Hi,

On Wed, 13 Jan 2021 at 03:21, Marek Szyprowski m.szyprowski@samsung.com wrote:

Hi,

<trim>

This patch landed in linux next-20210112 as commit eaa7995c529b ("regulator: core: avoid regulator_resolve_supply() race condition"). I found that it triggers a following lockdep warning during the DWC3 driver registration on some Exynos based boards (this log is from Samsung Exynos5420-based Peach-Pit board):

====================================================== WARNING: possible circular locking dependency detected 5.11.0-rc1-00008-geaa7995c529b #10095 Not tainted

---

swapper/0/1 is trying to acquire lock: c12e1b80 (regulator_list_mutex){+.+.}-{3:3}, at: regulator_lock_dependent+0x4c/0x2b0

but task is already holding lock: df7190c0 (regulator_ww_class_mutex){+.+.}-{3:3}, at: regulator_resolve_supply+0x44/0x318

LKFT testing also found this lockdep warning on arm64 - hi6220-hikey while booting.

[ 0.635532] WARNING: possible recursive locking detected [ 0.635558] 5.11.0-rc3-next-20210118 #1 Not tainted [ 0.635585] -------------------------------------------- [ 0.635611] swapper/0/1 is trying to acquire lock: [ 0.635636] ffff000000a13158 (regulator_ww_class_mutex){+.+.}-{3:3}, at: regulator_lock_recursive+0x9c/0x1e8 [ 0.635721] [ 0.635721] but task is already holding lock: [ 0.635749] ffff000000a13958 (regulator_ww_class_mutex){+.+.}-{3:3}, at: regulator_resolve_supply+0x70/0x2f0 [ 0.635817] [ 0.635817] other info that might help us debug this: [ 0.635847] Possible unsafe locking scenario: [ 0.635847] [ 0.635875] CPU0 [ 0.635892] ---- [ 0.635909] lock(regulator_ww_class_mutex); [ 0.635942] lock(regulator_ww_class_mutex); [ 0.635974] [ 0.635974] *** DEADLOCK *** [ 0.635974] [ 0.636002] May be due to missing lock nesting notation [ 0.636002] [ 0.636033] 4 locks held by swapper/0/1: [ 0.636057] #0: ffff000000a02988 (&dev->mutex){....}-{3:3}, at: __device_driver_lock+0x38/0x70 [ 0.636131] #1: ffff000000a13958 (regulator_ww_class_mutex){+.+.}-{3:3}, at: regulator_resolve_supply+0x70/0x2f0 [ 0.636205] #2: ffff800012b102c0 (regulator_list_mutex){+.+.}-{3:3}, at: regulator_lock_dependent+0x5c/0x290 [ 0.636280] #3: ffff8000137e3918 (regulator_ww_class_acquire){+.+.}-{0:0}, at: regulator_enable+0x40/0xe0 [ 0.636352] [ 0.636352] stack backtrace: [ 0.636378] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.11.0-rc3-next-20210118 #1 [ 0.636415] Hardware name: HiKey Development Board (DT) [ 0.636443] Call trace: [ 0.636460] dump_backtrace+0x0/0x1f0 [ 0.636490] show_stack+0x2c/0x80 [ 0.636516] dump_stack+0xf8/0x160 [ 0.636543] __lock_acquire+0xa3c/0x1718 [ 0.636571] lock_acquire+0x3d8/0x4f0 [ 0.636596] __ww_mutex_lock.constprop.14+0xbc/0xf68 [ 0.636628] ww_mutex_lock+0x6c/0x3e8 [ 0.636653] regulator_lock_recursive+0x9c/0x1e8 [ 0.636683] regulator_lock_dependent+0x198/0x290 [ 0.636713] regulator_enable+0x40/0xe0 [ 0.636739] regulator_resolve_supply+0x1e8/0x2f0 [ 0.636767] regulator_register_resolve_supply+0x24/0x80 [ 0.636797] class_for_each_device+0x78/0xf8 [ 0.636825] regulator_register+0x840/0xbb0 [ 0.636851] devm_regulator_register+0x50/0xa8 [ 0.636879] reg_fixed_voltage_probe+0x224/0x410 [ 0.636908] platform_probe+0x6c/0xd8 [ 0.636932] really_probe+0x2b8/0x520 [ 0.636960] driver_probe_device+0xf4/0x168 [ 0.636988] device_driver_attach+0x74/0x98 [ 0.637014] __driver_attach+0xc4/0x178 [ 0.637039] bus_for_each_dev+0x84/0xd8 [ 0.637066] driver_attach+0x30/0x40 [ 0.637092] bus_add_driver+0x170/0x258 [ 0.637119] driver_register+0x64/0x118 [ 0.637144] __platform_driver_register+0x34/0x40 [ 0.637172] regulator_fixed_voltage_init+0x20/0x28 [ 0.637205] do_one_initcall+0x94/0x4a0 [ 0.637231] kernel_init_freeable+0x2f0/0x344 [ 0.637261] kernel_init+0x18/0x120

Reported-by: Naresh Kamboju naresh.kamboju@linaro.org

Full boot log here: https://qa-reports.linaro.org/lkft/linux-next-master/build/next-20210118/tes...

metadata: git branch: master git repo: https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git git describe: next-20210112 kernel-config: http://snapshots.linaro.org/openembedded/lkft/lkft/sumo/hikey/lkft/linux-nex...