[i845G] Fatal server error: Couldn't bind memory for BO front buffer (Jaunty)

I do have grub.cfg without the xforcevesa if that could be an issue?

Thank you for reporting this bug. Could you please also attche the following:
- /etc/X11/xorg.conf
- output of `lspci -vvnn`

Also, have you installed the libdrm-intel1 package? (see bug 303177)

I have installed libdrm-intel1 package.

Attaching the output of lspci --vvnn

Attaching the output of /etc/X11/xorg.conf

My apt term.log

Could be a 845-specific issue (in the forums link, one user saying he had the same issue also saw it on 845; the other user didn't indicate his hardware so who knows.) Anyone on hardware other than 845 able to reproduce it?

Anyway, I've forwarded it upstream here; please subscribe, etc.:
https://bugs.freedesktop.org/show_bug.cgi?id=18974

Attaching the output of lspci --vvnn .

I have the same 'Fatal error' messege, intel865G chipset. Stuck on 'vesa'
attached ispci as well

Thanks, so 865 is affected as well apparently. Anyone else seeing this same exact error?

I am seeing exactly this behaviour with Jaunty and 845 chipset on an IBM machine.

On a Dell OptiPlex GX270 (bios A07), with a current Jaunty (including libdrm-intel1 2.4.1-0ubuntu5), I have the same problems here:

(EE) intel(0): Failed to pin front buffer: Cannot allocate memory

Fatal server error:
Couldn't bind memory for BO front buffer

00:02.0 VGA compatible controller [0300]: Intel Corporation 82865G Integrated Graphics Controller [8086:2572] (rev 02) (prog-if 00 [VGA])
 Subsystem: Dell Device [1028:0151]
 Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
 Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
 Latency: 0
 Interrupt: pin A routed to IRQ 16
 Region 0: Memory at f0000000 (32-bit, prefetchable) [size=128M]
 Region 1: Memory at feb80000 (32-bit, non-prefetchable) [size=512K]
 Region 2: I/O ports at ed98 [size=8]
 Capabilities: [d0] Power Management version 1
  Flags: PMEClk- DSI+ D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
  Status: D0 PME-Enable- DSel=0 DScale=0 PME-
 Kernel modules: intelfb

I was having this problem too.

I downgraded my kernel to 2.6.27 and things are working perfectly fine.

I suspect the problem is that libdrm is not correctly detecting that the kernel is not doing mode setting.

I confirm that downgrading to 2.6.27-8 also works for me. The desktop takes forever to load but I suspect that that is another issue.

I can confirm this bug with 855GM chipset.
For me, kernel 2.6.27 was working also. I cannot attach any logs cause I allready reformatted and removed Jaunty. I saw that this error still is in Alpha2 (https://wiki.ubuntu.com/JauntyJackalope/TechnicalOverview#Known%20Issues) so I wait testing Jaunty until at least X is working correctly. ;)

Still seeing this issue, adding new Xorg log.

Same issue with 'Intel Corporation 82852/855GM Integrated Graphics Device rev 2'.

I can confirm this bug with i845 chipset

I was experiencing this problem with an old uniwell laptop (intel graphics), and have a partial resolution by changing the color depth from 24 to 16 bit, i.e. add
DefaultDepth 16
to the screen section of /etc/X11/xorg.conf

Partial resolution means I can achieve default resolution 1280x800 at 16 bit color using the intel driver

I tried setting the DefaultDepth to 16, but that caused X to lock up (mouse worked, could not type in the log in prompt) after a few seconds (less than 15).

I tried adding NoAccel and that is allowing me to use X at 1920x1200, without compiz.

Both logs are attached.

The errors and warnings in the NoAccel log I attached are:

(WW) intel(0): libpciaccess reported 0 rom size, guessing 64kB
(WW) intel(0): Register 0x70024 (PIPEASTAT) changed from 0x80000207 to 0x00000000
(WW) intel(0): PIPEASTAT before: status: FIFO_UNDERRUN VSYNC_INT_STATUS SVBLANK_INT_STATUS VBLANK_INT_STATUS OREG_UPDATE_STATUS
(WW) intel(0): PIPEASTAT after: status:
(EE) intel(0): Failed to init memory manager

That first error message yields some google matches. I thought EE was a dead end for X, but this log is from the active session.

Marques Johansso, can you put up your xorg.conf ?
--
          Regards,
          Shirish Agarwal
  This email is licensed under http://creativecommons.org/licenses/by-nc/3.0/
http://flossexperiences.wordpress.com
065C 6D79 A68C E7EA 52B3 8D70 950D 53FB 729A 8B17

It's the vanilla xorg.conf with 'Option "NoAccel" "true"' added. I have some other junk in there, but that's all it is.

Section "Device"
 Identifier "Configured Video Device"
# when intel doesn't work, modprobe intelfb and use these instead of 'driver "intel"'
# Driver "fbdev"
# Option "UseFBDev" "true"
 Driver "intel"

 Option "NoAccel" "true" # works around LPB 304871 - probably disables the next two
 Option "XvMC" "true" # Enabled for no particular reason
 Option "TripleBuffer" "true" # Enabled for no particular reason
 Option "ModeDebug" "true" # More Verbose logging
 Option "FallbackDebug" "true" # More Verbose logging
EndSection

Section "Monitor"
 Identifier "Configured Monitor"
EndSection

Section "Screen"
# 16BPP is a possible work around for LPB 304871 that only locks me up at
# 2.6.28-3-generic, xserver-xorg-video-intel 2:2.5.1-1ubuntu7
# DefaultDepth 16
 Identifier "Default Screen"
 Monitor "Configured Monitor"
 Device "Configured Video Device"
EndSection

I've also used a daily Jaunty live CD while this bug has existed, and strangely enough, the CD which uses the latest xorg and 2.6.28 kernel doesn't give me this problem. Maybe it disables acceleration also - I didn't check.

The log shows that my XvMC is disabled, but if you are going to copy my config line per line, you may also want to make sure the file "/etc/X11/XvMCConfig" has this single line:
libIntelXvMC.so.1

I also get few warnings

(WW) The directory "/usr/share/fonts/X11/cyrillic" does not exist.
(WW) intel(0): libpciaccess reported 0 rom size, guessing 64kB
(WW) intel(0): Register 0x70024 (PIPEASTAT) changed from 0x80000207 to 0x00000000
(WW) intel(0): PIPEASTAT before: status: FIFO_UNDERRUN VSYNC_INT_STATUS SVBLANK_INT_STATUS VBLANK_INT_STATUS OREG_UPDATE_STATUS
(WW) intel(0): PIPEASTAT after: status:
(WW) config/hal: device AT Translated Set 2 keyboard already added. Ignoring.

Pretty sure I'm hitting this too with an Intel 915GM/GMS/910GML chipset. Can supply logs if useful in addition to what is already attached.

Temp workaround for me was to switch to Vesa driver.

i confirm this too, i can not use jaunty jackalope because of this bug, i am using:
Intel(r) 82845G/GL/GE/PE/GV Graphic Controller

Sure, intel 915GM is affected too. Also on vesa/fbdev.

I added the Option "NoAccel"
to Section "Device"
worked for me too (I even forgot to put "true" so that must be default)

I have a DELL 2400 that ran fine with 8.10, but failed after upgrade to Jaunty. lspci reveals

00:00.0 Host bridge: Intel Corporation 82845G/GL[Brookdale-G]/GE/PE DRAM Controller/Host-Hub Interface (rev 01)
00:02.0 VGA compatible controller: Intel Corporation 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device (rev 01)
00:1d.0 USB Controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) USB UHCI Controller #1 (rev 01)
00:1d.1 USB Controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) USB UHCI Controller #2 (rev 01)
00:1d.2 USB Controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) USB UHCI Controller #3 (rev 01)
00:1d.7 USB Controller: Intel Corporation 82801DB/DBM (ICH4/ICH4-M) USB2 EHCI Controller (rev 01)
00:1e.0 PCI bridge: Intel Corporation 82801 PCI Bridge (rev 81)
00:1f.0 ISA bridge: Intel Corporation 82801DB/DBL (ICH4/ICH4-L) LPC Interface Bridge (rev 01)
00:1f.1 IDE interface: Intel Corporation 82801DB (ICH4) IDE Controller (rev 01)
00:1f.3 SMBus: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) SMBus Controller (rev 01)
00:1f.5 Multimedia audio controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) AC'97 Audio Controller (rev 01)
01:09.0 Ethernet controller: Broadcom Corporation BCM4401 100Base-T (rev 01)

and the simple fix works ie adding

option "NoAccel" "true"

to Section "Device" in the stock xorg.conf seems to work well for me.

