Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Sasha Levin <sashal@kernel.org>
To:     linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc:     Xiang Chen <chenxiang66@hisilicon.com>,
        John Garry <john.garry@huawei.com>,
        "Martin K . Petersen" <martin.petersen@oracle.com>,
        Sasha Levin <sashal@kernel.org>, linux-scsi@vger.kernel.org
Subject: [PATCH AUTOSEL 5.0 026/262] scsi: hisi_sas: Fix a timeout race of driver internal and SMP IO
Date:   Wed, 27 Mar 2019 13:58:01 -0400
Message-Id: <20190327180158.10245-26-sashal@kernel.org>
In-Reply-To: <20190327180158.10245-1-sashal@kernel.org>
References: <20190327180158.10245-1-sashal@kernel.org>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

From: Xiang Chen <chenxiang66@hisilicon.com>

[ Upstream commit 4790595723d4b833b18c994973d39f9efb842887 ]

For internal IO and SMP IO, there is a time-out timer for them. In the
timer handler, it checks whether IO is done according to the flag
task->task_state_lock.

There is an issue which may cause system suspended: internal IO or SMP IO
is sent, but at that time because of hardware exception (such as inject
2Bit ECC error), so IO is not completed and also not timeout. But, at that
time, the SAS controller reset occurs to recover system. It will release
the resource and set the status of IO to be SAS_TASK_STATE_DONE, so when IO
timeout, it will never complete the completion of IO and wait for ever.

[  729.123632] Call trace:
[  729.126791] [<ffff00000808655c>] __switch_to+0x94/0xa8
[  729.133106] [<ffff000008d96e98>] __schedule+0x1e8/0x7fc
[  729.138975] [<ffff000008d974e0>] schedule+0x34/0x8c
[  729.144401] [<ffff000008d9b000>] schedule_timeout+0x1d8/0x3cc
[  729.150690] [<ffff000008d98218>] wait_for_common+0xdc/0x1a0
[  729.157101] [<ffff000008d98304>] wait_for_completion+0x28/0x34
[  729.165973] [<ffff000000dcefb4>] hisi_sas_internal_task_abort+0x2a0/0x424 [hisi_sas_test_main]
[  729.176447] [<ffff000000dd18f4>] hisi_sas_abort_task+0x244/0x2d8 [hisi_sas_test_main]
[  729.185258] [<ffff000008971714>] sas_eh_handle_sas_errors+0x1c8/0x7b8
[  729.192391] [<ffff000008972774>] sas_scsi_recover_host+0x130/0x398
[  729.199237] [<ffff00000894d8a8>] scsi_error_handler+0x148/0x5c0
[  729.206009] [<ffff0000080f4118>] kthread+0x10c/0x138
[  729.211563] [<ffff0000080855dc>] ret_from_fork+0x10/0x18

To solve the issue, callback function task_done of those IOs need to be
called when on SAS controller reset.

Signed-off-by: Xiang Chen <chenxiang66@hisilicon.com>
Signed-off-by: John Garry <john.garry@huawei.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/hisi_sas/hisi_sas_main.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/scsi/hisi_sas/hisi_sas_main.c b/drivers/scsi/hisi_sas/hisi_sas_main.c
index 88ae415e907a..62d158574281 100644
--- a/drivers/scsi/hisi_sas/hisi_sas_main.c
+++ b/drivers/scsi/hisi_sas/hisi_sas_main.c
@@ -873,7 +873,8 @@ static void hisi_sas_do_release_task(struct hisi_hba *hisi_hba, struct sas_task
 		spin_lock_irqsave(&task->task_state_lock, flags);
 		task->task_state_flags &=
 			~(SAS_TASK_STATE_PENDING | SAS_TASK_AT_INITIATOR);
-		task->task_state_flags |= SAS_TASK_STATE_DONE;
+		if (!slot->is_internal && task->task_proto != SAS_PROTOCOL_SMP)
+			task->task_state_flags |= SAS_TASK_STATE_DONE;
 		spin_unlock_irqrestore(&task->task_state_lock, flags);
 	}
 
-- 
2.19.1