On 8 December 2011 13:46, Vinod Koul vinod.koul@intel.com wrote:
On Thu, 2011-12-08 at 16:43 +0900, Kukjin Kim wrote:
Vinod Koul wrote:
On Tue, 2011-12-06 at 16:15 +0530, Tushar Behera wrote:
amba_probe() now calls pm_runtime_get_noresume() and pm_runtime_enable() for the devices before the device probe is called. Hence we don't need to call pm_runtime_get_xxx and pm_runtime_enable() in device probe again. In the same way, since amba_remove() calls the respective pm_runtime functions, those functions need not be called from device remove.
This patch fixes following run time error with pl330 driver.
dma-pl330 dma-pl330.0: Unbalanced pm_runtime_enable! dma-pl330 dma-pl330.0: failed to get runtime pm
Signed-off-by: Giridhar Maruthy giridhar.maruthy@linaro.org Signed-off-by: Tushar Behera tushar.behera@linaro.org
Looks fine to me. Do you want these to go thru slave-dma or samsung tree.
Hi Vinod,
I think, this patch can be sent to upstream via slave-dma tree and second one via Samsung tree separately and you can add my and Boojin Kim's ack(actually, she replied) on this when you apply.
Okay, I have applied this one only
Guys, any reason to keep me, the author of the driver, out of loop ? I almost lost this patch, had it not for chance.
Also, I noticed 42bc9cf45939c2 'DMA: PL330: Add DMA_CYCLIC capability' while implements my correction, doesn't carry the ack http://www.spinics.net/lists/linux-samsung-soc/msg06386.html
While, unlike some, I am not interested in claiming territory by pissing around acks, I would sure like to log my acks if I spent personal time maintaining the code I am supposed to be responsible for.