Skip to contentSkip to conversation

mei, make modules.alias UUID information easier to read

PATCH In review Prarit Bhargava 4 participants 9 messages Aug 14, 2015 – Sep 21, 2015

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 mei, make modules.alias UUID information easier to read

2nd try on this ...

P.

---8<---

scripts/mod/file2alias.c:add_uuid() munges a UUID into a single string which does not conform to the standard little endian UUID. This patch changes add_uuid() to use the UUID correctly so that future drivers which use UUID matches can simply use the %pUL format, and modifies the mei driver with those changes.

Cc Tomas Winkler <[email protected]>
Cc Greg Kroah-Hartman <[email protected]>
Cc Joe Perches <[email protected]>
Cc "David S. Miller" <[email protected]>
Cc Jiri Kosina <[email protected]>
Cc Sharon Dvir <[email protected]>
Cc "Suthikulpanit, Suravee" <[email protected]>
Cc Heikki Krogerus <[email protected]>
Cc James Hogan <[email protected]>
Cc Daniel Thompson <[email protected]>
Cc Michael Opdenacker <[email protected]>
Cc David Cohen <[email protected]>
Cc Felipe Balbi <[email protected]>
Cc Ralf Baechle <[email protected]>
---
 drivers/misc/mei/bus.c          |    6 ++----
 include/linux/mod_devicetable.h |    4 ----
 scripts/mod/file2alias.c        |    9 ++++++---
 3 files changed, 8 insertions(+), 11 deletions(-)
drivers/misc/mei/bus.c +2 -4
index 458aa5a..457e132 100644
@@ -141,8 +141,7 @@ static ssize_t modalias_show(struct device *dev, struct device_attribute *a,
 	const uuid_le *uuid = mei_me_cl_uuid(device->me_cl);
 	size_t len;
 
-	len = snprintf(buf, PAGE_SIZE, "mei:%s:" MEI_CL_UUID_FMT ":",
-		device->name, MEI_CL_UUID_ARGS(uuid->b));
+	len = snprintf(buf, PAGE_SIZE, "mei:%s:%pUl:", device->name, uuid);
 
 	return (len >= PAGE_SIZE) ? (PAGE_SIZE - 1) : len;
 }
@@ -167,8 +166,7 @@ static int mei_cl_uevent(struct device *dev, struct kobj_uevent_env *env)
 	if (add_uevent_var(env, "MEI_CL_NAME=%s", device->name))
 		return -ENOMEM;
 
-	if (add_uevent_var(env, "MODALIAS=mei:%s:" MEI_CL_UUID_FMT ":",
-		device->name, MEI_CL_UUID_ARGS(uuid->b)))
+	if (add_uevent_var(env, "MODALIAS=mei:%s:%pUl:", device->name, uuid))
 		return -ENOMEM;
 
 	return 0;
include/linux/mod_devicetable.h +0 -4
index 34f25b7..becda6f 100644
@@ -601,10 +601,6 @@ struct ipack_device_id {
 
 #define MEI_CL_MODULE_PREFIX "mei:"
 #define MEI_CL_NAME_SIZE 32
-#define MEI_CL_UUID_FMT "%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x"
-#define MEI_CL_UUID_ARGS(_u) \
-	_u[0], _u[1], _u[2], _u[3], _u[4], _u[5], _u[6], _u[7], \
-	_u[8], _u[9], _u[10], _u[11], _u[12], _u[13], _u[14], _u[15]
 
 /**
  * struct mei_cl_device_id - MEI client device identifier
scripts/mod/file2alias.c +6 -3
index 5f20882..8a46c60 100644
@@ -137,10 +137,13 @@ static inline void add_wildcard(char *str)
 static inline void add_uuid(char *str, uuid_le uuid)
 {
 	int len = strlen(str);
-	int i;
 
-	for (i = 0; i < 16; i++)
-		sprintf(str + len + (i << 1), "%02x", uuid.b[i]);
+	sprintf(str + len,
+		"%02x%02x%02x%02x-%02x%02x-%02x%02x-%02x%02x-%02x%02x%02x%02x%02x%02x",
+		uuid.b[3], uuid.b[2], uuid.b[1], uuid.b[0],
+		uuid.b[5], uuid.b[4], uuid.b[7], uuid.b[6],
+		uuid.b[8], uuid.b[9], uuid.b[10], uuid.b[11],
+		uuid.b[12], uuid.b[13], uuid.b[14], uuid.b[15]);
 }
 
 /**
Signature · 1 line
1.7.9.3
#1 lore raw

Heikki, Tomas?

P.

Quoted 95 lines · Prarit Bhargava

2nd try on this ...

P.