Upstream said that DRI2 will fix this bug.

When someone is able to boot with DRI2 on jaunty, please let us know if this is the case or not.

I tested Jaunty Daily Build 1/27/2009 and the bug still occurs. Is DRI2 live yet? If not, do we have an ETA? I am happy to re-test when it makes sense.

klaxian, why didn't you attach your Xorg.0.log?

I can confirm the problem is still there as of my latest updates 30/jan/2009 15:gmt.
uname -a ==>
Linux everest 2.6.28-6-generic #16-Ubuntu SMP Mon Jan 26 20:16:00 UTC 2009 i686 GNU/Linux

My Xorg.0.log is attached (there is also an Xorg.failsafe.log as the system tries to start in low res mode).

Sorry Bryce. I figured there were already enough logs attached, but I would be happy to include mine as well. Since I'm only testing with a LiveCD and it won't boot into X, how do you suggest I go about getting this log? Is there an easy method? Thanks.

I was also able to get the driver working by adding 'Option "DRI" "false"' instead of 'Option "NoAccel" "true"'. It sounds like a more favorable work-around for the time being.

Section "Device"
        Identifier "Configured Video Device"
        Driver "intel"
        Option "XvMC" "true" # /etc/X11/XvMCConfig should contain "libIntelXvMC.so.1"
        Option "DRI" "false"
        Option "ModeDebug" "true"
        Option "FallbackDebug" "true"
EndSection

Section "Monitor"
        Identifier "Configured Monitor"
EndSection

Section "Screen"
        Identifier "Default Screen"
        Monitor "Configured Monitor"
        Device "Configured Video Device"
EndSection

In Jaunty, it is now possible to use DRI2 by using UXA. This can be done by specifying Option "AccelMethod" "UXA" in your xorg.conf.

Note that it will *probably* break your xorg very badly, however I don't know anyone who has tested it on 8xx chips, so who knows. A few people have found that with UXA it resolves all performance issues and works great. Other people have found it freezes on displaying the login screen, and other problems, so please be forewarned.

Once we have someone's testing of UXA on 8xx, we can re-ping upstream about this issue.

OK I have the 82845G chipset and UXA works up to a point in that I see the GDM splash screen and the mouse works, but I find my keyboard locked whilst trying to login (User name appeared after a pause, but could not enter). I attach the Xorg.0.conf

I have also tried Marques Johansson method and that works and seems better than just disabling all acceleration.

i can confirm, that on my laptop it is fixed on latest jaunty updates

i have:
Intel Corporation 82852/855GM Integrated Graphics Device [8086:3582]

To confirm, Option "AccelMethod" "UXA" worked for me on a Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller.

Option "AccelMethod" "UXA" did not work for
00:02.0 VGA compatible controller: Intel Corporation 82865G Integrated Graphics Controller (rev 02)

The video did come up and immediately had no keyboard. Restarted GDM with no effect. Tried to reboot computer, it just ignored me, ssh was still working so I umounted my raid and did a hard reboot. It booted again and now I had keyboard and mouse. I logged in but before gnome fully came up, it locked up again. I tried all the above steps nothing, so i changed back to NoAccel and reboot the hard way and bqack to normal.

Using the following hardware:

00:02.0 VGA compatible controller: Intel Corporation 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device (rev 01)

Option "AccelMethod" "UXA" will boot to graphical login prompt but is incredibly slow and not usable. Had to kill -9 X from ssh session and reboot.

Option "DRI" "false" and option "Noaccel" "true" are equally usable.

UXA also locks my X up after a few seconds.

00:02.0 VGA compatible controller: Intel Corporation 82865G Integrated Graphics Controller (rev 02) (prog-if 00 [VGA])
 Subsystem: Dell Device 0151
 Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
 Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
 Latency: 0
 Interrupt: pin A routed to IRQ 16
 Region 0: Memory at e8000000 (32-bit, prefetchable) [size=128M]
 Region 1: Memory at feb80000 (32-bit, non-prefetchable) [size=512K]
 Region 2: I/O ports at ed98 [size=8]
 Capabilities: [d0] Power Management version 1
  Flags: PMEClk- DSI+ D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
  Status: D0 PME-Enable- DSel=0 DScale=0 PME-
 Kernel modules: intelfb

I have tested Kubuntu Jaunty amd64 Desktop from 03-feb-2009 http://cdimage.ubuntu.com/kubuntu/daily-live/20090203/jaunty-desktop-amd64.iso with 82865G. Using the live CD I could start X and use KDE with some graphics corruptions and Kwin compositor reported 100 FPS, with Intrepid KDE 4.2 I get only 85, so I get a performance boost.

Also bug 295934 get fixed in Jaunty.

00:02.0 VGA compatible controller: Intel Corporation 82865G Integrated Graphics Controller (rev 02)

UXA locks my laptop with i830m graphics:
00:02.0 VGA compatible controller: Intel Corporation 82830 CGC [Chipset Graphics Controller] (rev 04)

With Option ""NoAccel" "true" i get a usable (but sluggish) desktop, with just ~50 FPS in glxgears.

With following hardware:
00:02.1 Display controller: Intel Corporation Mobile 945GM/GMS/GME, 943/940GML Express Integrated Graphics Controller (rev 03)

and UXA method works but with aprox 300FPS in glxgear.

After upgrading from Ubuntu 8.10, upon restart I got after the progression bar with Ubuntu logo a black (blueish) screen for gdm (with usual sound). Ctrl+Alt+F1 seemed to do something, but with the same screen. At least Ctrl+Alt+Del would restart the system. Choosing 2.6.27 in grub didn't make a difference. After some more restarts, I was proposed failsafe on server 1 (0 beeing busy. Failsave proposed to archive the logs, which I did (attachment). This failsafe is in 1280*1024 (the screen's best)which allows me to write this normally.

Video is Intel 945G. (Full lspci -vvnn attached)
00:02.0 VGA compatible controller [0300]: Intel Corporation 82945G/GZ Integrated Graphics Controller [8086:2772] (rev 02)
 Subsystem: ASUSTeK Computer Inc. Device [1043:817a]
 Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
 Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
 Latency: 0
 Interrupt: pin A routed to IRQ 11
 Region 0: Memory at dfe00000 (32-bit, non-prefetchable) [size=512K]
 Region 1: I/O ports at 8800 [size=8]
 Region 2: Memory at e0000000 (32-bit, prefetchable) [size=256M]
 Region 3: Memory at dfe80000 (32-bit, non-prefetchable) [size=256K]
 Capabilities: <access denied>
 Kernel modules: intelfb

UXA locks up my pc, im running latest packages
00:00.0 Host bridge: Intel Corporation 82865G/PE/P DRAM Controller/Host-Hub Interface (rev 02)

Section "Device"
        Identifier "Configured Video Device"
        Driver "intel"
        Option "ModeDebug" "true"
        Option "FallbackDebug" "true"
        Option "AccelMethod" "UXA"
        Option "XvMC" "true"
        Option "DRI" "false"
EndSection

and the Xorg log... sorry i couldn't tell how to attach two log files in one post.

Confirmed here on Dell Dimension 4500s

This xorg.conf, with 16 bit and no acel worked at a temp fix
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/304871/comments/27

put very poor refresh rate/fps

I came here seeing a warning about Intel under http://www.ubuntu.com/testing/jaunty/alpha4#Known%20Issues but I realise that my problem (above) has little to do with this bug. (I do not get "Fatal server error: Couldn't bind memory for BO front buffer".) Should I file this as another bug?
I keep getting things like
[ 0.148993] (II) intel(0): Output VGA using monitor section Monitor0
[ 0.149006] (**) intel(0): Option "PreferredMode" "1280x1024"
[ 1.201850] (II) intel(0): Output VGA disconnected
[ 1.201890] (WW) intel(0): No outputs definitely connected, trying again...
[ 1.201901] (II) intel(0): Output VGA disconnected
[ 1.201912] (WW) intel(0): Unable to find initial modes
[ 1.201923] (EE) intel(0): No valid modes.
[ 1.499698] (II) UnloadModule: "intel"
[ 1.499738] (II) UnloadModule: "vgahw"
[ 1.499749] (II) Unloading /usr/lib/xorg/modules//libvgahw.so
[ 1.499768] (EE) Screen(s) found, but none have a usable configuration.

