[Bug 1104616] New: kernel 4.12.14-lp150.12.13-default: snd_hda_intel oops on load in virtualbox vm
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 Bug ID: 1104616 Summary: kernel 4.12.14-lp150.12.13-default: snd_hda_intel oops on load in virtualbox vm Classification: openSUSE Product: openSUSE Distribution Version: Leap 15.0 Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Kernel Assignee: kernel-maintainers@forge.provo.novell.com Reporter: seife@novell.slipkontur.de QA Contact: qa-bugs@suse.de CC: tiwai@suse.com Found By: --- Blocker: --- Since todays update to 4.12.14-lp150.12.13-default, on boot I get this oopses: [ 7.439097] BUG: unable to handle kernel NULL pointer dereference at 0000000000000034 [ 7.439136] IP: unlink_anon_vmas+0x3a/0x190 [ 7.439151] PGD 0 P4D 0 [ 7.439193] Oops: 0002 [#1] SMP PTI [ 7.439214] Modules linked in: snd_hda_intel snd_hda_codec snd_hda_core snd_hwdep snd_pcm snd_timer snd pcspkr joydev virtio_net video soundcore i2c_piix4 button battery ac dm_crypt sr_mod cdrom ata_generic hid_generic usbhid crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel ata_piix vboxvideo(O) aesni_intel ohci_pci ehci_pci ohci_hcd drm_kms_helper ehci_hcd ttm aes_x86_64 crypto_simd cryptd glue_helper serio_raw ahci drm drm_panel_orientation_quirks syscopyarea libahci sysfillrect sysimgblt usbcore fb_sys_fops vboxguest(O) dm_mirror dm_region_hash dm_log sg dm_multipath dm_mod scsi_dh_rdac scsi_dh_emc scsi_dh_alua [ 7.439414] CPU: 1 PID: 769 Comm: systemd-udevd Tainted: G O 4.12.14-lp150.12.13-default #1 openSUSE Leap 15.0 [ 7.439456] Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006 [ 7.439480] task: ffff88007ac28140 task.stack: ffffc900007d4000 [ 7.439504] RIP: 0010:unlink_anon_vmas+0x3a/0x190 [ 7.439530] RSP: 0018:ffffc900007d7c68 EFLAGS: 00010246 [ 7.439549] RAX: 0000000000000000 RBX: ffff88007c16ae00 RCX: 0000000000000000 [ 7.439572] RDX: 0000000000000000 RSI: ffff88007c1c0b38 RDI: 0000000000000001 [ 7.439597] RBP: 0000000000000000 R08: 00007f414e7bf000 R09: 0000000000000000 [ 7.439621] R10: ffff88007c16a120 R11: 000000000000000c R12: ffff880036b7d5f0 [ 7.439644] R13: ffff880036b7d578 R14: ffff88007c16a100 R15: ffff88007c1c0af8 [ 7.439702] FS: 0000000000000000(0000) GS:ffff88007fd00000(0000) knlGS:0000000000000000 [ 7.439749] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 7.439769] CR2: 0000000000000034 CR3: 000000007bf72003 CR4: 00000000000606e0 [ 7.439795] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 7.439823] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 7.439847] Call Trace: [ 7.439867] free_pgtables+0xb6/0x110 [ 7.439883] exit_mmap+0xa2/0x180 [ 7.439897] mmput+0x4a/0x130 [ 7.439920] flush_old_exec+0x4cc/0x6a0 [ 7.439942] load_elf_binary+0x24a/0x1050 [ 7.439959] ? _copy_from_user+0x2f/0x60 [ 7.440644] search_binary_handler+0x90/0x1c0 [ 7.441299] do_execveat_common.isra.36+0x687/0x810 [ 7.441933] SyS_execve+0x2c/0x40 [ 7.442594] do_syscall_64+0x7b/0x150 [ 7.443265] entry_SYSCALL_64_after_hwframe+0x3d/0xa2 [ 7.443890] RIP: 0033:0x7f415112f477 [ 7.444606] RSP: 002b:00007fff33740848 EFLAGS: 00000202 ORIG_RAX: 000000000000003b [ 7.445288] RAX: ffffffffffffffda RBX: 00007f4152304b70 