---8<---

scripts/mod/file2alias.c:add_uuid() munges a UUID into a single string which does not conform to the standard little endian UUID. This patch changes add_uuid() to use the UUID correctly so that future drivers which use UUID matches can simply use the %pUL format, and modifies the mei driver with those changes.

Cc Tomas Winkler <[email protected]>
Cc Greg Kroah-Hartman <[email protected]>
Cc Joe Perches <[email protected]>
Cc "David S. Miller" <[email protected]>
Cc Jiri Kosina <[email protected]>
Cc Sharon Dvir <[email protected]>
Cc "Suthikulpanit, Suravee" <[email protected]>
Cc Heikki Krogerus <[email protected]>
Cc James Hogan <[email protected]>
Cc Daniel Thompson <[email protected]>
Cc Michael Opdenacker <[email protected]>
Cc David Cohen <[email protected]>
Cc Felipe Balbi <[email protected]>
Cc Ralf Baechle <[email protected]>
---
 drivers/misc/mei/bus.c          |    6 ++----
 include/linux/mod_devicetable.h |    4 ----
 scripts/mod/file2alias.c        |    9 ++++++---
 3 files changed, 8 insertions(+), 11 deletions(-)
drivers/misc/mei/bus.c +2 -4
index 458aa5a..457e132 100644
@@ -141,8 +141,7 @@ static ssize_t modalias_show(struct device *dev, struct device_attribute *a,
 	const uuid_le *uuid = mei_me_cl_uuid(device->me_cl);
 	size_t len;
 
-	len = snprintf(buf, PAGE_SIZE, "mei:%s:" MEI_CL_UUID_FMT ":",
-		device->name, MEI_CL_UUID_ARGS(uuid->b));
+	len = snprintf(buf, PAGE_SIZE, "mei:%s:%pUl:", device->name, uuid);
 
 	return (len >= PAGE_SIZE) ? (PAGE_SIZE - 1) : len;
 }
@@ -167,8 +166,7 @@ static int mei_cl_uevent(struct device *dev, struct kobj_uevent_env *env)
 	if (add_uevent_var(env, "MEI_CL_NAME=%s", device->name))
 		return -ENOMEM;
 
-	if (add_uevent_var(env, "MODALIAS=mei:%s:" MEI_CL_UUID_FMT ":",
-		device->name, MEI_CL_UUID_ARGS(uuid->b)))
+	if (add_uevent_var(env, "MODALIAS=mei:%s:%pUl:", device->name, uuid))
 		return -ENOMEM;
 
 	return 0;
include/linux/mod_devicetable.h +0 -4
index 34f25b7..becda6f 100644
@@ -601,10 +601,6 @@ struct ipack_device_id {
 
 #define MEI_CL_MODULE_PREFIX "mei:"
 #define MEI_CL_NAME_SIZE 32
-#define MEI_CL_UUID_FMT "%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x"
-#define MEI_CL_UUID_ARGS(_u) \
-	_u[0], _u[1], _u[2], _u[3], _u[4], _u[5], _u[6], _u[7], \
-	_u[8], _u[9], _u[10], _u[11], _u[12], _u[13], _u[14], _u[15]
 
 /**
  * struct mei_cl_device_id - MEI client device identifier
scripts/mod/file2alias.c +6 -3
index 5f20882..8a46c60 100644
@@ -137,10 +137,13 @@ static inline void add_wildcard(char *str)
 static inline void add_uuid(char *str, uuid_le uuid)
 {
 	int len = strlen(str);
-	int i;
 
-	for (i = 0; i < 16; i++)
-		sprintf(str + len + (i << 1), "%02x", uuid.b[i]);
+	sprintf(str + len,
+		"%02x%02x%02x%02x-%02x%02x-%02x%02x-%02x%02x-%02x%02x%02x%02x%02x%02x",
+		uuid.b[3], uuid.b[2], uuid.b[1], uuid.b[0],
+		uuid.b[5], uuid.b[4], uuid.b[7], uuid.b[6],
+		uuid.b[8], uuid.b[9], uuid.b[10], uuid.b[11],
+		uuid.b[12], uuid.b[13], uuid.b[14], uuid.b[15]);
 }
 
 /**
#2 lore raw

Hi Prarit,

Quoted 1 line · Prarit Bhargava

Heikki, Tomas?

I'm afraid I don't know much about Intel's Management Engine Interface. Looks like the driver is from Samuel (CC'd) so I'm guessing he is the person you wanted comments from and not me.

cheers,

Quoted 95 lines · Prarit Bhargava
Quoted 94 lines

2nd try on this ...

P.

