From: Mukesh Savaliya <mukesh.savaliya@oss.qualcomm.com>
To: Jyothi Kumar Seerapu <jyothi.seerapu@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 1/2] dmaengine: qcom-gpi: Add I2C High-Speed mode configuration support
Date: Fri, 28 Aug 2026 00:56:41 +0530 [thread overview]
Message-ID: <f69d4f56-82c7-403e-98b9-f78b79ac5906@oss.qualcomm.com> (raw)
In-Reply-To: <20260824-i2c-hs-v1-1-8dd94e5c6113@oss.qualcomm.com>
On 8/24/2026 4:55 PM, Jyothi Kumar Seerapu wrote:
> Add support in the Qualcomm GPI (Generic Packet Interface) DMA engine
> to configure I2C High-Speed (HS) mode transfers.
>
> Introduce support for CONFIG1 Transfer Ring Element (TRE) to convey
> HS-specific timing parameters to the hardware. Define a new
> gpi_i2c_config1 structure containing tcycle_cnt and tlow_cnt fields,
> with default values of 28 and 38 respectively, as required for 3.4 MHz
> HS mode operation.
>
Little more enhanced -
dmaengine: qcom: gpi: Add I2C High-Speed mode frequency support
Introduce CONFIG1 Transfer Ring Element (TRE) support to program
I2C High-Speed (HS) mode timing parameters through the Qualcomm
GPI DMA engine.
Add a new gpi_i2c_config1 structure containing tcycle_cnt and
tlow_cnt fields. The fields are initialized to the hardware
recommended values of 28 and 38, enabling 3.4 MHz HS-mode I2C
transfers.
> Signed-off-by: Jyothi Kumar Seerapu <jyothi.seerapu@oss.qualcomm.com>
> ---
> drivers/dma/qcom/gpi.c | 54 +++++++++++++++++++++++++++++++++++-----
> include/linux/dma/qcom-gpi-dma.h | 28 ++++++++++++++++++++-
> 2 files changed, 75 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/dma/qcom/gpi.c b/drivers/dma/qcom/gpi.c
> index a5055a6273af..52a7800ed4f5 100644
> --- a/drivers/dma/qcom/gpi.c
> +++ b/drivers/dma/qcom/gpi.c
> @@ -21,6 +21,7 @@
> #define TRE_TYPE_IMMEDIATE_DMA 0x11
> #define TRE_TYPE_GO 0x20
> #define TRE_TYPE_CONFIG0 0x22
> +#define TRE_TYPE_CONFIG1 0x23
>
> /* TRE flags */
> #define TRE_FLAGS_CHAIN BIT(0)
> @@ -58,9 +59,14 @@
> #define TRE_I2C_C0_TX_PACK BIT(24)
> #define TRE_I2C_C0_RX_PACK BIT(25)
>
> +/* I2C Config1 WD0 */
> +#define TRE_I2C_C1_TLOW GENMASK(9, 0)
> +#define TRE_I2C_C1_TCYCLE GENMASK(19, 10)
> +
Not sure why above two not looking aligned by tab
> /* I2C GO WD0 */
> #define TRE_I2C_GO_CMD GENMASK(4, 0)
[...]
>
> - /* create the GO tre for Tx */
> - if (i2c->op == I2C_WRITE) {
> + /* Create CONFIG1 TRE if requested */
> + if (i2c->set_config1) {
can we move this set_config1 into else if part of set_config, above ?>
tre = &desc->tre[tre_idx];
> tre_idx++;
>
> + /* CONFIG1 TRE with timing parameters */
> + tre->dword[0] = u32_encode_bits(i2c->config1.tlow_cnt, TRE_I2C_C1_TLOW);
> + tre->dword[0] |= u32_encode_bits(i2c->config1.tcycle_cnt, TRE_I2C_C1_TCYCLE);
> + tre->dword[1] = 0;
> + tre->dword[2] = 0;
> + tre->dword[3] = u32_encode_bits(TRE_TYPE_CONFIG1, TRE_FLAGS_TYPE);
> + tre->dword[3] |= u32_encode_bits(1, TRE_FLAGS_CHAIN);
> + }
> +
> + /* create the GO tre for Tx */
> + if (i2c->op == I2C_WRITE || i2c->op == I2C_HS_WRITE) {
> + u8 master_code = 0; /* Default master code for Linux EE */
why for Linux EE ? what else could be ?> + u32 cmd_opcode;
> + bool is_hs_mode = false;
> +
> + tre = &desc->tre[tre_idx];
> + tre_idx++;
> +
> + is_hs_mode = (i2c->op == I2C_HS_WRITE || i2c->op == I2C_HS_READ);
> +
> + /* Determine the command opcode based on HS mode and multi_msg flag */
Comment is mostly restates the code, would write as -
/* Select HS-mode or standard I2C opcode. */
> if (i2c->multi_msg)
> - tre->dword[0] = u32_encode_bits(I2C_READ, TRE_I2C_GO_CMD);
> + cmd_opcode = is_hs_mode ? I2C_HS_READ : I2C_READ;
can you add explanation of cmd opcode into commit log too ? That would
make story describing changes.> else
> - tre->dword[0] = u32_encode_bits(i2c->op, TRE_I2C_GO_CMD);
> + /* I2C HS write vs Regular I2C write */
> + cmd_opcode = is_hs_mode ? I2C_HS_WRITE : i2c->op;
> +
what if i write as below, will be simplified ?
