[ofa-general] ofed-1.5-daily, linux kernel 2.6.30, and failure to read the uverbs ABI version

[Andrew Shewmaker]

Hi,

Thanks in advance for any info and/or help.

My understanding of the ofed-1.5-daily code was that it was based on
2.6.30 kernel code.  However, after building the daily from yesterday,
I see errors when loading the ib_umad module.  IPOIB appears to work,
but RDMA and verbs commands report "libibverbs: Fatal: couldn't read
uverbs ABI version."

What should I expect to work?

I'm not actually expecting this to work yet, but I was testing pnfs
and thought I would try ofed too.

BTW, dmesg shows this upon loading the ib_umad module:

kernel BUG at drivers/infiniband/core/verbs.c:91!
invalid opcode: 0000 [#1] SMP
last sysfs file: /sys/devices/pci0000:80/0000:80:0e.0/class
CPU 1
Modules linked in: ib_umad(+) nfsd lockd exportfs nfs_acl auth_rpcgss
sunrpc autofs4 ib_ipoib ib_cm ib_sa ipv6 ipt_LOG xt_limit xt_multiport
dm_multipath snd_pcsp ppdev snd_
seq_dummy parport_pc snd_seq_oss snd_seq_midi_event snd_seq
snd_seq_device snd_pcm_oss snd_mixer_oss parport floppy snd_pcm
ib_mthca ib_mad snd_timer i2c_nforce2 snd ib_core
 i2c_core soundcore forcedeth snd_page_alloc k8temp hwmon sata_nv
pata_acpi pata_amd ata_generic [last unloaded: scsi_wait_scan]
Pid: 481, comm: modprobe Tainted: G        W  2.6.30-6.pnfs.fc9.x86_64 #1 H8DCE
RIP: 0010:[<ffffffffa0054855>]  [<ffffffffa0054855>]
rdma_node_get_transport+0x2d/0x47 [ib_core]
RSP: 0018:ffff880042f0dbc8  EFLAGS: 00010297
RAX: 0000000000000000 RBX: ffff88007e26c000 RCX: 0000000000170001
RDX: 0000000000000000 RSI: ffffffff814be723 RDI: 0000000000000000
RBP: ffff880042f0dbd8 R08: 000000004e519652 R09: 0000000000000000
R10: ffff88013ae3b958 R11: 000000004e519652 R12: ffffffffa02b3900
R13: 0000000000000000 R14: 0000000001294120 R15: 0000000001294138
FS:  00007fff455526f0(0000) GS:ffffc200001d7000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00007fff3d7a8bf8 CR3: 0000000139971000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process modprobe (pid: 481, threadinfo ffff880042f0c000, task ffff88007e1c47c0)
Stack:
 000000004e519652 000000004e519652 ffff880042f0dc58 ffffffffa02b1afb
 000000004e519652 000000004e519652 0000000000000246 ffff88007e26c090
 ffff880042f0dc58 ffffffffa0057b7e 0000000001294120 0000000001294138
Call Trace:
 [<ffffffffa02b1afb>] ib_umad_add_one+0x37/0x480 [ib_umad]
 [<ffffffffa0057b7e>] ? add_client_context+0xdf/0x104 [ib_core]
 [<ffffffffa0057c13>] ib_register_client+0x70/0xc2 [ib_core]
 [<ffffffffa02b60aa>] ib_umad_init+0xaa/0xf6 [ib_umad]
 [<ffffffff810c248d>] ? marker_update_probe_range+0x3a/0x276
 [<ffffffffa02b6000>] ? ib_umad_init+0x0/0xf6 [ib_umad]
 [<ffffffff8100a0b6>] do_one_initcall+0x84/0x1ad
 [<ffffffff8108a23a>] ? print_lock_contention_bug+0x2d/0x133
 [<ffffffff814bcbfb>] ? __mutex_lock_common+0x362/0x3c3
 [<ffffffff81241ccd>] ? __up_read+0x2d/0xaf
 [<ffffffff81088761>] ? trace_hardirqs_on_caller+0x32/0x172
 [<ffffffff810888c1>] ? trace_hardirqs_on+0x20/0x36
 [<ffffffff81241ccd>] ? __up_read+0x2d/0xaf
 [<ffffffff81241ccd>] ? __up_read+0x2d/0xaf
 [<ffffffff81088761>] ? trace_hardirqs_on_caller+0x32/0x172
 [<ffffffff810888c1>] ? trace_hardirqs_on+0x20/0x36
 [<ffffffff81241d2f>] ? __up_read+0x8f/0xaf
 [<ffffffff8107a218>] ? up_read+0x3a/0x55
 [<ffffffff8107b177>] ? __blocking_notifier_call_chain+0x72/0x98
 [<ffffffff810956c5>] sys_init_module+0xbf/0x1fd
 [<ffffffff81013082>] system_call_fastpath+0x16/0x1b
Code: 89 e5 48 83 ec 10 0f 1f 44 00 00 65 48 8b 04 25 28 00 00 00 48
89 45 f8 31 c0 83 ff 01 72 0c 83 ff 03 76 0b 83 ff 04 b0 01 74 04 <0f>
0b eb fe 48 8b 55 f8 65 48 33 14 25 28 00 00 00 74 05 e8 fb
RIP  [<ffffffffa0054855>] rdma_node_get_transport+0x2d/0x47 [ib_core]
 RSP <ffff880042f0dbc8>
---[ end trace 4e46d3e076693c7d ]---

Thanks again.

-- 
Andrew Shewmaker