Hi Hans, On Tue, Aug 27, 2019 at 12:51:08PM +0300, Laurent Pinchart wrote: > On Tue, Aug 27, 2019 at 11:07:56AM +0200, Hans Verkuil wrote: > > Hi Laurent, > > > > I tried this series with my Pandaboard, but it broke my Pandaboard. After > > doing a git bisect it ended up with this patch as the culprit. > > > > If I boot my Pandaboard I get this: > > > > [ 3.271881] omapdss_dss 58000000.dss: 58000000.dss supply vdda_video not found, using dummy regulator > > [ 3.285369] omapdss_dss 58000000.dss: 58000000.dss supply vdda_video not found, using dummy regulator > > [ 24.286773] rcu: INFO: rcu_sched self-detected stall on CPU > > [ 24.292388] rcu: 0-...!: (1306 ticks this GP) idle=6b6/1/0x40000002 softirq=85/85 fqs=19 > > [ 24.300689] (t=2100 jiffies g=-1063 q=10) > > [ 24.304809] rcu: rcu_sched kthread starved for 2058 jiffies! g-1063 f0x0 RCU_GP_WAIT_FQS(5) ->state=0x402 ->cpu=1 > > [ 24.315124] rcu: RCU grace-period kthread stack dump: > > [ 24.315124] rcu_sched I 0 10 2 0x00000000 > > [ 24.325805] [<c09b7130>] (__schedule) from [<c09b73a8>] (schedule+0x40/0xc0) > > [ 24.332885] [<c09b73a8>] (schedule) from [<c09ba9d0>] (schedule_timeout+0x174/0x2a8) > > [ 24.340698] [<c09ba9d0>] (schedule_timeout) from [<c01943a4>] (rcu_gp_kthread+0x49c/0x9f4) > > [ 24.349121] [<c01943a4>] (rcu_gp_kthread) from [<c01559f8>] (kthread+0x13c/0x148) > > [ 24.356658] [<c01559f8>] (kthread) from [<c01010e8>] (ret_from_fork+0x14/0x2c) > > [ 24.363922] Exception stack(0xe70a1fb0 to 0xe70a1ff8) > > [ 24.363922] 1fa0: 00000000 00000000 00000000 00000000 > > [ 24.363922] 1fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 > > [ 24.385437] 1fe0: 00000000 00000000 00000000 00000000 00000013 00000000 > > [ 24.392089] NMI backtrace for cpu 0 > > [ 24.392120] CPU: 0 PID: 105 Comm: kworker/0:2 Not tainted 5.3.0-rc3-arm #152 > > [ 24.392120] Hardware name: Generic OMAP4 (Flattened Device Tree) > > [ 24.408721] Workqueue: events deferred_probe_work_func > > [ 24.408721] [<c010fd70>] (unwind_backtrace) from [<c010b7e8>] (show_stack+0x10/0x14) > > [ 24.421783] [<c010b7e8>] (show_stack) from [<c09a054c>] (dump_stack+0x84/0x9c) > > [ 24.429077] [<c09a054c>] (dump_stack) from [<c09a6ecc>] (nmi_cpu_backtrace+0x8c/0xc0) > > [ 24.436950] [<c09a6ecc>] (nmi_cpu_backtrace) from [<c09a7014>] (nmi_trigger_cpumask_backtrace+0x114/0x12c) > > [ 24.446655] [<c09a7014>] (nmi_trigger_cpumask_backtrace) from [<c01961e0>] (rcu_dump_cpu_stacks+0xa4/0xcc) > > [ 24.456359] [<c01961e0>] (rcu_dump_cpu_stacks) from [<c01952bc>] (rcu_sched_clock_irq+0x608/0x7f8) > > [ 24.465362] [<c01952bc>] (rcu_sched_clock_irq) from [<c019b21c>] (update_process_times+0x34/0x6c) > > [ 24.465362] [<c019b21c>] (update_process_times) from [<c01acfcc>] (tick_sched_timer+0x4c/0xa8) > > [ 24.483001] [<c01acfcc>] (tick_sched_timer) from [<c019c470>] (__hrtimer_run_queues+0x154/0x1fc) > > [ 24.491851] [<c019c470>] (__hrtimer_run_queues) from [<c019c84c>] (hrtimer_interrupt+0x11c/0x2ac) > > [ 24.500762] [<c019c84c>] (hrtimer_interrupt) from [<c010f3ac>] (twd_handler+0x30/0x38) > > [ 24.500762] [<c010f3ac>] (twd_handler) from [<c018959c>] (handle_percpu_devid_irq+0x78/0x138) > > [ 24.517333] [<c018959c>] (handle_percpu_devid_irq) from [<c0183a60>] (generic_handle_irq+0x24/0x34) > > [ 24.526458] [<c0183a60>] (generic_handle_irq) from [<c0184028>] (__handle_domain_irq+0x5c/0xb4) > > [ 24.526458] [<c0184028>] (__handle_domain_irq) from [<c03bbf2c>] (gic_handle_irq+0x58/0x9c) > > [ 24.526458] [<c03bbf2c>] (gic_handle_irq) from [<c0101a8c>] (__irq_svc+0x6c/0x90) > > [ 24.526458] Exception stack(0xe6439d68 to 0xe6439db0) > > [ 24.551177] 9d60: c0f783e8 00000000 00000000 efdb49d0 e647f2e8 efdb7f74 > > [ 24.551177] 9d80: e647f2e8 00000000 e647f294 e647f2e8 00000000 efdb5a98 00000000 e6439db8 > > [ 24.551177] 9da0: c04e9830 c04e9858 20000153 ffffffff > > [ 24.577789] [<c0101a8c>] (__irq_svc) from [<c04e9858>] (of_drm_find_bridge+0x40/0x84) > > [ 24.577789] [<c04e9858>] (of_drm_find_bridge) from [<c0526608>] (omapdss_device_init_output+0x38/0x140) > > [ 24.595153] [<c0526608>] (omapdss_device_init_output) from [<c0530294>] (dpi_init_port+0x208/0x294) > > [ 24.604248] [<c0530294>] (dpi_init_port) from [<c05278c4>] (dss_probe+0x2b4/0x550) > > [ 24.604248] [<c05278c4>] (dss_probe) from [<c0576c1c>] (platform_drv_probe+0x48/0x98) > > [ 24.604248] [<c0576c1c>] (platform_drv_probe) from [<c0574eb0>] (really_probe+0xf0/0x2c4) > > [ 24.627990] [<c0574eb0>] (really_probe) from [<c057520c>] (driver_probe_device+0x60/0x168) > > [ 24.627990] [<c057520c>] (driver_probe_device) from [<c0573410>] (bus_for_each_drv+0x84/0xc8) > > [ 24.644866] [<c0573410>] (bus_for_each_drv) from [<c0574d4c>] (__device_attach+0xd4/0x140) > > [ 24.644866] [<c0574d4c>] (__device_attach) from [<c05740f8>] (bus_probe_device+0x84/0x8c) > > [ 24.661376] [<c05740f8>] (bus_probe_device) from [<c057456c>] (deferred_probe_work_func+0x64/0x90) > > [ 24.670379] [<c057456c>] (deferred_probe_work_func) from [<c014f960>] (process_one_work+0x204/0x41c) > > [ 24.670410] [<c014f960>] (process_one_work) from [<c01510d4>] (worker_thread+0x2a8/0x5bc) > > [ 24.687805] [<c01510d4>] (worker_thread) from [<c01559f8>] (kthread+0x13c/0x148) > > [ 24.687805] [<c01559f8>] (kthread) from [<c01010e8>] (ret_from_fork+0x14/0x2c) > > [ 24.702484] Exception stack(0xe6439fb0 to 0xe6439ff8) > > [ 24.702484] 9fa0: 00000000 00000000 00000000 00000000 > > [ 24.702484] 9fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 > > [ 24.723999] 9fe0: 00000000 00000000 00000000 00000000 00000013 00000000 > > > > I can send my .config upon request if needed. > > > > CONFIG_OMAP2_DSS_DPI is 'y', but there is nothing connected to the DPI port. > > I've retested this on my pandaboard and couldn't reproduce the issue. As > I suspected a problem related to the kernel configuration, I tried > building the drivers in the kernel image instead of using modules, and I > then got a similar looking crash. I'm investigating it now. Fix pushed to git://linuxtv.org/pinchartl/media.git omapdrm/bridge/devel The issue was caused by incorrect cleanup on probe deferral, fixed by a combination of "drm/omap: dss: Cleanup DSS ports on initialisation failure" (new patch) and changes to "drm/omap: dpi: Register a drm_bridge" and "drm/omap: spi: Register a drm_bridge". -- Regards, Laurent Pinchart _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/dri-devel
