* Re: [PATCH 1/7] ARM: add GPMI support for imx23/imx28 [not found] ` <19840.43943.592336.854865@ipc1.ka-ro> @ 2011-03-17 2:19 ` Huang Shijie 2011-03-17 10:36 ` Lothar Waßmann 0 siblings, 1 reply; 21+ messages in thread From: Huang Shijie @ 2011-03-17 2:19 UTC (permalink / raw) To: Lothar Waßmann, linux-arm-kernel, linux-mtd Hi: > Huang Shijie writes: >> Hi Lothar: >>> Hi, >>> >>> All in all I do not like this driver in any way. I already vomitted >>> over it, when I saw it in the Freescale BSP. >> :) I have already removed a lot of code. >> >>> There is no need for the multiple layers of abstraction that are >>> papered over the MTD layer. >> I think this is good. Make the code clear. >> > I'll comment when the remaining patch has got through. > thanks a lot . >>> There is no need for spilling the driver over a dozen files. >> I can merge some files into one. thanks. >> > IMO all the source code could be kept in one file (except for the > include files of course) and the gpmi-nfc subdirectory removed. I really can merge some files, but IMHO, it's not a good solution to keep all in one file. The reasons are : [1] The GPMI support the NAND boot mode,but the imx23 and imx28 is a little different in the NAND boot procedure due the different firmwares in the chips. If I merge the rom-imx23.c rom-imx28.c to one file, the code will very mess. ugly enough! [2]The GPMI driver is just a mini version against the version in the FREESCALE BSP. The original GPMI driver also support the IMX508 which is different from the imx23 and imx28. The imx508 supports the ONFI NAND and TOGGLE NAND which are DDR mode NAND.So the implement of ECC read_page/write_page is much different from the IMX23/IMX28. Because I want to support the IMX508 in future, this is why the hal-imx23.c and hal-imx28.c share most the same code,but i still remain them in separate, the reason is simple : I will add a hal-imx508.c in future. [3]The GPMI driver is a little complicated. The GPMI controller needs other two hardware module to work :the BCH module, and the DMA mode. Yes, i can merge the DMA code into other files, but the BCH module should occupy a separate file to make the logic clear. [4] If i packet all the code in one, the gentlemen of MTD mailist, such as David Woodhouse, Artem Bityutskiy , will kill me, :). what's more the big boy even can not be emailed to arm-kernel maillist. > I've been working on this too, but had to abandone it for other tasks > several times. :( > I really apreciate for your review and comments. I just want to make the drive tidy and grace. Please point out the ugly part that you regards in free. I am very glad you had ever tried to work on this. :) Best Regards Huang Shijie ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 1/7] ARM: add GPMI support for imx23/imx28 2011-03-17 2:19 ` [PATCH 1/7] ARM: add GPMI support for imx23/imx28 Huang Shijie @ 2011-03-17 10:36 ` Lothar Waßmann 2011-03-18 2:06 ` Huang Shijie 0 siblings, 1 reply; 21+ messages in thread From: Lothar Waßmann @ 2011-03-17 10:36 UTC (permalink / raw) To: Huang Shijie; +Cc: linux-mtd, linux-arm-kernel Hi, Huang Shijie writes: > Hi: > > > Huang Shijie writes: > >> Hi Lothar: > >>> Hi, > >>> > >>> All in all I do not like this driver in any way. I already vomitted > >>> over it, when I saw it in the Freescale BSP. > >> :) I have already removed a lot of code. > >> > >>> There is no need for the multiple layers of abstraction that are > >>> papered over the MTD layer. > >> I think this is good. Make the code clear. > >> > > I'll comment when the remaining patch has got through. > > > thanks a lot . > > > >>> There is no need for spilling the driver over a dozen files. > >> I can merge some files into one. thanks. > >> > > IMO all the source code could be kept in one file (except for the > > include files of course) and the gpmi-nfc subdirectory removed. > I really can merge some files, but IMHO, it's not a good solution to > keep all in one file. > > The reasons are : > [1] The GPMI support the NAND boot mode,but the imx23 and imx28 is a > little different in > the NAND boot procedure due the different firmwares in the chips. If I > merge the rom-imx23.c > rom-imx28.c to one file, the code will very mess. ugly enough! > You can use platform_ids to differentiate between the versions and have separate functions for imx23 and imx28 where appropriate and select between those depending on the platform_id. Lothar Waßmann -- ___________________________________________________________ Ka-Ro electronics GmbH | Pascalstraße 22 | D - 52076 Aachen Phone: +49 2408 1402-0 | Fax: +49 2408 1402-10 Geschäftsführer: Matthias Kaussen Handelsregistereintrag: Amtsgericht Aachen, HRB 4996 www.karo-electronics.de | info@karo-electronics.de ___________________________________________________________ ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 1/7] ARM: add GPMI support for imx23/imx28 2011-03-17 10:36 ` Lothar Waßmann @ 2011-03-18 2:06 ` Huang Shijie 0 siblings, 0 replies; 21+ messages in thread From: Huang Shijie @ 2011-03-18 2:06 UTC (permalink / raw) To: Lothar Waßmann; +Cc: linux-mtd, linux-arm-kernel hi: >>> IMO all the source code could be kept in one file (except for the >>> include files of course) and the gpmi-nfc subdirectory removed. >> I really can merge some files, but IMHO, it's not a good solution to >> keep all in one file. >> >> The reasons are : >> [1] The GPMI support the NAND boot mode,but the imx23 and imx28 is a >> little different in >> the NAND boot procedure due the different firmwares in the chips. If I >> merge the rom-imx23.c >> rom-imx28.c to one file, the code will very mess. ugly enough! >> > You can use platform_ids to differentiate between the versions and > have separate functions for imx23 and imx28 where appropriate and > select between those depending on the platform_id. > > thanks a lot. Best Regards. Huang Shijie ^ permalink raw reply [flat|nested] 21+ messages in thread
[parent not found: <1300239773-4222-6-git-send-email-b32955@freescale.com>]
[parent not found: <19848.39439.380861.195051@ipc1.ka-ro>]
* Re: [PATCH 5/7] add GPMI support for imx28 [not found] ` <19848.39439.380861.195051@ipc1.ka-ro> @ 2011-03-23 3:11 ` Huang Shijie 2011-03-23 14:56 ` Lothar Waßmann 0 siblings, 1 reply; 21+ messages in thread From: Huang Shijie @ 2011-03-23 3:11 UTC (permalink / raw) To: Lothar Waßmann; +Cc: linux, linux-mtd, dwmw2, linux-arm-kernel Hi Lothar: Thanks a lot for your carefully review. > Hi, > > some general comments: > - Why are you not using the existing nand_ids but inventing your own > database? > The nand_ids{} contains poor information for me. Such as : [1]The nand_ids does not have the enough information for the page size,oob size for some new NANDs. And you can not get the information from parsing the NAND ids, such as some Micron NANDs. [2]I need the timing information of the NAND. The nand_ids DOES not have it. I have to read the datasheet of the NAND, and add it to my database. [3]I need the information about the error Management for BCH. The BCH module needs to know the ECC information of the NAND, the nand_ids{} does not have it. Different NANDs may have different ECC requires such as "Minimum required ECC". > - What is the purpose of the 'rom_helpers'? To me it looks like they > are doing something that should be done in the mtd partition > handlers, not in the flash chip driver. > We have saled a lot products based the imx23 or imx28. All the products support the nand boot mode by flashing the kernel to the kernel partitions of the nands. In order to compliant to the old product, the layout of the nand is constraint to the current situation: [1] first a boot partition for the kernel. [2] the rest is for other uses, such as rootfs. So I have to use the rom_helpers to split the kernel partition before the real MTD partition initializations. > Huang Shijie writes: >> These files contain the code to implement the GPMI in the imx28. >> >> Signed-off-by: Huang Shijie<b32955@freescale.com> >> --- >> drivers/mtd/nand/gpmi-nfc/bch-regs-imx28.h | 557 +++++++++++++++++++++++++++ >> drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx28.h | 421 ++++++++++++++++++++ >> drivers/mtd/nand/gpmi-nfc/hal-imx28.c | 503 ++++++++++++++++++++++++ >> drivers/mtd/nand/gpmi-nfc/rom-imx28.c | 66 ++++ >> 4 files changed, 1547 insertions(+), 0 deletions(-) >> create mode 100644 drivers/mtd/nand/gpmi-nfc/bch-regs-imx28.h >> create mode 100644 drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx28.h >> create mode 100644 drivers/mtd/nand/gpmi-nfc/hal-imx28.c >> create mode 100644 drivers/mtd/nand/gpmi-nfc/rom-imx28.c >> >> diff --git a/drivers/mtd/nand/gpmi-nfc/bch-regs-imx28.h b/drivers/mtd/nand/gpmi-nfc/bch-regs-imx28.h >> new file mode 100644 >> index 0000000..692db08 >> --- /dev/null >> +++ b/drivers/mtd/nand/gpmi-nfc/bch-regs-imx28.h >> @@ -0,0 +1,557 @@ >> +/* >> + * Freescale GPMI NFC NAND Flash Driver >> + * >> + * Copyright 2008-2010 Freescale Semiconductor, Inc. All Rights Reserved. >> + * >> + * This program is free software; you can redistribute it and/or modify >> + * it under the terms of the GNU General Public License as published by >> + * the Free Software Foundation; either version 2 of the License, or >> + * (at your option) any later version. >> + * >> + * This program is distributed in the hope that it will be useful, >> + * but WITHOUT ANY WARRANTY; without even the implied warranty of >> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the >> + * GNU General Public License for more details. >> + * >> + * You should have received a copy of the GNU General Public License >> + * along with this program; if not, write to the Free Software >> + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA >> + * >> + * Xml Revision: 2.5 >> + * Template revision: 26195 >> + */ >> + >> +#ifndef __GPMI_NFC_BCH_REGS_H >> +#define __GPMI_NFC_BCH_REGS_H >> + >> +/*============================================================================*/ >> + >> +#define HW_BCH_CTRL (0x00000000) >> +#define HW_BCH_CTRL_SET (0x00000004) >> +#define HW_BCH_CTRL_CLR (0x00000008) >> +#define HW_BCH_CTRL_TOG (0x0000000c) >> > No need for parens around bare numbers. > Ok. Frankly speaking, the file is auto created by some tools from the hardware info. >> +#define BM_BCH_CTRL_SFTRST 0x80000000 >> +#define BV_BCH_CTRL_SFTRST__RUN 0x0 >> +#define BV_BCH_CTRL_SFTRST__RESET 0x1 >> +#define BM_BCH_CTRL_CLKGATE 0x40000000 >> +#define BV_BCH_CTRL_CLKGATE__RUN 0x0 >> +#define BV_BCH_CTRL_CLKGATE__NO_CLKS 0x1 >> +#define BP_BCH_CTRL_RSVD5 23 >> +#define BM_BCH_CTRL_RSVD5 0x3F800000 >> +#define BF_BCH_CTRL_RSVD5(v) \ >> + (((v)<< 23)& BM_BCH_CTRL_RSVD5) >> +#define BM_BCH_CTRL_DEBUGSYNDROME 0x00400000 >> +#define BP_BCH_CTRL_RSVD4 20 >> +#define BM_BCH_CTRL_RSVD4 0x00300000 >> +#define BF_BCH_CTRL_RSVD4(v) \ >> + (((v)<< 20)& BM_BCH_CTRL_RSVD4) >> +#define BP_BCH_CTRL_M2M_LAYOUT 18 >> +#define BM_BCH_CTRL_M2M_LAYOUT 0x000C0000 >> +#define BF_BCH_CTRL_M2M_LAYOUT(v) \ >> + (((v)<< 18)& BM_BCH_CTRL_M2M_LAYOUT) >> +#define BM_BCH_CTRL_M2M_ENCODE 0x00020000 >> +#define BM_BCH_CTRL_M2M_ENABLE 0x00010000 >> +#define BP_BCH_CTRL_RSVD3 11 >> +#define BM_BCH_CTRL_RSVD3 0x0000F800 >> +#define BF_BCH_CTRL_RSVD3(v) \ >> + (((v)<< 11)& BM_BCH_CTRL_RSVD3) >> +#define BM_BCH_CTRL_DEBUG_STALL_IRQ_EN 0x00000400 >> +#define BM_BCH_CTRL_RSVD2 0x00000200 >> +#define BM_BCH_CTRL_COMPLETE_IRQ_EN 0x00000100 >> +#define BP_BCH_CTRL_RSVD1 4 >> +#define BM_BCH_CTRL_RSVD1 0x000000F0 >> +#define BF_BCH_CTRL_RSVD1(v) \ >> + (((v)<< 4)& BM_BCH_CTRL_RSVD1) >> +#define BM_BCH_CTRL_BM_ERROR_IRQ 0x00000008 >> +#define BM_BCH_CTRL_DEBUG_STALL_IRQ 0x00000004 >> +#define BM_BCH_CTRL_RSVD0 0x00000002 >> +#define BM_BCH_CTRL_COMPLETE_IRQ 0x00000001 >> + > [...] > Messed up indentation. Also, use TAB instead of spaces for indentation. > ditto >> diff --git a/drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx28.h b/drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx28.h >> new file mode 100644 >> index 0000000..dcb3b7d >> --- /dev/null >> +++ b/drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx28.h >> @@ -0,0 +1,421 @@ >> +/* >> + * Freescale GPMI NFC NAND Flash Driver >> + * >> + * Copyright 2008-2010 Freescale Semiconductor, Inc. All Rights Reserved. >> + * >> + * This program is free software; you can redistribute it and/or modify >> + * it under the terms of the GNU General Public License as published by >> + * the Free Software Foundation; either version 2 of the License, or >> + * (at your option) any later version. >> + * >> + * This program is distributed in the hope that it will be useful, >> + * but WITHOUT ANY WARRANTY; without even the implied warranty of >> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the >> + * GNU General Public License for more details. >> + * >> + * You should have received a copy of the GNU General Public License >> + * along with this program; if not, write to the Free Software >> + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA >> + * >> + * Xml Revision: 2.2 >> + * Template revision: 26195 >> + */ >> + >> +#ifndef __GPMI_NFC_GPMI_REGS_H >> +#define __GPMI_NFC_GPMI_REGS_H >> + >> +/*============================================================================*/ >> + >> +#define HW_GPMI_CTRL0 (0x00000000) >> +#define HW_GPMI_CTRL0_SET (0x00000004) >> +#define HW_GPMI_CTRL0_CLR (0x00000008) >> +#define HW_GPMI_CTRL0_TOG (0x0000000c) >> + >> +#define BM_GPMI_CTRL0_SFTRST 0x80000000 >> +#define BV_GPMI_CTRL0_SFTRST__RUN 0x0 >> +#define BV_GPMI_CTRL0_SFTRST__RESET 0x1 >> +#define BM_GPMI_CTRL0_CLKGATE 0x40000000 >> +#define BV_GPMI_CTRL0_CLKGATE__RUN 0x0 >> +#define BV_GPMI_CTRL0_CLKGATE__NO_CLKS 0x1 >> +#define BM_GPMI_CTRL0_RUN 0x20000000 >> +#define BV_GPMI_CTRL0_RUN__IDLE 0x0 >> +#define BV_GPMI_CTRL0_RUN__BUSY 0x1 >> +#define BM_GPMI_CTRL0_DEV_IRQ_EN 0x10000000 >> +#define BM_GPMI_CTRL0_LOCK_CS 0x08000000 >> +#define BV_GPMI_CTRL0_LOCK_CS__DISABLED 0x0 >> +#define BV_GPMI_CTRL0_LOCK_CS__ENABLED 0x1 >> +#define BM_GPMI_CTRL0_UDMA 0x04000000 >> +#define BV_GPMI_CTRL0_UDMA__DISABLED 0x0 >> +#define BV_GPMI_CTRL0_UDMA__ENABLED 0x1 >> +#define BP_GPMI_CTRL0_COMMAND_MODE 24 >> +#define BM_GPMI_CTRL0_COMMAND_MODE 0x03000000 >> +#define BF_GPMI_CTRL0_COMMAND_MODE(v) \ >> + (((v)<< 24)& BM_GPMI_CTRL0_COMMAND_MODE) >> +#define BV_GPMI_CTRL0_COMMAND_MODE__WRITE 0x0 >> +#define BV_GPMI_CTRL0_COMMAND_MODE__READ 0x1 >> +#define BV_GPMI_CTRL0_COMMAND_MODE__READ_AND_COMPARE 0x2 >> +#define BV_GPMI_CTRL0_COMMAND_MODE__WAIT_FOR_READY 0x3 >> +#define BM_GPMI_CTRL0_WORD_LENGTH 0x00800000 >> +#define BV_GPMI_CTRL0_WORD_LENGTH__16_BIT 0x0 >> +#define BV_GPMI_CTRL0_WORD_LENGTH__8_BIT 0x1 >> +#define BP_GPMI_CTRL0_CS 20 >> +#define BM_GPMI_CTRL0_CS 0x00700000 >> +#define BF_GPMI_CTRL0_CS(v) \ >> + (((v)<< 20)& BM_GPMI_CTRL0_CS) >> +#define BP_GPMI_CTRL0_ADDRESS 17 >> +#define BM_GPMI_CTRL0_ADDRESS 0x000E0000 >> +#define BF_GPMI_CTRL0_ADDRESS(v) \ >> + (((v)<< 17)& BM_GPMI_CTRL0_ADDRESS) >> +#define BV_GPMI_CTRL0_ADDRESS__NAND_DATA 0x0 >> +#define BV_GPMI_CTRL0_ADDRESS__NAND_CLE 0x1 >> +#define BV_GPMI_CTRL0_ADDRESS__NAND_ALE 0x2 >> +#define BM_GPMI_CTRL0_ADDRESS_INCREMENT 0x00010000 >> +#define BV_GPMI_CTRL0_ADDRESS_INCREMENT__DISABLED 0x0 >> +#define BV_GPMI_CTRL0_ADDRESS_INCREMENT__ENABLED 0x1 >> +#define BP_GPMI_CTRL0_XFER_COUNT 0 >> +#define BM_GPMI_CTRL0_XFER_COUNT 0x0000FFFF >> +#define BF_GPMI_CTRL0_XFER_COUNT(v) \ >> + (((v)<< 0)& BM_GPMI_CTRL0_XFER_COUNT) >> > same as above. > >> diff --git a/drivers/mtd/nand/gpmi-nfc/hal-imx28.c b/drivers/mtd/nand/gpmi-nfc/hal-imx28.c >> new file mode 100644 >> index 0000000..ff87d7f >> --- /dev/null >> +++ b/drivers/mtd/nand/gpmi-nfc/hal-imx28.c >> @@ -0,0 +1,503 @@ >> +/* >> + * Freescale GPMI NFC NAND Flash Driver >> + * >> + * Copyright (C) 2011 Freescale Semiconductor, Inc. >> + * Copyright (C) 2008 Embedded Alley Solutions, Inc. >> + * >> + * This program is free software; you can redistribute it and/or modify >> + * it under the terms of the GNU General Public License as published by >> + * the Free Software Foundation; either version 2 of the License, or >> + * (at your option) any later version. >> + * >> + * This program is distributed in the hope that it will be useful, >> + * but WITHOUT ANY WARRANTY; without even the implied warranty of >> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the >> + * GNU General Public License for more details. >> + * >> + * You should have received a copy of the GNU General Public License along >> + * with this program; if not, write to the Free Software Foundation, Inc., >> + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. >> + */ >> +#include "gpmi-nfc.h" >> +#include "gpmi-regs-imx28.h" >> +#include "bch-regs-imx28.h" >> + > [...] >> +static int send_command(struct gpmi_nfc_data *this) >> +{ >> + struct dma_chan *channel = get_dma_chan(this); >> + struct mil *mil =&this->mil; >> + struct dma_async_tx_descriptor *desc; >> + struct scatterlist *sgl; >> + u32 pio[3]; >> + >> + /* [1] send out the PIO words */ >> + pio[0] = BF_GPMI_CTRL0_COMMAND_MODE(BV_GPMI_CTRL0_COMMAND_MODE__WRITE) >> + | BM_GPMI_CTRL0_WORD_LENGTH >> + | BF_GPMI_CTRL0_CS(mil->current_chip) >> + | BF_GPMI_CTRL0_ADDRESS(BV_GPMI_CTRL0_ADDRESS__NAND_CLE) >> + | BM_GPMI_CTRL0_ADDRESS_INCREMENT >> + | BF_GPMI_CTRL0_XFER_COUNT(mil->command_length); >> + pio[1] = pio[2] = 0; >> + desc = channel->device->device_prep_slave_sg(channel, >> + (struct scatterlist *)pio, >> + ARRAY_SIZE(pio), DMA_NONE, 0); >> + if (!desc) { >> + log("step 1 error"); >> + return -1; >> + } >> + >> + /* [2] send out the COMMAND + ADDRESS string stored in @buffer */ >> + sgl =&mil->cmd_sgl; >> + >> + dma_map_sg(this->dev, sgl, 1, DMA_TO_DEVICE); >> + sgl->length = mil->command_length; >> > this has to be done _before_ dma_map_sg(). With CONFIG_DMA_API_DEBUG > enabled you would get: > |gpmi-nfc gpmi-nfc: DMA-API: device driver frees DMA memory with different size [device address=0x0000000046e10000] [map size=4096 bytes] [unmap size=1 bytes] > thanks. Huang Shijie ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-23 3:11 ` [PATCH 5/7] add GPMI support for imx28 Huang Shijie @ 2011-03-23 14:56 ` Lothar Waßmann 2011-03-23 15:16 ` Florian Fainelli 2011-03-24 3:03 ` Huang Shijie 0 siblings, 2 replies; 21+ messages in thread From: Lothar Waßmann @ 2011-03-23 14:56 UTC (permalink / raw) To: Huang Shijie; +Cc: linux, linux-mtd, dwmw2, linux-arm-kernel Hi, Huang Shijie writes: > Hi Lothar: > > Thanks a lot for your carefully review. > > > > Hi, > > > > some general comments: > > - Why are you not using the existing nand_ids but inventing your own > > database? > > > The nand_ids{} contains poor information for me. > Such as : > [1]The nand_ids does not have the enough information for the page > size,oob size for some new NANDs. > And you can not get the information from parsing the NAND ids, such > as some Micron NANDs. > You could use it for the standard chips where you do not need additional information. That way most NAND chips could be supported out of the box without having to modify the driver. > [2]I need the timing information of the NAND. The nand_ids DOES not have > it. I have to > read the datasheet of the NAND, and add it to my database. > Do we really need exact timing data for each individual chip? Or couldn't we live with some sane timing that works for most chips and provide some means to specify a different timing via platform_data? > > - What is the purpose of the 'rom_helpers'? To me it looks like they > > are doing something that should be done in the mtd partition > > handlers, not in the flash chip driver. > > > We have saled a lot products based the imx23 or imx28. > All the products support the nand boot mode by flashing the kernel to > the kernel partitions of the nands. > In order to compliant to the old product, the layout of the nand is > constraint to the current situation: > [1] first a boot partition for the kernel. > [2] the rest is for other uses, such as rootfs. > > So I have to use the rom_helpers to split the kernel partition before > the real MTD partition initializations. > Why can't you just create appropriate partitions using the standard partition handlers? Lothar Waßmann -- ___________________________________________________________ Ka-Ro electronics GmbH | Pascalstraße 22 | D - 52076 Aachen Phone: +49 2408 1402-0 | Fax: +49 2408 1402-10 Geschäftsführer: Matthias Kaussen Handelsregistereintrag: Amtsgericht Aachen, HRB 4996 www.karo-electronics.de | info@karo-electronics.de ___________________________________________________________ ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-23 14:56 ` Lothar Waßmann @ 2011-03-23 15:16 ` Florian Fainelli 2011-03-24 3:07 ` Huang Shijie 2011-03-24 3:03 ` Huang Shijie 1 sibling, 1 reply; 21+ messages in thread From: Florian Fainelli @ 2011-03-23 15:16 UTC (permalink / raw) To: linux-arm-kernel; +Cc: Huang Shijie, linux, dwmw2, linux-mtd Hello, On Wednesday 23 March 2011 15:56:24 Lothar Waßmann wrote: > Hi, > > Huang Shijie writes: > > Hi Lothar: > > Thanks a lot for your carefully review. > > > > > > Hi, > > > > > > some general comments: > > > - Why are you not using the existing nand_ids but inventing your own > > > > > > database? > > > > The nand_ids{} contains poor information for me. > > Such as : > > [1]The nand_ids does not have the enough information for the page > > size,oob size for some new NANDs. > > > > And you can not get the information from parsing the NAND ids, such > > > > as some Micron NANDs. > > You could use it for the standard chips where you do not need > additional information. That way most NAND chips could be supported > out of the box without having to modify the driver. > > > [2]I need the timing information of the NAND. The nand_ids DOES not have > > it. I have to > > > > read the datasheet of the NAND, and add it to my database. > > Do we really need exact timing data for each individual chip? > Or couldn't we live with some sane timing that works for most chips > and provide some means to specify a different timing via > platform_data? I suppose this is the way it should be done, but then you might some need some sort of NAND detection available before registering the driver. Eventually, the bootloader might have configured the appropriate timings already. Have you considered using ONFI chips? They have standard timings at least. > > > > - What is the purpose of the 'rom_helpers'? To me it looks like they > > > > > > are doing something that should be done in the mtd partition > > > handlers, not in the flash chip driver. > > > > We have saled a lot products based the imx23 or imx28. > > All the products support the nand boot mode by flashing the kernel to > > the kernel partitions of the nands. > > In order to compliant to the old product, the layout of the nand is > > constraint to the current situation: > > [1] first a boot partition for the kernel. > > [2] the rest is for other uses, such as rootfs. > > > > So I have to use the rom_helpers to split the kernel partition before > > the real MTD partition initializations. > > Why can't you just create appropriate partitions using the standard > partition handlers? > > > Lothar Waßmann ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-23 15:16 ` Florian Fainelli @ 2011-03-24 3:07 ` Huang Shijie 0 siblings, 0 replies; 21+ messages in thread From: Huang Shijie @ 2011-03-24 3:07 UTC (permalink / raw) To: Florian Fainelli; +Cc: linux-mtd, linux, dwmw2, linux-arm-kernel Hi: >>>> Hi, >>>> >>>> some general comments: >>>> - Why are you not using the existing nand_ids but inventing your own >>>> >>>> database? >>> The nand_ids{} contains poor information for me. >>> Such as : >>> [1]The nand_ids does not have the enough information for the page >>> size,oob size for some new NANDs. >>> >>> And you can not get the information from parsing the NAND ids, such >>> >>> as some Micron NANDs. >> You could use it for the standard chips where you do not need >> additional information. That way most NAND chips could be supported >> out of the box without having to modify the driver. >> >>> [2]I need the timing information of the NAND. The nand_ids DOES not have >>> it. I have to >>> >>> read the datasheet of the NAND, and add it to my database. >> Do we really need exact timing data for each individual chip? >> Or couldn't we live with some sane timing that works for most chips >> and provide some means to specify a different timing via >> platform_data? > I suppose this is the way it should be done, but then you might some need some > sort of NAND detection available before registering the driver. Eventually, > the bootloader might have configured the appropriate timings already. The bootloader does not set it. > Have you considered using ONFI chips? They have standard timings at least. > The imx23/imx28 do not support ONFI chips in the DDR mode due to the hardware limits. The the imx508 does support the ONFI and TOGGLE nand in the DDR mode. Best Regards Huang Shijie ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-23 14:56 ` Lothar Waßmann 2011-03-23 15:16 ` Florian Fainelli @ 2011-03-24 3:03 ` Huang Shijie 2011-03-24 7:34 ` Lothar Waßmann 1 sibling, 1 reply; 21+ messages in thread From: Huang Shijie @ 2011-03-24 3:03 UTC (permalink / raw) To: Lothar Waßmann; +Cc: linux, linux-mtd, dwmw2, linux-arm-kernel Hi Lothar: > >>> some general comments: >>> - Why are you not using the existing nand_ids but inventing your own >>> database? >>> >> The nand_ids{} contains poor information for me. >> Such as : >> [1]The nand_ids does not have the enough information for the page >> size,oob size for some new NANDs. >> And you can not get the information from parsing the NAND ids, such >> as some Micron NANDs. >> > You could use it for the standard chips where you do not need > additional information. That way most NAND chips could be supported > out of the box without having to modify the driver. > What the meaning of "standard chips"? There are many nands in the world. Every vendor has its own rules, some even does has :) Unluckily, the imx23/imx28 supports many nands that the nand_ids{} does not support. I have many nands by my hand. I will add it gradually. I want to get the page size/oob size from my own database which can not get from the nand_ids. >> [2]I need the timing information of the NAND. The nand_ids DOES not have >> it. I have to >> read the datasheet of the NAND, and add it to my database. >> > Do we really need exact timing data for each individual chip? > Or couldn't we live with some sane timing that works for most chips > and provide some means to specify a different timing via > platform_data? > Most of the time, the timing is really based on a safe timing setting. But in the original GPMI driver in the FREESCALE BSP, there exits some nands need to be set with their own timing setting. So I do not use the safe timing for _ALL_ the nand, and i'd better get it from the database. >>> - What is the purpose of the 'rom_helpers'? To me it looks like they >>> are doing something that should be done in the mtd partition >>> handlers, not in the flash chip driver. >>> >> We have saled a lot products based the imx23 or imx28. >> All the products support the nand boot mode by flashing the kernel to >> the kernel partitions of the nands. >> In order to compliant to the old product, the layout of the nand is >> constraint to the current situation: >> [1] first a boot partition for the kernel. >> [2] the rest is for other uses, such as rootfs. >> >> So I have to use the rom_helpers to split the kernel partition before >> the real MTD partition initializations. >> > Why can't you just create appropriate partitions using the standard > partition handlers? > thanks , I will try to change the code according to your advice. Huang Shijie ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-24 3:03 ` Huang Shijie @ 2011-03-24 7:34 ` Lothar Waßmann 2011-03-24 8:26 ` Jason Liu 2011-03-24 8:51 ` Huang Shijie 0 siblings, 2 replies; 21+ messages in thread From: Lothar Waßmann @ 2011-03-24 7:34 UTC (permalink / raw) To: Huang Shijie; +Cc: linux, linux-mtd, dwmw2, linux-arm-kernel Hi, Huang Shijie writes: > Hi Lothar: > > > >>> some general comments: > >>> - Why are you not using the existing nand_ids but inventing your own > >>> database? > >>> > >> The nand_ids{} contains poor information for me. > >> Such as : > >> [1]The nand_ids does not have the enough information for the page > >> size,oob size for some new NANDs. > >> And you can not get the information from parsing the NAND ids, such > >> as some Micron NANDs. > >> > > You could use it for the standard chips where you do not need > > additional information. That way most NAND chips could be supported > > out of the box without having to modify the driver. > > > What the meaning of "standard chips"? > Those chips that are supported by other drivers for long time. > There are many nands in the world. Every vendor has its own rules, some > even does has :) > > Unluckily, the imx23/imx28 supports many nands that the nand_ids{} does > not support. > That's no reason to scrap support for all chips that every other driver supports. > I have many nands by my hand. I will add it gradually. > So why not add them to the generic database? > I want to get the page size/oob size from my own database which can not > get from the nand_ids. > If there are parameters needed that cannot be obtained from the generic database, it might be worth upgrading that database instead of creating your own local database. > >> [2]I need the timing information of the NAND. The nand_ids DOES not have > >> it. I have to > >> read the datasheet of the NAND, and add it to my database. > >> > > Do we really need exact timing data for each individual chip? > > Or couldn't we live with some sane timing that works for most chips > > and provide some means to specify a different timing via > > platform_data? > > > Most of the time, the timing is really based on a safe timing setting. > But in the original GPMI driver in the FREESCALE BSP, there exits some > nands need to be set with their own timing setting. > > So I do not use the safe timing for _ALL_ the nand, and i'd better get > it from the > database. > It should be sufficient to provide timing info from platform_data in special cases instead of bloating the nand id database with that stuff. Platforms might need to adjust the timing because of peculiarities in the HW. Thus the timing info should be provided from there, not from the chip database. Lothar Waßmann -- ___________________________________________________________ Ka-Ro electronics GmbH | Pascalstraße 22 | D - 52076 Aachen Phone: +49 2408 1402-0 | Fax: +49 2408 1402-10 Geschäftsführer: Matthias Kaussen Handelsregistereintrag: Amtsgericht Aachen, HRB 4996 www.karo-electronics.de | info@karo-electronics.de ___________________________________________________________ ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-24 7:34 ` Lothar Waßmann @ 2011-03-24 8:26 ` Jason Liu 2011-03-24 8:32 ` Wolfram Sang 2011-03-24 8:33 ` Lothar Waßmann 2011-03-24 8:51 ` Huang Shijie 1 sibling, 2 replies; 21+ messages in thread From: Jason Liu @ 2011-03-24 8:26 UTC (permalink / raw) To: Lothar Waßmann Cc: linux-arm-kernel, Huang Shijie, linux, dwmw2, linux-mtd 2011/3/24 Lothar Waßmann <LW@karo-electronics.de>: > Hi, [...] >> Most of the time, the timing is really based on a safe timing setting. >> But in the original GPMI driver in the FREESCALE BSP, there exits some >> nands need to be set with their own timing setting. >> >> So I do not use the safe timing for _ALL_ the nand, and i'd better get >> it from the >> database. >> > It should be sufficient to provide timing info from platform_data in > special cases instead of bloating the nand id database with that > stuff. Platforms might need to adjust the timing because of > peculiarities in the HW. Thus the timing info should be provided from > there, not from the chip database. No, we can't. since some boards only provide one NAND socket, customer can place any NAND flash chip they want. And with the NAND tech evolution, some NAND flash will end-of-line and customer need change another new NAND. As we all know that RAW NAND flash vendor does not follow the same rule to encode ID related information such as page size/oob size/block size etc. Even with same vendor, take Sumsung as example, there is not one generic rule to encode the id information. If you do NAND support with many customers, you will find that currenly linux NAND driver (nand_base.c/nand_ids.c) provide very poor support for that. Jason > > > Lothar Waßmann > -- > ___________________________________________________________ > > Ka-Ro electronics GmbH | Pascalstraße 22 | D - 52076 Aachen > Phone: +49 2408 1402-0 | Fax: +49 2408 1402-10 > Geschäftsführer: Matthias Kaussen > Handelsregistereintrag: Amtsgericht Aachen, HRB 4996 > > www.karo-electronics.de | info@karo-electronics.de > ___________________________________________________________ > > _______________________________________________ > linux-arm-kernel mailing list > linux-arm-kernel@lists.infradead.org > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel > ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-24 8:26 ` Jason Liu @ 2011-03-24 8:32 ` Wolfram Sang 2011-03-24 8:33 ` Lothar Waßmann 1 sibling, 0 replies; 21+ messages in thread From: Wolfram Sang @ 2011-03-24 8:32 UTC (permalink / raw) To: Jason Liu; +Cc: linux, Huang Shijie, linux-mtd, dwmw2, linux-arm-kernel [-- Attachment #1: Type: text/plain, Size: 415 bytes --] > If you do NAND support with many customers, you will find that > currenly linux NAND driver > (nand_base.c/nand_ids.c) provide very poor support for that. Then what about improving that instead of doing a private layer? Regards, Wolfram -- Pengutronix e.K. | Wolfram Sang | Industrial Linux Solutions | http://www.pengutronix.de/ | [-- Attachment #2: Digital signature --] [-- Type: application/pgp-signature, Size: 198 bytes --] ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-24 8:26 ` Jason Liu 2011-03-24 8:32 ` Wolfram Sang @ 2011-03-24 8:33 ` Lothar Waßmann 1 sibling, 0 replies; 21+ messages in thread From: Lothar Waßmann @ 2011-03-24 8:33 UTC (permalink / raw) To: Jason Liu; +Cc: linux-arm-kernel, Huang Shijie, linux, dwmw2, linux-mtd Hi, Jason Liu writes: > 2011/3/24 Lothar Waßmann <LW@karo-electronics.de>: > > Hi, > [...] > >> Most of the time, the timing is really based on a safe timing setting. > >> But in the original GPMI driver in the FREESCALE BSP, there exits some > >> nands need to be set with their own timing setting. > >> > >> So I do not use the safe timing for _ALL_ the nand, and i'd better get > >> it from the > >> database. > >> > > It should be sufficient to provide timing info from platform_data in > > special cases instead of bloating the nand id database with that > > stuff. Platforms might need to adjust the timing because of > > peculiarities in the HW. Thus the timing info should be provided from > > there, not from the chip database. > > No, we can't. since some boards only provide one NAND socket, customer > can place any NAND flash chip they want. And with the NAND tech evolution, some > NAND flash will end-of-line and customer need change another new NAND. > As we all know that RAW NAND flash vendor does not follow the same > rule to encode > ID related information such as page size/oob size/block size etc. Even > with same vendor, > take Sumsung as example, there is not one generic rule to encode the > id information. > If you do NAND support with many customers, you will find that > currenly linux NAND driver > (nand_base.c/nand_ids.c) provide very poor support for that. > That's a reason to improve that support, not to invent private fixes for each chip driver. Lothar Waßmann -- ___________________________________________________________ Ka-Ro electronics GmbH | Pascalstraße 22 | D - 52076 Aachen Phone: +49 2408 1402-0 | Fax: +49 2408 1402-10 Geschäftsführer: Matthias Kaussen Handelsregistereintrag: Amtsgericht Aachen, HRB 4996 www.karo-electronics.de | info@karo-electronics.de ___________________________________________________________ ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-24 7:34 ` Lothar Waßmann 2011-03-24 8:26 ` Jason Liu @ 2011-03-24 8:51 ` Huang Shijie 2011-03-24 13:54 ` Florian Fainelli 1 sibling, 1 reply; 21+ messages in thread From: Huang Shijie @ 2011-03-24 8:51 UTC (permalink / raw) To: Lothar Waßmann; +Cc: linux, linux-mtd, dwmw2, linux-arm-kernel Hi Lothar: > Hi, > > Huang Shijie writes: >> Hi Lothar: >>>>> some general comments: >>>>> - Why are you not using the existing nand_ids but inventing your own >>>>> database? >>>>> >>>> The nand_ids{} contains poor information for me. >>>> Such as : >>>> [1]The nand_ids does not have the enough information for the page >>>> size,oob size for some new NANDs. >>>> And you can not get the information from parsing the NAND ids, such >>>> as some Micron NANDs. >>>> >>> You could use it for the standard chips where you do not need >>> additional information. That way most NAND chips could be supported >>> out of the box without having to modify the driver. >>> >> What the meaning of "standard chips"? >> > Those chips that are supported by other drivers for long time. > see belowing. >> There are many nands in the world. Every vendor has its own rules, some >> even does has :) >> >> Unluckily, the imx23/imx28 supports many nands that the nand_ids{} does >> not support. >> > That's no reason to scrap support for all chips that every other > driver supports. > Frankly speaking, Which nand we should to support depends on the customer's requires. So we really do not want to support other nands that the customer does not have. >> I have many nands by my hand. I will add it gradually. >> > So why not add them to the generic database? > I ever thought to change the generic database. But I found it would cost a long time. The parsing code in nand_get_flash_type() can not parse out the page size/oob size in some cases. And IMHO I do not think to get the page size/oob size by parsing the ids is a good way, this really makes the code mess (sorry, David, I really think the current code is ugly.). IMHO, the best solution is add a database like mine. Using the id bytes as the keyword to match the nand. then to get the page size /oob size , etc, from the database. I will send a patch about this in future, but now, I do not want to be stumbled by this problem. >> I want to get the page size/oob size from my own database which can not >> get from the nand_ids. >> > If there are parameters needed that cannot be obtained from the > generic database, it might be worth upgrading that database instead of > creating your own local database. > >>>> [2]I need the timing information of the NAND. The nand_ids DOES not have >>>> it. I have to >>>> read the datasheet of the NAND, and add it to my database. >>>> >>> Do we really need exact timing data for each individual chip? >>> Or couldn't we live with some sane timing that works for most chips >>> and provide some means to specify a different timing via >>> platform_data? >>> >> Most of the time, the timing is really based on a safe timing setting. >> But in the original GPMI driver in the FREESCALE BSP, there exits some >> nands need to be set with their own timing setting. >> >> So I do not use the safe timing for _ALL_ the nand, and i'd better get >> it from the >> database. >> > It should be sufficient to provide timing info from platform_data in > special cases instead of bloating the nand id database with that > stuff. Platforms might need to adjust the timing because of > peculiarities in the HW. Thus the timing info should be provided from > there, not from the chip database. The timing information is needed by the GPMI module, not