linux-kernel - Re: [PATCH 2/2] soc: bcm: bcm2835-pm: Fix error paths of initialization.

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] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 15 Feb 2019 11:29:52 -0800
From:   Florian Fainelli <f.fainelli@...il.com>
To:     Stefan Wahren <stefan.wahren@...e.com>,
        Eric Anholt <eric@...olt.net>,
        Florian Fainelli <f.fainelli@...il.com>
Cc:     linux-rpi-kernel@...ts.infradead.org,
        linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org,
        bcm-kernel-feedback-list@...adcom.com
Subject: Re: [PATCH 2/2] soc: bcm: bcm2835-pm: Fix error paths of
 initialization.

On 2/13/19 2:33 PM, Stefan Wahren wrote:
> 
>> Eric Anholt <eric@...olt.net> hat am 13. Februar 2019 um 19:28 geschrieben:
>>
>>
>> Stefan Wahren <stefan.wahren@...e.com> writes:
>>
>>> Hi Eric,
>>>
>>> Am 13.02.19 um 01:33 schrieb Eric Anholt:
>>>> The clock driver may probe after ours and so we need to pass the
>>>> -EPROBE_DEFER out.  Fix the other error path while we're here.
>>>>
>>>> Signed-off-by: Eric Anholt <eric@...olt.net>
>>>> Fixes: 670c672608a1 ("soc: bcm: bcm2835-pm: Add support for power domains under a new binding.")
>>>> ---
>>>>  drivers/soc/bcm/bcm2835-power.c | 30 +++++++++++++++++++++++++-----
>>>>  1 file changed, 25 insertions(+), 5 deletions(-)
>>>>
>>>> diff --git a/drivers/soc/bcm/bcm2835-power.c b/drivers/soc/bcm/bcm2835-power.c
>>>> index 4a1b99b773c0..11f9469423f7 100644
>>>> --- a/drivers/soc/bcm/bcm2835-power.c
>>>> +++ b/drivers/soc/bcm/bcm2835-power.c
>>>> @@ -485,7 +485,7 @@ static int bcm2835_power_pd_power_off(struct generic_pm_domain *domain)
>>>>  	}
>>>>  }
>>>>  
>>>> -static void
>>>> +static int
>>>>  bcm2835_init_power_domain(struct bcm2835_power *power,
>>>>  			  int pd_xlate_index, const char *name)
>>>>  {
>>>> @@ -493,6 +493,12 @@ bcm2835_init_power_domain(struct bcm2835_power *power,
>>>>  	struct bcm2835_power_domain *dom = &power->domains[pd_xlate_index];
>>>>  
>>>>  	dom->clk = devm_clk_get(dev->parent, name);
>>>> +	if (IS_ERR(dom->clk)) {
>>>> +		int ret = PTR_ERR(dom->clk);
>>>> +
>>>> +		if (ret == -EPROBE_DEFER)
>>>> +			return ret;
>>> is it safe to proceed in the other error cases?
>>> Even it would be more consistent with clk_prepare_enable() to print an
>>> error here.
>>
>> Yes, not all domains have a clk, so we want to ignore the other error.
> 
> But shouldn't we set dom->clk to NULL instead of keeping the error pointer? AFAIK clk_prepare_enable is aware of NULL instead of error pointer.

If the clock is really optional, then yes, this should be the way to go.
-- 
Florian