Originally on bug #198043: I can confirm: Thinkpad X41 with: 00:02.1 Display controller: Intel Corporation Mobile 915GM/GMS/910GML Express Graphics Controller (rev 03) Results in a 'panning' 1280x1024 screen within a 1024x768 display. Here are first few lines of /var/log/gdm/:0.log: X Window System Version 7.1.0 Release Date: 22 May 2006 X Protocol Version 11, Revision 0, Release 7.1 Build Operating System: Linux 2.6.9-34.ELsmp i686 Red Hat, Inc. Current Operating System: Linux localhost.localdomain 2.6.17-1.2358.fc6 #1 SMP Fri Jul 7 04:02:07 EDT 2006 i686 Build Date: 07 July 2006 Before reporting problems, check http://wiki.x.org to make sure that you have the latest version. Module Loader present Markers: (--) probed, (**) from config file, (==) default setting, (++) from command line, (!!) notice, (II) informational, (WW) warning, (EE) error, (NI) not implemented, (??) unknown. (==) Log file: "/var/log/Xorg.0.log", Time: Sat Jul 8 10:35:06 2006 (==) Using config file: "/etc/X11/xorg.conf" (WW) I810: No matching Device section for instance (BusID PCI:0:2:1) found (EE) I810(0): Unable to read from SDVOCTRL_E Slave 112. (EE) I810(0): Unable to read from SDVOCTRL_E Slave 114. (EE) I810(0): DDC Analog 0, 00005010 (EE) I810(0): DDC LVDS 1, 00005018 Executing (ax == 0x5f01) BIOS call at i830_driver.c:617 Couldn't get established mode 5 Couldn't get established mode 12 (EE) I810(0): Modeline "1280x1024"x75.0 135.00 1280 1296 1440 1688 1024 1025 1028 1066 +hsync +vsync (EE) I810(0): Modeline "1280x1024"x75.0 135.00 1280 1296 1440 1688 1024 1025 1028 1066 +hsync +vsync (EE) I810(0): Modeline "1280x1024"x60.0 108.78 1280 1360 1496 1712 1024 1025 1028 1059 Display 'returns to normal' (i.e., 1280x1024 display as 1280x1024) when I revert to xorg-x11-drv-i810-1.6.0-5.i386.rpm. Here is :0.log when I run with 1.6.0-5: X Window System Version 7.1.0 Release Date: 22 May 2006 X Protocol Version 11, Revision 0, Release 7.1 Build Operating System: Linux 2.6.9-34.ELsmp i686 Red Hat, Inc. Current Operating System: Linux localhost.localdomain 2.6.17-1.2358.fc6 #1 SMP Fri Jul 7 04:02:07 EDT 2006 i686 Build Date: 07 July 2006 Before reporting problems, check http://wiki.x.org to make sure that you have the latest version. Module Loader present Markers: (--) probed, (**) from config file, (==) default setting, (++) from command line, (!!) notice, (II) informational, (WW) warning, (EE) error, (NI) not implemented, (??) unknown. (==) Log file: "/var/log/Xorg.0.log", Time: Sat Jul 8 10:46:03 2006 (==) Using config file: "/etc/X11/xorg.conf" (WW) I810: No matching Device section for instance (BusID PCI:0:2:1) found (EE) AIGLX error: dlopen of /usr/lib/dri/i915_dri.so failed (/usr/lib/dri/i915_dri.so: undefined symbol: _glapi_tls_Context) (EE) AIGLX: reverting to software rendering xkb_keycodes { include "xfree86+aliases(qwerty)" }; xkb_types { include "complete" }; xkb_compatibility { include "complete" }; xkb_symbols { include "pc(pc105)+us" }; xkb_geometry { include "pc(pc105)" }; AUDIT: Sat Jul 8 10:46:04 2006: 3478 Xorg: client 1 rejected from local host
Please attach the complete /var/log/Xorg.0.log from both 1.6.0-5 and 1.6.0-6.modeset*, and your xorg.conf.
Created attachment 132135 [details] xorg.conf
Created attachment 132136 [details] Xorg.0.log from 1.6.0-5 version of i810 driver
Created attachment 132137 [details] Xorg.0.log from 1.6.0-6.modeset version of i810 driver
Yeah, I see it too, except it's much worse for me, since my Mactel Mini now insists on always going for the 2560x1600 on a 1680x1050 screen, forcing a virtual display. I'll see if I can find a previous RPM to revert to.
This appears to be the problem: (II) I810(0): Found panel of size 1024x768 in BIOS VBT tables Although this is also pretty hilarious: (II) I810(0): Not using default mode "640x350" (height too large for virtual size) I'm looking into it.
Yes, it looks like all modes are discarded with "height too large for virtual size" and then the largest defined in bios is used.
The new modesetting driver is doing interesting things on my Dell Latitude D620 too. The old driver (xorg-x11-drv-i810-1.6.0-4) gives me a stretched 1024x768 and the new one seems to detect the laptop screen properly but I get garbled stuff on the screen. Will attach the two log files and the xorg logs and config.
Created attachment 132170 [details] xorg config on Dell D620
Created attachment 132171 [details] The xorg log file for the modesetting driverr
Created attachment 132172 [details] xorg log pre modesetting driver
I did the equivalent of System->Preferences->Screen Resolution (I have it as an icon on my top panel) and selected a smaller size than what the screen displays (selected 1024x768 on a screen that's 1280x800). The new screen was 1024x768 stretched to 1280x800. I aborted that mode, that the screen switched to 1280x800 without the extra panning screen area.
All: As step one, please comment out the following entries in your xorg.conf: - Any Modes lines in the Screen section - Any DisplaySize, HorizSync, or VertRefresh commands in the Monitor section and make sure you're running at least 1.6.0-11*, previous versions did hilariously wrong things in mode validation. pbrobinson@: I'm not sure why your config requests 1400x1050, the panel certainly seems to be 1440x900. Following the above hints should do something more sensible. selinux@: Everything I'm finding about the Thinkpad X41 indicates that it really is 1024x768. Try this for me, when you have X running: - Xnest -ac -noreset :2 - DISPLAY=:2 xsetroot -gray The Xnest window should fill with the traditional X root weave background. If it looks like alternating white and black squares, you're running at the native panel resolution (or a power-of-two factor, I suppose); if it looks like a head-exploding striped interference pattern, you're not.
