Bug 39216 - [G45] SDVO_LVDS can't be lighted
Summary: [G45] SDVO_LVDS can't be lighted
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: XOrg git
Hardware: All Linux (All)
: medium normal
Assignee: Chris Wilson
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2011-07-13 23:23 UTC by bo.b.wang
Modified: 2017-09-04 10:07 UTC (History)
4 users (show)

See Also:
i915 platform:
i915 features:


Attachments
drm.debug=0xe dmesg info (73.41 KB, text/plain)
2011-07-13 23:23 UTC, bo.b.wang
no flags Details
LVDS panels are digital as well! (3.05 KB, patch)
2011-07-14 01:28 UTC, Chris Wilson
no flags Details | Splinter Review
boot info with patch (70.33 KB, text/plain)
2011-07-14 19:02 UTC, bo.b.wang
no flags Details
after done testdisplay info with patch (86.49 KB, text/plain)
2011-07-14 19:03 UTC, bo.b.wang
no flags Details
dump info with testdisplay -i (515 bytes, text/plain)
2011-07-14 19:04 UTC, bo.b.wang
no flags Details

Description bo.b.wang 2011-07-13 23:23:48 UTC
Created attachment 49068 [details]
drm.debug=0xe dmesg info

System Environment:
--------------------------
Kernel:	(drm-intel-fixes) 1083694ab052e5ba38a8f9a057c4745448977837
port: SDVO_LVDS 
platform: G45

Bug detailed description:
-------------------------
Booting G45 machine with SDVO_LVDS port, it is normal in bios and bootloader stage, After the system has started all the services and switched into console mode, the LVDS can't be lighted but the backlight is on. And I use testdisplay -i to dump info ,it is all disconnected. 
Attachment is the drm.debug info

Reproduce steps:
----------------
1. SDVO_LVDS
2. testdisplay -i
Comment 1 Chris Wilson 2011-07-14 01:28:25 UTC
Created attachment 49069 [details] [review]
LVDS panels are digital as well!

Another patch that slipped through the cracks...
Comment 2 bo.b.wang 2011-07-14 19:02:02 UTC
Created attachment 49112 [details]
boot info with patch
Comment 3 bo.b.wang 2011-07-14 19:03:43 UTC
Created attachment 49113 [details]
after done testdisplay info with patch
Comment 4 bo.b.wang 2011-07-14 19:04:54 UTC
Created attachment 49114 [details]
dump info with testdisplay -i
Comment 5 bo.b.wang 2011-07-14 19:05:26 UTC
(In reply to comment #1)
> Created an attachment (id=49069) [details]
> LVDS panels are digital as well!
> 
> Another patch that slipped through the cracks...

Hello Chris.I have tried your patch. Sometimes(sometimes SDVO_LVDS can't be lighted and backlight is off except I powoff the machine for a while) the SDVO_LVDS can work well with your patch. after I have done the preferred mode testing  with testdisplay command  in the console ,  the backlight is off , SDVO_LVDS can't be lighted ,but testdisplay can detect the SDVO_LVDS info. 
Attachment is the dmesg info
Comment 6 Chris Wilson 2011-07-16 04:55:00 UTC
Right the patch fixes a false negative for detection, but it sounds like we have the usual testdisplay issues.
Comment 7 Jesse Barnes 2011-08-01 11:17:25 UTC
I guess testdisplay needs backlight support?  If you modify the backlight by directly echoing values into /sys/class/backlight/.../brightness do things work?
Comment 8 bo.b.wang 2011-08-09 00:29:40 UTC
(In reply to comment #7)
> I guess testdisplay needs backlight support?  If you modify the backlight by
> directly echoing values into /sys/class/backlight/.../brightness do things
> work?

hello Jesse. There is nothing in /sys/class/backlight when it is just using SDVO_LVDS port
Comment 9 bo.b.wang 2011-08-11 23:23:13 UTC
Kernel: (drm-intel-fixes)d7d39e20cc5e44f915d6c5687f65839632c44381
Some additional commit info:
Author: Jesse Barnes <jbarnes@virtuousgeek.org>
Date:   Wed Aug 3 12:59:21 2011 -0700

    drm/i915: enable/disable spread spectrum as needed on mode set

this issue also exists in this commit
Comment 10 Georg Grabler 2011-09-05 02:56:41 UTC
Hi guys,

As I get it, the patch did not completely fix the problem. From a distribution side (Chakra), I'd like to ask the following:
- In which repository / branch are the patches?
- Does the last comment mean this code has the same problem or fixes it?
- If it's solved (it obviously isn't, our users still report this in our latest testing release with 3.0.4), will we see a backport to intel-fixes "for linus" or do we have to do that on our own or wait for 3.1?
- Are there any traces needed we could ask our users for if it's reported (I don't have a backlight-lid display to test it myself, sorry).

For the workaround (setpci -s 00:02.0 f4.b=00)
Do we have any chance to "check" system-side if this problem exists for the specified card? We could just add a rc-file as workaround if necessary from the distribution side.
Comment 11 Chris Wilson 2011-12-19 16:11:08 UTC
Keith finally sent the patch upstream in

commit 2cfab8d74ebfbe06ce2947117945c4f45a5915ec
Merge: 5885b9b 3b5c78a
Author: Linus Torvalds <torvalds@linux-foundation.org>
Date:   Fri Dec 16 11:27:56 2011 -0800

    Merge branch 'drm-intel-fixes' of git://people.freedesktop.org/~keithp/linux

a mere 6 months after being sent the regression fix. This should fix the original problem, anything that remains (such as misconfigured ACPI) should be addressed elsewhere.
Comment 12 bo.b.wang 2011-12-19 18:34:24 UTC
(In reply to comment #11)
> Keith finally sent the patch upstream in
> 
> commit 2cfab8d74ebfbe06ce2947117945c4f45a5915ec
> Merge: 5885b9b 3b5c78a
> Author: Linus Torvalds <torvalds@linux-foundation.org>
> Date:   Fri Dec 16 11:27:56 2011 -0800
> 
>     Merge branch 'drm-intel-fixes' of
> git://people.freedesktop.org/~keithp/linux
> 
> a mere 6 months after being sent the regression fix. This should fix the
> original problem, anything that remains (such as misconfigured ACPI) should be
> addressed elsewhere.


Hello. This patch has fixed the issue.
Tested-by: Wang Bo <bo.b.wang@intel.com>
Comment 13 Florian Mickler 2012-01-12 14:17:29 UTC
A patch referencing this bug report has been merged in Linux v3.2-rc6:

commit 522200858093a6f31af9830672109f6d9807dd1e
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Mon Jun 20 14:45:50 2011 +0100

    drm/i915/sdvo: Include LVDS panels for the IS_DIGITAL check
Comment 14 Jari Tahvanainen 2017-09-04 10:07:16 UTC
Closing old verified+fixed.


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.