Linux kernel and device drivers for NXP i.MX platforms
 help / color / mirror / Atom feed
* [PATCH] i2c: imx: add I2C_M_REV_DIR_ADDR and I2C_M_NOSTART
@ 2026-08-06 22:37 Vincent Jardin via B4 Relay
  2026-08-06 22:48 ` sashiko-bot
  0 siblings, 1 reply; 3+ messages in thread
From: Vincent Jardin via B4 Relay @ 2026-08-06 22:37 UTC (permalink / raw)
  To: Oleksij Rempel, Pengutronix Kernel Team, Andi Shyti, Frank Li,
	Sascha Hauer, Fabio Estevam
  Cc: linux-i2c, imx, linux-arm-kernel, linux-kernel, Vincent Jardin

From: Vincent Jardin <vjardin@free.fr>

The Realtek RTL8366SE SMI read frame per:

  S Addr Rd [A] reg[7:0] [A] reg[15:8] [A] [data[7:0]] A [data[15:8]] NA P

Linux support it using I2C_M_REV_DIR_ADDR on a write message
that inverts the transmitted R/W bit, and I2C_M_NOSTART on the following
read message that continues the frame without re-addressing.

This NXP i2c-imx is missing such support, so on an LX2160A the Realtek switch
could not be used.

Signed-off-by: Vincent Jardin <vjardin@free.fr>
---
Follow the logics from i2c-algo-bit and i2c-s3c2410 (addr ^= 1).

Documentation/i2c/i2c-protocol.rst says I2C_M_NOSTART "may also be used
between direction changes by some rare devices"; the RTL8366SE is one of
such device !

- The NOSTART read turns the bus around without waiting for a completion,
  because the controller stretches SCL after the previous byte and the
  frame is therefore still open. That is what lets i2c_imx_read() enter
  the state the ISR would otherwise have moved to.

- I2C_FUNC_PROTOCOL_MANGLING also covers I2C_M_IGNORE_NAK,
  I2C_M_NO_RD_ACK and I2C_M_STOP, which this driver does not implement.
  They are rejected with -EOPNOTSUPP rather than advertised and silently
  ignored. I2C_M_NOSTART on the first message is rejected as well.
---
 drivers/i2c/busses/i2c-imx.c | 143 +++++++++++++++++++++++++++++++------------
 1 file changed, 103 insertions(+), 40 deletions(-)

diff --git a/drivers/i2c/busses/i2c-imx.c b/drivers/i2c/busses/i2c-imx.c
index 9477d814fde9..81c1bd461606 100644
--- a/drivers/i2c/busses/i2c-imx.c
+++ b/drivers/i2c/busses/i2c-imx.c
@@ -1000,16 +1000,10 @@ static inline int i2c_imx_isr_write(struct imx_i2c_struct *i2c_imx)
 	return 1;
 }
 
-static inline int i2c_imx_isr_read(struct imx_i2c_struct *i2c_imx)
+static inline void i2c_imx_setup_read(struct imx_i2c_struct *i2c_imx)
 {
-	int result;
 	unsigned int temp;
 
-	result = i2c_imx_isr_acked(i2c_imx);
-	if (result)
-		return result;
-
-	/* setup bus to read data */
 	temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2CR);
 	temp &= ~I2CR_MTX;
 	if ((i2c_imx->msg->len - 1) || (i2c_imx->msg->flags & I2C_M_RECV_LEN))
@@ -1017,6 +1011,18 @@ static inline int i2c_imx_isr_read(struct imx_i2c_struct *i2c_imx)
 
 	imx_i2c_write_reg(temp, i2c_imx, IMX_I2C_I2CR);
 	imx_i2c_read_reg(i2c_imx, IMX_I2C_I2DR); /* dummy read */
+}
+
+static inline int i2c_imx_isr_read(struct imx_i2c_struct *i2c_imx)
+{
+	int result;
+
+	result = i2c_imx_isr_acked(i2c_imx);
+	if (result)
+		return result;
+
+	/* setup bus to read data */
+	i2c_imx_setup_read(i2c_imx);
 
 	return 0;
 }
@@ -1172,6 +1178,16 @@ static irqreturn_t i2c_imx_isr(int irq, void *dev_id)
 	return i2c_imx_master_isr(i2c_imx, status);
 }
 
