Re: Trace with v3.17-rc1 while trying to enable power saving for pci device

[Julian Wollrath]

Hi,

> I just tried out the current git master (commit 5317821c08) and got
> the following trace:
> [   25.637181] ------------[ cut here ]------------
> [   25.637191] WARNING: CPU: 0 PID: 756 at fs/proc/generic.c:521
> remove_proc_entry+0x170/0x180() [   25.637194] remove_proc_entry:
> removing non-empty directory 'irq/23', leaking at least
> 'ehci_hcd:usb4' [   25.637196] Modules linked in: ecb btusb bluetooth
> crc16 arc4 iwldvm x86_pkg_temp_thermal mac80211 intel_powerclamp
> snd_hda_codec_hdmi snd_hda_codec_realtek i915 snd_hda_codec_generic
> snd_hda_intel intel_rapl coretemp kvm_intel snd_hda_controller
> snd_hda_codec kvm snd_hwdep iwlwifi psmouse snd_pcm thinkpad_acpi
> uvcvideo cfg80211 evdev serio_raw nvram snd_timer i2c_algo_bit
> videobuf2_vmalloc videobuf2_memops drm_kms_helper videobuf2_core snd
> v4l2_common drm soundcore wmi pcspkr tpm_tis tpm rfkill videodev
> battery ac video i2c_i801 i2ccore button processor sch_fq fuse
> autofs4 dm_crypt dm_mod sd_mod crc_t10dif crct10dif_generic
> crct10dif_pclmul crct10dif_common crc32_pclmul crc32c_intel
> ghash_clmulni_intel ahci libahci libata scsi_mod ehci_pci ehci_hcd
> xhci_hcd sdhci_pci e1000e sdhci ptp pps_core mmc_core usbcore thermal
> usb_common thermal_sys [   25.637257] CPU: 0 PID: 756 Comm: rc.local
> Not tainted 3.17.0-rc1+ #1 [   25.637259] Hardware name: LENOVO
> 23250BM/23250BM, BIOS G2ETA1WW (2.61 ) 04/22/2014 [   25.637261]
> 0000000000000009 ffffffff814b56ef ffff8800c9b7fc30 ffffffff8106eb42
> [   25.637265]  ffff880037648c40 ffff8800c9b7fc80 0000000000000002
> ffff8800c9b7fce6 [   25.637268]  ffff8800c9b7fce6 ffffffff8106eba7
> ffffffff81719f08 ffff880000000030 [   25.637271] Call Trace:
> [   25.637277]  [<ffffffff814b56ef>] ? dump_stack+0x41/0x51
> [   25.637283]  [<ffffffff8106eb42>] ? warn_slowpath_common+0x72/0x90
> [   25.637286]  [<ffffffff8106eba7>] ? warn_slowpath_fmt+0x47/0x50
> [   25.637291]  [<ffffffff811da561>] ? proc_entry_rundown+0x41/0x80
> [   25.637294]  [<ffffffff811de740>] ? remove_proc_entry+0x170/0x180
> [   25.637300]  [<ffffffff810b7db7>] ? unregister_irq_proc+0xb7/0xc0
> [   25.637303]  [<ffffffff810b2ab2>] ? free_desc+0x22/0x60
> [   25.637307]  [<ffffffff810b2b22>] ? irq_free_descs+0x32/0x80
> [   25.637312]  [<ffffffff810335f4>] ? mp_unmap_irq+0x74/0xb0
> [   25.637316]  [<ffffffff812d733c>] ?
> do_pci_disable_device+0x4c/0x60 [   25.637319]
> [<ffffffff812d7390>] ? pci_disable_device+0x40/0xb0 [   25.637330]
> [<ffffffffa0038fc2>] ? suspend_common+0x52/0x150 [usbcore]
> [   25.637335]  [<ffffffff812d8ff7>] ?
> pci_pm_runtime_suspend+0x57/0x150 [   25.637340]
> [<ffffffff8137e636>] ? __rpm_callback+0x26/0x60 [   25.637343]
> [<ffffffff8137e688>] ? rpm_callback+0x18/0x80 [   25.637346]
> [<ffffffff8137f137>] ? rpm_suspend+0x107/0x600 [   25.637350]
> [<ffffffff810a1c6f>] ? __wake_up+0x2f/0x50 [   25.637353]
> [<ffffffff8137fcad>] ? pm_runtime_allow+0x4d/0x50 [   25.637357]
> [<ffffffff8137c75d>] ? control_store+0xbd/0xc0 [   25.637361]
> [<ffffffff811e91bb>] ? kernfs_fop_write+0xdb/0x160 [   25.637365]
> [<ffffffff811835ed>] ? vfs_write+0xad/0x1e0 [   25.637368]
> [<ffffffff811840fd>] ? SyS_write+0x3d/0xa0 [   25.637373]
> [<ffffffff814bb052>] ? system_call_fastpath+0x16/0x1b [   25.637375]
> ---[ end trace 40fa9ae9acda7087 ]---
> 
> I get the trace due to enabling power saving via
> 'echo 'auto' > /sys/bus/pci/devices/0000:00:1d.0/power/control'
> with 00:1d.0 being the USB controller (according to lspci):
> 00:1d.0 USB controller [0c03]: Intel Corporation 7 Series/C210 Series
> Chipset Family USB Enhanced Host Controller #1 [8086:1e26] (rev 04)
> 
> If you need more information, please do not hesitate to ask.

Inspired by commit c5946f9d286, which mentions a similar problem, I
tried the following:
diff --git a/drivers/usb/core/hcd-pci.c b/drivers/usb/core/hcd-pci.c
index efc953119ce2..c5160bcaa790 100644
--- a/drivers/usb/core/hcd-pci.c
+++ b/drivers/usb/core/hcd-pci.c
@@ -361,6 +361,7 @@ void usb_hcd_pci_remove(struct pci_dev *dev)
        }
 
        usb_put_hcd(hcd);
+       free_irq(hcd->irq, hcd);
        pci_disable_device(dev);
 }
 EXPORT_SYMBOL_GPL(usb_hcd_pci_remove);

But this did not work, so maybe somebody, who actually knows this code
should have a look at this.


With best regards,
Julian Wollrath
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/