Re: [PATCH 08/12] ath10k: Add QMI CAP request support

[Bjorn Andersson <bjorn.andersson@linaro.org>]

On Sun 25 Mar 22:40 PDT 2018, Govind Singh wrote:

> Add support for target capablity request
> qmi message for wcn3990 target.
> 
> Signed-off-by: Govind Singh <govinds@codeaurora.org>
> ---
>  drivers/net/wireless/ath/ath10k/qmi.c | 91 +++++++++++++++++++++++++++++++++++
>  drivers/net/wireless/ath/ath10k/qmi.h | 25 ++++++++++
>  2 files changed, 116 insertions(+)
> 
> diff --git a/drivers/net/wireless/ath/ath10k/qmi.c b/drivers/net/wireless/ath/ath10k/qmi.c
> index 763b812..65a43af 100644
> --- a/drivers/net/wireless/ath/ath10k/qmi.c
> +++ b/drivers/net/wireless/ath/ath10k/qmi.c
> @@ -270,6 +270,93 @@ static int ath10k_qmi_msa_ready_send_sync_msg(struct ath10k_qmi *qmi)
>  	return ret;
>  }
>  
> +static int ath10k_qmi_cap_send_sync_msg(struct ath10k_qmi *qmi)
> +{
> +	struct wlfw_cap_resp_msg_v01 *resp;

I think this is 207 bytes, so it makes sense to allocate this on the
heap.

> +	struct wlfw_cap_req_msg_v01 *req;

But this is 1 bytes, so keep it on the stack.

> +	struct qmi_txn txn;
> +	int ret;
> +
> +	req = kzalloc(sizeof(*req), GFP_KERNEL);
> +	if (!req)
> +		return -ENOMEM;
> +
> +	resp = kzalloc(sizeof(*resp), GFP_KERNEL);
> +	if (!resp) {
> +		kfree(req);
> +		return -ENOMEM;
> +	}
> +
> +	ret = qmi_txn_init(&qmi->qmi_hdl, &txn, wlfw_cap_resp_msg_v01_ei, resp);
> +	if (ret < 0) {
> +		pr_err("fail to init txn for capability resp %d\n", ret);

Same comments as in previous patches.

> +		goto out;
> +	}
> +
> +	ret = qmi_send_request(&qmi->qmi_hdl, NULL, &txn,
> +			       QMI_WLFW_CAP_REQ_V01,
> +			       WLFW_CAP_REQ_MSG_V01_MAX_MSG_LEN,
> +			       wlfw_cap_req_msg_v01_ei, req);
> +	if (ret < 0) {
> +		qmi_txn_cancel(&txn);
> +		pr_err("fail to send capability req %d\n", ret);
> +		goto out;
> +	}
> +
> +	ret = qmi_txn_wait(&txn, WLFW_TIMEOUT * HZ);
> +	if (ret < 0)
> +		goto out;
> +
> +	if (resp->resp.result != QMI_RESULT_SUCCESS_V01) {
> +		pr_err("qmi capability request rejected, result:%d error:%d\n",
> +		       resp->resp.result, resp->resp.error);
> +		ret = -resp->resp.result;
> +		goto out;
> +	}
> +
> +	if (resp->chip_info_valid) {
> +		qmi->chip_info.chip_id = resp->chip_info.chip_id;
> +		qmi->chip_info.chip_family = resp->chip_info.chip_family;
> +	}
> +
> +	if (resp->board_info_valid)
> +		qmi->board_info.board_id = resp->board_info.board_id;
> +	else
> +		qmi->board_info.board_id = 0xFF;
> +
> +	if (resp->soc_info_valid)
> +		qmi->soc_info.soc_id = resp->soc_info.soc_id;
> +
> +	if (resp->fw_version_info_valid) {
> +		qmi->fw_version_info.fw_version =
> +			resp->fw_version_info.fw_version;
> +		strlcpy(qmi->fw_version_info.fw_build_timestamp,
> +			resp->fw_version_info.fw_build_timestamp,
> +			MAX_TIMESTAMP_LEN + 1);

Use sizeof(qmi->fw_version_info.fw_build_timestamp) to remove the risk
of them getting out of sync.

