After fresh Fedora install (from Fedora repo only) I found in "/var/log/messages" records as:
Oct 1 07:08:26 ns kernel: ------------[ cut here ]------------ Oct 1 07:08:26 ns kernel: WARNING: at lib/dma-debug.c:791 check_unmap+0x7a/0x59b() Oct 1 07:08:26 ns kernel: Hardware name: EP45-DQ6 Oct 1 07:08:26 ns kernel: dl2k 0000:09:00.0: DMA-API: device driver tries to free an invalid DMA memory address Oct 1 07:08:26 ns kernel: Modules linked in: bluetooth rfkill nf_nat_ftp nf_conntrack_ftp iptable_nat nf_nat xt_u32 ipt_LOG xt_limit uinput iTCO_wdt i2c_i801 iTCO_vendor_support r8169 mii i2c_core dl2k microcode raid1 [last unloaded: hwmon_vid] Oct 1 07:08:26 ns kernel: Pid: 2454, comm: ip Tainted: G W 2.6.35.4-28.fc14.i686.PAE #1 Oct 1 07:08:26 ns kernel: Call Trace: Oct 1 07:08:26 ns kernel: [<c0441f39>] warn_slowpath_common+0x6a/0x7f Oct 1 07:08:26 ns kernel: [<c05dfbc6>] ? check_unmap+0x7a/0x59b Oct 1 07:08:26 ns kernel: [<c0441fc1>] warn_slowpath_fmt+0x2b/0x2f Oct 1 07:08:26 ns kernel: [<c05dfbc6>] check_unmap+0x7a/0x59b Oct 1 07:08:26 ns kernel: [<c0467ae1>] ? print_lock_contention_bug+0x11/0xa8 Oct 1 07:08:26 ns kernel: [<c0467dad>] ? lock_acquired+0x235/0x23d Oct 1 07:08:26 ns kernel: [<c05daf43>] ? debug_object_deactivate+0x28/0x92 Oct 1 07:08:26 ns kernel: [<c05e0195>] debug_dma_unmap_page+0x5a/0x62 Oct 1 07:08:26 ns kernel: [<f7e6b2d5>] pci_unmap_single+0x58/0x63 [dl2k] Oct 1 07:08:26 ns kernel: [<f7e6cd5c>] rio_close+0x9e/0x114 [dl2k] Oct 1 07:08:26 ns kernel: [<c075058a>] __dev_close+0x70/0x85 Oct 1 07:08:26 ns kernel: [<c07519d9>] __dev_change_flags+0x98/0x10d Oct 1 07:08:26 ns kernel: [<c0751abe>] dev_change_flags+0x18/0x44 Oct 1 07:08:26 ns kernel: [<c075a44f>] do_setlink+0x253/0x52e Oct 1 07:08:26 ns kernel: [<c0467ae1>] ? print_lock_contention_bug+0x11/0xa8 Oct 1 07:08:26 ns kernel: [<c0467dad>] ? lock_acquired+0x235/0x23d Oct 1 07:08:26 ns kernel: [<c04d0bf9>] ? handle_mm_fault+0x342/0x8f4 Oct 1 07:08:26 ns kernel: [<c075af7a>] rtnl_newlink+0x218/0x3a6 Oct 1 07:08:26 ns kernel: [<c075ae18>] ? rtnl_newlink+0xb6/0x3a6 Oct 1 07:08:26 ns kernel: [<c0467dad>] ? lock_acquired+0x235/0x23d Oct 1 07:08:26 ns kernel: [<c075a0f0>] ? rtnl_lock+0x14/0x16 Oct 1 07:08:26 ns kernel: [<c075a0f0>] ? rtnl_lock+0x14/0x16 Oct 1 07:08:26 ns kernel: [<c07e3e85>] ? __mutex_lock_common+0x2d0/0x2da Oct 1 07:08:26 ns kernel: [<c0467dad>] ? lock_acquired+0x235/0x23d Oct 1 07:08:26 ns kernel: [<c075ad62>] ? rtnl_newlink+0x0/0x3a6 Oct 1 07:08:26 ns kernel: [<c075abb7>] rtnetlink_rcv_msg+0x19c/0x1ab Oct 1 07:08:26 ns kernel: [<c075aa1b>] ? rtnetlink_rcv_msg+0x0/0x1ab Oct 1 07:08:26 ns kernel: [<c076a19a>] netlink_rcv_skb+0x37/0x78 Oct 1 07:08:26 ns kernel: [<c075a112>] rtnetlink_rcv+0x20/0x27 Oct 1 07:08:26 ns kernel: [<c0769d7b>] netlink_unicast+0xc6/0x121 Oct 1 07:08:26 ns kernel: [<c076a004>] netlink_sendmsg+0x22e/0x23b Oct 1 07:08:26 ns kernel: [<c0740d65>] __sock_sendmsg+0x56/0x5f Oct 1 07:08:26 ns kernel: [<c0741056>] sock_sendmsg+0x98/0xac Oct 1 07:08:26 ns kernel: [<c04ce553>] ? might_fault+0x4c/0x86 Oct 1 07:08:26 ns kernel: [<c0469b00>] ? lock_release+0x17f/0x186 Oct 1 07:08:26 ns kernel: [<c07497a1>] ? copy_from_user+0xd/0xf Oct 1 07:08:26 ns kernel: [<c0749ad7>] ? verify_iovec+0x43/0x70 Oct 1 07:08:26 ns kernel: [<c0742b45>] sys_sendmsg+0x182/0x1e6 Oct 1 07:08:26 ns kernel: [<c0467ae1>] ? print_lock_contention_bug+0x11/0xa8 Oct 1 07:08:26 ns kernel: [<c0467dad>] ? lock_acquired+0x235/0x23d Oct 1 07:08:26 ns kernel: [<c04cec8e>] ? __do_fault+0x206/0x396 Oct 1 07:08:26 ns kernel: [<c04cedf2>] ? __do_fault+0x36a/0x396 Oct 1 07:08:26 ns kernel: [<c0469bb8>] ? lock_acquire+0xb1/0xbc Oct 1 07:08:26 ns kernel: [<c04ce553>] ? might_fault+0x4c/0x86 Oct 1 07:08:26 ns kernel: [<c0469b00>] ? lock_release+0x17f/0x186 Oct 1 07:08:26 ns kernel: [<c0742f43>] sys_socketcall+0x168/0x1a8 Oct 1 07:08:26 ns kernel: [<c05d5d74>] ? trace_hardirqs_on_thunk+0xc/0x10 Oct 1 07:08:26 ns kernel: [<c04089df>] sysenter_do_call+0x12/0x38 Oct 1 07:08:26 ns kernel: ---[ end trace a70d91368aa35943 ]---
Oops itself is probably suited for bug report, but what isn't clear for me is line Oct 1 07:08:26 ns kernel: Pid: 2454, comm: ip Tainted: G W 2.6.35.4-28.fc14.i686.PAE #1
Which module caused that kernel is tainted? I suppose Fedora contains only open SW including kernel modules, isn't then there something wrong?