other module. So the best way to get this is from the chip database. Setting the GPMI with the proper timing will make the NAND runs in high performance. Thanks for your advice, I will use the safe timing for all the nand. Huang Shijie ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-24 8:51 ` Huang Shijie @ 2011-03-24 13:54 ` Florian Fainelli 2011-03-25 2:50 ` Huang Shijie 0 siblings, 1 reply; 21+ messages in thread From: Florian Fainelli @ 2011-03-24 13:54 UTC (permalink / raw) To: linux-arm-kernel; +Cc: Huang Shijie, linux, dwmw2, linux-mtd Hello Huang, On Thursday 24 March 2011 09:51:40 Huang Shijie wrote: > Hi Lothar: > > Hi, > > > > Huang Shijie writes: > >> Hi Lothar: > >>>>> some general comments: > >>>>> - Why are you not using the existing nand_ids but inventing your own > >>>>> > >>>>> database? > >>>> > >>>> The nand_ids{} contains poor information for me. > >>>> Such as : > >>>> [1]The nand_ids does not have the enough information for the page > >>>> size,oob size for some new NANDs. > >>>> > >>>> And you can not get the information from parsing the NAND ids, > >>>> such > >>>> > >>>> as some Micron NANDs. > >>> > >>> You could use it for the standard chips where you do not need > >>> additional information. That way most NAND chips could be supported > >>> out of the box without having to modify the driver. > >> > >> What the meaning of "standard chips"? > > > > Those chips that are supported by other drivers for long time. > > see belowing. > > >> There are many nands in the world. Every vendor has its own rules, some > >> even does has :) > >> > >> Unluckily, the imx23/imx28 supports many nands that the nand_ids{} does > >> not support. > > > > That's no reason to scrap support for all chips that every other > > driver supports. > > Frankly speaking, Which nand we should to support depends on the > customer's requires. > > So we really do not want to support other nands that the customer does > not have. > > >> I have many nands by my hand. I will add it gradually. > > > > So why not add them to the generic database? > > I ever thought to change the generic database. But I found it would cost > a long time. > > The parsing code in nand_get_flash_type() can not parse out the page > size/oob size > in some cases. And IMHO I do not think to get the page size/oob size by > parsing the ids is a good way, > this really makes the code mess (sorry, David, I really think the > current code is ugly.). > > IMHO, the best solution is add a database like mine. Using the id bytes > as the keyword to match the nand. > then to get the page size /oob size , etc, from the database. > > I will send a patch about this in future, but now, I do not want to be > stumbled by this problem. > > >> I want to get the page size/oob size from my own database which can not > >> get from the nand_ids. > > > > If there are parameters needed that cannot be obtained from the > > generic database, it might be worth upgrading that database instead of > > creating your own local database. > > > >>>> [2]I need the timing information of the NAND. The nand_ids DOES not > >>>> have it. I have to > >>>> > >>>> read the datasheet of the NAND, and add it to my database. > >>> > >>> Do we really need exact timing data for each individual chip? > >>> Or couldn't we live with some sane timing that works for most chips > >>> and provide some means to specify a different timing via > >>> platform_data? > >> > >> Most of the time, the timing is really based on a safe timing setting. > >> But in the original GPMI driver in the FREESCALE BSP, there exits some > >> nands need to be set with their own timing setting. > >> > >> So I do not use the safe timing for _ALL_ the nand, and i'd better get > >> it from the > >> database. > > > > It should be sufficient to provide timing info from platform_data in > > special cases instead of bloating the nand id database with that > > stuff. Platforms might need to adjust the timing because of > > peculiarities in the HW. Thus the timing info should be provided from > > there, not from the chip database. > > The timing information is needed by the GPMI module, not other module. > So the > best way to get this is from the chip database. Such a database already exists in drivers/mtd/nand/nand_ids.c. The idea here would be for you to get the NAND geometry characteristics from nands_ids.c, which benefits to all drivers in general, and let platform code define specific timings, which is specific to your driver. > > Setting the GPMI with the proper timing will make the NAND runs in high > performance. > > Thanks for your advice, I will use the safe timing for all the nand. If platform code does not define any timings, then yes, use the safest timings possible, if provided, let the platform configure its timings. -- Florian ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-24 13:54 ` Florian Fainelli @ 2011-03-25 2:50 ` Huang Shijie 0 siblings, 0 replies; 21+ messages in thread From: Huang Shijie @ 2011-03-25 2:50 UTC (permalink / raw) To: Florian Fainelli; +Cc: linux-mtd, linux, dwmw2, linux-arm-kernel Hi Florian: > Hello Huang, > > On Thursday 24 March 2011 09:51:40 Huang Shijie wrote: >> Hi Lothar: >>> Hi, >>> >>> Huang Shijie writes: >>>> Hi Lothar: >>>>>>> some general comments: >>>>>>> - Why are you not using the existing nand_ids but inventing your own >>>>>>> >>>>>>> database? >>>>>> The nand_ids{} contains poor information for me. >>>>>> Such as : >>>>>> [1]The nand_ids does not have the enough information for the page >>>>>> size,oob size for some new NANDs. >>>>>> >>>>>> And you can not get the information from parsing the NAND ids, >>>>>> such >>>>>> >>>>>> as some Micron NANDs. >>>>> You could use it for the standard chips where you do not need >>>>> additional information. That way most NAND chips could be supported >>>>> out of the box without having to modify the driver. >>>> What the meaning of "standard chips"? >>> Those chips that are supported by other drivers for long time. >> see belowing. >> >>>> There are many nands in the world. Every vendor has its own rules, some >>>> even does has :) >>>> >>>> Unluckily, the imx23/imx28 supports many nands that the nand_ids{} does >>>> not support. >>> That's no reason to scrap support for all chips that every other >>> driver supports. >> Frankly speaking, Which nand we should to support depends on the >> customer's requires. >> >> So we really do not want to support other nands that the customer does >> not have. >> >>>> I have many nands by my hand. I will add it gradually. >>> So why not add them to the generic database? >> I ever thought to change the generic database. But I found it would cost >> a long time. >> >> The parsing code in nand_get_flash_type() can not parse out the page >> size/oob size >> in some cases. And IMHO I do not think to get the page size/oob size by >> parsing the ids is a good way, >> this really makes the code mess (sorry, David, I really think the >> current code is ugly.). >> >> IMHO, the best solution is add a database like mine. Using the id bytes >> as the keyword to match the nand. >> then to get the page size /oob size , etc, from the database. >> >> I will send a patch about this in future, but now, I do not want to be >> stumbled by this problem. >> >>>> I want to get the page size/oob size from my own database which can not >>>> get from the nand_ids. >>> If there are parameters needed that cannot be obtained from the >>> generic database, it might be worth upgrading that database instead of >>> creating your own local database. >>> >>>>>> [2]I need the timing information of the NAND. The nand_ids DOES not >>>>>> have it. I have to >>>>>> >>>>>> read the datasheet of the NAND, and add it to my database. >>>>> Do we really need exact timing data for each individual chip? >>>>> Or couldn't we live with some sane timing that works for most chips >>>>> and provide some means to specify a different timing via >>>>> platform_data? >>>> Most of the time, the timing is really based on a safe timing setting. >>>> But in the original GPMI driver in the FREESCALE BSP, there exits some >>>> nands need to be set with their own timing setting. >>>> >>>> So I do not use the safe timing for _ALL_ the nand, and i'd better get >>>> it from the >>>> database. >>> It should be sufficient to provide timing info from platform_data in >>> special cases instead of bloating the nand id database with that >>> stuff. Platforms might need to adjust the timing because of >>> peculiarities in the HW. Thus the timing info should be provided from >>> there, not from the chip database. >> The timing information is needed by the GPMI module, not other module. >> So the >> best way to get this is from the chip database. > Such a database already exists in drivers/mtd/nand/nand_ids.c. The idea here > would be for you to get the NAND geometry characteristics from nands_ids.c, > which benefits to all drivers in general, and let platform code define specific > timings, which is specific to your driver. > I will remove the timing fields. thanks. ^ permalink raw reply [flat|nested] 21+ messages in thread
[parent not found: <1300239773-4222-3-git-send-email-b32955@freescale.com>]
[parent not found: <19848.38860.624803.339369@ipc1.ka-ro>]
* Re: [PATCH 2/7] add the common code for GPMI driver [not found] ` <19848.38860.624803.339369@ipc1.ka-ro> @ 2011-03-23 3:41 ` Huang Shijie 0 siblings, 0 replies; 21+ messages in thread From: Huang Shijie @ 2011-03-23 3:41 UTC (permalink / raw) To: Lothar Waßmann; +Cc: linux, linux-mtd, dwmw2, linux-arm-kernel Hi: >> + >> +static int acquire_register_block(struct gpmi_nfc_data *this, >> + const char *resource_name, void **reg_block_base) >> +{ >> + struct platform_device *pdev = this->pdev; >> + struct resource *r; >> + void *p; >> + >> + r = platform_get_resource_byname(pdev, IORESOURCE_MEM, resource_name); >> + if (!r) { >> + log("Can't get resource information for '%s'", resource_name); >> + return -ENXIO; >> + } >> + >> + /* remap the register block */ >> + p = ioremap(r->start, r->end - r->start + 1); > ^^^^^^^^^^^^^^^^^^^^^ > resource_size(r) > thanks. > >> +static int acquire_interrupt(struct gpmi_nfc_data *this, >> + const char *resource_name, >> + irq_handler_t interrupt_handler, int *lno, int *hno) >> +{ >> + struct platform_device *pdev = this->pdev; >> + struct resource *r; >> + int err = 0; >> > Useless initialization. > ok. >> + r = platform_get_resource_byname(pdev, IORESOURCE_IRQ, resource_name); >> + if (!r) { >> + log("Can't get resource information for '%s'", resource_name); >> + return -ENXIO; >> + } >> + >> + BUG_ON(r->start != r->end); >> + err = request_irq(r->start, interrupt_handler, 0, resource_name, this); >> + if (err) { >> + log("Can't own %s", resource_name); >> + return -EIO; >> > Promote the error code from request_irq(). > thanks. >> + } >> + >> + *lno = r->start; >> + *hno = r->end; >> + return 0; >> +} >> + >> +static void release_interrupt(struct gpmi_nfc_data *this, >> + int low_interrupt_number, int high_interrupt_number) >> +{ >> + int i; >> + for (i = low_interrupt_number; i<= high_interrupt_number; i++) >> + free_irq(i, this); >> +} >> + >> >> + >> +static void release_dma_channels(struct gpmi_nfc_data *this) >> +{ >> + unsigned int i; >> + for (i = 0; i<= DMA_CHANS; i++) >> > Obiwan error! should be 'i< DMA_CHANS' > thanks a lot! >> + if (this->dma_chans[i]) { >> + dma_release_channel(this->dma_chans[i]); >> + this->dma_chans[i] = NULL; >> + } >> +} >> + >> >> +static inline int acquire_clock(struct gpmi_nfc_data *this, >> + const char *clock_name, struct clk **clock) >> +{ >> + struct clk *c; >> + >> + c = clk_get(this->dev, clock_name); >> + if (IS_ERR(c)) { >> + log("Can't own clock %s", clock_name); >> + return PTR_ERR(c); >> + } >> + *clock = c; >> + return 0; >> +} >> + >> +static void release_clock(struct gpmi_nfc_data *this, struct clk *clock) >> +{ >> + clk_disable(clock); >> > Unbalanced clk_disable(). thanks. >> + clk_put(clock); >> +} >> + >> +static int acquire_resources(struct gpmi_nfc_data *this) >> +{ >> + struct gpmi_nfc_platform_data *pdata = this->pdata; >> + struct resources *resources =&this->resources; >> + int error = 0; >> > Useless initialization. ok. >> + >> + /* Attempt to acquire our clock. */ >> + error = acquire_clock(this, pdata->clock_name,&resources->clock); >> > Abuse of the clock API. Don't pass clock names via platform_data. > thanks, I will change it in next version. >> +static int set_up_nfc_hal(struct gpmi_nfc_data *this) >> +{ >> + struct gpmi_nfc_platform_data *pdata = this->pdata; >> + struct nfc_hal *nfc; >> + int error = 0; >> > Useless initialization. > >> + /* >> + * This structure contains the "safe" GPMI timing that should succeed >> + * with any NAND Flash device >> + * (although, with less-than-optimal performance). >> + */ >> + static struct nand_timing safe_timing = { >> + .data_setup_in_ns = 80, >> + .data_hold_in_ns = 60, >> + .address_setup_in_ns = 25, >> + .gpmi_sample_delay_in_ns = 6, >> + .tREA_in_ns = -1, >> + .tRLOH_in_ns = -1, >> + .tRHOH_in_ns = -1, >> + }; >> + >> + switch (pdata->chip_version) { >> + case CHIP_VERSION_IMX23: >> + nfc =&gpmi_nfc_hal_imx23; >> + break; >> + case CHIP_VERSION_IMX28: >> + nfc =&gpmi_nfc_hal_imx28; >> + break; >> + default: >> + log("Unkown NFC version %u", pdata->chip_version); >> + return -ENXIO; >> + } >> > Use platform_ids to distinguish between HW variants. > ok. >> + >> + this->nfc = nfc; >> + >> + /* Initialize the NFC HAL. */ >> + error = nfc->init(this); >> + if (error) >> + return error; >> + >> + /* Set up safe timing. */ >> + nfc->set_timing(this,&safe_timing); >> + return 0; >> +} >> + >> +static int set_up_boot_rom_helper(struct gpmi_nfc_data *this) >> +{ >> + struct gpmi_nfc_platform_data *pdata = this->pdata; >> + struct boot_rom_helper *rom; >> + >> + switch (pdata->chip_version) { >> + case CHIP_VERSION_IMX23: >> + rom =&gpmi_nfc_boot_rom_imx23; >> + break; >> + case CHIP_VERSION_IMX28: >> + rom =&gpmi_nfc_boot_rom_imx28; >> + break; >> + default: >> + log("Unkown NFC version %u", pdata->chip_version); >> + return -ENXIO; >> + } >> + >> + pr_info("Boot ROM: Version %u, %s\n", rom->version, rom->description); >> + this->rom = rom; >> + return 0; >> +} >> + >> +/* Creates/Removes sysfs files for this device.