On 10/09, Tejun Heo wrote: >Hello, > >On Sat, Sep 30, 2017 at 01:21:13PM -0700, Linus Torvalds wrote: >> Tejun, any ideas? The original report is at >> >> https://lkml.org/lkml/2017/9/20/939 >> >> in case you don't see it in your inbox from lkml. > >So, in the full log, there is an earlier warning. > > [ 60.198729] ------------[ cut here ]------------ > [ 60.199708] WARNING: CPU: 0 PID: 1 at drivers/ata/libata-core.c:6533 > ata_host_detach+0x65/0xef > [ 60.201642] CPU: 0 PID: 1 Comm: swapper Tainted: G S > 4.12.0-rc5-00126-g47e0fb46 #1 > [ 60.203370] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS > 1.9.3-20161025_171302-gandalf 04/01/2014 > [ 60.205276] task: d9b10040 task.stack: d9b0a000 > [ 60.206377] EIP: ata_host_detach+0x65/0xef > [ 60.207254] EFLAGS: 00210246 CPU: 0 > [ 60.208051] EAX: fffffffd EBX: d5f80000 ECX: 00000000 EDX: d57d6a80 > [ 60.209158] ESI: 00200282 EDI: 00000000 EBP: d9b0be70 ESP: d9b0be60 > [ 60.210262] DS: 007b ES: 007b FS: 0000 GS: 00e0 SS: 0068 > [ 60.211264] CR0: 80050033 CR2: 00000000 CR3: 02a75000 CR4: 001406b0 > [ 60.212370] Call Trace: > [ 60.213060] ata_pci_remove_one+0xe/0x10 > [ 60.213914] piix_remove_one+0x25/0x28 > [ 60.214750] pci_device_remove+0x1c/0x44 > [ 60.215607] driver_probe_device+0x188/0x34d > [ 60.216501] __driver_attach+0x64/0xb4 > [ 60.217341] bus_for_each_dev+0x4f/0x63 > [ 60.218175] driver_attach+0x14/0x16 > [ 60.218988] ? driver_probe_device+0x34d/0x34d > [ 60.219895] bus_add_driver+0xba/0x1d4 > [ 60.220721] driver_register+0x6e/0xa0 > [ 60.221562] __pci_register_driver+0x28/0x2b > [ 60.222465] ? pdc_sata_pci_driver_init+0x16/0x16 > [ 60.223406] piix_init+0x14/0x24 > [ 60.224178] do_one_initcall+0x89/0x12d > [ 60.225016] ? parse_args+0x102/0x298 > [ 60.225915] sd 1:0:0:0: [sda] 16384 512-byte logical blocks: (8.39 MB/8.00 > MiB) > [ 60.227523] kernel_init_freeable+0xec/0x168 > [ 60.228421] ? rest_init+0x6a/0x6a > [ 60.229225] kernel_init+0x8/0xcb > [ 60.230003] ret_from_fork+0x19/0x30 > >It looks like somebody is trying to detach the device while async >probing is in progress, making later async operations free-after-use. >There was a driver core driver detach autotest feature which >incorrectly invoked driver detach while async probes are in progress >and this looks like that. This has already been fixed. Xialong, can >you repro this on the latest kernel?
I can reproduce this warning on commit ff5abbe799e29099695cb8b5b2f198dd8b8bdf26 for 5 times run. See dmesg attached. Thanks, Xialong > >Thanks. > >-- >tejun
dmesg-ff5abbe7.xz
Description: Binary data