Fatal server error:
no screens found

After some restarts, I once got a regular mode. Most of the time nothing. Sometimes failsafe.

I would like to ask that all subscribers to this bug with i8xx hardware please test activating UXA. Directions and a place to record your findings are at this page:

   https://wiki.ubuntu.com/X/UxaTesting

The reason I ask is that there have been just two 8xx users reporting there (both indicating very positive findings), and I'd like to obtain broader test coverage, particularly so we will know about problems.

@Dominique: yes, your issue is probably a separate bug, please open a new bug report, preferably by using the ubuntu-bug tool as described in https://wiki.ubuntu.com/X/Reporting . Thanks!

Bryce, I had a minute to try out the latest alpha live CD on my 855 with UXA. It was a big improvement for me. I've updated the matrix. It wasn't extensive testing. I had compiz and UXA running for as long as it takes to install Jaunty + an additional 20 minutes or so. No lockups.

On Tue, Feb 10, 2009 at 10:20 PM, Bryce Harrington
<email address hidden> wrote:
> I would like to ask that all subscribers to this bug with i8xx hardware
> please test activating UXA. Directions and a place to record your
> findings are at this page:
>
> https://wiki.ubuntu.com/X/UxaTesting
>
> The reason I ask is that there have been just two 8xx users reporting
> there (both indicating very positive findings), and I'd like to obtain
> broader test coverage, particularly so we will know about problems.

Done.

On a 82852/855GM is is nice to see native resolution again.
Unfortunately the blue squares replacing various text pixmaps maps the
system unusable within 30 minutes.

I also see some icon corruption as well.

I have not taken into account any kind of performance and in order to
avoid a lock-up you must disable visual effects as well.

Thanks,
Anand

Bryce,

I have added my findings on the wiki. I'm also attaching the latest Xorg.0.log. I want to reaffirm that only the vesa driver doesn't cause my system to lock up.

Thanks,
Aram

Added dmesg and Xorg.0.log using UXA, and screenshots showing corruption with EXA and UXA. Also added findings on the wiki.

On my system (Thinkpad X40), UXA works nicely and gives the best performance when desktop effects are activated, but it shows some random screen corruption (blanked characters, red-filled icons, horizontal lines...).

lspci:
00:02.0 VGA compatible controller: Intel Corporation 82852/855GM Integrated Graphics Device (rev 02)
00:02.1 Display controller: Intel Corporation 82852/855GM Integrated Graphics Device (rev 02)

$ apt-cache show xserver-xorg-video-intel | grep Version
Version: 2:2.6.1-1ubuntu2

$ uname -a
Linux x40 2.6.28-5-generic #15-Ubuntu SMP Thu Jan 22 21:21:04 UTC 2009 i686 GNU/Linux

(I cannot use the latest kernel because usplash doesn't ask for crypt-setup password)

uxa gives a blue screen with white top and bottom borders, then no response to any keyboard action. Mouse pointer moves however no response to clicks.

IBM ThinkCentre A30

Intel Corporation 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device (rev 01)

Jaunty uname -a
linux 2.6.28-7-generic #20-Ubuntu SMP Mon Feb 9 15:43:21 UTC 2009 i686 GNU/Linux

Since the system is dead altogether no logs or dmesg. Only thing that works is Option "NoAccel" which is pretty slow.

Jerry

Thanks everyone, this fleshes in the table a lot better. It confirms what I suspected at the beginning - it's a mixed bag, great for some, poor for others, and unfortunately there's not a strong correlation between success and chip family, so black/whitelisting doesn't seem to be an option for us.

From what I gather, upstream is going to be sticking with UXA. Earlier on it sounded like they would backport stuff to EXA once it was ready, but recent comments make it sound like that is no longer the plan. So going forward I anticipate that Intel will be prioritizing support for UXA issues over EXA, which means we are going to be faced with the situation of a) user for whom UXA is better than EXA will be pushing us to move to UXA by default, and b) people for whom EXA has been working ok but UXA brings problems will be upset to see the regressions.

In my ideal world, what I would like to see is for us the distro to hold off on moving to UXA until it is more stable, and have technically able users such as yourselves take the initiative of working with upstream directly at getting the issues you've found investigated and solved. This would help Ubuntu out a lot in paving a path of stability with UXA that we can follow for Jaunty+1 (and maybe even Jaunty if enough work gets done in the next few weeks).

So I would like to encourage each of you to forward the issues you find upstream. Here is a link to use in filing your bugs:

  https://bugs.freedesktop.org/enter_bug.cgi?bug_severity=major&component=Driver%2Fintel&op_sys=Linux (All)&priority=high&product=xorg&version=7.4

When filing your bugs upstream, please always include your Xorg.0.log (and/or Xorg.0.log.old) after reproducing the issue with Option "ModeDebug" "true" in your xorg.conf. Also include your PCI IDs, which you can get by running the command `lspci -vvnn | grep -A2 "VGA compat"`.

Additional guidelines on filing bugs upstream to Intel are at http://intellinuxgraphics.org/how_to_report_bug.html - please be concise and specific in your reports, and be sure to follow up; this will help ensure the issue gets examined and addressed in a timely fashion for us. I'm particularly concerned for 8xx since it seems to get ignored too much, and I'd like to get awareness for the known issues upstream ASAP.

Once you've filed your bug, it would be wonderful if you could also include a link to the bug in your comments on the UxaTesting wiki page - https://wiki.ubuntu.com/X/UxaTesting - as this will help others with similar bugs, and make it easier for me to keep tabs on how stabilization is going. In particular, update the Result section once the issues you've encountered are no longer occurring on your system.

Perhaps it would make sense to track version 2.7 of the intel driver for jaunty (which would in effect mean the master branch for now, I guess). After all, intel releases on a quarterly basis. Last release was in q4 2008, so 2.7 should be released in q1 2009, i.e. in time for the jaunty release. And I'm sure it'll have a lot of UXA stabilization going in. In either event, it's hard to imagine 2.7 being less stable and less performant than 2.6.x :-p

> In either event, it's hard to imagine 2.7 being less stable and less performant than 2.6.x :-p

You're new here, aren't you. ;-)

Bryce, I assume this model worked with XAA? Is it possible to switch between XAA and EXA on a per-model basis?

Martin, yes it is possible to switch between XAA and EXA on a per-model basis. The trade-off is that Intel will *only* support EXA, so we would not be able to upstream bugs for any chips that use XAA. They asked us last year to stop using XAA and move to EXA for all chips, in order to make it easier for them to provide support.

We've actually had a series of problems with 8xx chips ever since moving from -i810 to -intel in the Hardy timeframe. There have been a series of performance and stabilization regressions on these old chips as upstream has moved forward, that seem to not be getting attention. Indeed, they have completely dropped support on i815 and some of the other very early 8xx chips.

So, if XAA makes things better on i845, we can consider moving to it, but understand that once we do so we probably won't be able to get any other bugs fixed on this chipset.

XAA not much luck on this i845. Sometimes it will run for a minute or two, today it wouldn't even complete booting complaining about low graphics mode with 2.6.28-8.8. Neither is UXA or EXA I could generate Xorg.0.logs if they are of interest.

Presently for jaunty I'm using Driver "vesa" which has much much better scroll than Driver "Intel" Option "NoAccel".

By far the best is intrepid ibex with acpid=noirq.

Ubuntu seems to be well on the path of "latest, high end equipment only" which is the way Microsoft goes. If it weren't for the jaunty video the performance of this 2 gHz Celeron is just fine for internet news, flash videos, Office Write & spreadsheet, mail, ... No I don't do games.

So I may well be stuck on Intrepid for IBM Think Centre A30 2 gHz Celeron and IBM Thinkpad R31 1 gHz Celeron. The IBM Thinkpad T40 and Compaq Presario 3.3 gHz Celeron are O.K. with jaunty, who knows about Koala....

Thanks for your efforts on this. Jerry

> Ubuntu seems to be well on the path of "latest, high end equipment only" which is the way Microsoft goes.

Please, that is hardly a fair statement to make.

The purpose for putting out new Ubuntu releases is *primarily* to provide users with easy access to the latest code from upstream. If upstream chooses to drop support for some old piece of hardware, there is often little we can do about it at the distro level; if it's something our OEMs care about or that has a ton of users, Canonical may be able to justify investing resources there, but that is rarely the case with legacy graphics hardware. If an older release works, stick with it.

Ultimately, part of the reason we all want to see hardware vendors putting out their drivers as open source instead of proprietary, is so that when the company decides to cut off support for a chip, at least the community has the option to provide support for it themselves. We're in that situation now for i815 and i825; it stands to reason eventually we'll be there with i845 too.

I don't like seeing support for old Intel chips deteriorate any more than owners of the chip do, but I think at some point the only way we're going to keep up support on the old chips is if some community heroes step up to the plate and take a swing at it themselves. The good news is that with -intel, most of the code is already there, and experts are easily at hand for advice and guidance; if you have the hardware, ample free time, and know a little C coding, you can do it.

On Fri, Jan 30, 2009 at 2:48 AM, Bryce Harrington
<email address hidden> wrote:
> Upstream said that DRI2 will fix this bug.
>
> When someone is able to boot with DRI2 on jaunty, please let us know if
> this is the case or not.

Hi,

I am running:

eve[~]% uname -a
Linux eve 2.6.29-020629rc6-generic #020629rc6 SMP Mon Feb 23 12:53:58
UTC 2009 i686 GNU/Linux

And everything is now working as expected.

So it seems:
 2.6.27 (via linux-ports) works but does not have audio modules
 2.6.28 (9.04 kernel) does not work (glyph corruption, eventual machine hang)
 2.6.29-rc6 (and later, presumably) will work

What is the process from here? As 2.6.28 will be the shipping kernel
anyone upgrading with i845 / i855 hardware is going to experience an
issue.

Cheers,
Anand

I can confirm my Intel i815 integrated graphics has this problem, x starts then its says something about Intel freezing and I never can get to a log-in screen, so then I can't use Ubuntu 9.04.I hope this is fixed so I can upgrade to 9.04.and I also have to say since I've been using Ubuntu since 6.06 I have never had a problem with any alpha's or beta with the i815 graphics driver.email me ,please when there is a workaround this.

thanks

keep up the good work.

Hi Edmond,

If you have had a chance to read the various things stated, there are
three workarounds:

- use the 2.6.27 kernel
- use the 2.6.28 kernel
- skip the upgrade to 9.04

Cheers,
Anand

On Sat, Feb 28, 2009 at 5:27 AM, Edmond Thiebaud <email address hidden> wrote:
> I can confirm my Intel i815 integrated graphics has this problem, x
> starts then its says something about Intel freezing and I never can get
> to a log-in screen, so then I can't use Ubuntu 9.04.I hope this is fixed
> so I can upgrade to 9.04.and I also have to say since I've been using
> Ubuntu since 6.06 I have never had a problem with any alpha's or beta
> with the i815 graphics driver.email me ,please when there is a
> workaround this.
>
> thanks
>
> keep up the good work.
>
> --
> [i845G] Fatal server error: Couldn't bind memory for BO front buffer (Jaunty)
> https://bugs.launchpad.net/bugs/304871
> You received this bug notification because you are a direct subscriber
> of the bug.
>
>

Status update:

I've seen there are a bunch of BO fixes included in the 2.6.3 version of the driver, however that version has new dependencies. I've done the merge of the 2.6.3 -intel, however it won't build until the dependencies are satisfied. I've packaged and uploaded libdrm 2.4.5 that it requires, and this is included in Alpha6. It also requires a small kernel change, which I've spoken to the kernel team about but that won't go in until after Alpha6 at the earliest.

Also, the vast bulk of the 2.6.3 changes are for features we do not enable by default yet, so I'm not certain if the benefit is going to outweigh the risk. We'll want to make sure ample testing is done on it before we make the decision.

I only get the message "Fatal server error: Couldn't bind memory for BO front buffer" when trying to use XAA in a i865G, but not with EXA nor with UXA.

Using Intel 2.7 RC1, Kernel 2.6.29 RC7, and the Xserver + Mesa from the repository I don't get so many corruptions on screen, but still some, especially with wine, virtualbox and flash videos. Using composition those corruptions go away, but scrolling text e.g. with Firefox is slower.

Also UXA is more stable, it doesn't crash any more with opengl apps like glxgears, but it is not possible to distinguish the gears, it is like blur.

With my little experience, I think that we should try intel 2.6.3.

I have good news and bad news, before i wasn't able to get into 9.04, but
just a few hours ago i upgraded to 9.04 again and was not able to login to
9.04, but i was able to log in fluxbox.

On Wed, Mar 11, 2009 at 4:30 PM, G. Christ <email address hidden> wrote:

> I only get the message "Fatal server error: Couldn't bind memory for BO
> front buffer" when trying to use XAA in a i865G, but not with EXA nor
> with UXA.
>
> Using Intel 2.7 RC1, Kernel 2.6.29 RC7, and the Xserver + Mesa from the
> repository I don't get so many corruptions on screen, but still some,
> especially with wine, virtualbox and flash videos. Using composition
> those corruptions go away, but scrolling text e.g. with Firefox is
> slower.
>
> Also UXA is more stable, it doesn't crash any more with opengl apps like
> glxgears, but it is not possible to distinguish the gears, it is like
> blur.
>
> With my little experience, I think that we should try intel 2.6.3.
>
> --
> [i845G] Fatal server error: Couldn't bind memory for BO front buffer
> (Jaunty)
> https://bugs.launchpad.net/bugs/304871
> You received this bug notification because you are a direct subscriber
> of the bug.
>

Since I'm the guy who suggested we track 2.7, obviously I think 2.6.3 is a good idea :-p

Seriously, the fixes that have gone into 2.6.3 might not be relevant for what's gonna be enabled by default in Jaunty, but if things don't improve, 855GM users are gonna see forum posts that say "add accelmethod UXA to xorg.conf and you'll be fine." Which brings me to my point: with 2.6.3 I have had no server crash on my 855GM with UXA enabled, whereas on 2.6.1 I had a couple of hard xserver hangs.

So +1 for trying out 2.6.3 :-)

Hi,

> So +1 for trying out 2.6.3 :-)

Me too. I also hits freeze on GMA965 on last intel-GPU MacBook. Looks
like the problem does not touch only old hardwares.

Regards,
Étienne
--
E Ultreïa !

I've been getting the "Fatal server error: Couldn't bind memory for BO front buffer" error as well on Jaunty with my old laptop's "Intel Corporation 82852/855GM Integrated Graphics Device". The error goes away for me by adding

Option "Legacy3D" "false"

to the Device section.

Nolari, with Legacy3d false, does it work in AccelMethod UXA, EXA, or both?

Marques: seems to work with both.

There is a fix for this issue in the newest driver-version :

http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=73db44e7ac524e84e5f0fda2d60069a9e954ad1b

"Drop Legacy3D option, only use fixed texture space with non-gem."

Together with two other git-commits (see Bug #316588) backported to the driver version 2.6.1, the driver now works perfectly on my 855-chipset.

O.K., downloaded
xf86-video-intel-2.6.3
any hints on how to use this directory?
There seems to be a Makefile.am in it and an autogen.sh however I don't know the command line commands to do anything with them.

Thanks for hints,

Jerry

You can test the latest intel (from git) driver and other stuff from the "PPA for xorg crack pushers" https://launchpad.net/~xorg-edgers/+archive/ppa and the latest kernel from https://wiki.ubuntu.com/KernelTeam/KnowledgeBase

I'm using this right now with linux 2.6.29 rc7 and everything seems stable and with less problems.

I've uploaded 2.6.3 a couple days ago, which in theory should fix this problem.

@Shirish, mind updating to latest jaunty and re-testing?

I have updated to 2.6.3 but it has made no difference at all to the
reported problem.

Thanks for the update.

If any one or combination of these lines is remarked in the xorg.conf file
the system throws the error or fails to boot.

    Option "DRI" "False"
    Option "NoAccel" "True"
    Option "AccelMethod" "UXA"

2009/3/18 Bryce Harrington <email address hidden>

> I've uploaded 2.6.3 a couple days ago, which in theory should fix this
> problem.
>
> @Shirish, mind updating to latest jaunty and re-testing?
>
> --
> [i845G] Fatal server error: Couldn't bind memory for BO front buffer
> (Jaunty)
> https://bugs.launchpad.net/bugs/304871
> You received this bug notification because you are a direct subscriber
> of a duplicate bug.
>
> Status in X.org xf86-video-intel: Confirmed
> Status in “xserver-xorg-video-intel” source package in Ubuntu: Triaged
> Status in xserver-xorg-video-intel in Ubuntu Jaunty: Triaged
> Status in “xserver-xorg-video-intel” source package in Fedora: Fix Released
>
> Bug description:
> Since the change from the 2.4.1 driver to 2.5.1, Jaunty now hangs during
> boot, with "Fatal server error: Couldn't bind memory for BO front buffer"
> shown in the log. So far, this has been seen by two people with i845
> graphics and two with i865.
>
> Also discussed at this forum:
> https://bugs.freedesktop.org/show_bug.cgi?id=18974
>
> [Original Report]
> Binary package hint: xserver-xorg-video-intel
>
> Hi all,
> This is in jaunty, I get no screens.
>
>
>
> [lspci]
> 00:00.0 Host bridge [0600]: Intel Corporation 82845G/GL[Brookdale-G]/GE/PE
> DRAM Controller/Host-Hub Interface [8086:2560] (rev 03)
> 00:02.0 VGA compatible controller [0300]: Intel Corporation
> 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device [8086:2562]
> (rev 03)
> Subsystem: Intel Corporation 82845G/GL[Brookdale-G]/GE Chipset
> Integrated Graphics Device [8086:2562]
>
>

Hi,

I've also upgraded, done the same tests and experienced the same failures.

Anand

On Thu, Mar 19, 2009 at 3:00 AM, wilbur <email address hidden> wrote:
> I  have updated to 2.6.3 but it has made no difference at all to the
> reported problem.
>
> Thanks for the update.
>
> If any one or combination of these lines is remarked in the xorg.conf file
> the system throws the error or fails to boot.
>
>    Option      "DRI" "False"
>    Option      "NoAccel" "True"
>    Option      "AccelMethod" "UXA"
>
>
> 2009/3/18 Bryce Harrington <email address hidden>
>
>> I've uploaded 2.6.3 a couple days ago, which in theory should fix this
>> problem.
>>
>> @Shirish, mind updating to latest jaunty and re-testing?
>>
>> --
>> [i845G] Fatal server error: Couldn't bind memory for BO front buffer
>> (Jaunty)
>> https://bugs.launchpad.net/bugs/304871
>> You received this bug notification because you are a direct subscriber
>> of a duplicate bug.
>>
>> Status in X.org xf86-video-intel: Confirmed
>> Status in “xserver-xorg-video-intel” source package in Ubuntu: Triaged
>> Status in xserver-xorg-video-intel in Ubuntu Jaunty: Triaged
>> Status in “xserver-xorg-video-intel” source package in Fedora: Fix Released
>>
>> Bug description:
>> Since the change from the 2.4.1 driver to 2.5.1, Jaunty now hangs during
>> boot, with "Fatal server error: Couldn't bind memory for BO front buffer"
>> shown in the log.  So far, this has been seen by two people with i845
>> graphics and two with i865.
>>
>> Also discussed at this forum:
>> https://bugs.freedesktop.org/show_bug.cgi?id=18974
>>
>> [Original Report]
>> Binary package hint: xserver-xorg-video-intel
>>
>> Hi all,
>>  This is in jaunty, I get no screens.
>>
>>
>>
>> [lspci]
>> 00:00.0 Host bridge [0600]: Intel Corporation 82845G/GL[Brookdale-G]/GE/PE
>> DRAM Controller/Host-Hub Interface [8086:2560] (rev 03)
>> 00:02.0 VGA compatible controller [0300]: Intel Corporation
>> 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device [8086:2562]
>> (rev 03)
>>        Subsystem: Intel Corporation 82845G/GL[Brookdale-G]/GE Chipset
>> Integrated Graphics Device [8086:2562]
>>
>>
>
> --
> [i845G] Fatal server error: Couldn't bind memory for BO front buffer (Jaunty)
> https://bugs.launchpad.net/bugs/304871
> You received this bug notification because you are a direct subscriber
> of the bug.
>
>

2.6.3 isn't a solution for me either. UXA locks X (the keyboard and display locks during the gnome-session startup - the mouse cursor works however). I think the BO aspect of the xserver-xorg-video-intel problems may be out of the way.

This log was generated when I tried EXA with NoLegacy3D:

X.Org X Server 1.6.0
Release Date: 2009-2-25
X Protocol Version 11, Revision 0
Build Operating System: Linux 2.6.24-19-server i686 Ubuntu
Current Operating System: Linux bang 2.6.28-10-generic #33-Ubuntu SMP Tue Mar 17 12:27:50 UTC 2009 i686
Build Date: 07 March 2009 02:18:57AM
xorg-server 2:1.6.0-0ubuntu1 (<email address hidden>)
        Before reporting problems, check http://wiki.x.org
        to make sure that you have the latest version.
Markers: [ 0.011339] (--) probed, [ 0.011360] (**) from config file, [ 0.011375] (==) default setting,
        [ 0.011390] (++) from command line, [ 0.011404] (!!) notice, [ 0.011418] (II) informational,
        [ 0.011433] (WW) warning, [ 0.011447] (EE) error, [ 0.011461] (NI) not implemented, [ 0.011476] (??) unknown.

[ 0.011590] (==) Log file: "/var/log/Xorg.0.log", Time: Wed Mar 18 08:37:21 2009
[ 0.028686] (==) Using config file: "/etc/X11/xorg.conf"
exaCopyDirty: Pending damage region empty!
error setting MTRR (base = 0xe8000000, size = 0x08000000, type = 1) Invalid argument (22)
 ddxSigGiveUp: Closing log

This is what I got from UXA without default options. (I can't use UXA with DRI off. X still locks with Legacy3D)

X.Org X Server 1.6.0
Release Date: 2009-2-25
X Protocol Version 11, Revision 0
Build Operating System: Linux 2.6.24-19-server i686 Ubuntu
Current Operating System: Linux bang 2.6.28-10-generic #33-Ubuntu SMP Tue Mar 17 12:27:50 UTC 2009 i686
Build Date: 07 March 2009 02:18:57AM
xorg-server 2:1.6.0-0ubuntu1 (<email address hidden>)
 Before reporting problems, check http://wiki.x.org
 to make sure that you have the latest version.
Markers: [ 0.017446] (--) probed, [ 0.017469] (**) from config file, [ 0.017484] (==) default setting,
 [ 0.017498] (++) from command line, [ 0.017513] (!!) notice, [ 0.017527] (II) informational,
 [ 0.017541] (WW) warning, [ 0.017556] (EE) error, [ 0.017570] (NI) not implemented, [ 0.017584] (??) unknown.
[ 0.017701] (==) Log file: "/var/log/Xorg.0.log", Time: Wed Mar 18 08:28:57 2009
[ 0.034666] (==) Using config file: "/etc/X11/xorg.conf"
get fences failed: -1
param: 6, val: 0
get fences failed: -1
param: 6, val: 0
[mi] EQ overflowing. The server is probably stuck in an infinite loop.

