Oracle Grid Infrastructure crashes after migrating to SLES12 SP3

This document (7022277) is provided subject to the disclaimer at the end of this document.

Environment

SUSE Linux Enterprise Server 12 Service Pack 3
Oracle Grid Infrastructure 12c (12.2.0.1)
Oracle RDBMS Database 12c (12.1.0.2)

Situation

After migrating to SUSE Linux Enterprise Server 12 Service Pack 3, the server crashes with the following back-trace:

PID: 4982   TASK: ffff880425559340  CPU: 0   COMMAND: "oracle"
 #0 [ffff880425463c08] panic at ffffffff8118bf57
 #1 [ffff880425463c80] oops_end at ffffffff8101a579
 #2 [ffff880425463cb8] bio_put at ffffffff812e7beb
 #3 [ffff880425463d50] invalid_op at ffffffff8160bede
    [exception RIP: bio_put+27]
    RIP: ffffffff812e7beb  RSP: ffff880425463e00  RFLAGS: 00010246
    RAX: 0000000000000000  RBX: ffff8800363ce880  RCX: ffffea0000d81c20
    RDX: 0000000000000072  RSI: ffffea0005357300  RDI: ffff8800363ce880
    RBP: 0000000000000000   R8: ffff8800363ce880   R9: 0000000000000000
    R10: 0000000000000000  R11: 0000000000000246  R12: 0000000000000000
    R13: ffff8801ca820740  R14: ffffffff81f559c0  R15: 0000000013f86670
    ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
 #4 [ffff880425463e00] blk_rq_unmap_user at ffffffff812f509d
 #5 [ffff880425463e20] asm_cleanup_bios at ffffffffa04c6b3a [oracleasm]
 #6 [ffff880425463e50] asmfs_file_read at ffffffffa04c8b9e [oracleasm]
 #7 [ffff880425463e70] __vfs_read at ffffffff81205b73
 #8 [ffff880425463ee8] vfs_read at ffffffff812060ea
 #9 [ffff880425463f18] sys_read at ffffffff81207202
#10 [ffff880425463f50] entry_SYSCALL_64_fastpath at ffffffff8160a22e
    RIP: 00007ffff3bcfd10  RSP: 00007fffffff3628  RFLAGS: 00000246
    RAX: ffffffffffffffda  RBX: 00007ffff27f04c0  RCX: 00007ffff3bcfd10
    RDX: 0000000000000050  RSI: 00007fffffff3650  RDI: 000000000000002d
    RBP: 000000000000002e   R8: 0000000000000001   R9: 0000000000000000
    R10: 0000000000000000  R11: 0000000000000246  R12: 0000000000000000
    R13: ffff8801cb1a8800  R14: 0000000013f86670  R15: 00007ffff27f06d8
    ORIG_RAX: 0000000000000000  CS: 0033  SS: 002b


The crash is triggered when the Oracle ASM disks are accessed.

Resolution

Update oracleasm to to patch 1994 (version 2.0.8_k4.4.92_6.18-3.6.1) released on 2017-12-01

Cause

Upstream kernel development has introduced commit f9cc4472c963e64493261d834155974c48c0ff88 ("block: unexport various bio mapping helpers").

As a consequence of this, Oracle ASM must now use the function 'blk_rq_unmap_user(r->r_bio)' instead of 'bio_unmap_user(r->r_bio)'.

Additional Information

A support agreement is needed in order to obtain the patch mentioned in this document.

Disclaimer

This Support Knowledgebase provides a valuable tool for SUSE customers and parties interested in our products and solutions to acquire information, ideas and learn from one another. Materials are provided for informational, personal or non-commercial use within your organization and are presented "AS IS" WITHOUT WARRANTY OF ANY KIND.