[WHL HDA] firmware boot failure on WHL platform.

[Jiangxinx]

summary:
firmware boot failure on WHL platform. According to the call stack information,it seems to be caused by commit 9c6b980

Step:
1.Aplay -l

Output:

aplay: device_list:270: no soundcards found...

Dmesg:

[    3.474102] BUG: unable to handle kernel NULL pointer dereference at 0000000000000008
[    3.474102] #PF error: [normal kernel read fault]
[    3.474102] PGD 0 P4D 0
[    3.474104] Oops: 0000 [#1] SMP NOPTI
[    3.474106] CPU: 1 PID: 2076 Comm: irq/144-AudioDS Not tainted 5.0.0-daily-hda-20190402 #f93c4ee9
[    3.474106] Hardware name: Intel Corporation CoffeeLake Client Platform/WhiskeyLake U DDR4 ERB, BIOS CNLSFWR1.R00.X137.B00.1803280218 03/28/2018
[    3.474110] RIP: 0010:hda_dsp_ipc_get_reply+0x35/0x130 [snd_sof_intel_hda_common]
[    3.474110] Code: 53 48 89 fb 48 83 ec 18 48 8b af 88 01 00 00 4c 89 e7 65 48 8b 04 25 28 00 00 00 48 89 44 24 10 31 c0 e8 6e a8 ea cf 49 89 c6 <48> 8b 45 08 81 78 04 00 00 01 40 0f 85 8d 00 00 00 48 8b 55 20 48
[    3.474111] RSP: 0018:ffffb9be41017e28 EFLAGS: 00010046
[    3.474112] RAX: 0000000000000246 RBX: ffff9f3cdf881028 RCX: ffffffffc0304240
[    3.474112] RDX: 0000000000000001 RSI: 0000000000000286 RDI: ffff9f3cdf881030
[    3.474113] RBP: 0000000000000000 R08: 0000000000000002 R09: ffff9f3ce3a60a40
[    3.474114] R10: ffffb9be412a7e08 R11: 0000000000000004 R12: ffff9f3cdf881030
[    3.474114] R13: 0000000000000002 R14: 0000000000000246 R15: ffff9f3cdf5f0000
[    3.474115] FS:  0000000000000000(0000) GS:ffff9f3ce3a40000(0000) knlGS:0000000000000000
[    3.474116] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    3.474116] CR2: 0000000000000008 CR3: 00000002606c8005 CR4: 00000000003606e0
[    3.474117] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[    3.474117] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[    3.474117] Call Trace:
[    3.474121]  ? snd_sof_dsp_update_bits+0x51/0x70 [snd_sof]
[    3.474123]  cnl_ipc_irq_thread+0x149/0x9f0 [snd_sof_intel_hda_common]
[    3.474126]  ? irq_forced_thread_fn+0x70/0x70
[    3.474127]  irq_thread_fn+0x1c/0x60
[    3.474129]  irq_thread+0xe2/0x160
[    3.474130]  ? wake_threads_waitq+0x30/0x30
[    3.474132]  ? irq_thread_dtor+0x90/0x90
[    3.474133]  kthread+0x10e/0x130
[    3.474134]  ? kthread_park+0x80/0x80
[    3.474136]  ret_from_fork+0x35/0x40
[    3.474137] Modules linked in: snd_soc_hdac_hdmi(+) snd_hda_codec_realtek snd_hda_codec_generic snd_soc_dmic sof_pci_dev snd_sof_intel_hda_common snd_soc_hdac_hda iwlmvm(+) snd_sof_intel_hda snd_sof_intel_byt snd_sof_xtensa_dsp snd_sof snd_soc_acpi_intel_match snd_soc_acpi snd_hda_ext_core snd_soc_core snd_hda_codec snd_hwdep snd_hda_core snd_pcm x86_pkg_temp_thermal iwlwifi intel_lpss_pci intel_lpss mfd_core efivarfs sdhci_pci xhci_pci cqhci sdhci xhci_hcd
[    3.474148] CR2: 0000000000000008
[    3.474149] ---[ end trace b2191f1416986bb1 ]---
[    3.474151] RIP: 0010:hda_dsp_ipc_get_reply+0x35/0x130 [snd_sof_intel_hda_common]
[    3.474151] Code: 53 48 89 fb 48 83 ec 18 48 8b af 88 01 00 00 4c 89 e7 65 48 8b 04 25 28 00 00 00 48 89 44 24 10 31 c0 e8 6e a8 ea cf 49 89 c6 <48> 8b 45 08 81 78 04 00 00 01 40 0f 85 8d 00 00 00 48 8b 55 20 48
[    3.474152] RSP: 0018:ffffb9be41017e28 EFLAGS: 00010046
[    3.474153] RAX: 0000000000000246 RBX: ffff9f3cdf881028 RCX: ffffffffc0304240
[    3.474153] RDX: 0000000000000001 RSI: 0000000000000286 RDI: ffff9f3cdf881030
[    3.474154] RBP: 0000000000000000 R08: 0000000000000002 R09: ffff9f3ce3a60a40
[    3.474154] R10: ffffb9be412a7e08 R11: 0000000000000004 R12: ffff9f3cdf881030
[    3.474155] R13: 0000000000000002 R14: 0000000000000246 R15: ffff9f3cdf5f0000
[    3.474156] FS:  0000000000000000(0000) GS:ffff9f3ce3a40000(0000) knlGS:0000000000000000