*/ >> +static void manage_sysfs_files(struct gpmi_nfc_data *this, int create) >> +{ >> + struct device *dev = this->dev; >> + int error; >> + unsigned int i; >> + struct device_attribute **attr; >> + >> + for (i = 0, attr = device_attributes; >> + i< ARRAY_SIZE(device_attributes); i++, attr++) { >> + >> + if (create) { >> + error = device_create_file(dev, *attr); >> + if (error) { >> + while (--attr>= device_attributes) >> + device_remove_file(dev, *attr); >> + return; >> + } >> + } else { >> + device_remove_file(dev, *attr); >> + } >> + } >> +} >> + >> +static int gpmi_nfc_probe(struct platform_device *pdev) >> +{ >> + struct gpmi_nfc_platform_data *pdata = pdev->dev.platform_data; >> + struct gpmi_nfc_data *this = 0; >> + int error = 0; >> > Useless initialization (NB: pointers are initialized with 'NULL' not > '0'). > thanks. I have already changed a lot, but i missed this one. >> +/* This structure represents this driver to the platform management system. */ >> +static struct platform_driver gpmi_nfc_driver = { >> + .driver = { >> + .name = GPMI_NFC_DRIVER_NAME, >> + }, >> + .probe = gpmi_nfc_probe, >> + .remove = __exit_p(gpmi_nfc_remove), >> + .suspend = gpmi_nfc_suspend, >> + .resume = gpmi_nfc_resume, >> +}; >> + >> +static int __init gpmi_nfc_init(void) >> +{ >> + printk(KERN_INFO "GPMI NFC driver registered. (IMX)\n"); >> + if (platform_driver_register(&gpmi_nfc_driver) != 0) { >> + pr_err("i.MX GPMI NFC driver registration failed\n"); >> + return -ENODEV; >> > Promote the error code from platform_driver_register(). > ok >> + } >> + return 0; >> +} >> + >> +static void __exit gpmi_nfc_exit(void) >> +{ >> + platform_driver_unregister(&gpmi_nfc_driver); >> +} >> + >> +module_init(gpmi_nfc_init); >> +module_exit(gpmi_nfc_exit); >> + >> +MODULE_AUTHOR("Freescale Semiconductor, Inc."); >> +MODULE_DESCRIPTION("i.MX GPMI NAND Flash Controller Driver"); >> +MODULE_LICENSE("GPL"); >> diff --git a/drivers/mtd/nand/gpmi-nfc/gpmi-nfc.h b/drivers/mtd/nand/gpmi-nfc/gpmi-nfc.h >> new file mode 100644 >> index 0000000..b5a2d77 >> --- /dev/null >> +++ b/drivers/mtd/nand/gpmi-nfc/gpmi-nfc.h >> @@ -0,0 +1,520 @@ >> +/* >> + * Freescale GPMI NFC NAND Flash Driver >> + * >> + * Copyright (C) 2010-2011 Freescale Semiconductor, Inc. >> + * Copyright (C) 2008 Embedded Alley Solutions, Inc. >> + * >> + * This program is free software; you can redistribute it and/or modify >> + * it under the terms of the GNU General Public License as published by >> + * the Free Software Foundation; either version 2 of the License, or >> + * (at your option) any later version. >> + * >> + * This program is distributed in the hope that it will be useful, >> + * but WITHOUT ANY WARRANTY; without even the implied warranty of >> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the >> + * GNU General Public License for more details. >> + * >> + * You should have received a copy of the GNU General Public License along >> + * with this program; if not, write to the Free Software Foundation, Inc., >> + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. >> + */ >> + >> +#ifndef __DRIVERS_MTD_NAND_GPMI_NFC_H >> +#define __DRIVERS_MTD_NAND_GPMI_NFC_H >> + >> +/* Linux header files. */ >> +#include<linux/err.h> >> +#include<linux/init.h> >> +#include<linux/module.h> >> +#include<linux/io.h> >> +#include<linux/interrupt.h> >> +#include<linux/clk.h> >> +#include<linux/delay.h> >> +#include<linux/platform_device.h> >> +#include<linux/dma-mapping.h> >> +#include<linux/mtd/mtd.h> >> +#include<linux/mtd/nand.h> >> +#include<linux/mtd/partitions.h> >> +#include<linux/mtd/concat.h> >> +#include<linux/dmaengine.h> >> +#include<asm/sizes.h> >> + >> +/* Platform header files. */ >> +#include<mach/mxs.h> >> +#include<mach/common.h> >> +#include<mach/dma.h> >> +#include<mach/gpmi-nfc.h> >> +#include<mach/system.h> >> +#include<mach/clock.h> >> + >> +/* Driver header files. */ >> +#include "nand_device_info.h" >> + >> +/* >> + *------------------------------------------------------------------------------ >> + * Fundamental Data Structures >> + *------------------------------------------------------------------------------ >> + */ >> + >> +/** >> + * struct resources - The collection of resources the driver needs. >> + * >> + * @gpmi_regs: A pointer to the GPMI registers. >> + * @bch_regs: A pointer to the BCH registers. >> + * @bch_interrupt: The BCH interrupt number. >> + * @dma_low_channel: The low DMA channel. >> + * @dma_high_channel: The high DMA channel. >> + * @clock: A pointer to the struct clk for the NFC's clock. >> + */ >> +struct resources { >> + void *gpmi_regs; >> + void *bch_regs; >> + unsigned int bch_low_interrupt; >> + unsigned int bch_high_interrupt; >> + unsigned int dma_low_channel; >> + unsigned int dma_high_channel; >> + struct clk *clock; >> +}; >> + >> +/** >> + * struct mil - State for the MTD Interface Layer. >> + * >> + * @nand: The NAND Flash MTD data structure that represents >> + * the NAND Flash medium. >> + * @mtd: The MTD data structure that represents the NAND >> + * Flash medium. >> + * @oob_layout: A structure that describes how bytes are laid out >> + * in the OOB. >> + * @general_use_mtd: A pointer to an MTD we export for general use. >> + * This *may* simply be a pointer to the mtd field, if >> + * we've been instructed NOT to protect the boot >> + * areas. >> + * @partitions: A pointer to a set of partitions applied to the >> + * general use MTD. >> + * @partition_count: The number of partitions. >> + * @current_chip: The chip currently selected by the NAND Fash MTD >> + * code. A negative value indicates that no chip is >> + * selected. >> + * @command_length: The length of the command that appears in the >> + * command buffer (see cmd_virt, below). >> + * @ignore_bad_block_marks: Indicates we are ignoring bad block marks. >> + * @saved_bbt: A saved pointer to the in-memory NAND Flash MTD bad >> + * block table. See show_device_ignorebad() for more >> + * details. >> + * @marking_a_bad_block: Indicates the caller is marking a bad block. See >> + * mil_ecc_write_oob() for details. >> + * @hooked_block_markbad: A pointer to the block_markbad() function we >> + * we "hooked." See mil_ecc_write_oob() for details. >> + * @upper_buf: The buffer passed from upper layer. >> + * @upper_len: The buffer len passed from upper layer. >> + * @direct_dma_map_ok: Is the direct DMA map is good for the upper_buf? >> + * @cmd_sgl/cmd_buffer: For NAND command. >> + * @data_sgl/data_buffer_dma:For NAND DATA ops. >> + * @page_buffer_virt: A pointer to a DMA-coherent buffer we use for >> + * reading and writing pages. This buffer includes >> + * space for both the payload data and the auxiliary >> + * data (including status bytes, but not syndrome >> + * bytes). >> + * @page_buffer_phys: The physical address for the page_buffer_virt >> + * buffer. >> + * @page_buffer_size: The size of the page buffer. >> + * @payload_virt: A pointer to a location in the page buffer used >> + * for payload bytes. The size of this buffer is >> + * determined by struct nfc_geometry. >> + * @payload_phys: The physical address for payload_virt. >> + * @auxiliary_virt: A pointer to a location in the page buffer used >> + * for auxiliary bytes. The size of this buffer is >> + * determined by struct nfc_geometry. >> + * @auxiliary_phys: The physical address for auxiliary_virt. >> + */ >> +struct mil { >> + /* MTD Data Structures */ >> + struct nand_chip nand; >> + struct mtd_info mtd; >> + struct nand_ecclayout oob_layout; >> + >> + /* Partitioning and Boot Area Protection */ >> + struct mtd_info *general_use_mtd; >> + struct mtd_partition *partitions; >> + unsigned int partition_count; >> + >> + /* General-use Variables */ >> + int current_chip; >> + unsigned int command_length; >> + int ignore_bad_block_marks; >> + void *saved_bbt; >> + >> + /* MTD Function Pointer Hooks */ >> + int marking_a_bad_block; >> + int (*hooked_block_markbad)(struct mtd_info *mtd, >> + loff_t ofs); >> + >> + /* from upper layer */ >> + uint8_t *upper_buf; >> + int upper_len; >> + >> + /* DMA */ >> + bool direct_dma_map_ok; >> + >> + struct scatterlist cmd_sgl; >> + char *cmd_buffer; >> + >> + struct scatterlist data_sgl; >> + char *data_buffer_dma; >> + >> + void *page_buffer_virt; >> + dma_addr_t page_buffer_phys; >> + unsigned int page_buffer_size; >> + >> + void *payload_virt; >> + dma_addr_t payload_phys; >> + >> + void *auxiliary_virt; >> + dma_addr_t auxiliary_phys; >> +}; >> > IMO this struct should be removed and the members integrated into the > struct gpmi_nfc_data. > >> + >> +/** >> + * struct nfc_geometry - NFC geometry description. >> + * >> + * This structure describes the NFC's view of the medium geometry. >> + * >> + * @ecc_algorithm: The human-readable name of the ECC algorithm >> + * (e.g., "Reed-Solomon" or "BCH"). >> + * @ecc_strength: A number that describes the strength of the ECC >> + * algorithm. >> + * @page_size_in_bytes: The size, in bytes, of a physical page, including >> + * both data and OOB. >> + * @metadata_size_in_bytes: The size, in bytes, of the metadata. >> + * @ecc_chunk_size_in_bytes: The size, in bytes, of a single ECC chunk. Note >> + * the first chunk in the page includes both data and >> + * metadata, so it's a bit larger than this value. >> + * @ecc_chunk_count: The number of ECC chunks in the page, >> + * @payload_size_in_bytes: The size, in bytes, of the payload buffer. >> + * @auxiliary_size_in_bytes: The size, in bytes, of the auxiliary buffer. >> + * @auxiliary_status_offset: The offset into the auxiliary buffer at which >> + * the ECC status appears. >> + * @block_mark_byte_offset: The byte offset in the ECC-based page view at >> + * which the underlying physical block mark appears. >> + * @block_mark_bit_offset: The bit offset into the ECC-based page view at >> + * which the underlying physical block mark appears. >> + */ >> +struct nfc_geometry { >> + char *ecc_algorithm; >> + unsigned int ecc_strength; >> + unsigned int page_size_in_bytes; >> + unsigned int metadata_size_in_bytes; >> + unsigned int ecc_chunk_size_in_bytes; >> + unsigned int ecc_chunk_count; >> + unsigned int payload_size_in_bytes; >> + unsigned int auxiliary_size_in_bytes; >> + unsigned int auxiliary_status_offset; >> + unsigned int block_mark_byte_offset; >> + unsigned int block_mark_bit_offset; >> +}; > Most of these parameters are already available in the mtd_info > structure. No need to duplicate them here! I will remove some ones. >> + >> +/** >> + * struct boot_rom_geometry - Boot ROM geometry description. >> + * >> + * This structure encapsulates decisions made by the Boot ROM Helper. >> + * >> + * @boot_area_count: The number of boot areas. The first boot area >> + * appears at the beginning of chip 0, the next >> + * at the beginning of chip 1, etc. >> + * @boot_area_size_in_bytes: The size, in bytes, of each boot area. >> + * @stride_size_in_pages: The size of a boot block stride, in pages. >> + * @search_area_stride_exponent: The logarithm to base 2 of the size of a >> + * search area in boot block strides. >> + */ >> +struct boot_rom_geometry { >> + unsigned int boot_area_count; >> + unsigned int boot_area_size_in_bytes; >> + unsigned int stride_size_in_pages; >> + unsigned int search_area_stride_exponent; >> +}; >> > IMO the whole boot_rom stuff should not be part of an mtd chip driver, > but has its place in the mtd partition handlers. > I have explained this in preview email. thanks >> +/* Can we use the upper's buffer directly for DMA? */ >> +void prepare_data_dma(struct gpmi_nfc_data *this, enum dma_data_direction dr) >> +{ >> + struct mil *mil =&this->mil; >> + struct scatterlist *sgl =&mil->data_sgl; >> + int ret = 0; >> + >> + mil->direct_dma_map_ok = true; >> + >> + /* first try to map the upper buffer directly */ >> + sg_init_one(sgl, mil->upper_buf, mil->upper_len); >> + ret = dma_map_sg(this->dev, sgl, 1, dr); >> + if (ret == 0) { >> + /* We have to use our own DMA buffer. */ >> + sg_init_one(sgl, mil->data_buffer_dma, PAGE_SIZE); >> + ret = dma_map_sg(this->dev, sgl, 1, dr); >> + BUG_ON(ret == 0); >> + >> + if (dr == DMA_TO_DEVICE) >> + memcpy(mil->data_buffer_dma, mil->upper_buf, >> + mil->upper_len); >> + mil->direct_dma_map_ok = false; >> + } >> + sgl->length = mil->upper_len; >> > This has already