On Mon, Dec 24, 2018 at 10:30:26AM -0800, Mike Kravetz wrote: > Adding people more familiar with this code and changes that went into 4.20. > > On 12/24/18 3:02 AM, Heiner Kallweit wrote: > > I just got the following error. It's for the first time that I see it. > > It happened whilst machine was idle, nothing special running. > > > > See also second error below. Not sure whether both errors may be related. > > > > [17932.571487] invalid opcode: 0000 [#1] SMP > > [17932.571518] CPU: 2 PID: 203 Comm: khugepaged Not tainted 4.20.0-rc7-next-20181221+ #2 > > [17932.571550] Hardware name: NA ZBOX-CI327NANO-GS-01/ZBOX-CI327NANO-GS-01, BIOS 5.12 04/26/2018 > > [17932.571614] RIP: 0010:khugepaged+0x2a2/0x2280 > > [17932.571640] Code: c0 48 8b 4d d0 65 48 33 0c 25 28 00 00 00 0f 85 22 1e 00 00 48 8d 65 d8 5b 41 5c 41 5d 41 5e 41 5f 5d c3 e8 30 43 ec ff e9 1e <fe> ff ff 48 8d 45 a0 48 8b 15 88 cc cc 00 49 c7 c4 90 f5 aa 85 48 > > [17932.571721] RSP: 0018:ffffab4b801f7dc0 EFLAGS: 00010286 > > [17932.571751] RAX: 0000000000000000 RBX: 0000000000002710 RCX: 0000000000000000 > > [17932.571786] RDX: 0000000000000000 RSI: 0000000000000001 RDI: ffff94c23a8c2a80 > > [17932.571820] RBP: ffffab4b801f7f9a R08: 0000000000000000 R09: 0000000000000000 > > [17932.571855] R10: 0000000000000000 R11: 0000000000000000 R12: ffffab4b801f7e20 > > [17932.571890] R13: ffffffff86862720 R14: 0000000000000000 R15: 00000000000000d0 > > [17932.571928] FS: 0000000000000000(0000) GS:ffff94c23bb00000(0000) knlGS:0000000000000000 > > [17932.571967] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 > > [17932.571997] CR2: 00007f20c458f012 CR3: 0000000171213000 CR4: 00000000003406e0 > > [17932.572034] Call Trace: > > [17932.572061] ? wait_woken+0xa0/0xa0 > > [17932.572088] ? kthread+0x126/0x140 > > [17932.572111] ? __collapse_huge_page_swapin+0x540/0x540 > > [17932.572141] ? kthread_create_on_node+0x60/0x60 > > [17932.572172] ? ret_from_fork+0x3a/0x50 > > [17932.572197] Modules linked in: snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_codec_generic vfat fat x86_pkg_temp_thermal realtek snd_hda_intel i2c_i801 i915 snd_hda_codec r8169 snd_hda_core intel_gtt i2c_algo_bit snd_pcm libphy drm_kms_helper snd_timer syscopyarea sysfillrect sysimgblt fb_sys_fops snd mei_me drm mei usb_storage crypto_user efivarfs ipv6 serio_raw atkbd libps2 xhci_pci xhci_hcd usbcore usb_common i8042 serio ext4 crc32c_intel mbcache jbd2 ahci libahci libata All code ======== 0: c0 48 8b 4d rorb $0x4d,-0x75(%rax) 4: d0 65 48 shlb 0x48(%rbp) 7: 33 0c 25 28 00 00 00 xor 0x28,%ecx e: 0f 85 22 1e 00 00 jne 0x1e36 14: 48 8d 65 d8 lea -0x28(%rbp),%rsp 18: 5b pop %rbx 19: 41 5c pop %r12 1b: 41 5d pop %r13 1d: 41 5e pop %r14 1f: 41 5f pop %r15 21: 5d pop %rbp 22:* c3 retq <-- trapping instruction 23: e8 30 43 ec ff callq 0xffffffffffec4358 28: e9 1e fe ff ff jmpq 0xfffffffffffffe4b 2d: 48 8d 45 a0 lea -0x60(%rbp),%rax 31: 48 8b 15 88 cc cc 00 mov 0xcccc88(%rip),%rdx # 0xccccc0 38: 49 c7 c4 90 f5 aa 85 mov $0xffffffff85aaf590,%r12 3f: 48 rex.W Code starting with the faulting instruction =========================================== 0: fe (bad) 1: ff (bad) 2: ff 48 8d decl -0x73(%rax) 5: 45 a0 48 8b 15 88 cc rex.RB movabs 0x4900cccc88158b48,%al c: cc 00 49 f: c7 c4 90 f5 aa 85 mov $0x85aaf590,%esp 15: 48 rex.W It looks like somebody jumped into the middle of the instruction. I have no idea why and how. Maybe stack corruption? -- Kirill A. Shutemov