Also please make sure you are _not_ using the 855resolution or 915resolution hacks, they're known to interact badly.
I installed Xnest, but I get a seg fault when I try to run it from gnome/gdm. Also, the LCD panel of the X41 is indeed 1024x768, but the adapter supports higher for external monitors. I typically run with the laptop-lid shut, only driving the external monitor at 1280x1024. I'm installing latest xorg-x11-drv-i810 driver. I've made the xorg.conf changes, and will report after install is complete.
(In reply to comment #15) > I installed Xnest, but I get a seg fault when I try to run it from gnome/gdm. Can you open a new bug for this please?
Filed against 'xorg-x11-server' (couldn't find bz dropdown for Xnest) https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=198930
OK, installed 'new' i810 driver and made changes to xorg.conf as describe in comment #13. 'Fails' the same way as before: 1024x768 'panning' mode, instead of 1280x1024. Xorg.0.log appears the same as with previous modeset driver: Lots of 'height too large for virtural ....' I've reverted back to xorg-x11-drv-i810-1.6.0-5
xorg-x11-drv-i810-1.6.0-11.20060713modeset.fc6 Confirmed working on Mactel Mini. It now correctly goes into 1680x1050 without having to resort to 915resolution or other hacks. Thanks bunches! Logs/configs available upon request.
Created attachment 132517 [details] xorg-x11-drv-i810-1.6.0-11.20060713modeset.fc6 log files on Dell 640m/e1405 xorg-x11-drv-i810-1.6.0-11.20060713modeset.fc6 don't work for me. In fact, all modeset version don't work, when I'm running start& the screen goes a kind of black. Driver prior to that works with 915resolution. I'm not running 915resolution this time. In the attachment there is the xorg.conf, Xorg.0.log, lspci-v, dmesg, etc.
(In reply to comment #18) > OK, installed 'new' i810 driver and made changes to xorg.conf as describe in > comment #13. > > 'Fails' the same way as before: 1024x768 'panning' mode, instead of 1280x1024. > > Xorg.0.log appears the same as with previous modeset driver: Lots of 'height > too large for virtural ....' So, the behaviour you'd prefer is that the external monitor be run at 1280x1024 native, and the internal panel either in 10x7 pan-and-scan or turned off, correct? Either of those are doable I think, I'm just not sure which option to prefer. (Sadly it's not possible to use the panel's scaler to filter _down_ from 1280x1024 to 1024x768.)
To be as clear as possible, here is the behavior I think is 'expected. (Please reset expectations if I have them wrong....): 1. When there is no external monitor connected, or when the laptop display is 'selected', the display should be in max supported/configured resolution. In my case (Thinkpad X41) this should be 1024x768, no? 2. When there is an external monitor connected and 'active', and when the laptop display is not 'active', the resolution should be in the max supported/configured resolution. I typically connect to 2 different LCD displays, one supports a max of 1280x1024, the other goes higher. Regardless, at least 1280x1024 should be selected. I do not want pan and scan in either case (laptop LCD active, or external monitor active). Previous drivers never came up in pan and scan. Also, once the pan and scan mode 'starts', I can't seem to get out of it except to System->Admin->Display and display a particular resolution (e.g., 1280x1024) that gets put in xorg.conf, and to restart Xorg. Am I just thinking about this wrong? tom
Created attachment 132545 [details] Xorg log file for Aopen mini pc
driver: xorg-x11-drv-i810-1.6.0-11.20060713modeset.fc6 Hardware: Aopen MiniPC with built in i915 video card. External 19" display is a Hyundai L90D+, connection type DVI. Previous driver detected 1280x1024 mode with no problem. The current driver does not. I will attach Xorg.0.log and xorg.conf to this bug
Created attachment 132546 [details] Xorg conf file for Aopen mini pc
switching the driver on my MiniPC back to xorg-x11-drv-i810-1.6.0-4 and not changing anything in the xorg.conf file, allows the default rez of 1280x1024 to work again. Attaching working Xorg.0.log file
Created attachment 132548 [details] Log file for older driver on Aopen MiniPC that works correctly
xorg-x11-drv-i810-1.6.0-11.20060713modeset.fc6: This works for me without having to edit xorg.conf (although I did run system-config-display while I had problem, but without positive results). Both GDM and my session start with the correct native resolution for my notebook's display (1280x800). But System->Preferences->Screen Resolution no longer offers the 1024x768 resolution. It does offer 800x600 though. Is the lack of 1024x768 normal? I'd like to hookup a projector and press F8 while in 1024x768 (the native resolution for the projector). Most of the time, I don't have access to the projector, so I can't play around with settings. I like the 1024x768 setting with unused sides of the screen being black to preset my system.
Exactly the same thing as comment #20 on Dell Inspiron 6400 with 1680x1050 LCD panel. Blank screen with X running. Driver shipped with FC6T1 (and the one from FC5 AIGLX repository) works fine with 915resolution.
