Re: [PATCH v4 2/5] btrfs: initial fsverity support

From: David Sterba <dsterba@suse.cz>
To: Boris Burkov <boris@bur.io>
Cc: dsterba@suse.cz, linux-btrfs@vger.kernel.org,
	linux-fscrypt@vger.kernel.org, kernel-team@fb.com
Subject: Re: [PATCH v4 2/5] btrfs: initial fsverity support
Date: Mon, 17 May 2021 23:40:23 +0200	[thread overview]
Message-ID: <20210517214023.GR7604@twin.jikos.cz> (raw)
In-Reply-To: <YJ17yjaMwJTH+GtU@zen>

On Thu, May 13, 2021 at 12:19:38PM -0700, Boris Burkov wrote:
> On Tue, May 11, 2021 at 10:31:43PM +0200, David Sterba wrote:
> > On Wed, May 05, 2021 at 12:20:40PM -0700, Boris Burkov wrote:
> > > +	/* zero fill any bytes we didn't write into the page */
> > > +	if (ret < PAGE_SIZE) {
> > > +		char *kaddr = kmap_atomic(p);
> > > +
> > > +		memset(kaddr + ret, 0, PAGE_SIZE - ret);
> > > +		kunmap_atomic(kaddr);
> > 
> > There's helper memzero_page wrapping the kmap
> 
> FWIW, that helper uses kmap_atomic, not kmap_local_page. Would you
> prefer I use the helper or not introduce new uses of kmap_atomic?

Please use the memzero_page helper, see d048b9c2a737eb791a5e9 "btrfs:
use memzero_page() instead of open coded kmap pattern" for more details.