Thanks, Franta Hanzlík
Frantisek Hanzlik wrote:
After fresh Fedora install (from Fedora repo only) I found in "/var/log/messages" records as:
Oct 1 07:08:26 ns kernel: ------------[ cut here ]------------ Oct 1 07:08:26 ns kernel: WARNING: at lib/dma-debug.c:791 check_unmap+0x7a/0x59b() Oct 1 07:08:26 ns kernel: Hardware name: EP45-DQ6 Oct 1 07:08:26 ns kernel: dl2k 0000:09:00.0: DMA-API: device driver tries to free an invalid DMA memory address Oct 1 07:08:26 ns kernel: Modules linked in: bluetooth rfkill nf_nat_ftp nf_conntrack_ftp iptable_nat nf_nat xt_u32 ipt_LOG xt_limit uinput iTCO_wdt i2c_i801 iTCO_vendor_support r8169 mii i2c_core dl2k microcode raid1 [last unloaded: hwmon_vid] Oct 1 07:08:26 ns kernel: Pid: 2454, comm: ip Tainted: G W 2.6.35.4-28.fc14.i686.PAE #1 Oct 1 07:08:26 ns kernel: Call Trace: Oct 1 07:08:26 ns kernel: [<c0441f39>] warn_slowpath_common+0x6a/0x7f Oct 1 07:08:26 ns kernel: [<c05dfbc6>] ? check_unmap+0x7a/0x59b Oct 1 07:08:26 ns kernel: [<c0441fc1>] warn_slowpath_fmt+0x2b/0x2f Oct 1 07:08:26 ns kernel: [<c05dfbc6>] check_unmap+0x7a/0x59b Oct 1 07:08:26 ns kernel: [<c0467ae1>] ? print_lock_contention_bug+0x11/0xa8 Oct 1 07:08:26 ns kernel: [<c0467dad>] ? lock_acquired+0x235/0x23d Oct 1 07:08:26 ns kernel: [<c05daf43>] ? debug_object_deactivate+0x28/0x92 Oct 1 07:08:26 ns kernel: [<c05e0195>] debug_dma_unmap_page+0x5a/0x62 Oct 1 07:08:26 ns kernel: [<f7e6b2d5>] pci_unmap_single+0x58/0x63 [dl2k] Oct 1 07:08:26 ns kernel: [<f7e6cd5c>] rio_close+0x9e/0x114 [dl2k] Oct 1 07:08:26 ns kernel: [<c075058a>] __dev_close+0x70/0x85 Oct 1 07:08:26 ns kernel: [<c07519d9>] __dev_change_flags+0x98/0x10d Oct 1 07:08:26 ns kernel: [<c0751abe>] dev_change_flags+0x18/0x44 Oct 1 07:08:26 ns kernel: [<c075a44f>] do_setlink+0x253/0x52e Oct 1 07:08:26 ns kernel: [<c0467ae1>] ? print_lock_contention_bug+0x11/0xa8 Oct 1 07:08:26 ns kernel: [<c0467dad>] ? lock_acquired+0x235/0x23d Oct 1 07:08:26 ns kernel: [<c04d0bf9>] ? handle_mm_fault+0x342/0x8f4 Oct 1 07:08:26 ns kernel: [<c075af7a>] rtnl_newlink+0x218/0x3a6 Oct 1 07:08:26 ns kernel: [<c075ae18>] ? rtnl_newlink+0xb6/0x3a6 Oct 1 07:08:26 ns kernel: [<c0467dad>] ? lock_acquired+0x235/0x23d Oct 1 07:08:26 ns kernel: [<c075a0f0>] ? rtnl_lock+0x14/0x16 Oct 1 07:08:26 ns kernel: [<c075a0f0>] ? rtnl_lock+0x14/0x16 Oct 1 07:08:26 ns kernel: [<c07e3e85>] ? __mutex_lock_common+0x2d0/0x2da Oct 1 07:08:26 ns kernel: [<c0467dad>] ? lock_acquired+0x235/0x23d Oct 1 07:08:26 ns kernel: [<c075ad62>] ? rtnl_newlink+0x0/0x3a6 Oct 1 07:08:26 ns kernel: [<c075abb7>] rtnetlink_rcv_msg+0x19c/0x1ab Oct 1 07:08:26 ns kernel: [<c075aa1b>] ? rtnetlink_rcv_msg+0x0/0x1ab Oct 1 07:08:26 ns kernel: [<c076a19a>] netlink_rcv_skb+0x37/0x78 Oct 1 07:08:26 ns kernel: [<c075a112>] rtnetlink_rcv+0x20/0x27 Oct 1 07:08:26 ns kernel: [<c0769d7b>] netlink_unicast+0xc6/0x121 Oct 1 07:08:26 ns kernel: [<c076a004>] netlink_sendmsg+0x22e/0x23b