Re: [PATCH] bus: mhi: core: Resume success on SYS_ERR

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 2021-05-31 01:30 AM, Loic Poulain wrote:
When a crash occurs while in suspended state M3/D3hot, the host
only discovers SYS_ERR pm-state change on resume. Handle this
state as valid transition for resume procedure, giving a chance
to the MHI core to manage the state (sbl transition, etc...).

Signed-off-by: Loic Poulain <loic.poulain@xxxxxxxxxx>
---
 drivers/bus/mhi/core/pm.c | 21 ++++++++++++++++++++-
 1 file changed, 20 insertions(+), 1 deletion(-)

diff --git a/drivers/bus/mhi/core/pm.c b/drivers/bus/mhi/core/pm.c
index 704a5e2..e8eae4f 100644
--- a/drivers/bus/mhi/core/pm.c
+++ b/drivers/bus/mhi/core/pm.c
@@ -898,8 +898,27 @@ int mhi_pm_resume(struct mhi_controller *mhi_cntrl)
 	if (MHI_PM_IN_ERROR_STATE(mhi_cntrl->pm_state))
 		return -EIO;

-	if (mhi_get_mhi_state(mhi_cntrl) != MHI_STATE_M3)
+	if (mhi_get_mhi_state(mhi_cntrl) == MHI_STATE_SYS_ERR) {
+		/* If we are in SYS_ERR state, let MHI stack manages the error
Nitpick: manage (instead of manages)
+		 * and resume successfully.
+		 */
+		dev_warn(dev, "Entered error while suspended\n");
+
+		write_lock_irq(&mhi_cntrl->pm_lock);
+		cur_state = mhi_tryset_pm_state(mhi_cntrl, MHI_PM_SYS_ERR_DETECT);
+		write_unlock_irq(&mhi_cntrl->pm_lock);
+
+		if (cur_state == MHI_PM_SYS_ERR_DETECT) {
+			mhi_pm_sys_err_handler(mhi_cntrl);
+			return 0;
+		}
+	}
+
+	if (mhi_get_mhi_state(mhi_cntrl) != MHI_STATE_M3) {
+		dev_warn(dev, "Is not in M3 state, resume failed %d\n",
Maybe this log can be improved to say, "Resume failed. Current state: %d, expected M3.\n"
+			 mhi_get_mhi_state(mhi_cntrl));
 		return -EINVAL;
+	}

 	/* Notify clients about exiting LPM */
 	list_for_each_entry_safe(itr, tmp, &mhi_cntrl->lpm_chans, node) {

We can add RDDM handling here as well since this hole exists in RDDM handling too. This means we can include the below within if (cur_state == MHI_PM_SYS_ERR_DETECT):

if (mhi_cntrl->rddm_image && mhi_is_active(mhi_cntrl)) {
	mhi_cntrl->status_cb(mhi_cntrl, MHI_CB_EE_RDDM);
	mhi_cntrl->ee = ee;
	wake_up_all(&mhi_cntrl->state_event);
	return 0;
}

It can be added in a separate patch as well following this one if you're fine
with that. In its current state + nitpick addressed, you can add my:

Reviewed-by: Bhaumik Bhatt <bbhatt@xxxxxxxxxxxxxx>

Thanks,
Bhaumik
---
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project



[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [Linux for Sparc]     [IETF Annouce]     [Security]     [Bugtraq]     [Linux MIPS]     [ECOS]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux