|
Description
Michael Fu
2009-03-11 20:31:32 UTC
Let's use this instead of bug# 17823 to track this sdvo vga dual head issue. (change the topic) I never have a SDVO VGA card to test, do we have any ADD2 card with VGA? Created attachment 23784 [details]
xorg.conf
Created attachment 23785 [details]
xorg.log
(EE) intel(0): First SDVOB output reported failure to sync No EDID data from SDVO VGA. Created attachment 23788 [details] [review] Reset states before new output setup Please try this patch against master. It fixes the missing state reset before output setup, otherwise we might take wrong mode setting operations later. Pushed this patch. Please try current git master. Shih Ming， are you able to verify if the patch fix your dual display issue ( no compiz, please ). Should be fixed in git master. Reopen if it's not. commit dc3ff0b514b609448025680778f0e95e1980a5d8 Author: Zhenyu Wang <zhenyu.z.wang@intel.com> Date: Thu Mar 12 16:32:02 2009 +0800 Revert "SDVO: Switch control bus only before DDC access" This reverts commit ddedf19f889da2ce6d69a3afce4665e2245682fa. After i2c STOP, control bus will return back to internal registers. So this brings back to origin code that we switch to DDC bus before START. But it's ideal to only issue DDC bus switch after STOP, not before every START, which might eliminate some complains from SDVO device, that will be another patch later. commit c6b0135d209bdad3dbc641d0e264596eaf6f99d3 Author: Zhenyu Wang <zhenyu.z.wang@intel.com> Date: Thu Mar 12 16:31:01 2009 +0800 SDVO: reset privates before output setup For multifunction encoder, forget to reset these values will cause wrong output type in later mode setting operations. Created attachment 23898 [details] log file after patching reset sdvo I had collected the log file after using attachment 23788 [details] [review] patch on the git master driver. VGA-1 still display black screen. Looks like the behavior is still same, not sure if you have successfully applied my patch. As I have pushed it, please just git pull current git master. (Michael would suggest to add some debug output in sdvo detect and get_modes function, I believe. ;) (In reply to comment #11) > Looks like the behavior is still same, not sure if you have successfully > applied my patch. As I have pushed it, please just git pull current git master. > (Michael would suggest to add some debug output in sdvo detect and get_modes > function, I believe. ;) > Hmm...actually, I am thinking the same thing as you :) - Shih Ming maybe doesn't get c6b0135d209bdad3dbc641d0e264596eaf6f99d3. The trace shows that even after output is renamed VGA-1, get_modes still use the is_tv branch... -- (II) intel(0): Output VGA-1 using monitor section Monitor[2] (**) intel(0): Option "PreferredMode" "1024x768" (II) intel(0): SDVOB: W: 28 (SDVO_CMD_GET_TV_FORMAT) (II) intel(0): SDVOB: R: 01 00 00 00 00 00 (Success) (II) intel(0): SDVOB: W: 83 01 00 00 (SDVO_CMD_GET_SDTV_RESOLUTION_SUPPORT) (II) intel(0): SDVOB: R: FF FF 07 (Success) -- Could it be some configuration settings that I should include/exclude from the xorg.conf file? Anyway, I just recieved a new machine for testing. Will do a fresh install with SLES11 GM (without Compiz) and will pull from git master. Thanks (1) Install SLES11GMC to the terminal (2) downloaded the package from http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=dc3ff0b514b609448025680778f0e95e1980a5d8 (3) tar -xvzf xf86-video-intel-dc3ff0b514b609448025680778f0e95e1980a5d8.tar.gz (4)run into this error when execute ./autogen.sh put in the required package that prompt along the way from http://download.opensuse.org/factory/repo/oss/suse/i586/ And now see this error, could you let me know what other packages are needed : "checking for XORG... configure: error: Package requirements (xorg-server xproto fontsproto xineramaproto randrproto renderproto xextproto x11 xextproto) were not met: Package xcb-xlib was not found in the pkg-config search path. Perhaps you should add the directory containing `xcb-xlib.pc' to the PKG_CONFIG_PATH environment variable Package 'xcb-xlib', required by 'X11', not found" I guess you don't have the dev environment or something... Stefan or Quanxin? Guek, I suggest to contact Lew how to build intel driver on SLE11. Thanks. Created attachment 23985 [details]
xorg.conf, xorg.0.log and pictures of porblem when set to 1024x768
I managed to compile the intel driver after installing addition packages xorg.