[    3.474157] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    3.474157] CR2: 0000000000000008 CR3: 00000002606c8005 CR4: 00000000003606e0
[    3.474158] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[    3.474158] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[    3.474161] BUG: unable to handle kernel NULL pointer dereference at 0000000000000000
[    3.474161] #PF error: [INSTR]
[    3.474161] PGD 0 P4D 0
[    3.474163] Oops: 0010 [#2] SMP NOPTI
[    3.474164] CPU: 1 PID: 2076 Comm: irq/144-AudioDS Tainted: G      D           5.0.0-daily-hda-20190402 #f93c4ee9
[    3.474164] Hardware name: Intel Corporation CoffeeLake Client Platform/WhiskeyLake U DDR4 ERB, BIOS CNLSFWR1.R00.X137.B00.1803280218 03/28/2018
[    3.474165] RIP: 0010:          (null)
[    3.474167] Code: Bad RIP value.
[    3.474167] RSP: 0018:ffffb9be41017ea8 EFLAGS: 00010282
[    3.474168] RAX: 0000000000000000 RBX: ffff9f3cdf5f0000 RCX: 0000000000000000
[    3.474169] RDX: ffffb9be41017ec8 RSI: 0000000000000000 RDI: ffffb9be41017ec8
[    3.474169] RBP: ffffffff912d69c0 R08: 0000000000000000 R09: 0000000000000000
[    3.474169] R10: 0000000000000246 R11: 0000000000000000 R12: 0000000000000000
[    3.474170] R13: ffff9f3cdf5f0704 R14: 0000000000000001 R15: ffff9f3cdf5f06d0
[    3.474171] FS:  0000000000000000(0000) GS:ffff9f3ce3a40000(0000) knlGS:0000000000000000
[    3.474171] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    3.474172] CR2: ffffffffffffffd6 CR3: 00000002606c8005 CR4: 00000000003606e0
[    3.474172] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[    3.474173] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[    3.474173] Call Trace:
[    3.474174]  ? task_work_run+0x79/0xa0
[    3.474176]  ? do_exit+0x2ca/0xbc0
[    3.474178]  ? irq_thread_dtor+0x90/0x90
[    3.474179]  ? kthread+0x10e/0x130
[    3.474180]  ? rewind_stack_do_exit+0x17/0x20
[    3.474181] Modules linked in: snd_soc_hdac_hdmi(+) snd_hda_codec_realtek snd_hda_codec_generic snd_soc_dmic sof_pci_dev snd_sof_intel_hda_common snd_soc_hdac_hda iwlmvm(+) snd_sof_intel_hda snd_sof_intel_byt snd_sof_xtensa_dsp snd_sof snd_soc_acpi_intel_match snd_soc_acpi snd_hda_ext_core snd_soc_core snd_hda_codec snd_hwdep snd_hda_core snd_pcm x86_pkg_temp_thermal iwlwifi intel_lpss_pci intel_lpss mfd_core efivarfs sdhci_pci xhci_pci cqhci sdhci xhci_hcd
[    3.474186] CR2: 0000000000000000
[    3.474187] ---[ end trace b2191f1416986bb2 ]---
[    3.474188] RIP: 0010:hda_dsp_ipc_get_reply+0x35/0x130 [snd_sof_intel_hda_common]
[    3.474189] Code: 53 48 89 fb 48 83 ec 18 48 8b af 88 01 00 00 4c 89 e7 65 48 8b 04 25 28 00 00 00 48 89 44 24 10 31 c0 e8 6e a8 ea cf 49 89 c6 <48> 8b 45 08 81 78 04 00 00 01 40 0f 85 8d 00 00 00 48 8b 55 20 48
[    3.474189] RSP: 0018:ffffb9be41017e28 EFLAGS: 00010046
[    3.474190] RAX: 0000000000000246 RBX: ffff9f3cdf881028 RCX: ffffffffc0304240
[    3.474190] RDX: 0000000000000001 RSI: 0000000000000286 RDI: ffff9f3cdf881030
[    3.474191] RBP: 0000000000000000 R08: 0000000000000002 R09: ffff9f3ce3a60a40
[    3.474191] R10: ffffb9be412a7e08 R11: 0000000000000004 R12: ffff9f3cdf881030
[    3.474192] R13: 0000000000000002 R14: 0000000000000246 R15: ffff9f3cdf5f0000
[    3.474193] FS:  0000000000000000(0000) GS:ffff9f3ce3a40000(0000) knlGS:0000000000000000
[    3.474193] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    3.474194] CR2: ffffffffffffffd6 CR3: 00000002606c8005 CR4: 00000000003606e0
[    3.474194] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[    3.474195] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[    3.474195] Fixing recursive fault but reboot is needed!
[    3.478124] HDMI HDA Codec ehdaudio0D2: Max dais supported: 3

