linux/drivers/dma/cppi41.c, branch v4.9

dmaengine: cppi41: More PM runtime fixes

2016-11-17T10:39:23Z

Fix use of u32 instead of int for checking for negative errors values as pointed out by Dan Carpenter . And while testing the PM runtime error path by randomly returning failed values in runtime resume, I noticed two more places that need fixing: - If pm_runtime_get_sync() fails in probe, we still need to do pm_runtime_put_sync() to keep the use count happy. We could call pm_runtime_put_noidle() on the error path, but we're just going to call pm_runtime_disable() after that so pm_runtime_put_sync() will do what we want - We should print an error if pm_runtime_get_sync() fails in cppi41_dma_alloc_chan_resources() so we know where it happens Reported-by: Dan Carpenter Fixes: 740b4be3f742 ("dmaengine: cpp41: Fix handling of error path") Signed-off-by: Tony Lindgren Signed-off-by: Vinod Koul

dmaengine: cpp41: Fix handling of error path

2016-11-14T08:30:16Z

If we return early on pm_runtime_get() error, we need to also call pm_runtime_put_noidle() as pointed out in a musb related thread by Johan Hovold . This is to keep the PM runtime use counts happy. Fixes: fdea2d09b997 ("dmaengine: cppi41: Add basic PM runtime support") Cc: Johan Hovold Signed-off-by: Tony Lindgren Signed-off-by: Vinod Koul

dmaengine: cppi41: Fix unpaired pm runtime when only a USB hub is connected

2016-11-14T05:27:32Z

On am335x with musb host we can end up with unpaired pm runtime calls if a hub with no devices is connected and disconnected. This is because of the conditional pm runtime calls which are always a bad idea. Let's fix the issue by making them unconditional and paired in each function. Fixes: fdea2d09b997 ("dmaengine: cppi41: Add basic PM runtime support") Signed-off-by: Tony Lindgren Signed-off-by: Vinod Koul

dmaengine: cppi41: Fix list not empty warning on module removal

2016-11-14T05:27:21Z

If musb controller is configured with USB peripherals and we have enumerated with a USB host, we can get warnings on removal of the modules: WARNING: CPU: 0 PID: 1269 at drivers/dma/cppi41.c:391 cppi41_dma_free_chan_resources Fix the issue by adding the missing pm_runtime_get to cppi41_dma_free_chan_resources to make sure the pending work list is cleared on removal. Fixes: fdea2d09b997 ("dmaengine: cppi41: Add basic PM runtime support") Signed-off-by: Tony Lindgren Signed-off-by: Vinod Koul

Merge branch 'topic/err_reporting' into for-linus

2016-10-03T03:47:33Z

Signed-off-by: Vinod Koul Conflicts: drivers/dma/cppi41.c

dmaengine: cppi41: Ignore EINPROGRESS for PM runtime

2016-09-14T13:43:22Z

We can occasionally get -EINPROGRESS for pm_runtime_get. In that case we can just continue as we're queueing transfers anyways when pm_runtime_active is not set. Fixes: fdea2d09b997 ("dmaengine: cppi41: Add basic PM runtime support") Signed-off-by: Tony Lindgren Signed-off-by: Vinod Koul

dmaengine: cppi41: mark PM functions as __maybe_unused

2016-09-07T08:08:57Z

When CONFIG_PM_SLEEP is disabled, we get a build error in the cppi41 dmaengine driver, since the runtime-pm functions are hidden within the wrong #ifdef: drivers/dma/cppi41.c:1158:21: error: 'cppi41_runtime_suspend' undeclared here (not in a function) This removes the #ifdef and instead uses __maybe_unused annotations that cannot have this problem. Signed-off-by: Arnd Bergmann Fixes: fdea2d09b997 ("dmaengine: cppi41: Add basic PM runtime support") Signed-off-by: Vinod Koul

dmaengine: cppi41: Add basic PM runtime support

2016-08-31T15:54:24Z

Let's keep the device enabled between cppi41_dma_issue_pending() and dmaengine_desc_get_callback_invoke() and rely on the PM runtime autoidle timeout elsewhere. As the PM runtime is for whole device, not for each channel, we need to queue pending transfers if the device is PM runtime suspended. Then we start the pending transfers in PM runtime resume. Signed-off-by: Tony Lindgren Signed-off-by: Vinod Koul

dmaengine: cppi41: Prepare to add PM runtime support

2016-08-31T04:52:19Z

Let's just move code from cppi41_dma_issue_pending() to push_desc_queue() as that's the only call to push_desc_queue(). We want to do this for PM runtime as we need to call push_desc_queue() also for pending queued transfers from PM runtime resume. No functional changes, just moves code around. Signed-off-by: Tony Lindgren Signed-off-by: Vinod Koul

dmaengine: cppi41: convert callback to helper function

2016-08-08T02:41:38Z

This is in preperation of moving to a callback that provides results to the callback for the transaction. The conversion will maintain current behavior and the driver must convert to new callback mechanism at a later time in order to receive results. Signed-off-by: Dave Jiang Reviewed-by: Lars-Peter Clausen Signed-off-by: Vinod Koul