linux/drivers/base/regmap/regcache.c, branch v3.18

Merge remote-tracking branch 'regmap/topic/cache' into regmap-next

2014-09-29T19:49:42Z

regmap: cache: Do not fail silently from regcache_sync calls

2014-09-16T16:53:40Z

Call stack of regcache_sync calls may not emit any error message even if operation was cancelled due an error in I/O driver. One such a silent error is for instance if I2C bus driver doesn't receive ACK from the I2C device and returns -EREMOTEIO. Since many users of regcache_sync() don't check and print the error there is no any indication that HW registers are potentially out of sync. Signed-off-by: Jarkko Nikula Signed-off-by: Mark Brown

regmap: Don't attempt block writes when syncing cache on single_rw devices

2014-08-31T12:22:37Z

If the device can't support block writes then don't attempt to use raw syncing which will automatically generate block writes for adjacent registers, use the existing _single() block syncing implementation. Reported-by: Jarkko Nikula Tested-by: Jarkko Nikula Signed-off-by: Mark Brown Cc: stable@vger.kernel.org

Merge remote-tracking branches 'regmap/topic/cache', 'regmap/topic/irq', 'regmap/topic/lock', 'regmap/topic/mmio', 'regmap/topic/nodev', 'regmap/topic/parse-val' and 'regmap/topic/patch' into regmap-next

2014-03-28T11:50:43Z

regmap: cache: Don't attempt to sync non-writeable registers

2014-03-19T10:38:56Z

In the regcache_default_sync, if a register isn't writeable, then _regmap_write will return an error and the rest of the sync will be aborted. Avoid this by checking if a register is writeable before trying to sync it. Signed-off-by: Dylan Reid Signed-off-by: Mark Brown

regmap: cache: Step by stride in default sync

2014-03-19T10:38:54Z

The default sync operation was still assuming a stride of one, fix it to respect the reg_stride set in the map. Signed-off-by: Dylan Reid Signed-off-by: Mark Brown

regmap: Check stride of register patch as we register it

2014-02-22T02:15:13Z

Currently, we check the registers in the patch are aligned to the register stride everytime we sync the cache and the first time the patch is written out is unchecked. This patch checks the register patch when we first register it so the first writes are no longer unchecked and then doesn't check on subsequent syncs as the patch will be unchanged. Signed-off-by: Charles Keepax Signed-off-by: Mark Brown

regmap: cache: Handle stride > 1 in sync_block_raw_flush

2014-01-27T18:25:11Z

regcache_sync_block_raw_flush takes the address of the base register and the address of one past the last register to write to. "count" is the number of registers in the range, not the number of bytes, it should be (end addr - start addr) / stride. Without accounting for strides greater than one, registers past the end might be synced or the writeable_reg callback at the beginning of _regmap_raw_write will fail and nothing will be written. Signed-off-by: Dylan Reid Signed-off-by: Mark Brown

regmap: Use async I/O during cache sync

2013-10-10T20:15:23Z

Try to speed up I/O a little by not synchronising until we are finished scheduling writes. A brief survey of existing users suggests we have none that would currently benefit from an async cache sync. Signed-off-by: Mark Brown

regmap: Simplify the initiation of async I/O

2013-10-09T13:05:24Z

Rather than passing a flag around through the entire call stack store it in the regmap struct and read it when required. This minimises the visibility of the feature through the API, minimising the code updates needed to use it more widely. Signed-off-by: Mark Brown