Bug 7559

Summary:

radeon x700/acer ferrari 4000 anything other than LVDS,None has issues

Product:

xorg

Reporter:

aaron.dewell

Component:

Driver/Radeon

Assignee:

Xorg Project Team <xorg-team>

Status:

RESOLVED DUPLICATE

QA Contact:

Severity:

normal

Priority:

high

CC:

alan.coopersmith, alexdeucher, henry.zhao

Version:

7.0.0

Hardware:

x86 (IA32)

OS:

Linux (All)

Whiteboard:

i915 platform:

i915 features:

Attachments:

Description	Flags
xorg.conf from attempt 1	none
Xorg.0.log from attempt 1	none
xorg.conf from attempt 2	none
Xorg.0.log from attempt 2	none
xorg.conf from attempt 3	none
Xorg.0.log from attempt 3	none
xorg.conf from attempt 4	none
Xorg.0.log from attempt 4	none
xorg.conf from attempt 5	none
Xorg.0.log from attempt 5	none
xorg.conf from attempt 6	none
Xorg.0.log from attempt 6	none
xorg.conf from attempt 7	none
Xorg.0.log from attempt 7	none
xorg.conf from attempt 8	none
Xorg.0.log from attempt 8	none
requested rom file	none
unpatched log 6.6.0 driver	none
patch #1 log driver version 6.6.0	none
attempt to fix the ATOM bios parser	none

Description aaron.dewell 2006-07-18 08:50:10 UTC

I have tried various options to use a TMDS (DVI) panel, but having no success
(except at the internal panel's highest resolution, even that is weird, see
below).  Here is a summary of what I have tried:

Attempt #1:  TMDS,None
Result #1:   wobbly on internal panel, nothing on external panel (cannot display
this mode)

Attempt #2:  TMDS,LVDS (not expected to work at this time, but included for
completeness)
Result #2:   same as #1, with an added green bar about 2/3 of the way down the
screen

The green bar might have been there on Attempt #1, I didn't note it, however.

Attempt #3:  LVDS,TMDS (also not expected to work at this time)
Result #3:   No screens found, Screen 0 deleted due to no matching config section

Attempt #4:  CRT,LVDS
Result #4:   black on both internal and external panels

Note:  The external panel is connected via both VGA and DVI.  It requires a
button press on the panel to change, and I checked both for all of these tests.

Attempt #5:  CRT,None
Result #5:   black on both internal and external panels

Attempt #6:  LVDS,CRT
Result #6:   Starts, normal display at low resolution on internal panel.  CRT
has signal, but is black.

Attempt #7:  TMDS,CRT
Result #7:   wobbly with green bar on internal panel.  CRT has signal, but is
black.  DVI has no signal.

Attempt #8:  TMDS,None with reduced resolution (removed 1920x1200)
Result #8:   Internal panel:  rolling upwards and squished into 8 or 10 copies.
 warped on the right side.  External panel:  works normally.

