6.1.30: thunderbolt: Clear registers properly when auto clear isn't in use cause call trace after resume

Upgrade to 6.1.30, got crash message after resume, but looks still running normally

After revert
e16629c639d429e48c849808e59f1efcce886849
thunderbolt: Clear registers properly when auto clear isn't in use
This error was gone.

kernel config attached, system is Slackware 15.0 on XPS 9700

May 27 13:55:39 devel kernel: ------------[ cut here ]------------
May 27 13:55:39 devel kernel: thunderbolt 0000:07:00.0: interrupt for
TX ring 0 is already enabled
May 27 13:55:39 devel kernel: WARNING: CPU: 15 PID: 21394 at
drivers/thunderbolt/nhi.c:137 ring_interrupt_active+0x1ff/0x250
[thunderbolt]
May 27 13:55:39 devel kernel: Modules linked in: squashfs

nls_iso8859_1 nls_cp437 tun fuse 8021q garp mrp iptable_nat xt_MASQUERADE nf_nat nf_conntrack nf_defrag_ipv4 ip_tables x_tables efivarfs binfmt_misc snd_ctl_led snd_soc_sof_sdw snd_soc_intel_hda_dsp_common snd_soc_intel_sof_maxim_common snd_sof_probes snd_soc_rt715 snd_soc_rt711 snd_soc_rt1308_sdw regmap_sdw snd_soc_dmic snd_sof_pci_intel_cnl snd_sof_intel_hda_common snd_sof_pci soundwire_intel soundwire_generic_allocation soundwire_cadence snd_sof_intel_hda snd_sof snd_sof_utils snd_sof_xtensa_dsp snd_soc_acpi_intel_match snd_soc_acpi snd_soc_hdac_hda soundwire_bus snd_hda_ext_core snd_hda_codec_hdmi snd_soc_core coretemp snd_compress ac97_bus nouveau intel_tcc_cooling snd_hda_intel x86_pkg_temp_thermal dell_smm_hwmon hid_multitouch iwlmvm hwmon intel_powerclamp snd_intel_dspcfg mxm_wmi i915 i2c_designware_platform snd_intel_sdw_acpi rtsx_pci_sdmmc drm_ttm_helper i2c_designware_core mac80211 drm_buddy i2c_algo_bit dell_laptop snd_hda_codec

May 27 13:55:39 devel kernel:  ucsi_ccg dell_wmi mmc_core hid_generic

drm_display_helper ledtrig_audio sparse_keymap libarc4 snd_hwdep intel_rapl_msr dell_smbios uvcvideo ttm snd_hda_core dell_wmi_sysman kvm_intel videobuf2_vmalloc firmware_attributes_class dell_wmi_descriptor wmi_bmof intel_wmi_thunderbolt dcdbas processor_thermal_device_pci_legacy drm_kms_helper videobuf2_memops iwlwifi intel_soc_dts_iosf kvm btusb r8153_ecm btrtl videobuf2_v4l2 snd_pcm syscopyarea processor_thermal_device irqbypass cdc_ether btbcm evdev usbnet psmouse intel_lpss_pci btintel processor_thermal_rfim snd_timer videobuf2_common crc32c_intel ucsi_acpi sysfillrect ghash_clmulni_intel serio_raw cfg80211 efi_pstore r8152 typec_ucsi bluetooth sysimgblt videodev processor_thermal_mbox intel_gtt intel_lpss fb_sys_fops processor_thermal_rapl i2c_i801 roles snd i2c_nvidia_gpu drm i2c_smbus ecdh_generic idma64 i2c_hid_acpi mii usbhid thunderbolt mc soundcore rtsx_pci ecc agpgart i2c_ccgx_ucsi rfkill intel_rapl_common mfd_core

May 27 13:55:39 devel kernel:  intel_pch_thermal i2c_hid typec video
button battery hid int3403_thermal int340x_thermal_zone
pinctrl_cannonlake pinctrl_intel wmi int3400_thermal intel_pmc_core
acpi_pad acpi_thermal_rel acpi_tad ac usb_storage
May 27 13:55:39 devel kernel: CPU: 15 PID: 21394 Comm: kworker/u32:15
Tainted: G        W          6.1.30-dell-2 #1
May 27 13:55:39 devel kernel: Hardware name: Dell Inc. XPS 17
9700/0P1CHN, BIOS 1.11.1 11/18/2021
May 27 13:55:39 devel kernel: Workqueue: events_unbound async_run_entry_fn
May 27 13:55:39 devel kernel: RIP:
0010:ring_interrupt_active+0x1ff/0x250 [thunderbolt]
May 27 13:55:39 devel kernel: Code: 24 04 e8 24 2b 3c e1 4c 8b 4c 24
08 44 8b 44 24 04 48 c7 c7 50 c7 29 a0 48 8b 4c 24 10 48 8b 54 24 18
48 89 c6 e8 71 34 e4 e0 <0f> 0b 45 84 ed 0f 85 09 ff ff ff 48 8b 43 08
f6 40 70 01 0f 85 38
May 27 13:55:39 devel kernel: RSP: 0018:ffffc90000517c48 EFLAGS: 00010082
May 27 13:55:39 devel kernel: RAX: 0000000000000000 RBX:
ffff888101dab800 RCX: 0000000000000000
May 27 13:55:39 devel kernel: RDX: 0000000000000004 RSI:
0000000000000086 RDI: 00000000ffffffff
May 27 13:55:39 devel kernel: RBP: 0000000000000000 R08:
80000000ffffe7b4 R09: 0000000082999bac
May 27 13:55:39 devel kernel: R10: ffffffffffffffff R11:
ffffffff82999ba1 R12: 0000000000001001
May 27 13:55:39 devel kernel: R13: 0000000000000001 R14:
0000000000038200 R15: 0000000000000001
May 27 13:55:39 devel kernel: FS:  0000000000000000(0000)
GS:ffff88887d7c0000(0000) knlGS:0000000000000000
May 27 13:55:39 devel kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
May 27 13:55:39 devel kernel: CR2: 00007f745c010b00 CR3:
000000000220a005 CR4: 00000000007706e0
May 27 13:55:39 devel kernel: PKRU: 55555554
May 27 13:55:39 devel kernel: Call Trace:
May 27 13:55:39 devel kernel:  <TASK>
May 27 13:55:39 devel kernel:  tb_ring_start+0x141/0x230 [thunderbolt]
May 27 13:55:39 devel kernel:  tb_ctl_start+0x1f/0x70 [thunderbolt]
May 27 13:55:39 devel kernel:  ? pci_pm_restore_noirq+0xc0/0xc0
May 27 13:55:39 devel kernel:  tb_domain_runtime_resume+0x15/0x30 [thunderbolt]
May 27 13:55:39 devel kernel:  __rpm_callback+0x41/0x110
May 27 13:55:39 devel kernel:  ? pci_pm_restore_noirq+0xc0/0xc0
May 27 13:55:39 devel kernel:  rpm_callback+0x59/0x70
May 27 13:55:39 devel kernel:  rpm_resume+0x4b3/0x7f0
May 27 13:55:39 devel kernel:  ? _raw_spin_unlock_irq+0x13/0x30
May 27 13:55:39 devel kernel:  ? __wait_for_common+0x171/0x1a0
May 27 13:55:39 devel kernel:  ? usleep_range_state+0x90/0x90
May 27 13:55:39 devel kernel:  ? preempt_count_add+0x68/0xa0
May 27 13:55:39 devel kernel:  __pm_runtime_resume+0x4a/0x80
May 27 13:55:39 devel kernel:  pci_pm_suspend+0x60/0x170
May 27 13:55:39 devel kernel:  ? pci_pm_freeze+0xb0/0xb0
May 27 13:55:39 devel kernel:  dpm_run_callback+0x3f/0x150
May 27 13:55:39 devel kernel:  ? _raw_spin_lock_irqsave+0x19/0x40
May 27 13:55:39 devel kernel:  __device_suspend+0x130/0x4d0
May 27 13:55:39 devel kernel:  async_suspend+0x1b/0x90
May 27 13:55:39 devel kernel:  async_run_entry_fn+0x1a/0xa0

May 27 13:55:39 devel kernel:  process_one_work+0x1bd/0x3c0
May 27 13:55:39 devel kernel:  worker_thread+0x4d/0x3c0
May 27 13:55:39 devel kernel:  ? process_one_work+0x3c0/0x3c0
May 27 13:55:39 devel kernel:  kthread+0xe5/0x110
May 27 13:55:39 devel kernel:  ? kthread_complete_and_exit+0x20/0x20
May 27 13:55:39 devel kernel:  ret_from_fork+0x1f/0x30
May 27 13:55:39 devel kernel:  </TASK>
May 27 13:55:39 devel kernel: ---[ end trace 0000000000000000 ]---

Quoted 2 lines · beld zhang

Upgrade to 6.1.30, got crash message after resume, but looks still running normally

Quoted 5 lines

After revert
e16629c639d429e48c849808e59f1efcce886849
thunderbolt: Clear registers properly when auto clear isn't in use
This error was gone.

