DMA Engine development
 help / color / mirror / Atom feed
From: Jyothi Kumar Seerapu <jyothi.seerapu@oss.qualcomm.com>
To: Mukesh Savaliya <mukesh.savaliya@oss.qualcomm.com>,
	Vinod Koul <vkoul@kernel.org>, Frank Li <Frank.Li@kernel.org>,
	Viken Dadhaniya <viken.dadhaniya@oss.qualcomm.com>,
	Andi Shyti <andi.shyti@kernel.org>
Cc: linux-arm-msm@vger.kernel.org, dmaengine@vger.kernel.org,
	linux-kernel@vger.kernel.org, linux-i2c@vger.kernel.org
Subject: Re: [PATCH 2/2] i2c: qcom-geni: Add support for I2C High-Speed mode
Date: Mon, 7 Sep 2026 21:44:30 +0530	[thread overview]
Message-ID: <649f44a6-740b-46b3-86f4-990878ea0968@oss.qualcomm.com> (raw)
In-Reply-To: <419902d5-cbb6-4d8f-acb6-d9640adb56d9@oss.qualcomm.com>



On 8/28/2026 1:12 AM, Mukesh Savaliya wrote:
> 
> 
Hi Mukesh, Thanks for the review comments.> On 8/24/2026 4:55 PM, Jyothi 
Kumar Seerapu wrote:
> [...]>   struct geni_i2c_err_log {
>> @@ -162,9 +180,14 @@ static const struct geni_i2c_err_log gi2c_log[] = {
>>   struct geni_i2c_clk_fld {
>>       u32    clk_freq_out;
>>       u8    clk_div;
>> -    u8    t_high_cnt;
>> -    u8    t_low_cnt;
>> -    u8    t_cycle_cnt;
>> +    /*
>> +     * In normal mode, these counter values fit within 8 bits.
>> +     * In High-Speed mode, HS_TLOW_COUNT and HS_TCYCLE_COUNT are
>> +     * 10-bit fields, so u16 is required.
>> +     */
> Now, normal mode doesn't matter once switched to 16 bit.
> May be comment also doesn't matter once change is merged, you can add 
> some information in commit log though. Will be taken as default size.> 
Sure, will remove these comments in V2.+
> u16    t_high_cnt;
>> +    u16    t_low_cnt;
>> +    u16    t_cycle_cnt;
>>   };
>>   /*
>> @@ -194,10 +217,30 @@ static const struct geni_i2c_clk_fld 
>> geni_i2c_clk_map_32mhz[] = {
>>       {}
>>   };
>> +/* source_clock = 100 MHz */
>> +static const struct geni_i2c_clk_fld geni_i2c_clk_map_100mhz[] = {
>> +    { I2C_MAX_STANDARD_MODE_FREQ, 1, 449, 548, 998 },
>> +    { I2C_MAX_FAST_MODE_FREQ, 1, 76, 167, 248 },
>> +    { I2C_MAX_FAST_MODE_PLUS_FREQ, 1, 23, 59, 98 },
>> +    {}
>> +};
>> +
>>   static int geni_i2c_clk_map_idx(struct geni_i2c_dev *gi2c)
>>   {
>>       const struct geni_i2c_clk_fld *itr;
>> +    /* Check if HS mode is requested */
>> +    if (gi2c->clk_freq_out == I2C_HS_MODE_FREQ) {
>> +        gi2c->is_hs_mode = true;
>> +        /* For HS mode, source clock should be 100 MHz */
>> +        itr = geni_i2c_clk_map_100mhz;
>> +        /* For HS mode, start with 1MHz for master code */
>> +        gi2c->clk_fld = &itr[2];
>> +        return 0;
>> +    }
>> +
> else part ?
No, else is not required here, as the if condition for I2C HS mode case 
returns from the function. For non-HS mode case the below  will execute.

