[PATCH] mpt3sas: Do scsi_remove_host() before deleting SAS PHY objects
From: Calvin Owens
Date: Fri May 13 2016 - 16:28:45 EST
On the hardware I'm testing on, simply removing the mpt3sas module
triggers a litany of WARNs culminating in an OOPS:
------------[ cut here ]------------
WARNING: CPU: 5 PID: 13348 at lib/kobject.c:244 kobject_add_internal+0x359/0x8a0
kobject_add_internal failed for ArrayDevice09 (error: -2 parent: 6:0:15:0)
CPU: 5 PID: 13348 Comm: rmmod Not tainted 4.6.0-rc2-mpt3sas-debug-00001-g7e7e6f4 #2
Hardware name: Wiwynn HoneyBadger/PantherPlus, BIOS HBP6.6 11/20/2015
ffffffff82b88ec0 ffff8806ce76f820 ffffffff81deff03 ffff8806ce76f898
0000000000000000 ffff8806ce76f868 ffffffff811191e2 ffff880749819af8
00000000000000f4 ffffed00d9cedf0f ffff880749819b08 ffff88074c9b8168
Call Trace:
[<ffffffff81deff03>] dump_stack+0x67/0x94
[<ffffffff811191e2>] __warn+0x172/0x1b0
[<ffffffff811192b7>] warn_slowpath_fmt+0x97/0xb0
[<ffffffff81df72d9>] kobject_add_internal+0x359/0x8a0
[<ffffffff81df792e>] kobject_add+0x10e/0x1c0
[<ffffffff820b4cda>] device_add+0x30a/0x1490
[<ffffffffa036a222>] enclosure_remove_device+0x172/0x1cc [enclosure]
[<ffffffffa0390254>] ses_intf_remove+0x1c4/0x270 [ses]
[<ffffffff820b1b3b>] device_del+0x2ab/0x680
[<ffffffff820b1f22>] device_unregister+0x12/0x30
[<ffffffff821264c5>] __scsi_remove_device+0x1d5/0x250
[<ffffffff821225ec>] scsi_forget_host+0x12c/0x1e0
[<ffffffff820ff4dc>] scsi_remove_host+0x10c/0x300
[<ffffffffa0066c31>] scsih_remove+0x321/0x680 [mpt3sas]
[<ffffffff81ebe5e0>] pci_device_remove+0x70/0x110
[<ffffffff820bc7a0>] __device_release_driver+0x160/0x3a0
[<ffffffff820bdd33>] driver_detach+0x183/0x200
[<ffffffff820bbbbf>] bus_remove_driver+0xdf/0x200
[<ffffffff820be847>] driver_unregister+0x67/0xa0
[<ffffffff81ebc50e>] pci_unregister_driver+0x1e/0xe0
[<ffffffffa0093e0a>] _mpt3sas_exit+0x23/0x219 [mpt3sas]
[<ffffffff81291e7e>] SyS_delete_module+0x2ee/0x390
[<ffffffff82924aa5>] entry_SYSCALL_64_fastpath+0x18/0xa8
---[ end trace fe163024b624f4af ]---
general protection fault: 0000 [#1] SMP KASAN
CPU: 6 PID: 17388 Comm: rmmod Tainted: G W 4.6.0-rc2-00001-g7e7e6f4 #1
Hardware name: Wiwynn HoneyBadger/PantherPlus, BIOS HBP6.6 11/20/2015
task: ffff880753731740 ti: ffff8806896f0000 task.ti: ffff8806896f0000
RIP: 0010:[<ffffffff8290972f>] [<ffffffff8290972f>] klist_put+0x1f/0x160
RSP: 0018:ffff8806896f7a10 EFLAGS: 00010202
RAX: dffffc0000000000 RBX: 0000000000000000 RCX: ffff880753731f48
RDX: 000000000000000b RSI: 0000000000000001 RDI: 0000000000000058
RBP: ffff8806896f7a30 R08: 0000000000000006 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000000 R12: ffff880752bec000
R13: 0000000000000001 R14: dffffc0000000000 R15: ffff880752bec2b0