i2c->op = cmd_opcode ; //OR can store directly into i2c->op.
tre->dword[0] = u32_encode_bits(i2c->op, TRE_I2C_GO_CMD);
> + tre->dword[0] = u32_encode_bits(cmd_opcode, TRE_I2C_GO_CMD);
> +
> + if (is_hs_mode)
> + tre->dword[0] |= u32_encode_bits(master_code, TRE_I2C_GO_MASTER_CODE);
>
> tre->dword[0] |= u32_encode_bits(i2c->addr, TRE_I2C_GO_ADDR);
> tre->dword[0] |= u32_encode_bits(i2c->stretch, TRE_I2C_GO_STRETCH);
> @@ -1674,7 +1708,7 @@ static int gpi_create_i2c_tre(struct gchan *chan, struct gpi_desc *desc,
> tre->dword[3] |= u32_encode_bits(1, TRE_FLAGS_CHAIN);
> }
>
> - if (i2c->op == I2C_READ || i2c->multi_msg == false) {
> + if (i2c->op == I2C_READ || i2c->op == I2C_HS_READ || i2c->multi_msg == false) {
> /* create the DMA TRE */
> tre = &desc->tre[tre_idx];
> tre_idx++;
> @@ -1826,6 +1860,14 @@ gpi_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
> if (direction == DMA_DEV_TO_MEM) /* rx */
> nr_tre = 1;
>
> + /* I2C High-Speed mode sends an extra CONFIG1 TRE ahead of the GO TRE */
> + if (gchan->protocol == QCOM_GPI_I2C) {
> + struct gpi_i2c_config *i2c = gchan->config;
> +
> + if (i2c->set_config1)
> + nr_tre++;
> + }
> +
> /* calculate # of elements required & available */
> nr = gpi_ring_num_elements_avail(ch_ring);
> if (nr < nr_tre) {
> diff --git a/include/linux/dma/qcom-gpi-dma.h b/include/linux/dma/qcom-gpi-dma.h
> index 332be28427e4..6e9f09fc109a 100644
> --- a/include/linux/dma/qcom-gpi-dma.h
> +++ b/include/linux/dma/qcom-gpi-dma.h
[...]
>
> /**
> @@ -62,15 +82,19 @@ enum i2c_op {
> * @high_count: high period of clock
> * @low_count: low period of clock
> * @clk_div: source clock divider
> + * @clk_src: serial clock
source clock right ?> * @addr: i2c bus address
> * @stretch: stretch the clock at eot
> - * @set_config: set peripheral config
> + * @set_config: set peripheral config (CONFIG0)
> + * @set_config1: set peripheral config1 (CONFIG1)
> + * @config1: I2C HS mode timing configuration (CONFIG1 TRE parameters)
> * @rx_len: receive length for buffer
> * @op: i2c cmd
> * @multi_msg: is part of multi i2c r-w msgs
> */
> struct gpi_i2c_config {
> u8 set_config;
> + u8 set_config1;
> u8 pack_enable;
> u8 cycle_count;
> u8 high_count;
> @@ -78,6 +102,8 @@ struct gpi_i2c_config {
> u8 addr;
> u8 stretch;
> u16 clk_div;
> + u32 clk_src;
> + struct gpi_i2c_config1 config1;
> u32 rx_len;
> enum i2c_op op;
> bool multi_msg;
>
next prev parent reply other threads:[~2026-08-27 19:26 UTC|newest]
Thread overview: 9+ 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
2026-08-27 19:26 ` Mukesh Savaliya [this message]
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
2026-08-27 19:42 ` Mukesh Savaliya
2026-09-07 16:14 ` Jyothi Kumar Seerapu
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=f69d4f56-82c7-403e-98b9-f78b79ac5906@oss.qualcomm.com \
--to=mukesh.savaliya@oss.qualcomm.com \
--cc=Frank.Li@kernel.org \
--cc=andi.shyti@kernel.org \
--cc=dmaengine@vger.kernel.org \
--cc=jyothi.seerapu@oss.qualcomm.com \
--cc=linux-arm-msm@vger.kernel.org \
--cc=linux-i2c@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--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
all inboxes | Powered by JetHome®