Re: [PATCH v2 4/7] remoteproc: Use kstrdup_const() rather than kstrup()

From: Bjorn Andersson <bjorn.andersson@linaro.org>
To: Suman Anna <s-anna@ti.com>
Cc: Alex Elder <elder@linaro.org>,
	Mathieu Poirier <mathieu.poirier@linaro.org>,
	ohad@wizery.com, Markus.Elfring@web.de,
	linux-remoteproc@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2 4/7] remoteproc: Use kstrdup_const() rather than kstrup()
Date: Sun, 19 Apr 2020 22:21:50 -0700	[thread overview]
Message-ID: <20200420052147.GF1516868@builder.lan> (raw)
In-Reply-To: <3f0a602a-64ff-8092-3cca-b63bab8a79f0@ti.com>

On Fri 17 Apr 06:44 PDT 2020, Suman Anna wrote:

> On 4/15/20 4:25 PM, Alex Elder wrote:
> > On 4/15/20 3:48 PM, Mathieu Poirier wrote:
> > > For cases where @firmware is declared "const char *", use function
> > > kstrdup_const() to avoid needlessly creating another copy on the
> > > heap.
> > > 
> > > Suggested-by: Bjorn Andersson <bjorn.andersson@linaro.org>
> > > Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
> > 
> > Looks good.
> > 
> > Reviewed-by: Alex Elder <elder@linaro.org>
> > 
> > > ---
> > >   drivers/remoteproc/remoteproc_core.c | 4 ++--
> > >   include/linux/remoteproc.h           | 2 +-
> > >   2 files changed, 3 insertions(+), 3 deletions(-)
> > > 
> > > diff --git a/drivers/remoteproc/remoteproc_core.c b/drivers/remoteproc/remoteproc_core.c
> > > index 9899467fa1cf..ebaff496ef81 100644
> > > --- a/drivers/remoteproc/remoteproc_core.c
> > > +++ b/drivers/remoteproc/remoteproc_core.c
> > > @@ -1982,7 +1982,7 @@ static const struct device_type rproc_type = {
> > >   static int rproc_alloc_firmware(struct rproc *rproc,
> > >   				const char *name, const char *firmware)
> > >   {
> > > -	char *p;
> > > +	const char *p;
> > >   	if (!firmware)
> > >   		/*
> > > @@ -1991,7 +1991,7 @@ static int rproc_alloc_firmware(struct rproc *rproc,
> > >   		 */
> > >   		p = kasprintf(GFP_KERNEL, "rproc-%s-fw", name);
> 
> So, to be consistent for both paths, should we be using kvasprintf_const()
> here and kfree_const() in release.

Given that the second argument is a "proper" format string
kvasprintf_const() is really just kasprintf() - but with the requirement
that we set up a va_list. So I prefer that we stick with this.

> The kfree_const() is needed to account
> for the kstrdup_const below for sure.
> 

You are correct Suman, this patch needs to also change the kfree() to a
kfree_const() or bad things will happen after a visit to
rproc_type_release().

Regards,
Bjorn

> regards
> Suman
> 
> > >   	else
> > > -		p = kstrdup(firmware, GFP_KERNEL);
> > > +		p = kstrdup_const(firmware, GFP_KERNEL);
> > >   	if (!p)
> > >   		return -ENOMEM;
> > > diff --git a/include/linux/remoteproc.h b/include/linux/remoteproc.h
> > > index 9c07d7958c53..38607107b7cb 100644
> > > --- a/include/linux/remoteproc.h
> > > +++ b/include/linux/remoteproc.h
> > > @@ -489,7 +489,7 @@ struct rproc {
> > >   	struct list_head node;
> > >   	struct iommu_domain *domain;
> > >   	const char *name;
> > > -	char *firmware;
> > > +	const char *firmware;
> > >   	void *priv;
> > >   	struct rproc_ops *ops;
> > >   	struct device dev;
> > > 
> > 
> 