Retrieving the supported versions of a command is a fairly common
operation. Provide a helper for it.

If the command is not supported at all the EC returns
-EINVAL/EC_RES_INVALID_PARAMS.

This error is translated into an empty version mask as that is easier to
handle for callers and they don't need to know about the error details.

Signed-off-by: Thomas Weißschuh <linux@weissschuh.net>
---
 drivers/platform/chrome/cros_ec_proto.c     | 26 ++++++++++++++++++++++++++
 include/linux/platform_data/cros_ec_proto.h |  2 ++
 2 files changed, 28 insertions(+)

diff --git a/drivers/platform/chrome/cros_ec_proto.c b/drivers/platform/chrome/cros_ec_proto.c
index fe68be66ee98..9cfe885a5301 100644
--- a/drivers/platform/chrome/cros_ec_proto.c
+++ b/drivers/platform/chrome/cros_ec_proto.c
@@ -1069,3 +1069,29 @@ int cros_ec_cmd_readmem(struct cros_ec_device *ec_dev, u8 offset, u8 size, void
 			   &params, sizeof(params), dest, size);
 }
 EXPORT_SYMBOL_GPL(cros_ec_cmd_readmem);
+
+/**
+ * cros_ec_cmd_versions - Get supported version mask.
+ *
+ * @ec_dev: EC device
+ * @cmd: Command to test
+ *
+ * Return: version mask on success, negative error number on failure.
+ */
+int cros_ec_cmd_versions(struct cros_ec_device *ec_dev, u16 cmd)
+{
+	struct ec_params_get_cmd_versions_v1 req = {};
+	struct ec_response_get_cmd_versions resp;
+	int ret;
+
+	req.cmd = cmd;
+	ret = cros_ec_cmd(ec_dev, 1, EC_CMD_GET_CMD_VERSIONS,
+			  &req, sizeof(req), &resp, sizeof(resp));
+	if (ret == -EINVAL)
+		return 0; /* Command not implemented */
+	else if (ret < 0)
+		return ret;
+	else
+		return resp.version_mask;
+}
+EXPORT_SYMBOL_GPL(cros_ec_cmd_versions);
diff --git a/include/linux/platform_data/cros_ec_proto.h b/include/linux/platform_data/cros_ec_proto.h
index 6e9225bdf903..98ab5986f543 100644
--- a/include/linux/platform_data/cros_ec_proto.h
+++ b/include/linux/platform_data/cros_ec_proto.h
@@ -263,6 +263,8 @@ int cros_ec_cmd(struct cros_ec_device *ec_dev, unsigned int version, int command
 
 int cros_ec_cmd_readmem(struct cros_ec_device *ec_dev, u8 offset, u8 size, void *dest);
 
+int cros_ec_cmd_versions(struct cros_ec_device *ec_dev, u16 cmd);
+
 /**
  * cros_ec_get_time_ns() - Return time in ns.
  *

-- 
2.45.2

On Mon, Jun 10, 2024 at 05:51:08PM +0200, Thomas Weißschuh wrote:
> +/**
> + * cros_ec_cmd_versions - Get supported version mask.

I guess we would like to call it something like "cros_ec_get_cmd_versions".

> + *
> + * @ec_dev: EC device
> + * @cmd: Command to test
> + *
> + * Return: version mask on success, negative error number on failure.
> + */
> +int cros_ec_cmd_versions(struct cros_ec_device *ec_dev, u16 cmd)

Could it support a "version" parameter as existing EC_CMD_GET_CMD_VERSIONS
usages use both versions?  An `u16 cmd` parameter and down-cast to u8 for v0
should be fine. (ec_params_get_cmd_versions vs. ec_params_get_cmd_versions_v1)

On 2024-06-11 11:35:39+0000, Tzung-Bi Shih wrote:
> On Mon, Jun 10, 2024 at 05:51:08PM +0200, Thomas Weißschuh wrote:
> > +/**
> > + * cros_ec_cmd_versions - Get supported version mask.
> 
> I guess we would like to call it something like "cros_ec_get_cmd_versions".

Ack.

> > + *
> > + * @ec_dev: EC device
> > + * @cmd: Command to test
> > + *
> > + * Return: version mask on success, negative error number on failure.
> > + */
> > +int cros_ec_cmd_versions(struct cros_ec_device *ec_dev, u16 cmd)
> 
> Could it support a "version" parameter as existing EC_CMD_GET_CMD_VERSIONS
> usages use both versions?  An `u16 cmd` parameter and down-cast to u8 for v0
> should be fine. (ec_params_get_cmd_versions vs. ec_params_get_cmd_versions_v1)

Ack.

Or we could automatically use v0 if cmd <= U8_MAX?

On Tue, Jun 11, 2024 at 02:14:33PM +0200, Thomas Weißschuh wrote:
> On 2024-06-11 11:35:39+0000, Tzung-Bi Shih wrote:
> > On Mon, Jun 10, 2024 at 05:51:08PM +0200, Thomas Weißschuh wrote:
> > > + *
> > > + * @ec_dev: EC device
> > > + * @cmd: Command to test
> > > + *
> > > + * Return: version mask on success, negative error number on failure.
> > > + */
> > > +int cros_ec_cmd_versions(struct cros_ec_device *ec_dev, u16 cmd)
> > 
> > Could it support a "version" parameter as existing EC_CMD_GET_CMD_VERSIONS
> > usages use both versions?  An `u16 cmd` parameter and down-cast to u8 for v0
> > should be fine. (ec_params_get_cmd_versions vs. ec_params_get_cmd_versions_v1)
> 
> Ack.
> 
> Or we could automatically use v0 if cmd <= U8_MAX?

Ack.  Looks feasible.

On Mon, Jun 10, 2024 at 05:51:08PM +0200, Thomas Weißschuh wrote:
> If the command is not supported at all the EC returns
> -EINVAL/EC_RES_INVALID_PARAMS.
> 
> This error is translated into an empty version mask as that is easier to
> handle for callers and they don't need to know about the error details.

I'm not sure whether the behavior is what we want or not as existing
EC_CMD_GET_CMD_VERSIONS usages don't have it.

On 2024-06-11 06:32:38+0000, Tzung-Bi Shih wrote:
> On Mon, Jun 10, 2024 at 05:51:08PM +0200, Thomas Weißschuh wrote:
> > If the command is not supported at all the EC returns
> > -EINVAL/EC_RES_INVALID_PARAMS.
> > 
> > This error is translated into an empty version mask as that is easier to
> > handle for callers and they don't need to know about the error details.
> 
> I'm not sure whether the behavior is what we want or not as existing
> EC_CMD_GET_CMD_VERSIONS usages don't have it.

At least the caller of cros_ec_get_host_command_version_mask() expects
it:

ret = cros_ec_get_host_command_version_mask(..., &ver_mask);
if (ret < 0 || ver_mask == 0)
...

ver_mask == 0 will never happen as in that case -EINVAL would have been
returned.

Others, like cros_ec_cec_get_write_cmd_version(), expect the current
semantic of ver_mask != 0 but log spurious errors in case of -EINVAL.
cros_pchg_cmd_ver_check(), works with both semantics, but currently also
logs a spurious error message.

To me the new semantic looks more obvious and much easier to handle.
For each command version a bit is set. no command versions -> no bits.

On Tue, Jun 11, 2024 at 09:23:24AM +0200, Thomas Weißschuh wrote:
> On 2024-06-11 06:32:38+0000, Tzung-Bi Shih wrote:
> > On Mon, Jun 10, 2024 at 05:51:08PM +0200, Thomas Weißschuh wrote:
> > > If the command is not supported at all the EC returns
> > > -EINVAL/EC_RES_INVALID_PARAMS.
> > > 
> > > This error is translated into an empty version mask as that is easier to
> > > handle for callers and they don't need to know about the error details.
> > 
> > I'm not sure whether the behavior is what we want or not as existing
> > EC_CMD_GET_CMD_VERSIONS usages don't have it.
> 
> At least the caller of cros_ec_get_host_command_version_mask() expects
> it:
> 
> ret = cros_ec_get_host_command_version_mask(..., &ver_mask);
> if (ret < 0 || ver_mask == 0)
> ...
> 
> ver_mask == 0 will never happen as in that case -EINVAL would have been
> returned.
> 
> Others, like cros_ec_cec_get_write_cmd_version(), expect the current
> semantic of ver_mask != 0 but log spurious errors in case of -EINVAL.
> cros_pchg_cmd_ver_check(), works with both semantics, but currently also
> logs a spurious error message.
> 
> To me the new semantic looks more obvious and much easier to handle.
> For each command version a bit is set. no command versions -> no bits.

Ack.