Backtrace:
0: /usr/X11R6/bin/X(xorg_backtrace+0x3b) [0x8134f0b]
1: /usr/X11R6/bin/X(mieqEnqueue+0x2d9) [0x8114ab9]
2: /usr/X11R6/bin/X(xf86PostKeyboardEvent+0xc5) [0x80d81c5]
3: /usr/lib/xorg/modules/input//evdev_drv.so [0xaf47a3db]
4: /usr/X11R6/bin/X [0x80c7dc7]
5: /usr/X11R6/bin/X [0x80b82ec]
6: [0xb7eef400]
7: /usr/lib/xorg/modules/drivers//intel_drv.so [0xb781b810]
8: /usr/lib/xorg/modules/drivers//intel_drv.so(uxa_prepare_access+0x64) [0xb7832574]
9: /usr/lib/xorg/modules/drivers//intel_drv.so(uxa_copy_n_to_n+0x2db) [0xb78349fb]
10: /usr/lib/xorg/modules//libfb.so(fbCopyRegion+0x1d6) [0xb7725796]
11: /u...

Friday March 20th 2009 with latest updates.

With the following entry in the xorg.conf file Ubuntu now boots and runs
normally. It hangs with "DRI" "True".

Section "Device"
    VendorName "Intel Corporation"
    Identifier "82865G Integrated Graphics Controller"
    Driver "Intel"
    BusID "PCI:0:2:0"
    Option "DRI" "False"
EndSection

2009/3/19 Marques Johansson <email address hidden>

> 2.6.3 isn't a solution for me either. UXA locks X (the keyboard and
> display locks during the gnome-session startup - the mouse cursor works
> however). I think the BO aspect of the xserver-xorg-video-intel
> problems may be out of the way.
>
> This log was generated when I tried EXA with NoLegacy3D:
>
> X.Org X Server 1.6.0
> Release Date: 2009-2-25
> X Protocol Version 11, Revision 0
> Build Operating System: Linux 2.6.24-19-server i686 Ubuntu
> Current Operating System: Linux bang 2.6.28-10-generic #33-Ubuntu SMP Tue
> Mar 17 12:27:50 UTC 2009 i686
> Build Date: 07 March 2009 02:18:57AM
> xorg-server 2:1.6.0-0ubuntu1 (<email address hidden>)
> Before reporting problems, check http://wiki.x.org
> to make sure that you have the latest version.
> Markers: [ 0.011339] (--) probed, [ 0.011360] (**) from config file,
> [ 0.011375] (==) default setting,
> [ 0.011390] (++) from command line, [ 0.011404] (!!) notice, [
> 0.011418] (II) informational,
> [ 0.011433] (WW) warning, [ 0.011447] (EE) error, [
> 0.011461] (NI) not implemented, [ 0.011476] (??) unknown.
>
> [ 0.011590] (==) Log file: "/var/log/Xorg.0.log", Time: Wed Mar 18
> 08:37:21 2009
> [ 0.028686] (==) Using config file: "/etc/X11/xorg.conf"
> exaCopyDirty: Pending damage region empty!
> error setting MTRR (base = 0xe8000000, size = 0x08000000, type = 1) Invalid
> argument (22)
> ddxSigGiveUp: Closing log
>
>
> This is what I got from UXA without default options. (I can't use UXA with
> DRI off. X still locks with Legacy3D)
>
> X.Org X Server 1.6.0
> Release Date: 2009-2-25
> X Protocol Version 11, Revision 0
> Build Operating System: Linux 2.6.24-19-server i686 Ubuntu
> Current Operating System: Linux bang 2.6.28-10-generic #33-Ubuntu SMP Tue
> Mar 17 12:27:50 UTC 2009 i686
> Build Date: 07 March 2009 02:18:57AM
> xorg-server 2:1.6.0-0ubuntu1 (<email address hidden>)
> Before reporting problems, check http://wiki.x.org
> to make sure that you have the latest version.
> Markers: [ 0.017446] (--) probed, [ 0.017469] (**) from config file,
> [ 0.017484] (==) default setting,
> [ 0.017498] (++) from command line, [ 0.017513] (!!) notice, [
> 0.017527] (II) informational,
> [ 0.017541] (WW) warning, [ 0.017556] (EE) error, [
> 0.017570] (NI) not implemented, [ 0.017584] (??) unknown.
> [ 0.017701] (==) Log file: "/var/log/Xorg.0.log", Time: Wed Mar 18
> 08:28:57 2009
> [ 0.034666] (==) Using config file: "/etc/X11/xorg.conf"
> get fences failed: -1
> param: 6, val: 0
> get fences failed: -1
> param: 6, val: 0
> [mi] EQ overflowing. The server is probably stuck in an infinite loop.
>
> Backtrace:
> 0: /usr/X11R6/bin/X(xorg_backtrace+0x3b) [0x8134f0b]
> 1: ...

Friday March 20th 2009 with latest updates.

With the following entry in the xorg.conf file Ubuntu now boots and runs
normally. It hangs with "DRI" "True" and *Running a video causes a reboot or
kills X*.

Section "Device"
    VendorName "Intel Corporation"
    Identifier "82865G Integrated Graphics Controller"
    Driver "Intel"
    BusID "PCI:0:2:0"
    Option "DRI" "False"
EndSection

Score! kernel 2.6.28-11-generic with xorg.......intel updates works on i830 on IBM Thinkpad R31. No changes to xorg.conf, just as is upgraded today.

First jaunty that's run without "NoAccel" on the R31!

Kudo's to whoever fixed it!l Hope it stays fixed, I'll use dual boot to keep this jaunty around.

Thanks much, Jerry

p.s. I'll try my i845 next when I swap a display onto it. It was too sluggish with jaunty before.

Try running a video which takes down my i865. Much better screen scrolling
as long as I don't run a video.

wilbur told me to change the option section in xorg.conf to "DRI" "False"
and then i saved it, upgraded to 9.04 and eveything boots just perfect, the
only thing is for some reason i cant uninstall Ubuntu Studio, but thats not
big deal, and also the fonts are too small in ubuntu, but i use xrandr to
set my correct resolution, its only a temp fix, everytime i log in to
ubuntu to use 800x600 i have to do xrandr -s 800x600 -r 85, but thats also
not big problem because i always had to do that in fluxbox.Thanks Wilbur for
that fix for me, i'am not going to touch anything on this system.

On Fri, Mar 20, 2009 at 9:05 PM, wilbur <wilburgr@gI follmail.com> wrote:

> Try running a video which takes down my i865. Much better screen scrolling
> as long as I don't run a video.
>
> --
> [i845G] Fatal server error: Couldn't bind memory for BO front buffer
> (Jaunty)
> https://bugs.launchpad.net/bugs/304871
> You received this bug notification because you are a direct subscriber
> of the bug.
>

I have no problem playing video

Dell2400Dimension-Intel 82845G/GL/ICH4
Have today march 21 2009 successfully installed Ubuntu Jaunty 9.04 daily build live-cd. Perfect audio/video.
I have missed Ubuntu since 11 September 2008 due to i8xx chipset issue.
Thanks

Video YouTube Leona Lewis "Run" and scrolled while the video was going.

No problem. Video, audio, & scroll all at the same time. Must be time for a new update!

