linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
To: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>,
	vkoul@kernel.org
Cc: yung-chuan.liao@linux.intel.com, sanyog.r.kale@intel.com,
	linux-kernel@vger.kernel.org, alsa-devel@alsa-project.org
Subject: Re: [PATCH v2 1/5] soundwire: qcom: add support to missing transport params
Date: Tue, 2 Mar 2021 10:13:12 +0000	[thread overview]
Message-ID: <31eca99d-9f84-e088-443b-9aae782c2df2@linaro.org> (raw)
In-Reply-To: <00842f73-c0fb-5728-3ed0-c0a1fd75f94e@linux.intel.com>

Thanks Pierre for reviewing the patches!

On 26/02/2021 16:45, Pierre-Louis Bossart wrote:
> 
> 
> On 2/26/21 9:58 AM, Srinivas Kandagatla wrote:
>> Some of the transport parameters derived from device tree
>> are not fully parsed by the driver.
>>
>> This patch adds support to parse those missing parameters.
>>
>> Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
>> ---
>>   drivers/soundwire/qcom.c | 99 ++++++++++++++++++++++++++++++++++++++--
>>   1 file changed, 95 insertions(+), 4 deletions(-)
>>
>> diff --git a/drivers/soundwire/qcom.c b/drivers/soundwire/qcom.c
>> index 6d22df01f354..fee7465c72c2 100644
>> --- a/drivers/soundwire/qcom.c
>> +++ b/drivers/soundwire/qcom.c
>> @@ -54,7 +54,13 @@
>>   #define SWRM_MCP_SLV_STATUS                    0x1090
>>   #define SWRM_MCP_SLV_STATUS_MASK                GENMASK(1, 0)
>>   #define SWRM_DP_PORT_CTRL_BANK(n, m)    (0x1124 + 0x100 * (n - 1) + 
>> 0x40 * m)
>> +#define SWRM_DP_PORT_CTRL_2_BANK(n, m)    (0x1128 + 0x100 * (n - 1) + 
>> 0x40 * m)
>> +#define SWRM_DP_BLOCK_CTRL_1(n)        (0x112C + 0x100 * (n - 1))
>> +#define SWRM_DP_BLOCK_CTRL2_BANK(n, m)    (0x1130 + 0x100 * (n - 1) + 
>> 0x40 * m)
>> +#define SWRM_DP_PORT_HCTRL_BANK(n, m)    (0x1134 + 0x100 * (n - 1) + 
>> 0x40 * m)
>>   #define SWRM_DP_BLOCK_CTRL3_BANK(n, m)    (0x1138 + 0x100 * (n - 1) 
>> + 0x40 * m)
>> +#define SWRM_DIN_DPn_PCM_PORT_CTRL(n)    (0x1054 + 0x100 * (n - 1))
>> +
>>   #define SWRM_DP_PORT_CTRL_EN_CHAN_SHFT                0x18
>>   #define SWRM_DP_PORT_CTRL_OFFSET2_SHFT                0x10
>>   #define SWRM_DP_PORT_CTRL_OFFSET1_SHFT                0x08
>> @@ -73,12 +79,20 @@
>>   #define QCOM_SDW_MAX_PORTS    14
>>   #define DEFAULT_CLK_FREQ    9600000
>>   #define SWRM_MAX_DAIS        0xF
>> +#define SWR_INVALID_PARAM 0xFF
>> +#define SWR_HSTOP_MAX_VAL 0xF
>> +#define SWR_HSTART_MIN_VAL 0x0
>>   struct qcom_swrm_port_config {
>>       u8 si;
>>       u8 off1;
>>       u8 off2;
>>       u8 bp_mode;
>> +    u8 hstart;
>> +    u8 hstop;
>> +    u8 word_length;
>> +    u8 blk_group_count;
>> +    u8 lane_control;
>>   };
>>   struct qcom_swrm_ctrl {
>> @@ -396,7 +410,13 @@ static int qcom_swrm_port_params(struct sdw_bus 
>> *bus,
>>                    struct sdw_port_params *p_params,
>>                    unsigned int bank)
>>   {
>> -    /* TBD */
>> +    struct qcom_swrm_ctrl *ctrl = to_qcom_sdw(bus);
>> +
>> +    if (p_params->bps != SWR_INVALID_PARAM)
> 
> this is odd. sdw_port_params is a structure in 
> include/linux/soundwire/sdw.h, but here you are comparing the value with 
> a private qualcomm-defined value.
> 
> The Word length of the port is limited by the standard (64), so your 
> choice of 0xFF for SWR_INVALID_PARAM is legit, but it should be a 
> 'public' define.
> 
Am not sure if this really make sense to add it as a public define!

This is more of how the parameters are marked as not applicable for some 
ports while passing these values from device tree.


But I agree with your comments, now I have modified the code to look at 
the qcom_swrm_port_config instead of checking the "struct 
sdw_port_params" or "struct sdw_transport_params" which should make this 
INVALID flag not relevant to these data-structures anymore!


