On 2014?06?30? 17:59, Caizhiyong wrote:

quoted

-----Original Message-----
From: Arnd Bergmann [mailto:arnd at arndb.de]
Sent: Monday, June 30, 2014 5:01 PM
To: linux-arm-kernel at lists.infradead.org
Cc: Zhou Wang; Rob Herring; Pawel Moll; Mark Rutland; Ian Campbell; Kumar
Gala; Russell King; David Woodhouse; Brian Norris; Grant Likely; Ezequiel
Garcia; Pekon Gupta; Artem Bityutskiy; Alexander Shiyan; Ivan Khoronzhuk;
Jussi Kivilinna; Joern Engel; Randy Dunlap; devicetree at vger.kernel.org;
linux-mtd at lists.infradead.org; linux-doc at vger.kernel.org;
linux-kernel at vger.kernel.org; Caizhiyong; Wangzhou (B)
Subject: Re: [PATCH 2/3] mtd: hisilicon: add a new nand controller driver for
hisilicon hip04 Soc

quoted

+	struct device		*dev;
+	void __iomem		*iobase;
+	struct completion       cmd_complete;
+	unsigned int		offset;
+	unsigned int		command;
+	int			chipselect;
+	unsigned int		addr_cycle;
+	unsigned int		addr_value[2];
+	unsigned int		cache_addr_value[2];
+	char			*buffer;
+	dma_addr_t		dma_buffer;
+	dma_addr_t		dma_oob;
+	int			version;
+	unsigned int            ecc_bits;
+	unsigned int            irq_status; /* interrupt status */
+
+	int (*send_cmd_pageprog)(struct hinfc_host *host);
+	int (*send_cmd_status)(struct hinfc_host *host);
+	int (*send_cmd_readstart)(struct hinfc_host *host);
+	int (*send_cmd_erase)(struct hinfc_host *host);
+	int (*send_cmd_readid)(struct hinfc_host *host);
+	int (*send_cmd_reset)(struct hinfc_host *host, int chipselect);
+};

Why do you need function pointers here? The current version of the
driver you posted always assigns these to the same functions, so
it would be more efficient to just call those directly.

I agree with you.
This feature was originally designed to support a variety version of NAND controllers.
But in fact, this feature is not used now.

I will move them out the host, thanks for your opinion!