Comment 1 aaron.dewell 2006-07-18 08:51:18 UTC

Created attachment 6254 [details]
xorg.conf from attempt 1

Comment 2 aaron.dewell 2006-07-18 08:51:43 UTC

Created attachment 6255 [details]
Xorg.0.log from attempt 1

Comment 3 aaron.dewell 2006-07-18 08:53:08 UTC

Created attachment 6256 [details]
xorg.conf from attempt 2

Comment 4 aaron.dewell 2006-07-18 08:53:27 UTC

Created attachment 6257 [details]
Xorg.0.log from attempt 2

Comment 5 aaron.dewell 2006-07-18 08:54:01 UTC

Created attachment 6258 [details]
xorg.conf from attempt 3

Comment 6 aaron.dewell 2006-07-18 08:54:16 UTC

Created attachment 6259 [details]
Xorg.0.log from attempt 3

Comment 7 aaron.dewell 2006-07-18 08:55:21 UTC

Created attachment 6260 [details]
xorg.conf from attempt 4

Comment 8 aaron.dewell 2006-07-18 08:55:39 UTC

Created attachment 6261 [details]
Xorg.0.log from attempt 4

Comment 9 aaron.dewell 2006-07-18 08:56:26 UTC

Created attachment 6262 [details]
xorg.conf from attempt 5

Comment 10 aaron.dewell 2006-07-18 08:56:42 UTC

Created attachment 6263 [details]
Xorg.0.log from attempt 5

Comment 11 aaron.dewell 2006-07-18 08:57:18 UTC

Created attachment 6264 [details]
xorg.conf from attempt 6

Comment 12 aaron.dewell 2006-07-18 08:57:35 UTC

Created attachment 6265 [details]
Xorg.0.log from attempt 6

Comment 13 aaron.dewell 2006-07-18 08:58:16 UTC

Created attachment 6266 [details]
xorg.conf from attempt 7

Comment 14 aaron.dewell 2006-07-18 08:58:37 UTC

Created attachment 6267 [details]
Xorg.0.log from attempt 7

Comment 15 aaron.dewell 2006-07-18 08:59:12 UTC

Created attachment 6268 [details]
xorg.conf from attempt 8

Comment 16 aaron.dewell 2006-07-18 08:59:31 UTC

Created attachment 6269 [details]
Xorg.0.log from attempt 8

Comment 17 aaron.dewell 2006-07-18 09:02:49 UTC

BTW, the ATI driver is 6.5.7.  I have updated that separately from X.org main
(which is 7.0.0 from debian unstable).

Comment 18 Alex Deucher 2006-07-18 09:47:12 UTC

It's currently not possible to use LVDS and TMDS at the same time so for now
you'll need to use either TMDS,NONE (for DVI) or LVDS,NONE (for internal panel).
 you might also want to force off the use of the second controller:
Option "MergedFB" "FALSE"

Comment 19 aaron.dewell 2006-07-18 22:01:13 UTC

No change in behavior with MergedFB False.

Comment 20 Dave Airlie 2006-10-09 13:47:53 UTC

can sosmeone email me the BIOS from one of those laptops?

cd /sys/bus/pci/devices/<pci id of card>

echo 1 > rom
cat rom > /tmp/acerrom
echo 0 > rom

I think the ATOM tables are garbage and I want to fix the driver to deal with it..

Comment 21 aaron.dewell 2006-10-10 10:09:52 UTC

Created attachment 7328 [details]
requested rom file

also emailed directly yesterday

Comment 22 henry.zhao@oracle.com 2006-10-18 16:21:14 UTC

Refer [Bug 5473], and try the latest patch from there without using a config file.

Comment 23 aaron.dewell 2006-10-26 07:57:14 UTC

With the latest version (6.6.0), downloaded 10/18, unpatched, works much better
without the panel connected.  (I am traveling with no access to my external
display right now.)  "works much better" in this case defined as that I can have
the right configuration (with LVDS,TMDS, and don't have to comment out the panel
portion).  Log attached.

Patched with patch #1, pretty much the same result.  Again, no panel to test
with until next weekend.  Log attached.

Patch #2: patches src/radeon_display.c, and I have no radeon_display.c in 6.6.0.
 Not sure if the file was renamed or what.

I will try to borrow an external display where I'm working to test before.

I ran into another issue during testing:  the keyboard only works when X is
first started.  If I start X a second time without rebooting, all keys act as
mode switch, changing resolution.  Makes it rather hard to use. :)  Console acts
normally.  reboot fixes it, but I'd like a less drastic solution.

Still trying to get DRI/DRM working, but that's probably a kernel issue.

