linux-kernel - Re: [PATCH] regulator: Use IRQF

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: <20210325210543.000001d4@163.com>
Date:   Thu, 25 Mar 2021 21:05:43 +0800
From:   Jian Dong <dj0227@....com>
To:     Felipe Balbi <balbi@...nel.org>
Cc:     Krzysztof Kozlowski <krzysztof.kozlowski@...onical.com>,
        matthias.bgg@...il.com, agross@...nel.org,
        bjorn.andersson@...aro.org, linux-kernel@...r.kernel.org,
        linux-arm-kernel@...ts.infradead.org,
        linux-mediatek@...ts.infradead.org, linux-arm-msm@...r.kernel.org,
        huyue2@...ong.com, Jian Dong <dongjian@...ong.com>
Subject: Re: [PATCH] regulator: Use IRQF_ONESHOT

On Thu, 25 Mar 2021 14:29:02 +0200
Felipe Balbi <balbi@...nel.org> wrote:

> Hi,
> 
> Krzysztof Kozlowski <krzysztof.kozlowski@...onical.com> writes:
> > On 23/03/2021 13:12, Jian Dong wrote:  
> >> From: Jian Dong <dongjian@...ong.com>
> >> 
> >> Fixes coccicheck error:
> >> 
> >> drivers/regulator/mt6360-regulator.c:388:8-33: ERROR:
> >> drivers/regulator/pca9450-regulator.c:781:7-32: ERROR:
> >> drivers/regulator/slg51000-regulator.c:480:8-33: ERROR:
> >> drivers/regulator/qcom-labibb-regulator.c:364:8-33: ERROR:
> >> Threaded IRQ with no primary handler requested without IRQF_ONESHOT
> >> 
> >> Signed-off-by: Jian Dong <dongjian@...ong.com>
> >> ---
> >>  drivers/regulator/mt6360-regulator.c      | 4 ++--
> >>  drivers/regulator/pca9450-regulator.c     | 2 +-
> >>  drivers/regulator/qcom-labibb-regulator.c | 3 ++-
> >>  drivers/regulator/slg51000-regulator.c    | 4 ++--
> >>  4 files changed, 7 insertions(+), 6 deletions(-)
> >> 
> >> diff --git a/drivers/regulator/mt6360-regulator.c
> >> b/drivers/regulator/mt6360-regulator.c index 15308ee..947350d
> >> 100644 --- a/drivers/regulator/mt6360-regulator.c
> >> +++ b/drivers/regulator/mt6360-regulator.c
> >> @@ -385,8 +385,8 @@ static int
> >> mt6360_regulator_irq_register(struct platform_device *pdev, return
> >> irq; }
> >>  
> >> -		ret = devm_request_threaded_irq(&pdev->dev, irq,
> >> NULL, irq_desc->handler, 0,
> >> -						irq_desc->name,
> >> rdev);
> >> +		ret = devm_request_threaded_irq(&pdev->dev, irq,
> >> NULL, irq_desc->handler,
> >> +					IRQF_ONESHOT,
> >> irq_desc->name, rdev);  
> >
> > This does not look like trivial rename/replace fix. This should be
> > tested but it looks that you just did what coccinelle asked for,
> > without testing.  
> 
> Right, but it must be done. If things work today, they work out of
> sheer luck. Also, which evidence is there that $subject wasn't tested?
> 
it just use coccinelle to check again, the warning didn't rise after
modify

> >> diff --git a/drivers/regulator/pca9450-regulator.c
> >> b/drivers/regulator/pca9450-regulator.c index 2f7ee21..d4bc1c3
> >> 100644 --- a/drivers/regulator/pca9450-regulator.c
> >> +++ b/drivers/regulator/pca9450-regulator.c
> >> @@ -780,7 +780,7 @@ static int pca9450_i2c_probe(struct i2c_client
> >> *i2c, 
> >>  	ret = devm_request_threaded_irq(pca9450->dev,
> >> pca9450->irq, NULL, pca9450_irq_handler,
> >> -					(IRQF_TRIGGER_FALLING |
> >> IRQF_ONESHOT),
> >> +					IRQF_TRIGGER_FALLING |
> >> IRQF_ONESHOT,  
> >
> > How this is related to the missing IRQF_ONESHOT?  
> 
> agreed.
> 
> >> diff --git a/drivers/regulator/slg51000-regulator.c
> >> b/drivers/regulator/slg51000-regulator.c index 75a941f..3f310ab
> >> 100644 --- a/drivers/regulator/slg51000-regulator.c
> >> +++ b/drivers/regulator/slg51000-regulator.c
> >> @@ -479,8 +479,8 @@ static int slg51000_i2c_probe(struct
> >> i2c_client *client) if (chip->chip_irq) {
> >>  		ret = devm_request_threaded_irq(dev,
> >> chip->chip_irq, NULL, slg51000_irq_handler,
> >> -
> >> (IRQF_TRIGGER_HIGH |
> >> -						IRQF_ONESHOT),
> >> +						IRQF_TRIGGER_HIGH
> >> |
> >> +						IRQF_ONESHOT,
> >>  						"slg51000-irq",
> >> chip);  
> >
> > How this is related to the missing IRQF_ONESHOT?  
> 
> agreed.
> 
I thnik it maybe the result of the coccinelle rule. it need to check 
value explicit. another wanring like this:  define irq_flag
= IRQF_ONESHOT, then reference irq_flag in fuction, it also will trigger
coccinelle warning. it seems not the code error but coccicheck bug.