Re: [RFT][PATCH 2/3] regulator: ab8500-ext: Don't update info->update_val if set_mode() fails

[Bengt Jönsson]

On 04/10/2013 02:55 PM, Axel Lin wrote:
This ensures info->update_val status is still correct if set_mode() call fails.
Otherwise, get_mode() may return wrong status if a set_mode() call fails.
Signed-off-by: Axel Lin <axel....@ingics.com>
---
  drivers/regulator/ab8500-ext.c |   19 ++++++++++++-------
  1 file changed, 12 insertions(+), 7 deletions(-)

diff --git a/drivers/regulator/ab8500-ext.c b/drivers/regulator/ab8500-ext.c
index f0a1bbf..1efe702 100644
--- a/drivers/regulator/ab8500-ext.c
+++ b/drivers/regulator/ab8500-ext.c
@@ -181,6 +181,7 @@ static int ab8500_ext_regulator_set_mode(struct 
regulator_dev *rdev,
  {
        int ret = 0;
        struct ab8500_ext_regulator_info *info = rdev_get_drvdata(rdev);
+       u8 regval;
  
  	if (info == NULL) {
                dev_err(rdev_get_dev(rdev), "regulator info null pointer\n");
@@ -189,14 +190,14 @@ static int ab8500_ext_regulator_set_mode(struct 
regulator_dev *rdev,
  
  	switch (mode) {
        case REGULATOR_MODE_NORMAL:
-               info->update_val = info->update_val_hp;
+               regval = info->update_val_hp;
                break;
        case REGULATOR_MODE_IDLE:
                /* Always on with high power mode if info->cfg->hwreq is set */
                if (info->cfg && info->cfg->hwreq)
                        return -EINVAL;
  
-		info->update_val = info->update_val_lp;
+               regval = info->update_val_lp;
                break;
  
  	default:
@@ -204,12 +205,14 @@ static int ab8500_ext_regulator_set_mode(struct 
regulator_dev *rdev,
        }
  
  	if (ab8500_ext_regulator_is_enabled(rdev)) {
-               u8 regval;
-
-               ret = enable(info, &regval);
-               if (ret < 0)
+               ret = abx500_mask_and_set_register_interruptible(info->dev,
+                                       info->update_bank, info->update_reg,
+                                       info->update_mask, regval);
+               if (ret < 0) {
                        dev_err(rdev_get_dev(rdev),
                                "Could not set regulator mode.\n");
+                       return ret;
+               }
  
  		dev_dbg(rdev_get_dev(rdev),
                        "%s-set_mode (bank, reg, mask, value): "
@@ -218,7 +221,9 @@ static int ab8500_ext_regulator_set_mode(struct 
regulator_dev *rdev,
                        info->update_mask, regval);
        }
  
-	return ret;
+       info->update_val = regval;
+
+       return 0;
  }
  
  static unsigned int ab8500_ext_regulator_get_mode(struct regulator_dev *rdev)
This patch is needed to fix the error path but it partly depends on the 
previous patch so I would like to get that one sorted out before ack-ing 
this patch.

Regards,

Bengt
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/