Created attachment 41354 [details] reg dump Ubuntu 10.10 kernel: 2.6.37-10-generic xorg-intel driver: 2.13.902 git20101220 Machine: Asus eeePC 1000HE agpgart-intel 0000:00:00.0: Intel 945GME Chipset I was trying the git version of the driver and got the error in the message summary repeatedly in Xorg.0.log. When gdm attempts to start, I get a blank screen with a mouse cursor. Eventually it fails. https://bugs.launchpad.net/ubuntu/+source/unity/+bug/657976/comments/74
Created attachment 41355 [details] xorg log
Created attachment 41356 [details] dmesg
Ignore this, it was a fun bug with driver 2.12 or drm on the stock ubuntu kernel/xorg. I didn't mean to include it in this bug report. > https://bugs.launchpad.net/ubuntu/+source/unity/+bug/657976/comments/74
Something is very upset, can you step through and see why we're provoking the EINVAL? The only thing I can immediately think of is trying to use the BLT ring which is just so very wrong.
Just give let me know how you want me to step through it.
Attach gdb to the xserver, b intel_batch_submit keep pressing step until you see a return -EINVAL and the print out the locals and the bt. Or perhaps more simply do a breakpoint on drm_intel_gem_bo_mrb_exec2 and print out the locals. At any rate the only way I can think of for an EINVAL to be returned by, as it appears to be, libdrm is for us to be trying to use the BLT ring. That is impossible in the upstream sources, unless your gen3 device is masquerading as a gen6.
(gdb) step intel_batch_submit (scrn=0x94108e0, flush=1) at ../../src/intel_batchbuffer.c:199 199 ../../src/intel_batchbuffer.c: No such file or directory. in ../../src/intel_batchbuffer.c (gdb) step 210 in ../../src/intel_batchbuffer.c (gdb) step 211 in ../../src/intel_batchbuffer.c (gdb) step 215 in ../../src/intel_batchbuffer.c (gdb) step 212 in ../../src/intel_batchbuffer.c (gdb) step 218 in ../../src/intel_batchbuffer.c (gdb) step 219 in ../../src/intel_batchbuffer.c (gdb) step 230 in ../../src/intel_batchbuffer.c (gdb) step xf86DrvMsg (scrnIndex=0, type=X_ERROR, format=0x371b48 "Failed to submit batch buffer, expect rendering corruption or even a frozen display: %s.\n") at ../../../../hw/xfree86/common/xf86Helper.c:1254 1254 ../../../../hw/xfree86/common/xf86Helper.c: No such file or directory. in ../../../../hw/xfree86/common/xf86Helper.c (gdb) bt #0 xf86DrvMsg (scrnIndex=0, type=X_ERROR, format=0x371b48 "Failed to submit batch buffer, expect rendering corruption or even a frozen display: %s.\n") at ../../../../hw/xfree86/common/xf86Helper.c:1254 #1 0x0033d753 in intel_batch_submit (scrn=0x94108e0, flush=1) at ../../src/intel_batchbuffer.c:230 #2 0x003447e7 in intel_uxa_block_handler (intel=0x94111d0) at ../../src/intel_uxa.c:947 #3 0x00341c4e in I830BlockHandler (i=0, blockData=0x0, pTimeout=0xbf87b8fc, pReadmask=0x8204c80) at ../../src/intel_driver.c:736 #4 0x081bce3b in AnimCurScreenBlockHandler (screenNum=0, blockData=0x0, pTimeout=0xbf87b8fc, pReadmask=0x8204c80) at ../../render/animcur.c:194 #5 0x080eacee in compBlockHandler (i=0, blockData=0x0, pTimeout=0xbf87b8fc, pReadmask=0x8204c80) at ../../composite/compinit.c:157 #6 0x08073368 in BlockHandler (pTimeout=0xbf87b8fc, pReadmask=0x8204c80) at ../../dix/dixutils.c:385 #7 0x0809fcbc in WaitForSomething (pClientsReady=0x96cc300) at ../../os/WaitFor.c:216 #8 0x0806dd3e in Dispatch () at ../../dix/dispatch.c:368 #9 0x080625ba in main (argc=8, argv=0xbf87ba64, envp=0xbf87ba88) at ../../dix/main.c:291 (gdb)
Hallo. I'm sorry, but I've same problems/symptoms on my EeeBox B202. DRM, intel driver are latest git versions. Distro is Ubuntu 10.10. Kernel is vanilla 2.6.37. I've attached lspci -v, dmesg and xorg.log outputs. With 2.13 driver X works, but i get some dmesg: [drm:i915_gem_mmap_gtt_ioctl] *ERROR* Attempting to mmap a purgeable buffer [drm:i915_gem_mmap_gtt_ioctl] *ERROR* Attempting to mmap a purgeable buffer [drm:i915_gem_mmap_gtt_ioctl] *ERROR* Attempting to mmap a purgeable buffer [drm:i915_gem_mmap_gtt_ioctl] *ERROR* Attempting to mmap a purgeable buffer and xorg errors: [ 970.039] (WW) intel(0): intel_uxa_prepare_access: bo map failed: Invalid argument [ 970.053] (WW) intel(0): intel_uxa_prepare_access: bo map failed: Invalid argument [ 970.053] (WW) intel(0): intel_uxa_prepare_access: bo map failed: Invalid argument [ 970.075] (WW) intel(0): intel_uxa_prepare_access: bo map failed: Invalid argument and sometimes X restarts unexpectedly, but that is another story. Thank you, kind regards Milan Oravec
Created attachment 41702 [details] xorg.log
Created attachment 41703 [details] dmesg
Created attachment 41704 [details] lspci
(In reply to comment #8) > Hallo. I'm sorry, but I've same problems/symptoms on my EeeBox B202. No, you don't. That's a completely different bug.
(In reply to comment #8) > and xorg errors: > > [ 970.039] (WW) intel(0): intel_uxa_prepare_access: bo map failed: Invalid > argument > [ 970.053] (WW) intel(0): intel_uxa_prepare_access: bo map failed: Invalid > argument > [ 970.053] (WW) intel(0): intel_uxa_prepare_access: bo map failed: Invalid > argument > [ 970.075] (WW) intel(0): intel_uxa_prepare_access: bo map failed: Invalid > argument Where are those errors? Otherwise the log file is the same symptom of something screwed up on your system (libdrm falsely reporting an EINVAL).
(In reply to comment #12) > (In reply to comment #8) > > Hallo. I'm sorry, but I've same problems/symptoms on my EeeBox B202. > > No, you don't. That's a completely different bug. Is it reported (known) as an bug?
(In reply to comment #13) > (In reply to comment #8) > > and xorg errors: > > > > [ 970.039] (WW) intel(0): intel_uxa_prepare_access: bo map failed: Invalid > > argument > > [ 970.053] (WW) intel(0): intel_uxa_prepare_access: bo map failed: Invalid > > argument > > [ 970.053] (WW) intel(0): intel_uxa_prepare_access: bo map failed: Invalid > > argument > > [ 970.075] (WW) intel(0): intel_uxa_prepare_access: bo map failed: Invalid > > argument > > Where are those errors? Otherwise the log file is the same symptom of something > screwed up on your system (libdrm falsely reporting an EINVAL). it is from xorg.log with latest drm from git. I'll attach complete log.
Created attachment 41711 [details] Xorg log with 2.13 ver
hallo, is there any way to point out what causes this failure? With 2.14 final driver is it the same, only mouse cursor appears and sometimes some fragments of windows from saved session. I've tried 2.6.36.3 kernel too. Than you very much, best regards Milan
Everyone should have had time to update their systems by now...
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.