Hi Eugeniu
thanks for your test & comments and adding more people for review
I will add necessary backtrace information to description and
rephrase commit summary in V2 patch
Thanks,
Jiada
On 2019/04/17 2:48, Eugeniu Rosca wrote:
Hi Jiada,
Adding below people, since they've made recent contributions to the
driver and might be interested in your patch:
git log master --since="1 year" -- drivers/thermal/rcar_gen3_thermal.c \
| grep -o "\-by:.*" | sed 's/\-by: //' | sort | uniq -c | sort -rn
7 Eduardo Valentin <edubez...@gmail.com>
6 Simon Horman <horms+rene...@verge.net.au>
5 Niklas Söderlund <niklas.soderlund+rene...@ragnatech.se>
2 Geert Uytterhoeven <geert+rene...@glider.be>
1 Sergei Shtylyov <sergei.shtyl...@cogentembedded.com>
1 Marek Vasut <marek.vasut+rene...@gmail.com>
1 Kuninori Morimoto <kuninori.morimoto...@renesas.com>
1 Hien Dang <hien.dang...@renesas.com>
1 Fabrizio Castro <fabrizio.cas...@bp.renesas.com>
1 Dien Pham <dien.pham...@renesas.com>
1 Daniel Lezcano <daniel.lezc...@linaro.org>
1 Biju Das <biju....@bp.renesas.com>
I confirm that loading and unloading the rcar3 thermal driver in a
loop produces soft lockup using v5.1-rc5-10-g618d919cae2f on
H3-ES2.0-Salvator-X.
Full log and .config can be found here:
https://gist.github.com/erosca/1f76b6dd897cdc39581fca475155e363
I post an excerpt from the above [1] (why not including it in the
description?). Also, why not rephrasing the commit summary line in such
a way that everybody understands this patch fixes a severe issue, e.g.
"thermal: rcar_gen3_thermal: Fix soft lockup on probe" ?
BTW, with this patch applied I left the thermal driver being
loaded/unloaded on the target for over one hour w/o seeing the issue
reproduced. So, while there might be slight variations in how the final
solution looks like, I think the patch already deserves a:
Tested-by: Eugeniu Rosca <ero...@de.adit-jv.com>
[1] Soft lockup reproduced with v5.1-rc5-10-g618d919cae2f
root@rcar-gen3:~# while true; do rmmod rcar_gen3_thermal; modprobe
rcar_gen3_thermal; done
[ 43.439043] rcar_gen3_thermal e6198000.thermal: TSC0: Loaded 0 trip points
[ 43.451670] rcar_gen3_thermal e6198000.thermal: TSC1: Loaded 0 trip points
[ 43.463974] rcar_gen3_thermal e6198000.thermal: TSC2: Loaded 0 trip points
[..]
[ 553.966104] rcar_gen3_thermal e6198000.thermal: TSC0: Loaded 0 trip points
[ 553.978759] rcar_gen3_thermal e6198000.thermal: TSC1: Loaded 0 trip points
[ 553.991058] rcar_gen3_thermal e6198000.thermal: TSC2: Loaded 0 trip points
[ 562.235306] renesas_sdhi_internal_dmac ee100000.sd: timeout waiting for
hardware interrupt (CMD25)
[ 567.353336] renesas_sdhi_internal_dmac ee100000.sd: timeout waiting for
hardware interrupt (CMD13)
[ 572.473318] renesas_sdhi_internal_dmac ee100000.sd: timeout waiting for
hardware interrupt (CMD13)
[ 577.593328] renesas_sdhi_internal_dmac ee100000.sd: timeout waiting for
hardware interrupt (CMD12)
[ 579.189148] rcu: INFO: rcu_preempt self-detected stall on CPU
[ 579.195329] rcu: 0-....: (1 GPs behind) idle=b76/1/0x4000000000000004
softirq=263851/263851 fqs=6251 last_accelerate: e095/4240, Nonlazy posted: ...
[ 579.209711] rcu: (t=25008 jiffies g=346801 q=468)
[ 579.214801] Task dump for CPU 0:
[ 579.218178] modprobe R running task 0 6337 1420 0x0000002a
[ 579.225514] Call trace:
[ 579.228103] dump_backtrace+0x0/0x1dc
[ 579.231934] show_stack+0x24/0x30
[ 579.235410] sched_show_task+0x31c/0x36c
[ 579.239507] dump_cpu_task+0xb0/0xc0
[ 579.243248] rcu_dump_cpu_stacks+0x220/0x238
[ 579.247702] rcu_sched_clock_irq+0x8a4/0x141c
[ 579.252249] update_process_times+0x34/0x64
[ 579.256617] tick_sched_handle+0x80/0x98
[ 579.260714] tick_sched_timer+0x64/0xbc
[ 579.264722] __hrtimer_run_queues+0x5c0/0xb84
[ 579.269266] hrtimer_interrupt+0x1ec/0x454
[ 579.273547] arch_timer_handler_phys+0x40/0x58
[ 579.278185] handle_percpu_devid_irq+0x174/0x6e8
[ 579.282999] generic_handle_irq+0x3c/0x54
[ 579.287185] __handle_domain_irq+0x114/0x118
[ 579.291639] gic_handle_irq+0x70/0xac
[ 579.295465] el1_irq+0xbc/0x180
[ 579.298756] __asan_load8+0x8c/0x9c
[ 579.302403] rcu_is_watching+0x80/0x8c
[ 579.306322] rebalance_domains+0x12c/0x584
[ 579.310599] run_rebalance_domains+0x1f4/0x298
[ 579.315231] __do_softirq+0x4c0/0xab8
[ 579.319061] irq_exit+0x148/0x1d8
[ 579.322530] __handle_domain_irq+0xc0/0x118
[ 579.326894] gic_handle_irq+0x70/0xac
[ 579.330720] el1_irq+0xbc/0x180
[ 579.334012] lock_is_held_type+0xec/0x144
[ 579.338201] rcu_read_lock_sched_held+0x90/0x98
[ 579.342927] kmem_cache_alloc+0x328/0x3e0
[ 579.347114] create_object+0x5c/0x39c
[ 579.350944] kmemleak_alloc+0x54/0x88
[ 579.354774] __kmalloc_track_caller+0x1c8/0x434
[ 579.359499] devres_alloc_node+0x40/0x8c
[ 579.363597] __devm_request_region+0x48/0xc8
[ 579.368055] devm_ioremap_resource+0xcc/0x148
[ 579.372626] rcar_gen3_thermal_probe+0x288/0x618 [rcar_gen3_thermal]
[ 579.379231] platform_drv_probe+0x70/0xe4
[ 579.383420] really_probe+0x2d8/0x3d8
[ 579.387249] driver_probe_device+0x154/0x164
[ 579.391705] device_driver_attach+0x98/0xa0
[ 579.396070] __driver_attach+0xf0/0xf4
[ 579.399987] bus_for_each_dev+0x114/0x13c
[ 579.404173] driver_attach+0x38/0x44
[ 579.407912] bus_add_driver+0x234/0x288
[ 579.411919] driver_register+0x148/0x190
[ 579.416015] __platform_driver_register+0x84/0x90
[ 579.420931] rcar_gen3_thermal_driver_init+0x28/0x1000 [rcar_gen3_thermal]
[ 579.428074] do_one_initcall+0x124/0x68c
[ 579.432173] do_init_module+0xb4/0x300
[ 579.436090] load_module+0x2c90/0x2f18
[ 579.440008] __se_sys_finit_module+0x128/0x148
[ 579.444642] __arm64_sys_finit_module+0x4c/0x5c
[ 579.449367] el0_svc_common+0xd0/0x16c
[ 579.453283] el0_svc_handler+0x94/0xa0
[ 579.457200] el0_svc+0x8/0xc
[ 582.713314] renesas_sdhi_internal_dmac ee100000.sd: timeout waiting for
hardware interrupt (CMD12)
[ 587.833305] renesas_sdhi_internal_dmac ee100000.sd: timeout waiting for
hardware interrupt (CMD12)
[ 592.953323] renesas_sdhi_internal_dmac ee100000.sd: timeout waiting for
hardware interrupt (CMD12)
[ 598.073430] renesas_sdhi_internal_dmac ee100000.sd: timeout waiting for
hardware interrupt (CMD12)
[ 603.193306] renesas_sdhi_internal_dmac ee100000.sd: timeout waiting for
hardware interrupt (CMD12)
[ 604.242120] watchdog: BUG: soft lockup - CPU#0 stuck for 22s! [modprobe:6337]
[..]
Best regards,
Eugeniu.