been done by sg_init_one(). > thanks. >> +} >> + >> +/* This will be called after the DMA operation is finished. */ >> +static void dma_irq_callback(void *param) >> +{ >> + struct gpmi_nfc_data *this = param; >> + struct nfc_hal *nfc = this->nfc; >> + struct mil *mil =&this->mil; >> + >> + complete(&nfc->dma_done); >> + >> + switch (this->dma_type) { >> + case DMA_FOR_COMMAND: >> + dma_unmap_sg(this->dev,&mil->cmd_sgl, 1, DMA_TO_DEVICE); >> + break; >> + >> + case DMA_FOR_READ_DATA: >> + if (mil->direct_dma_map_ok == false) >> + memcpy(mil->upper_buf, (char *)mil->data_buffer_dma, >> + mil->upper_len); >> + dma_unmap_sg(this->dev,&mil->data_sgl, 1, DMA_FROM_DEVICE); >> + break; >> + >> + case DMA_FOR_WRITE_DATA: >> + dma_unmap_sg(this->dev,&mil->data_sgl, 1, DMA_TO_DEVICE); >> + break; >> + >> + case DMA_FOR_READ_ECC_PAGE: >> + case DMA_FOR_WRITE_ECC_PAGE: >> + break; >> + >> + default: >> + BUG(); >> + } >> +} >> + >> >> +static uint8_t mil_read_byte(struct mtd_info *mtd) >> +{ >> + uint8_t byte = 0; >> + mil_read_buf(mtd, (uint8_t *)&byte, 1); >> > With CONFIG_DMA_API_DEBUG enabled this would blow up right in your > face: > |gpmi-nfc gpmi-nfc: DMA-API: device driver maps memory fromstack [addr=c732bd3f] > Thanks a lot. >> + return byte; >> +} >> + >> +/** >> + * mil_handle_block_mark_swapping() - Handles block mark swapping. >> + * >> + * Note that, when this function is called, it doesn't know whether it's >> + * swapping the block mark, or swapping it *back* -- but it doesn't matter >> + * because the the operation is the same. >> + * >> + * @this: Per-device data. >> + * @payload: A pointer to the payload buffer. >> + * @auxiliary: A pointer to the auxiliary buffer. >> + */ >> +static void mil_handle_block_mark_swapping(struct gpmi_nfc_data *this, >> + void *payload, void *auxiliary) >> +{ >> + struct nfc_geometry *nfc_geo =&this->nfc_geometry; >> + struct boot_rom_helper *rom = this->rom; >> + unsigned char *p; >> + unsigned char *a; >> + unsigned int bit; >> + unsigned char mask; >> + unsigned char from_data; >> + unsigned char from_oob; >> + >> + /* Check if we're doing block mark swapping. */ >> + if (!rom->swap_block_mark) >> + return; >> + >> + /* >> + * If control arrives here, we're swapping. Make some convenience >> + * variables. >> + */ >> + bit = nfc_geo->block_mark_bit_offset; >> + p = ((unsigned char *) payload) + nfc_geo->block_mark_byte_offset; >> > Useless cast. ok. >> + a = auxiliary; >> + >> + /* >> + * Get the byte from the data area that overlays the block mark. Since >> + * the ECC engine applies its own view to the bits in the page, the >> + * physical block mark won't (in general) appear on a byte boundary in >> + * the data. >> + */ >> + from_data = (p[0]>> bit) | (p[1]<< (8 - bit)); >> + >> + /* Get the byte from the OOB. */ >> + from_oob = a[0]; >> + >> + /* Swap them. */ >> + a[0] = from_data; >> + >> + mask = (0x1<< bit) - 1; >> + p[0] = (p[0]& mask) | (from_oob<< bit); >> + >> + mask = ~0<< bit; >> + p[1] = (p[1]& mask) | (from_oob>> (8 - bit)); >> +} >> + >> +static int mil_ecc_read_page(struct mtd_info *mtd, struct nand_chip *nand, >> + uint8_t *buf, int page) >> +{ >> + struct gpmi_nfc_data *this = nand->priv; >> + struct nfc_hal *nfc = this->nfc; >> + struct nfc_geometry *nfc_geo =&this->nfc_geometry; >> + struct mil *mil =&this->mil; >> + void *payload_virt = 0; >> + dma_addr_t payload_phys = ~0; >> + void *auxiliary_virt = 0; >> + dma_addr_t auxiliary_phys = ~0; > Useless initialization. > ok. >> + unsigned int i; >> + unsigned char *status; >> + unsigned int failed; >> + unsigned int corrected; >> + int error = 0; >> + > Dto. > >> + error = read_page_prepare(this, buf, mtd->writesize, >> + mil->payload_virt, mil->payload_phys, >> + nfc_geo->payload_size_in_bytes, >> + &payload_virt,&payload_phys); >> + if (error) { >> + log("Inadequate DMA buffer"); >> + error = -ENOMEM; >> + return error; >> + } >> + auxiliary_virt = mil->auxiliary_virt; >> + auxiliary_phys = mil->auxiliary_phys; >> + >> + /* ask the NFC */ >> + error = nfc->read_page(this, payload_phys, auxiliary_phys); >> + if (error) { >> + log("Error in ECC-based read: %d", error); >> + goto exit_nfc; >> + } >> + >> + /* handle the block mark swapping */ >> + mil_handle_block_mark_swapping(this, payload_virt, auxiliary_virt); >> + >> + /* Loop over status bytes, accumulating ECC status. */ >> + failed = 0; >> + corrected = 0; >> + status = ((unsigned char *) auxiliary_virt) + >> + nfc_geo->auxiliary_status_offset; >> > Useless cast. >> + >> + for (i = 0; i< nfc_geo->ecc_chunk_count; i++, status++) { >> + if ((*status == STATUS_GOOD) || (*status == STATUS_ERASED)) >> + continue; >> + >> + if (*status == STATUS_UNCORRECTABLE) { >> + failed++; >> + continue; >> + } >> + corrected += *status; >> + } >> + >> + /* >> + * Propagate ECC status to the owning MTD only when failed or >> + * corrected times nearly reaches our ECC correction threshold. >> + */ >> + if (failed || corrected>= (nfc_geo->ecc_strength - 1)) { >> + mtd->ecc_stats.failed += failed; >> + mtd->ecc_stats.corrected += corrected; >> + } >> + >> + /* >> + * It's time to deliver the OOB bytes. See mil_ecc_read_oob() for >> + * details about our policy for delivering the OOB. >> + * >> + * We fill the caller's buffer with set bits, and then copy the block >> + * mark to th caller's buffer. Note that, if block mark swapping was >> + * necessary, it has already been done, so we can rely on the first >> + * byte of the auxiliary buffer to contain the block mark. >> + */ >> + memset(nand->oob_poi, ~0, mtd->oobsize); >> + nand->oob_poi[0] = ((uint8_t *) auxiliary_virt)[0]; >> + >> +exit_nfc: >> + read_page_end(this, buf, mtd->writesize, >> + mil->payload_virt, mil->payload_phys, >> + nfc_geo->payload_size_in_bytes, >> + payload_virt, payload_phys); >> + return error; >> +} >> + >> +static void mil_ecc_write_page(struct mtd_info *mtd, >> + struct nand_chip *nand, const uint8_t *buf) >> +{ >> + struct gpmi_nfc_data *this = nand->priv; >> + struct nfc_hal *nfc = this->nfc; >> + struct nfc_geometry *nfc_geo =&this->nfc_geometry; >> + struct boot_rom_helper *rom = this->rom; >> + struct mil *mil =&this->mil; >> + const void *payload_virt = 0; >> + dma_addr_t payload_phys = ~0; >> + const void *auxiliary_virt = 0; >> + dma_addr_t auxiliary_phys = ~0; >> > Useless initialization. > >> + int error; >> + >> + if (rom->swap_block_mark) { >> + /* >> + * If control arrives here, we're doing block mark swapping. >> + * Since we can't modify the caller's buffers, we must copy them >> + * into our own. >> + */ >> + memcpy(mil->payload_virt, buf, mtd->writesize); >> + payload_virt = mil->payload_virt; >> + payload_phys = mil->payload_phys; >> + >> + memcpy(mil->auxiliary_virt, nand->oob_poi, >> + nfc_geo->auxiliary_size_in_bytes); >> + auxiliary_virt = mil->auxiliary_virt; >> + auxiliary_phys = mil->auxiliary_phys; >> + >> + /* Handle block mark swapping. */ >> + mil_handle_block_mark_swapping(this, >> + (void *) payload_virt, (void *) auxiliary_virt); >> + } else { >> + /* >> + * If control arrives here, we're not doing block mark swapping, >> + * so we can to try and use the caller's buffers. >> + */ >> + error = send_page_prepare(this, >> + buf, mtd->writesize, >> + mil->payload_virt, mil->payload_phys, >> + nfc_geo->payload_size_in_bytes, >> + &payload_virt,&payload_phys); >> + if (error) { >> + log("Inadequate payload DMA buffer"); >> + return; >> + } >> + >> + error = send_page_prepare(this, >> + nand->oob_poi, mtd->oobsize, >> + mil->auxiliary_virt, mil->auxiliary_phys, >> + nfc_geo->auxiliary_size_in_bytes, >> + &auxiliary_virt,&auxiliary_phys); >> + if (error) { >> + log("Inadequate auxiliary DMA buffer"); >> + goto exit_auxiliary; >> + } >> + } >> + >> + /* Ask the NFC. */ >> + error = nfc->send_page(this, payload_phys, auxiliary_phys); >> + if (error) >> + log("Error in ECC-based write: %d", error); >> + >> + if (!rom->swap_block_mark) { >> + send_page_end(this, nand->oob_poi, mtd->oobsize, >> + mil->auxiliary_virt, mil->auxiliary_phys, >> + nfc_geo->auxiliary_size_in_bytes, >> + auxiliary_virt, auxiliary_phys); >> +exit_auxiliary: >> + send_page_end(this, buf, mtd->writesize, >> + mil->payload_virt, mil->payload_phys, >> + nfc_geo->payload_size_in_bytes, >> + payload_virt, payload_phys); >> + } >> +} >> + >> +/** >> + * mil_hook_block_markbad() - Hooked MTD Interface block_markbad(). >> + * >> + * This function is a veneer that replaces the function originally installed by >> + * the NAND Flash MTD code. See the description of the marking_a_bad_block field >> + * in struct mil for more information about this. >> + * >> + * @mtd: A pointer to the MTD. >> + * @ofs: Byte address of the block to mark. >> + */ >> +static int mil_hook_block_markbad(struct mtd_info *mtd, loff_t ofs) >> +{ >> + register struct nand_chip *chip = mtd->priv; >> + struct gpmi_nfc_data *this = chip->priv; >> + struct mil *mil =&this->mil; >> + int ret; >> + >> + mil->marking_a_bad_block = true; >> + ret = mil->hooked_block_markbad(mtd, ofs); >> + mil->marking_a_bad_block = false; >> + return ret; >> +} >> + >> +/** >> + * mil_ecc_read_oob() - MTD Interface ecc.read_oob(). >> + * >> + * There are several places in this driver where we have to handle the OOB and >> + * block marks. This is the function where things are the most complicated, so >> + * this is where we try to explain it all. All the other places refer back to >> + * here. >> + * >> + * These are the rules, in order of decreasing importance: >> + * >> + * 1) Nothing the caller does can be allowed to imperil the block mark, so all >> + * write operations take measures to protect it. >> + * >> + * 2) In read operations, the first byte of the OOB we return must reflect the >> + * true state of the block mark, no matter where that block mark appears in >> + * the physical page. >> + * >> + * 3) ECC-based read operations return an OOB full of set bits (since we never >> + * allow ECC-based writes to the OOB, it doesn't matter what ECC-based reads >> + * return). >> + * >> + * 4) "Raw" read operations return a direct view of the physical bytes in the >> + * page, using the conventional definition of which bytes are data and which >> + * are OOB. This gives the caller a way to see the actual, physical bytes >> + * in the page, without the distortions applied by our ECC engine. >> + * >> + * >> + * What we do for this specific read operation depends on two questions: >> + * >> + * 1) Are we doing a "raw" read, or an ECC-based read? >> + * >> + * 2) Are we using block mark swapping or transcription? >> + * >> + * There are four cases, illustrated by the following Karnaugh map: >> + * >> + * | Raw | ECC-based | >> + * -------------+-------------------------+-------------------------+ >> + * | Read the conventional | | >> + * | OOB at the end of the | | >> + * Swapping | page and return it. It | | >> + * | contains exactly what | | >> + * | we want. | Read the block mark and | >> + * -------------+-------------------------+ return it in a buffer | >> + * | Read the conventional | full of set bits. | >> + * | OOB at the end of the | | >> + * | page and also the block | | >> + * Transcribing | mark in the metadata. | | >> + * | Copy the block mark | | >> + * | into the first byte of | | >> + * | the OOB. | | >> + * -------------+-------------------------+-------------------------+ >> + * >> + * Note that we break rule #4 in the Transcribing/Raw case because we're not >> + * giving an accurate view of the actual, physical bytes in the page (we're >> + * overwriting the block mark). That's OK because it's more important to follow >> + * rule #2. >> + * >> + * It turns out that knowing whether we want an "ECC-based" or "raw" read is not >> + * easy. When reading a page, for example, the NAND Flash MTD code calls our >> + * ecc.read_page or ecc.read_page_raw function. Thus, the fact that MTD wants an >> + * ECC-based or raw view of the page is implicit in which function it calls >> + * (there is a similar pair of ECC-based/raw functions for writing). >> + * >> + * Since MTD assumes the OOB is not covered by ECC, there is no pair of >> + * ECC-based/raw functions for reading or or writing the OOB. The fact that the >> + * caller wants an ECC-based or raw view of the page is not propagated down to >> + * this driver. >> + * >> + * @mtd: A pointer to the owning MTD. >> + * @nand: A pointer to the owning NAND Flash MTD. >> + * @page: The page number to read. >> + * @sndcmd: Indicates this function should send a command to the chip before >> + * reading the out-of-band bytes. This is only false for small page >> + * chips that support auto-increment. >> + */ >> +static int mil_ecc_read_oob(struct mtd_info *mtd, struct nand_chip *nand, >> + int page, int sndcmd) >> +{ >> + struct gpmi_nfc_data *this = nand->priv; >> + struct boot_rom_helper *rom = this->rom; >> + >> + /* clear the OOB buffer */ >> + memset(nand->oob_poi, ~0, mtd->oobsize); >> + >> + /* Read out the conventional OOB. */ >> + nand->cmdfunc(mtd, NAND_CMD_READ0, mtd->writesize, page); >> + nand->read_buf(mtd, nand->oob_poi, mtd->oobsize); >> + >> + /* >> + * Now, we want to make sure the block mark is correct. In the >> + * Swapping/Raw case, we already have it. Otherwise, we need to >> + * explicitly read it. >> + */ >> + if (!rom->swap_block_mark) { >> + /* Read