hi Laurent and all who it may concern, Ever since I switches to using imx7-media-csi on imx8mq I have seen a bug that hangs /freezes the system so that I can't really save a trace or get logs for it. It happens not every time when starting streaming, but sometimes - when starting to stream frames. It kind of looks like a lock not released or something similar, but some race condition might happen since it's happening every 2 to 10 times. For me, when it fails, it fails the first time when starting to stream after rebooting. If it succeeds that first time, it continues to succeed, until I reboot. The last logs I can see currently in the failing case are [ 37.811693] videobuf2_common: [cap-00000000e2303470] __setup_offsets: buffer 0, plane 0 offset 0x00000000 [ 37.824161] videobuf2_common: [cap-00000000e2303470] __setup_offsets: buffer 1, plane 0 offset 0x003d0000 [ 37.836844] videobuf2_common: [cap-00000000e2303470] __setup_offsets: buffer 2, plane 0 offset 0x007a0000 [ 37.849406] videobuf2_common: [cap-00000000e2303470] __setup_offsets: buffer 3, plane 0 offset 0x00b70000 [ 37.860091] videobuf2_common: [cap-00000000e2303470] __vb2_queue_alloc: allocated 4 buffers, 1 plane(s) each [ 37.870770] videobuf2_common: [cap-00000000e2303470] vb2_mmap: buffer 0, plane 0 successfully mapped [ 37.880155] videobuf2_common: [cap-00000000e2303470] vb2_mmap: buffer 1, plane 0 successfully mapped [ 37.889532] videobuf2_common: [cap-00000000e2303470] vb2_mmap: buffer 2, plane 0 successfully mapped [ 37.898880] videobuf2_common: [cap-00000000e2303470] vb2_mmap: buffer 3, plane 0 successfully mapped [ 37.908127] videobuf2_common: [cap-00000000e2303470] vb2_core_qbuf: qbuf of buffer 0 succeeded [ 37.916949] videobuf2_common: [cap-00000000e2303470] vb2_core_qbuf: qbuf of buffer 1 succeeded [ 37.925659] videobuf2_common: [cap-00000000e2303470] vb2_core_qbuf: qbuf of buffer 2 succeeded [ 37.934395] videobuf2_common: [cap-00000000e2303470] vb2_core_qbuf: qbuf of buffer 3 succeeded [ 38.077613] videobuf2_common: [cap-00000000e2303470] vb2_core_streamon: successful [ 38.085523] videobuf2_common: [cap-00000000e2303470] __vb2_wait_for_done_vb: will sleep waiting for buffers but sometimes I see it hang a few lines later: [ 60.996573] videobuf2_common: [cap-00000000eb3bac3a] __vb2_wait_for_done_vb: will sleep waiting for b uffers [ 61.048564] imx7_csi_vb2_buf_done: start [ 61.052514] videobuf2_common: vb2_buffer_done: start [ 61.057478] videobuf2_common: [cap-00000000eb3bac3a] vb2_buffer_done: done processing on buffer 0, st ate: done [ 61.067479] videobuf2_common: [cap-00000000eb3bac3a] __vb2_buf_mem_finish: call_memop(0, finish) [ 61.076268] videobuf2_common: vb2_buffer_done: waking up done_wq the vb2 logs with advanced debugging preceding these, are identical in the success and failing case. With the exact same sensor driver but the nxp CSI driver https://source.codeaurora.org/external/imx/linux-imx/tree/drivers/media/platform/mxc/capture/mx6s_capture.c?h=imx_5.4.0_8dxlphantom_er is never happened (and the mipi driver is most likely not the issue since it's controller config that should be exactly the same as well). In the success case then imx7_csi_vb2_buf_done() gets called then. I'm quite confident now that it's a bug in imx7-media-csi since it uses the buffers. So what's happening? I really don't yet know. Any idea? thank you, martin and the success case continues like so (with some more debug prints that in mainline): [ 26.493415] videobuf2_common: [cap-00000000500f079b] __vb2_wait_for_done_vb: will sleep waiting for b uffers [ 26.531273] imx7_csi_vb2_buf_done: start [ 26.535238] videobuf2_common: [cap-00000000500f079b] vb2_buffer_done: done processing on buffer 0, st ate: done [ 26.545261] videobuf2_common: vb2_buffer_done: waking up done_wq [ 26.551284] imx_media_capture_device_next_buf: start [ 26.551294] videobuf2_common: [cap-00000000500f079b] __vb2_wait_for_done_vb: loop done [ 26.556259] imx_media_capture_device_next_buf: ok [ 26.563269] imx7_csi_vb2_buf_done: start and so on