>> +        return ctrl->reg_write(ctrl,
>> +                       SWRM_DP_BLOCK_CTRL_1(p_params->num),
>> +                       p_params->bps - 1);
>> +
>>       return 0;
>>   }
>> @@ -415,10 +435,32 @@ static int qcom_swrm_transport_params(struct 
>> sdw_bus *bus,
>>       ret = ctrl->reg_write(ctrl, reg, value);
>> -    if (!ret && params->blk_pkg_mode) {
>> -        reg = SWRM_DP_BLOCK_CTRL3_BANK(params->port_num, bank);
>> +    if (params->lane_ctrl != SWR_INVALID_PARAM) {
>> +        reg = SWRM_DP_PORT_CTRL_2_BANK(params->port_num, bank);
>> +        value = params->lane_ctrl;
>> +        ret = ctrl->reg_write(ctrl, reg, value);
>> +    }
>> -        ret = ctrl->reg_write(ctrl, reg, 1);
>> +    if (params->blk_grp_ctrl != SWR_INVALID_PARAM) {
>> +        reg = SWRM_DP_BLOCK_CTRL2_BANK(params->port_num, bank);
>> +        value = params->blk_grp_ctrl;
>> +        ret = ctrl->reg_write(ctrl, reg, value);
>> +    }
>> +
>> +    if (params->hstart != SWR_INVALID_PARAM
>> +            && params->hstop != SWR_INVALID_PARAM) {
>> +        reg = SWRM_DP_PORT_HCTRL_BANK(params->port_num, bank);
>> +        value = (params->hstop << 4) | params->hstart;
>> +        ret = ctrl->reg_write(ctrl, reg, value);
>> +    } else {
>> +        reg = SWRM_DP_PORT_HCTRL_BANK(params->port_num, bank);
>> +        value = (SWR_HSTOP_MAX_VAL << 4) | SWR_HSTART_MIN_VAL;
>> +        ret = ctrl->reg_write(ctrl, reg, value);
>> +    }
>> +
>> +    if (params->blk_pkg_mode != SWR_INVALID_PARAM) {
>> +        reg = SWRM_DP_BLOCK_CTRL3_BANK(params->port_num, bank);
>> +        ret = ctrl->reg_write(ctrl, reg, params->blk_pkg_mode);
>>       }
> 
> same comments here, you should define a public define for all those 
> parameters.
> 
>>       return ret;
>> @@ -470,6 +512,17 @@ static int qcom_swrm_compute_params(struct 
>> sdw_bus *bus)
>>               p_rt->transport_params.offset1 = pcfg->off1;
>>               p_rt->transport_params.offset2 = pcfg->off2;
>>               p_rt->transport_params.blk_pkg_mode = pcfg->bp_mode;
>> +            p_rt->transport_params.blk_grp_ctrl = pcfg->blk_group_count;
>> +            p_rt->transport_params.hstart = pcfg->hstart;
>> +            p_rt->transport_params.hstop = pcfg->hstop;
>> +            p_rt->transport_params.lane_ctrl = pcfg->lane_control;
>> +            if (pcfg->word_length != SWR_INVALID_PARAM) {
>> +                sdw_fill_port_params(&p_rt->port_params,
>> +                         p_rt->num,  pcfg->word_length + 1,
>> +                         SDW_PORT_FLOW_MODE_ISOCH,
>> +                         SDW_PORT_DATA_MODE_NORMAL);
>> +            }
>> +
>>           }
>>           list_for_each_entry(s_rt, &m_rt->slave_rt_list, m_rt_node) {
>> @@ -481,6 +534,18 @@ static int qcom_swrm_compute_params(struct 
>> sdw_bus *bus)
>>                   p_rt->transport_params.offset1 = pcfg->off1;
>>                   p_rt->transport_params.offset2 = pcfg->off2;
>>                   p_rt->transport_params.blk_pkg_mode = pcfg->bp_mode;
>> +                p_rt->transport_params.blk_grp_ctrl = 
>> pcfg->blk_group_count;
>> +
>> +                p_rt->transport_params.hstart = pcfg->hstart;
>> +                p_rt->transport_params.hstop = pcfg->hstop;
>> +                p_rt->transport_params.lane_ctrl = pcfg->lane_control;
>> +                if (pcfg->word_length != SWR_INVALID_PARAM) {
>> +                    sdw_fill_port_params(&p_rt->port_params,
>> +                             p_rt->num,
>> +                             pcfg->word_length + 1,
>> +                             SDW_PORT_FLOW_MODE_ISOCH,
>> +                             SDW_PORT_DATA_MODE_NORMAL);
>> +                }
>>                   i++;
>>               }
>>           }
>> @@ -728,6 +793,11 @@ static int qcom_swrm_get_port_config(struct 
>> qcom_swrm_ctrl *ctrl)
>>       u8 off2[QCOM_SDW_MAX_PORTS];
>>       u8 si[QCOM_SDW_MAX_PORTS];
>>       u8 bp_mode[QCOM_SDW_MAX_PORTS] = { 0, };
>> +    u8 hstart[QCOM_SDW_MAX_PORTS];
>> +    u8 hstop[QCOM_SDW_MAX_PORTS];
>> +    u8 word_length[QCOM_SDW_MAX_PORTS];
>> +    u8 blk_group_count[QCOM_SDW_MAX_PORTS];
>> +    u8 lane_control[QCOM_SDW_MAX_PORTS];
>>       int i, ret, nports, val;
>>       ctrl->reg_read(ctrl, SWRM_COMP_PARAMS, &val);
>> @@ -772,11 +842,32 @@ static int qcom_swrm_get_port_config(struct 
>> qcom_swrm_ctrl *ctrl)
>>       ret = of_property_read_u8_array(np, "qcom,ports-block-pack-mode",
>>                       bp_mode, nports);
>> +
>> +    memset(hstart, SWR_INVALID_PARAM, QCOM_SDW_MAX_PORTS);
>> +    of_property_read_u8_array(np, "qcom,ports-hstart", hstart, nports);
>> +
>> +    memset(hstop, SWR_INVALID_PARAM, QCOM_SDW_MAX_PORTS);
>> +    of_property_read_u8_array(np, "qcom,ports-hstop", hstop, nports);
>> +
>> +    memset(word_length, SWR_INVALID_PARAM, QCOM_SDW_MAX_PORTS);
>> +    of_property_read_u8_array(np, "qcom,ports-word-length", 
>> word_length, nports);
>> +
>> +    memset(blk_group_count, SWR_INVALID_PARAM, QCOM_SDW_MAX_PORTS);
>> +    of_property_read_u8_array(np, "qcom,ports-block-group-count", 
>> blk_group_count, nports);
>> +
>> +    memset(lane_control, SWR_INVALID_PARAM, QCOM_SDW_MAX_PORTS);
>> +    of_property_read_u8_array(np, "qcom,ports-lane-control", 
>> lane_control, nports);
>> +
>>       for (i = 0; i < nports; i++) {
>>           ctrl->pconfig[i].si = si[i];
>>           ctrl->pconfig[i].off1 = off1[i];
>>           ctrl->pconfig[i].off2 = off2[i];
>>           ctrl->pconfig[i].bp_mode = bp_mode[i];
>> +        ctrl->pconfig[i].hstart = hstart[i];
>> +        ctrl->pconfig[i].hstop = hstop[i];
>> +        ctrl->pconfig[i].word_length = word_length[i];
>> +        ctrl->pconfig[i].blk_group_count = blk_group_count[i];
>> +        ctrl->pconfig[i].lane_control = lane_control[i];
>>       }
> 
> I don't get why you test the values parsed from DT before writing the 
> registers. Why do test them here? if some values are incorrect it's much 
> better to provide an error log instead of writing a partially valid 
> setup to hardware, no?

from DT we pass parameters for all the master ports, however some of 
these parameters are not really applicable for some of the ports! so the 
way we handle this is by marking them as 0xFF which means these values 
are not applicable for those ports! Having said that I think I should 
probably redefine SWR_INVALID_PARAM to QCOM_SWR_PARAM_NA or something on 
those lines!


--srini
>>       return 0;
>>

  reply	other threads:[~2021-03-02 10:38 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-02-26 15:58 [PATCH v2 0/5] soundwire: qcom: various improvements Srinivas Kandagatla
2021-02-26 15:58 ` [PATCH v2 1/5] soundwire: qcom: add support to missing transport params Srinivas Kandagatla
2021-02-26 16:45   ` Pierre-Louis Bossart
2021-03-02 10:13     ` Srinivas Kandagatla [this message]
2021-03-02 14:29       ` Pierre-Louis Bossart
2021-03-03  9:36         ` Srinivas Kandagatla
2021-03-03 16:16           ` Pierre-Louis Bossart
2021-02-26 15:58 ` [PATCH v2 2/5] soundwire: qcom: set continue execution flag for ignored commands Srinivas Kandagatla
2021-02-26 15:58 ` [PATCH v2 3/5] soundwire: qcom: start the clock during initialization Srinivas Kandagatla
2021-02-26 15:58 ` [PATCH v2 4/5] soundwire: qcom: update register read/write routine Srinivas Kandagatla
2021-02-26 15:58 ` [PATCH v2 5/5] soundwire: qcom: add support to new interrupts Srinivas Kandagatla

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=31eca99d-9f84-e088-443b-9aae782c2df2@linaro.org \
    --to=srinivas.kandagatla@linaro.org \
    --cc=alsa-devel@alsa-project.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pierre-louis.bossart@linux.intel.com \
    --cc=sanyog.r.kale@intel.com \
    --cc=vkoul@kernel.org \
    --cc=yung-chuan.liao@linux.intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).