with the attached xorg.conf that has resolution of 800x600, i am able to see standard display on both screen.
However will run into seeing in porportion display on screen2 when the xorg.conf is set to resolution of 1024x768. see attached picture.
In addition to this pls guide me on how to setup the xinerama and clone display for this intel card. a sample xorg.conf will definetly be helpful.
dont' attach zip, pls attach files one by one here... thanks. Created attachment 23988 [details]
xorg.conf that has 1024x768 setup
Created attachment 23989 [details]
Xorg.0.log collected for 1024x768 setup
Created attachment 23990 [details]
picture taken of screen 1 and 2 with 1024X768 setup
attached in comment#21, monitor on the left is connected to port 2. Monitor on the right is connected to port 1 Created attachment 23991 [details]
xorg.conf with 800X600 setup that display ok for standard setup.
Created attachment 23992 [details]
picture taken from xorg.conf that has 800x600 setup,
Picture shows that display output from port 1(screen on the Right) and 2(screen on the left) is ok with 800x600
Thanks for testing. It looks like native VGA can't get to use 1024x768 mode? And SDVO VGA has displayed 1024x768 correctly. Could you turn on ModeDebug option and send log for 1024x768 config? Created attachment 24026 [details]
Xorg.0.log1024x768 with Modebug on
Forget to turn on ModeDebug option? looks no debug output in log. Created attachment 24032 [details]
this the xorg.conf that i used. is with ModeDebug=yes and 1024x768
Add this line Option "ModeDebug" "yes" to "Device" section. Created attachment 24033 [details]
Xorg.0.log with ModeDebug and 1024x768.
here is the one. Previously attached at place the ModeDebug in wrong section. Sorry!
Try to remove HorizSync and VertRefresh setting in "Monitor[0]" for VGA, looks it filters out default 1024x768 mode. Ya removing the HorizSync and VertRefresh both monitors will turn out ok for standard display in both monitors connected to port 1 and 2 respectively. With this, meaning is going to me a manual workaround? can we have it auto? Section "Monitor" Option "CalcAlgorithm" "XServerPool" DisplaySize 345 259 #HorizSync 30-40 Identifier "Monitor[0]" ModelName "Unknown" Option "DPMS" Option "PreferredMode" "1024x768" VendorName "Unknown" #VertRefresh 50-75 UseModes "Modes[0]" EndSection You don't have reason to set them, actually it's your setting made default modes be filtered out, as it's not within the range in your sync setting. Remove them, and let driver probe the modes for you. Close this one. Fixed by commits note in comment 9. We have not resolve the Xinerama and clone view setting yet. so pls do not close the bug. Created attachment 24034 [details]
xorg.conf used to setup xinerama display on 2 screen.
I am seeing two blank blank screen when turn on xinerama for 2 screen display
Created attachment 24035 [details]
Xorg.0.log when turn on xinerama for 2screen
need to resolve the xinerama problem reported in comment#35 Why not using RandR? see intel and xorg.conf man page on how to config it, intel man should have a section on multihead configurations. As both outputs display fine, left problem is your own display config issue, so this bug should be closed. Please try to use randr to config your outputs. As origin SDVO VGA blank issue is fixed in master, close this one.
Here's an example to put SDVO-VGA on the right of VGA, it looks your origin xorg.conf is broken.
