get a kernel crash with kernel 3.10.0-1.el6.elrepo.x86_64

get a kernel crash with kernel 3.10.0-1.el6.elrepo.x86_64

[Stefan Sabolowitsch]

Hi all,
I get a kernel crash with 3.10.0-1.el6.elrepo.x86_64 (on two machines the
same problem).
Never seen on v3.8.x / v3.9x

Any idea ?

Thx
Stefan

-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-

abrt_version:   2.0.8
cmdline:        ro root=/dev/mapper/vg_ipd2-lv_root rd_NO_LUKS rd_NO_MD
SYSFONT=latarcyrheb-sun16 crashkernel=auto rd_LVM_LV=vg_ipd2/lv_swap
rd_NO_DM LANG=de_DE.UTF-8  KEYBOARDTYPE=pc KEYTABLE=de
rd_LVM_LV=vg_ipd2/lv_root rhgb quiet
kernel:         3.10.0-1.el6.elrepo.x86_64
time:           Do 04 Jul 2013 16:18:29 GMT

sosreport.tar.xz: Binary file, 1665652 bytes

backtrace:
:WARNING: at drivers/dma/ioat/dca.c:697 ioat3_dca_init+0x1f2/0x200
[ioatdma]()
:ioatdma 0000:00:16.0: APICID_TAG_MAP set incorrectly by BIOS, disabling
DCA
:Modules linked in: ioatdma(+) i7core_edac edac_core igb hwmon dca ptp
pps_core ext4 jbd2 mbcache sd_mod crc_t10dif ahci libahci dm_mirror
dm_region_hash dm_log dm_mod
:CPU: 0 PID: 1943 Comm: modprobe Not tainted 3.10.0-1.el6.elrepo.x86_64 #1
:Hardware name: NEXCOM NSA7110/NSB7110, BIOS 1.43 11/10/2010
:ffffffffa016cbfc ffff8802333efa68 ffffffff815f50e8 ffff8802333efaa8
:ffffffff8105b2f0 ffffffffa016e9c0 ffff8802345da7c0 ffff880234660de0
:ffff880234679000 0000000000000002 0000000000000100 ffff8802333efb08
:Call Trace:
:[<ffffffff815f50e8>] dump_stack+0x19/0x21
:[<ffffffff8105b2f0>] warn_slowpath_common+0x70/0xa0
:[<ffffffff8105b37f>] warn_slowpath_fmt_taint+0x3f/0x50
:[<ffffffffa01139a1>] ? alloc_dca_provider+0x21/0x40 [dca]
:[<ffffffffa016b142>] ioat3_dca_init+0x1f2/0x200 [ioatdma]
:[<ffffffffa01673e0>] ioat3_dma_probe+0x310/0x460 [ioatdma]
:[<ffffffff812ce4cb>] ? __pci_set_master+0x2b/0x90
:[<ffffffffa0163241>] ioat_pci_probe+0x1b1/0x1d0 [ioatdma]
:[<ffffffff812d0e0e>] local_pci_probe+0x4e/0x90
:[<ffffffff812d0f21>] __pci_device_probe+0xd1/0xe0
:[<ffffffff812d2262>] ? pci_dev_get+0x22/0x30
:[<ffffffff812d22aa>] pci_device_probe+0x3a/0x60
:[<ffffffff813cddc1>] really_probe+0x61/0x210
:[<ffffffff813cdfae>] driver_probe_device+0x3e/0x60
:[<ffffffff813ce07b>] __driver_attach+0xab/0xb0
:[<ffffffff813cdfd0>] ? driver_probe_device+0x60/0x60
:[<ffffffff813cdfd0>] ? driver_probe_device+0x60/0x60
:[<ffffffff813cc264>] bus_for_each_dev+0x94/0xb0
:[<ffffffff813cdc0e>] driver_attach+0x1e/0x20
:[<ffffffff813cd658>] bus_add_driver+0x1e8/0x250
:[<ffffffffa0157000>] ? 0xffffffffa0156fff
:[<ffffffff813ce714>] driver_register+0x74/0x160
:[<ffffffffa0157000>] ? 0xffffffffa0156fff
:[<ffffffff812d23bc>] __pci_register_driver+0x4c/0x50
:[<ffffffffa015706d>] ioat_init_module+0x6d/0x1000 [ioatdma]
:[<ffffffff81002042>] do_one_initcall+0x42/0x170
:[<ffffffff810c3a30>] do_init_module+0x80/0x1f0
:[<ffffffff810c5540>] load_module+0x390/0x500
:[<ffffffff810c3110>] ? mod_sysfs_teardown+0x150/0x150
:[<ffffffff810c31a7>] ? copy_module_from_user+0x67/0xc0
:[<ffffffff810c2830>] ? module_sect_show+0x30/0x30
:[<ffffffff810c57f3>] SyS_init_module+0x93/0xa0
:[<ffffffff816014d9>] system_call_fastpath+0x16/0x1b

Re: get a kernel crash with kernel 3.10.0-1.el6.elrepo.x86_64

[Borislav Petkov]

On Fri, Jul 05, 2013 at 09:30:53AM +0000, Stefan Sabolowitsch wrote:
> Hi all,
> I get a kernel crash with 3.10.0-1.el6.elrepo.x86_64 (on two machines the
> same problem).
> Never seen on v3.8.x / v3.9x
> 
> Any idea ?
> 
> Thx
> Stefan
> 
> -#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-
> 
> abrt_version:   2.0.8
> cmdline:        ro root=/dev/mapper/vg_ipd2-lv_root rd_NO_LUKS rd_NO_MD
> SYSFONT=latarcyrheb-sun16 crashkernel=auto rd_LVM_LV=vg_ipd2/lv_swap
> rd_NO_DM LANG=de_DE.UTF-8  KEYBOARDTYPE=pc KEYTABLE=de
> rd_LVM_LV=vg_ipd2/lv_root rhgb quiet
> kernel:         3.10.0-1.el6.elrepo.x86_64
> time:           Do 04 Jul 2013 16:18:29 GMT
> 
> sosreport.tar.xz: Binary file, 1665652 bytes
> 
> backtrace:
> :WARNING: at drivers/dma/ioat/dca.c:697 ioat3_dca_init+0x1f2/0x200
> [ioatdma]()
> :ioatdma 0000:00:16.0: APICID_TAG_MAP set incorrectly by BIOS, disabling
> DCA

It basically says that your BIOS is b0rked
(f3c78f8515dd4cb9e26040ebb54cf589d21137ee) and disables the Direct Cache
Access.

-- 
Regards/Gruss,
    Boris.

Sent from a fat crate under my desk. Formatting is fine.
--