[coreboot] Resume problems on Lenovo X60t

Sat Apr 22 09:11:50 CEST 2017

Dear coreboot folks,

Am Freitag, den 21.04.2017, 09:32 +0200 schrieb Paul Menzel:

> Am Donnerstag, den 20.04.2017, 21:35 +0200 schrieb Paul Menzel:
> 
> > With Linux 4.11-rcX sometimes the Lenovo X60t doesn’t correctly resume
> > anymore. I still have to do some more tests, but I believe I am unable
> > to reproduce this issue with Linux 4.10.8. But as I also do not know
> > how to reproduce it, despite doing suspend and resuming, and see how it
> > goes, I cannot know for sure. With Linux 4.11-rcX, I’d say the issue
> > happens one out of ten or 15 times.
> > 
> > ```
> > 12.951: [  350.898287] BUG: unable to handle kernel paging request at f8281008
> > 12.951: [  350.898346] IP: gen2_write32+0x62/0x130 [i915]
> > 12.951: [  350.898347] *pde = 34c88067 
> > 12.951: [  350.898349] *pte = 00000000 
> > 12.951: [  350.898349] 
> > 12.951: [  350.898352] Oops: 0002 [#1] SMP
> > 12.951: [  350.898353] Modules linked in: joydev(E) wacom_w8001(E) serport(E) cpufreq_powersave(E) cpufreq_conservative(E) cpufreq_userspace(E) iTCO_wdt(E) iTCO_vendor_support(E) acpi_cpufreq(E) coretemp(E) arc4
> > (E) kvm_intel(E) lpc_ich(E) kvm(E) mfd_core(E) irqbypass(E) iwl3945(E) iwlegacy(E) i915(E) evdev(E) snd_hda_codec_analog(E) snd_hda_codec_generic(E) snd_pcsp(E) pcmcia(E) serio_raw(E) mac80211(E) rng_core(E) snd
> > _hda_intel(E) yenta_socket(E) pcmcia_rsrc(E) drm_kms_helper(E) pcmcia_core(E) snd_hda_codec(E) snd_hda_core(E) snd_hwdep(E) thinkpad_acpi(E) snd_pcm(E) drm(E) cfg80211(E) battery(E) nvram(E) i2c_algo_bit(E) snd_
> > timer(E) fb_sys_fops(E) syscopyarea(E) sysfillrect(E) snd(E) rfkill(E) sysimgblt(E) soundcore(E) video(E) ac(E) button(E) shpchp(E) tpm_tis(E) tpm_tis_core(E) tpm(E) fuse(E) parport_pc(E)
> > 12.951: [  350.898397]  ppdev(E) lp(E) parport(E) autofs4(E) ext4(E) crc16(E) jbd2(E) fscrypto(E) mbcache(E) ecb(E) cbc(E) algif_skcipher(E) af_alg(E) dm_crypt(E) dm_mod(E) sg(E) sr_mod(E) sd_mod(E) cdrom(E) ata
> > _generic(E) sdhci_pci(E) psmouse(E) sdhci(E) uhci_hcd(E) e1000e(E) ata_piix(E) ahci(E) ehci_pci(E) firewire_ohci(E) libahci(E) i2c_i801(E) libata(E) ehci_hcd(E) ptp(E) mmc_core(E) firewire_core(E) crc_itu_t(E) s
> > csi_mod(E) usbcore(E) pps_core(E) thermal(E)
> > 12.951: [  350.898429] CPU: 1 PID: 1113 Comm: kworker/u4:32 Tainted: G            E   4.11.0-rc7 #23
> > 12.951: [  350.898431] Hardware name: LENOVO 636338U/636338U, BIOS CBET4000 4.5-1596-gccdb801 04/19/2017
> > 12.951: [  350.898436] Workqueue: events_unbound async_run_entry_fn
> > 12.951: [  350.898438] task: f2588480 task.stack: f258c000
> > 12.951: [  350.898483] EIP: gen2_write32+0x62/0x130 [i915]
> > 12.951: [  350.898484] EFLAGS: 00210282 CPU: 1
> > 12.951: [  350.898486] EAX: f8281008 EBX: f8d80eb0 ECX: 00000001 EDX: f8180000
> > 12.951: [  350.898487] ESI: f658403c EDI: 00000001 EBP: f258de44 ESP: f258de14
> > 12.951: [  350.898488]  DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