Section "Monitor"
Identifier "VGA"
Option "Position" "0 0"
EndSection
Section "Monitor"
Identifier "SDVO-VGA"
Option "RightOf" "VGA"
EndSection
Section "Device"
Identifier "Intel Driver"
Driver "intel"
Option "monitor-VGA" "VGA"
Option "monitor-VGA-1" "SDVO-VGA"
EndSection
I had got information from the http://intellinuxgraphics.org/dualhead.html for help on the set up on the dualhead configuration. Managed to setup the xinerama setup for 2 monitors. VGA connected to 1024X768 monitor and VGA-1 connected to 800x600 monitor. Xinerama is working well. I had the following question: (1) question on the 'Option "Position"'. Am i able to make use of the 'option "Position"' to achieve the screen layout to display portionally? how? Example in this case, i am not able acheive portionally view for the monitors which is 1024x768 and another 800x600. (2)How do i enable clone view? My understanding of clone view is that i should see same view including lauching bar on both screen. However i only see the lauching bar on monitor(1024x768) connected to VGA but did not see luanching bar on the monitor(800x600) connected to VGA-1. Created attachment 24153 [details]
xinerama setup.
xorg.conf used for xinerama ie extended dual display setup.
VGA=1024x768, VGA-1 =800x600
Created attachment 24155 [details]
this is the clone setup i used. didnot see the launching bar on both screen
Does this card support clone and Standard Twin view? If yes could you help to clarify on the setup for Clone and Twin view. Once this clarify, then we can proceed to close this bug. (In reply to comment #43) > Does this card support clone and Standard Twin view? If yes could you help to > clarify on the setup for Clone and Twin view. > > Once this clarify, then we can proceed to close this bug. > Guek Wu Neo, Novell people would be more exeprience than us on how to play with xorg.conf... Why would you use one monitor to be 1024x768 while the other to be 800x600? Can both set to be 800x600 for your usage senario? If that's acceptable, then your questions in comment# 40 are gone... Does this card support clone and Standard Twin view? In Intel definition, does clone configuration support different resolution of monitors for this particular intel card? And how is twinview different from clone? I have tested 2 same 1024x768 monitor as well. I got the xinerama setup and working fine. However i still not able to verify if the the clone and twin view is working as it supposed to be because i see no difference in clone and twinview. Thus i would appreciate if you can help to clarify my doubt on the clone and standard twin view setup. (In reply to comment #45) > Does this card support clone and Standard Twin view? > yes. > In Intel definition, does clone configuration support different resolution of > monitors for this particular intel card? > yes. > And how is twinview different from clone? no idea. > > I have tested 2 same 1024x768 monitor as well. I got the xinerama setup and > working fine. However i still not able to verify if the the clone and twin view > is working as it supposed to be because i see no difference in clone and > twinview. so pls close the bug. > > Thus i would appreciate if you can help to clarify my doubt on the clone and > standard twin view setup. > I had read from this chipset information from http://www.intel.com/Products/Notebook/Chipsets/915GM/915GM-overview.htm It states " The Mobile Intel® 915GM Express chipset is a component of the Intel® Centrino® processor technology. Featuring the Intel® Graphics Media Accelerator 900, the 915GM chipset enables .... " With the featuring of Intel® Graphics Media Accelerator 900 in this 915GM chipset, this terminal should have the Intel dual clone, Intel dual twin and Intel dual extended support as listed in http://www.intel.com/design/chipsets/applnots/30262403.pdf Thus, i look forward support to the clarification on the setup of clone and twinview as posted in comment#40 and 45. Guek, RANDR 1.2 setup and desktop issues are really not part of this bugreport. Use xrandr to do the RANDR 1.2 setup mirror: xrandr --output VGA --same-as LVDS side-by-side: xrandr --output --right-of LVDS See 'xrandr --help' and 'man xrandr' for more details. For a side-by-side setup you need to have specified an appropriate Virtual line in your xorg.conf (see 'man xorg.conf'). For 1024x768 and 800x600 it would be 'Virtual 1824 768'. SaX2 writes such a line by default, but I think you can't use SaX2 in this case. > side-by-side: xrandr --output --right-of LVDS
side-by-side: xrandr --output VGA --right-of LVDS
Guek, in your case you probably need to replace LVDS with VGA-1. But you can figure the output names with 'xrandr'. using xrandr only helps to tune the monitors to output per what is specified under the command. The monitor output is gone once is the terminal reboot. How can get setup the clone or twin or extended setup in xorg.conf? Also i did not see the the launching bar VGA-1 which is 800x600 when set to 'xrandr --output VGA --same-as VGA-1' Stefan since you mentioned desktop setup in not under this bug. could you ask advice me on the setup in Novell bugzilla #445998? (In reply to comment #51) > using xrandr only helps to tune the monitors to output per what is specified > under the command. The monitor output is gone once is the terminal reboot. It's for runtime change, yes. But apparently gnome-display-properties (GNOME RANDR 1.2 tool) notice the RANDR 1.2 settings and apply it again the next time you login. If this really works reliably is another question. Honestly I always had issues with that. > How can get setup the clone or twin or extended setup in xorg.conf? Let's do this in Novell's bugzilla. > Also i did not see the the launching bar VGA-1 which is 800x600 when set to > 'xrandr --output VGA --same-as VGA-1' Sure, when you have 1024x768 + 800x600 in mirrored mode there's always something missing, right? Move the 'launching bar' to another position or use an offset for the second monitor. You need to resize it as well. > since you mentioned desktop setup in not under this bug. could you ask advice > me on the setup in Novell bugzilla #445998? I'll try. (In reply to comment #51) > using xrandr only helps to tune the monitors to output per what is specified > under the command. The monitor output is gone once is the terminal reboot. >It's for runtime change, yes. But apparently gnome-display-properties (GNOME >RANDR 1.2 tool) notice the RANDR 1.2 settings and apply it again >the next time you login. If this really works reliably is another question. >Honestly I always had issues with that. so you are suggesting to use the GNOME RANDR 1.2 tool and if it does not raise separate bug for it? Using gnome-display-properties would be the easiest way, yes. by the way, to confirm 'gnome-display-properties' that you refering to is the monitor icon on the launching bar with half side yellow that say 'configure display setting' when point the cursor to it. am i right? Yes, I think so. thanks. Stefan. Fixes are in comment# 9. Intel, Pls let stefan knows where to get the package. He would need a patch which fixes this issue only. This the reply from Stefan in Novell Bugzilla: "Package included? I would need a patch, which fixes only that issue." (In reply to comment #59) > comment 9 The second commit doesn't apply to xf86-video-intel 2.5.0. I'm not sure how to adjust it. Can you help here? commit c6b0135d209bdad3dbc641d0e264596eaf6f99d3 Author: Zhenyu Wang <zhenyu.z.wang@intel.com> Date: Thu Mar 12 16:31:01 2009 +0800 SDVO: reset privates before output setup For multifunction encoder, forget to reset these values will cause wrong output type in later mode setting operations. Without SDVO multifunction device patch, you can't apply that to 2.5, so you have to backport SDVO multifunction patch too, commit 0d20bbbc2005a51f427a9ae6b6a66dbbb101dbab (In reply to comment #61) > Without SDVO multifunction device patch, you can't apply that to 2.5, so you > have to backport SDVO multifunction patch too, commit > 0d20bbbc2005a51f427a9ae6b6a66dbbb101dbab For which you also need 37c67088a887e6380571e6eec8a8f058e3e24717. Hi Zhenyu, Does this driver supports ACPI in particular to S3. I have problem putting the terminal to suspend which was OK before the patches (fresh install of SLED). Thanks I had reverified the S3 support on Sles11GM with this same patch. Is able to resume from S3 on both pri and secondary monitor upon wake from USB,PS2 and power button. |
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.