RCX: 00007f415112f477 [ 7.445909] RDX: 00005617fcbda0c0 RSI: 00007fff33740900 RDI: 00007fff33741100 [ 7.446548] RBP: 0000000000000010 R08: 00005617fcbda0c0 R09: 000000000000001a [ 7.447177] R10: 0000000000000008 R11: 0000000000000202 R12: 0000000000000012 [ 7.447834] R13: 00007fff33740900 R14: 000000000000000f R15: 00007fff33742640 [ 7.448468] Code: 67 78 55 53 49 89 fd 48 83 ec 08 48 8b 57 78 49 39 d4 48 8b 02 0f 84 4c 01 00 00 4c 8d 72 f0 48 8d 58 f0 31 ff eb 1f 49 8b 47 38 <83> 68 34 01 48 8d 53 10 48 8b 4b 10 49 89 de 48 89 ef 4c 39 e2 [ 7.450271] RIP: unlink_anon_vmas+0x3a/0x190 RSP: ffffc900007d7c68 [ 7.450817] CR2: 0000000000000034 [ 7.451434] ---[ end trace 877104692e2cb656 ]--- [ 7.458225] BUG: unable to handle kernel NULL pointer dereference at 0000000000000034 [ 7.458979] IP: unlink_anon_vmas+0x3a/0x190 [ 7.459607] PGD 0 P4D 0 [ 7.460266] Oops: 0002 [#2] SMP PTI [ 7.460860] Modules linked in: snd_hda_intel snd_hda_codec snd_hda_core snd_hwdep snd_pcm snd_timer snd pcspkr joydev virtio_net video soundcore i2c_piix4 button battery ac dm_crypt sr_mod cdrom ata_generic hid_generic usbhid crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel ata_piix vboxvideo(O) aesni_intel ohci_pci ehci_pci ohci_hcd drm_kms_helper ehci_hcd ttm aes_x86_64 crypto_simd cryptd glue_helper serio_raw ahci drm drm_panel_orientation_quirks syscopyarea libahci sysfillrect sysimgblt usbcore fb_sys_fops vboxguest(O) dm_mirror dm_region_hash dm_log sg dm_multipath dm_mod scsi_dh_rdac scsi_dh_emc scsi_dh_alua [ 7.465051] CPU: 1 PID: 767 Comm: apparmor.system Tainted: G D O 4.12.14-lp150.12.13-default #1 openSUSE Leap 15.0 [ 7.466334] Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006 [ 7.466976] task: ffff88007bc50040 task.stack: ffffc900007b4000 [ 7.467631] RIP: 0010:unlink_anon_vmas+0x3a/0x190 [ 7.468316] RSP: 0018:ffffc900007b7b88 EFLAGS: 00010246 [ 7.469107] RAX: 0000000000000000 RBX: ffff88007c101980 RCX: 0000000000000000 [ 7.469818] RDX: 0000000000000000 RSI: ffff88007c1c0f70 RDI: 0000000000000001 [ 7.470508] RBP: 0000000000000000 R08: 00007f7acf87b000 R09: 0000000000000000 [ 7.471298] R10: ffff88007c101e60 R11: ffff8800370fbbe0 R12: ffff88007990b2d0 [ 7.471968] R13: ffff88007990b258 R14: ffff88007c101e40 R15: ffff88007c1c0f30 [ 7.472641] FS: 0000000000000000(0000) GS:ffff88007fd00000(0000) knlGS:0000000000000000 [ 7.473372] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 7.474069] CR2: 0000000000000034 CR3: 000000007b2da002 CR4: 00000000000606e0 [ 7.474770] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 7.475470] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 7.476138] Call Trace: [ 7.477422] free_pgtables+0xb6/0x110 [ 7.478183] exit_mmap+0xa2/0x180 [ 7.478842] mmput+0x4a/0x130 [ 7.479426] flush_old_exec+0x4cc/0x6a0 [ 7.479959] load_elf_binary+0x24a/0x1050 [ 7.480575] search_binary_handler+0x90/0x1c0 [ 7.481260] load_script+0x208/0x230 [ 7.481990] ? _copy_from_user+0x2f/0x60 [ 7.482624] search_binary_handler+0x90/0x1c0 [ 7.483222] do_execveat_common.isra.36+0x687/0x810 [ 7.483827] SyS_execve+0x2c/0x40 [ 7.484424] do_syscall_64+0x7b/0x150 [ 7.485012] entry_SYSCALL_64_after_hwframe+0x3d/0xa2 [ 7.485600] RIP: 