On 05/04/2015 04:05 PM, Kevin Cernekee wrote:
On Sun, May 3, 2015 at 11:38 PM, Lars-Peter Clausen lars@metafoo.de wrote:
diff --git a/drivers/base/regmap/regcache.c b/drivers/base/regmap/regcache.c index 7eb7b3b98794..63af3103d0c6 100644 --- a/drivers/base/regmap/regcache.c +++ b/drivers/base/regmap/regcache.c @@ -253,6 +253,9 @@ static int regcache_default_sync(struct regmap *map, unsigned int min, unsigned int max) { unsigned int reg;
bool no_sync_defaults = map->no_sync_defaults;map->no_sync_defaults = false;This needs to be done at the end in regcache_sync(), the same place where dirty is set to false.
But map->cache_dirty means "any register is dirty," not "all registers are dirty." So it can only be cleared after a successful flush.
If one of the writes fails and regcache_sync() has to return prematurely, we probably don't want no_sync_defaults to stay true because some of the HW registers might not match map->reg_defaults anymore.
Makes sense. But it should still be done in a central place rather than repeating it in every sync implementation. You can still put it at the end of regcache_sync(), just clear it regardless of whether it succeeded or not.