Env:
sof master: acf2bd5
kernel sof-dev: f93c4ee
tplg: sof-hda-generic.tplg

[lgirdwood]

@Jiangxinx can you attach full log. This may be a FW issue if we dont have ROM errors.

[emilchudzik]

Double check issue on my side:

python tests in windows HDA works without issues. (FW: master/d7b36bf3).

• loading FW
• basic playback and record on HDA (loopback on codec ALC700)

linux Env:
FW: master/d7b36bf3
tplg: master/d7b36bf3 -> sof-hda-generic.tplg
kernel: sof-dev/e29122d
HW: WHL (HD-A mode) with ALC700

The same as I described on thesofproject/sof#1173 (comment)

• loading is ok
• aplay -l return devices from topology as expected
• run "aplay" command return error => unable to playback

[wenqingfu]

it seems to be caused by commit 9c6b980

9c6b980 is part of #740 patchset, so @lyakh

[ranj063]

@Jiangxinx @emilchudzik can you please check if linux PR #775 helps?

[xiulipan]

@emilchudzik @ranj063 @Jiangxinx
e29122d is some older version than 9c6b980

So it should be OK.

f3adfd6 (HEAD -> sof-dev, sof/topic/sof-dev) ASoC:sof: remove duplicate posn message in kernel log
0967e07 ASoC: SOF: control: fix a PM put missing at error
d6d661c ASoC: SOF: Intel: fix period_bytes calculation at hw_params()
f93c4ee Merge pull request #740 from lyakh/ipc-20190321
f304b5c ASoC: sof: use iopoll.h macro for polled register reads
1a30233 ASoC: SOF: topology: use the default hw_config in DAI link init
c091fce ASoC: SOF: core: fix a typo in a comment
5c78516 ASoC: SOF: skl: remove the .cmd_done platform driver method
3dcbf59 ASoC: SOF: hsw: remove the .cmd_done platform driver method
a74c774 ASoC: SOF: spi: remove the .cmd_done platform driver method
2c787c5 ASoC: SOF: hsw: eliminate redundant mailbox reads
7f7d265 ASoC: SOF: skl: remove .get_reply()
bb7890c ASoC: SOF: hsw: remove .get_reply()
63f6059 ASoC: SOF: spi: remove .get_reply()
dc5f5f0 ASoC: SOF: hsw: remove an always true condition check
ae4e51e ASoC: SOF: hsw: print sizes in decimal format
d435088 ASoC: SOF: hsw: simplify .get_reply() implementations
9c6b980 ASoC: SOF: ipc: remove the .cmd_done platform driver method
04654c8 ASoC: SOF: Intel: eliminate redundant mailbox reads
df70b78 ASoC: SOF: ipc: remove .get_reply()
9bfc74c ASoC: SOF: Intel: remove an always true condition check
344bfbb ASoC: SOF: ipc: eliminate a trivial function
64b1050 ASoC: SOF: Intel: print sizes in decimal format
bedc1d3 ASoC: SOF: Intel: simplify multiple .get_reply() implementations
e29122d ASoC: Intel: sof_rt5682: test devm_ and return -ENOMEM