+    gi2c->is_hs_mode = false;
>> +
>>       if (clk_get_rate(gi2c->se.clk) == 32 * HZ_PER_MHZ)
>>           itr = geni_i2c_clk_map_32mhz;
>>       else
>> @@ -219,7 +262,12 @@ static int qcom_geni_i2c_conf(struct geni_se *se, 
>> unsigned long freq)
>>       const struct geni_i2c_clk_fld *itr = gi2c->clk_fld;
>>       u32 val;
>> -    writel_relaxed(0, gi2c->se.base + SE_GENI_CLK_SEL);
>> +    if (gi2c->is_hs_mode) {
>> +        writel_relaxed(I2C_HS_TCYCLE_CNT, gi2c->se.base + 
>> SE_I2C_HS_TCYCLE_REG);
>> +        writel_relaxed(I2C_HS_TLOW_CNT, gi2c->se.base + 
>> SE_I2C_HS_TLOW_REG);
>> +    }
>> +
>> +    writel_relaxed(gi2c->dfs_index, gi2c->se.base + SE_GENI_CLK_SEL);
>>       val = (itr->clk_div << CLK_DIV_SHFT) | SER_CLK_EN;
>>       writel_relaxed(val, gi2c->se.base + GENI_SER_M_CLK_CFG);
>> @@ -500,7 +548,11 @@ static int geni_i2c_rx_one_msg(struct 
>> geni_i2c_dev *gi2c, struct i2c_msg *msg,
>>           geni_se_select_mode(se, GENI_SE_FIFO);
>>       writel_relaxed(len, se->base + SE_I2C_RX_TRANS_LEN);
>> -    geni_se_setup_m_cmd(se, I2C_READ, m_param);
>> +
>> +    if (gi2c->is_hs_mode) 
> gi2c->is_hs_mode ? i2c->op = I2C_HS_READ : I2C_READ;
> geni_se_setup_m_cmd(se, i2c->op, m_param);> +        
> geni_se_setup_m_cmd(se, I2C_HS_READ, m_param);
>> +    else
>> +        geni_se_setup_m_cmd(se, I2C_READ, m_param);
>>       if (dma_buf && geni_se_rx_dma_prep(se, dma_buf, len, &rx_dma)) {
>>           geni_se_select_mode(se, GENI_SE_FIFO);
>> @@ -539,7 +591,11 @@ static int geni_i2c_tx_one_msg(struct 
>> geni_i2c_dev *gi2c, struct i2c_msg *msg,
>>           geni_se_select_mode(se, GENI_SE_FIFO);
>>       writel_relaxed(len, se->base + SE_I2C_TX_TRANS_LEN);
>> -    geni_se_setup_m_cmd(se, I2C_WRITE, m_param);
>> +
>> +    if (gi2c->is_hs_mode)
>> +        geni_se_setup_m_cmd(se, I2C_HS_WRITE, m_param);
>> +    else
>> +        geni_se_setup_m_cmd(se, I2C_WRITE, m_param);
> same as above>
Yes this can be set like below:
geni_se_setup_m_cmd(se, gi2c->is_hs_mode ? I2C_HS_WRITE : I2C_WRITE, 
m_param);