VGA compatible controller: Intel Corporation 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device (rev 01)

Linux version 2.6.28-11-generic (buildd@rothera) (gcc version 4.3.3 (Ubuntu 4.3.3-5ubuntu4) ) #36-Ubuntu SMP Fri Mar 20 19:40:40 UTC 2009

Jerry

On Sat, Mar 21, 2009 at 2:04 PM, jerrylamos <email address hidden> wrote:

> Video YouTube Leona Lewis "Run" and scrolled while the video was going.
> The only problem besides the small fonts is now i cant log-in with the
> newest kernal, but the older kernal i'am able to log-in.
> No problem. Video, audio, & scroll all at the same time. Must be time
> for a new update!
>
> VGA compatible controller: Intel Corporation 82845G/GL[Brookdale-G]/GE
> Chipset Integrated Graphics Device (rev 01)
>
> Linux version 2.6.28-11-generic (buildd@rothera) (gcc version 4.3.3
> (Ubuntu 4.3.3-5ubuntu4) ) #36-Ubuntu SMP Fri Mar 20 19:40:40 UTC 2009
>
> Jerry
>
> --
> [i845G] Fatal server error: Couldn't bind memory for BO front buffer
> (Jaunty)
> https://bugs.launchpad.net/bugs/304871
> You received this bug notification because you are a direct subscriber
> of the bug.
>

@ubuntu:~$ lspci
00:00.0 Host bridge: Intel Corporation 82865G/PE/P DRAM Controller/Host-Hub Interface (rev 02)
00:02.0 VGA compatible controller: Intel Corporation 82865G Integrated Graphics Controller (rev 02)
00:1d.0 USB Controller: Intel Corporation 82801EB/ER (ICH5/ICH5R) USB UHCI Controller #1 (rev 02)
00:1d.1 USB Controller: Intel Corporation 82801EB/ER (ICH5/ICH5R) USB UHCI Controller #2 (rev 02)
00:1d.2 USB Controller: Intel Corporation 82801EB/ER (ICH5/ICH5R) USB UHCI Controller #3 (rev 02)
00:1d.7 USB Controller: Intel Corporation 82801EB/ER (ICH5/ICH5R) USB2 EHCI Controller (rev 02)
00:1e.0 PCI bridge: Intel Corporation 82801 PCI Bridge (rev c2)
00:1f.0 ISA bridge: Intel Corporation 82801EB/ER (ICH5/ICH5R) LPC Interface Bridge (rev 02)
00:1f.1 IDE interface: Intel Corporation 82801EB/ER (ICH5/ICH5R) IDE Controller

as others above i could only get into kde by setting DRI to false within the device section of xorg.conf.

I boot into the cli and then startx. No kdm. before the DRI fix, it would typically freeze during the splash sequence, though sometimes it would make it to the desktop with mouse moving but otherwise not functioning. No keyboard either.

it is 3/22/09 and i am updated to all the latest jaunty versions of kernel and intel and xorg and libdrm; and still not working for the 865g without a hack.

Anyway! Yes running a video will crash the desktop. Most of the time i end up at the cli (no kdm) with xorg messages there in print to read. the xorg.0.log message says "could not pin xv". so i startx into kde and open mplayer and and go to preferences to set video from xv to gl. then to real player to turn xv off. then to electricsheep screensaver to set it from xv to gl. I control alt backspace and then startx again and yes i can do the electricsheep without crashing the desktop. Oh, by the way, electricsheep uses mplayer to run the sheep.

Those interested, please note: in systemsettings within the desktop icon/desktop effects/advanced tab; I have compositing type set to "xrender". I had to change it from the default "opengl" to get transparent effects in the panel and for konsole and some plasmoids.

Those still interested please also note: I have been running jaunty for months. I have been using the xorg packages from intrepid. Everything on my system is the latest jaunty versions, except xorg. everything has worked "fine" "slow but functional" for months. if interested and AT YOUR OWN risk, just go to packages.ubuntu.com and download the xorgs you need from intrepid. then do "apt-get --purge remove xserver-xorg". now go to directory containg all those xserver-xorg packages and do a "dpkg -i xserver*". i must have gone back and forth with this a number of times. every time the repositories show a new version of xorg-intel i do a full upgrade and crash the desktop. lukily i got my little stash of intrepid xorg to get me back in.

THANKS guys for showing me the way with DRI false. Thanks.

Hello

for Device :

00:02.0 VGA compatible controller: Intel Corporation 82852/855GM Integrated Graphics Device (rev 02)

bug fixed with next xorg.conf config Video Device:

Section "Device"
        Identifier "Configured Video Device"
        Driver "intel"
        Option "XvMC" "true"
        Option "AccelMethod" "UXA"
EndSection

I have acceleration and not have problem with video ... ;-)

THANKS

Thanks for the input.

With an Intel 82865G my system locks up tight using UXA.

The following works except for some video:

Section "Device"
    VendorName "Intel Corporation"
    BoardName "Intel 82865G"
    Identifier "82865G Integrated Graphics Controller"
    Driver "intel"
    BusID "PCI:0:2:0"
    Option "DRI" "False"
# Option "NoAccel" "True"
EndSection

To run some video the 'NoAccel' line has to be:
    Option "NoAccel" "True"

We have to hope that this does get fixed at some point.

2009/3/25 segato <email address hidden>

> Hello
>
> for Device :
>
> 00:02.0 VGA compatible controller: Intel Corporation 82852/855GM
> Integrated Graphics Device (rev 02)
>
> bug fixed with next xorg.conf config Video Device:
>
> Section "Device"
> Identifier "Configured Video Device"
> Driver "intel"
> Option "XvMC" "true"
> Option "AccelMethod" "UXA"
> EndSection
>
> I have acceleration and not have problem with video ... ;-)
>
> THANKS
>
> --
> [i845G] Fatal server error: Couldn't bind memory for BO front buffer
> (Jaunty)
> https://bugs.launchpad.net/bugs/304871
> You received this bug notification because you are a direct subscriber
> of a duplicate bug.
>
> Status in X.org xf86-video-intel: Confirmed
> Status in “xserver-xorg-video-intel” source package in Ubuntu: Triaged
> Status in xserver-xorg-video-intel in Ubuntu Jaunty: Triaged
> Status in “xserver-xorg-video-intel” source package in Fedora: Fix Released
>
> Bug description:
> Since the change from the 2.4.1 driver to 2.5.1, Jaunty now hangs during
> boot, with "Fatal server error: Couldn't bind memory for BO front buffer"
> shown in the log. So far, this has been seen by two people with i845
> graphics and two with i865.
>
> Also discussed at this forum:
> https://bugs.freedesktop.org/show_bug.cgi?id=18974
>
> [Original Report]
> Binary package hint: xserver-xorg-video-intel
>
> Hi all,
> This is in jaunty, I get no screens.
>
>
>
> [lspci]
> 00:00.0 Host bridge [0600]: Intel Corporation 82845G/GL[Brookdale-G]/GE/PE
> DRAM Controller/Host-Hub Interface [8086:2560] (rev 03)
> 00:02.0 VGA compatible controller [0300]: Intel Corporation
> 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device [8086:2562]
> (rev 03)
> Subsystem: Intel Corporation 82845G/GL[Brookdale-G]/GE Chipset
> Integrated Graphics Device [8086:2562]
>
>

Hi,

There are two possible solutions for this bug (at least for all 855GM-based hardware owners) :

Add the following line to your Device-Section in the "xorg.conf"-file :

Option "Legacy3d" "false"

With this option i can use EXA an 3D-acceleration an my DELL Latitude D505.

Second option :

Include the following git-commit into the driver version 2.6.3 :

"Drop Legacy3D option, only use fixed texture space with non-gem."

http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=73db44e7ac524e84e5f0fda2d60069a9e954ad1b

This patch works perfectly with driver version 2.6.3 on my notebook and the "Legacy3D" option is no longer needed.

Hello

with this xorg.conf :

Section "Device"
        Identifier "Configured Video Device"
        Driver "intel"
        Option "XvMC" "true"
        Option "AccelMethod" "UXA"
EndSection

I have problem with de VGA output, and this VGA config found use this config :

Section "Device"
        Identifier "Configured Video Device"
        Driver "intel"
       Option "DRI" "false"
EndSection

I do change this config to use the VGA in my notebook ...

VGA output BUG NOT CLOSED

:'(

Saludos from Chile

Using the 9.04 Beta Live CD it starts on my Intel 865GBF motherboard.

0000:00:02.0 VGA compatible controller: Intel Corporation 82865G Integrated Graphics Controller (rev 02) (prog-if 00 [VGA])
        Subsystem: Intel Corporation: Unknown device 4246
        Flags: bus master, fast devsel, latency 0, IRQ 11
        Memory at f0000000 (32-bit, prefetchable) [size=128M]
        Memory at ffa00000 (32-bit, non-prefetchable) [size=512K]
        I/O ports at ec00 [size=8]
        Capabilities: [d0] Power Management version 1

Hrmph, no word from upstream, no word from the original reporter.

Of the other commenters, some say the issue is fixed. Others (mostly 845G and 865G users) say it's still there but goes away once DRI is disabled.

It is easy enough to force DRI off for all 845G and 865G chips. I've put a patch to do this into my ppa. I would ask that 845 and 865 users please test it and verify it fixes the issue, without requiring setting anything in your xorg.conf. If it does, I can upload it.

