2020-01-23 15:48:47

by Daniel Baluta (OSS)

[permalink] [raw]
Subject: [PATCH] lib: devres: Export devm_ioremap_resource_wc

From: Daniel Baluta <[email protected]>

So that modules can also use it.

Fixes: b873af620e58863b ("lib: devres: provide devm_ioremap_resource_wc()")
Signed-off-by: Daniel Baluta <[email protected]>
---
lib/devres.c | 1 +
1 file changed, 1 insertion(+)

diff --git a/lib/devres.c b/lib/devres.c
index 6ef51f159c54..7fe2bd75dfa3 100644
--- a/lib/devres.c
+++ b/lib/devres.c
@@ -182,6 +182,7 @@ void __iomem *devm_ioremap_resource_wc(struct device *dev,
{
return __devm_ioremap_resource(dev, res, DEVM_IOREMAP_WC);
}
+EXPORT_SYMBOL(devm_ioremap_resource_wc);

/*
* devm_of_iomap - Requests a resource and maps the memory mapped IO
--
2.17.1


2020-01-23 17:58:09

by Greg Kroah-Hartman

[permalink] [raw]
Subject: Re: [PATCH] lib: devres: Export devm_ioremap_resource_wc

On Thu, Jan 23, 2020 at 03:47:21PM +0000, Daniel Baluta (OSS) wrote:
> From: Daniel Baluta <[email protected]>
>
> So that modules can also use it.
>
> Fixes: b873af620e58863b ("lib: devres: provide devm_ioremap_resource_wc()")
> Signed-off-by: Daniel Baluta <[email protected]>
> ---
> lib/devres.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/lib/devres.c b/lib/devres.c
> index 6ef51f159c54..7fe2bd75dfa3 100644
> --- a/lib/devres.c
> +++ b/lib/devres.c
> @@ -182,6 +182,7 @@ void __iomem *devm_ioremap_resource_wc(struct device *dev,
> {
> return __devm_ioremap_resource(dev, res, DEVM_IOREMAP_WC);
> }
> +EXPORT_SYMBOL(devm_ioremap_resource_wc);

EXPORT_SYMBOL_GPL() perhaps?

What in-tree driver needs this?

thanks,

greg k-h

2020-01-23 18:18:02

by Daniel Baluta

[permalink] [raw]
Subject: Re: [PATCH] lib: devres: Export devm_ioremap_resource_wc

On Thu, Jan 23, 2020 at 7:57 PM [email protected]
<[email protected]> wrote:
>
> On Thu, Jan 23, 2020 at 03:47:21PM +0000, Daniel Baluta (OSS) wrote:
> > From: Daniel Baluta <[email protected]>
> >
> > So that modules can also use it.
> >
> > Fixes: b873af620e58863b ("lib: devres: provide devm_ioremap_resource_wc()")
> > Signed-off-by: Daniel Baluta <[email protected]>
> > ---
> > lib/devres.c | 1 +
> > 1 file changed, 1 insertion(+)
> >
> > diff --git a/lib/devres.c b/lib/devres.c
> > index 6ef51f159c54..7fe2bd75dfa3 100644
> > --- a/lib/devres.c
> > +++ b/lib/devres.c
> > @@ -182,6 +182,7 @@ void __iomem *devm_ioremap_resource_wc(struct device *dev,
> > {
> > return __devm_ioremap_resource(dev, res, DEVM_IOREMAP_WC);
> > }
> > +EXPORT_SYMBOL(devm_ioremap_resource_wc);
>
> EXPORT_SYMBOL_GPL() perhaps?
>
> What in-tree driver needs this?

I was experimenting with an out of tree driver and I also was using it wrong :D.
Indeed looks like there is no real potential user so far in the kernel tree.

Perhaps: drivers/net/ethernet/amazon/ena/ena_netdev.c

Also, I chose EXPORT_SYMBOL because the same way its cousin
devm_ioremap_resource.

Daniel.

2020-01-23 18:23:14

by Greg Kroah-Hartman

[permalink] [raw]
Subject: Re: [PATCH] lib: devres: Export devm_ioremap_resource_wc

On Thu, Jan 23, 2020 at 08:16:35PM +0200, Daniel Baluta wrote:
> On Thu, Jan 23, 2020 at 7:57 PM [email protected]
> <[email protected]> wrote:
> >
> > On Thu, Jan 23, 2020 at 03:47:21PM +0000, Daniel Baluta (OSS) wrote:
> > > From: Daniel Baluta <[email protected]>
> > >
> > > So that modules can also use it.
> > >
> > > Fixes: b873af620e58863b ("lib: devres: provide devm_ioremap_resource_wc()")
> > > Signed-off-by: Daniel Baluta <[email protected]>
> > > ---
> > > lib/devres.c | 1 +
> > > 1 file changed, 1 insertion(+)
> > >
> > > diff --git a/lib/devres.c b/lib/devres.c
> > > index 6ef51f159c54..7fe2bd75dfa3 100644
> > > --- a/lib/devres.c
> > > +++ b/lib/devres.c
> > > @@ -182,6 +182,7 @@ void __iomem *devm_ioremap_resource_wc(struct device *dev,
> > > {
> > > return __devm_ioremap_resource(dev, res, DEVM_IOREMAP_WC);
> > > }
> > > +EXPORT_SYMBOL(devm_ioremap_resource_wc);
> >
> > EXPORT_SYMBOL_GPL() perhaps?
> >
> > What in-tree driver needs this?
>
> I was experimenting with an out of tree driver and I also was using it wrong :D.
> Indeed looks like there is no real potential user so far in the kernel tree.

Great, we can drop it then :)

thanks,

greg k-h