Skip to contentSkip to conversation

mfd: tps6586x: Remove device ID check

PATCH In review Stephen Warren 3 participants 6 messages Dec 2, 2010 – Dec 19, 2010 Multifunction devices (MFD)

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 mfd: tps6586x: Remove device ID check

... and convert it to a dev_info print at probe time.

There are many variants of this chip with different values of VERSIONCRC. The set of values is large, and not useful to enumerate. All are SW compatible. The difference lies in default settings of the various power rails, and other similar differences. The driver, or clients of the driver, shouldn't be affected by this, since all rails should be programmed into the desired state in all cases for correct operation.

---
 drivers/mfd/tps6586x.c |    6 +-----
 1 files changed, 1 insertions(+), 5 deletions(-)
drivers/mfd/tps6586x.c +1 -5
index b4931ab..cb482ba 100644
@@ -498,11 +498,7 @@ static int __devinit tps6586x_i2c_probe(struct i2c_client *client,
 		return -EIO;
 	}
 
-	if ((ret != TPS658621A_VERSIONCRC) &&
-	    (ret != TPS658621C_VERSIONCRC)) {
-		dev_err(&client->dev, "Unsupported chip ID: %x\n", ret);
-		return -ENODEV;
-	}
+	dev_info(&client->dev, "VERSIONCRC is %02x\n", ret);
 
 	tps6586x = kzalloc(sizeof(struct tps6586x), GFP_KERNEL);
 	if (tps6586x == NULL)
Signature · 2 lines
1.7.0.4
nvpublic
#1 lore raw
Quoted 19 lines

-----Original Message-----
From: Stephen Warren [mailto:[email protected]]
Sent: Thursday, December 02, 2010 10:33 AM
To: [email protected]
Cc: [email protected]; Andrew Chew; [email protected]; Stephen Warren
Subject: [PATCH] mfd: tps6586x: Remove device ID check

... and convert it to a dev_info print at probe time.

There are many variants of this chip with different values of VERSIONCRC.
The set of values is large, and not useful to enumerate. All are SW compatible. The difference lies in default settings of the various power
rails, and other similar differences. The driver, or clients of the driver, shouldn't be affected by this, since all rails should be programmed into the desired state in all cases for correct operation.

Quoted 2 lines

LGTM.

Acked by: Andrew Chew <[email protected]>

#2 lore raw
In reply to Stephen Warren

Hi Stephen,

Quoted 8 lines · Stephen Warren

... and convert it to a dev_info print at probe time.

There are many variants of this chip with different values of VERSIONCRC. The set of values is large, and not useful to enumerate. All are SW compatible. The difference lies in default settings of the various power rails, and other similar differences. The driver, or clients of the driver, shouldn't be affected by this, since all rails should be programmed into the desired state in all cases for correct operation.

Fair enough, but then you can remove the TPS658621A_VERSIONCRC and TPS658621C_VERSIONCRC defines.

Cheers,
Samuel.

Signature · 2 lines
Intel Open Source Technology Centre
http://oss.intel.com/
#3 lore raw

PATCH 1/1 mfd: tps6586x: Remove device ID check

... and convert it to a dev_info print at probe time.

There are many variants of this chip with different values of VERSIONCRC. The set of values is large, and not useful to enumerate. All are SW compatible. The difference lies in default settings of the various power rails, and other similar differences. The driver, or clients of the driver, shouldn't be affected by this, since all rails should be programmed into the desired state in all cases for correct operation.

---
 drivers/mfd/tps6586x.c |    8 +-------
 1 files changed, 1 insertions(+), 7 deletions(-)
drivers/mfd/tps6586x.c +1 -7
index b4931ab..3575739 100644
@@ -46,8 +46,6 @@
 
 /* device id */
 #define TPS6586X_VERSIONCRC	0xcd
-#define TPS658621A_VERSIONCRC	0x15
-#define TPS658621C_VERSIONCRC	0x2c
 
 struct tps6586x_irq_data {
 	u8	mask_reg;
@@ -498,11 +496,7 @@ static int __devinit tps6586x_i2c_probe(struct i2c_client *client,
 		return -EIO;
 	}
 
-	if ((ret != TPS658621A_VERSIONCRC) &&
-	    (ret != TPS658621C_VERSIONCRC)) {
-		dev_err(&client->dev, "Unsupported chip ID: %x\n", ret);
-		return -ENODEV;
-	}
+	dev_info(&client->dev, "VERSIONCRC is %02x\n", ret);
 
 	tps6586x = kzalloc(sizeof(struct tps6586x), GFP_KERNEL);
 	if (tps6586x == NULL)
Signature · 1 line
1.7.0.4
#4 lore raw
Quoted 4 lines

From: Stephen Warren [mailto:[email protected]]
Sent: Thursday, December 09, 2010 10:30 AM

... and convert it to a dev_info print at probe time.

Samuel,

Sorry to ping you, but does this updated patch look OK and address your concerns? Thanks.

Quoted 41 lines

There are many variants of this chip with different values of VERSIONCRC. The set of values is large, and not useful to enumerate. All are SW compatible. The difference lies in default settings of the various power rails, and other similar differences. The driver, or clients of the driver, shouldn't be affected by this, since all rails should be programmed into the desired state in all cases for correct operation.

---
 drivers/mfd/tps6586x.c |    8 +-------
 1 files changed, 1 insertions(+), 7 deletions(-)
drivers/mfd/tps6586x.c +1 -8
index b4931ab..3575739 100644
@@ -46,8 +46,6 @@
 
 /* device id */
 #define TPS6586X_VERSIONCRC	0xcd
-#define TPS658621A_VERSIONCRC	0x15
-#define TPS658621C_VERSIONCRC	0x2c
 
 struct tps6586x_irq_data {
 	u8	mask_reg;
@@ -498,11 +496,7 @@ static int __devinit tps6586x_i2c_probe(struct i2c_client *client,
 		return -EIO;
 	}
 
-	if ((ret != TPS658621A_VERSIONCRC) &&
-	    (ret != TPS658621C_VERSIONCRC)) {
-		dev_err(&client->dev, "Unsupported chip ID: %x\n", ret);
-		return -ENODEV;
-	}
+	dev_info(&client->dev, "VERSIONCRC is %02x\n", ret);
 
 	tps6586x = kzalloc(sizeof(struct tps6586x), GFP_KERNEL);
 	if (tps6586x == NULL)
--

1.7.0.4

Signature · 1 line
nvpublic
#5 lore raw
In reply to Stephen Warren

Hi Stephen,

Quoted 11 lines · Stephen Warren

... and convert it to a dev_info print at probe time.

There are many variants of this chip with different values of VERSIONCRC. The set of values is large, and not useful to enumerate. All are SW compatible. The difference lies in default settings of the various power rails, and other similar differences. The driver, or clients of the driver, shouldn't be affected by this, since all rails should be programmed into the desired state in all cases for correct operation.

Patch applied, thanks a lot.

Cheers,
Samuel.

Signature · 2 lines
Intel Open Source Technology Centre
http://oss.intel.com/