> > 12.951: [  350.898490] CR0: 80050033 CR2: f8281008 CR3: 17589000 CR4: 000006d0
> > 12.951: [  350.898492] Call Trace:
> > 12.951: [  350.898537]  ? i915_gem_restore_gtt_mappings+0x1ca/0x290 [i915]
> > 12.951: [  350.898581]  ? gen9_decoupled_read64+0x270/0x270 [i915]
> > 12.951: [  350.898622]  ? gen6_ggtt_invalidate+0x25/0x30 [i915]
> > 12.951: [  350.898665]  ? i915_gem_resume+0x2d/0x80 [i915]
> > 12.951: [  350.898701]  ? i915_drm_resume+0x38/0x180 [i915]
> > 12.952: [  350.898705]  ? pci_pm_resume+0x4b/0xc0
> > 12.952: [  350.898709]  ? dpm_run_callback+0x53/0x150
> > 12.952: [  350.898713]  ? wait_for_completion+0x2a/0x140
> > 12.952: [  350.898716]  ? pci_pm_thaw+0x90/0x90
> > 12.952: [  350.898718]  ? device_resume+0x87/0x170
> > 12.952: [  350.898720]  ? async_resume+0x1e/0x50
> > 12.952: [  350.898722]  ? async_run_entry_fn+0x35/0x190
> > 12.952: [  350.898726]  ? process_one_work+0x15f/0x3a0
> > 12.952: [  350.898729]  ? worker_thread+0x39/0x470
> > 12.952: [  350.898732]  ? kthread+0xdb/0x110
> > 12.952: [  350.898734]  ? process_one_work+0x3a0/0x3a0
> > 12.952: [  350.898736]  ? kthread_create_on_node+0x30/0x30
> > 12.952: [  350.898739]  ? ret_from_fork+0x1c/0x28
> > 12.952: [  350.898741] Code: d4 a6 e1 f8 bb 02 00 00 00 89 4c 24 08 89 5c 24 04 c7 04 24 49 d1 e0 f8 e8 bc 92 c8 ff 8b 97 d4 03 00 00 8b 45 e8 8b 7d e4 01 d0 <89> 38 83 c4 24 5b 5e 5f 5d c3 8d 74 26 00 64 8b 15 00 31 57 d7
> > 12.952: [  350.898813] EIP: gen2_write32+0x62/0x130 [i915] SS:ESP: 0068:f258de14
> > 12.952: [  350.898814] CR2: 00000000f8281008
> > 12.952: [  350.898817] ---[ end trace 478b15034b0b3e6a ]---
> > ```
> > 
> > Ticket #100739 in the Freedesktop.org Bugzilla [1] tracks this issue.
> > 
> > Chris Wilson replied already.
> > 
> > > The stacktrace is mostly garbage. A register mmio goes wrong. One of
> > > the suggestions is that the ioremap of the PCI bar is invalid upon
> > > resume.
> > 
> > coreboot has the TPM patches applied, and is built with native graphics
> > initialization.
> > 
> > Has anybody else experienced any issues on the Lenovo X60t?
> > 
> > In #coreboot at irc.freenode.net it was mentioned that there might be low
> > memory corruptions on the Intel 945 devices.
> > 
> > I’d welcome help also from “normal users” to reproduce this issue.
> 
> With the TPM changes (up for review) [2][3] reverted, I was unable to
> reproduce this issue with Linux 4.11-rc7+ today in ten suspend/resume
> cycles.

This morning the problem manifested itself with plain coreboot, that
means commit 8795feb (mb/lenovo/t420: Enable libgfxinit) with no TPM
changes.

Unfortunately, Linux wasn’t configured to use the serial console, so I
didn’t get a stacktrace.

Thanks,

Paul

> > [1] https://bugs.freedesktop.org/show_bug.cgi?id=100739
> [2] https://review.coreboot.org/19235/
> [3] https://review.coreboot.org/13410/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 195 bytes
Desc: This is a digitally signed message part
URL: <http://mail.coreboot.org/pipermail/coreboot/attachments/20170422/6f7e07df/attachment.asc>