> +	}
> +
> +	if (resp->fw_build_id_valid)
> +		strlcpy(qmi->fw_build_id, resp->fw_build_id,
> +			MAX_BUILD_ID_LEN + 1);
> +
> +	pr_debug("chip_id: 0x%x, chip_family: 0x%x, board_id: 0x%x, soc_id: 0x%x, fw_version: 0x%x, fw_build_timestamp: %s, fw_build_id: %s",
> +		 qmi->chip_info.chip_id, qmi->chip_info.chip_family,
> +		 qmi->board_info.board_id, qmi->soc_info.soc_id,
> +		 qmi->fw_version_info.fw_version,
> +		 qmi->fw_version_info.fw_build_timestamp,
> +		 qmi->fw_build_id);

Use dev_dbg(), but perhaps this could even be a dev_info()?

> +
> +	pr_debug("target cap request completed\n");
> +	kfree(resp);
> +	kfree(req);
> +
> +	return 0;
> +out:
> +	kfree(resp);
> +	kfree(req);
> +	return ret;
> +}
> +
>  static int
>  ath10k_qmi_ind_register_send_sync_msg(struct ath10k_qmi *qmi)
>  {
> @@ -425,6 +512,10 @@ static void ath10k_qmi_event_server_arrive(struct work_struct *work)
>  	if (ret)
>  		goto err_setup_msa;
>  
> +	ret = ath10k_qmi_cap_send_sync_msg(qmi);
> +	if (ret)
> +		goto err_setup_msa;
> +
>  	return;
>  
>  err_setup_msa:
> diff --git a/drivers/net/wireless/ath/ath10k/qmi.h b/drivers/net/wireless/ath/ath10k/qmi.h
> index 47af020..09d20a0 100644
> --- a/drivers/net/wireless/ath/ath10k/qmi.h
> +++ b/drivers/net/wireless/ath/ath10k/qmi.h
> @@ -17,6 +17,8 @@
>  #define _QMI_H_
>  
>  #define MAX_NUM_MEMORY_REGIONS			2
> +#define MAX_TIMESTAMP_LEN			32
> +#define MAX_BUILD_ID_LEN			128
>  
>  enum ath10k_qmi_driver_event_type {
>  	ATH10K_QMI_EVENT_SERVER_ARRIVE,
> @@ -31,6 +33,24 @@ struct ath10k_msa_mem_region_info {
>  	u8 secure_flag;
>  };
>  
> +struct ath10k_qmi_chip_info {
> +	u32 chip_id;
> +	u32 chip_family;
> +};
> +
> +struct ath10k_qmi_board_info {
> +	u32 board_id;
> +};
> +
> +struct ath10k_qmi_soc_info {
> +	u32 soc_id;
> +};
> +
> +struct ath10k_qmi_fw_version_info {
> +	u32 fw_version;
> +	char fw_build_timestamp[MAX_TIMESTAMP_LEN + 1];
> +};

I don't see the benefit off lugging these around as separate structs,
why not just put them in the ath10k_qmi struct directly.

But in particular, I don't see a reason for stashing the data that will
never be used, store the parts that you're actually using.

> +
>  struct ath10k_qmi {
>  	struct platform_device *pdev;
>  	struct qmi_handle qmi_hdl;
> @@ -47,5 +67,10 @@ struct ath10k_qmi {
>  	phys_addr_t msa_pa;
>  	u32 msa_mem_size;
>  	void *msa_va;
> +	struct ath10k_qmi_chip_info chip_info;
> +	struct ath10k_qmi_board_info board_info;
> +	struct ath10k_qmi_soc_info soc_info;
> +	struct ath10k_qmi_fw_version_info fw_version_info;
> +	char fw_build_id[MAX_BUILD_ID_LEN + 1];
>  };
>  #endif /* _QMI_H_ */

Regards,
Bjorn