I've been getting oopses (and sometimes panics) in what looks to be the memory manager page freeing functions. This has been happening in 2.6.9, 11, and 12. I cannot completely rule out faulty hardware, but x86 kernels did not have these issues. The machine is not overclocked and does not overheat. The oopses appear every few days. The oopses appear with or without adm8211 and nvidia, so I don't think they are at fault. Jul 9 03:12:46 reactor Bad page state at free_hot_cold_page (in process 'kswapd0', page ffff8100018e24d8) Jul 9 03:12:46 reactor flags:0x20000000 mapping:0000000000000000 mapcount:-16777216 count:0 Jul 9 03:12:46 reactor Backtrace: Jul 9 03:12:46 reactor Jul 9 03:12:46 reactor Call Trace:<ffffffff80157432>{bad_page+114} <ffffffff80157b87>{free_hot_cold_page+135} Jul 9 03:12:46 reactor <ffffffff80157c25>{__pagevec_free+37} <ffffffff8015cb95>{release_pages+277} Jul 9 03:12:46 reactor <ffffffff8015ce59>{__pagevec_release+25} <ffffffff8015d2ee>{invalidate_mapping_pages+174} Jul 9 03:12:46 reactor <ffffffff80189249>{shrink_icache_memory+281} <ffffffff8015d721>{shrink_slab+193} Jul 9 03:12:46 reactor <ffffffff8015eb04>{balance_pgdat+628} <ffffffff8015ed87>{kswapd+295} Jul 9 03:12:46 reactor <ffffffff80145660>{autoremove_wake_function+0} <ffffffff80145660>{autoremove_wake_function+0} Jul 9 03:12:46 reactor <ffffffff8010f163>{child_rip+8} <ffffffff8015ec60>{kswapd+0} Jul 9 03:12:46 reactor <ffffffff8010f15b>{child_rip+0} Jul 9 03:12:46 reactor Trying to fix it up, but a reboot is needed _____________________________________________________________ Jul 9 14:50:19 reactor Unable to handle kernel NULL pointer dereference at 0000000000000058 RIP: Jul 9 14:50:19 reactor <ffffffff801740e6>{try_to_release_page+38} Jul 9 14:50:19 reactor PGD 2ad53067 PUD 1433e067 PMD 0 Jul 9 14:50:19 reactor Oops: 0000 [1] Jul 9 14:50:19 reactor CPU 0 Jul 9 14:50:19 reactor Modules linked in: arc4 adm8211 nvidia Jul 9 14:50:19 reactor Pid: 212, comm: kswapd0 Tainted: P B 2.6.12-gentoo Jul 9 14:50:19 reactor RIP: 0010:[<ffffffff801740e6>] <ffffffff801740e6>{try_to_release_page+38} Jul 9 14:50:19 reactor RSP: 0018:ffff81003fdb9b70 EFLAGS: 00010286 Jul 9 14:50:19 reactor RAX: 0000000000000008 RBX: ffff81003e85cb48 RCX: ffff810001d25ee0 Jul 9 14:50:19 reactor RDX: ffff81003e85cb48 RSI: 00000000000000d0 RDI: ffff810001d25eb8 Jul 9 14:50:19 reactor RBP: ffff810001d25eb8 R08: 0000000000000002 R09: ffff81003fdb9a16 Jul 9 14:50:19 reactor R10: ffff81003fdb9a90 R11: 0000000000000020 R12: ffffffff8058d6c0 Jul 9 14:50:19 reactor R13: ffff81003fdb9e38 R14: 0000000000000000 R15: ffff81003fdb9d18 Jul 9 14:50:19 reactor FS: 0000000001655ae0(0000) GS:ffffffff806e6900(0000) knlGS:0000000000000000 Jul 9 14:50:19 reactor CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b Jul 9 14:50:19 reactor CR2: 0000000000000058 CR3: 0000000003a69000 CR4: 00000000000006e0 Jul 9 14:50:19 reactor Process kswapd0 (pid: 212, threadinfo ffff81003fdb8000, task ffff81003fcf4750) Jul 9 14:50:19 reactor Stack: ffffffff8015e32c ffffffff8058d840 ffffffff8058d840 ffffffff8058d850 Jul 9 14:50:19 reactor 0000000100000292 ffff81003e85cb48 0000000b00000256 ffffffff00000001 Jul 9 14:50:19 reactor 0000000000000017 0000000000000099 Jul 9 14:50:19 reactor Call Trace:<ffffffff8015e32c>{shrink_zone+2780} <ffffffff80157c25>{__pagevec_free+37} Jul 9 14:50:19 reactor <ffffffff8015a9b6>{kmem_freepages+198} <ffffffff8015b304>{free_block+228} Jul 9 14:50:19 reactor <ffffffff8015b491>{cache_flusharray+113} <ffffffff8015eae7>{balance_pgdat+599} Jul 9 14:50:19 reactor <ffffffff8015ed87>{kswapd+295} <ffffffff80145660>{autoremove_wake_function+0} Jul 9 14:50:19 reactor <ffffffff80145660>{autoremove_wake_function+0} <ffffffff8010f163>{child_rip+8} Jul 9 14:50:19 reactor <ffffffff8015ec60>{kswapd+0} <ffffffff8010f15b>{child_rip+0} Jul 9 14:50:19 reactor Jul 9 14:50:19 reactor Jul 9 14:50:19 reactor Code: 48 8b 40 50 48 85 c0 74 06 49 89 c3 41 ff e3 e9 36 ff ff ff Jul 9 14:50:19 reactor RIP <ffffffff801740e6>{try_to_release_page+38} RSP <ffff81003fdb9b70> Jul 9 14:50:19 reactor CR2: 0000000000000058 Reproducible: Sometimes Steps to Reproduce:
Created attachment 63041 [details] kernel config
Have you tested your memory with memtest recently?