Oct 1 07:08:26 ns kernel: [<c0740d65>] __sock_sendmsg+0x56/0x5f Oct 1 07:08:26 ns kernel: [<c0741056>] sock_sendmsg+0x98/0xac Oct 1 07:08:26 ns kernel: [<c04ce553>] ? might_fault+0x4c/0x86 Oct 1 07:08:26 ns kernel: [<c0469b00>] ? lock_release+0x17f/0x186 Oct 1 07:08:26 ns kernel: [<c07497a1>] ? copy_from_user+0xd/0xf Oct 1 07:08:26 ns kernel: [<c0749ad7>] ? verify_iovec+0x43/0x70 Oct 1 07:08:26 ns kernel: [<c0742b45>] sys_sendmsg+0x182/0x1e6 Oct 1 07:08:26 ns kernel: [<c0467ae1>] ? print_lock_contention_bug+0x11/0xa8 Oct 1 07:08:26 ns kernel: [<c0467dad>] ? lock_acquired+0x235/0x23d Oct 1 07:08:26 ns kernel: [<c04cec8e>] ? __do_fault+0x206/0x396 Oct 1 07:08:26 ns kernel: [<c04cedf2>] ? __do_fault+0x36a/0x396 Oct 1 07:08:26 ns kernel: [<c0469bb8>] ? lock_acquire+0xb1/0xbc Oct 1 07:08:26 ns kernel: [<c04ce553>] ? might_fault+0x4c/0x86 Oct 1 07:08:26 ns kernel: [<c0469b00>] ? lock_release+0x17f/0x186 Oct 1 07:08:26 ns kernel: [<c0742f43>] sys_socketcall+0x168/0x1a8 Oct 1 07:08:26 ns kernel: [<c05d5d74>] ? trace_hardirqs_on_thunk+0xc/0x10 Oct 1 07:08:26 ns kernel: [<c04089df>] sysenter_do_call+0x12/0x38 Oct 1 07:08:26 ns kernel: ---[ end trace a70d91368aa35943 ]---
Oops itself is probably suited for bug report, but what isn't clear for me is line Oct 1 07:08:26 ns kernel: Pid: 2454, comm: ip Tainted: G W 2.6.35.4-28.fc14.i686.PAE #1
Which module caused that kernel is tainted? I suppose Fedora contains only open SW including kernel modules, isn't then there something wrong?
Thanks, Franta Hanzlík
Oops I reported to Bugzilla, and regarding tainted kernel - I found in /var/log/messages interesting thing, from 11 oopses recorded there for last three days was two (4. and 7.) marked as "Tainted" and remaining as "Not tainted"! Without (un-)installing or reconfiguring anything. It seems as system is occasionally cracked :(
Frantisek Hanzlik <franta <at> hanzlici.cz> writes:
...
Oct 1 07:08:26 ns kernel: Pid: 2454, comm: ip Tainted: G W 2.6.35.4-28.fc14.i686.PAE #1
Hi, This is a kernel from Fedora 14. Devs and testers on test mailing list are on the look out for errors like that and will help you quickly. JB
JB wrote:
Frantisek Hanzlik<franta<at> hanzlici.cz> writes:
...
Oct 1 07:08:26 ns kernel: Pid: 2454, comm: ip Tainted: G W 2.6.35.4-28.fc14.i686.PAE #1
Hi, This is a kernel from Fedora 14. Devs and testers on test mailing list are on the look out for errors like that and will help you quickly. JB
This was rather general, not F14, question, how from messages line:
kernel: Pid: 2454, comm: ip Tainted: G W 2.6.35.4-28.fc14.i686.PAE
possible determine, why is kernel tainted and which modul caused it.
Differences betveen 2.6.34 (F13 stable) and 2.6.35 (F14 soon) probably isn't too big (controll groups are only one notable thing). And, of course, I'm too lazy to register at development/testers ML for this one question only. Rather will wait for F14 release... Or LKML...
