Skip to contentSkip to conversation

component: add support for releasing match data

REGRESSION Maarten Lankhorst 3 participants 6 messages Jan 26, 2016

Subsystems (2)
Driver core Driver component framework
lore
Share & tools
lore

Maarten Lankhorst

author
#0 lore raw

Hey,

Commit ce657b1cddf1f88c56 ("component: add support for releasing match data") causes a general protection fault when unloading snd-hda-intel with the i915 module loaded on a recent skylake machine.

This breaks one of the i915 acceptance tests that performs a module unload/reload on snd-hda-intel and i915.

# modprobe -r snd-hda-intel
[  268.635792] general protection fault: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC 
[  268.635879] Modules linked in: fuse snd_hda_codec_hdmi

snd_hda_codec_realtek snd_hda_codec_generic snd_hda_intel(-) snd_hda_codec snd_hwdep snd_hda_core snd_pcm snd_seq_midi snd_seq_midi_event snd_rawmidi snd_seq input_leds intel_rapl led_class x86_pkg_temp_thermal snd_seq_device snd_timer thermal fan wmi snd soundcore pinctrl_sunrisepoint pinctrl_intel acpi_cpufreq acpi_pad i915 processor parport_pc ppdev parport efivarfs autofs4

[  268.636072] CPU: 3 PID: 2156 Comm: modprobe Tainted: G     U  W       4.5.0-rc1+ #4626
[  268.636101] Hardware name: Intel Corporation Skylake Client 
platform/Skylake DT DDR4
[  268.636142] task: ffff88044242ab80 ti: ffff880089c7c000 task.ti: ffff880089c7c000
[  268.636168] RIP: 0010:[<ffffffff8141f865>]  [<ffffffff8141f865>] devm_component_match_release+0x35/0x50
[  268.636205] RSP: 0018:ffff880089c7fd88  EFLAGS: 00010202
[  268.636224] RAX: 6b6b6b6b6b6b6b6b RBX: 0000000000000000 RCX: 0000000000000001
[  268.636248] RDX: ffff880453b71260 RSI: 6b6b6b6b6b6b6b6b RDI: ffff8804592545b8
[  268.636272] RBP: ffff880089c7fda0 R08: ffff88045c0039d0 R09: 0000000000000026
[  268.636296] R10: 00000000000000d2 R11: 0000000000002c81 R12: ffff8804592545b8
[  268.636320] R13: ffff8804582181b0 R14: 0000000000000002 R15: ffff880089c7fdd0
[  268.636345] FS:  00007fc3b6f74700(0000) GS:ffff88046c6c0000(0000) knlGS:0000000000000000
[  268.636373] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  268.636393] CR2: 0000559dae9b10e0 CR3: 000000045566e000 CR4: 00000000003406e0
[  268.636417] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[  268.636442] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[  268.636468] Stack:
[  268.636477]  ffff880458218188 ffff8804592545b8 ffff880089c7fdd0 ffff880089c7fe08
[  268.636509]  ffffffff81428f50 0000000000000286 ffff880459254818 ffffffff00000000
[  268.636542]  ffff880459254818 ffff880458218188 ffff880453b71238 ffff8804592545b8
[  268.636574] Call Trace:
[  268.636587]  [<ffffffff81428f50>] release_nodes+0x120/0x1e0
[  268.636608]  [<ffffffff81429107>] devres_release_all+0x37/0x60
[  268.636630]  [<ffffffff81424c74>] __device_release_driver+0xa4/0x150
[  268.636654]  [<ffffffff8142561c>] driver_detach+0xac/0xb0
[  268.636674]  [<ffffffff814247a0>] bus_remove_driver+0x60/0xb0
[  268.636695]  [<ffffffff81425bf7>] driver_unregister+0x27/0x50
[  268.636717]  [<ffffffff81320535>] pci_unregister_driver+0x25/0x70
[  268.636740]  [<ffffffffc022a9a1>] azx_driver_exit+0x10/0x66f [snd_hda_intel]
[  268.636767]  [<ffffffff810f445b>] SyS_delete_module+0x17b/0x240
[  268.636789]  [<ffffffff817396d7>] entry_SYSCALL_64_fastpath+0x12/0x6a
[  268.636811] Code: f5 41 54 49 89 fc 53 31 db 48 83 7e 08 00 74 29 48 
8d 14 80 49 8b 45 10 48 8d 14 d0 48 8b 42 10 48 85 c0 74 08 48 8b 32 4c 
89 e7 <ff> d0 8d 43 01 49 3b 45 08 48 89 c3 72 d7 5b 41 5c 41 5d 
5d c3 
[  268.636993] RIP  [<ffffffff8141f865>] devm_component_match_release+0x35/0x50
[  268.637021]  RSP <ffff880089c7fd88>
[  268.637043] ---[ end trace 31731dfc9d95562b ]---

Looking at the code, I see it only happens on newer i915 chips with power well support, so haswell/broadwell/skylake are probably affected in the functions snd_hdac_i915_init and snd_hdac_i915_exit.

~Maarten

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#1 lore raw
Quoted 3 lines · Maarten Lankhorst

Commit ce657b1cddf1f88c56 ("component: add support for releasing match data") causes a general protection fault when unloading snd-hda-intel with the i915 module loaded on a recent skylake machine.

I'm no good at interpreting x86 oopses. Can I have some help with this one? What causes a "general protection fault"? What does the code line disassemble to? What was the code doing at the point it caused this fault? Was it calling mc->release() or was it trying to access some data?

Quoted 58 lines

This breaks one of the i915 acceptance tests that performs a module unload/reload on snd-hda-intel and i915.

# modprobe -r snd-hda-intel
[  268.635792] general protection fault: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC 
[  268.635879] Modules linked in: fuse snd_hda_codec_hdmi

snd_hda_codec_realtek snd_hda_codec_generic snd_hda_intel(-) snd_hda_codec snd_hwdep snd_hda_core snd_pcm snd_seq_midi snd_seq_midi_event snd_rawmidi snd_seq input_leds intel_rapl led_class x86_pkg_temp_thermal snd_seq_device snd_timer thermal fan wmi snd soundcore pinctrl_sunrisepoint pinctrl_intel acpi_cpufreq acpi_pad i915 processor parport_pc ppdev parport efivarfs autofs4

[  268.636072] CPU: 3 PID: 2156 Comm: modprobe Tainted: G     U  W       4.5.0-rc1+ #4626
[  268.636101] Hardware name: Intel Corporation Skylake Client 
platform/Skylake DT DDR4
[  268.636142] task: ffff88044242ab80 ti: ffff880089c7c000 task.ti: ffff880089c7c000
[  268.636168] RIP: 0010:[<ffffffff8141f865>]  [<ffffffff8141f865>] devm_component_match_release+0x35/0x50
[  268.636205] RSP: 0018:ffff880089c7fd88  EFLAGS: 00010202
[  268.636224] RAX: 6b6b6b6b6b6b6b6b RBX: 0000000000000000 RCX: 0000000000000001
[  268.636248] RDX: ffff880453b71260 RSI: 6b6b6b6b6b6b6b6b RDI: ffff8804592545b8
[  268.636272] RBP: ffff880089c7fda0 R08: ffff88045c0039d0 R09: 0000000000000026
[  268.636296] R10: 00000000000000d2 R11: 0000000000002c81 R12: ffff8804592545b8
[  268.636320] R13: ffff8804582181b0 R14: 0000000000000002 R15: ffff880089c7fdd0
[  268.636345] FS:  00007fc3b6f74700(0000) GS:ffff88046c6c0000(0000) knlGS:0000000000000000
[  268.636373] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  268.636393] CR2: 0000559dae9b10e0 CR3: 000000045566e000 CR4: 00000000003406e0
[  268.636417] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[  268.636442] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[  268.636468] Stack:
[  268.636477]  ffff880458218188 ffff8804592545b8 ffff880089c7fdd0 ffff880089c7fe08
[  268.636509]  ffffffff81428f50 0000000000000286 ffff880459254818 ffffffff00000000
[  268.636542]  ffff880459254818 ffff880458218188 ffff880453b71238 ffff8804592545b8
[  268.636574] Call Trace:
[  268.636587]  [<ffffffff81428f50>] release_nodes+0x120/0x1e0
[  268.636608]  [<ffffffff81429107>] devres_release_all+0x37/0x60
[  268.636630]  [<ffffffff81424c74>] __device_release_driver+0xa4/0x150
[  268.636654]  [<ffffffff8142561c>] driver_detach+0xac/0xb0
[  268.636674]  [<ffffffff814247a0>] bus_remove_driver+0x60/0xb0
[  268.636695]  [<ffffffff81425bf7>] driver_unregister+0x27/0x50
[  268.636717]  [<ffffffff81320535>] pci_unregister_driver+0x25/0x70
[  268.636740]  [<ffffffffc022a9a1>] azx_driver_exit+0x10/0x66f [snd_hda_intel]
[  268.636767]  [<ffffffff810f445b>] SyS_delete_module+0x17b/0x240
[  268.636789]  [<ffffffff817396d7>] entry_SYSCALL_64_fastpath+0x12/0x6a
[  268.636811] Code: f5 41 54 49 89 fc 53 31 db 48 83 7e 08 00 74 29 48 
8d 14 80 49 8b 45 10 48 8d 14 d0 48 8b 42 10 48 85 c0 74 08 48 8b 32 4c 
89 e7 <ff> d0 8d 43 01 49 3b 45 08 48 89 c3 72 d7 5b 41 5c 41 5d 
5d c3 
[  268.636993] RIP  [<ffffffff8141f865>] devm_component_match_release+0x35/0x50
[  268.637021]  RSP <ffff880089c7fd88>
[  268.637043] ---[ end trace 31731dfc9d95562b ]---

Looking at the code, I see it only happens on newer i915 chips with power well support, so haswell/broadwell/skylake are probably affected in the functions snd_hdac_i915_init and snd_hdac_i915_exit.

~Maarten

Signature · 7 lines
RMK's Patch system: http://www.arm.linux.org.uk/developer/patches/
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
author
#2 lore raw

Op 26-01-16 om 15:15 schreef Russell King - ARM Linux:

Quoted 9 lines · Maarten Lankhorst
Quoted 3 lines

Commit ce657b1cddf1f88c56 ("component: add support for releasing match data") causes a general protection fault when unloading snd-hda-intel with the i915 module loaded on a recent skylake machine.

I'm no good at interpreting x86 oopses. Can I have some help with this one? What causes a "general protection fault"? What does the code line disassemble to? What was the code doing at the point it caused this fault? Was it calling mc->release() or was it trying to access some data?

Something similar to a segfault. It's trying to call 0x6b6b6b6b6b which is POISON_FREE.

mc appears to be freed already, so calling mc->release would jump to invalid data.

Quoted 58 lines
Quoted 58 lines

This breaks one of the i915 acceptance tests that performs a module unload/reload on snd-hda-intel and i915.

# modprobe -r snd-hda-intel
[  268.635792] general protection fault: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC 
[  268.635879] Modules linked in: fuse snd_hda_codec_hdmi

snd_hda_codec_realtek snd_hda_codec_generic snd_hda_intel(-) snd_hda_codec snd_hwdep snd_hda_core snd_pcm snd_seq_midi snd_seq_midi_event snd_rawmidi snd_seq input_leds intel_rapl led_class x86_pkg_temp_thermal snd_seq_device snd_timer thermal fan wmi snd soundcore pinctrl_sunrisepoint pinctrl_intel acpi_cpufreq acpi_pad i915 processor parport_pc ppdev parport efivarfs autofs4

[  268.636072] CPU: 3 PID: 2156 Comm: modprobe Tainted: G     U  W       4.5.0-rc1+ #4626
[  268.636101] Hardware name: Intel Corporation Skylake Client 
platform/Skylake DT DDR4
[  268.636142] task: ffff88044242ab80 ti: ffff880089c7c000 task.ti: ffff880089c7c000
[  268.636168] RIP: 0010:[<ffffffff8141f865>]  [<ffffffff8141f865>] devm_component_match_release+0x35/0x50
[  268.636205] RSP: 0018:ffff880089c7fd88  EFLAGS: 00010202
[  268.636224] RAX: 6b6b6b6b6b6b6b6b RBX: 0000000000000000 RCX: 0000000000000001
[  268.636248] RDX: ffff880453b71260 RSI: 6b6b6b6b6b6b6b6b RDI: ffff8804592545b8
[  268.636272] RBP: ffff880089c7fda0 R08: ffff88045c0039d0 R09: 0000000000000026
[  268.636296] R10: 00000000000000d2 R11: 0000000000002c81 R12: ffff8804592545b8
[  268.636320] R13: ffff8804582181b0 R14: 0000000000000002 R15: ffff880089c7fdd0
[  268.636345] FS:  00007fc3b6f74700(0000) GS:ffff88046c6c0000(0000) knlGS:0000000000000000
[  268.636373] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  268.636393] CR2: 0000559dae9b10e0 CR3: 000000045566e000 CR4: 00000000003406e0
[  268.636417] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[  268.636442] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[  268.636468] Stack:
[  268.636477]  ffff880458218188 ffff8804592545b8 ffff880089c7fdd0 ffff880089c7fe08
[  268.636509]  ffffffff81428f50 0000000000000286 ffff880459254818 ffffffff00000000
[  268.636542]  ffff880459254818 ffff880458218188 ffff880453b71238 ffff8804592545b8
[  268.636574] Call Trace:
[  268.636587]  [<ffffffff81428f50>] release_nodes+0x120/0x1e0
[  268.636608]  [<ffffffff81429107>] devres_release_all+0x37/0x60
[  268.636630]  [<ffffffff81424c74>] __device_release_driver+0xa4/0x150
[  268.636654]  [<ffffffff8142561c>] driver_detach+0xac/0xb0
[  268.636674]  [<ffffffff814247a0>] bus_remove_driver+0x60/0xb0
[  268.636695]  [<ffffffff81425bf7>] driver_unregister+0x27/0x50
[  268.636717]  [<ffffffff81320535>] pci_unregister_driver+0x25/0x70
[  268.636740]  [<ffffffffc022a9a1>] azx_driver_exit+0x10/0x66f [snd_hda_intel]
[  268.636767]  [<ffffffff810f445b>] SyS_delete_module+0x17b/0x240
[  268.636789]  [<ffffffff817396d7>] entry_SYSCALL_64_fastpath+0x12/0x6a
[  268.636811] Code: f5 41 54 49 89 fc 53 31 db 48 83 7e 08 00 74 29 48 
8d 14 80 49 8b 45 10 48 8d 14 d0 48 8b 42 10 48 85 c0 74 08 48 8b 32 4c 
89 e7 <ff> d0 8d 43 01 49 3b 45 08 48 89 c3 72 d7 5b 41 5c 41 5d 
5d c3 
[  268.636993] RIP  [<ffffffff8141f865>] devm_component_match_release+0x35/0x50
[  268.637021]  RSP <ffff880089c7fd88>
[  268.637043] ---[ end trace 31731dfc9d95562b ]---