Just ran memtest for a few hours on that machine. No errors in the standard test configuration.
Some more data: -The machine is one-cpu, no smp in kernel -Reproducible on different video cards -Reproducible with preemptible kernel on or off -Reproducible with ondemand governor/cool'n'quiet on or off -Happened on Fedora Core 3 too (2.6.9) -1 GB of RAM (2x512), socket 939 (two memory controllers) -ASUS A8V Deluxe with default BIOS configuration
Is this reproducable on vanilla-sources-2.6.13_rc3?
(In reply to comment #5) > Is this reproducable on vanilla-sources-2.6.13_rc3? I haven't been able to reproduce it yet after a day of running 2.6.13. I'll report if I get it again but that might take a while as this machine is in the process of moving.
The bug is reproduceable on 2.6.13-rc3: Jul 19 16:22:36 reactor Unable to handle kernel NULL pointer dereference at 0000000000000050 RIP: Jul 19 16:22:36 reactor <ffffffff80176156>{try_to_release_page+38} Jul 19 16:22:36 reactor PGD dc4067 PUD ce9067 PMD 0 Jul 19 16:22:36 reactor Oops: 0000 [1] Jul 19 16:22:36 reactor CPU 0 Jul 19 16:22:36 reactor Modules linked in: nvidia Jul 19 16:22:36 reactor Pid: 215, comm: kswapd0 Tainted: P 2.6.13-rc3 Jul 19 16:22:36 reactor RIP: 0010:[<ffffffff80176156>] <ffffffff80176156>{try_to_release_page+38} Jul 19 16:22:36 reactor RSP: 0000:ffff81003fdb9b70 EFLAGS: 00010286 Jul 19 16:22:36 reactor RAX: 0000000000000000 RBX: ffff81003e875248 RCX: ffff810001d03ea0 Jul 19 16:22:36 reactor RDX: ffff81003e875248 RSI: 00000000000000d0 RDI: ffff810001d03e78 Jul 19 16:22:36 reactor RBP: ffff810001d03e78 R08: 0000000000000000 R09: ffff81003fdb9a16 Jul 19 16:22:36 reactor R10: ffff81003fdb9a90 R11: ffffffff801d6380 R12: ffffffff80596868 Jul 19 16:22:36 reactor R13: ffff81003fdb9e38 R14: 0000000000000000 R15: 0000000000000001 Jul 19 16:22:36 reactor FS: 0000000000524ae0(0000) GS:ffffffff806fa800(0000) knlGS:0000000000000000 Jul 19 16:22:36 reactor CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b Jul 19 16:22:36 reactor CR2: 0000000000000050 CR3: 0000000000d29000 CR4: 00000000000006e0 Jul 19 16:22:36 reactor Process kswapd0 (pid: 215, threadinfo ffff81003fdb8000, task ffff81003fcf4760) Jul 19 16:22:36 reactor Stack: ffffffff8015f9ac ffff8100012c9d28 ffffffff805969e8 ffffffff805969e8 Jul 19 16:22:36 reactor ffffffff805969f8 ffff81003e875248 0000000e00000256 ffffffff00000000 Jul 19 16:22:36 reactor 0000000000000012 000000000000002a Jul 19 16:22:36 reactor Call Trace:<ffffffff8015f9ac>{shrink_zone+2796} <ffffffff80158f15>{__pagevec_free+37} Jul 19 16:22:36 reactor <ffffffff8015e20f>{release_pages+319} <ffffffff8015c8e4>{free_block+228} Jul 19 16:22:36 reactor <ffffffff8015ca71>{cache_flusharray+113} <ffffffff8016017c>{balance_pgdat+588} Jul 19 16:22:36 reactor <ffffffff80145f12>{prepare_to_wait+66} <ffffffff80160417>{kswapd+295} Jul 19 16:22:36 reactor <ffffffff80145fc0>{autoremove_wake_function+0} <ffffffff80145fc0>{autoremove_wake_function+0} Jul 19 16:22:36 reactor <ffffffff8010f323>{child_rip+8} <ffffffff801602f0>{kswapd+0} Jul 19 16:22:36 reactor <ffffffff8010f31b>{child_rip+0} Jul 19 16:22:36 reactor Jul 19 16:22:36 reactor Code: 48 8b 40 50 48 85 c0 74 06 49 89 c3 41 ff e3 e9 36 ff ff ff Jul 19 16:22:36 reactor RIP <ffffffff80176156>{try_to_release_page+38} RSP <ffff81003fdb9b70> Jul 19 16:22:36 reactor CR2: 0000000000000050
Looks like an upstream bug then. Please report this at http://bugzilla.kernel.org and post the new bug URL here.