0033:0x7f7ad083d477 [ 7.486173] RSP: 002b:00007fffc993ae88 EFLAGS: 00000206 ORIG_RAX: 000000000000003b [ 7.486744] RAX: ffffffffffffffda RBX: 0000562307578570 RCX: 00007f7ad083d477 [ 7.487234] RDX: 0000562307570940 RSI: 00005623075b1540 RDI: 0000562307578570 [ 7.487703] RBP: 00005623075b1540 R08: 00005623075b5ac0 R09: 0000000000000000 [ 7.488442] R10: 00000000000005b8 R11: 0000000000000206 R12: 00005623075b7360 [ 7.488991] R13: 0000562307570940 R14: 0000000000000000 R15: 00005623075b79c0 [ 7.489524] Code: 67 78 55 53 49 89 fd 48 83 ec 08 48 8b 57 78 49 39 d4 48 8b 02 0f 84 4c 01 00 00 4c 8d 72 f0 48 8d 58 f0 31 ff eb 1f 49 8b 47 38 <83> 68 34 01 48 8d 53 10 48 8b 4b 10 49 89 de 48 89 ef 4c 39 e2 [ 7.491172] RIP: unlink_anon_vmas+0x3a/0x190 RSP: ffffc900007b7b88 [ 7.491693] CR2: 0000000000000034 [ 7.492261] ---[ end trace 877104692e2cb657 ]--- last known good kernel is 4.12.14-lp150.12.7-default, I'm right now trying the in-between versions. This is in a Virtualbox 5.2.16 VM guest, Windows 10 host seife:~ # hwinfo --sound 12: PCI 05.0: 0403 Audio device [Created at pci.378] Unique ID: CvwD.LlSaw+h3QM7 SysFS ID: /devices/pci0000:00/0000:00:05.0 SysFS BusID: 0000:00:05.0 Hardware Class: sound Model: "Intel 82801FB/FBM/FR/FW/FRW (ICH6 Family) High Definition Audio Controller" Vendor: pci 0x8086 "Intel Corporation" Device: pci 0x2668 "82801FB/FBM/FR/FW/FRW (ICH6 Family) High Definition Audio Controller" SubVendor: pci 0x8384 "SigmaTel" SubDevice: pci 0x7680 Revision: 0x01 Driver: "snd_hda_intel" Driver Modules: "snd_hda_intel" Memory Range: 0xf0404000-0xf0407fff (rw,non-prefetchable) IRQ: 21 (8638 events) Module Alias: "pci:v00008086d00002668sv00008384sd00007680bc04sc03i00" Driver Info #0: Driver Status: snd_hda_intel is active Driver Activation Cmd: "modprobe snd_hda_intel" Config Status: cfg=new, avail=yes, need=no, active=unknown -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c1 --- Comment #1 from Takashi Iwai <tiwai@suse.com> --- There is no big changes in the audio side, and the stack trace looks really like a mm issue, or more likely some memory corruption. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c2 --- Comment #2 from Stefan Seyfried <seife@novell.slipkontur.de> --- 4.12.14-lp150.12.10-default does not exhibit this bug -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c3 --- Comment #3 from Stefan Seyfried <seife@novell.slipkontur.de> --- (In reply to Takashi Iwai from comment #1)
There is no big changes in the audio side, and the stack trace looks really like a mm issue, or more likely some memory corruption.
Any hints on how I could help to debug this? Magic boot parameters maybe? ;-) -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c4 --- Comment #4 from Takashi Iwai <tiwai@suse.com> --- (In reply to Stefan Seyfried from comment #2)
4.12.14-lp150.12.10-default does not exhibit this bug
Could you give the output of "rpm -qi kernel-default-4.12.14-lp150.12.10 | grep GIT" ? (In reply to Stefan Seyfried from comment #3)
(In reply to Takashi Iwai from comment #1)
There is no big changes in the audio side, and the stack trace looks really like a mm issue, or more likely some memory corruption.
