From f05ac2e5930506c980817a59a47f903e25f91429 Mon Sep 17 00:00:00 2001 From: Yaniv Gardi Date: Mon, 1 Feb 2016 15:02:42 +0200 Subject: [PATCH] scsi: ufs: avoid exception event handler racing with PM callbacks If device raises the exception event in the response to the commands sent during the runtime/system PM callbacks, exception event handler might run in parallel with PM callbacks and may see unclocked register accesses. This change fixes this issue by not scheduling the exception event handler while PM callbacks are running. Reviewed-by: Gilad Broner Reviewed-by: Dolev Raviv Signed-off-by: Subhash Jadavani Signed-off-by: Yaniv Gardi Signed-off-by: Martin K. Petersen --- drivers/scsi/ufs/ufshcd.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-) diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c index e74c53a..e374b79 100644 --- a/drivers/scsi/ufs/ufshcd.c +++ b/drivers/scsi/ufs/ufshcd.c @@ -3144,7 +3144,20 @@ ufshcd_transfer_rsp_status(struct ufs_hba *hba, struct ufshcd_lrb *lrbp) scsi_status = result & MASK_SCSI_STATUS; result = ufshcd_scsi_cmd_status(lrbp, scsi_status); - if (ufshcd_is_exception_event(lrbp->ucd_rsp_ptr)) + /* + * Currently we are only supporting BKOPs exception + * events hence we can ignore BKOPs exception event + * during power management callbacks. BKOPs exception + * event is not expected to be raised in runtime suspend + * callback as it allows the urgent bkops. + * During system suspend, we are anyway forcefully + * disabling the bkops and if urgent bkops is needed + * it will be enabled on system resume. Long term + * solution could be to abort the system suspend if + * UFS device needs urgent BKOPs. + */ + if (!hba->pm_op_in_progress && + ufshcd_is_exception_event(lrbp->ucd_rsp_ptr)) schedule_work(&hba->eeh_work); break; case UPIU_TRANSACTION_REJECT_UPIU: -- 2.7.4