[Jiangxinx]

@ranj063 PR #775 doesn't help, got the same dmesg log.
Git diff:

commit ad88ef1313311b45dc97bd037a6bcaaccfeca353
Author: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Date:   Tue Apr 2 17:22:02 2019 -0700

    ASoC: SOF: intel: ipc: don't read mailbox for CTX_SAVE

    If the reply from the DSP is for a CTX_SAVE ipc, don't
    read the mailbox, just return after setting the reply attributes.

    Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>

commit f3adfd66c97c730a9db99dafabab7a94f6bcf552
Author: Rander Wang <rander.wang@linux.intel.com>
Date:   Fri Mar 29 15:46:04 2019 +0800

    ASoC:sof: remove duplicate posn message in kernel log

    There is a lot of posn offset message in kernel log. Actually
    the posn in mailbox is never changed after it is set in hw_params.
    So now just print it once in hw_params and make kernel message lesser

    dmesg log example:
    sof-audio-pci 0000:00:1f.3: posn mailbox: posn offset is 0xc104c
    sof-audio-pci 0000:00:1f.3: posn : host 0x6c00 dai 0xcc660 wall 0x31e4fbd
    sof-audio-pci 0000:00:1f.3: posn mailbox: posn offset is 0xc104c
    sof-audio-pci 0000:00:1f.3: posn : host 0xab00 dai 0xd4460 wall 0x33d12bc
    sof-audio-pci 0000:00:1f.3: posn mailbox: posn offset is 0xc104c
    sof-audio-pci 0000:00:1f.3: posn : host 0xea00 dai 0xdc260 wall 0x35bd5bd
    sof-audio-pci 0000:00:1f.3: posn mailbox: posn offset is 0xc104c

    Signed-off-by: Rander Wang <rander.wang@linux.intel.com>

commit 0967e0766216940cbdcd2d1d3662ea9e5beb239d
Author: Keyon Jie <yang.jie@linux.intel.com>
Date:   Tue Apr 2 14:01:52 2019 +0800

    ASoC: SOF: control: fix a PM put missing at error

    We still need call pm_runtime_put_autosuspend() to release dev at
    failure of copy_to_user(), here correct it.

    Signed-off-by: Keyon Jie <yang.jie@linux.intel.com>

commit d6d661c7521a72c2bce6f2ced11b812f449f8e43
Author: Keyon Jie <yang.jie@linux.intel.com>
Date:   Tue Apr 2 14:06:27 2019 +0800

    ASoC: SOF: Intel: fix period_bytes calculation at hw_params()

    We should use params_period_bytes for hdac_stream period_bytes
    calculation, it used params_period_size so actually wrong period bytes
    have being used for long time, here correct it.

    Signed-off-by: Keyon Jie <yang.jie@linux.intel.com>

[emilchudzik]

@Jiangxinx @emilchudzik can you please check if linux PR #775 helps?

@ranj063 is #775 should fix booting issue?
Or should fix my observations? I can boot firmware, but I have this issue: #774

[Jiangxinx]

@emilchudzik do you use the latest kernel commit to test the boot issue? e29122d is older than my test version.

[emilchudzik]

With the latest sof-dev on linux repo (f3adfd6) I can't boot to OS on WHL.
FW and tplg: master/d7b36bf3 (2 April)

On e29122d loading to OS and loading above firmware was done correctly.

I see panic while trying to load the system. I made quick screen shot: