From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753189Ab1HOUvN (ORCPT ); Mon, 15 Aug 2011 16:51:13 -0400 Received: from ogre.sisk.pl ([217.79.144.158]:43167 "EHLO ogre.sisk.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753109Ab1HOUvL convert rfc822-to-8bit (ORCPT ); Mon, 15 Aug 2011 16:51:11 -0400 From: "Rafael J. Wysocki" To: linux-usb@vger.kernel.org Subject: [Regression] GPF in usb_driver_release_interface during/after resume Date: Mon, 15 Aug 2011 22:52:49 +0200 User-Agent: KMail/1.13.6 (Linux/3.1.0-rc2+; KDE/4.6.0; x86_64; ; ) Cc: LKML , Greg KH , Linux PM mailing list , Alan Stern , Marcel Holtmann MIME-Version: 1.0 Content-Type: Text/Plain; charset="utf-8" Content-Transfer-Encoding: 8BIT Message-Id: <201108152252.49863.rjw@sisk.pl> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi, I get the following general protection fault sometimes (but sufficiently often for it to be annoying) during resume from system suspend on Toshiba Portege R500: [ 330.839651] PM: Finishing wakeup. [ 330.839656] Restarting tasks ... [ 330.839742] e1000e 0000:01:00.0: PME# enabled [ 330.840121] usb 2-2: USB disconnect, device number 3 [ 330.841626] option1 ttyUSB0: GSM modem (1-port) converter now disconnected from ttyUSB0 [ 330.841835] option 2-2:1.0: device disconnected [ 330.842626] option1 ttyUSB1: GSM modem (1-port) converter now disconnected from ttyUSB1 [ 330.842770] option 2-2:1.1: device disconnected [ 330.843506] option1 ttyUSB2: GSM modem (1-port) converter now disconnected from ttyUSB2 [ 330.843648] option 2-2:1.2: device disconnected [ 330.857550] done. [ 330.857591] video LNXVIDEO:00: Restoring backlight state [ 330.915075] option1 ttyUSB3: GSM modem (1-port) converter now disconnected from ttyUSB3 [ 330.924372] option 2-2:1.3: device disconnected [ 330.964520] usb 5-2: USB disconnect, device number 4 [ 331.218859] general protection fault: 0000 [#1] PREEMPT SMP [ 331.218889] CPU 0 [ 331.218894] Modules linked in: bnep cryptd aes_x86_64 aes_generic xt_tcpudp xt_pkttype af_packet snd_pcm_oss snd_mixer_oss snd_seq snd_seq_device ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 ip6table_raw xt_NOTRACK ipt_REJECT iptable_raw iptable_filter ip6table_mangle nf_conntrack_netbios_ns nf_conntrack_broadcast nf_conntrack_ipv4 nf_defrag_ipv4 ip_tables xt_conntrack nf_conntrack ip6table_filter ip6_tables x_tables ipv6 cpufreq_conservative cpufreq_ondemand cpufreq_userspace cpufreq_powersave acpi_cpufreq microcode freq_table mperf fuse dm_mod sr_mod cdrom arc4 joydev btusb bluetooth option usb_wwan iwl4965 pcmcia iwl_legacy mac80211 psmouse snd_hda_codec_realtek usbhid serio_raw sdhci_pci usb_storage hid usbserial sdhci firewire_ohci snd_hda_intel yenta_socket snd_hda_codec pcmcia_rsrc sg firewire_core mmc_core snd_hwdep pcmcia_core crc_itu_t cfg80211 snd_pcm iTCO_wdt iTCO_vendor_support snd_timer snd soundcore e1000e rfkill snd_page_alloc toshiba_bluetooth battery ac uhci_hcd sd_mod crc_t10dif ehci_hcd usbcore rtc_cmos ext3 jbd ahci libahci libata fan processor thermal [ 331.219207] [ 331.219216] Pid: 906, comm: khubd Not tainted 2.6.39+ #388 TOSHIBA PORTEGE R500/Portable PC [ 331.219237] RIP: 