From mboxrd@z Thu Jan 1 00:00:00 1970 From: Julia Lawall Subject: Re: [PATCH 14/27] mtd: nand: use the mtd instance embedded in struct nand_chip Date: Tue, 17 Nov 2015 10:05:03 +0100 (CET) Message-ID: References: <1447681080-31232-1-git-send-email-boris.brezillon@free-electrons.com> <1447681080-31232-15-git-send-email-boris.brezillon@free-electrons.com> <20151117030019.GY8456@google.com> <20151117093836.052c86b5@bbrezillon> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20151117093836.052c86b5@bbrezillon> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: driverdev-devel-bounces@linuxdriverproject.org Sender: "devel" To: Boris Brezillon Cc: Andrew Lunn , Krzysztof Kozlowski , linux-doc@vger.kernel.org, Tony Lindgren , Stefan Agner , linux-sunxi@googlegroups.com, linux-mtd@lists.infradead.org, Robert Jarzmik , Alexander Clouter , devel@driverdev.osuosl.org, Jesper Nilsson , linux-samsung-soc@vger.kernel.org, Maxim Levitsky , Jonathan Corbet , Marek Vasut , Chen-Yu Tsai , Kukjin Kim , Ezequiel Garcia , Josh Wu , Sebastian Hesselbarth , Jason Cooper , Wan ZongShun , Steven Miao , adi-buildroot-devel@lists.sourceforge.net, Ha List-Id: linux-omap@vger.kernel.org > > (This isn't the worst one, but it just happens to be one of the first.) > > There are many cases where the typical style would be to declare a new > > variable at the top of the function, where you perform the > > macro/function-call to convert from one abstraction to another. Like > > > > static int nfc_set_sram_bank(struct atmel_nand_host *host, unsigned int bank) > > { > > struct mtd_info *mtd = nand_to_mtd(&hot->nand_chip); > > ... > > > > and then use it later. Can that be done very easily? > > > > > return -EINVAL; > > > nfc_writel(host->nfc->hsmc_regs, BANK, ATMEL_HSMC_NFC_BANK1); > > > } else { > > > > ... > > Honestly, I don't know how to do that with a coccinelle script, and it > will probably take me more time to find how to do it than addressing > those problems manually. > > Julia, could you give us some hint? Probably something like the following would be easiest. You can just run it after your other transformations: @r exists@ identifier f; expression e; @@ f(...) { <+... nand_to_mtd(e) ...+> } @@ identifier r.f; expression r.e; @@ f(...) { + struct mtd_info *mtd = nand_to_mtd(e); ... } This won't work if there is more than one possible value of e. If that is likely, then I could come up with something more complex. It also assumes that you want to convert all such calls. If you only want to convert calls that occur in a particular context, eg a field reference, then you could enhance the pattern inside the <+... ...+> in the first rule. julia