Re: [PATCH] clocksource: Add driver for the Ingenic JZ47xx OST

[Daniel Lezcano]

On 07/11/2019 21:34, Paul Cercueil wrote:
> 
> 
> Le jeu., nov. 7, 2019 at 21:22, Daniel Lezcano
> <daniel.lezcano@xxxxxxxxxx> a Ãcrit :
>> On 07/11/2019 20:57, Paul Cercueil wrote:
>>>
>>>
>>> ÂLe jeu., nov. 7, 2019 at 20:39, Daniel Lezcano
>>> Â<daniel.lezcano@xxxxxxxxxx> a Ãcrit :
>>>> ÂOn 07/11/2019 16:56, Paul Cercueil wrote:
>>>>> Â Hi Daniel,
>>>>>
>>>>> Â On 2019-08-16 16:54, Daniel Lezcano wrote:
>>>>>> Â On 09/08/2019 14:38, Paul Cercueil wrote:
>>>>>>> Â From: Maarten ter Huurne <maarten@xxxxxxxxxxxxxx>
>>>>>>>
>>>>>>> Â OST is the OS Timer, a 64-bit timer/counter with buffered reading.
>>>>>>>
>>>>>>> Â SoCs before the JZ4770 had (if any) a 32-bit OST; the JZ4770 and
>>>>>>> Â JZ4780 have a 64-bit OST.
>>>>>>>
>>>>>>> Â This driver will register both a clocksource and a sched_clock
>>>>>>> to the
>>>>>>> Â system.
>>>>>>>
>>>>>>> Â Signed-off-by: Maarten ter Huurne <maarten@xxxxxxxxxxxxxx>
>>>>>>> Â Signed-off-by: Paul Cercueil <paul@xxxxxxxxxxxxxxx>
>>>>>>> Â Tested-by: Mathieu Malaterre <malat@xxxxxxxxxx>
>>>>>>> Â Tested-by: Artur Rojek <contact@xxxxxxxxxxxxxx>
>>>>>>> Â ---
>>>>>>>
>>>>>>
>>>>>> Â [ ... ]
>>>>>>
>>>>>>> Â +ÂÂÂ err = clocksource_register_hz(cs, rate);
>>>>>>> Â +ÂÂÂ if (err) {
>>>>>>> Â +ÂÂÂÂÂÂÂ dev_err(dev, "clocksource registration failed: %d\n",
>>>>>>> err);
>>>>>>> Â +ÂÂÂÂÂÂÂ clk_disable_unprepare(ost->clk);
>>>>>>> Â +ÂÂÂÂÂÂÂ return err;
>>>>>>> Â +ÂÂÂ }
>>>>>>> Â +
>>>>>>> Â +ÂÂÂ /* Cannot register a sched_clock with interrupts on */
>>>>>>
>>>>>> Â Aren't they already disabled?
>>>>>
>>>>> Â Sorry for the late reply.
>>>>>
>>>>> Â No, they are not already disabled; this is what I get if I
>>>>> comment out
>>>>> Â the local_irq_save()/local_irq_restore():
>>>>>
>>>>> Â [ÂÂÂ 0.361014] clocksource: ingenic-ost: mask: 0xffffffff
>>>>> max_cycles:
>>>>> Â 0xffffffff, max_idle_ns: 159271703898 ns
>>>>> Â [ÂÂÂ 0.361515] clocksource: Switched to clocksource ingenic-ost
>>>>> Â [ÂÂÂ 0.361686] ------------[ cut here ]------------
>>>>> Â [ÂÂÂ 0.361893] WARNING: CPU: 0 PID: 1 at
>>>>> kernel/time/sched_clock.c:179
>>>>> Â sched_clock_register+0x7c/0x2e4
>>>>> Â [ÂÂÂ 0.362174] CPU: 0 PID: 1 Comm: swapper Not tainted 5.4.0-rc5+
>>>>> #461
>>>>> Â [ÂÂÂ 0.362330] Stack : 80744558 80069b44 80770000 00000000 00000000
>>>>> Â 00dfd7a7 806e6db4 8106bb74
>>>>> Â [ÂÂÂ 0.362619]ÂÂÂÂÂÂÂÂ 806f0000 81067ca4 806f31c7 80769478 00000020
>>>>> Â 10000400 8106bb20 00dfd7a7
>>>>> Â [ÂÂÂ 0.362906]ÂÂÂÂÂÂÂÂ 00000000 00000000 80780000 00000000 00000007
>>>>> Â 00000001 00000049 3563722d
>>>>> Â [ÂÂÂ 0.363191]ÂÂÂÂÂÂÂÂ 8106ba61 00000000 ffffffff 00000010 806f0000
>>>>> Â 00000000 00000000 806f0000
>>>>> Â [ÂÂÂ 0.363477]ÂÂÂÂÂÂÂÂ 00000020 00000000 80714534 80770000 00000002
>>>>> Â 80319154 00000000 80770000
>>>>> Â [ÂÂÂ 0.363762]ÂÂÂÂÂÂÂÂ ...
>>>>> Â [ÂÂÂ 0.363906] Call Trace:
>>>>> Â [ÂÂÂ 0.364087] [<8001af14>] show_stack+0x40/0x128
>>>>> Â [ÂÂÂ 0.364289] [<8002fd88>] __warn+0xb8/0xe0
>>>>> Â [ÂÂÂ 0.364478] [<8002fe14>] warn_slowpath_fmt+0x64/0xc0
>>>>> Â [ÂÂÂ 0.364678] [<8072b1c8>] sched_clock_register+0x7c/0x2e4
>>>>> Â [ÂÂÂ 0.364895] [<8073c874>] ingenic_ost_probe+0x224/0x248
>>>>> Â [ÂÂÂ 0.365090] [<803d5394>] platform_drv_probe+0x40/0x94
>>>>> Â [ÂÂÂ 0.365526] [<803d362c>] really_probe+0x104/0x374
>>>>> Â [ÂÂÂ 0.365743] [<803d3ff0>] device_driver_attach+0x78/0x80
>>>>> Â [ÂÂÂ 0.365938] [<803d4070>] __driver_attach+0x78/0x118
>>>>> Â [ÂÂÂ 0.366129] [<803d1700>] bus_for_each_dev+0x7c/0xc8
>>>>> Â [ÂÂÂ 0.366318] [<803d226c>] bus_add_driver+0x1bc/0x204
>>>>> Â [ÂÂÂ 0.366513] [<803d4878>] driver_register+0x84/0x14c
>>>>> Â [ÂÂÂ 0.366717] [<8073a144>] __platform_driver_probe+0x98/0x140
>>>>> Â [ÂÂÂ 0.366931] [<80724e38>] do_one_initcall+0x84/0x1b4
>>>>> Â [ÂÂÂ 0.367126] [<807250cc>] kernel_init_freeable+0x164/0x240
>>>>> Â [ÂÂÂ 0.367318] [<805df75c>] kernel_init+0x10/0xf8
>>>>> Â [ÂÂÂ 0.367510] [<8001542c>] ret_from_kernel_thread+0x14/0x1c
>>>>> Â [ÂÂÂ 0.367722] ---[ end trace 7fedf00408fa3bed ]---
>>>>> Â [ÂÂÂ 0.367985] sched_clock: 32 bits at 12MHz, resolution 83ns, wraps
>>>>> Â every 178956970966ns
>>>>>
>>>>> Â At kernel/time/sched_clock.c:179 there is:
>>>>> Â WARN_ON(!irqs_disabled());
>>>>
>>>> ÂThat is strange, no drivers is doing that and no warning is appearing.
>>>>
>>>> ÂIsn't missing a local_irq_disable in the code path in the stack above?
>>>
>>> ÂI think it comes to the fact that the other drivers are probed much
>>> Âearlier in the boot process, while this one is probed as a regular
>>> Âplatform device driver.
>>
>>
>> There are other drivers doing and nobody is disabling the interrupt:
>>
>> em_sti.c:static struct platform_driver em_sti_device_driver = {
>> ingenic-timer.c:static struct platform_driver ingenic_tcu_driver = {
>> sh_cmt.c:static struct platform_driver sh_cmt_device_driver = {
>> sh_mtu2.c:static struct platform_driver sh_mtu2_device_driver = {
>> sh_tmu.c:static struct platform_driver sh_tmu_device_driver = {
>> timer-ti-dm.c:static struct platform_driver omap_dm_timer_driver = {
> 
> Yes, but they don't register a sched_clock at all (except for
> ingenic-timer, but it does probe using TIMER_OF_DECLARE), and this
> warning is in sched_clock_register().
> 
>>
>>>>>>> Â +ÂÂÂ local_irq_save(flags);
>>>>>>> Â +ÂÂÂ if (soc_info->is64bit)
>>>>>>> Â +ÂÂÂÂÂÂÂ sched_clock_register(ingenic_ost_read_cntl, 32, rate);
>>>>>>> Â +ÂÂÂ else
>>>>>>> Â +ÂÂÂÂÂÂÂ sched_clock_register(ingenic_ost_read_cnth, 32, rate);
>>>>>>> Â +ÂÂÂ local_irq_restore(flags);
>>>>>>> Â +
>>>>>>> Â +ÂÂÂ return 0;
>>>>>>> Â +}
>>>>>>> Â +

Thomas,

the local_irq_save/restore calls shouldn't be in the
sched_clock_register() function instead of showing a warning?


-- 
 <http://www.linaro.org/> Linaro.org â Open source software for ARM SoCs

Follow Linaro:  <http://www.facebook.com/pages/Linaro> Facebook |
<http://twitter.com/#!/linaroorg> Twitter |
<http://www.linaro.org/linaro-blog/> Blog