Any hints on how I could help to debug this? Magic boot parameters maybe? ;-)
It'd be great if you could boot with kernel-debug, but I'm afraid that some KMP missing for vbox... -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c5 --- Comment #5 from Stefan Seyfried <seife@novell.slipkontur.de> --- seife@seife:~> rpm -qi kernel-default-4.12.14-lp150.12.10.1|grep GIT GIT Revision: da1ff4b723edfd60a96352f78b2a8bb3482369ac GIT Branch: openSUSE-15.0_EMBARGO Trying kerneld-debug right now (it should work without vbox-kmp, just no fancy graphics etc) -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c6 --- Comment #6 from Stefan Seyfried <seife@novell.slipkontur.de> --- it does not happen with kernel-debug. I'll try blacklisting virtualbox modules in default kernel, maybe that fixes it, too. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c7 Stefan Seyfried <seife@novell.slipkontur.de> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |Larry.Finger@gmail.com --- Comment #7 from Stefan Seyfried <seife@novell.slipkontur.de> --- I cannot clearly reproduce this after rebooting into the debug kernel, maybe also an issue with the virtualbox hypervisor. What I'm getting now on some boots is: [ 7.509824] BUG: Bad rss-counter state mm:ffff880036903b00 idx:0 val:910 [ 7.509854] BUG: Bad rss-counter state mm:ffff880036903b00 idx:1 val:102 [ 7.509877] BUG: non-zero nr_ptes on freeing mm: 20 with differing values and differing number of lines. When I get the BUG during boot, I'll get some backtraces shortly before shutdown/reboot. I do not get theses if I blacklist the virtualbox modules, even though they are loades later by the init scripts / systemd services. So "modprobe.blacklist=vboxguest" is enough to make this work for me. The virtualbox-guest-kmp-default is still built for an older kernel: seife@seife:~> rpm -q virtualbox-guest-kmp-default virtualbox-guest-kmp-default-5.2.16_k4.12.14_lp150.12.7-lp150.4.9.1.x86_64 so maybe just rebuilding it against the current kernel might help? -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c8 --- Comment #8 from Takashi Iwai <tiwai@suse.com> --- Let's try, although I'm not sure whether it's a problem in vbox driver. Since you didn't get a problem with the older kernel, something had changed apparently in the kernel side that triggers the problem... -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c9 --- Comment #9 from Takashi Iwai <tiwai@suse.com> --- BTW, to which kernel version was vbox KMP built? Please show the exact package version/release number of vbox kmp. It's possible that vbox package was built for some buggy kernel before the correction, and it resulted in a bad effect on the fixed kernel. If so, rebuilding vbox with the latest kernel should cure. Let's see. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c10 --- Comment #10 from Stefan Seyfried <seife@novell.slipkontur.de> --- seife@seife:~> rpm -q virtualbox-guest-kmp-default virtualbox-guest-kmp-default-5.2.16_k4.12.14_lp150.12.7-lp150.4.9.1.x86_64 -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c11 --- Comment #11 from Larry Finger <Larry.Finger@gmail.com> --- I have updated my Leap 15.0 VM running on a Leap 15.0 host. With "zypper dup", I got the following message: The following package update will NOT be installed: kernel-default-base-4.12.14-lp150.12.13.1 The following 4 NEW packages are going to be installed: kernel-default-4.12.14-lp150.12.13.1 kernel-default-base-4.12.14-lp150.12.10.1 kernel-default-devel-4.12.14-lp150.12.13.1 kernel-devel-4.12.14-lp150.12.13.1 Investigating as to why that kernel package was not being installed, "zypper up kernel-default-base-4.12.14-lp150.12.13.1" resulted in Problem: kernel-default-base-4.12.14-lp150.12.13.1.x86_64 conflicts with kernel-default = 4.12.14-lp150.12.13 provided by kernel-default-4.12.14-lp150.12.13.1.x86_64 Solution 1: deinstallation of kernel-default-4.12.14-lp150.12.13.1.x86_64 Solution 2: do not install kernel-default-base-4.12.14-lp150.12.13.1.x86_64 Choose from above solutions by number or cancel [1/2/c] (c): @Takashi: Do you understand the conflict here? After rebooting, "dmesg | grep vbox" resulted in the following: linux-o0pd:~ # dmesg -t | grep vbox vboxguest: loading out-of-tree module taints kernel. fb: switching to vboxdrmfb from VESA VGA fbcon: vboxdrmfb (fb0) is primary device vboxvideo 0000:00:02.0: fb0: vboxdrmfb frame buffer device [drm] Initialized vboxvideo 1.0.0 20130823 for 0000:00:02.0 on minor 0 vboxguest: misc device minor 57, IRQ 20, I/O port d020, MMIO at 00000000f0400000 (size 0x400000) vboxguest: Successfully loaded version 5.2.16_SUSE (interface 0x00010004) vboxsf: Successfully loaded version 5.2.16_SUSE (interface 0x00010004) The vbox guest versions are as follows: linux-o0pd:~ rpm -q virtualbox-guest-kmp-default virtualbox-guest-kmp-default-5.2.16_k4.12.14_lp150.12.7-lp150.4.9.1.x86_64 linux-o0pd:~ # rpm -q virtualbox-guest-tools virtualbox-guest-tools-5.2.16-lp150.4.9.1.x86_64 linux-o0pd:~ # rpm -q virtualbox-guest-x11 virtualbox-guest-x11-5.2.16-lp150.4.9.1.x86_64 As to the kernel, "uname -r" results in 4.12.14-lp150.12.13-default In short, my system is behaving as expected. @Stephan: What host OS are you running, and what version of VirtualBox? -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c12 --- Comment #12 from Stefan Seyfried <seife@novell.slipkontur.de> --- Larry, kernel-default-base is not supposed to be installed if you install kernel-default, it's a "smaller version" of kernel-default, so just leave it out. I'm running virtualbox 5.2.16 on Windows 10 (no idea which exact version). I'll not be in the office tomorrow, so can only get more info on Wednesday. The original oops from comment#1 is no longer reproducible, now I'm seeing the BUG: messages from comment#7 *if* i let the vboxguest/vboxsf load by udev or whatever loads them early. If vboxguest is blacklisted and thus only loaded later by init script/systemd service, then everything is fine. As Takashi suggested the original oops might be from memory corruption and the now appearing BUG: messages also point to memory corruption problems (at least to my untrained eye...), my guess was that this is some hidden incompatibility of the virtualbox-guest-kmp-default built against a slightly older kernel version that might have changed some internal KABI unnoticed or exposed a now fixed bug, thus my suggestion to rebuild virtualbox-guest-kmp against the current update kernel. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c13 --- Comment #13 from Larry Finger <Larry.Finger@gmail.com> --- Stephan, thanks for the info about kernel-default-base. If the kernel modules load, they should be OK. As long as neither the headers or gcc change between kernel updates, there should not be a problem. There should be no problem running virtualbox-guest-kmp-default-5.2.16_k4.12.14_lp150.12.7-lp150.4.9.1.x86_64 with kernel 4.12.14-lp150.12.13-default. Note that modules vboxsf and vboxvideo are built into kernel 4.12.14 in Leap 15. Only vboxguest is outside the kernel. That is why it taints the kernel. I used my Windows 10 Home Version 1803 to create a new Leap 15 VM using the DVD iso to install a KDE desktop. That VM was then updated to today's latest code, and the VM rebooted. None of the bugs you report show up. I then updated the Windows machine with KB4340917, which is the latest one published. After rebooting the host, the guest came up with no errors. The test also included mounting a shared folder. Finally, I tried all 3 types of audio emulation. None of them showed any bugs of the type you report. I am not sure where the problem lies, nor even how to try to debug it. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c14 --- Comment #14 from Larry Finger <Larry.Finger@gmail.com> --- Oracle has released Version 5.2.16 for VirtualBox with some fixes. That new version has been pushed to Leap 15.0 Update. I tested it with Windows 10 Home running 5.2.16 with a Leap 15.0 guest. Everything seemed to be working. Please test on your machine. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=1104616 http://bugzilla.opensuse.org/show_bug.cgi?id=1104616#c15 Stefan Seyfried <seife@novell.slipkontur.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution|--- |FIXED --- Comment #15 from Stefan Seyfried <seife@novell.slipkontur.de> --- No longer reproducible with the new (and built against current kernel) virtualbox-kmp package. Thanks! :-) -- You are receiving this mail because: You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply＠novell.com