+static u8 i2c_imx_addr_byte(struct i2c_msg *msg)
+{
+	u8 addr = i2c_8bit_addr_from_msg(msg);
+
+	if (msg->flags & I2C_M_REV_DIR_ADDR)
+		addr ^= 1;
+
+	return addr;
+}
+
 static int i2c_imx_dma_write(struct imx_i2c_struct *i2c_imx,
 					struct i2c_msg *msgs)
 {
@@ -1200,7 +1216,7 @@ static int i2c_imx_dma_write(struct imx_i2c_struct *i2c_imx,
 	 * Write slave address.
 	 * The first byte must be transmitted by the CPU.
 	 */
-	imx_i2c_write_reg(i2c_8bit_addr_from_msg(msgs), i2c_imx, IMX_I2C_I2DR);
+	imx_i2c_write_reg(i2c_imx_addr_byte(msgs), i2c_imx, IMX_I2C_I2DR);
 	time_left = wait_for_completion_timeout(
 				&i2c_imx->dma->cmd_complete,
 				msecs_to_jiffies(DMA_TIMEOUT));
@@ -1242,14 +1258,20 @@ static int i2c_imx_prepare_read(struct imx_i2c_struct *i2c_imx,
 	int result;
 	unsigned int temp = 0;
 
-	/* write slave address */
-	imx_i2c_write_reg(i2c_8bit_addr_from_msg(msgs), i2c_imx, IMX_I2C_I2DR);
-	result = i2c_imx_trx_complete(i2c_imx, !use_dma);
-	if (result)
-		return result;
-	result = i2c_imx_acked(i2c_imx);
-	if (result)
-		return result;
+	/*
+	 * I2C_M_NOSTART continues a frame that is already open, so there is
+	 * no address phase: go straight to turning the bus around.
+	 */
+	if (!(msgs->flags & I2C_M_NOSTART)) {
+		/* write slave address */
+		imx_i2c_write_reg(i2c_imx_addr_byte(msgs), i2c_imx, IMX_I2C_I2DR);
+		result = i2c_imx_trx_complete(i2c_imx, !use_dma);
+		if (result)
+			return result;
+		result = i2c_imx_acked(i2c_imx);
+		if (result)
+			return result;
+	}
 
 	dev_dbg(&i2c_imx->adapter.dev, "<%s> setup bus\n", __func__);
 
@@ -1371,16 +1393,18 @@ static int i2c_imx_atomic_write(struct imx_i2c_struct *i2c_imx,
 	int i, result;
 
 	dev_dbg(&i2c_imx->adapter.dev, "<%s> write slave address: addr=0x%x\n",
-		__func__, i2c_8bit_addr_from_msg(msgs));
+		__func__, i2c_imx_addr_byte(msgs));
 
-	/* write slave address */
-	imx_i2c_write_reg(i2c_8bit_addr_from_msg(msgs), i2c_imx, IMX_I2C_I2DR);
-	result = i2c_imx_trx_complete(i2c_imx, true);
-	if (result)
-		return result;
-	result = i2c_imx_acked(i2c_imx);
-	if (result)
-		return result;
+	if (!(msgs->flags & I2C_M_NOSTART)) {
+		/* write slave address */
+		imx_i2c_write_reg(i2c_imx_addr_byte(msgs), i2c_imx, IMX_I2C_I2DR);
+		result = i2c_imx_trx_complete(i2c_imx, true);
+		if (result)
+			return result;
+		result = i2c_imx_acked(i2c_imx);
+		if (result)
+			return result;
+	}
 	dev_dbg(&i2c_imx->adapter.dev, "<%s> write data\n", __func__);
 
 	/* write data */
@@ -1402,7 +1426,7 @@ static int i2c_imx_atomic_write(struct imx_i2c_struct *i2c_imx,
 static int i2c_imx_write(struct imx_i2c_struct *i2c_imx, struct i2c_msg *msgs)
 {
 	dev_dbg(&i2c_imx->adapter.dev, "<%s> write slave address: addr=0x%x\n",
-		__func__, i2c_8bit_addr_from_msg(msgs));
+		__func__, i2c_imx_addr_byte(msgs));
 
 	i2c_imx->state = IMX_I2C_STATE_WRITE;
 	i2c_imx->msg = msgs;
@@ -1411,8 +1435,16 @@ static int i2c_imx_write(struct imx_i2c_struct *i2c_imx, struct i2c_msg *msgs)
 	/*
 	 * By writing the device address we start the state machine in the ISR.
 	 * The ISR will report when it is done or when it fails.
+	 *
+	 * I2C_M_NOSTART continues a frame that is already open and so has no
+	 * address byte: push the first data byte instead. That raises the same
+	 * interrupt and the ISR carries on from the second byte.
 	 */
-	imx_i2c_write_reg(i2c_8bit_addr_from_msg(msgs), i2c_imx, IMX_I2C_I2DR);
+	if (msgs->flags & I2C_M_NOSTART)
+		imx_i2c_write_reg(msgs->buf[i2c_imx->msg_buf_idx++], i2c_imx,
+				  IMX_I2C_I2DR);
+	else
+		imx_i2c_write_reg(i2c_imx_addr_byte(msgs), i2c_imx, IMX_I2C_I2DR);
 	wait_event_timeout(i2c_imx->queue,
 			   i2c_imx->state == IMX_I2C_STATE_DONE ||
 			   i2c_imx->state == IMX_I2C_STATE_FAILED,
@@ -1529,22 +1561,29 @@ static int i2c_imx_read(struct imx_i2c_struct *i2c_imx, struct i2c_msg *msgs,
 
 	dev_dbg(&i2c_imx->adapter.dev,
 		"<%s> write slave address: addr=0x%x\n",
-		__func__, i2c_8bit_addr_from_msg(msgs));
+		__func__, i2c_imx_addr_byte(msgs));
 
 	i2c_imx->is_lastmsg = is_lastmsg;
 
-	if (block_data)
-		i2c_imx->state = IMX_I2C_STATE_READ_BLOCK_DATA;
-	else
-		i2c_imx->state = IMX_I2C_STATE_READ;
 	i2c_imx->msg = msgs;
 	i2c_imx->msg_buf_idx = 0;
 
-	/*
-	 * By writing the device address we start the state machine in the ISR.
-	 * The ISR will report when it is done or when it fails.
-	 */
-	imx_i2c_write_reg(i2c_8bit_addr_from_msg(msgs), i2c_imx, IMX_I2C_I2DR);
+	if (msgs->flags & I2C_M_NOSTART) {
+		i2c_imx->state = block_data ? IMX_I2C_STATE_READ_BLOCK_DATA_LEN
+					    : IMX_I2C_STATE_READ_CONTINUE;
+		i2c_imx_setup_read(i2c_imx);
+	} else {
+		if (block_data)
+			i2c_imx->state = IMX_I2C_STATE_READ_BLOCK_DATA;
+		else
+			i2c_imx->state = IMX_I2C_STATE_READ;
+
+		/*
+		 * By writing the device address we start the state machine in the ISR.
+		 * The ISR will report when it is done or when it fails.
+		 */
+		imx_i2c_write_reg(i2c_imx_addr_byte(msgs), i2c_imx, IMX_I2C_I2DR);
+	}
 	wait_event_timeout(i2c_imx->queue,
 			   i2c_imx->state == IMX_I2C_STATE_DONE ||
 			   i2c_imx->state == IMX_I2C_STATE_FAILED,
@@ -1583,6 +1622,27 @@ static int i2c_imx_xfer_common(struct i2c_adapter *adapter,
 	struct imx_i2c_struct *i2c_imx = i2c_get_adapdata(adapter);
 	int use_dma = 0;
 
+	for (i = 0; i < num; i++) {
+		/*
+		 * I2C_FUNC_PROTOCOL_MANGLING is advertised for
+		 * I2C_M_REV_DIR_ADDR only; reject the flags it also covers but
+		 * that this driver does not implement, rather than silently
+		 * transferring something the caller did not ask for.
+		 */
+		if (msgs[i].flags & (I2C_M_IGNORE_NAK | I2C_M_NO_RD_ACK |
+				     I2C_M_STOP))
+			return -EOPNOTSUPP;
+
+		if (!(msgs[i].flags & I2C_M_NOSTART))
+			continue;
+
+		/* Nothing to continue from, and a zero-length continuation
+		 * would leave the controller with no byte to clock.
+		 */
+		if (i == 0 || msgs[i].len == 0)
+			return -EOPNOTSUPP;
+	}
+
 	/* Start I2C transfer */
 	result = i2c_imx_start(i2c_imx, atomic);
 	if (result) {
@@ -1604,7 +1664,7 @@ static int i2c_imx_xfer_common(struct i2c_adapter *adapter,
 		if (i == num - 1)
 			is_lastmsg = true;
 
-		if (i) {
+		if (i && !(msgs[i].flags & I2C_M_NOSTART)) {
 			dev_dbg(&i2c_imx->adapter.dev,
 				"<%s> repeated start\n", __func__);
 			temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2CR);
@@ -1636,7 +1696,8 @@ static int i2c_imx_xfer_common(struct i2c_adapter *adapter,
 #endif
 
 		use_dma = i2c_imx->dma && msgs[i].len >= DMA_THRESHOLD &&
-			msgs[i].flags & I2C_M_DMA_SAFE;
+			msgs[i].flags & I2C_M_DMA_SAFE &&
+			!(msgs[i].flags & I2C_M_NOSTART);
 		if (msgs[i].flags & I2C_M_RD) {
 			int block_data = msgs->flags & I2C_M_RECV_LEN;
 
@@ -1730,7 +1791,9 @@ static int i2c_imx_init_recovery_info(struct imx_i2c_struct *i2c_imx,
 static u32 i2c_imx_func(struct i2c_adapter *adapter)
 {
 	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL
-		| I2C_FUNC_SMBUS_READ_BLOCK_DATA;
+		| I2C_FUNC_SMBUS_READ_BLOCK_DATA
+		| I2C_FUNC_NOSTART
+		| I2C_FUNC_PROTOCOL_MANGLING;
 }
 
 static const struct i2c_algorithm i2c_imx_algo = {

---
base-commit: 6c68fa601b49683ecb04eded993a71dfa8b2ba0b
change-id: 20260807-for-upstream-i2c-imx-lx2160-reverse-7a94b9561188

Best regards,
-- 
Vincent Jardin <vjardin@free.fr>



^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH] i2c: imx: add I2C_M_REV_DIR_ADDR and I2C_M_NOSTART
  2026-08-06 22:37 [PATCH] i2c: imx: add I2C_M_REV_DIR_ADDR and I2C_M_NOSTART Vincent Jardin via B4 Relay
@ 2026-08-06 22:48 ` sashiko-bot
  2026-08-07  6:22   ` Vincent Jardin
  0 siblings, 1 reply; 3+ messages in thread
From: sashiko-bot @ 2026-08-06 22:48 UTC (permalink / raw)
  To: Vincent Jardin; +Cc: imx, Frank.Li

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] Zero-length I2C reads cause a NULL pointer dereference and kernel panic in the interrupt handler due to integer underflow.
--

--- Patch [1]: [PATCH] i2c: imx: add I2C_M_REV_DIR_ADDR and I2C_M_NOSTART ---
commit 600d2bffc8575545709ea03b9e6d37e2e2b2e812
Author: Vincent Jardin <vjardin@free.fr>

i2c: imx: add I2C_M_REV_DIR_ADDR and I2C_M_NOSTART
    
This commit adds support for I2C_M_REV_DIR_ADDR and I2C_M_NOSTART flags in
the i2c-imx driver. It modifies the read/write paths and ISR to handle
messages that continue frames without re-addressing or invert the R/W bit.

> diff --git a/drivers/i2c/busses/i2c-imx.c b/drivers/i2c/busses/i2c-imx.c
> --- a/drivers/i2c/busses/i2c-imx.c
> +++ b/drivers/i2c/busses/i2c-imx.c
[ ... ]
> -static inline int i2c_imx_isr_read(struct imx_i2c_struct *i2c_imx)
> +static inline void i2c_imx_setup_read(struct imx_i2c_struct *i2c_imx)
>  {
> -	int result;
>  	unsigned int temp;
>  
> -	result = i2c_imx_isr_acked(i2c_imx);
> -	if (result)
> -		return result;
> -
> -	/* setup bus to read data */
>  	temp = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2CR);
>  	temp &= ~I2CR_MTX;
>  	if ((i2c_imx->msg->len - 1) || (i2c_imx->msg->flags & I2C_M_RECV_LEN))
> @@ -1017,6 +1011,18 @@ static inline int i2c_imx_isr_read(struct imx_i2c_struct *i2c_imx)
>  
>  	imx_i2c_write_reg(temp, i2c_imx, IMX_I2C_I2CR);
>  	imx_i2c_read_reg(i2c_imx, IMX_I2C_I2DR); /* dummy read */
> +}

[Severity: High]
This is a pre-existing issue, but can a zero-length I2C read cause an integer
underflow here?

When an application uses the I2C_RDWR ioctl with a read message of length 0,
would (i2c_imx->msg->len - 1) underflow to a non-zero value and bypass the
intended logic to NACK the read?

If this underflows, the driver appears to clear I2CR_TXAK and issues a dummy
read to start data reception. Later, in i2c_imx_isr_read_continue(), the
underflowed length values bypass the loop termination conditions:

    if ((i2c_imx->msg->len - 1) == i2c_imx->msg_buf_idx) {
        ...
    } else if (i2c_imx->msg_buf_idx == (i2c_imx->msg->len - 2)) {
        ...
    }

    i2c_imx->msg->buf[i2c_imx->msg_buf_idx++] = imx_i2c_read_reg(i2c_imx, IMX_I2C_I2DR);

Since the I2C core sets msg->buf to NULL for zero-length messages, does this
result in the driver writing to a NULL pointer and causing a kernel panic?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260807-for-upstream-i2c-imx-lx2160-reverse-v1-1-f276ce8cb660@free.fr?part=1

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] i2c: imx: add I2C_M_REV_DIR_ADDR and I2C_M_NOSTART
  2026-08-06 22:48 ` sashiko-bot
@ 2026-08-07  6:22   ` Vincent Jardin
  0 siblings, 0 replies; 3+ messages in thread
From: Vincent Jardin @ 2026-08-07  6:22 UTC (permalink / raw)
  To: sashiko-reviews; +Cc: imx, Frank.Li

Hi Mr. Bot,

> Pre-existing issues:
> - [High] Zero-length I2C reads cause a NULL pointer dereference and kernel panic in the interrupt handler due to integer underflow.

This analysis seems to be correct but it is not the purpose of this feature
update which does only "move" this block of code for the needed new feature.

The pre-existing issue/topic may needs its own specific "Fix" commit.

> Since the I2C core sets msg->buf to NULL for zero-length messages, does this
> result in the driver writing to a NULL pointer and causing a kernel panic?

I think your analysis it not correct.

thanks for the reviews,
  Vincent

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2026-08-07  6:22 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-08-06 22:37 [PATCH] i2c: imx: add I2C_M_REV_DIR_ADDR and I2C_M_NOSTART Vincent Jardin via B4 Relay
2026-08-06 22:48 ` sashiko-bot
2026-08-07  6:22   ` Vincent Jardin

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox