On Thu, 7 Apr 2011, Andrew Morton wrote: > On Mon, 14 Mar 2011 22:00:25 +0100 > Julia Lawall <julia@xxxxxxx> wrote: > > > The memory allocated using request_mem_region should be released using > > release_mem_region, not release_region. > > > > The semantic patch that fixes part of this problem is as follows: > > (http://coccinelle.lip6.fr/) > > > > // <smpl> > > @@ > > expression E1,E2,E3; > > @@ > > > > request_mem_region(E1,E2,E3) > > ... > > ?- release_region(E1,E2) > > + release_mem_region(E1,E2) > > // </smpl> > > > > Signed-off-by: Julia Lawall <julia@xxxxxxx> > > > > --- > > drivers/rtc/rtc-mrst.c | 4 ++-- > > 1 file changed, 2 insertions(+), 2 deletions(-) > > > > diff --git a/drivers/rtc/rtc-mrst.c b/drivers/rtc/rtc-mrst.c > > index b86bc32..558dd7d 100644 > > --- a/drivers/rtc/rtc-mrst.c > > +++ b/drivers/rtc/rtc-mrst.c > > @@ -379,7 +379,7 @@ cleanup1: > > mrst_rtc.dev = NULL; > > rtc_device_unregister(mrst_rtc.rtc); > > cleanup0: > > - release_region(iomem->start, iomem->end + 1 - iomem->start); > > + release_mem_region(iomem->start, iomem->end + 1 - iomem->start); > > dev_err(dev, "rtc-mrst: unable to initialise\n"); > > return retval; > > } > > @@ -405,7 +405,7 @@ static void __exit rtc_mrst_do_remove(struct device *dev) > > mrst->rtc = NULL; > > > > iomem = mrst->iomem; > > - release_region(iomem->start, iomem->end + 1 - iomem->start); > > + release_mem_region(iomem->start, iomem->end + 1 - iomem->start); > > mrst->iomem = NULL; > > > > mrst->dev = NULL; > > We may as well clean this up too. > > --- a/drivers/rtc/rtc-mrst.c~drivers-rtc-rtc-mrstc-use-release_mem_region-after-request_mem_region-fix > +++ a/drivers/rtc/rtc-mrst.c > @@ -380,7 +380,7 @@ cleanup1: > cleanup0: > dev_set_drvdata(dev, NULL); > mrst_rtc.dev = NULL; > - release_mem_region(iomem->start, iomem->end + 1 - iomem->start); > + release_mem_region(iomem->start, resource_size(iomem)); > dev_err(dev, "rtc-mrst: unable to initialise\n"); > return retval; > } > @@ -406,7 +406,7 @@ static void __devexit rtc_mrst_do_remove > mrst->rtc = NULL; > > iomem = mrst->iomem; > - release_mem_region(iomem->start, iomem->end + 1 - iomem->start); > + release_mem_region(iomem->start, resource_size(iomem)); > mrst->iomem = NULL; > > mrst->dev = NULL; > _ > > which makes one wonder why we don't have > > void release_resource_mem(struct resource *r) > { > release_mem_region(r->start, resource_size(r)); > } Perhaps the name is not quite good, because release resource is something different from: void release_resource(struct resource *r) { release_region(r->start, resource_size(r)); } It would be really nice to have such a wrapper for both of them, since the code is so common, and the size computation is so error prone. julia -- To unsubscribe from this list: send the line "unsubscribe kernel-janitors" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