the block mark into the first byte of the OOB buffer. */ >> + nand->cmdfunc(mtd, NAND_CMD_READ0, 0, page); >> + nand->oob_poi[0] = nand->read_byte(mtd); >> + } >> + >> + /* >> + * Return true, indicating that the next call to this function must send >> + * a command. >> + */ >> + return true; >> +} >> + >> +static int mil_ecc_write_oob(struct mtd_info *mtd, >> + struct nand_chip *nand, int page) >> +{ >> + struct gpmi_nfc_data *this = nand->priv; >> + struct device *dev = this->dev; >> + struct mil *mil =&this->mil; >> + struct boot_rom_helper *rom = this->rom; >> + uint8_t block_mark = 0; >> + int block_mark_column; >> + int status; >> + int error = 0; >> + >> + /* >> + * There are fundamental incompatibilities between the i.MX GPMI NFC and >> + * the NAND Flash MTD model that make it essentially impossible to write >> + * the out-of-band bytes. >> + * >> + * We permit *ONE* exception. If the *intent* of writing the OOB is to >> + * mark a block bad, we can do that. >> + */ >> + if (!mil->marking_a_bad_block) { >> + dev_emerg(dev, "This driver doesn't support writing the OOB\n"); >> + WARN_ON(1); >> + error = -EIO; >> + goto exit; >> + } >> + >> + /* >> + * If control arrives here, we're marking a block bad. First, figure out >> + * where the block mark is. >> + * >> + * If we're using swapping, the block mark is in the conventional >> + * location. Otherwise, we're using transcription, and the block mark >> + * appears in the first byte of the page. >> + */ >> + if (rom->swap_block_mark) >> + block_mark_column = mtd->writesize; >> + else >> + block_mark_column = 0; >> + >> + /* Write the block mark. */ >> + nand->cmdfunc(mtd, NAND_CMD_SEQIN, block_mark_column, page); >> + nand->write_buf(mtd,&block_mark, 1); >> > DMA mapping memory from stack again. thanks. >> + nand->cmdfunc(mtd, NAND_CMD_PAGEPROG, -1, -1); >> + >> + status = nand->waitfunc(mtd, nand); >> + >> + /* Check if it worked. */ >> + if (status& NAND_STATUS_FAIL) >> + error = -EIO; >> +exit: >> + return error; >> +} >> + >> +/** >> + * mil_block_bad - Claims all blocks are good. >> + * >> + * In principle, this function is *only* called when the NAND Flash MTD system >> + * isn't allowed to keep an in-memory bad block table, so it is forced to ask >> + * the driver for bad block information. >> + * >> + * In fact, we permit the NAND Flash MTD system to have an in-memory BBT, so >> + * this function is *only* called when we take it away. >> + * >> + * We take away the in-memory BBT when the user sets the "ignorebad" parameter, >> + * which indicates that all blocks should be reported good. >> + * >> + * Thus, this function is only called when we want *all* blocks to look good, >> + * so it *always* return success. >> + * >> + * @mtd: Ignored. >> + * @ofs: Ignored. >> + * @getchip: Ignored. >> + */ >> +static int mil_block_bad(struct mtd_info *mtd, loff_t ofs, int getchip) >> +{ >> + return 0; >> +} >> + >> +/* Set up the Boot ROM Helper geometry. */ >> +static int mil_set_boot_rom_helper_geometry(struct gpmi_nfc_data *this) >> +{ >> + struct boot_rom_helper *rom = this->rom; >> + struct boot_rom_geometry *geo =&this->rom_geometry; >> + >> + if (rom->set_geometry(this)) >> + return !0; >> + >> + pr_info("--------------------------------------------\n"); >> + pr_info(" Boot ROM Geometry\n"); >> + pr_info("--------------------------------------------\n"); >> + pr_info("Boot Area Count : %u\n", geo->boot_area_count); >> + pr_info("Boot Area Size in Bytes : %u (0x%x)\n", >> + geo->boot_area_size_in_bytes, >> + geo->boot_area_size_in_bytes); >> + pr_info("Stride Size in Pages : %u\n", geo->stride_size_in_pages); >> + pr_info("Search Area Stride Exponent: %u\n", >> + geo->search_area_stride_exponent); >> + return 0; >> +} >> + >> +static int mil_set_geometry(struct gpmi_nfc_data *this) >> +{ >> + struct nfc_hal *nfc = this->nfc; >> + struct nfc_geometry *geo =&this->nfc_geometry; >> + >> + /* Free the temporary DMA memory for read ID case */ >> + mil_free_dma_buffer(this); >> + >> + /* Set up the NFC geometry which is used by BCH. */ >> + if (nfc->set_geometry(this)) >> + return -ENXIO; >> > Promote the error code returned by set_geometry(). > ok. thanks. >> + pr_info("---------------------------------------\n"); >> + pr_info(" NFC Geometry (used by BCH)\n"); >> + pr_info("---------------------------------------\n"); >> + pr_info("ECC Algorithm : %s\n", geo->ecc_algorithm); >> + pr_info("ECC Strength : %u\n", geo->ecc_strength); >> + pr_info("Page Size in Bytes : %u\n", geo->page_size_in_bytes); >> + pr_info("Metadata Size in Bytes : %u\n", geo->metadata_size_in_bytes); >> + pr_info("ECC Chunk Size in Bytes: %u\n", geo->ecc_chunk_size_in_bytes); >> + pr_info("ECC Chunk Count : %u\n", geo->ecc_chunk_count); >> + pr_info("Payload Size in Bytes : %u\n", geo->payload_size_in_bytes); >> + pr_info("Auxiliary Size in Bytes: %u\n", geo->auxiliary_size_in_bytes); >> + pr_info("Auxiliary Status Offset: %u\n", geo->auxiliary_status_offset); >> + pr_info("Block Mark Byte Offset : %u\n", geo->block_mark_byte_offset); >> + pr_info("Block Mark Bit Offset : %u\n", geo->block_mark_bit_offset); >> + >> + /* Alloc the new DMA buffers according to the pagesize and oobsize */ >> + return mil_alloc_dma_buffer(this); >> +} >> + >> +static int mil_pre_bbt_scan(struct gpmi_nfc_data *this) >> +{ >> + struct boot_rom_helper *rom = this->rom; >> + int error = 0; >> + >> + if (mil_set_boot_rom_helper_geometry(this)) >> + return -ENXIO; >> > Promote the error code returned by mil_set_boot_rom_helper_geometry(). ok. Huang Shijie ^ permalink raw reply [flat|nested] 21+ messages in thread
* [PATCH 0/7] add the GPMI controller driver for IMX23/IMX28 @ 2011-03-16 1:55 Huang Shijie 2011-03-16 1:55 ` [PATCH 5/7] add GPMI support for imx28 Huang Shijie 0 siblings, 1 reply; 21+ messages in thread From: Huang Shijie @ 2011-03-16 1:55 UTC (permalink / raw) To: linux-arm-kernel; +Cc: linux, Huang Shijie, linux-mtd, dwmw2, David.Woodhouse The general-purpose media interface(GPMI) controller is a flexible interface to up to several NAND flashs. The Bose Ray-Choudhury Hocquenghem(BCH) module is a hardware ECC accelerator. With the help of BCH, the GPMI controller can choose to do the hardware ECC or not. This driver is based the Shawn Guo's DMA patches for IMX23/IMX28, please refer to : http://git.infradead.org/users/vkoul/slave-dma.git/commit/a580b8c5429a624d120cd603e1498bf676e2b4da add the mtd maillist. Huang Shijie (7): ARM: add GPMI support for imx23/imx28 add the common code for GPMI driver add the database for the NANDs add GPMI support for imx23 add GPMI support for imx28 dmaengine: change the flags of request_irq() MTD : add GPMI driver in the config and Makefile arch/arm/mach-mxs/Kconfig | 2 + arch/arm/mach-mxs/clock-mx23.c | 3 + arch/arm/mach-mxs/clock-mx28.c | 3 + arch/arm/mach-mxs/devices-mx23.h | 3 + arch/arm/mach-mxs/devices-mx28.h | 3 + arch/arm/mach-mxs/devices/Kconfig | 3 + arch/arm/mach-mxs/devices/Makefile | 1 + arch/arm/mach-mxs/devices/platform-gpmi.c | 144 +++ arch/arm/mach-mxs/include/mach/devices-common.h | 4 + arch/arm/mach-mxs/include/mach/gpmi-nfc.h | 75 ++ arch/arm/mach-mxs/mach-mx23evk.c | 37 + arch/arm/mach-mxs/mach-mx28evk.c | 37 + drivers/dma/mxs-dma.c | 2 +- drivers/mtd/nand/Kconfig | 10 + drivers/mtd/nand/Makefile | 1 + drivers/mtd/nand/gpmi-nfc/Makefile | 10 + drivers/mtd/nand/gpmi-nfc/bch-regs-imx23.h | 550 +++++++++++ drivers/mtd/nand/gpmi-nfc/bch-regs-imx28.h | 557 ++++++++++++ drivers/mtd/nand/gpmi-nfc/gpmi-nfc-main.c | 524 +++++++++++ drivers/mtd/nand/gpmi-nfc/gpmi-nfc.h | 520 +++++++++++ drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx23.h | 416 +++++++++ drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx28.h | 421 +++++++++ drivers/mtd/nand/gpmi-nfc/hal-common.c | 816 +++++++++++++++++ drivers/mtd/nand/gpmi-nfc/hal-imx23.c | 556 ++++++++++++ drivers/mtd/nand/gpmi-nfc/hal-imx28.c | 503 +++++++++++ drivers/mtd/nand/gpmi-nfc/mil.c | 1104 +++++++++++++++++++++++ drivers/mtd/nand/gpmi-nfc/nand_device_info.c | 181 ++++ drivers/mtd/nand/gpmi-nfc/nand_device_info.h | 145 +++ drivers/mtd/nand/gpmi-nfc/rom-common.c | 48 + drivers/mtd/nand/gpmi-nfc/rom-imx23.c | 300 ++++++ drivers/mtd/nand/gpmi-nfc/rom-imx28.c | 66 ++ 31 files changed, 7044 insertions(+), 1 deletions(-) create mode 100644 arch/arm/mach-mxs/devices/platform-gpmi.c create mode 100644 arch/arm/mach-mxs/include/mach/gpmi-nfc.h create mode 100644 drivers/mtd/nand/gpmi-nfc/Makefile create mode 100644 drivers/mtd/nand/gpmi-nfc/bch-regs-imx23.h create mode 100644 drivers/mtd/nand/gpmi-nfc/bch-regs-imx28.h create mode 100644 drivers/mtd/nand/gpmi-nfc/gpmi-nfc-main.c create mode 100644 drivers/mtd/nand/gpmi-nfc/gpmi-nfc.h create mode 100644 drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx23.h create mode 100644 drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx28.h create mode 100644 drivers/mtd/nand/gpmi-nfc/hal-common.c create mode 100644 drivers/mtd/nand/gpmi-nfc/hal-imx23.c create mode 100644 drivers/mtd/nand/gpmi-nfc/hal-imx28.c create mode 100644 drivers/mtd/nand/gpmi-nfc/mil.c create mode 100644 drivers/mtd/nand/gpmi-nfc/nand_device_info.c create mode 100644 drivers/mtd/nand/gpmi-nfc/nand_device_info.h create mode 100644 drivers/mtd/nand/gpmi-nfc/rom-common.c create mode 100644 drivers/mtd/nand/gpmi-nfc/rom-imx23.c create mode 100644 drivers/mtd/nand/gpmi-nfc/rom-imx28.c ^ permalink raw reply [flat|nested] 21+ messages in thread
* [PATCH 5/7] add GPMI support for imx28 2011-03-16 1:55 [PATCH 0/7] add the GPMI controller driver for IMX23/IMX28 Huang Shijie @ 2011-03-16 1:55 ` Huang Shijie 2011-03-31 9:47 ` Artem Bityutskiy ` (2 more replies) 0 siblings, 3 replies; 21+ messages in thread From: Huang Shijie @ 2011-03-16 1:55 UTC (permalink / raw) To: linux-arm-kernel; +Cc: linux, Huang Shijie, linux-mtd, dwmw2, David.Woodhouse These files contain the code to implement the GPMI in the imx28. Signed-off-by: Huang Shijie <b32955@freescale.com> --- drivers/mtd/nand/gpmi-nfc/bch-regs-imx28.h | 557 +++++++++++++++++++++++++++ drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx28.h | 421 ++++++++++++++++++++ drivers/mtd/nand/gpmi-nfc/hal-imx28.c | 503 ++++++++++++++++++++++++ drivers/mtd/nand/gpmi-nfc/rom-imx28.c | 66 ++++ 4 files changed, 1547 insertions(+), 0 deletions(-) create mode 100644 drivers/mtd/nand/gpmi-nfc/bch-regs-imx28.h create mode 100644 drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx28.h create mode 100644 drivers/mtd/nand/gpmi-nfc/hal-imx28.c create mode 100644 drivers/mtd/nand/gpmi-nfc/rom-imx28.c diff --git a/drivers/mtd/nand/gpmi-nfc/bch-regs-imx28.h b/drivers/mtd/nand/gpmi-nfc/bch-regs-imx28.h new file mode 100644 index 0000000..692db08 --- /dev/null +++ b/drivers/mtd/nand/gpmi-nfc/bch-regs-imx28.h @@ -0,0 +1,557 @@ +/* + * Freescale GPMI NFC NAND Flash Driver + * + * Copyright 2008-2010 Freescale Semiconductor, Inc. All Rights Reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * Xml Revision: 2.5 + * Template revision: 26195 + */ + +#ifndef __GPMI_NFC_BCH_REGS_H +#define __GPMI_NFC_BCH_REGS_H + +/*============================================================================*/ + +#define HW_BCH_CTRL (0x00000000) +#define HW_BCH_CTRL_SET (0x00000004) +#define HW_BCH_CTRL_CLR (0x00000008) +#define HW_BCH_CTRL_TOG (0x0000000c) + +#define BM_BCH_CTRL_SFTRST 0x80000000 +#define BV_BCH_CTRL_SFTRST__RUN 0x0 +#define BV_BCH_CTRL_SFTRST__RESET 0x1 +#define BM_BCH_CTRL_CLKGATE 0x40000000 +#define BV_BCH_CTRL_CLKGATE__RUN 0x0 +#define BV_BCH_CTRL_CLKGATE__NO_CLKS 0x1 +#define BP_BCH_CTRL_RSVD5 23 +#define BM_BCH_CTRL_RSVD5 0x3F800000 +#define BF_BCH_CTRL_RSVD5(v) \ + (((v) << 23) & BM_BCH_CTRL_RSVD5) +#define BM_BCH_CTRL_DEBUGSYNDROME 0x00400000 +#define BP_BCH_CTRL_RSVD4 20 +#define BM_BCH_CTRL_RSVD4 0x00300000 +#define BF_BCH_CTRL_RSVD4(v) \ + (((v) << 20) & BM_BCH_CTRL_RSVD4) +#define BP_BCH_CTRL_M2M_LAYOUT 18 +#define BM_BCH_CTRL_M2M_LAYOUT 0x000C0000 +#define BF_BCH_CTRL_M2M_LAYOUT(v) \ + (((v) << 18) & BM_BCH_CTRL_M2M_LAYOUT) +#define BM_BCH_CTRL_M2M_ENCODE 0x00020000 +#define BM_BCH_CTRL_M2M_ENABLE 0x00010000 +#define BP_BCH_CTRL_RSVD3 11 +#define BM_BCH_CTRL_RSVD3 0x0000F800 +#define BF_BCH_CTRL_RSVD3(v) \ + (((v) << 11) & BM_BCH_CTRL_RSVD3) +#define BM_BCH_CTRL_DEBUG_STALL_IRQ_EN 0x00000400 +#define BM_BCH_CTRL_RSVD2 0x00000200 +#define BM_BCH_CTRL_COMPLETE_IRQ_EN 0x00000100 +#define BP_BCH_CTRL_RSVD1 4 +#define BM_BCH_CTRL_RSVD1 0x000000F0 +#define BF_BCH_CTRL_RSVD1(v) \ + (((v) << 4) & BM_BCH_CTRL_RSVD1) +#define BM_BCH_CTRL_BM_ERROR_IRQ 0x00000008 +#define BM_BCH_CTRL_DEBUG_STALL_IRQ 0x00000004 +#define BM_BCH_CTRL_RSVD0 0x00000002 +#define BM_BCH_CTRL_COMPLETE_IRQ 0x00000001 + +/*============================================================================*/ + +#define HW_BCH_STATUS0 (0x00000010) + +#define BP_BCH_STATUS0_HANDLE 20 +#define BM_BCH_STATUS0_HANDLE 0xFFF00000 +#define BF_BCH_STATUS0_HANDLE(v) \ + (((v) << 20) & BM_BCH_STATUS0_HANDLE) +#define BP_BCH_STATUS0_COMPLETED_CE 16 +#define BM_BCH_STATUS0_COMPLETED_CE 0x000F0000 +#define BF_BCH_STATUS0_COMPLETED_CE(v) \ + (((v) << 16) & BM_BCH_STATUS0_COMPLETED_CE) +#define BP_BCH_STATUS0_STATUS_BLK0 8 +#define BM_BCH_STATUS0_STATUS_BLK0 0x0000FF00 +#define BF_BCH_STATUS0_STATUS_BLK0(v) \ + (((v) << 8) & BM_BCH_STATUS0_STATUS_BLK0) +#define