Frantisek Hanzlik <franta <at> hanzlici.cz> writes:
After fresh Fedora install (from Fedora repo only) I found in "/var/log/messages" records as:
Oct 1 07:08:26 ns kernel: ------------[ cut here ]------------ Oct 1 07:08:26 ns kernel: WARNING: at lib/dma-debug.c:791 check_unmap+0x7a/0x59b() Oct 1 07:08:26 ns kernel: Hardware name: EP45-DQ6 Oct 1 07:08:26 ns kernel: dl2k 0000:09:00.0: DMA-API: device driver tries to free an invalid DMA memory address Oct 1 07:08:26 ns kernel: Modules linked in: bluetooth rfkill nf_nat_ftp nf_conntrack_ftp iptable_nat nf_nat xt_u32 ipt_LOG xt_limit uinput iTCO_wdt i2c_i801 iTCO_vendor_support r8169 mii i2c_core dl2k microcode raid1 [last unloaded: hwmon_vid] Oct 1 07:08:26 ns kernel: Pid: 2454, comm: ip Tainted: G W 2.6.35.4-28.fc14.i686.PAE #1 Oct 1 07:08:26 ns kernel: Call Trace: Oct 1 07:08:26 ns kernel: [<c0441f39>] warn_slowpath_common+0x6a/0x7f ... Oct 1 07:08:26 ns kernel: ---[ end trace a70d91368aa35943 ]---
Oops itself is probably suited for bug report, but what isn't clear for me is line Oct 1 07:08:26 ns kernel: Pid: 2454, comm: ip Tainted: G W 2.6.35.4-28.fc14.i686.PAE #1
Which module caused that kernel is tainted? I suppose Fedora contains only open SW including kernel modules, isn't then there something wrong?
Thanks, Franta Hanzlík
Hi, an answer from test mailing list.
From kernel doc
" 1: 'G' if all modules loaded have a GPL or compatible license, 'P' if any proprietary module has been loaded. Modules without a MODULE_LICENSE or with a MODULE_LICENSE that is not recognised by insmod as GPL compatible are assumed to be proprietary. "
" 10: 'W' if a warning has previously been issued by the kernel. (Though some warnings may set more specific taint flags.) "
JB
JB wrote:
Frantisek Hanzlik<franta<at> hanzlici.cz> writes:
After fresh Fedora install (from Fedora repo only) I found in "/var/log/messages" records as:
Oct 1 07:08:26 ns kernel: ------------[ cut here ]------------ Oct 1 07:08:26 ns kernel: WARNING: at lib/dma-debug.c:791 check_unmap+0x7a/0x59b() Oct 1 07:08:26 ns kernel: Hardware name: EP45-DQ6 Oct 1 07:08:26 ns kernel: dl2k 0000:09:00.0: DMA-API: device driver tries to free an invalid DMA memory address Oct 1 07:08:26 ns kernel: Modules linked in: bluetooth rfkill nf_nat_ftp nf_conntrack_ftp iptable_nat nf_nat xt_u32 ipt_LOG xt_limit uinput iTCO_wdt i2c_i801 iTCO_vendor_support r8169 mii i2c_core dl2k microcode raid1 [last unloaded: hwmon_vid] Oct 1 07:08:26 ns kernel: Pid: 2454, comm: ip Tainted: G W 2.6.35.4-28.fc14.i686.PAE #1 Oct 1 07:08:26 ns kernel: Call Trace: Oct 1 07:08:26 ns kernel: [<c0441f39>] warn_slowpath_common+0x6a/0x7f ... Oct 1 07:08:26 ns kernel: ---[ end trace a70d91368aa35943 ]---
Oops itself is probably suited for bug report, but what isn't clear for me is line Oct 1 07:08:26 ns kernel: Pid: 2454, comm: ip Tainted: G W 2.6.35.4-28.fc14.i686.PAE #1
Which module caused that kernel is tainted? I suppose Fedora contains only open SW including kernel modules, isn't then there something wrong?
Thanks, Franta Hanzlík
Hi, an answer from test mailing list.
From kernel doc
" 1: 'G' if all modules loaded have a GPL or compatible license, 'P' if any proprietary module has been loaded. Modules without a MODULE_LICENSE or with a MODULE_LICENSE that is not recognised by insmod as GPL compatible are assumed to be proprietary. "
" 10: 'W' if a warning has previously been issued by the kernel. (Though some warnings may set more specific taint flags.) "
JB
Hello Miss JB, i just found your answer, as you wrote it, in kernel-doc Fedora RPM package (/usr/share/doc/stable.knl/Documentation/oops-tracing.txt). I previous had some phantasy how should be "desperate" tainting module dislosured - but it was probable other kernel in other universe. Thanks
-
Franta Hanzlík -
JB