xorg-x11-drv-i810-1.6.0-12.20060717modeset.fc6 fails on my Thinkpad X41. I get server crash with xorg.conf containing mode/sync lines. Removing them 'cures' crash, but all I get is 'pan and scan' mode again. Reverting to xorg-x11-drv-i810-1.6.0-5 restores 1280x1024 function again (after uncommenting mode/sync lines in xorg.conf)
Content of gdm/:0.log for server crash: X Window System Version 7.1.1 Release Date: 12 May 2006 X Protocol Version 11, Revision 0, Release 7.1.1 Build Operating System: Linux 2.6.9-34.ELsmp i686 Red Hat, Inc. Current Operating System: Linux localhost.localdomain 2.6.17-1.2445.fc6 #1 SMP Mon Jul 24 22:43:31 EDT 2006 i686 Build Date: 24 July 2006 Before reporting problems, check http://wiki.x.org to make sure that you have the latest version. Module Loader present Markers: (--) probed, (**) from config file, (==) default setting, (++) from command line, (!!) notice, (II) informational, (WW) warning, (EE) error, (NI) not implemented, (??) unknown. (==) Log file: "/var/log/Xorg.0.log", Time: Tue Jul 25 07:19:12 2006 (==) Using config file: "/etc/X11/xorg.conf" (WW) I810: No matching Device section for instance (BusID PCI:0:2:1) found (EE) I810(0): Unable to read from SDVOCTRL_E Slave 112. (EE) I810(0): Unable to read from SDVOCTRL_E Slave 114. (EE) I810(0): DDC Analog 0, 00005010 (EE) I810(0): DDC LVDS 1, 00005018 Executing (ax == 0x5f01) BIOS call at i830_driver.c:617 Couldn't get established mode 5 Couldn't get established mode 12 (EE) I810(0): Modeline "1280x1024x75"x75.0 135.00 1280 1296 1440 1688 1024 1025 1028 1066 +hsync +vsync (80.0 kHz) (EE) I810(0): Modeline "1280x1024x75"x75.0 135.00 1280 1296 1440 1688 1024 1025 1028 1066 +hsync +vsync (80.0 kHz) (EE) I810(0): Modeline "1280x1024x60"x60.0 108.78 1280 1360 1496 1712 1024 1025 1028 1059 (63.5 kHz) Fatal server error: No modes found
I agree the new driver does NOT fix the problem on my MiniPC as well. It has a virtual size of 1280x1024 but an actual screen size of 800x600 so it goes into a pan and scan mode. The driver should detect my screen size as 1280x1024 and use that resolution. Reverting to the -4 version of the driver does not have this problem.
The 1.6.0-12.20060717modeset.fc6 version also makes no difference to the previous modeset versions on my Dell D620 either.
Are there any updates to this bug? This driver kinda sucks at the moment. Any updates would be appreciated. Is there any information I can provide to help to debug this issue.
Created attachment 133556 [details] Xorg.0.log from 1.6.0-13 version of i810 driver
Created attachment 133557 [details] Xorg.setup.log
Hmmm.... BZ dropped my comments for comment 35 and 36. xorg-x11-drv-i810-1.6.0-13.20060717modeset.fc6 still fails badly on my Thinkpad X41. I had to revert to xorg-x11-drv-i810-1.6.0-5. Above two attachments are Xorg.0.log and Xorg.setup.log.
xorg-x11-drv-i810-1.6.0-13.20060717modeset.fc6 is still failing on my MiniPC... Attaching Xorg.0.log
Created attachment 133567 [details] Xorg.0.log for failing 1.6-13 driver
(In reply to comment #22) > To be as clear as possible, here is the behavior I think is 'expected. (Please > reset expectations if I have them wrong....): > > 1. When there is no external monitor connected, or when the laptop display is > 'selected', the display should be in max supported/configured resolution. In my > case (Thinkpad X41) this should be 1024x768, no? I suppose. How do you determine "selected" ? > 2. When there is an external monitor connected and 'active', and when the laptop > display is not 'active', the resolution should be in the max > supported/configured resolution. I typically connect to 2 different LCD > displays, one supports a max of 1280x1024, the other goes higher. Regardless, at > least 1280x1024 should be selected. Again, define "active". > Also, once the pan and scan mode 'starts', I can't seem to get out of it except > to System->Admin->Display and display a particular resolution (e.g., 1280x1024) > that gets put in xorg.conf, and to restart Xorg. System / Preferences / Screen Resolution should allow you to set to a smaller mode.
(In reply to comment #24) > driver: xorg-x11-drv-i810-1.6.0-11.20060713modeset.fc6 > > Hardware: Aopen MiniPC with built in i915 video card. External 19" display is a > Hyundai L90D+, connection type DVI. > > Previous driver detected 1280x1024 mode with no problem. The current driver does > not. I will attach Xorg.0.log and xorg.conf to this bug Just acquired this box, actually. The bug is that the BIOS claims there's an LFP connected, and so the driver attempts to clamp the mode pool to the size of the LFP. Except that there's no LFP on that box, so you lose. You should be able to work around it for the moment by saying: Option "MonitorLayout" "DFP,NONE" in the appropriate Device section of your xorg.conf, but I'll probably end up blacklisting that hardware to ignore BIOS-reported LFP (fortunately the subvendor/subdevice pair identifies it as an AOpen board).
(In reply to comment #28) > But System->Preferences->Screen Resolution no longer offers the 1024x768 > resolution. It does offer 800x600 though. Is the lack of 1024x768 normal? I'd > like to hookup a projector and press F8 while in 1024x768 (the native resolution > for the projector). Most of the time, I don't have access to the projector, so > I can't play around with settings. I like the 1024x768 setting with unused > sides of the screen being black to preset my system. What _should_ happen now is that if you start the Screen Resolution applet after connecting the panel, the driver will re-probe all outputs, and the 10x7 mode should magically appear in the list.
(In reply to comment #35) > Created an attachment (id=133556) [edit] > Xorg.0.log from 1.6.0-6.modeset version of i810 driver This concerns me: (II) I810(0): Ranges: V min: 30 V max: 85 Hz, H min: 31 H max: 80 kHz, PixClock max 140 MHz ... (II) I810(0): Ignoring EDID Ranges. Using configured ranges. That would seem to indicate that your config file still has sync ranges listed for the monitor, which shouldn't be necessary. There's definitely at least one bug in the way the new code is handling sync ranges though; I'm building xorg-x11-server-1.1.1-15 with a fix. Also I realized that you're defining "LFP active" to mean something like "lid open and enabled through BIOS hotkeys", which is sensible. The hotkey thing should be easy enough, the lid detection might be harder, I know how to get edge triggers for it from ACPI but not status level yet.
