David, I'm seeing that it seams like you already fix a while back in cachefiles. This is the commit in mainline, http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=c4d6d8dbf335c7fa47341654a37c53a512b519bb However, I'm still leaving a similar error with 3.10 plus your latest 3.11 patches. The reason I say similar because it's triggered by readahead and is caught when trying to free the pages in the pagelist. Do you have any pointers for me on how to go about debugging this? Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752455] BUG: Bad page state in process petabucket pfn:eb5d0e Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752471] page:ffffea003ad74380 count:0 mapcount:0 mapping: (null) index:0x0 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752478] page flags: 0x200000000001000(private_2) Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752486] Modules linked in: ceph libceph cachefiles ghash_clmulni_intel aesni_intel ablk_helper cryptd lrw gf128mul glue_helper aes_x86_64 microcode auth_rpcgss oid_registry nfsv4 nfs fscache lockd sunrpc raid10 raid456 async_pq async_xor async_memcpy async_raid6_recov async_tx raid1 multipath linear btrfs raid6_pq lzo_compress raid0 xor zlib_deflate libcrc32c Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752518] CPU: 4 PID: 1999 Comm: petabucket Not tainted 3.10.0-virtual #23 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752521] 0000000000000001 ffff880ecac4f978 ffffffff8154f6cc ffff880ecac4f998 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752526] ffffffff8111d8b7 ffff880ecac4f9b8 ffffea003ad74380 ffff880ecac4f9d8 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752531] ffffffff8111de6e 0000000000000000 ffffea003ad74380 0200000000001000 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752537] Call Trace: Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752550] [<ffffffff8154f6cc>] dump_stack+0x19/0x1b Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752557] [<ffffffff8111d8b7>] bad_page+0xc7/0x120 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752562] [<ffffffff8111de6e>] free_pages_prepare+0x10e/0x120 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752565] [<ffffffff8111f670>] free_hot_cold_page+0x40/0x170 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752568] [<ffffffff81122f57>] __put_single_page+0x27/0x30 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752571] [<ffffffff81123875>] put_page+0x25/0x40 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752578] [<ffffffffa02e8158>] ceph_readpages+0x2a8/0x400 [ceph] Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752583] [<ffffffff811224f2>] __do_page_cache_readahead+0x1b2/0x260 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752586] [<ffffffff81122911>] ra_submit+0x21/0x30 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752589] [<ffffffff811188a7>] filemap_fault+0x267/0x4a0 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752593] [<ffffffff8113a59f>] __do_fault+0x6f/0x4c0 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752599] [<ffffffff810092c3>] ? pte_mfn_to_pfn+0x93/0x110 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752603] [<ffffffff8113d450>] handle_pte_fault+0x90/0x860 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752606] [<ffffffff810092c3>] ? pte_mfn_to_pfn+0x93/0x110 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752609] [<ffffffff8100935e>] ? xen_pmd_val+0xe/0x10 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752612] [<ffffffff81005469>] ? __raw_callee_save_xen_pmd_val+0x11/0x1e Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752615] [<ffffffff8113eeb1>] handle_mm_fault+0x241/0x340 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752618] [<ffffffff815592fa>] __do_page_fault+0x1aa/0x550 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752622] [<ffffffff81004ec2>] ? xen_mc_flush+0xb2/0x1c0 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752625] [<ffffffff8100483d>] ? xen_clts+0x8d/0x190 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752627] [<ffffffff815596ae>] do_page_fault+0xe/0x10 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752632] [<ffffffff81555b18>] page_fault+0x28/0x30 Jul 15 12:31:58 pbnode-0e019a6f kernel: [4637542.752634] Disabling lock debugging due to kernel taint Thanks, -- Milosz -- Linux-cachefs mailing list Linux-cachefs@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/linux-cachefs
