Re: [PATCH 3/9] remoteproc: add support to skip firmware load when recovery

[Arnaud POULIQUEN]

On 2/19/20 4:40 PM, Peng Fan wrote:
> 
>> Subject: Re: [PATCH 3/9] remoteproc: add support to skip firmware load when
>> recovery
>>
>> Hi,
>>
>> On 2/19/20 8:27 AM, peng.fan@xxxxxxx wrote:
>>> From: Peng Fan <peng.fan@xxxxxxx>
>>>
>>> Remote processor such as M4 inside i.MX8QXP is not handled by Linux
>>> when it is configured to run inside its own hardware partition by
>>> system control unit(SCU). So even remote processor crash reset, it is
>>> handled by SCU, not linux. To such case, firmware load should be
>>> ignored, So introduce skip_fw_load_recovery and platform driver should
>>> set it if needed.
>>>
>>> Signed-off-by: Peng Fan <peng.fan@xxxxxxx>
>>> ---
>>>  drivers/remoteproc/remoteproc_core.c | 19 +++++++++++--------
>>>  include/linux/remoteproc.h           |  1 +
>>>  2 files changed, 12 insertions(+), 8 deletions(-)
>>>
>>> diff --git a/drivers/remoteproc/remoteproc_core.c
>>> b/drivers/remoteproc/remoteproc_core.c
>>> index 876b5420a32b..ca310e3582bf 100644
>>> --- a/drivers/remoteproc/remoteproc_core.c
>>> +++ b/drivers/remoteproc/remoteproc_core.c
>>> @@ -1678,20 +1678,23 @@ int rproc_trigger_recovery(struct rproc *rproc)
>>>  	if (ret)
>>>  		goto unlock_mutex;
>>>
>>> -	/* generate coredump */
>>> -	rproc_coredump(rproc);
>>> +	if (!rproc->skip_fw_load_recovery) {
>>> +		/* generate coredump */
>>> +		rproc_coredump(rproc);
>>>
>>> -	/* load firmware */
>>> -	ret = request_firmware(&firmware_p, rproc->firmware, dev);
>>> -	if (ret < 0) {
>>> -		dev_err(dev, "request_firmware failed: %d\n", ret);
>>> -		goto unlock_mutex;
>>> +		/* load firmware */
>>> +		ret = request_firmware(&firmware_p, rproc->firmware, dev);
>>> +		if (ret < 0) {
>>> +			dev_err(dev, "request_firmware failed: %d\n", ret);
>>> +			goto unlock_mutex;
>>> +		}
>>
>> Any specific reason to not reuse skip_fw_load here?
> 
> Just thought firmware needs to be loaded by Linux when remote
> processor crash, even if it initially booted ealy.
> 
> skip_fw_load just handles first boot which no need firmware.
> But if recovery boot needs firwarem, skip_fw_load will not handle.
> 
> So I add this new bool.
> 
> Actually to my platform, skip_fw_load could work when recovery,
> I just think other platforms might need firware load when recovery.

Thanks for the clarification. For the ST platform, we can have both usecase
By clearing skip_fw_load in case of crash we s should be able to handle this usecase.
Anyway as discussion is still on going for the preloaded firmware topic, we have at least to take into account your
use case.

Thanks,
Arnaud

 
> 
> Regards,
> Peng. 
> 
>> FYI i'm reworking the Loic's patch and i plan to implement the recovery part
>> using skip_fw_load...
> 
> 
> 
>>
>> Regards
>> Arnaud
>>
>>>  	}
>>>
>>>  	/* boot the remote processor up again */
>>>  	ret = rproc_start(rproc, firmware_p);
>>>
>>> -	release_firmware(firmware_p);
>>> +	if (!rproc->skip_fw_load_recovery)
>>> +		release_firmware(firmware_p);
>>>
>>>  unlock_mutex:
>>>  	mutex_unlock(&rproc->lock);
>>> diff --git a/include/linux/remoteproc.h b/include/linux/remoteproc.h
>>> index 4fd5bedab4fa..fe6ee253b385 100644
>>> --- a/include/linux/remoteproc.h
>>> +++ b/include/linux/remoteproc.h
>>> @@ -514,6 +514,7 @@ struct rproc {
>>>  	bool has_iommu;
>>>  	bool auto_boot;
>>>  	bool skip_fw_load;
>>> +	bool skip_fw_load_recovery;
>>>  	struct list_head dump_segments;
>>>  	int nb_vdev;
>>>  };
>>>