Can you check latest mainline to see if this regression still happens?

Quoted 97 lines

kernel config attached, system is Slackware 15.0 on XPS 9700

May 27 13:55:39 devel kernel: ------------[ cut here ]------------
May 27 13:55:39 devel kernel: thunderbolt 0000:07:00.0: interrupt for
TX ring 0 is already enabled
May 27 13:55:39 devel kernel: WARNING: CPU: 15 PID: 21394 at
drivers/thunderbolt/nhi.c:137 ring_interrupt_active+0x1ff/0x250
[thunderbolt]
May 27 13:55:39 devel kernel: Modules linked in: squashfs

nls_iso8859_1 nls_cp437 tun fuse 8021q garp mrp iptable_nat xt_MASQUERADE nf_nat nf_conntrack nf_defrag_ipv4 ip_tables x_tables efivarfs binfmt_misc snd_ctl_led snd_soc_sof_sdw snd_soc_intel_hda_dsp_common snd_soc_intel_sof_maxim_common snd_sof_probes snd_soc_rt715 snd_soc_rt711 snd_soc_rt1308_sdw regmap_sdw snd_soc_dmic snd_sof_pci_intel_cnl snd_sof_intel_hda_common snd_sof_pci soundwire_intel soundwire_generic_allocation soundwire_cadence snd_sof_intel_hda snd_sof snd_sof_utils snd_sof_xtensa_dsp snd_soc_acpi_intel_match snd_soc_acpi snd_soc_hdac_hda soundwire_bus snd_hda_ext_core snd_hda_codec_hdmi snd_soc_core coretemp snd_compress ac97_bus nouveau intel_tcc_cooling snd_hda_intel x86_pkg_temp_thermal dell_smm_hwmon hid_multitouch iwlmvm hwmon intel_powerclamp snd_intel_dspcfg mxm_wmi i915 i2c_designware_platform snd_intel_sdw_acpi rtsx_pci_sdmmc drm_ttm_helper i2c_designware_core mac80211 drm_buddy i2c_algo_bit dell_laptop snd_hda_codec

May 27 13:55:39 devel kernel:  ucsi_ccg dell_wmi mmc_core hid_generic

drm_display_helper ledtrig_audio sparse_keymap libarc4 snd_hwdep intel_rapl_msr dell_smbios uvcvideo ttm snd_hda_core dell_wmi_sysman kvm_intel videobuf2_vmalloc firmware_attributes_class dell_wmi_descriptor wmi_bmof intel_wmi_thunderbolt dcdbas processor_thermal_device_pci_legacy drm_kms_helper videobuf2_memops iwlwifi intel_soc_dts_iosf kvm btusb r8153_ecm btrtl videobuf2_v4l2 snd_pcm syscopyarea processor_thermal_device irqbypass cdc_ether btbcm evdev usbnet psmouse intel_lpss_pci btintel processor_thermal_rfim snd_timer videobuf2_common crc32c_intel ucsi_acpi sysfillrect ghash_clmulni_intel serio_raw cfg80211 efi_pstore r8152 typec_ucsi bluetooth sysimgblt videodev processor_thermal_mbox intel_gtt intel_lpss fb_sys_fops processor_thermal_rapl i2c_i801 roles snd i2c_nvidia_gpu drm i2c_smbus ecdh_generic idma64 i2c_hid_acpi mii usbhid thunderbolt mc soundcore rtsx_pci ecc agpgart i2c_ccgx_ucsi rfkill intel_rapl_common mfd_core

