linux-kernel - Re: [PATCH] soc: rockchip: power-domain: Handle errors from of_genpd_add_provider

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <4f94a0bf-921c-cacb-a035-8111a969d156@collabora.com>
Date:   Fri, 16 Sep 2016 07:59:02 +0200
From:   Tomeu Vizoso <tomeu.vizoso@...labora.com>
To:     Heiko Stübner <heiko@...ech.de>
Cc:     linux-kernel@...r.kernel.org, Caesar Wang <wxt@...k-chips.com>,
        linux-arm-kernel@...ts.infradead.org,
        linux-rockchip@...ts.infradead.org
Subject: Re: [PATCH] soc: rockchip: power-domain: Handle errors from
 of_genpd_add_provider_onecell

On 09/16/2016 12:31 AM, Heiko Stübner wrote:
> Am Donnerstag, 15. September 2016, 16:39:34 schrieb Heiko Stübner:
>> Am Donnerstag, 15. September 2016, 12:43:41 schrieb Tomeu Vizoso:
>>> It was a bit surprising that the device was reported to have probed just
>>> fine, but the provider hadn't been registered.
>>>
>>> So handle any errors when registering the provider and fail the probe
>>> accordingly.
>>>
>>> Signed-off-by: Tomeu Vizoso <tomeu.vizoso@...labora.com>
>>> Cc: Caesar Wang <wxt@...k-chips.com>
>>> ---
>>>
>>>  drivers/soc/rockchip/pm_domains.c | 6 +++++-
>>>  1 file changed, 5 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/drivers/soc/rockchip/pm_domains.c
>>> b/drivers/soc/rockchip/pm_domains.c index 7acd1517dd37..57e920128cb2
>>> 100644
>>> --- a/drivers/soc/rockchip/pm_domains.c
>>> +++ b/drivers/soc/rockchip/pm_domains.c
>>> @@ -627,7 +627,11 @@ static int rockchip_pm_domain_probe(struct
>>> platform_device *pdev) goto err_out;
>>>
>>>  	}
>>>
>>> -	of_genpd_add_provider_onecell(np, &pmu->genpd_data);
>>> +	error = of_genpd_add_provider_onecell(np, &pmu->genpd_data);
>>> +	if (error) {
>>> +		dev_err(dev, "failed to add provider: %d\n", error);
>>> +		goto err_out;
>>> +	}
>>>
>>>  	return 0;
>>
>> Looks good in itself, but seems to trigger some issue in the genpd code
>> when applied alone on top of linux-next-20160915. Looks like genpd
>> is missing counter-initialization somewhere, as I'm seeing now:
>>
>> [    1.664744] genpd_poweroff_unused disabling (null)
>> [    1.669553] ------------[ cut here ]------------
>> [    1.674169] WARNING: CPU: 0 PID: 1 at ../kernel/workqueue.c:1440
>> __queue_work+0x2b8/0x3f8 [    1.682337] Modules linked in:
>> [    1.685401] CPU: 0 PID: 1 Comm: swapper/0 Not tainted
>> 4.8.0-rc6-next-20160915-00001-g7432710-dirty #5 [    1.694608] Hardware
>> name: Rockchip (Device Tree)
>> [    1.699312] [<c0310388>] (unwind_backtrace) from [<c030ba04>]
>> (show_stack+0x10/0x14) [    1.707050] [<c030ba04>] (show_stack) from
>> [<c0599c90>] (dump_stack+0x90/0xa4) [    1.714267] [<c0599c90>]
>> (dump_stack) from [<c0341bc0>] (__warn+0xe8/0x100) [    1.721221]
>> [<c0341bc0>] (__warn) from [<c0341c88>] (warn_slowpath_null+0x20/0x28) [   
>> 1.728785] [<c0341c88>] (warn_slowpath_null) from [<c0355b28>]
>> (__queue_work+0x2b8/0x3f8) [    1.737041] [<c0355b28>] (__queue_work) from
>> [<c0355ca8>] (queue_work_on+0x40/0x4c) [    1.744692] [<c0355ca8>]
>> (queue_work_on) from [<c103e8b8>] (genpd_poweroff_unused+0x78/0x9c) [   
>> 1.753123] [<c103e8b8>] (genpd_poweroff_unused) from [<c0301e78>]
>> (do_one_initcall+0x40/0x170) [    1.761815] [<c0301e78>] (do_one_initcall)
>> from [<c1000dc8>] (kernel_init_freeable+0x15c/0x1fc) [    1.770507]
>> [<c1000dc8>] (kernel_init_freeable) from [<c0bb7ba0>]
>> (kernel_init+0x8/0x114) [    1.778678] [<c0bb7ba0>] (kernel_init) from
>> [<c0307df8>] (ret_from_fork+0x14/0x3c) [    1.786240] ---[ end trace
>> b38c51ace1463add ]---
>> [    1.790875] genpd_poweroff_unused disabling ������
>> [    1.795856] genpd_poweroff_unused disabling ������
>> [    1.800830] ------------[ cut here ]------------
>> [    1.805443] WARNING: CPU: 0 PID: 1 at ../kernel/workqueue.c:1440
>> __queue_work+0x2b8/0x3f8 [    1.813610] Modules linked in:
>> [    1.816673] CPU: 0 PID: 1 Comm: swapper/0 Tainted: G        W      
>> 4.8.0-rc6-next-20160915-00001-g7432710-dirty #5 [    1.827094] Hardware
>> name: Rockchip (Device Tree)
>> [    1.831795] [<c0310388>] (unwind_backtrace) from [<c030ba04>]
>> (show_stack+0x10/0x14) [    1.839532] [<c030ba04>] (show_stack) from
>> [<c0599c90>] (dump_stack+0x90/0xa4) [    1.846747] [<c0599c90>]
>> (dump_stack) from [<c0341bc0>] (__warn+0xe8/0x100) [    1.853700]
>> [<c0341bc0>] (__warn) from [<c0341c88>] (warn_slowpath_null+0x20/0x28) [   
>> 1.861264] [<c0341c88>] (warn_slowpath_null) from [<c0355b28>]
>> (__queue_work+0x2b8/0x3f8) [    1.869521] [<c0355b28>] (__queue_work) from
>> [<c0355ca8>] (queue_work_on+0x40/0x4c) [    1.877170] [<c0355ca8>]
>> (queue_work_on) from [<c103e8b8>] (genpd_poweroff_unused+0x78/0x9c) [   
>> 1.885600] [<c103e8b8>] (genpd_poweroff_unused) from [<c0301e78>]
>> (do_one_initcall+0x40/0x170) [    1.894290] [<c0301e78>] (do_one_initcall)
>> from [<c1000dc8>] (kernel_init_freeable+0x15c/0x1fc) [    1.902981]
>> [<c1000dc8>] (kernel_init_freeable) from [<c0bb7ba0>]
>> (kernel_init+0x8/0x114) [    1.911152] [<c0bb7ba0>] (kernel_init) from
>> [<c0307df8>] (ret_from_fork+0x14/0x3c) [    1.918713] ---[ end trace
>> b38c51ace1463ade ]---
>> [    1.923338] genpd_poweroff_unused disabling ������
>> [    1.928325] genpd_poweroff_unused disabling ������
>>
>> [+ millions more of those]
> 
> just for completenes sake, I've included your patch in a series of my
> own [0] which adds the necessary  pm_genpd_remove prequisite to prevent
> the errors shown above.

That's awesome, thanks. For some reason I hadn't noticed those.

Regards,

Tomeu

> 
> Heiko
> 
> [0] "[PATCH 0/2] soc: rockchip: fix probe error path in power-domain driver"
>