STATUS: FAILED BUILD: master: 8b7116d25ed6255a03895d835e5a0560858ab496 ENVIRONMENT: genericx86-64 on Sugarbay STEPS TO REPRODUCE: 1.run " dmesg | grep -i error" in the terminal. 2.check if there is any error log printed. OUTCOME: [ 2.328502] v86d[103]: segfault at 0 ip 0000000000401749 sp 00007fffcab15990 error 4 in v86d[400000+19000] [ 7.330763] uvesafb: probe of uvesafb.0 failed with error -22
which image was used here? Please include that in all bug reports.
A core-image-sato-sdk.
Appears also on 1.8.M2.RC2 (22c38e2d8a058619e3d05534b5e830b750aa4241)
I have not spent much time on it. This needs further debugging.
Created attachment 2429 [details] More errors found in the logs.
I just checked this on the Sugarbay machine I was using to look at the tar related issue and did not see this error with that build, please verify this is no longer an issue
This issue appears on 1.9_M1.rc2: f9ea480e6c26fea46c3011c4e9d0e2de3d5c81b4, genericx86-64 on NUC and MinnowMax and on genericx86 on MinnowMax. Log: dmesg_output.log ----------------------- Central error: [ 6.843864] uvesafb: vbe_init() failed with -22 *********************** [ 3.261093] scsi host4: usb-storage 1-4:1.0 [ 4.367385] scsi 4:0:0:0: Direct-Access Kingston DataTraveler 3.0 PMAP PQ: 0 ANSI: 6 [ 4.883531] sd 4:0:0:0: [sdb] 15155200 512-byte logical blocks: (7.75 GB/7.22 GiB) [ 4.885440] sd 4:0:0:0: [sdb] Write Protect is off [ 4.885447] sd 4:0:0:0: [sdb] Mode Sense: 23 00 00 00 [ 4.887245] sd 4:0:0:0: [sdb] No Caching mode page found [ 4.887283] sd 4:0:0:0: [sdb] Assuming drive cache: write through [ 4.968561] sdb: [ 4.972900] sd 4:0:0:0: [sdb] Attached SCSI removable disk [ 6.843790] uvesafb: Getting VBE info block failed (eax=0x4f00, err=1) [ 6.843864] uvesafb: vbe_init() failed with -22 [ 6.843916] uvesafb: probe of uvesafb.0 failed with error -22 [ 6.963980] udevd[119]: starting version 182 [ 7.012946] snd_hda_intel 0000:00:03.0: enabling device (0000 -> 0002) [ 7.013086] snd_hda_intel 0000:00:1b.0: enabling device (0000 -> 0002) [ 7.043234] sound hdaudioC1D0: autoconfig: line_outs=1 (0x21/0x0/0x0/0x0/0x0) type:hp [ 7.043238] sound hdaudioC1D0: speaker_outs=0 (0x0/0x0/0x0/0x0/0x0) [ 7.043240] sound hdaudioC1D0: hp_outs=0 (0x0/0x0/0x0/0x0/0x0) [ 7.043242] sound hdaudioC1D0: mono: mono_out=0x0 [ 7.043244] sound hdaudioC1D0: inputs: [ 7.043246] sound hdaudioC1D0: Mic=0x19 *********************** Central error: [ 6.843790] uvesafb: Getting VBE info block failed (eax=0x4f00, err=1) *********************** [ 3.260626] usb-storage 1-4:1.0: USB Mass Storage device detected [ 3.261093] scsi host4: usb-storage 1-4:1.0 [ 4.367385] scsi 4:0:0:0: Direct-Access Kingston DataTraveler 3.0 PMAP PQ: 0 ANSI: 6 [ 4.883531] sd 4:0:0:0: [sdb] 15155200 512-byte logical blocks: (7.75 GB/7.22 GiB) [ 4.885440] sd 4:0:0:0: [sdb] Write Protect is off [ 4.885447] sd 4:0:0:0: [sdb] Mode Sense: 23 00 00 00 [ 4.887245] sd 4:0:0:0: [sdb] No Caching mode page found [ 4.887283] sd 4:0:0:0: [sdb] Assuming drive cache: write through [ 4.968561] sdb: [ 4.972900] sd 4:0:0:0: [sdb] Attached SCSI removable disk [ 6.843790] uvesafb: Getting VBE info block failed (eax=0x4f00, err=1) [ 6.843864] uvesafb: vbe_init() failed with -22 [ 6.843916] uvesafb: probe of uvesafb.0 failed with error -22 [ 6.963980] udevd[119]: starting version 182 [ 7.012946] snd_hda_intel 0000:00:03.0: enabling device (0000 -> 0002) [ 7.013086] snd_hda_intel 0000:00:1b.0: enabling device (0000 -> 0002) [ 7.043234] sound hdaudioC1D0: autoconfig: line_outs=1 (0x21/0x0/0x0/0x0/0x0) type:hp [ 7.043238] sound hdaudioC1D0: speaker_outs=0 (0x0/0x0/0x0/0x0/0x0) [ 7.043240] sound hdaudioC1D0: hp_outs=0 (0x0/0x0/0x0/0x0/0x0) [ 7.043242] sound hdaudioC1D0: mono: mono_out=0x0 [ 7.043244] sound hdaudioC1D0: inputs:
reopening based on Cristina's comments
Has this issue been seen again? I have not been able to reproduce this.