May 27 13:55:39 devel kernel:  intel_pch_thermal i2c_hid typec video
button battery hid int3403_thermal int340x_thermal_zone
pinctrl_cannonlake pinctrl_intel wmi int3400_thermal intel_pmc_core
acpi_pad acpi_thermal_rel acpi_tad ac usb_storage
May 27 13:55:39 devel kernel: CPU: 15 PID: 21394 Comm: kworker/u32:15
Tainted: G        W          6.1.30-dell-2 #1
May 27 13:55:39 devel kernel: Hardware name: Dell Inc. XPS 17
9700/0P1CHN, BIOS 1.11.1 11/18/2021
May 27 13:55:39 devel kernel: Workqueue: events_unbound async_run_entry_fn
May 27 13:55:39 devel kernel: RIP:
0010:ring_interrupt_active+0x1ff/0x250 [thunderbolt]
May 27 13:55:39 devel kernel: Code: 24 04 e8 24 2b 3c e1 4c 8b 4c 24
08 44 8b 44 24 04 48 c7 c7 50 c7 29 a0 48 8b 4c 24 10 48 8b 54 24 18
48 89 c6 e8 71 34 e4 e0 <0f> 0b 45 84 ed 0f 85 09 ff ff ff 48 8b 43 08
f6 40 70 01 0f 85 38
May 27 13:55:39 devel kernel: RSP: 0018:ffffc90000517c48 EFLAGS: 00010082
May 27 13:55:39 devel kernel: RAX: 0000000000000000 RBX:
ffff888101dab800 RCX: 0000000000000000
May 27 13:55:39 devel kernel: RDX: 0000000000000004 RSI:
0000000000000086 RDI: 00000000ffffffff
May 27 13:55:39 devel kernel: RBP: 0000000000000000 R08:
80000000ffffe7b4 R09: 0000000082999bac
May 27 13:55:39 devel kernel: R10: ffffffffffffffff R11:
ffffffff82999ba1 R12: 0000000000001001
May 27 13:55:39 devel kernel: R13: 0000000000000001 R14:
0000000000038200 R15: 0000000000000001
May 27 13:55:39 devel kernel: FS:  0000000000000000(0000)
GS:ffff88887d7c0000(0000) knlGS:0000000000000000
May 27 13:55:39 devel kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
May 27 13:55:39 devel kernel: CR2: 00007f745c010b00 CR3:
000000000220a005 CR4: 00000000007706e0
May 27 13:55:39 devel kernel: PKRU: 55555554
May 27 13:55:39 devel kernel: Call Trace:
May 27 13:55:39 devel kernel:  <TASK>
May 27 13:55:39 devel kernel:  tb_ring_start+0x141/0x230 [thunderbolt]
May 27 13:55:39 devel kernel:  tb_ctl_start+0x1f/0x70 [thunderbolt]
May 27 13:55:39 devel kernel:  ? pci_pm_restore_noirq+0xc0/0xc0
May 27 13:55:39 devel kernel:  tb_domain_runtime_resume+0x15/0x30 [thunderbolt]
May 27 13:55:39 devel kernel:  __rpm_callback+0x41/0x110
May 27 13:55:39 devel kernel:  ? pci_pm_restore_noirq+0xc0/0xc0
May 27 13:55:39 devel kernel:  rpm_callback+0x59/0x70
May 27 13:55:39 devel kernel:  rpm_resume+0x4b3/0x7f0
May 27 13:55:39 devel kernel:  ? _raw_spin_unlock_irq+0x13/0x30
May 27 13:55:39 devel kernel:  ? __wait_for_common+0x171/0x1a0
May 27 13:55:39 devel kernel:  ? usleep_range_state+0x90/0x90
May 27 13:55:39 devel kernel:  ? preempt_count_add+0x68/0xa0
May 27 13:55:39 devel kernel:  __pm_runtime_resume+0x4a/0x80
May 27 13:55:39 devel kernel:  pci_pm_suspend+0x60/0x170
May 27 13:55:39 devel kernel:  ? pci_pm_freeze+0xb0/0xb0
May 27 13:55:39 devel kernel:  dpm_run_callback+0x3f/0x150
May 27 13:55:39 devel kernel:  ? _raw_spin_lock_irqsave+0x19/0x40
May 27 13:55:39 devel kernel:  __device_suspend+0x130/0x4d0
May 27 13:55:39 devel kernel:  async_suspend+0x1b/0x90
May 27 13:55:39 devel kernel:  async_run_entry_fn+0x1a/0xa0

Quoted 8 lines

May 27 13:55:39 devel kernel:  process_one_work+0x1bd/0x3c0
May 27 13:55:39 devel kernel:  worker_thread+0x4d/0x3c0
May 27 13:55:39 devel kernel:  ? process_one_work+0x3c0/0x3c0
May 27 13:55:39 devel kernel:  kthread+0xe5/0x110
May 27 13:55:39 devel kernel:  ? kthread_complete_and_exit+0x20/0x20
May 27 13:55:39 devel kernel:  ret_from_fork+0x1f/0x30
May 27 13:55:39 devel kernel:  </TASK>
May 27 13:55:39 devel kernel: ---[ end trace 0000000000000000 ]---

Anyway, I'm adding it to regzbot (as stable-specific regression for now):

#regzbot ^introduced: e16629c639d429
#regzbot title: Properly clearing Thunderbolt registers when not autoclearing triggers ring_interrupt_active crash on resume

Thanks.

This is specific resuming from s2idle, doesn't happen at boot?

Does it happen with hot-plugging or hot-unplugging a TBT3 or USB4 dock too?

In addition to checking mainline, can you please attach a full dmesg to somewhere ephemeral like a kernel bugzilla with thunderbolt.dyndbg='+p' on the kernel command line set?

---------- Forwarded message ---------
From: beld zhang
Date: Sun, May 28, 2023 at 2:07 PM
Subject: Re: 6.1.30: thunderbolt: Clear registers properly when auto clear isn't in use cause call trace after resume
To: Mario Limonciello

Quoted 9 lines · Mario Limonciello

This is specific resuming from s2idle, doesn't happen at boot?

Does it happen with hot-plugging or hot-unplugging a TBT3 or USB4 dock too?

In addition to checking mainline, can you please attach a full dmesg to somewhere ephemeral like a kernel bugzilla with thunderbolt.dyndbg='+p' on the kernel command line set?

6.4-rc4:
*) test 1~4 was done with usb hub with ethernet plugged-in
model: UE330, usb 3.0 3-port hub & GIgabit Ether adapter a rapoo wireless mouse in one of the ports
1) no crash at boot
until [169.099024]
2) no crash after plug an extra usb dock
from [297.004691]
3) no crash after remove it
from [373.273511]
4) crash after suspend/resume: 2 call-stacks
from [438.356253]
5) removed that hub(only ac-power left): NO crash after resume from [551.820333]
6) plug in the hub(no mouse): NO crash after resume from [1250.256607]
7) put on mouse: CRASH after resume
from [1311.400963]
mouse model: Rapoo Wireless Optical Mouse 1620

sorry I have no idea how to fill a proper bug report at kernel bugzilla, hope these shared links work.
btw I have no TB devices to test.

dmesg:
https://drive.google.com/file/d/1bUWnV7q2ziM4tdTzmuGiVuvEzaLcdfKm/view?usp=sharing

config:
https://drive.google.com/file/d/1It75_AV5tOzfkXXBAX5zAiZMoeJAe0Au/view?usp=sharing

Quoted 9 lines

sorry I have no idea how to fill a proper bug report at kernel bugzilla, hope these shared links work.
btw I have no TB devices to test.

dmesg:
https://drive.google.com/file/d/1bUWnV7q2ziM4tdTzmuGiVuvEzaLcdfKm/view?usp=sharing

config:
https://drive.google.com/file/d/1It75_AV5tOzfkXXBAX5zAiZMoeJAe0Au/view?usp=sharing

<formletter>

This is not the correct way to submit patches for inclusion in the stable kernel tree. Please read:
https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html
for how to do this properly.

</formletter>

Quoted 17 lines · Greg KH

Quoted 9 lines

sorry I have no idea how to fill a proper bug report at kernel bugzilla, hope these shared links work.
btw I have no TB devices to test.

dmesg:
https://drive.google.com/file/d/1bUWnV7q2ziM4tdTzmuGiVuvEzaLcdfKm/view?usp=sharing

config:
https://drive.google.com/file/d/1It75_AV5tOzfkXXBAX5zAiZMoeJAe0Au/view?usp=sharing

<formletter>

This is not the correct way to submit patches for inclusion in the stable kernel tree. Please read:
https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html
for how to do this properly.

Quoted 1 line

</formletter>

Hi Greg,

You confuse attaching bug artifacts (here dmesg and .config) with stable kernel patches. The reporter doesn't have any patches to help fixing this regression at the moment.

---------- Forwarded message ---------
From: beld zhang
Date: Sun, May 28, 2023 at 2:07 PM
Subject: Re: 6.1.30: thunderbolt: Clear registers properly when auto clear isn't in use cause call trace after resume
To: Mario Limonciello

Quoted 9 lines · Mario Limonciello

This is specific resuming from s2idle, doesn't happen at boot?

Does it happen with hot-plugging or hot-unplugging a TBT3 or USB4 dock too?

In addition to checking mainline, can you please attach a full dmesg to somewhere ephemeral like a kernel bugzilla with thunderbolt.dyndbg='+p' on the kernel command line set?

6.4-rc4:
*) test 1~4 was done with usb hub with ethernet plugged-in
model: UE330, usb 3.0 3-port hub & GIgabit Ether adapter a rapoo wireless mouse in one of the ports
1) no crash at boot
until [169.099024]
2) no crash after plug an extra usb dock
from [297.004691]
3) no crash after remove it
from [373.273511]
4) crash after suspend/resume: 2 call-stacks

from [438.356253]
5) removed that hub(only ac-power left): NO crash after resume from [551.820333]
6) plug in the hub(no mouse): NO crash after resume from [1250.256607]
7) put on mouse: CRASH after resume
from [1311.400963]
mouse model: Rapoo Wireless Optical Mouse 1620

sorry I have no idea how to fill a proper bug report at kernel bugzilla, hope these shared links work.
btw I have no TB devices to test.

dmesg:
https://drive.google.com/file/d/1bUWnV7q2ziM4tdTzmuGiVuvEzaLcdfKm/view?usp=sharing

config:
https://drive.google.com/file/d/1It75_AV5tOzfkXXBAX5zAiZMoeJAe0Au/view?usp=sharing

Quoted 30 lines · beld zhang

---------- Forwarded message ---------
From: beld zhang
Date: Sun, May 28, 2023 at 2:07 PM
Subject: Re: 6.1.30: thunderbolt: Clear registers properly when auto clear isn't in use cause call trace after resume
To: Mario Limonciello

Quoted 9 lines · Mario Limonciello

This is specific resuming from s2idle, doesn't happen at boot?

Does it happen with hot-plugging or hot-unplugging a TBT3 or USB4 dock too?

In addition to checking mainline, can you please attach a full dmesg to somewhere ephemeral like a kernel bugzilla with thunderbolt.dyndbg='+p' on the kernel command line set?

6.4-rc4:
*) test 1~4 was done with usb hub with ethernet plugged-in
model: UE330, usb 3.0 3-port hub & GIgabit Ether adapter a rapoo wireless mouse in one of the ports
1) no crash at boot
until [169.099024]
2) no crash after plug an extra usb dock
from [297.004691]
3) no crash after remove it
from [373.273511]
4) crash after suspend/resume: 2 call-stacks

Quoted 8 lines

from [438.356253]
5) removed that hub(only ac-power left): NO crash after resume from [551.820333]
6) plug in the hub(no mouse): NO crash after resume from [1250.256607]
7) put on mouse: CRASH after resume
from [1311.400963]
mouse model: Rapoo Wireless Optical Mouse 1620

Before suspend, is attaching your mouse not crashing your system?

Quoted 10 lines

sorry I have no idea how to fill a proper bug report at kernel bugzilla, hope these shared links work.
btw I have no TB devices to test.

dmesg:
https://drive.google.com/file/d/1bUWnV7q2ziM4tdTzmuGiVuvEzaLcdfKm/view?usp=sharing

config:
https://drive.google.com/file/d/1It75_AV5tOzfkXXBAX5zAiZMoeJAe0Au/view?usp=sharing

There is a functionality on Bugzilla to attach above files. Use it.

Thanks.

--
An old man doll... just what I always wanted! - Clara

Quoted 3 lines · Bagas Sanjaya

Quoted 2 lines · beld zhang

Upgrade to 6.1.30, got crash message after resume, but looks still running normally

This is specific resuming from s2idle, doesn't happen at boot?

Does it happen with hot-plugging or hot-unplugging a TBT3 or USB4 dock too?

Quoted 8 lines · Mario Limonciello

Quoted 3 lines · Bagas Sanjaya

Quoted 2 lines · beld zhang

Upgrade to 6.1.30, got crash message after resume, but looks still running normally

This is specific resuming from s2idle, doesn't happen at boot?

Does it happen with hot-plugging or hot-unplugging a TBT3 or USB4 dock too?