If anyone does not think we should do this change, now's the time to speak up, but provide an explanation why.

PPA with this fix:
https://edge.launchpad.net/%7Ebryceharrington/+archive/ppa/+sourcepub/535882/+listing-archive-extra

Intel 82865G user.

Before doing the patch I would like to know what the DRI actually does and
what will be lost by patching it out. I am one of those for whom the problem
is definitely not resolved. Things like flash and video are not working
correctly or not working at all to the point where I am off looking at other
distos.

Option "XvMC" "true" does not seem to help
Option "DRI" "False" has to be false or lockup
Option "NoAccel" "True" has to be true to play videos
makes the system unpleasant to use for other things.
Option "AccelMethod" "UXA" causes lockup EXA seems OK
Option "DPMS" "false" no change noted
Option "Legacy3d" "false" no change noted

So most of the time I run with the 'NoAccel' option commented out but
activate it and restart X if I want to run a video. Some flash will not run,
for example the Dow Jones chart at
http://www.google.com/finance?q=INDEXDJX:.DJI which used to work perfectly
in 8.10 and seems to work well in a number of other distros I have tried.

At the moment 9.04 is a crippled OS with the Intel 82865G. I hope it gets
fixed and that we are not left stranded.

Many thanks for the DRI patch I will give it a try and report back.

Here to report that Stefan Glasenhardt's workaround works perfectly.

I added the line:
Option "Legacy3d" "false"

to xorg.conf, which enabled hardware acceleration, allowing me to get to the desktop WITH compiz. Perfect!

I've attached the requested outputs.

VGA compatible controller: Intel Corporation 82865G Integrated Graphics Controller (rev 02)

am running the latest from the jaunty repositories and still no satisfaction with electricsheep screensaver, that uses mplayer to run an avi. It will only show up in a window and not full screen as it used to do with xorg from intrepid repositories. will not work when running a 2.6.29 kernel either.

but the good news is that all flash and other videos work, I only miss my full screen electricsheep.

I do have to set DRI to false within xorg.conf because otherwise i will never make it to the kde4 desktop. AND I have to set all video apps to not use "xv" because otherwise it will crash the desktop when i run a video. i currently have mplayer set to use "x11". "gl" is too slow.

do i need to buy an nvidea card!!!!!? I will not leave ubuntu!!

March 12, 11:37pm
I installed the update with an interesting result. Now when I run a video it
is exactly the same result as pressing Ctl-Alt-Bksp. Without any warning X
is gone and after a black screen the login screen comes up and then X. So
now running a video is a software X restart. If I can figure out what is the
trigger then it will be posted. Tomorrow I will try changing some of the
Options in xorg.conf.

2009/3/27 Bryce Harrington <email address hidden>

> PPA with this fix:
>
> https://edge.launchpad.net/%7Ebryceharrington/+archive/ppa/+sourcepub/535882/+listing-archive-extra
>
> ** Changed in: xserver-xorg-video-intel (Ubuntu Jaunty)
> Status: Triaged => In Progress
>
> --
> [i845G] Fatal server error: Couldn't bind memory for BO front buffer
> (Jaunty)
> https://bugs.launchpad.net/bugs/304871
> You received this bug notification because you are a direct subscriber
> of a duplicate bug.
>
> Status in X.org xf86-video-intel: Confirmed
> Status in “xserver-xorg-video-intel” source package in Ubuntu: In Progress
> Status in xserver-xorg-video-intel in Ubuntu Jaunty: In Progress
> Status in “xserver-xorg-video-intel” source package in Fedora: Fix Released
>
> Bug description:
> Since the change from the 2.4.1 driver to 2.5.1, Jaunty now hangs during
> boot, with "Fatal server error: Couldn't bind memory for BO front buffer"
> shown in the log. So far, this has been seen by two people with i845
> graphics and two with i865.
>
> Also discussed at this forum:
> https://bugs.freedesktop.org/show_bug.cgi?id=18974
>
> [Original Report]
> Binary package hint: xserver-xorg-video-intel
>
> Hi all,
> This is in jaunty, I get no screens.
>
>
>
> [lspci]
> 00:00.0 Host bridge [0600]: Intel Corporation 82845G/GL[Brookdale-G]/GE/PE
> DRAM Controller/Host-Hub Interface [8086:2560] (rev 03)
> 00:02.0 VGA compatible controller [0300]: Intel Corporation
> 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device [8086:2562]
> (rev 03)
> Subsystem: Intel Corporation 82845G/GL[Brookdale-G]/GE Chipset
> Integrated Graphics Device [8086:2562]
>
>

Intel Corporation 82865G Integrated Graphics Controller (rev 02)

the latest xorg upgrade from the jaunty repository now allows me to get into the kde4 desktop without having to set DRI to false. Yipee!

BUT, running mplayer with the "xv" driver, still, will knock me right out of kde4. and "xv" is the only way i can get electricsheep to run full screen. I have to set the driver to "x11" for videos to work. But as stated in previous posts, my electricsheep will only run in a window, not fullscreen, if set to "x11". It will run in fullscreen if I set mplayer to use "xv" driver and i revert to xorg from the intrepid repository.

anyway. congrads. you are half way there.

fix the "xv". please. I want my eye candy fullscreen.

and why is electricshhep not a default option for kde4? I had to hack it in with trial and error making a electricsheep.desktop file.

There seem to be different problems with Intel 865G video.

One is that DRI has to be disabled with
Option "DRI" "False" in /etc/xorg.conf
or with the patch from Bryce (Thanks)
http://launchpadlibrarian.net/24434513/116_846g_865g_disable_dri.patch
https://bugs.launchpad.net/bugs/304871.

Another is the acceleration. To play video (badly)
Option "NoAccel" "True" has to be set.

This setting is not practical for other use and not much good for video
other than it does not crash the system.

There are problems with all sorts of video, flash, mplayer, slide shows to
name three. Some lead to an instant restart of X, a software Ctl-Alt-Bksp,
while others simply do not play at all, cause lock up or display badly.

The system boots to lockup with UXA enabled.

Nothing additional in the logs.

If anyone has any ideas to try or tests to run please post them.

Alright, there are just waaaay too many different things going on with this bug.

The original problem which is targeted for jaunty is the fatal server error on 845.

The problem people are having with video, requiring NoAccel to be used is a separate bug. File a new report on that; don't comment further about it here.

From the comments, there are reporters for 845, 855, 865, and other chips, and the symptoms/fixes for each seem to be different.

845: From the last couple comments from 845 users jerrylamos and j-de-cuba it seems the issue on 845 is solved with the latest updates. Other 845 users including the original reporter have not commented further; I guess we have to assume they're not commenting because the issue is gone. So 845 is solved. Great.

855: This chip seems to be in much better shape. Many users find UXA solves the issue. Others not so much. The Legacy3D patch seems necessary for this. However I think we need to handle the 855 bugs separately from the 845 ones. So this needs to go to a different bug report.

865: From comments from randister and kitterman this may now be fixed with latest updates. Otherwise, disabling DRI solves the problem. Video is still broken but as I mentioned, that should be a separate bug.

bryce i just tested todays daily jaunty 32bits on an Dell GX270 with Intel 865 and it failed to boot the installer.
Only using Safe Graphics and having VESA on xorg.cong

FYI adding DRI false seems to work

I've unduped the non-845 bugs from this one. Non-845 users should unsub from this one and sub to the appropriate one:

bug 304871 - [i845] couldn't bind memory for BO backbuffer on 845

bug 322646 - [i855] couldn't bind memory for BO backbuffer on 855

bug 317457 - [i865] freeze, works only with dri false

bug 328528 - [i865] UXA locks up system (UXA bug)

If your issue doesn't fit into one of the above bugs, file a NEW one.

As per the last two comments from the two most recent 845 reporters, the issue on i845G chips has been resolved with latest updates as of 2009-03-21.

Please do not reopen this bug unless you have i845 graphics and see the exact error message "Couldn't bind memory for BO front buffer".

Thank you Bryce, et al.! Today's released fix worked perfectly. I can finally empty my xorg.conf file :)

Well, after this update the scrolling and window management behavior sucks.