It appears also on 1.9_M3.rc1: ea562de57590c966cd5a75fda8defecd397e6436 poky commit, core-image-sato-sdk image for genericx86-64 tested on NUC and MinnowMax and genericx86 on MinnowMax. This is the error: Central error: [ 6.904144] uvesafb: probe of uvesafb.0 failed with error -22 *********************** [ 3.971505] scsi 4:0:0:0: Direct-Access Kingston DataTraveler 3.0 PMAP PQ: 0 ANSI: 6 [ 4.509839] sd 4:0:0:0: [sdb] 15155200 512-byte logical blocks: (7.75 GB/7.22 GiB) [ 4.511750] sd 4:0:0:0: [sdb] Write Protect is off [ 4.511756] sd 4:0:0:0: [sdb] Mode Sense: 23 00 00 00 [ 4.513485] sd 4:0:0:0: [sdb] No Caching mode page found [ 4.513505] sd 4:0:0:0: [sdb] Assuming drive cache: write through [ 4.598734] sdb: [ 4.602942] sd 4:0:0:0: [sdb] Attached SCSI removable disk [ 6.904025] uvesafb: Getting VBE info block failed (eax=0x4f00, err=1) [ 6.904094] uvesafb: vbe_init() failed with -22 [ 6.904144] uvesafb: probe of uvesafb.0 failed with error -22 [ 7.034071] udevd[119]: starting version 182 [ 7.085258] snd_hda_intel 0000:00:03.0: enabling device (0000 -> 0002) [ 7.085395] snd_hda_intel 0000:00:1b.0: enabling device (0000 -> 0002) [ 7.085417] snd_hda_intel 0000:00:03.0: bound 0000:00:02.0 (ops i915_audio_component_bind_ops) [ 7.136159] snd_hda_codec_realtek hdaudioC1D0: autoconfig for ALC283: line_outs=1 (0x21/0x0/0x0/0x0/0x0) type:hp [ 7.136163] snd_hda_codec_realtek hdaudioC1D0: speaker_outs=0 (0x0/0x0/0x0/0x0/0x0) [ 7.136166] snd_hda_codec_realtek hdaudioC1D0: hp_outs=0 (0x0/0x0/0x0/0x0/0x0) [ 7.136168] snd_hda_codec_realtek hdaudioC1D0: mono: mono_out=0x0 [ 7.136169] snd_hda_codec_realtek hdaudioC1D0: inputs: [ 7.136172] snd_hda_codec_realtek hdaudioC1D0: Mic=0x19
After some further investigation and ensuring reproducible, the root cause seems to be that the generix86* BSPs uses the v86d in MACHINE_EXTRA_RRECOMMENDS and when used with systems that have non-vesa framebuffers will fail. Since the v86d package installs either a fbsetup script (for sysvinit) or a modprobe.d/uvesafb.conf for systemd. These need to be modified to check for an existing framebuffer driver is installed.
- "It appears also on 1.9_M3.rc1: ea562de57590c966cd5a75fda8defecd397e6436 poky commit, core-image-sato-sdk image for genericx86-64 tested on NUC and MinnowMax and genericx86 on MinnowMax." Will checkout this head and do some debugging with generic-x86 & MinnowMax for a solution.
This can also reproduced on jethro. Maybe a short-term solution is just disable uvesafb module on non-emu machines as we discussed. I am doing some debugging today towards to these steps and more clues in logs: rmmod uvesafb modprobe uvesafb root@genericx86-64:~# cat /var/log/messages Nov 24 12:03:25 genericx86-64 user.warn v86d: EBDA too big (38800), truncating. Nov 24 12:03:25 genericx86-64 user.err v86d: Video BIOS not found at c0000. Nov 24 12:03:25 genericx86-64 user.err v86d: v86 memory initialization failed. Nov 24 12:03:30 genericx86-64 user.err kernel: uvesafb: Getting VBE info block failed (eax=0x4f00, err=1) Nov 24 12:03:30 genericx86-64 user.err kernel: uvesafb: vbe_init() failed with -22 Nov 24 12:03:30 genericx86-64 user.warn kernel: uvesafb: probe of uvesafb.0 failed with error -22 Nov 24 12:03:34 genericx86-64 authpriv.notice login[758]: ROOT LOGIN on '/dev/ttyS0'
After some investigation, the root cause for this issue is: Vx86d tries to access to physical address 0xc0000 for video bios and also expects region from 0xa0000 as video ram. Clarified by FW team, Minnow Board Max BIOS is pure EFI-based and legacy free firmware. That means it doesn't support video bios and ram at these addresses. I think we can disable uvesa on minnow Max or degrade this bug since I don't see any functional impact from this error message.
patch has been submitted to poky mailing list today.
patch has been merged into poky master brand: http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=6af89812e8a9931ffed63768ed85367519bf7aef close it.
This should be backported to jethro branch.
Verified with master (m2_rc1) and jethro (2.0.1.rc6)
Tried on actual hardware minnowmax. same error log is comming. STATUS: FAILED BUILD: 2.3.4 RC1: ebb42af2829edfca1a23c7a51a431c656ffc2090 IMAGE: core-image-sato-sdk_generic-x86_64 ENVIRONMENT: generic-x86_64 on minnowmax STEPS TO REPRODUCE: Parse dmesg log and all log files in /var/log for k keywords like: "error" , "failed" , "cannot", "can't" etc. Filter out the known and invalid errors or errors that appear in certain situations and need to be ignored. I have attached test_image log.
Created attachment 4326 [details] log.do_testimage
This is a slightly different error message compared to the original bug. Here, the kernel has uvesafb in it but the userspace does not, its missing v86d. This is expected given the original change to remove it from the genericx86 images. This is therefore not a bug.