End of SLES install to FCoE hangs instead of rebooting

SUSE Linux Enterprise Server 15 SP6

SUSE Linux Enterprise Server 15 SP5

SUSE Linux Enterprise Server 15 SP4

SUSE Linux Enterprise Server 15 SP3

SUSE Linux Enterprise Server 15 SP2

SUSE Linux Enterprise Server 15 SP1

SUSE Linux Enterprise Server 15 SP0

During installation of SUSE Linux Enterprise Server 15 SP5 (and all previous SUSE Linux Enterprise Server 15 Service Packs) to Fiber Channel Over Ethernet (FCoE) storage using a Marvell FCoE controller and qedf driver, on some x86_64 servers the kernel crashes after the install completes and the system does not automatically reboot into the installed OS. 

[ 1340.805343][ T1135] BUG: kernel NULL pointer dereference, address: 0000000000000102
[ 1340.897641][ T1135] #PF: supervisor read access in kernel mode
[ 1340.967902][ T1135] #PF: error_code(0x0000) - not-present page
[ 1341.038168][ T1135] PGD 0 
[ 1341.070697][ T1135] Oops: 0000 [#1] PREEMPT SMP NOPTI
[ 1341.131531][ T1135] CPU: 28 PID: 1135 Comm: kworker/28:2 Tainted: G        W           5.14.21-150500.53-default #1 SLE15-SP5 3b90198179ad2dbddc570cfe6efd7895c9be3e4a
[ 1341.310813][ T1135] Hardware name: HPE Synergy 480 Gen11/Synergy 480 Gen11 Compute Module, BIOS 1.32 03/23/2023
[ 1341.432444][ T1135] Workqueue: qedf_0_ll2 qedf_ll2_process_skb [qedf]
[ 1341.510054][ T1135] RIP: 0010:__queue_work+0x1c/0x410
[ 1341.570884][ T1135] Code: 89 80 a8 00 00 00 c3 cc cc cc cc 0f 0b 90 0f 1f 44 00 00 41 57 41 56 41 55 41 54 49 89 d5 55 53 41 89 fc 48 89 f5 48 83 ec 10 <f6> 86 02 01 00 00 01 89 7c 24 04 0f 85 fc 02 00 00 48 bb eb 83 b5
[ 1341.805725][ T1135] RSP: 0018:ff6a742401437d28 EFLAGS: 00010086
[ 1341.877036][ T1135] RAX: ff16328307b80eb0 RBX: 0000000000000246 RCX: 0000000000000000
[ 1341.971411][ T1135] RDX: ff16328307b80ea8 RSI: 0000000000000000 RDI: 0000000000002000
[ 1342.065788][ T1135] RBP: 0000000000000000 R08: ff16328307b80eb0 R09: 0000000000000001
[ 1342.160158][ T1135] R10: 2064ec9b8b6496e7 R11: ff6a742401437ba8 R12: 0000000000002000
[ 1342.254530][ T1135] R13: ff16328307b80ea8 R14: 0000000000000014 R15: 0000000000000001
[ 1342.348903][ T1135] FS:  0000000000000000(0000) GS:ff163291ff500000(0000) knlGS:0000000000000000
[ 1342.454808][ T1135] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 1342.532411][ T1135] CR2: 0000000000000102 CR3: 0000000db3610006 CR4: 0000000000771ee0
[ 1342.626791][ T1135] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 1342.721159][ T1135] DR3: 0000000000000000 DR6: 00000000fffe07f0 DR7: 0000000000000400
[ 1342.815530][ T1135] PKRU: 55555554
[ 1342.856444][ T1135] Call Trace:
[ 1342.894213][ T1135]  <TASK>
[ 1342.927789][ T1135]  queue_delayed_work_on+0x3c/0x50
[ 1342.987574][ T1135]  qedf_ctx_soft_reset+0xbe/0x1c0 [qedf e776fbc03da01b8d6bbe6a4a20d9c1c4f1c163c6]
[ 1343.096628][ T1135]  ? qedf_fip_recv+0x77/0x590 [qedf e776fbc03da01b8d6bbe6a4a20d9c1c4f1c163c6]
[ 1343.201489][ T1135]  qedf_fip_recv+0x27f/0x590 [qedf e776fbc03da01b8d6bbe6a4a20d9c1c4f1c163c6]
[ 1343.305300][ T1135]  qedf_ll2_process_skb+0x2ac/0x5a0 [qedf e776fbc03da01b8d6bbe6a4a20d9c1c4f1c163c6]
[ 1343.416448][ T1135]  process_one_work+0x264/0x440
[ 1343.473087][ T1135]  ? process_one_work+0x440/0x440
[ 1343.531823][ T1135]  worker_thread+0x2d/0x3d0
[ 1343.584270][ T1135]  ? process_one_work+0x440/0x440
[ 1343.643009][ T1135]  kthread+0x154/0x180
[ 1343.690211][ T1135]  ? set_kthread_struct+0x50/0x50
[ 1343.748943][ T1135]  ret_from_fork+0x1f/0x30
[ 1343.800345][ T1135]  </TASK>

A patch fixing the controller driver to permit the automatic reboot after install has been released via the install medium (ISO) for SUSE Linux Enterprise Server 15 SP7.

Please note that the initial GM release of the SUSE Linux Enterprise Server 15 SP6 ISO remains affected by this issue, however, the subsequent QU ISO releases are not affected.

For any SUSE Linux Enterprise Server 15 SP5 or earlier install where the problem is still experienced, the install has successfully completed when the kernel crashes and is usable by power-cycling the host.

There was a race condition in the qedf driver during shutdown causing the kernel crash.

https://marc.info/?l=linux-scsi&m=171576408712922&w=2