From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Fri, 16 Sep 2016 14:19:18 +0200 From: Boris Brezillon To: Sascha Hauer Cc: linux-mtd@lists.infradead.org, kernel@pengutronix.de, Subject: Re: [PATCH v4] mtd: nand: automate NAND timings selection Message-ID: <20160916141918.0353f65f@bbrezillon> In-Reply-To: <1473928373-8680-1-git-send-email-s.hauer@pengutronix.de> References: <1473928373-8680-1-git-send-email-s.hauer@pengutronix.de> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Thu, 15 Sep 2016 10:32:44 +0200 Sascha Hauer wrote: > This series aims at automating the NAND timings selection which is > currently supposed to be done in each NAND controller driver, thus > simplifying drivers implementation. > > As suggested by Boris this version of the series introduces a nand_reset() > function which replaces the several open coded NAND_CMD_RESET commands > in the code. This makes sure we can apply the timing each time after > after reset. > > Also I have brought back the conversion patch for teh sunxi driver whic > was part of Boris initial posting. It's untested due to the lack of hardware, > so please test before applying. Applied after fixing a few coding style issues (to make checkpatch happy), and exporting the onfi_init_data_interface() function (not sure it's the best solution, but at least, it's consistent with the other functions defined in nand_timings.c). Thanks, Boris > > Sascha > > Changes since v4: > - Change onfi_init_data_interface() prototype to be more future proof as > requested by Boris > > Changes since v3: > - Bring back patch dropped in v3 > - Use statically allocated default timing for all chips and store one > optimized timing in struct nand_chip > > Changes since v2: > - Add accessor function to get the SDR timing from struct nand_data_interface > - Change nand_reset() argument to struct nand_chip > - Drop conversion of nand_timing array to struct nand_data_interface > - Recalculate timing whenever needed instead of storing a pointer in struct > nand_chip > - some more refactoring > > Changes since v1: > - create a nand_reset() function to create a single place to reset NAND > chips and to apply timings > - Add patch to convert sunxi driver for automated timing setup > - split into more patches > > Changes since the initial posting from Boris: > > - Integrate Feedback from Ezequiel Garcia > - When iterating over the chips calling onfi_set_features() for each > bail out when any of the calls fail, not only the last one. > - When one of the onfi_set_features() calls fail then reset the chipi > afterwards. > - Drop Sunxi example, add patch for the mxc_nand controller instead. > > ---------------------------------------------------------------- > > Boris Brezillon (1): > mtd: nand: automate NAND timings selection > > Sascha Hauer (8): > mtd: nand: Create a NAND reset function > mtd: nand: Introduce nand_data_interface > mtd: nand: convert ONFI mode into data interface > mtd: nand: Add function to convert ONFI mode to data_interface > mtd: nand: Expose data interface for ONFI mode 0 > mtd: nand: sunxi: switch from manual to automated timing config > mtd: nand: mxc: implement onfi get/set features > mtd: nand: mxc: Add timing setup for v2 controllers > > drivers/mtd/nand/mxc_nand.c | 133 ++++++++++++ > drivers/mtd/nand/nand_base.c | 179 ++++++++++++++- > drivers/mtd/nand/nand_timings.c | 469 ++++++++++++++++++++++------------------ > drivers/mtd/nand/sunxi_nand.c | 76 ++----- > include/linux/mtd/nand.h | 190 +++++++++++----- > 5 files changed, 721 insertions(+), 326 deletions(-) From mboxrd@z Thu Jan 1 00:00:00 1970 From: boris.brezillon@free-electrons.com (Boris Brezillon) Date: Fri, 16 Sep 2016 14:19:18 +0200 Subject: [PATCH v4] mtd: nand: automate NAND timings selection In-Reply-To: <1473928373-8680-1-git-send-email-s.hauer@pengutronix.de> References: <1473928373-8680-1-git-send-email-s.hauer@pengutronix.de> Message-ID: <20160916141918.0353f65f@bbrezillon> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Thu, 15 Sep 2016 10:32:44 +0200 Sascha Hauer wrote: > This series aims at automating the NAND timings selection which is > currently supposed to be done in each NAND controller driver, thus > simplifying drivers implementation. > > As suggested by Boris this version of the series introduces a nand_reset() > function which replaces the several open coded NAND_CMD_RESET commands > in the code. This makes sure we can apply the timing each time after > after reset. > > Also I have brought back the conversion patch for teh sunxi driver whic > was part of Boris initial posting. It's untested due to the lack of hardware, > so please test before applying. Applied after fixing a few coding style issues (to make checkpatch happy), and exporting the onfi_init_data_interface() function (not sure it's the best solution, but at least, it's consistent with the other functions defined in nand_timings.c). Thanks, Boris > > Sascha > > Changes since v4: > - Change onfi_init_data_interface() prototype to be more future proof as > requested by Boris > > Changes since v3: > - Bring back patch dropped in v3 > - Use statically allocated default timing for all chips and store one > optimized timing in struct nand_chip > > Changes since v2: > - Add accessor function to get the SDR timing from struct nand_data_interface > - Change nand_reset() argument to struct nand_chip > - Drop conversion of nand_timing array to struct nand_data_interface > - Recalculate timing whenever needed instead of storing a pointer in struct > nand_chip > - some more refactoring > > Changes since v1: > - create a nand_reset() function to create a single place to reset NAND > chips and to apply timings > - Add patch to convert sunxi driver for automated timing setup > - split into more patches > > Changes since the initial posting from Boris: > > - Integrate Feedback from Ezequiel Garcia > - When iterating over the chips calling onfi_set_features() for each > bail out when any of the calls fail, not only the last one. > - When one of the onfi_set_features() calls fail then reset the chipi > afterwards. > - Drop Sunxi example, add patch for the mxc_nand controller instead. > > ---------------------------------------------------------------- > > Boris Brezillon (1): > mtd: nand: automate NAND timings selection > > Sascha Hauer (8): > mtd: nand: Create a NAND reset function > mtd: nand: Introduce nand_data_interface > mtd: nand: convert ONFI mode into data interface > mtd: nand: Add function to convert ONFI mode to data_interface > mtd: nand: Expose data interface for ONFI mode 0 > mtd: nand: sunxi: switch from manual to automated timing config > mtd: nand: mxc: implement onfi get/set features > mtd: nand: mxc: Add timing setup for v2 controllers > > drivers/mtd/nand/mxc_nand.c | 133 ++++++++++++ > drivers/mtd/nand/nand_base.c | 179 ++++++++++++++- > drivers/mtd/nand/nand_timings.c | 469 ++++++++++++++++++++++------------------ > drivers/mtd/nand/sunxi_nand.c | 76 ++----- > include/linux/mtd/nand.h | 190 +++++++++++----- > 5 files changed, 721 insertions(+), 326 deletions(-)