Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   "Fabio M. De Francesco" <fmdefrancesco@gmail.com>
To:     Chris Mason <clm@fb.com>, Josef Bacik <josef@toxicpanda.com>,
        David Sterba <dsterba@suse.com>,
        Nick Terrell <terrelln@fb.com>,
        Chris Down <chris@chrisdown.name>,
        Filipe Manana <fdmanana@suse.com>, Qu Wenruo <wqu@suse.com>,
        Nikolay Borisov <nborisov@suse.com>,
        Gabriel Niebler <gniebler@suse.com>,
        linux-btrfs@vger.kernel.org, linux-kernel@vger.kernel.org
Cc:     Ira Weiny <ira.weiny@intel.com>
Subject: Re: [PATCH] btrfs: Convert zlib_compress_pages() to use kmap_local_page()
Date:   Sat, 25 Jun 2022 16:41:04 +0200
Message-ID: <3110135.5fSG56mABF@opensuse>
In-Reply-To: <20220618092752.25153-1-fmdefrancesco@gmail.com>
References: <20220618092752.25153-1-fmdefrancesco@gmail.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 7Bit
Content-Type: text/plain; charset="us-ascii"
Precedence: bulk

On sabato 18 giugno 2022 11:27:52 CEST Fabio M. De Francesco wrote:
> The use of kmap() is being deprecated in favor of kmap_local_page(). With
> kmap_local_page(), the mapping is per thread, CPU local and not globally
> visible.
> 
> Therefore, use kmap_local_page() / kunmap_local() in 
zlib_compress_pages()
> because in this function the mappings are per thread and are not visible
> in other contexts.
> 
> Tested with xfstests on QEMU + KVM 32-bit VM with 4GB of RAM and
> HIGHMEM64G enabled. This patch passes 26/26 tests of group "compress".
> 
> Cc: Qu Wenruo <wqu@suse.com>
> Suggested-by: Ira Weiny <ira.weiny@intel.com>
> Signed-off-by: Fabio M. De Francesco <fmdefrancesco@gmail.com>
> ---
> 
> This patch builds only on top of
> "[PATCH] btrfs: zlib: refactor how we prepare the input buffer" by Qu 
Wenruo".
> https://lore.kernel.org/linux-btrfs/
d0bfc791b5509df7b9ad44e41ada197d1b3149b3.1655519730.git.wqu@suse.com/
> 

I've seen that Qu sent a v2 of the above patch. However David thinks that 
it is better not to map pages allocated in zlib.c for output (out_page) 
since they cannot come from highmem because of "alloc_page(GFP_NOFS);".

@David:

I suppose that, since it builds _only_ on top of the refactor submitted by 
Qu, I'll have to wait and see what you decide. 

If you don't want kmap_local_page() and prefer using page_address() on 
"out_page", please drop this patch and let me know, so that I can send a 
new patch which will be in accordance to your preference.

Thanks,

Fabio