found 2.6.32-40 thanks Hi,
Hope it's okay for me to reassign #409349 and #611646 as kernel bugs. It may be only apcupsd users who experience this, but the process hangs due to a system call failing to return, and there were reports of a related error message in the kernel log (control queue full), which seems to have been a precursor to this hang. It's clear that this bug has existed for some time (the initial report was on 2.6.17). I'll have difficulty testing this on 3.2.x kernels because I require OpenVZ on the affected machine (and that's not ported to 3.x kernels yet), and I was unsuccessful so far in reproducing it on a spare box. I notice that most reports were from users of OHCI hardware. I've even seen this on a recent RHEL6-derived kernel just now (amd64 OpenVZ kernel 2.6.32-042stab049.6) : > [425881.271304] INFO: task apcupsd:7104 blocked for more than 120 seconds. > [425881.278546] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables > this message. > [425881.287407] apcupsd D ffff88004dc00680 0 7104 1 0 > 0x00000004 > [425881.295778] ffff880008f57d18 0000000000000082 0000000000000000 > 0000000000000286 > [425881.304301] 00000000fffffffe 0000000000000000 ffff88008f33b0a0 > ffff8800762fe200 > [425881.312839] ffff880008f57cb8 ffff88004dc00c38 ffff880008f57fd8 > ffff880008f57fd8 > [425881.321374] Call Trace: > [425881.324325] [<ffffffff81396ef5>] usb_kill_urb+0x85/0xc0 > [425881.330289] [<ffffffff81095310>] ? autoremove_wake_function+0x0/0x40 > [425881.337470] [<ffffffff81406d79>] usbhid_init_reports+0xb9/0x120 > [425881.344191] [<ffffffff81408a68>] hiddev_ioctl+0x418/0x710 > [425881.350357] [<ffffffff81154ba4>] ? handle_mm_fault+0x1e4/0x2b0 > [425881.356967] [<ffffffff81042b54>] ? __do_page_fault+0x1e4/0x480 > [425881.363595] [<ffffffff811a0182>] vfs_ioctl+0x22/0xa0 > [425881.369277] [<ffffffff811a0324>] do_vfs_ioctl+0x84/0x5b0 > [425881.375361] [<ffffffff8118e9f9>] ? __fput+0x1e9/0x280 > [425881.381140] [<ffffffff811a089f>] sys_ioctl+0x4f/0x80 > [425881.386822] [<ffffffff8100b182>] system_call_fastpath+0x16/0x1b Regards, -- Steven Chamberlain [email protected] -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected] Archive: http://lists.debian.org/[email protected]