(In reply to comment #41) > Option "MonitorLayout" "DFP,NONE" Ok, I added this to my xorg.conf file and it is better... but not quite right.. The modes from xrandr are odd.. Using driver: xorg-x11-drv-i810-1.6.0-13.20060717modeset.fc6 xrandr SZ: Pixels Physical Refresh *0 1280 x 1024 ( 373mm x 302mm ) *60 1 1024 x 768 ( 373mm x 302mm ) 60 2 800 x 600 ( 373mm x 302mm ) 60 3 640 x 480 ( 373mm x 302mm ) 60 4 1280 x 960 ( 373mm x 302mm ) 60 5 1152 x 864 ( 373mm x 302mm ) 60 6 1280 x 720 ( 373mm x 302mm ) 60 7 1280 x 800 ( 373mm x 302mm ) 60 8 1280 x 768 ( 373mm x 302mm ) 60 9 1440 x 900 ( 373mm x 302mm ) 60 10 1600 x 1024 ( 373mm x 302mm ) 60 Current rotation - normal Current reflection - none Rotations possible - normal left inverted right Reflections possible - none So the machine thinks my monitor is 1600x1024... vertical is right, but the display has a virtual desktop that is too big (monitor is only 1280x1024) and so the display "slides" left and right I'll attach the log..
Created attachment 133597 [details] Results after putting Option "MonitorLayout" "DFP,NONE" in xorg.conf Results after putting Option "MonitorLayout" "DFP,NONE" in xorg.conf for Aopen MiniPC 915-B
(In reply to comment #43) OK. I removed the mode lines from xorg.conf (sorry!) and restarted. The server no longer crashes, but I'm back to pan and scan on top of 800x600. I added Option "MonitorLayout" "DFP,NONE" to xorg.conf, and now get a usable system: [tbl@localhost ~]$ xrandr SZ: Pixels Physical Refresh *0 1280 x 1024 ( 342mm x 272mm ) *75 60 1 1024 x 768 ( 342mm x 272mm ) 75 70 60 2 800 x 600 ( 342mm x 272mm ) 72 75 60 56 3 640 x 480 ( 342mm x 272mm ) 73 75 60 4 1152 x 864 ( 342mm x 272mm ) 75 60 70 5 1280 x 960 ( 342mm x 272mm ) 60 6 832 x 624 ( 342mm x 272mm ) 75 7 1280 x 720 ( 342mm x 272mm ) 60 70 75 8 1280 x 800 ( 342mm x 272mm ) 60 70 75 9 1280 x 768 ( 342mm x 272mm ) 60 70 75 10 1400 x 1050 ( 342mm x 272mm ) 60 11 1440 x 900 ( 342mm x 272mm ) 60 12 1600 x 1024 ( 342mm x 272mm ) 60 13 720 x 400 ( 342mm x 272mm ) 70 14 1600 x 1050 ( 342mm x 272mm ) 75 Current rotation - normal Current reflection - none Rotations possible - normal left inverted right Reflections possible - none [tbl@localhost ~]$ [Previously was only getting 1024x768@59, 800x600@59 and something like 812x622@59.] I can now select 1280x1024, etc. I attach Xorg.0.log below.
Created attachment 133605 [details] Xorg.0.log from xorg-x11-drv-i810-1.6.0-13.20060717modeset.fc6
(In reply to comment #44) > So the machine thinks my monitor is 1600x1024... vertical is right, but the > display has a virtual desktop that is too big (monitor is only 1280x1024) and so > the display "slides" left and right Gah. The 1600x1024 mode raises its ugly head again. I'll move that to the i128 driver where it belongs, thanks for the catch.
(In reply to comment #46) > I added Option "MonitorLayout" "DFP,NONE" to xorg.conf, and now get a usable system: > > [tbl@localhost ~]$ xrandr > SZ: Pixels Physical Refresh > *0 1280 x 1024 ( 342mm x 272mm ) *75 60 > 1 1024 x 768 ( 342mm x 272mm ) 75 70 60 > 2 800 x 600 ( 342mm x 272mm ) 72 75 60 56 > 3 640 x 480 ( 342mm x 272mm ) 73 75 60 > 4 1152 x 864 ( 342mm x 272mm ) 75 60 70 > 5 1280 x 960 ( 342mm x 272mm ) 60 > 6 832 x 624 ( 342mm x 272mm ) 75 > 7 1280 x 720 ( 342mm x 272mm ) 60 70 75 > 8 1280 x 800 ( 342mm x 272mm ) 60 70 75 > 9 1280 x 768 ( 342mm x 272mm ) 60 70 75 > 10 1400 x 1050 ( 342mm x 272mm ) 60 > 11 1440 x 900 ( 342mm x 272mm ) 60 > 12 1600 x 1024 ( 342mm x 272mm ) 60 > 13 720 x 400 ( 342mm x 272mm ) 70 > 14 1600 x 1050 ( 342mm x 272mm ) 75 > Current rotation - normal > Current reflection - none > Rotations possible - normal left inverted right > Reflections possible - none > [tbl@localhost ~]$ That looks a bit better, although the virtual sizing is still messed up, and synthesizing a mode for 1600x1050 is clearly insane. Can you attach the X log from without the MonitorLayout option, so I can see how it's going wrong?
I don't know if I should put this here, or open a new bug. But with the latest driver the DPMS mode where the monitor goes into standby is now broken. It works fine in 1.6.0-4, but no longer works in 1.6.0-13.20060717modeset. I tried it with and without the MonitorLayout option and that does not seem to affect it. xdpyinfo does show the DPMS extension as being loaded. And xset shows it as being enabled. This problem showed up before in this driver and was fixed. Maybe we have a regression in that code.
(In reply to comment #49) Xorg.0.log attached below
Created attachment 133634 [details] Xorg.0.log with no MonitorLayout option in xorg.conf
(In reply to comment #50) > I don't know if I should put this here, or open a new bug. But with the latest > driver the DPMS mode where the monitor goes into standby is now broken. It works > fine in 1.6.0-4, but no longer works in 1.6.0-13.20060717modeset. I tried it > with and without the MonitorLayout option and that does not seem to affect it. New bug, please. It would be helpful if you could identify which revision introduced the bug.
Created attachment 133695 [details] Xorg.0.log xorg-x11-drv-i810-1.6.0-13.20060717modeset.fc6 randr: SZ: Pixels Physical Refresh *0 1280 x 800 ( 330mm x 211mm ) *64 1 1280 x 768 ( 330mm x 211mm ) 60 2 1280 x 720 ( 330mm x 211mm ) 60 3 1152 x 768 ( 330mm x 211mm ) 60 4 1024 x 768 ( 330mm x 211mm ) 60 5 832 x 624 ( 330mm x 211mm ) 60 6 800 x 600 ( 330mm x 211mm ) 60 7 1600 x 1024 ( 330mm x 211mm ) 64 Current rotation - normal Current reflection - none Rotations possible - normal left inverted right Reflections possible - none GDM starts with 1600x1024 (I believe), but after I login, the screen switches to 1280x800. I have an "NEC AccuSync 50" attached to this laptop. Both are displaying right now. The Xorg.0.log has a mark added by me where I switched the screen resolution from 1280x800 to 1024x768. The xorg-x11 wants to display 1024x768 on the notebook's LCD panel in stretch mode. Can't xorg-x11 display 1024x768 normally on the CRT and with black margins on the 1280x800 LCD panel? Why can't I have 640x480 (with black margins) for modes like TV output (the notebook has a SVideo output)? BTW - If I press Fn-F8 (the CRT/LCD toggler), xorg-x11 either crashes (goes back to GDM) or stops displaying. Can this chipset be used for dual-heading (LCD and CRT separately)?
On a fresh reboot, I'm no longer seeing the 1600x1024 entry.
Created attachment 133760 [details] i845 on Dell D620 Hmm, not sure why suddenly, but now the 1440x900 resolution has suddenly started working (I was previously stuck at a streches 1024.768). I had left my machine at work over the weekend and it had had some updates (Friday and Sat) without logout as well as possible a change to xorg.conf (comment out the following 2 lines). # HorizSync 31.5 - 100.0 # VertRefresh 59.0 - 75.0 xorg-x11-server-Xorg-1.1.1-14.fc6 xorg-x11-drv-i810-1.6.0-4 915resolution-0.5.2-5.fc5.90.at So its still without the modsetting driver but is at least running at the right res now. I had installed the 915resolution and put the following line in my rc.local but it hadn't really had much effect to date so I'm not sure what 'fixed' it. /usr/sbin/915resolution 5c 1440 900 & xorg.conf attached.... will also add xorg.0.log too
Created attachment 133761 [details] xorg log for i945 on Dell D620 attaching the log. Also note the comment 56 above is for a i945 (not i845 as mentioned.
Using a Toshiba Sat Pro (i845 chipset from memory). When I fire up with xorg-x11-drv-i810-1.60-13.20060717modeset, X fails to start and gives the following throwback EE : I810 : detecting sil164 EE : Unable to read from DVOIC2_E Slave 112 EE : detecting ch7xxx EE : Unable to read from DVOIC2_E Slave 236 EE : DDC Analogue, 00005010 EE : DDC LVDSE1, 00005018 EE : DDC DVO2, 00005012 Executing (ax == 0x5f01) BIOS CPU @ i830_drive.c: 617 Fatal server error No modes found on either pipe Reverting back to 1.6.0-4 and X works again. 0-5 also works, all of the modeset updates break X
Created attachment 133840 [details] Xorg.0.log on an LG T1 Dual Express with org-x11-drv-i810-1.60-13.20060717modeset The regular non-modesetting branch works fine with i915resolution. i915resolution was disabled for this test. I get what is a pretty cool almost organic decay of the screen when Xorg starts. Just colors warping no real outlines or garbage on the display.
Just some more info on this driver With 1.6.0-4 I can use CTRL-ALT-F1 to switch to a text console With 1.6.0-*modeset* I lose the text console (monitor gives me an invalid video mode message on the screen), but I can get it back by enabling the kernel frame buffer driver by adding vga=795 to the kernel line in grub. That setting is 1280x1024x32bit. I got the value from this page: http://ruslug.rutgers.edu/~mcgrof/HOWTOS/framebuffer/framebuffer.php Now I can switch to vt1 and fix things if X is broken.
Created attachment 134456 [details] xorg-x11-drv-i810-1.6.5-3.fc6 on Dell 640m/e1405 The "intel" driver still doesn't work for me, the screen turns grey when startx. The "i810" driver requires 915resolution to work, drm is ok.
What information can I provide that can help you make the modesetting driver work? Is the Xorg.log all that you need? I just tested version 1.6.5-5 of the intel driver from fc6 test3 on my fc5 install and it still doesn't work. I don't know what you are looking for so it is hard to provide helpful details. These are my lspci details: 00:02.0 VGA compatible controller: Intel Corporation Mobile Integrated Graphics Controller (rev 03) 00:02.1 Display controller: Intel Corporation Mobile Integrated Graphics Controller (rev 03) with the pci ids: 00:02.0 0300: 8086:27a2 (rev 03) 00:02.1 0380: 8086:27a6 (rev 03)
I get the scrambled and flickering screen with the current modesetting branch of version 1.72 of the driver. My system is an Acer Travelmate 6003 with i855GM chip. Native display resolution is 1400x1050 which is actually detected - the flickering lines are actually display lines and not interlaced, and Xorg.0.log shows me that it detected the resolution. Attaching part of the log. Btw, two constants are still not defined: http://lists.freedesktop.org/archives/xorg/2006-October/018652.html
Created attachment 139410 [details] flickering xorg.log modelines
FC6 These instructions helped me to get 1440x900 video resolution in my Toshiba Satellite laptop, everything looks great now: http://forums.fedoraforum.org/showthread.php?t=133789 Fedora Core 6 Toshiba Satellite P105-S6024 Video card: Intel 945GM Express Video drive: i810
Wow, long time since this one was touched. Anyone still having issues with this driver should try again with 1.6.5-17 or newer, which includes a _very_ new modesetting driver.
On my X60, I've been running with the driver set to i810 (not intel) for a long while with no problems. (Usually run the X60 in a dock connected to an LCD. Believe one LCD is quite old ....) Setting the driver to "intel" has not worked for me. Believe I have no mode setting lines in xorg.conf. This the expected behavior?
(In reply to comment #66) > Wow, long time since this one was touched. Anyone still having issues with this > driver should try again with 1.6.5-17 or newer, which includes a _very_ new > modesetting driver. "intel" driver works for me now. Dell 640m/e1405, 945G, xorg-x11-drv-i810-1.6.5-17.fc7 [root@mstar ~]# lspci -v 00:00.0 Host bridge: Intel Corporation Mobile 945GM/PM/GMS/940GML and 945GT Express Memory Controller Hub (rev 03) Subsystem: Dell Unknown device 01d8 Flags: bus master, fast devsel, latency 0 Capabilities: [e0] Vendor Specific Information 00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS/940GML Express Integrated Graphics Controller (rev 03) (prog-if 00 [VGA]) Subsystem: Dell Unknown device 01d8 Flags: bus master, fast devsel, latency 0, IRQ 16 Memory at eff00000 (32-bit, non-prefetchable) [size=512K] I/O ports at eff8 [size=8] Memory at d0000000 (32-bit, prefetchable) [size=256M] Memory at efec0000 (32-bit, non-prefetchable) [size=256K] Capabilities: [90] Message Signalled Interrupts: 64bit- Queue=0/0 Enable- Capabilities: [d0] Power Management version 2 00:02.1 Display controller: Intel Corporation Mobile 945GM/GMS/940GML Express Integrated Graphics Controller (rev 03) Subsystem: Dell Unknown device 01d8 Flags: bus master, fast devsel, latency 0 Memory at eff80000 (32-bit, non-prefetchable) [size=512K] Capabilities: [d0] Power Management version 2
Created attachment 151778 [details] Xorg.0.log from 'intel' run Updated to 1.6.5-17, and my Thinkpad X60 behaves differently than before. With new driver, and xorg.conf set to i810, all works. With new driver, and xorg.conf set to intel, both my LCD displays complain about the frequency settings (OCD displays, eventually, displays sleep). On my 'newer' display, using resolution of 1280x1024, it states that 'intel' is using H-ref of 83kHz, V-ref of 78 Hz. Here is 'xrandr' running under i810: [root@localhost ~]# xrandr Screen 0: minimum 640 x 480, current 1280 x 1024, maximum 1680 x 1050 default connected 1280x1024+0+0 (normal left inverted right) 0mm x 0mm 1680x1050 56.0 1280x1024 75.0* 1440x900 60.0 1280x800 75.0 1024x768 75.0 800x600 75.0 640x480 75.0 [root@localhost ~]# Below is my xorg.conf. I just 'toggle' i810/intel: # Xorg configuration created by system-config-display Section "ServerLayout" Identifier "single head configuration" Screen 0 "Screen0" 0 0 InputDevice "Keyboard0" "CoreKeyboard" EndSection Section "ServerFlags" Option "AllowMouseOpenFail" "yes" EndSection Section "InputDevice" Identifier "Keyboard0" Driver "kbd" Option "XkbModel" "pc105" Option "XkbLayout" "us" EndSection Section "Monitor" Identifier "Monitor0" ModelName "LCD Panel 1280x1024" ### Comment all HorizSync and VertSync values to use DDC: Option "dpms" EndSection Section "Device" Identifier "Videocard0" Driver "i810" EndSection Section "Screen" Identifier "Screen0" Device "Videocard0" Monitor "Monitor0" DefaultDepth 24 EndSection Attaching Xorg.0.log from both 'intel' and 'i810' runs.
Created attachment 151779 [details] Xorg.0.log from 'intel' run (previous was for i810, sorry)
I have a similar problem. With the i810 driver everything works fine for the most part. When I use the intel driver with everything else the same. My display will randomly go dark and come back (almost like it is losing sync and getting it back before the monitor can put up the error message). The display I am using in a 1680x1050 LCD connected over DVI. The machine I have is a Aopen Mini PC 915-B. Also the frequency of going black increases the longer I use the machine. I'm using a very minimal xorg.conf as generated by starting X. And adding the one line about XAA bitmaps. more /etc/X11/xorg.conf # Xorg configuration created by system-config-display Section "ServerLayout" Identifier "single head configuration" Screen 0 "Screen0" 0 0 InputDevice "Keyboard0" "CoreKeyboard" EndSection Section "InputDevice" Identifier "Keyboard0" Driver "kbd" Option "XkbModel" "pc105" Option "XkbLayout" "us" EndSection Section "Device" Identifier "Videocard0" Driver "i810" Option "XAANoOffscreenPixmaps" "true" EndSection Section "Screen" Identifier "Screen0" Device "Videocard0" DefaultDepth 24 SubSection "Display" Viewport 0 0 Depth 24 EndSubSection EndSection
I've got a tablet pc which doesn't allow me to run at the native 1024x768 resolution. This is mostly a question about the modesetting selection routine. As the list of modes below show, the 1024x600 mode is after the 1024x768 mode. It seems the driver always selects that mode even though I explicitly configure xorg.conf to only use 1024x768. Is there a faulty logic in the selection routine which assumes that the higher the mode number is the higher the resolution is. Or perhaps it only looks at the horizontal resolution. The default vga mode works for the text console, 800x600 works for X. 1024x600 works for X but is letterboxed (like 16:9 movies are on a 4:3 monitor). So the display is running 1024x768 but X is only using 1024x600. I've been testing this with the xorg server from the fedora development repository and with the xorg from FC6. Is there a way to force the driver to select one specific mode by id? Like mode 45 or 54. 855resolution version 0.4, by Alain Poirier Chipset: Unknown (id=0x35758086) VBIOS type: 3 VBIOS Version: 2756 Mode 30 : 640x480, 8 bits/pixel Mode 31 : 640x400, 8 bits/pixel Mode 32 : 800x600, 8 bits/pixel Mode 34 : 1024x768, 8 bits/pixel Mode 36 : 1024x600, 8 bits/pixel Mode 38 : 1280x1024, 8 bits/pixel Mode 3a : 1600x1200, 8 bits/pixel Mode 3c : 1920x1440, 8 bits/pixel Mode 40 : 640x480, 15 bits/pixel Mode 41 : 640x480, 16 bits/pixel Mode 42 : 800x600, 15 bits/pixel Mode 43 : 800x600, 16 bits/pixel Mode 44 : 1024x768, 15 bits/pixel Mode 45 : 1024x768, 16 bits/pixel Mode 46 : 1024x600, 15 bits/pixel Mode 47 : 1024x600, 16 bits/pixel Mode 48 : 1280x1024, 15 bits/pixel Mode 49 : 1280x1024, 16 bits/pixel Mode 4a : 1600x1200, 15 bits/pixel Mode 4b : 1600x1200, 16 bits/pixel Mode 4c : 1920x1440, 15 bits/pixel Mode 4d : 1920x1440, 16 bits/pixel Mode 50 : 640x480, 32 bits/pixel Mode 52 : 800x600, 32 bits/pixel Mode 54 : 1024x768, 32 bits/pixel Mode 56 : 1024x600, 32 bits/pixel Mode 58 : 1280x1024, 32 bits/pixel Mode 5a : 1600x1200, 32 bits/pixel Mode 5c : 1920x1440, 32 bits/pixel
I can get my Thinkpad X60 'working' using 'intel' instead of 'i810' by the following: It seems that with 'intel', the VertRefresh rate is getting set to 78 (instead of 60 or 75). Running 'randr --rate 60' seems to work: monitor is happy, etc. Here is output from two runs of 'randr'. The first, when the video was set at 78Hz (as reported by the little popup from my monitor), the second after running 'randr --rate 60'. [tbl@localhost ~]$ xrandr Screen 0: minimum 320 x 200, current 1280 x 1024, maximum 1280 x 1280 VGA connected 1280x1024+0+0 (normal left inverted right) 376mm x 301mm 1280x1024 60.0 + 75.0* 59.9 1152x864 75.0 1024x768 75.1 70.1 60.0 832x624 74.6 800x600 72.2 75.0 60.3 640x480 75.0 72.8 60.0 720x400 70.1 LVDS connected 1024x768+0+0 (normal left inverted right) 246mm x 185mm 1024x768 50.0*+ 60.0 40.0 800x600 60.3 640x480 60.0 59.9 TV disconnected (normal left inverted right) [tbl@localhost ~]$ xrandr --rate 60 [tbl@localhost ~]$ xrandr Screen 0: minimum 320 x 200, current 1280 x 1024, maximum 1280 x 1280 VGA connected 1280x1024+0+0 (normal left inverted right) 376mm x 301mm 1280x1024 60.0*+ 75.0 59.9 1152x864 75.0 1024x768 75.1 70.1 60.0 832x624 74.6 800x600 72.2 75.0 60.3 640x480 75.0 72.8 60.0 720x400 70.1 LVDS connected 1024x768+0+0 (normal left inverted right) 246mm x 185mm 1024x768 50.0*+ 60.0 40.0 800x600 60.3 640x480 60.0 59.9 TV disconnected (normal left inverted right) [tbl@localhost ~]$ I even tried putting 'VertRefresh 60,75' in xorg.conf to no avail. Here are first few lines from 'lspci': 00:00.0 Host bridge: Intel Corporation Mobile 945GM/PM/GMS, 943/940GML and 945GT Express Memory Controller Hub (rev 03) 00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller (rev 03) 00:02.1 Display controller: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller (rev 03)
'xrandr --rate 60' makes 'intel' work on my older LCD as well. 'intel' appears to set refresh rate to 77.8Hz instead of 75 or 60. [root@localhost ~]# xrandr Screen 0: minimum 320 x 200, current 1280 x 1024, maximum 1280 x 1280 VGA connected 1280x1024+0+0 (normal left inverted right) 359mm x 287mm 1280x1024 75.0 59.9* 1152x864 75.0 1024x768 75.1 70.1 60.0 832x624 74.6 800x600 72.2 75.0 60.3 56.2 640x480 75.0 72.8 66.7 60.0 LVDS connected (normal left inverted right) 1024x768 50.0 + 60.0 40.0 800x600 60.3 640x480 60.0 59.9 TV disconnected (normal left inverted right) [root@localhost ~]# Some of the above 'rates' seem funny ...
Running with 'intel', and after 'xrandr --rate 60'. Here is snippet of output from 'xrandr --verbose': 1280x1024 (0x5f) 135.0MHz h: width 1280 start 1296 end 1440 total 1688 skew 0 clock 80.0KHz v: height 1024 start 1025 end 1028 total 1066 clock 75.0Hz 1280x1024 (0x60) 109.0MHz h: width 1280 start 1368 end 1496 total 1712 skew 0 clock 63.7KHz v: height 1024 start 1027 end 1034 total 1063 clock 59.9Hz The 'on screen display' of the monitor reports that the clock is actually 109.6MHz, H is 64.0kHz and V is 60.1Hz. Is this just an apparent discrepency (109.0MHz vs. 109.6MHz, 63.7kHz vs. 64.0, 59.9Hz vs 60.1)? Could this 'rounding/truncation' be driving 75Hz to 77.8Hz (and 80KHz to 83KHz)?
