Re: [PATCH 01/18] mtd: rawnand: sunxi: Use devm_platform_get_and_ioremap_resource()

[Miquel Raynal]

miquel.raynal@xxxxxxxxxxx wrote on Wed, 12 Jul 2023 13:54:55 +0200:

> Hi Yangtao,
> 
> frank.li@xxxxxxxx wrote on Fri,  7 Jul 2023 12:06:05 +0800:
> 
> > Convert platform_get_resource(), devm_ioremap_resource() to a single
> > call to devm_platform_get_and_ioremap_resource(), as this is exactly
> > what this function does.
> > 
> > Signed-off-by: Yangtao Li <frank.li@xxxxxxxx>
> > ---
> >  drivers/mtd/nand/raw/sunxi_nand.c | 3 +--
> >  1 file changed, 1 insertion(+), 2 deletions(-)
> > 
> > diff --git a/drivers/mtd/nand/raw/sunxi_nand.c b/drivers/mtd/nand/raw/sunxi_nand.c
> > index 9884304634f6..db36bd755b8d 100644
> > --- a/drivers/mtd/nand/raw/sunxi_nand.c
> > +++ b/drivers/mtd/nand/raw/sunxi_nand.c
> > @@ -2087,8 +2087,7 @@ static int sunxi_nfc_probe(struct platform_device *pdev)
> >  	nand_controller_init(&nfc->controller);
> >  	INIT_LIST_HEAD(&nfc->chips);
> >  
> > -	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> > -	nfc->regs = devm_ioremap_resource(dev, r);
> > +	nfc->regs = devm_platform_get_and_ioremap_resource(pdev, 0, &r);
> 
> Why do you keep a reference over the resource? Why not just
> devm_platform_ioremap_resource(pdev, 0) ?
> 
> This comment is valid for almost all the cases in this series.

Actually 'r' is used at the bottom, so the patch is good. I reviewed
all the patches, 17 look right, I have a minor comment on one of them.
In order to avoid resending all the patches, I'll first apply the 17
good ones and ask you to resend just the last one.

> When the resource is only needed in an error printk, I am also in favor
> of modifying the error message to avoid having to grab the resource.
> 
> >  	if (IS_ERR(nfc->regs))
> >  		return PTR_ERR(nfc->regs);
> >  
> 
> These comments apply to the 18 patches.
> 
> Thanks,
> Miquèl


Thanks,
Miquèl