---8<---

scripts/mod/file2alias.c:add_uuid() munges a UUID into a single string which does not conform to the standard little endian UUID. This patch changes add_uuid() to use the UUID correctly so that future drivers which use UUID matches can simply use the %pUL format, and modifies the mei driver with those changes.

Cc Tomas Winkler <[email protected]>
Cc Greg Kroah-Hartman <[email protected]>
Cc Joe Perches <[email protected]>
Cc "David S. Miller" <[email protected]>
Cc Jiri Kosina <[email protected]>
Cc Sharon Dvir <[email protected]>
Cc "Suthikulpanit, Suravee" <[email protected]>
Cc Heikki Krogerus <[email protected]>
Cc James Hogan <[email protected]>
Cc Daniel Thompson <[email protected]>
Cc Michael Opdenacker <[email protected]>
Cc David Cohen <[email protected]>
Cc Felipe Balbi <[email protected]>
Cc Ralf Baechle <[email protected]>
---
 drivers/misc/mei/bus.c          |    6 ++----
 include/linux/mod_devicetable.h |    4 ----
 scripts/mod/file2alias.c        |    9 ++++++---
 3 files changed, 8 insertions(+), 11 deletions(-)
drivers/misc/mei/bus.c +2 -4
index 458aa5a..457e132 100644
@@ -141,8 +141,7 @@ static ssize_t modalias_show(struct device *dev, struct device_attribute *a,
 	const uuid_le *uuid = mei_me_cl_uuid(device->me_cl);
 	size_t len;
 
-	len = snprintf(buf, PAGE_SIZE, "mei:%s:" MEI_CL_UUID_FMT ":",
-		device->name, MEI_CL_UUID_ARGS(uuid->b));
+	len = snprintf(buf, PAGE_SIZE, "mei:%s:%pUl:", device->name, uuid);
 
 	return (len >= PAGE_SIZE) ? (PAGE_SIZE - 1) : len;
 }
@@ -167,8 +166,7 @@ static int mei_cl_uevent(struct device *dev, struct kobj_uevent_env *env)
 	if (add_uevent_var(env, "MEI_CL_NAME=%s", device->name))
 		return -ENOMEM;
 
-	if (add_uevent_var(env, "MODALIAS=mei:%s:" MEI_CL_UUID_FMT ":",
-		device->name, MEI_CL_UUID_ARGS(uuid->b)))
+	if (add_uevent_var(env, "MODALIAS=mei:%s:%pUl:", device->name, uuid))
 		return -ENOMEM;
 
 	return 0;
include/linux/mod_devicetable.h +0 -4
index 34f25b7..becda6f 100644
@@ -601,10 +601,6 @@ struct ipack_device_id {
 
 #define MEI_CL_MODULE_PREFIX "mei:"
 #define MEI_CL_NAME_SIZE 32
-#define MEI_CL_UUID_FMT "%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x"
-#define MEI_CL_UUID_ARGS(_u) \
-	_u[0], _u[1], _u[2], _u[3], _u[4], _u[5], _u[6], _u[7], \
-	_u[8], _u[9], _u[10], _u[11], _u[12], _u[13], _u[14], _u[15]
 
 /**
  * struct mei_cl_device_id - MEI client device identifier
scripts/mod/file2alias.c +6 -3
index 5f20882..8a46c60 100644
@@ -137,10 +137,13 @@ static inline void add_wildcard(char *str)
 static inline void add_uuid(char *str, uuid_le uuid)
 {
 	int len = strlen(str);
-	int i;
 
-	for (i = 0; i < 16; i++)
-		sprintf(str + len + (i << 1), "%02x", uuid.b[i]);
+	sprintf(str + len,
+		"%02x%02x%02x%02x-%02x%02x-%02x%02x-%02x%02x-%02x%02x%02x%02x%02x%02x",
+		uuid.b[3], uuid.b[2], uuid.b[1], uuid.b[0],
+		uuid.b[5], uuid.b[4], uuid.b[7], uuid.b[6],
+		uuid.b[8], uuid.b[9], uuid.b[10], uuid.b[11],
+		uuid.b[12], uuid.b[13], uuid.b[14], uuid.b[15]);
 }
 
 /**
Signature · 1 line
heikki
#3 lore raw
Quoted 10 lines

Hi Prarit,

Quoted 1 line · Prarit Bhargava

Heikki, Tomas?

I'm afraid I don't know much about Intel's Management Engine Interface. Looks like the driver is from Samuel (CC'd) so I'm guessing he is the person you wanted comments from and not me.

The patch was done against the master branch instead of char-misc-next so it doesn't apply. Anyhow I've rebased it already and I'm testing it. I will probably re-post it, with your permission with my other mei bus fixes.

