Message ID | CAN_5kQDTweB2GLvwgjN+BQuqnhFUp+MOcQr=ZW=8C6XSsC8XtA@mail.gmail.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
On Wed, Jul 06, 2011 at 07:19:59PM -0700, heechul Yun wrote: > ARM: mm: remove unnecessary cache flush on v6 copypage > > Signed-off-by: Heechul Yun <heechul@illinois.edu> > -- > arch/arm/mm/copypage-v6.c | 1 - > 1 files changed, 0 insertions(+), 1 deletions(-) > > diff --git a/arch/arm/mm/copypage-v6.c b/arch/arm/mm/copypage-v6.c > index bdba6c6..63cca00 100644 > --- a/arch/arm/mm/copypage-v6.c > +++ b/arch/arm/mm/copypage-v6.c > @@ -41,7 +41,6 @@ static void v6_copy_user_highpage_nonaliasing(struct page *to, > kfrom = kmap_atomic(from, KM_USER0); > kto = kmap_atomic(to, KM_USER1); > copy_page(kto, kfrom); > - __cpuc_flush_dcache_area(kto, PAGE_SIZE); > kunmap_atomic(kto, KM_USER1); > kunmap_atomic(kfrom, KM_USER0); > } Patch looks good, except that it's whitespace damaged (the tabs are replaced by spaces) which makes it hard to apply. Could you try to sort that out. Documentation/email-clients.txt in the kernel tree may help with that. I assume Catalin will (eventually) provide an ack - and at that point please submit it (in non-whitespace damaged form) to the patch system, thanks.
On Thursday, July 7, 2011, heechul Yun <heechul@illinois.edu> wrote: > ARM: mm: remove unnecessary cache flush on v6 copypage > > Signed-off-by: Heechul Yun <heechul@illinois.edu> I would clarify the commit log a bit stating that this flush is no longer necessary because of commit c0177800. We had a good reason for adding the flush initially. Otherwise, Acked-by: Catalin Marinas <catalin.marinas@arm.com>
On Thu, Jul 7, 2011 at 12:02 PM, Catalin Marinas <catalin.marinas@arm.com> wrote: > On Thursday, July 7, 2011, heechul Yun <heechul@illinois.edu> wrote: >> ARM: mm: remove unnecessary cache flush on v6 copypage I think you better mention that this is regarding v6 nonaliasing mode. >> >> Signed-off-by: Heechul Yun <heechul@illinois.edu> > > I would clarify the commit log a bit stating that this flush is no > longer necessary because of commit c0177800. We had a good reason for > adding the flush initially. I agree, but I would also mention that the flush was needed in order to coherent between dcache and icache, and now this is being handled by __sync_icache_dcache saeed
diff --git a/arch/arm/mm/copypage-v6.c b/arch/arm/mm/copypage-v6.c index bdba6c6..63cca00 100644 --- a/arch/arm/mm/copypage-v6.c +++ b/arch/arm/mm/copypage-v6.c @@ -41,7 +41,6 @@ static void v6_copy_user_highpage_nonaliasing(struct page *to, kfrom = kmap_atomic(from, KM_USER0); kto = kmap_atomic(to, KM_USER1); copy_page(kto, kfrom); - __cpuc_flush_dcache_area(kto, PAGE_SIZE); kunmap_atomic(kto, KM_USER1); kunmap_atomic(kfrom, KM_USER0); }
ARM: mm: remove unnecessary cache flush on v6 copypage Signed-off-by: Heechul Yun <heechul@illinois.edu> -- arch/arm/mm/copypage-v6.c | 1 - 1 files changed, 0 insertions(+), 1 deletions(-) -- Heechul