Hi All, I continue to work on my userspace-controlled CoW target for device-mapper and have almost completed qcow support. Recently, I've been hitting a BUG() when I do anything more than trivial reads and writes to my target (bug output included at the end of this mail). It seems as though the loopback driver (which is one of the places I'm sending bio's to) is calling bio_put() when the usage count is already zero. I'm not sure what I could be doing in my target that could be affecting the usage count on a bio. I would _really_ appreciate it if a few people could take a look at my kernel code (module/dm-cow.c) and see if anything pops out at them as incorrect: http://static.danplanet.com/dm-cow/dm_cow_04112006.tar.gz Anyone that wants to run it and see the issue for themselves can try running the 'copyusr' test, which might work with a healthy amount of luck. Thanks! -- Dan Smith IBM Linux Technology Center Open Hypervisor Team email: danms@xxxxxxxxxx kernel BUG at fs/bio.c:220! invalid opcode: 0000 [#1] SMP Modules linked in: loop dm_cow ipv6 parport_pc acl rfcomm l2cap bluetooth sunrpc video button i801 i2c_core shpchp tg3 dm_snapshot dm_zero dm_mirror ext3 jbd dm_mod mptspi mptscsih mptbase sd_mod scsi_mod CPU: 2 EIP: 0060:[<c1067b31>] Not tainted VLI EFLAGS: 00010246 (2.6.16.1 #2) EIP is at bio_put+0x24/0x2e eax: 00000000 ebx: cce658f8 ecx: f7af3d80 esi: d3e07d80 edi: 00000000 ebp: 00000000 ds: 007b es: 007b ss: 0068 Process loop0 (pid: 14450, threadinfo=f6ea6000 Stack: <0>f8885606 00000078 f88b9080 cce643ec 00000000 d3e07d80 00000200 f8885567 00000000 c10678b2 f6ea6f64 f8a58884 00000200 00000000 d3e07d80 f7fcafbc 00000200 00000000 f8a585ff f8a58884 f6ea6fc0 c1005e96 00000001 cda73124 Call Trace: [<f8885606>] clone_endio+0x9f/0xd2 [dm_mod] [<f8885567>] clone_endio+0x0/0xd2 [dm_mod] [<c10678b2>] bio_endio+0x3e/0x69 [<f8a58884>] lo_read_actor+0x0/0xe9 [loop] [<f8a585ff>] loop_thread+0x18e/0x413 [loop] [<f8a58884>] lo_read_actor+0x0/0xe9 [loop] [<c1005e96>] do_syscall_trace+0x196/0x1f5 [<c1003906>] syscall_exit_work+0x16/0x1c [<f8a58471>] loop_thread+0x0/0x413 [loop] [<c1001ac5>] kernel_thread_helper+0x5/0xb Code: c6 7c db 5b 5e 5f c3 89 c2 8b 40 38 85 c0 74 1b f0 ff 4a 38 0f 94 c0 84 c0 75 01 c3 c7 42 08 00 00 00 00 89 d0 ff 52 40 89 f6 c3 <0f> 0b dc 00 17 5a 23 c1 eb db 55 57 56 53 89 c7 8b 40 30 0f b7
Attachment:
pgp1by26GfuDR.pgp
Description: PGP signature
-- dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