FS: 00007feef8ea6700(0000) GS:ffff88075ef80000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007fe45044d020 CR3: 00000006f7092000 CR4: 00000000001006e0
Stack:
0000000000000000 ffff880752bec000 0000000000000000 dffffc0000000000
ffff8806896f7a40 ffffffff8290987e ffff8806896f7b00 ffffffff820aa3bd
ffff8806896f7aa0 1ffff100d12def4f ffff880752bec390 ffffffff829183ca
Call Trace:
[<ffffffff8290987e>] klist_del+0xe/0x10
[<ffffffff820aa3bd>] device_del+0x12d/0x680
[<ffffffff820aa922>] device_unregister+0x12/0x30
[<ffffffffa0368bb0>] enclosure_unregister+0xe0/0x170 [enclosure]
[<ffffffffa0390220>] ses_intf_remove+0x190/0x270 [ses]
[<ffffffff820aa53b>] device_del+0x2ab/0x680
[<ffffffff820aa922>] device_unregister+0x12/0x30
[<ffffffff8211ed65>] __scsi_remove_device+0x1d5/0x250
[<ffffffff8211ae8c>] scsi_forget_host+0x12c/0x1e0
[<ffffffff820f7d6c>] scsi_remove_host+0x10c/0x300
[<ffffffffa00668f1>] scsih_remove+0x321/0x680 [mpt3sas]
[<ffffffff81eb7170>] pci_device_remove+0x70/0x110
[<ffffffff820b51a0>] __device_release_driver+0x160/0x3a0
[<ffffffff820b6733>] driver_detach+0x183/0x200
[<ffffffff820b45bf>] bus_remove_driver+0xdf/0x200
[<ffffffff820b7247>] driver_unregister+0x67/0xa0
[<ffffffff81eb509e>] pci_unregister_driver+0x1e/0xe0
[<ffffffffa009359a>] _mpt3sas_exit+0x23/0xa89 [mpt3sas]
[<ffffffff81291e7e>] SyS_delete_module+0x2ee/0x390
[<ffffffff8291d1e5>] entry_SYSCALL_64_fastpath+0x18/0xa8
The issue is that enclosure_remove_device() expects to be able to re-add
the device that was previously enclosured: so with SES running, the order
we unwind things matters in a way it otherwise wouldn't.
Since mpt3sas deletes the SAS objects before the SCSI hosts, enclosure
ends up trying to re-parent the SCSI device from the enclosure to the SAS
PHY which has already been deleted. This obviously ends in sadness.
The fix appears to be simple: just call scsi_remove_host() before we call
sas_port_delete() and/or sas_remove_host().
Signed-off-by: Calvin Owens <calvinowens@xxxxxx>
---
drivers/scsi/mpt3sas/mpt3sas_scsih.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/scsi/mpt3sas/mpt3sas_scsih.c b/drivers/scsi/mpt3sas/mpt3sas_scsih.c
index e0e4920..4aa128a 100644
--- a/drivers/scsi/mpt3sas/mpt3sas_scsih.c
+++ b/drivers/scsi/mpt3sas/mpt3sas_scsih.c
@@ -8149,6 +8149,8 @@ void scsih_remove(struct pci_dev *pdev)
_scsih_raid_device_remove(ioc, raid_device);
}
+ scsi_remove_host(shost);
+
/* free ports attached to the sas_host */
list_for_each_entry_safe(mpt3sas_port, next_port,
&ioc->sas_hba.sas_port_list, port_list) {
@@ -8172,7 +8174,6 @@ void scsih_remove(struct pci_dev *pdev)
}
sas_remove_host(shost);
- scsi_remove_host(shost);
mpt3sas_base_detach(ioc);
spin_lock(&gioc_lock);
list_del(&ioc->list);
--
2.8.0.rc2