BTW, I took the inspiration from vmbus in regards to uuid representation so it if the uuid sting hurts you, you may fix vmbus as well. I don't have any vmbus setup and I try not to fix things I cannot check.

Thanks
Tomas

#4 lore raw
Quoted 6 lines · Tomas
Quoted 6 lines

Hi Prarit,

Quoted 1 line · Prarit Bhargava

Heikki, Tomas?

Quoted 7 lines
Quoted 4 lines

I'm afraid I don't know much about Intel's Management Engine Interface. Looks like the driver is from Samuel (CC'd) so I'm guessing he is the person you wanted comments from and not me.

The patch was done against the master branch instead of char-misc-next so it doesn't apply. Anyhow I've rebased it already and I'm testing it. I will probably re-post it, with your permission with my other mei bus fixes.

Tomas, can you send me a link to your tree?

Quoted 3 lines

BTW, I took the inspiration from vmbus in regards to uuid representation so it if the uuid sting hurts you, you may fix vmbus as well. I don't have any vmbus setup and I try not to fix things I cannot check.

I'll find a system that I can test vmbus on.

P.

Quoted 3 lines

Thanks
Tomas

#5 lore raw
Quoted 13 lines · Tomas
Quoted 11 lines
Quoted 10 lines

Hi Prarit,

Quoted 1 line · Prarit Bhargava

Heikki, Tomas?

I'm afraid I don't know much about Intel's Management Engine Interface. Looks like the driver is from Samuel (CC'd) so I'm guessing he is the person you wanted comments from and not me.

Quoted 5 lines
Quoted 1 line

The patch was done against the master branch instead of char-misc-next so it

doesn't apply. Anyhow I've rebased it already and I'm testing it.

Quoted 1 line

I will probably re-post it, with your permission with my other mei bus fixes.

Tomas, can you send me a link to your tree?

Sorry, it is not public. But it is very close to what is on git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc.git char-misc-next branch I will send new patches soon ... after 4.3-rc1 is out.
Thanks
Tomas

#6 lore raw
In reply to Winkler, Tomas
Quoted 6 lines · Tomas
Quoted 6 lines

Hi Prarit,

Quoted 1 line · Prarit Bhargava

Heikki, Tomas?

Quoted 8 lines
Quoted 4 lines

I'm afraid I don't know much about Intel's Management Engine Interface. Looks like the driver is from Samuel (CC'd) so I'm guessing he is the person you wanted comments from and not me.

The patch was done against the master branch instead of char-misc-next so it doesn't apply. Anyhow I've rebased it already and I'm testing it. I will probably re-post it, with your permission with my other mei bus fixes.

Yep, only requires some minor adjustments in order to apply to char-misc-next. Please repost when you can. FYI: this is needed to bring drivers/staging/unisys out of the staging directory.

Quoted 2 lines

BTW, I took the inspiration from vmbus in regards to uuid representation so it if the uuid sting hurts you, you may fix vmbus as well. I don't have any vmbus setup and I try not to fix things I cannot check.

I didn't recall other users of add_uuid() but took a look anway.

do_vmbus_entry() doesn't call add_uuid() like do_mei_entry() does. However, the vmbus code should also just simply use add_uuid(). I can do that after this patch gets into the kernel, as it isn't absolutely necessary to fix like do_mei_entry(). It should be a trivial drivers/hv patch to fix all that up.

P.

#7 lore raw
In reply to Prarit Bhargava
Quoted 1 line · Prarit Bhargava

2nd try on this ...

What changed from the first patch? I need some "version information" to figure out what is going on.

Can you resend it with that information?

thanks,

greg k-h

#8 lore raw
Quoted 12 lines

-----Original Message-----
From: Greg Kroah-Hartman [mailto:[email protected]]
Sent: Monday, September 21, 2015 05:25
To: Prarit Bhargava
Cc: [email protected]; Winkler, Tomas; Joe Perches; David S. Miller; Jiri Kosina; Sharon Dvir; Suthikulpanit, Suravee; Heikki Krogerus; James Hogan; Daniel Thompson; Michael Opdenacker; David Cohen; Felipe Balbi; Ralf Baechle
Subject: Re: [PATCH] mei, make modules.alias UUID information easier to read

On Fri, Aug 14, 2015 at 09:05:40AM -0400, Prarit Bhargava wrote:

Quoted 8 lines
Quoted 1 line

2nd try on this ...

What changed from the first patch? I need some "version information" to figure out what is going on.

Can you resend it with that information?

thanks,

I saw you've already merged it to the testing branch so I'm not resending,
Anyhow the original patch was done over Linus' master branch instead of char-misc-next so it needed a rebase.

Thanks.
Tomas