>>       if (dma_buf && geni_se_tx_dma_prep(se, dma_buf, len, &tx_dma)) {
>>           geni_se_select_mode(se, GENI_SE_FIFO);
>> @@ -700,7 +756,7 @@ static int geni_i2c_gpi(struct geni_i2c_dev *gi2c, 
>> struct i2c_msg msgs[],
>>           goto out;
>>       }
> 
> [...]
> 
>> @@ -1046,6 +1122,39 @@ static int geni_i2c_init(struct geni_i2c_dev 
>> *gi2c)
>>           return ret;
>>       }
>> +    /*
>> +     * For I2C High-Speed mode, first verify QUP HW version supports it
>> +     * (requires QUPv3 core >= 4.3 per HPG), then configure 100 MHz 
>> source clock.
>> +     */
> /* HS mode requires QUPv3 version >= 4.3 and source clock=100 MHz */
Sure, will update in V2.>> +    if (gi2c->is_hs_mode) {
>> +        u32 hw_ver = geni_se_get_qup_hw_version(&gi2c->se);
>> +        u32 major = GENI_SE_VERSION_MAJOR(hw_ver);
>> +        u32 minor = GENI_SE_VERSION_MINOR(hw_ver);
>> +
>> +        if (major < QUP_I2C_HS_MIN_MAJOR ||
>> +            (major == QUP_I2C_HS_MIN_MAJOR && minor < 
>> QUP_I2C_HS_MIN_MINOR)) {
>> +            dev_err(gi2c->se.dev,
>> +                "QUP HW v%u.%u does not support I2C HS mode (requires 
>> >= %u.%u)\n",
>> +                major, minor,
>> +                QUP_I2C_HS_MIN_MAJOR, QUP_I2C_HS_MIN_MINOR);
> then why it entered into is_hs_mode condition ?
Need to check QUP version only for I2C HS mode requests, for other I2C 
modes no need of QUP version check. > +            ret = -
> EOPNOTSUPP;
>> +            goto err;
>> +        }
>> +
>> +        ret = geni_se_clk_freq_match(&gi2c->se, I2C_HS_SRC_CLK_FREQ,
>> +                         &gi2c->dfs_index, &freq_out, false);
>> +        if (ret) {
>> +            dev_err(gi2c->se.dev, "Failed to get DFS index for HS 
>> mode: %d\n", ret);
>> +            goto err;
>> +        }
>> +
>> +        ret = clk_set_rate(gi2c->se.clk, freq_out);
>> +        if (ret) {
>> +            dev_err(gi2c->se.dev, "Failed to set HS mode clock rate: 
>> %d\n", ret);
>> +            goto err;
>> +        }
>> +    }
>> +
>>       proto = geni_se_read_proto(&gi2c->se);
>>       if (proto == GENI_SE_INVALID_PROTO) {
>>           ret = geni_load_se_firmware(&gi2c->se, GENI_SE_I2C);
>>
> 


  reply	other threads:[~2026-09-07 16:14 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-08-24 11:25 [PATCH 0/2] i2c: Add I2C High-Speed mode support for qcom-geni Jyothi Kumar Seerapu
2026-08-24 11:25 ` [PATCH 1/2] dmaengine: qcom-gpi: Add I2C High-Speed mode configuration support Jyothi Kumar Seerapu
2026-08-24 11:36   ` sashiko-bot
2026-08-25 20:31     ` Frank Li
2026-08-26  6:22       ` Jyothi Kumar Seerapu
2026-08-27 19:26   ` Mukesh Savaliya
2026-09-07 16:16     ` Jyothi Kumar Seerapu
2026-08-24 11:25 ` [PATCH 2/2] i2c: qcom-geni: Add support for I2C High-Speed mode Jyothi Kumar Seerapu
2026-08-24 11:39   ` sashiko-bot
2026-08-27 19:42   ` Mukesh Savaliya
2026-09-07 16:14     ` Jyothi Kumar Seerapu [this message]
2026-08-27 19:26 ` [PATCH 0/2] i2c: Add I2C High-Speed mode support for qcom-geni Mukesh Savaliya
2026-09-07 16:14   ` Jyothi Kumar Seerapu

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=649f44a6-740b-46b3-86f4-990878ea0968@oss.qualcomm.com \
    --to=jyothi.seerapu@oss.qualcomm.com \
    --cc=Frank.Li@kernel.org \
    --cc=andi.shyti@kernel.org \
    --cc=dmaengine@vger.kernel.org \
    --cc=linux-arm-msm@vger.kernel.org \
    --cc=linux-i2c@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mukesh.savaliya@oss.qualcomm.com \
    --cc=viken.dadhaniya@oss.qualcomm.com \
    --cc=vkoul@kernel.org \
    /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