Running with the intel driver on FC6 xorg-x11-drv-i810-1.6.5-10.fc6 xrandr reports SZ: Pixels Physical Refresh *0 1680 x 1050 ( 454mm x 281mm ) *60 1 1280 x 1024 ( 454mm x 281mm ) 75 60 2 1024 x 768 ( 454mm x 281mm ) 75 70 60 3 800 x 600 ( 454mm x 281mm ) 72 75 60 56 4 640 x 480 ( 454mm x 281mm ) 73 75 60 67 5 1152 x 864 ( 454mm x 281mm ) 75 60 70 6 1280 x 960 ( 454mm x 281mm ) 60 7 832 x 624 ( 454mm x 281mm ) 75 8 1280 x 720 ( 454mm x 281mm ) 60 70 75 9 1280 x 800 ( 454mm x 281mm ) 60 70 75 10 1280 x 768 ( 454mm x 281mm ) 60 70 75 11 1400 x 1050 ( 454mm x 281mm ) 60 70 12 1440 x 900 ( 454mm x 281mm ) 60 13 1600 x 1024 ( 454mm x 281mm ) 60 14 720 x 400 ( 454mm x 281mm ) 70 Current rotation - normal Current reflection - none Rotations possible - normal left inverted right Reflections possible - none BUT only if I specify my MonitorLayout. If I don't specify my monitor layout the setup is totally wrong. The display I am using is a Gateway FPD2185W (1680x1050) LCD connected over DVI. See Comment 45 above Section "Device" Identifier "Videocard0" Driver "intel" Option "XAANoOffscreenPixmaps" "true" Option "MonitorLayout" "DFP" EndSection The machine I have is a Aopen MiniPC 915-B 00:02.0 VGA compatible controller: Intel Corporation Mobile 915GM/GMS/910GML Express Graphics Controller (rev 04) (prog-if 00 [VGA]) Subsystem: AOPEN Inc. Unknown device 0589 Flags: bus master, fast devsel, latency 0, IRQ 19 Memory at d0180000 (32-bit, non-prefetchable) [size=512K] I/O ports at da00 [size=8] Memory at c0000000 (32-bit, prefetchable) [size=256M] Memory at d0200000 (32-bit, non-prefetchable) [size=256K] Capabilities: <access denied> 00:02.1 Display controller: Intel Corporation Mobile 915GM/GMS/910GML Express Graphics Controller (rev 04) Subsystem: AOPEN Inc. Unknown device 0589 Flags: bus master, fast devsel, latency 0 Memory at d0100000 (32-bit, non-prefetchable) [size=512K] Capabilities: <access denied>
Modesetting driver works fine on Dell Inspiron 6400 with 1280x800 panel here, as tested with Fedora 7 Test 4 Live CD (i386).
Ditto Dell Inspiron 6400 with 1680x1050 panel and F7T4 Live CD (i386).
With that higher res display (1680x1050), the display went crazy on my attempt to reboot (with barely visible white rectangles sprinkled over the black background). The box hung at that point. Had to press the button to turn it off.
this works much better for me with my Dell D620 now using Fedora 7 Test 4. So well in fact that after testing using the live CD I installed it. I can get a fully working screen on both the internal 1440x900 wide screen and the external screen. I can't seem to get it working properly across multiple screens but then I'm not sure what the status of xinerma support is. Basically the install (well yum upgrade) worked fine. Once running I then removed the 915resolution command run from rc.local and removed the xorg.conf file and regenerated it with the 'intel' driver and rebooted and it was all sweet!
I just tried the Fedora 7 test4 kde live cd, and have added some comments in bug #238171 (https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=238171) regarding some strange behavior on Dell Latitude X1 (widescreen 1280x768 internal display, intel 915GM/GMS/910GML) and an external monitor set to 1280x1024, using the 'intel' driver in my xorg.conf. Clone sort of works, but spanning desktop results in X freezing white. (sorry for not putting it here in the first place) -MartinG
I'm seeing exactly the same when trying to configure spanning. You configure spanning and restart X and you just get all white, it seems I can't even go to a terminal to reset it. I have the following config: Dell Latitude D620 Core Duo (not Core2) with a docking station with the following screens: - Laptop Panel (1440x900) - HP LP1965 connected to dock analog (1280x1024) - HP LP1965 connected to dock digital (1280x1024) Ideally I'd like to have the 2 HP screens spanned when connected to the dock and the 1440x900 when not. Not sure if there's enough support in the i945 to run all 3 at once. I suspect I can do this using xrandr commands manually when the screen doesn't turn white when xinerama is configured. So at the moment both HP screens are just mirrored when connected (laptop lid shut) which also brings up another interesting issue. I suspect its more to do with applications dealing with screens running at different resolutions but when you maximise some apps they grow to 1440x900 rather than 1280x1024, firefox seems OK but ForecastFox displays partway up the screen now. Evolution does some interesting stuff and the mouse pointer disappears off the right of the screen.
OK. The dual screen bit going white I think should be a different bug as its not related to modesetting which (for me a least) seem to be much better. A bit of reading through the X.org list and bugzilla revealed the following for dual screen reveals the bugzilla entry which indicates that it is supported but the config setup by the config gui tool does it wrong and it seems crashes the x driver, so when I get a sec I'm going to try doing it manually (either with xrandr or manual edit to the config file). There should probably be a bug done for system-config-display (or whatever the package is to fix this). https://bugs.freedesktop.org/show_bug.cgi?id=10706 Also saw some comments regarding my disparate display sizes with mice in this thread http://lists.freedesktop.org/archives/xorg/2007-May/024351.html
Current intel hardware is only able to drive to displays independently. So if you have 3 connected (laptop LCD + two monitors) at least two will have to be clones. It sounds like the actual modesetting is working ok here though, so maybe this bug should be closed out (that and there have been no adds for over 6 months).
There was only ever a maximum of 2 running at any one time. When it was connected to the docking station it was the 2 HP LCDs and the laptop lid was closed.
Changing version to '9' as part of upcoming Fedora 9 GA. More information and reason for this action is here: http://fedoraproject.org/wiki/BugZappers/HouseKeeping
I think this can now be closed. with the 2.x series of drivers this has been alot more stable. In Fedora 9 with a HP nx7400 1680x1050 LCD and a Dell 22inch 1680x1050 LCD attached this all just works. Got to love xrandr!
Closing as per comment #88. This bug is clearly tl;dr anyway. If anyone is still having problems with intel output setup, please file new bugs so we can track them separately rather than dogpiling on this one.