0010:[] [] usb_driver_release_interface+0x32/0x9f [usbcore] [ 331.219289] RSP: 0018:ffff880078899be0 EFLAGS: 00010282 [ 331.219301] RAX: 0000000000000000 RBX: 6b6b6b6b6b6b6b6b RCX: 0000000000000006 [ 331.219314] RDX: 0000000000000006 RSI: 00000000000001e3 RDI: ffffffffa0011020 [ 331.219327] RBP: ffff880078899c00 R08: ffff880078899be0 R09: 09f911029d74e35b [ 331.219341] R10: ffff880078899bb0 R11: 0000000000000000 R12: ffff88007a5e6548 [ 331.219354] R13: ffff880040a8d000 R14: ffffffffa0011020 R15: ffffffffa00110b8 [ 331.219368] FS: 0000000000000000(0000) GS:ffff88007f400000(0000) knlGS:0000000000000000 [ 331.219382] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 331.219394] CR2: 00007f4b2fd01080 CR3: 0000000079cf9000 CR4: 00000000000006f0 [ 331.219407] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 331.219420] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 331.219434] Process khubd (pid: 906, threadinfo ffff880078898000, task ffff88003798ee80) [ 331.219448] Stack: [ 331.219455] ffff880078899c00 ffff880037070388 ffff88007a5e6548 ffff880040a8d000 [ 331.219477] ffff880078899c30 ffffffffa000fdb4 ffff88007a5e6578 ffff88007a5e6578 [ 331.219497] ffff88007a5e6548 ffff88007856c188 ffff880078899c70 ffffffffa00c5af0 [ 331.219518] Call Trace: [ 331.219535] [] btusb_disconnect+0x9c/0xc2 [btusb] [ 331.219566] [] usb_unbind_interface+0x5a/0x109 [usbcore] [ 331.219590] [] __device_release_driver+0x81/0xca [ 331.219606] [] device_release_driver+0x19/0x26 [ 331.219622] [] bus_remove_device+0xc2/0xd3 [ 331.219637] [] device_del+0x12b/0x17a [ 331.219666] [] usb_disable_device+0x8a/0x176 [usbcore] [ 331.219695] [] usb_disconnect+0xcd/0x137 [usbcore] [ 331.219724] [] hub_thread+0x74b/0x113e [usbcore] [ 331.219745] [] ? get_parent_ip+0x11/0x41 [ 331.219760] [] ? need_resched+0x1e/0x28 [ 331.219777] [] ? wake_up_bit+0x25/0x25 [ 331.219804] [] ? usb_new_device+0x1ca/0x1ca [usbcore] [ 331.219821] [] kthread+0x7d/0x85 [ 331.219838] [] kernel_thread_helper+0x4/0x10 [ 331.219854] [] ? retint_restore_args+0xe/0xe [ 331.219870] [] ? __init_kthread_worker+0x56/0x56 [ 331.219885] [] ? gs_change+0xb/0xb [ 331.219896] Code: 54 53 48 89 f3 48 83 ec 08 be e3 01 00 00 48 85 db 74 0a 48 85 ff 75 13 be e6 01 00 00 48 c7 c7 9c 25 0d a0 e8 b6 6d f7 e0 eb 65 [ 331.219983] 8b 83 20 01 00 00 48 85 c0 74 59 48 81 c7 98 00 00 00 48 39 [ 331.220031] RIP [] usb_driver_release_interface+0x32/0x9f [usbcore] [ 331.220065] RSP [ 331.240771] ---[ end trace be6a0c945bcc56b9 ]--- This started to happen during the 3.0 development cycle, as far as I can say. According to GDB, usb_driver_release_interface+0x32 is line 484 in drivers/usb/core/driver.c: void usb_driver_release_interface(struct usb_driver *driver, struct usb_interface *iface) { struct device *dev = &iface->dev; /* this should never happen, don't release something that's not ours */ ---> if (!dev->driver || dev->driver != &driver->drvwrap.driver) so this looks like a use-after-free to me and since it doesn't occur every time, there seems to be a race condition in the USB stack. I tried to figure it out myself, but apparently it's too tricky to me. :-) Thanks, Rafael