Looking at the code, I see it only happens on newer i915 chips with power well support, so haswell/broadwell/skylake are probably affected in the functions snd_hdac_i915_init and snd_hdac_i915_exit.

~Maarten

#3 lore raw
Quoted 5 lines · Maarten Lankhorst

Something similar to a segfault. It's trying to call 0x6b6b6b6b6b which is POISON_FREE.

mc appears to be freed already, so calling mc->release would jump to invalid data.

It seems that my devm foo wasn't quite up to scratch. Quite why it doesn't show here while testing it (I have patches for the etnaviv GPU driver, which I'm regularly inserting/removing) I'm not sure.

Please test this patch - it seems "no worse" for me (in that it didn't crash before, and it still doesn't crash.)

Thanks.

 drivers/base/component.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/base/component.c +4 -2
index 89f5cf68d80a..05cd26c02449 100644
@@ -206,6 +206,8 @@ static void component_match_release(struct device *master,
 		if (mc->release)
 			mc->release(master, mc->data);
 	}
+
+	kfree(match->compare);
 }
 
 static void devm_component_match_release(struct device *dev, void *res)

@@ -221,14 +223,14 @@ static int component_match_realloc(struct device *dev,
 	if (match->alloc == num)
 		return 0;
 
-	new = devm_kmalloc_array(dev, num, sizeof(*new), GFP_KERNEL);
+	new = kmalloc_array(num, sizeof(*new), GFP_KERNEL);
 	if (!new)
 		return -ENOMEM;
 
 	if (match->compare) {
 		memcpy(new, match->compare, sizeof(*new) *
 					    min(match->num, num));
-		devm_kfree(dev, match->compare);
+		kfree(match->compare);
 	}
 	match->compare = new;
 	match->alloc = num;
Signature · 3 lines
RMK's Patch system: http://www.arm.linux.org.uk/developer/patches/
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.
author
#4 lore raw

Op 26-01-16 om 15:48 schreef Russell King - ARM Linux:

Quoted 12 lines · Maarten Lankhorst
Quoted 5 lines

Something similar to a segfault. It's trying to call 0x6b6b6b6b6b which is POISON_FREE.

mc appears to be freed already, so calling mc->release would jump to invalid data.

It seems that my devm foo wasn't quite up to scratch. Quite why it doesn't show here while testing it (I have patches for the etnaviv GPU driver, which I'm regularly inserting/removing) I'm not sure.

Please test this patch - it seems "no worse" for me (in that it didn't crash before, and it still doesn't crash.)

That one works, thanks. :)

~Maarten
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#5 lore raw
Quoted 7 lines · Maarten Lankhorst

Op 26-01-16 om 15:48 schreef Russell King - ARM Linux:

Quoted 6 lines · Maarten Lankhorst
Quoted 5 lines

Something similar to a segfault. It's trying to call 0x6b6b6b6b6b which is POISON_FREE.

mc appears to be freed already, so calling mc->release would jump to invalid data.

Quoted 7 lines
Quoted 6 lines

It seems that my devm foo wasn't quite up to scratch. Quite why it doesn't show here while testing it (I have patches for the etnaviv GPU driver, which I'm regularly inserting/removing) I'm not sure.

Please test this patch - it seems "no worse" for me (in that it didn't crash before, and it still doesn't crash.)

That one works, thanks. :)

Awesome, applied to the CI-only topic branch to unblock the robots, will fall out automatically once the proper patch lands upstream. -Daniel

Signature · 7 lines
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx