[PATCH v2 0/3] i3c: replace ENOTSUPP with SUSV4-compliant EOPNOTSUPP

[PATCH v2 0/3] i3c: replace ENOTSUPP with SUSV4-compliant EOPNOTSUPP

[Wolfram Sang]

Changes since v1:

* updated commit message

This checkpatch warning makes sense here: "ENOTSUPP is not a SUSV4 error
code, prefer EOPNOTSUPP" We don't have a userspace interface yet, but we
probably will get one. So, let's convert these error codes now, before
it will be forgotten in the future, and they could slip through to
applications.

This is based on 6.16-rc3 plus "[PATCH] i3c: don't fail if GETHDRCAP is
unsupported" which I sent earlier. During its development, the issue
became apparent. Automatically generated and build tested only.

Looking forward to comments.


Wolfram Sang (3):
  i3c: master: replace ENOTSUPP with SUSV4-compliant EOPNOTSUPP
  i3c: dw: replace ENOTSUPP with SUSV4-compliant EOPNOTSUPP
  i3c: master: cdns: replace ENOTSUPP with SUSV4-compliant EOPNOTSUPP

 drivers/i3c/master.c                 | 20 ++++++++++----------
 drivers/i3c/master/dw-i3c-master.c   |  8 ++++----
 drivers/i3c/master/i3c-master-cdns.c | 12 ++++++------
 3 files changed, 20 insertions(+), 20 deletions(-)

-- 
2.47.2


-- 
linux-i3c mailing list
linux-i3c@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-i3c

[PATCH v2 1/3] i3c: master: replace ENOTSUPP with SUSV4-compliant EOPNOTSUPP

[Wolfram Sang]

The checkpatch warning '"ENOTSUPP is not a SUSV4 error code, prefer
EOPNOTSUPP"' is correct. We never want this to slip to userspace.

Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
---
 drivers/i3c/master.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/drivers/i3c/master.c b/drivers/i3c/master.c
index e53c69d24873..354fef4a033c 100644
--- a/drivers/i3c/master.c
+++ b/drivers/i3c/master.c
@@ -837,14 +837,14 @@ static int i3c_master_send_ccc_cmd_locked(struct i3c_master_controller *master,
 		return -EINVAL;
 
 	if (!master->ops->send_ccc_cmd)
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	if ((cmd->id & I3C_CCC_DIRECT) && (!cmd->dests || !cmd->ndests))
 		return -EINVAL;
 
 	if (master->ops->supports_ccc_cmd &&
 	    !master->ops->supports_ccc_cmd(master, cmd))
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	ret = master->ops->send_ccc_cmd(master, cmd);
 	if (ret) {
@@ -1439,7 +1439,7 @@ static int i3c_master_retrieve_dev_info(struct i3c_dev_desc *dev)
 
 	if (dev->info.bcr & I3C_BCR_HDR_CAP) {
 		ret = i3c_master_gethdrcap_locked(master, &dev->info);
-		if (ret && ret != -ENOTSUPP)
+		if (ret && ret != -EOPNOTSUPP)
 			return ret;
 	}
 
@@ -2210,7 +2210,7 @@ of_i3c_master_add_i2c_boardinfo(struct i3c_master_controller *master,
 	 */
 	if (boardinfo->base.flags & I2C_CLIENT_TEN) {
 		dev_err(dev, "I2C device with 10 bit address not supported.");
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 	}
 
 	/* LVR is encoded in reg[2]. */
@@ -2340,13 +2340,13 @@ static int i3c_master_i2c_adapter_xfer(struct i2c_adapter *adap,
 		return -EINVAL;
 
 	if (!master->ops->i2c_xfers)
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	/* Doing transfers to different devices is not supported. */
 	addr = xfers[0].addr;
 	for (i = 1; i < nxfers; i++) {
 		if (addr != xfers[i].addr)
-			return -ENOTSUPP;
+			return -EOPNOTSUPP;
 	}
 
 	i3c_bus_normaluse_lock(&master->bus);
@@ -2766,7 +2766,7 @@ static int i3c_master_check_ops(const struct i3c_master_controller_ops *ops)
  *	    controller)
  * @ops: the master controller operations
  * @secondary: true if you are registering a secondary master. Will return
- *	       -ENOTSUPP if set to true since secondary masters are not yet
+ *	       -EOPNOTSUPP if set to true since secondary masters are not yet
  *	       supported
  *
  * This function takes care of everything for you:
@@ -2793,7 +2793,7 @@ int i3c_master_register(struct i3c_master_controller *master,
 
 	/* We do not support secondary masters yet. */
 	if (secondary)
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	ret = i3c_master_check_ops(ops);
 	if (ret)
@@ -2954,7 +2954,7 @@ int i3c_dev_do_priv_xfers_locked(struct i3c_dev_desc *dev,
 		return -EINVAL;
 
 	if (!master->ops->priv_xfers)
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	return master->ops->priv_xfers(dev, xfers, nxfers);
 }
@@ -3004,7 +3004,7 @@ int i3c_dev_request_ibi_locked(struct i3c_dev_desc *dev,
 	int ret;
 
 	if (!master->ops->request_ibi)
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	if (dev->ibi)
 		return -EBUSY;
-- 
2.47.2


-- 
linux-i3c mailing list
linux-i3c@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-i3c

[PATCH v2 2/3] i3c: dw: replace ENOTSUPP with SUSV4-compliant EOPNOTSUPP

[Wolfram Sang]

The checkpatch warning '"ENOTSUPP is not a SUSV4 error code, prefer
EOPNOTSUPP"' is correct. We never want this to slip to userspace.

Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
---
 drivers/i3c/master/dw-i3c-master.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/i3c/master/dw-i3c-master.c b/drivers/i3c/master/dw-i3c-master.c
index 611c22b72c15..91429d94a866 100644
--- a/drivers/i3c/master/dw-i3c-master.c
+++ b/drivers/i3c/master/dw-i3c-master.c
@@ -932,7 +932,7 @@ static int dw_i3c_master_priv_xfers(struct i3c_dev_desc *dev,
 		return 0;
 
 	if (i3c_nxfers > master->caps.cmdfifodepth)
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	for (i = 0; i < i3c_nxfers; i++) {
 		if (i3c_xfers[i].rnw)
@@ -943,7 +943,7 @@ static int dw_i3c_master_priv_xfers(struct i3c_dev_desc *dev,
 
 	if (ntxwords > master->caps.datafifodepth ||
 	    nrxwords > master->caps.datafifodepth)
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	xfer = dw_i3c_master_alloc_xfer(master, i3c_nxfers);
 	if (!xfer)
@@ -1093,7 +1093,7 @@ static int dw_i3c_master_i2c_xfers(struct i2c_dev_desc *dev,
 		return 0;
 
 	if (i2c_nxfers > master->caps.cmdfifodepth)
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	for (i = 0; i < i2c_nxfers; i++) {
 		if (i2c_xfers[i].flags & I2C_M_RD)
@@ -1104,7 +1104,7 @@ static int dw_i3c_master_i2c_xfers(struct i2c_dev_desc *dev,
 
 	if (ntxwords > master->caps.datafifodepth ||
 	    nrxwords > master->caps.datafifodepth)
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	xfer = dw_i3c_master_alloc_xfer(master, i2c_nxfers);
 	if (!xfer)
-- 
2.47.2


-- 
linux-i3c mailing list
linux-i3c@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-i3c

[PATCH v2 3/3] i3c: master: cdns: replace ENOTSUPP with SUSV4-compliant EOPNOTSUPP

[Wolfram Sang]

The checkpatch warning '"ENOTSUPP is not a SUSV4 error code, prefer
EOPNOTSUPP"' is correct. We never want this to slip to userspace.

Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
---
 drivers/i3c/master/i3c-master-cdns.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/i3c/master/i3c-master-cdns.c b/drivers/i3c/master/i3c-master-cdns.c
index fd3752cea654..e24d93a82e1c 100644
--- a/drivers/i3c/master/i3c-master-cdns.c
+++ b/drivers/i3c/master/i3c-master-cdns.c
@@ -742,7 +742,7 @@ static int cdns_i3c_master_priv_xfers(struct i3c_dev_desc *dev,
 
 	for (i = 0; i < nxfers; i++) {
 		if (xfers[i].len > CMD0_FIFO_PL_LEN_MAX)
-			return -ENOTSUPP;
+			return -EOPNOTSUPP;
 	}
 
 	if (!nxfers)
@@ -750,7 +750,7 @@ static int cdns_i3c_master_priv_xfers(struct i3c_dev_desc *dev,
 
 	if (nxfers > master->caps.cmdfifodepth ||
 	    nxfers > master->caps.cmdrfifodepth)
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	/*
 	 * First make sure that all transactions (block of transfers separated
@@ -765,7 +765,7 @@ static int cdns_i3c_master_priv_xfers(struct i3c_dev_desc *dev,
 
 	if (rxslots > master->caps.rxfifodepth ||
 	    txslots > master->caps.txfifodepth)
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	cdns_xfer = cdns_i3c_master_alloc_xfer(master, nxfers);
 	if (!cdns_xfer)
@@ -822,11 +822,11 @@ static int cdns_i3c_master_i2c_xfers(struct i2c_dev_desc *dev,
 	int i, ret = 0;
 
 	if (nxfers > master->caps.cmdfifodepth)
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	for (i = 0; i < nxfers; i++) {
 		if (xfers[i].len > CMD0_FIFO_PL_LEN_MAX)
-			return -ENOTSUPP;
+			return -EOPNOTSUPP;
 
 		if (xfers[i].flags & I2C_M_RD)
 			nrxwords += DIV_ROUND_UP(xfers[i].len, 4);
@@ -836,7 +836,7 @@ static int cdns_i3c_master_i2c_xfers(struct i2c_dev_desc *dev,
 
 	if (ntxwords > master->caps.txfifodepth ||
 	    nrxwords > master->caps.rxfifodepth)
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	xfer = cdns_i3c_master_alloc_xfer(master, nxfers);
 	if (!xfer)
-- 
2.47.2


-- 
linux-i3c mailing list
linux-i3c@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-i3c

Re: [PATCH v2 1/3] i3c: master: replace ENOTSUPP with SUSV4-compliant EOPNOTSUPP

[Frank Li]

On Fri, Jun 27, 2025 at 01:17:56PM +0200, Wolfram Sang wrote:
> The checkpatch warning '"ENOTSUPP is not a SUSV4 error code, prefer
> EOPNOTSUPP"' is correct. We never want this to slip to userspace.

what's means "We never want this to slip to userspace."?

especial what's 'this' here?

Frank
>
> Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
> ---
>  drivers/i3c/master.c | 20 ++++++++++----------
>  1 file changed, 10 insertions(+), 10 deletions(-)
>
> diff --git a/drivers/i3c/master.c b/drivers/i3c/master.c
> index e53c69d24873..354fef4a033c 100644
> --- a/drivers/i3c/master.c
> +++ b/drivers/i3c/master.c
> @@ -837,14 +837,14 @@ static int i3c_master_send_ccc_cmd_locked(struct i3c_master_controller *master,
>  		return -EINVAL;
>
>  	if (!master->ops->send_ccc_cmd)
> -		return -ENOTSUPP;
> +		return -EOPNOTSUPP;
>
>  	if ((cmd->id & I3C_CCC_DIRECT) && (!cmd->dests || !cmd->ndests))
>  		return -EINVAL;
>
>  	if (master->ops->supports_ccc_cmd &&
>  	    !master->ops->supports_ccc_cmd(master, cmd))
> -		return -ENOTSUPP;
> +		return -EOPNOTSUPP;
>
>  	ret = master->ops->send_ccc_cmd(master, cmd);
>  	if (ret) {
> @@ -1439,7 +1439,7 @@ static int i3c_master_retrieve_dev_info(struct i3c_dev_desc *dev)
>
>  	if (dev->info.bcr & I3C_BCR_HDR_CAP) {
>  		ret = i3c_master_gethdrcap_locked(master, &dev->info);
> -		if (ret && ret != -ENOTSUPP)
> +		if (ret && ret != -EOPNOTSUPP)
>  			return ret;
>  	}
>
> @@ -2210,7 +2210,7 @@ of_i3c_master_add_i2c_boardinfo(struct i3c_master_controller *master,
>  	 */
>  	if (boardinfo->base.flags & I2C_CLIENT_TEN) {
>  		dev_err(dev, "I2C device with 10 bit address not supported.");
> -		return -ENOTSUPP;
> +		return -EOPNOTSUPP;
>  	}
>
>  	/* LVR is encoded in reg[2]. */
> @@ -2340,13 +2340,13 @@ static int i3c_master_i2c_adapter_xfer(struct i2c_adapter *adap,
>  		return -EINVAL;
>
>  	if (!master->ops->i2c_xfers)
> -		return -ENOTSUPP;
> +		return -EOPNOTSUPP;
>
>  	/* Doing transfers to different devices is not supported. */
>  	addr = xfers[0].addr;
>  	for (i = 1; i < nxfers; i++) {
>  		if (addr != xfers[i].addr)
> -			return -ENOTSUPP;
> +			return -EOPNOTSUPP;
>  	}
>
>  	i3c_bus_normaluse_lock(&master->bus);
> @@ -2766,7 +2766,7 @@ static int i3c_master_check_ops(const struct i3c_master_controller_ops *ops)
>   *	    controller)
>   * @ops: the master controller operations
>   * @secondary: true if you are registering a secondary master. Will return
> - *	       -ENOTSUPP if set to true since secondary masters are not yet
> + *	       -EOPNOTSUPP if set to true since secondary masters are not yet
>   *	       supported
>   *
>   * This function takes care of everything for you:
> @@ -2793,7 +2793,7 @@ int i3c_master_register(struct i3c_master_controller *master,
>
>  	/* We do not support secondary masters yet. */
>  	if (secondary)
> -		return -ENOTSUPP;
> +		return -EOPNOTSUPP;
>
>  	ret = i3c_master_check_ops(ops);
>  	if (ret)
> @@ -2954,7 +2954,7 @@ int i3c_dev_do_priv_xfers_locked(struct i3c_dev_desc *dev,
>  		return -EINVAL;
>
>  	if (!master->ops->priv_xfers)
> -		return -ENOTSUPP;
> +		return -EOPNOTSUPP;
>
>  	return master->ops->priv_xfers(dev, xfers, nxfers);
>  }
> @@ -3004,7 +3004,7 @@ int i3c_dev_request_ibi_locked(struct i3c_dev_desc *dev,
>  	int ret;
>
>  	if (!master->ops->request_ibi)
> -		return -ENOTSUPP;
> +		return -EOPNOTSUPP;
>
>  	if (dev->ibi)
>  		return -EBUSY;
> --
> 2.47.2
>

-- 
linux-i3c mailing list
linux-i3c@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-i3c

Re: [PATCH v2 1/3] i3c: master: replace ENOTSUPP with SUSV4-compliant EOPNOTSUPP

[Wolfram Sang]

On Fri, Jun 27, 2025 at 04:46:37PM -0400, Frank Li wrote:
> On Fri, Jun 27, 2025 at 01:17:56PM +0200, Wolfram Sang wrote:
> > The checkpatch warning '"ENOTSUPP is not a SUSV4 error code, prefer
> > EOPNOTSUPP"' is correct. We never want this to slip to userspace.
> 
> what's means "We never want this to slip to userspace."?
> 
> especial what's 'this' here?

ENOTSUPP. If we were 100% sure to only use it it kernel space, we could
keep using it. But since this is fragile because we never know about how
error propagation will evolve, we chose to not use it.


-- 
linux-i3c mailing list
linux-i3c@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-i3c

Re: [PATCH v2 1/3] i3c: master: replace ENOTSUPP with SUSV4-compliant EOPNOTSUPP

[Frank Li]

On Sat, Jun 28, 2025 at 12:12:35PM +0200, Wolfram Sang wrote:
> On Fri, Jun 27, 2025 at 04:46:37PM -0400, Frank Li wrote:
> > On Fri, Jun 27, 2025 at 01:17:56PM +0200, Wolfram Sang wrote:
> > > The checkpatch warning '"ENOTSUPP is not a SUSV4 error code, prefer
> > > EOPNOTSUPP"' is correct. We never want this to slip to userspace.
> >
> > what's means "We never want this to slip to userspace."?
> >
> > especial what's 'this' here?
>
> ENOTSUPP. If we were 100% sure to only use it it kernel space, we could
> keep using it. But since this is fragile because we never know about how
> error propagation will evolve, we chose to not use it.

I understand this. But the sentence "we never want this(ENOTSUPP) to slip to
userspace" literal means (I understand at least).

we don't expect to export it to userspace at all.

Actually your means "we don't know if it will skip to user space in future"

Frank


>
>
> --
> linux-i3c mailing list
> linux-i3c@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-i3c

-- 
linux-i3c mailing list
linux-i3c@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-i3c

Re: [PATCH v2 1/3] i3c: master: replace ENOTSUPP with SUSV4-compliant EOPNOTSUPP

[Wolfram Sang]

> I understand this. But the sentence "we never want this(ENOTSUPP) to slip to
> userspace" literal means (I understand at least).
> 
> we don't expect to export it to userspace at all.
> 
> Actually your means "we don't know if it will skip to user space in future"

Sorry, I simply don't understand what you mean here. But I think it is
not worth the time to discuss this further anyhow. I will just use your
suggested commit message. It is not that important.


-- 
linux-i3c mailing list
linux-i3c@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-i3c