linux-kernel - Re: [PATCH v5 4/4] thermal: Add Tegra SOCTHERM thermal management driver

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:	Thu, 25 Sep 2014 07:59:55 +0200
From:	Thierry Reding <thierry.reding@...il.com>
To:	Mikko Perttunen <mikko.perttunen@...si.fi>
Cc:	Eduardo Valentin <edubezval@...il.com>,
	Mikko Perttunen <mperttunen@...dia.com>, rui.zhang@...el.com,
	swarren@...dotorg.org, linux-pm@...r.kernel.org,
	linux-tegra@...r.kernel.org, linux-kernel@...r.kernel.org,
	linux-arm-kernel@...ts.infradead.org, juha-matti.tilli@....fi
Subject: Re: [PATCH v5 4/4] thermal: Add Tegra SOCTHERM thermal management
 driver

On Wed, Sep 24, 2014 at 10:32:13PM +0300, Mikko Perttunen wrote:
> On 09/24/2014 10:18 PM, Eduardo Valentin wrote:
> >
> >Mikko,
> >
> >On Thu, Aug 21, 2014 at 01:17:22PM +0300, Mikko Perttunen wrote:
> >>...
> >>+
> >>+static int enable_tsensor(struct tegra_soctherm *tegra,
> >>+			  const struct tegra_tsensor *sensor,
> >>+			  struct tsensor_shared_calibration shared)
> >>+{
> >>+	void * __iomem base = tegra->regs + sensor->base;
> >
> >I get sparse complaining about this declaration. For the sake of keeping
> >a clean static checks, can you please:
> >-       void * __iomem base = tegra->regs + sensor->base;
> >+       void __iomem * base = tegra->regs + sensor->base;
> >
> 
> Sure.
> 
> >Can you also please check the remaining sparse errors?
> >   CHECK   drivers/thermal/tegra_soctherm.c
> >   drivers/thermal/tegra_soctherm.c:260:43: warning: incorrect type in
> >   initializer (different address spaces)
> >   drivers/thermal/tegra_soctherm.c:260:43:    expected void *[noderef]
> >   <asn:2>base
> >   drivers/thermal/tegra_soctherm.c:260:43:    got void [noderef]
> >   <asn:2>*
> >   drivers/thermal/tegra_soctherm.c:271:9: warning: incorrect type in
> >   argument 2 (different address spaces)
> >   drivers/thermal/tegra_soctherm.c:271:9:    expected void volatile
> >   [noderef] <asn:2>*addr
> >   drivers/thermal/tegra_soctherm.c:271:9:    got void *
> >   drivers/thermal/tegra_soctherm.c:279:9: warning: incorrect type in
> >   argument 2 (different address spaces)
> >   drivers/thermal/tegra_soctherm.c:279:9:    expected void volatile
> >   [noderef] <asn:2>*addr
> >   drivers/thermal/tegra_soctherm.c:279:9:    got void *
> >   drivers/thermal/tegra_soctherm.c:281:9: warning: incorrect type in
> >   argument 2 (different address spaces)
> >   drivers/thermal/tegra_soctherm.c:281:9:    expected void volatile
> >   [noderef] <asn:2>*addr
> >   drivers/thermal/tegra_soctherm.c:281:9:    got void *
> >   drivers/thermal/tegra_soctherm.c:347:25: warning: incorrect type in
> >   argument 1 (different address spaces)
> >   drivers/thermal/tegra_soctherm.c:347:25:    expected void const *ptr
> >   drivers/thermal/tegra_soctherm.c:347:25:    got void [noderef]
> >   <asn:2>*regs
> >   drivers/thermal/tegra_soctherm.c:349:37: warning: incorrect type in
> >   argument 1 (different address spaces)
> >   drivers/thermal/tegra_soctherm.c:349:37:    expected void const *ptr
> >   drivers/thermal/tegra_soctherm.c:349:37:    got void [noderef]
> >   <asn:2>*regs
> >   drivers/thermal/tegra_soctherm.c:271:9: warning: dereference of
> >   noderef expression
> >   drivers/thermal/tegra_soctherm.c:279:9: warning: dereference of
> >   noderef expression
> >   drivers/thermal/tegra_soctherm.c:281:9: warning: dereference of
> >   noderef expression
> >
> 
> Most of these seem to be caused by the above-mentioned swapping of __iomem
> and *. The ones on lines 347 and 349 are more peculiar, though. Apparently
> sparse doesn't like using IS_ERR and PTR_ERR on the void __iomem * pointer
> returned by devm_ioremap_resource. Looks like this has been discussed before
> (https://lkml.org/lkml/2013/6/17/216) and sparse should have been patched to
> ignore this situation, so I'm not sure why it's complaining about it.
> Anyway, there shouldn't be any issue here.

If I apply the change suggested by Eduardo above then all of the above
warnings go away for me. That's with sparse 0.5.0. According to the
email thread above a patch for this was applied in 0.4.5.

Thierry

Content of type "application/pgp-signature" skipped