Happens also when device is connected and do

  # rmmod thunderbolt
  # modprobe thunderbolt

I think it is because nhi_mask_interrupt() does not mask interrupt on Intel now.

Can you try the patch below? I'm unable to try myself because my test system has some booting issues at the moment.

drivers/thunderbolt/nhi.c +6 -3

index 4c9f2811d20d..a11650da40f9 100644
@@ -60,9 +60,12 @@ static int ring_interrupt_index(const struct tb_ring *ring)
 
 static void nhi_mask_interrupt(struct tb_nhi *nhi, int mask, int ring)
 {
-       if (nhi->quirks & QUIRK_AUTO_CLEAR_INT)
-               return;
-       iowrite32(mask, nhi->iobase + REG_RING_INTERRUPT_MASK_CLEAR_BASE + ring);
+       if (nhi->quirks & QUIRK_AUTO_CLEAR_INT) {
+               u32 val = ioread32(nhi->iobase + REG_RING_INTERRUPT_BASE + ring);
+               iowrite32(val & ~mask, nhi->iobase + REG_RING_INTERRUPT_BASE + ring);
+       } else {
+               iowrite32(mask, nhi->iobase + REG_RING_INTERRUPT_MASK_CLEAR_BASE + ring);
+       }
 }
 
 static void nhi_clear_interrupt(struct tb_nhi *nhi, int ring)

both

# rmmod thunderbolt
# modprobe thunderbolt
makes many crash logs on my hardware.

try to patch this to 6.1.30 and 6.4-rc4 but are all failed.

how about continue this on the kernel bugzilla, and post a patch here after it is resolved as Greg said ?

Quoted 8 lines · Mario Limonciello

Quoted 3 lines · Bagas Sanjaya

Quoted 2 lines · beld zhang

Upgrade to 6.1.30, got crash message after resume, but looks still running normally

This is specific resuming from s2idle, doesn't happen at boot?

Does it happen with hot-plugging or hot-unplugging a TBT3 or USB4 dock too?

Happens also when device is connected and do

   # rmmod thunderbolt
   # modprobe thunderbolt

I think it is because nhi_mask_interrupt() does not mask interrupt on Intel now.

Can you try the patch below? I'm unable to try myself because my test system has some booting issues at the moment.

drivers/thunderbolt/nhi.c +0 -2

index 4c9f2811d20d..a11650da40f9 100644
@@ -60,9 +60,12 @@ static int ring_interrupt_index(const struct tb_ring *ring)
 
  static void nhi_mask_interrupt(struct tb_nhi *nhi, int mask, int ring)
  {
-	if (nhi->quirks & QUIRK_AUTO_CLEAR_INT)
-		return;

Quoted 31 lines · Mika Westerberg

Quoted 8 lines · Mario Limonciello

Quoted 3 lines · Bagas Sanjaya

Quoted 2 lines · beld zhang

Upgrade to 6.1.30, got crash message after resume, but looks still running normally

This is specific resuming from s2idle, doesn't happen at boot?

Does it happen with hot-plugging or hot-unplugging a TBT3 or USB4 dock too?

Happens also when device is connected and do

   # rmmod thunderbolt
   # modprobe thunderbolt

I think it is because nhi_mask_interrupt() does not mask interrupt on Intel now.

Can you try the patch below? I'm unable to try myself because my test system has some booting issues at the moment.

drivers/thunderbolt/nhi.c +0 -2

index 4c9f2811d20d..a11650da40f9 100644
@@ -60,9 +60,12 @@ static int ring_interrupt_index(const struct tb_ring *ring)
 
  static void nhi_mask_interrupt(struct tb_nhi *nhi, int mask, int ring)
  {
-     if (nhi->quirks & QUIRK_AUTO_CLEAR_INT)
-             return;

drivers/thunderbolt/nhi.c static int ring_interrupt_index(const struct tb_ring *ring) · line 71

Quoted 10 lines

drivers/thunderbolt/nhi.c +6 -1

-     iowrite32(mask, nhi->iobase + REG_RING_INTERRUPT_MASK_CLEAR_BASE + ring);
+     if (nhi->quirks & QUIRK_AUTO_CLEAR_INT) {
+             u32 val = ioread32(nhi->iobase + REG_RING_INTERRUPT_BASE + ring);
+             iowrite32(val & ~mask, nhi->iobase + REG_RING_INTERRUPT_BASE + ring);
+     } else {
+             iowrite32(mask, nhi->iobase + REG_RING_INTERRUPT_MASK_CLEAR_BASE + ring);
+     }
  }
 
  static void nhi_clear_interrupt(struct tb_nhi *nhi, int ring)

Mika, that looks good for the issue, thanks!

