On 02-11-22, 11:07, Jerry Snitselaar wrote: > Hi Vinod, > > Thoughts on this patch? > > Maybe changing the summary to "dmaengine: Fix double increment of client_count in dma_chan_get()" > would be clearer? Yes that would be better > > On Sat, Oct 01, 2022 at 01:36:52AM +0800, Koba Ko wrote: > > If the passed client_count is 0, > > it would be incremented by balance_ref_count first > > then increment one more. > > This would cause client_count to 2. > > > > cat /sys/class/dma/dma0chan*/in_use > > 2 > > 2 > > 2 > > Would this be better? > > The first time dma_chan_get() is called for a channel the channel > client_count is incorrectly incremented twice for public channels, > first in balance_ref_count(), and again prior to returning. This > results in an incorrect client count which will lead to the > channel resources not being freed when they should be. A simple > test of repeated module load and unload of async_tx on a Dell > Power Edge R7425 also shows this resulting in a kref underflow > warning. Agree, also if you have the underflow warning handy, do add it to the log -- ~Vinod