Comment 24 aaron.dewell 2006-10-26 07:58:36 UTC

Created attachment 7532 [details]
unpatched log 6.6.0 driver

Comment 25 aaron.dewell 2006-10-26 07:59:30 UTC

Created attachment 7533 [details]
patch #1 log driver version 6.6.0

Comment 26 henry.zhao@oracle.com 2006-10-26 14:18:11 UTC

(In reply to comment #23)
> With the latest version (6.6.0), downloaded 10/18, unpatched, works much better
> without the panel connected.  (I am traveling with no access to my external
> display right now.)  "works much better" in this case defined as that I can have
> the right configuration (with LVDS,TMDS, and don't have to comment out the panel
> portion).  Log attached.
> 
> Patched with patch #1, pretty much the same result.  Again, no panel to test
> with until next weekend.  Log attached.
> 
> Patch #2: patches src/radeon_display.c, and I have no radeon_display.c in 6.6.0.
>  Not sure if the file was renamed or what.
> 
> I will try to borrow an external display where I'm working to test before.
> 
> I ran into another issue during testing:  the keyboard only works when X is
> first started.  If I start X a second time without rebooting, all keys act as
> mode switch, changing resolution.  Makes it rather hard to use. :)  Console acts
> normally.  reboot fixes it, but I'd like a less drastic solution.
> 
> Still trying to get DRI/DRM working, but that's probably a kernel issue.

(In reply to comment #23)
> With the latest version (6.6.0), downloaded 10/18, unpatched, works much better
> without the panel connected.  (I am traveling with no access to my external
> display right now.)  "works much better" in this case defined as that I can have
> the right configuration (with LVDS,TMDS, and don't have to comment out the panel
> portion).  Log attached.
> 
> Patched with patch #1, pretty much the same result.  Again, no panel to test
> with until next weekend.  Log attached.
> 
> Patch #2: patches src/radeon_display.c, and I have no radeon_display.c in 6.6.0.
>  Not sure if the file was renamed or what.
> 
> I will try to borrow an external display where I'm working to test before.
> 
> I ran into another issue during testing:  the keyboard only works when X is
> first started.  If I start X a second time without rebooting, all keys act as
> mode switch, changing resolution.  Makes it rather hard to use. :)  Console acts
> normally.  reboot fixes it, but I'd like a less drastic solution.
> 
> Still trying to get DRI/DRM working, but that's probably a kernel issue.

(1) In the patches you referred here, is patch #1 attach #5858 and patch #2
    attachement #7409 ? Attachment #7409 [details] uses latest modular source structure,
    where radeon_display.c is splitted out from radeon_driver.c. It has a 
    workaround for BIOS parsing error. Please try it.

(2) Please also try without a config file, to see if you can get a satisfactory
    result.

Comment 27 Dave Airlie 2006-11-13 21:35:52 UTC

Created attachment 7781 [details] [review]
attempt to fix the ATOM bios parser

this is a completely hacked up idea to fix the ATOM bios parser...

I've no idea if it is more or less correct than what is there..

can you test this please?

Comment 28 henry.zhao@oracle.com 2006-11-14 19:14:55 UTC

(In reply to comment #27)
> Created an attachment (id=7781) [edit]
> attempt to fix the ATOM bios parser
> 
> this is a completely hacked up idea to fix the ATOM bios parser...
> 
> I've no idea if it is more or less correct than what is there..
> 
> can you test this please?

I tried the patch on Ferrari 4000 with an external monitor connected, and found
out, DDCType for LCD is now detected, but incorrectly as 3 (DDC_VGA). Without
the patch, it is detected as 0. Detailed results:

Without patch:
port 0: DDCType, DACType, TMDSType = 0, 0, -1
port 1: DDCType, DACType, TMDSType = 0, -1, -1

With patch:
port 0: DDCType, DACType, TMDSType = 3, 0, 0
port 1: DDCType, DACType, TMDSType = 0, -1, -1

Correct or default setting (should be):
port 0: DDCType, DACType, TMDSType = 2, 1, 0
port 1: DDCType, DACType, TMDSType = 3, 0, 1

Comment 29 Dave Airlie 2006-11-15 00:54:07 UTC


*** This bug has been marked as a duplicate of 5473 ***

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.