Received-SPF: pass (google.com: domain of linux-ext4+bounces-1021-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
Date: Tue, 30 Jan 2024 09:12:52 +0100
From: Christoph Hellwig <hch@lst.de>
To: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>, linux-fsdevel@vger.kernel.org,
	linux-ext4@vger.kernel.org
Subject: Re: [PATCH 1/3] fs: Introduce buffered_write_operations
Message-ID: <20240130081252.GC22621@lst.de>
References: <20240130055414.2143959-1-willy@infradead.org> <20240130055414.2143959-2-willy@infradead.org>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20240130055414.2143959-2-willy@infradead.org>
User-Agent: Mutt/1.5.17 (2007-11-01)

> +struct buffered_write_operations {
> +	int (*write_begin)(struct file *, struct address_space *mapping,
> +			loff_t pos, size_t len, struct folio **foliop,
> +			void **fsdata);
> +	int (*write_end)(struct file *, struct address_space *mapping,
> +			loff_t pos, size_t len, size_t copied,
> +			struct folio *folio, void **fsdata);
> +};

Should write_begin simply return the folio or an ERR_PTR instead of
the return by reference?

I also wonder if the fsdata paramter should go away - if a fs needs
to pass forth and back fsdata, generic/filemap_perform_write is
probably the wrong abstraction for it.

Otherwise this looks sane to me.