You can add:

Reviewed-by Mario Limonciello <[email protected]>

When you submit it.

test passed both 6.1.30 and 6.4-rc4
comments at bugzilla.

A: http://en.wikipedia.org/wiki/Top_post
Q: Were do I find info about this thing called top-posting?
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

A: No.
Q: Should I include quotations after my reply?

http://daringfireball.net/2007/07/on_top

Quoted 29 lines · Mika Westerberg

Quoted 8 lines · Mario Limonciello

Quoted 3 lines · Bagas Sanjaya

Quoted 2 lines · beld zhang

Upgrade to 6.1.30, got crash message after resume, but looks still running normally

This is specific resuming from s2idle, doesn't happen at boot?

Does it happen with hot-plugging or hot-unplugging a TBT3 or USB4 dock too?

Happens also when device is connected and do

   # rmmod thunderbolt
   # modprobe thunderbolt

I think it is because nhi_mask_interrupt() does not mask interrupt on Intel now.

Can you try the patch below? I'm unable to try myself because my test system has some booting issues at the moment.

drivers/thunderbolt/nhi.c +0 -1

index 4c9f2811d20d..a11650da40f9 100644
@@ -60,9 +60,12 @@ static int ring_interrupt_index(const struct tb_ring *ring)
  static void nhi_mask_interrupt(struct tb_nhi *nhi, int mask, int ring)
  {
-	if (nhi->quirks & QUIRK_AUTO_CLEAR_INT)

drivers/thunderbolt/nhi.c static int ring_interrupt_index(const struct tb_ring *ring) · line 69

Quoted 10 lines

drivers/thunderbolt/nhi.c +6 -2

-		return;
-	iowrite32(mask, nhi->iobase + REG_RING_INTERRUPT_MASK_CLEAR_BASE + ring);
+	if (nhi->quirks & QUIRK_AUTO_CLEAR_INT) {
+		u32 val = ioread32(nhi->iobase + REG_RING_INTERRUPT_BASE + ring);
+		iowrite32(val & ~mask, nhi->iobase + REG_RING_INTERRUPT_BASE + ring);
+	} else {
+		iowrite32(mask, nhi->iobase + REG_RING_INTERRUPT_MASK_CLEAR_BASE + ring);
+	}
  }
  static void nhi_clear_interrupt(struct tb_nhi *nhi, int ring)

Mika, that looks good for the issue, thanks!

You can add:

Reviewed-by Mario Limonciello <[email protected]>

When you submit it.

On Mon, May 29, 2023 at 11:12:45PM -0500, Mario Limonciello wrote:

Thanks, submitted formal patch now here:

https://lore.kernel.org/linux-usb/[email protected]/

beld zhang, can you try it and see if it works on your system? It should apply on top of thunderbolt.git/fixes [1]. Thanks!

[1] git://git.kernel.org/pub/scm/linux/kernel/git/westeri/thunderbolt.git

Quoted 10 lines · Mika Westerberg

On Mon, May 29, 2023 at 11:12:45PM -0500, Mario Limonciello wrote:

Thanks, submitted formal patch now here:

https://lore.kernel.org/linux-usb/[email protected]/

beld zhang, can you try it and see if it works on your system? It should apply on top of thunderbolt.git/fixes [1]. Thanks!

Quoted 1 line

[1] git://git.kernel.org/pub/scm/linux/kernel/git/westeri/thunderbolt.git

tested fixes branch, applied patch:
1) boot with mouse..............: good
2) remove mouse then put on.....: good
3) rmmod / modprobe thunderbolt.: good
4) suspend / resume.............: good

Quoted 7 lines · beld zhang

Upgrade to 6.1.30, got crash message after resume, but looks still running normally

After revert
e16629c639d429e48c849808e59f1efcce886849
thunderbolt: Clear registers properly when auto clear isn't in use
This error was gone.

Can you check latest mainline to see if this regression still happens?
[...]
#regzbot ^introduced: e16629c639d429
#regzbot title: Properly clearing Thunderbolt registers when not autoclearing triggers ring_interrupt_active crash on resume