ath11k: use MHI provided APIs to allocate and free MHI controller

Use MHI provided APIs to allocate and free MHI controller to
improve MHI host driver handling. This also fixes a memory leak
as the MHI controller was allocated but never freed.

Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
---
 drivers/net/wireless/ath/ath11k/mhi.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

+ath11k list

On Mon, Nov 16, 2020 at 04:01:40PM -0800, Bhaumik Bhatt wrote:
> Use MHI provided APIs to allocate and free MHI controller to
> improve MHI host driver handling. This also fixes a memory leak
> as the MHI controller was allocated but never freed.
> 
> Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>

Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>

Thanks,
Mani

> ---
>  drivers/net/wireless/ath/ath11k/mhi.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/net/wireless/ath/ath11k/mhi.c b/drivers/net/wireless/ath/ath11k/mhi.c
> index aded9a7..1c9d9dc 100644
> --- a/drivers/net/wireless/ath/ath11k/mhi.c
> +++ b/drivers/net/wireless/ath/ath11k/mhi.c
> @@ -218,7 +218,7 @@ int ath11k_mhi_register(struct ath11k_pci *ab_pci)
>  	struct mhi_controller *mhi_ctrl;
>  	int ret;
>  
> -	mhi_ctrl = kzalloc(sizeof(*mhi_ctrl), GFP_KERNEL);
> +	mhi_ctrl = mhi_alloc_controller();
>  	if (!mhi_ctrl)
>  		return -ENOMEM;
>  
> @@ -234,7 +234,7 @@ int ath11k_mhi_register(struct ath11k_pci *ab_pci)
>  	ret = ath11k_mhi_get_msi(ab_pci);
>  	if (ret) {
>  		ath11k_err(ab, "failed to get msi for mhi\n");
> -		kfree(mhi_ctrl);
> +		mhi_free_controller(mhi_ctrl);
>  		return ret;
>  	}
>  
> @@ -252,7 +252,7 @@ int ath11k_mhi_register(struct ath11k_pci *ab_pci)
>  	ret = mhi_register_controller(mhi_ctrl, &ath11k_mhi_config);
>  	if (ret) {
>  		ath11k_err(ab, "failed to register to mhi bus, err = %d\n", ret);
> -		kfree(mhi_ctrl);
> +		mhi_free_controller(mhi_ctrl);
>  		return ret;
>  	}
>  
> @@ -265,6 +265,7 @@ void ath11k_mhi_unregister(struct ath11k_pci *ab_pci)
>  
>  	mhi_unregister_controller(mhi_ctrl);
>  	kfree(mhi_ctrl->irq);
> +	mhi_free_controller(mhi_ctrl);
>  }
>  
>  static char *ath11k_mhi_state_to_str(enum ath11k_mhi_state mhi_state)
> -- 
> The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
> a Linux Foundation Collaborative Project
>

Bhaumik Bhatt <bbhatt@codeaurora.org> writes:

> Use MHI provided APIs to allocate and free MHI controller to
> improve MHI host driver handling.

How does it improve the handling?

> This also fixes a memory leak as the MHI controller was allocated but
> never freed.
>
> Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
> ---
>  drivers/net/wireless/ath/ath11k/mhi.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)

For ath11k patches please CC ath11k and linux-wireless lists so that
patchwork sees it. So you need to resend this.

Is this a new API? I need to understand if there are any dependencies
between mhi and ath trees, or if I can apply this directly to my ath.git
tree.

On 2020-11-16 11:14 PM, Kalle Valo wrote:
> Bhaumik Bhatt <bbhatt@codeaurora.org> writes:
> 
>> Use MHI provided APIs to allocate and free MHI controller to
>> improve MHI host driver handling.
> 
> How does it improve the handling?
> 
Main reason is we want to ensure the MHI controller is zero-initialized 
and
we want to mandate it as it is better to have it under our control in 
case
we also want to allocate and track peripheral data/memory related to the 
MHI
controller.
>> This also fixes a memory leak as the MHI controller was allocated but
>> never freed.
>> 
>> Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
>> ---
>>  drivers/net/wireless/ath/ath11k/mhi.c | 7 ++++---
>>  1 file changed, 4 insertions(+), 3 deletions(-)
> 
> For ath11k patches please CC ath11k and linux-wireless lists so that
> patchwork sees it. So you need to resend this.
> 
> Is this a new API? I need to understand if there are any dependencies
> between mhi and ath trees, or if I can apply this directly to my 
> ath.git
> tree.

This one should be dependent on the patch [1] which exists on mainline 
and
should already part of your tree if you're on any 5.10 RC.

[1] 
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/drivers/bus/mhi?h=v5.10-rc4&id=f42dfbe8f712127031e7b9bc938a1c33cec2ff57

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