced onto the new kms intel driver, and I have been on the "vesa" driver ever since.

3. The chip is on the list the intel driver claims to support.
   The chip is not on the list of hardware the intel driver is verified with

4. I have the latest quantal-proposed kernel, dated 130207:
linux-image-extra-3.5.0-24-generic

5. X video driver: "intel" module version = 2.20.9

6. kernel kms active

7. hardware: lspci -k -nn -s 0:2
00:02.0 VGA compatible controller [0300]: Intel Corporation Mobile 915GM/GMS/910GML Express Graphics Controller [8086:2592] (rev 04)
 Subsystem: Holco Enterprise Co, Ltd/Shuttle Computer Device [1297:3041]
 Kernel modules: intelfb, i915
00:02.1 Display controller [0380]: Intel Corporation Mobile 915GM/GMS/910GML Express Graphics Controller [8086:2792] (rev 04)
 Subsystem: Holco Enterprise Co, Ltd/Shuttle Computer Device [1297:3041]

8. end of Xorg.0.log:
(everything dandy above here..., looks line-by-line like my other Intel graphics machine)
[  3295.346] (**) intel(0): SwapBuffers wait enabled
[  3295.346] (==) intel(0): video overlay key set to 0x101fe
[  3295.346] (EE) intel(0): failed to get resources: Invalid argument
[  3295.346] (II) UnloadModule: "intel"
[  3295.346] (EE) Screen(s) found, but none have a usable configuration.
[  3295.346] 
Fatal server error:
[  3295.346] no screens found
[  3295.346] (EE) 

9. Back in 2007, this worked! Here is what I could get out of randr then:
# Output TMDS-1 (DVI-x connector)
# Output VGA (VGA connector)
# Output LVDS (internal)
# Output TV disconnected
# 2 CRTCs, ie. 2 simultanous images
Comment 1 Chris Wilson 2013-02-17 09:35:05 UTC 
EINVAL here should only be returned if !KMS. So please attach both the full drm.debug=6 dmesg from boot and the full Xorg.0.log with the failure.
Comment 2 Harald Rudell 2013-02-17 15:44:00 UTC 
Created attachment 74987 [details]
dmesg
Comment 3 Harald Rudell 2013-02-17 15:45:29 UTC 
Created attachment 74988 [details]
xorg.0.log
Comment 4 Harald Rudell 2013-02-17 15:50:11 UTC 
Attached files are for boot with kms to X Fatal error

Hopefully they will provide clues to what is amiss?

What I did:

On boot, at grub menu, press e
append to the linuz line: drm.debug=6
press F10

The machine ends at a black screen but is otherwise functional, ie. ssh works great.

Thank you.
Comment 5 Daniel Vetter 2013-02-17 15:59:24 UTC 
You have kernel modesetting disabled and your version of the intel ddx does not support usermodesetting any more. Please remove any i915.modset=0 or nomodeset boot options. If kernelmodesetting doesn't work for you, then we need to fix that (please reopen in that case and provide dmesg with drm.debug=0xe and Xorg.log).
Comment 6 Harald Rudell 2013-02-17 16:22:38 UTC 
Thank you,

That's kind of correct. I had forgot to run

sudo update-initramfs -u

to propagate changes to the boot image. The boot somehow entered the pretty purple 160-column mode either way, but the X driver would not start.

It works now!
Comment 7 Daniel Vetter 2013-02-17 16:25:48 UTC 
Cool, thanks for reporting back, and I'm happy that kms works ;-)
Comment 8 Chris Wilson 2013-02-17 18:21:03 UTC 
Also note that I pushed

commit 43ba22ef4a4142f334e9ae2d926250988ecbe8bc
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Sun Feb 17 09:25:38 2013 +0000

    Confirm that i915.ko has KMS enabled before binding to the device

to the ddx so that we fail more gracefully here.

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.