BV_BCH_STATUS0_STATUS_BLK0__ZERO 0x00 +#define BV_BCH_STATUS0_STATUS_BLK0__ERROR1 0x01 +#define BV_BCH_STATUS0_STATUS_BLK0__ERROR2 0x02 +#define BV_BCH_STATUS0_STATUS_BLK0__ERROR3 0x03 +#define BV_BCH_STATUS0_STATUS_BLK0__ERROR4 0x04 +#define BV_BCH_STATUS0_STATUS_BLK0__UNCORRECTABLE 0xFE +#define BV_BCH_STATUS0_STATUS_BLK0__ERASED 0xFF +#define BP_BCH_STATUS0_RSVD1 5 +#define BM_BCH_STATUS0_RSVD1 0x000000E0 +#define BF_BCH_STATUS0_RSVD1(v) \ + (((v) << 5) & BM_BCH_STATUS0_RSVD1) +#define BM_BCH_STATUS0_ALLONES 0x00000010 +#define BM_BCH_STATUS0_CORRECTED 0x00000008 +#define BM_BCH_STATUS0_UNCORRECTABLE 0x00000004 +#define BP_BCH_STATUS0_RSVD0 0 +#define BM_BCH_STATUS0_RSVD0 0x00000003 +#define BF_BCH_STATUS0_RSVD0(v) \ + (((v) << 0) & BM_BCH_STATUS0_RSVD0) + +/*============================================================================*/ + +#define HW_BCH_MODE (0x00000020) + +#define BP_BCH_MODE_RSVD 8 +#define BM_BCH_MODE_RSVD 0xFFFFFF00 +#define BF_BCH_MODE_RSVD(v) \ + (((v) << 8) & BM_BCH_MODE_RSVD) +#define BP_BCH_MODE_ERASE_THRESHOLD 0 +#define BM_BCH_MODE_ERASE_THRESHOLD 0x000000FF +#define BF_BCH_MODE_ERASE_THRESHOLD(v) \ + (((v) << 0) & BM_BCH_MODE_ERASE_THRESHOLD) + +/*============================================================================*/ + +#define HW_BCH_ENCODEPTR (0x00000030) + +#define BP_BCH_ENCODEPTR_ADDR 0 +#define BM_BCH_ENCODEPTR_ADDR 0xFFFFFFFF +#define BF_BCH_ENCODEPTR_ADDR(v) (v) + +/*============================================================================*/ + +#define HW_BCH_DATAPTR (0x00000040) + +#define BP_BCH_DATAPTR_ADDR 0 +#define BM_BCH_DATAPTR_ADDR 0xFFFFFFFF +#define BF_BCH_DATAPTR_ADDR(v) (v) + +/*============================================================================*/ + +#define HW_BCH_METAPTR (0x00000050) + +#define BP_BCH_METAPTR_ADDR 0 +#define BM_BCH_METAPTR_ADDR 0xFFFFFFFF +#define BF_BCH_METAPTR_ADDR(v) (v) + +/*============================================================================*/ + +#define HW_BCH_LAYOUTSELECT (0x00000070) + +#define BP_BCH_LAYOUTSELECT_CS15_SELECT 30 +#define BM_BCH_LAYOUTSELECT_CS15_SELECT 0xC0000000 +#define BF_BCH_LAYOUTSELECT_CS15_SELECT(v) \ + (((v) << 30) & BM_BCH_LAYOUTSELECT_CS15_SELECT) +#define BP_BCH_LAYOUTSELECT_CS14_SELECT 28 +#define BM_BCH_LAYOUTSELECT_CS14_SELECT 0x30000000 +#define BF_BCH_LAYOUTSELECT_CS14_SELECT(v) \ + (((v) << 28) & BM_BCH_LAYOUTSELECT_CS14_SELECT) +#define BP_BCH_LAYOUTSELECT_CS13_SELECT 26 +#define BM_BCH_LAYOUTSELECT_CS13_SELECT 0x0C000000 +#define BF_BCH_LAYOUTSELECT_CS13_SELECT(v) \ + (((v) << 26) & BM_BCH_LAYOUTSELECT_CS13_SELECT) +#define BP_BCH_LAYOUTSELECT_CS12_SELECT 24 +#define BM_BCH_LAYOUTSELECT_CS12_SELECT 0x03000000 +#define BF_BCH_LAYOUTSELECT_CS12_SELECT(v) \ + (((v) << 24) & BM_BCH_LAYOUTSELECT_CS12_SELECT) +#define BP_BCH_LAYOUTSELECT_CS11_SELECT 22 +#define BM_BCH_LAYOUTSELECT_CS11_SELECT 0x00C00000 +#define BF_BCH_LAYOUTSELECT_CS11_SELECT(v) \ + (((v) << 22) & BM_BCH_LAYOUTSELECT_CS11_SELECT) +#define BP_BCH_LAYOUTSELECT_CS10_SELECT 20 +#define BM_BCH_LAYOUTSELECT_CS10_SELECT 0x00300000 +#define BF_BCH_LAYOUTSELECT_CS10_SELECT(v) \ + (((v) << 20) & BM_BCH_LAYOUTSELECT_CS10_SELECT) +#define BP_BCH_LAYOUTSELECT_CS9_SELECT 18 +#define BM_BCH_LAYOUTSELECT_CS9_SELECT 0x000C0000 +#define BF_BCH_LAYOUTSELECT_CS9_SELECT(v) \ + (((v) << 18) & BM_BCH_LAYOUTSELECT_CS9_SELECT) +#define BP_BCH_LAYOUTSELECT_CS8_SELECT 16 +#define BM_BCH_LAYOUTSELECT_CS8_SELECT 0x00030000 +#define BF_BCH_LAYOUTSELECT_CS8_SELECT(v) \ + (((v) << 16) & BM_BCH_LAYOUTSELECT_CS8_SELECT) +#define BP_BCH_LAYOUTSELECT_CS7_SELECT 14 +#define BM_BCH_LAYOUTSELECT_CS7_SELECT 0x0000C000 +#define BF_BCH_LAYOUTSELECT_CS7_SELECT(v) \ + (((v) << 14) & BM_BCH_LAYOUTSELECT_CS7_SELECT) +#define BP_BCH_LAYOUTSELECT_CS6_SELECT 12 +#define BM_BCH_LAYOUTSELECT_CS6_SELECT 0x00003000 +#define BF_BCH_LAYOUTSELECT_CS6_SELECT(v) \ + (((v) << 12) & BM_BCH_LAYOUTSELECT_CS6_SELECT) +#define BP_BCH_LAYOUTSELECT_CS5_SELECT 10 +#define BM_BCH_LAYOUTSELECT_CS5_SELECT 0x00000C00 +#define BF_BCH_LAYOUTSELECT_CS5_SELECT(v) \ + (((v) << 10) & BM_BCH_LAYOUTSELECT_CS5_SELECT) +#define BP_BCH_LAYOUTSELECT_CS4_SELECT 8 +#define BM_BCH_LAYOUTSELECT_CS4_SELECT 0x00000300 +#define BF_BCH_LAYOUTSELECT_CS4_SELECT(v) \ + (((v) << 8) & BM_BCH_LAYOUTSELECT_CS4_SELECT) +#define BP_BCH_LAYOUTSELECT_CS3_SELECT 6 +#define BM_BCH_LAYOUTSELECT_CS3_SELECT 0x000000C0 +#define BF_BCH_LAYOUTSELECT_CS3_SELECT(v) \ + (((v) << 6) & BM_BCH_LAYOUTSELECT_CS3_SELECT) +#define BP_BCH_LAYOUTSELECT_CS2_SELECT 4 +#define BM_BCH_LAYOUTSELECT_CS2_SELECT 0x00000030 +#define BF_BCH_LAYOUTSELECT_CS2_SELECT(v) \ + (((v) << 4) & BM_BCH_LAYOUTSELECT_CS2_SELECT) +#define BP_BCH_LAYOUTSELECT_CS1_SELECT 2 +#define BM_BCH_LAYOUTSELECT_CS1_SELECT 0x0000000C +#define BF_BCH_LAYOUTSELECT_CS1_SELECT(v) \ + (((v) << 2) & BM_BCH_LAYOUTSELECT_CS1_SELECT) +#define BP_BCH_LAYOUTSELECT_CS0_SELECT 0 +#define BM_BCH_LAYOUTSELECT_CS0_SELECT 0x00000003 +#define BF_BCH_LAYOUTSELECT_CS0_SELECT(v) \ + (((v) << 0) & BM_BCH_LAYOUTSELECT_CS0_SELECT) + +/*============================================================================*/ + +#define HW_BCH_FLASH0LAYOUT0 (0x00000080) + +#define BP_BCH_FLASH0LAYOUT0_NBLOCKS 24 +#define BM_BCH_FLASH0LAYOUT0_NBLOCKS 0xFF000000 +#define BF_BCH_FLASH0LAYOUT0_NBLOCKS(v) \ + (((v) << 24) & BM_BCH_FLASH0LAYOUT0_NBLOCKS) +#define BP_BCH_FLASH0LAYOUT0_META_SIZE 16 +#define BM_BCH_FLASH0LAYOUT0_META_SIZE 0x00FF0000 +#define BF_BCH_FLASH0LAYOUT0_META_SIZE(v) \ + (((v) << 16) & BM_BCH_FLASH0LAYOUT0_META_SIZE) +#define BP_BCH_FLASH0LAYOUT0_ECC0 12 +#define BM_BCH_FLASH0LAYOUT0_ECC0 0x0000F000 +#define BF_BCH_FLASH0LAYOUT0_ECC0(v) \ + (((v) << 12) & BM_BCH_FLASH0LAYOUT0_ECC0) +#define BV_BCH_FLASH0LAYOUT0_ECC0__NONE 0x0 +#define BV_BCH_FLASH0LAYOUT0_ECC0__ECC2 0x1 +#define BV_BCH_FLASH0LAYOUT0_ECC0__ECC4 0x2 +#define BV_BCH_FLASH0LAYOUT0_ECC0__ECC6 0x3 +#define BV_BCH_FLASH0LAYOUT0_ECC0__ECC8 0x4 +#define BV_BCH_FLASH0LAYOUT0_ECC0__ECC10 0x5 +#define BV_BCH_FLASH0LAYOUT0_ECC0__ECC12 0x6 +#define BV_BCH_FLASH0LAYOUT0_ECC0__ECC14 0x7 +#define BV_BCH_FLASH0LAYOUT0_ECC0__ECC16 0x8 +#define BV_BCH_FLASH0LAYOUT0_ECC0__ECC18 0x9 +#define BV_BCH_FLASH0LAYOUT0_ECC0__ECC20 0xA +#define BP_BCH_FLASH0LAYOUT0_DATA0_SIZE 0 +#define BM_BCH_FLASH0LAYOUT0_DATA0_SIZE 0x00000FFF +#define BF_BCH_FLASH0LAYOUT0_DATA0_SIZE(v) \ + (((v) << 0) & BM_BCH_FLASH0LAYOUT0_DATA0_SIZE) + +/*============================================================================*/ + +#define HW_BCH_FLASH0LAYOUT1 (0x00000090) + +#define BP_BCH_FLASH0LAYOUT1_PAGE_SIZE 16 +#define BM_BCH_FLASH0LAYOUT1_PAGE_SIZE 0xFFFF0000 +#define BF_BCH_FLASH0LAYOUT1_PAGE_SIZE(v) \ + (((v) << 16) & BM_BCH_FLASH0LAYOUT1_PAGE_SIZE) +#define BP_BCH_FLASH0LAYOUT1_ECCN 12 +#define BM_BCH_FLASH0LAYOUT1_ECCN 0x0000F000 +#define BF_BCH_FLASH0LAYOUT1_ECCN(v) \ + (((v) << 12) & BM_BCH_FLASH0LAYOUT1_ECCN) +#define BV_BCH_FLASH0LAYOUT1_ECCN__NONE 0x0 +#define BV_BCH_FLASH0LAYOUT1_ECCN__ECC2 0x1 +#define BV_BCH_FLASH0LAYOUT1_ECCN__ECC4 0x2 +#define BV_BCH_FLASH0LAYOUT1_ECCN__ECC6 0x3 +#define BV_BCH_FLASH0LAYOUT1_ECCN__ECC8 0x4 +#define BV_BCH_FLASH0LAYOUT1_ECCN__ECC10 0x5 +#define BV_BCH_FLASH0LAYOUT1_ECCN__ECC12 0x6 +#define BV_BCH_FLASH0LAYOUT1_ECCN__ECC14 0x7 +#define BV_BCH_FLASH0LAYOUT1_ECCN__ECC16 0x8 +#define BV_BCH_FLASH0LAYOUT1_ECCN__ECC18 0x9 +#define BV_BCH_FLASH0LAYOUT1_ECCN__ECC20 0xA +#define BP_BCH_FLASH0LAYOUT1_DATAN_SIZE 0 +#define BM_BCH_FLASH0LAYOUT1_DATAN_SIZE 0x00000FFF +#define BF_BCH_FLASH0LAYOUT1_DATAN_SIZE(v) \ + (((v) << 0) & BM_BCH_FLASH0LAYOUT1_DATAN_SIZE) + +/*============================================================================*/ + +#define HW_BCH_FLASH1LAYOUT0 (0x000000a0) + +#define BP_BCH_FLASH1LAYOUT0_NBLOCKS 24 +#define BM_BCH_FLASH1LAYOUT0_NBLOCKS 0xFF000000 +#define BF_BCH_FLASH1LAYOUT0_NBLOCKS(v) \ + (((v) << 24) & BM_BCH_FLASH1LAYOUT0_NBLOCKS) +#define BP_BCH_FLASH1LAYOUT0_META_SIZE 16 +#define BM_BCH_FLASH1LAYOUT0_META_SIZE 0x00FF0000 +#define BF_BCH_FLASH1LAYOUT0_META_SIZE(v) \ + (((v) << 16) & BM_BCH_FLASH1LAYOUT0_META_SIZE) +#define BP_BCH_FLASH1LAYOUT0_ECC0 12 +#define BM_BCH_FLASH1LAYOUT0_ECC0 0x0000F000 +#define BF_BCH_FLASH1LAYOUT0_ECC0(v) \ + (((v) << 12) & BM_BCH_FLASH1LAYOUT0_ECC0) +#define BV_BCH_FLASH1LAYOUT0_ECC0__NONE 0x0 +#define BV_BCH_FLASH1LAYOUT0_ECC0__ECC2 0x1 +#define BV_BCH_FLASH1LAYOUT0_ECC0__ECC4 0x2 +#define BV_BCH_FLASH1LAYOUT0_ECC0__ECC6 0x3 +#define BV_BCH_FLASH1LAYOUT0_ECC0__ECC8 0x4 +#define BV_BCH_FLASH1LAYOUT0_ECC0__ECC10 0x5 +#define BV_BCH_FLASH1LAYOUT0_ECC0__ECC12 0x6 +#define BV_BCH_FLASH1LAYOUT0_ECC0__ECC14 0x7 +#define BV_BCH_FLASH1LAYOUT0_ECC0__ECC16 0x8 +#define BV_BCH_FLASH1LAYOUT0_ECC0__ECC18 0x9 +#define BV_BCH_FLASH1LAYOUT0_ECC0__ECC20 0xA +#define BP_BCH_FLASH1LAYOUT0_DATA0_SIZE 0 +#define BM_BCH_FLASH1LAYOUT0_DATA0_SIZE 0x00000FFF +#define BF_BCH_FLASH1LAYOUT0_DATA0_SIZE(v) \ + (((v) << 0) & BM_BCH_FLASH1LAYOUT0_DATA0_SIZE) + +/*============================================================================*/ + +#define HW_BCH_FLASH1LAYOUT1 (0x000000b0) + +#define BP_BCH_FLASH1LAYOUT1_PAGE_SIZE 16 +#define BM_BCH_FLASH1LAYOUT1_PAGE_SIZE 0xFFFF0000 +#define BF_BCH_FLASH1LAYOUT1_PAGE_SIZE(v) \ + (((v) << 16) & BM_BCH_FLASH1LAYOUT1_PAGE_SIZE) +#define BP_BCH_FLASH1LAYOUT1_ECCN 12 +#define BM_BCH_FLASH1LAYOUT1_ECCN 0x0000F000 +#define BF_BCH_FLASH1LAYOUT1_ECCN(v) \ + (((v) << 12) & BM_BCH_FLASH1LAYOUT1_ECCN) +#define BV_BCH_FLASH1LAYOUT1_ECCN__NONE 0x0 +#define BV_BCH_FLASH1LAYOUT1_ECCN__ECC2 0x1 +#define BV_BCH_FLASH1LAYOUT1_ECCN__ECC4 0x2 +#define BV_BCH_FLASH1LAYOUT1_ECCN__ECC6 0x3 +#define BV_BCH_FLASH1LAYOUT1_ECCN__ECC8 0x4 +#define BV_BCH_FLASH1LAYOUT1_ECCN__ECC10 0x5 +#define BV_BCH_FLASH1LAYOUT1_ECCN__ECC12 0x6 +#define BV_BCH_FLASH1LAYOUT1_ECCN__ECC14 0x7 +#define BV_BCH_FLASH1LAYOUT1_ECCN__ECC16 0x8 +#define BV_BCH_FLASH1LAYOUT1_ECCN__ECC18 0x9 +#define BV_BCH_FLASH1LAYOUT1_ECCN__ECC20 0xA +#define BP_BCH_FLASH1LAYOUT1_DATAN_SIZE 0 +#define BM_BCH_FLASH1LAYOUT1_DATAN_SIZE 0x00000FFF +#define BF_BCH_FLASH1LAYOUT1_DATAN_SIZE(v) \ + (((v) << 0) & BM_BCH_FLASH1LAYOUT1_DATAN_SIZE) + +/*============================================================================*/ + +#define HW_BCH_FLASH2LAYOUT0 (0x000000c0) + +#define BP_BCH_FLASH2LAYOUT0_NBLOCKS 24 +#define BM_BCH_FLASH2LAYOUT0_NBLOCKS 0xFF000000 +#define BF_BCH_FLASH2LAYOUT0_NBLOCKS(v) \ + (((v) << 24) & BM_BCH_FLASH2LAYOUT0_NBLOCKS) +#define BP_BCH_FLASH2LAYOUT0_META_SIZE 16 +#define BM_BCH_FLASH2LAYOUT0_META_SIZE 0x00FF0000 +#define BF_BCH_FLASH2LAYOUT0_META_SIZE(v) \ + (((v) << 16) & BM_BCH_FLASH2LAYOUT0_META_SIZE) +#define BP_BCH_FLASH2LAYOUT0_ECC0 12 +#define BM_BCH_FLASH2LAYOUT0_ECC0 0x0000F000 +#define BF_BCH_FLASH2LAYOUT0_ECC0(v) \ + (((v) << 12) & BM_BCH_FLASH2LAYOUT0_ECC0) +#define BV_BCH_FLASH2LAYOUT0_ECC0__NONE 0x0 +#define BV_BCH_FLASH2LAYOUT0_ECC0__ECC2 0x1 +#define BV_BCH_FLASH2LAYOUT0_ECC0__ECC4 0x2 +#define BV_BCH_FLASH2LAYOUT0_ECC0__ECC6 0x3 +#define BV_BCH_FLASH2LAYOUT0_ECC0__ECC8 0x4 +#define BV_BCH_FLASH2LAYOUT0_ECC0__ECC10 0x5 +#define BV_BCH_FLASH2LAYOUT0_ECC0__ECC12 0x6 +#define BV_BCH_FLASH2LAYOUT0_ECC0__ECC14 0x7 +#define BV_BCH_FLASH2LAYOUT0_ECC0__ECC16 0x8 +#define BV_BCH_FLASH2LAYOUT0_ECC0__ECC18 0x9 +#define BV_BCH_FLASH2LAYOUT0_ECC0__ECC20 0xA +#define BP_BCH_FLASH2LAYOUT0_DATA0_SIZE 0 +#define BM_BCH_FLASH2LAYOUT0_DATA0_SIZE 0x00000FFF +#define BF_BCH_FLASH2LAYOUT0_DATA0_SIZE(v) \ + (((v) << 0) & BM_BCH_FLASH2LAYOUT0_DATA0_SIZE) + +/*============================================================================*/ + +#define HW_BCH_FLASH2LAYOUT1 (0x000000d0) + +#define BP_BCH_FLASH2LAYOUT1_PAGE_SIZE 16 +#define BM_BCH_FLASH2LAYOUT1_PAGE_SIZE 0xFFFF0000 +#define BF_BCH_FLASH2LAYOUT1_PAGE_SIZE(v) \ + (((v) << 16) & BM_BCH_FLASH2LAYOUT1_PAGE_SIZE) +#define BP_BCH_FLASH2LAYOUT1_ECCN 12 +#define BM_BCH_FLASH2LAYOUT1_ECCN 0x0000F000 +#define BF_BCH_FLASH2LAYOUT1_ECCN(v) \ + (((v) << 12) & BM_BCH_FLASH2LAYOUT1_ECCN) +#define BV_BCH_FLASH2LAYOUT1_ECCN__NONE 0x0 +#define BV_BCH_FLASH2LAYOUT1_ECCN__ECC2 0x1 +#define BV_BCH_FLASH2LAYOUT1_ECCN__ECC4 0x2 +#define BV_BCH_FLASH2LAYOUT1_ECCN__ECC6 0x3 +#define BV_BCH_FLASH2LAYOUT1_ECCN__ECC8 0x4 +#define BV_BCH_FLASH2LAYOUT1_ECCN__ECC10 0x5 +#define BV_BCH_FLASH2LAYOUT1_ECCN__ECC12 0x6 +#define BV_BCH_FLASH2LAYOUT1_ECCN__ECC14 0x7 +#define BV_BCH_FLASH2LAYOUT1_ECCN__ECC16 0x8 +#define BV_BCH_FLASH2LAYOUT1_ECCN__ECC18 0x9 +#define BV_BCH_FLASH2LAYOUT1_ECCN__ECC20 0xA +#define BP_BCH_FLASH2LAYOUT1_DATAN_SIZE 0 +#define BM_BCH_FLASH2LAYOUT1_DATAN_SIZE 0x00000FFF +#define BF_BCH_FLASH2LAYOUT1_DATAN_SIZE(v) \ + (((v) << 0) & BM_BCH_FLASH2LAYOUT1_DATAN_SIZE) + +/*============================================================================*/ + +#define HW_BCH_FLASH3LAYOUT0 (0x000000e0) + +#define BP_BCH_FLASH3LAYOUT0_NBLOCKS 24 +#define BM_BCH_FLASH3LAYOUT0_NBLOCKS 0xFF000000 +#define BF_BCH_FLASH3LAYOUT0_NBLOCKS(v) \ + (((v) << 24) & BM_BCH_FLASH3LAYOUT0_NBLOCKS) +#define BP_BCH_FLASH3LAYOUT0_META_SIZE 16 +#define BM_BCH_FLASH3LAYOUT0_META_SIZE 0x00FF0000 +#define BF_BCH_FLASH3LAYOUT0_META_SIZE(v) \ + (((v) << 16) & BM_BCH_FLASH3LAYOUT0_META_SIZE) +#define BP_BCH_FLASH3LAYOUT0_ECC0 12 +#define BM_BCH_FLASH3LAYOUT0_ECC0 0x0000F000 +#define BF_BCH_FLASH3LAYOUT0_ECC0(v) \ + (((v) << 12) & BM_BCH_FLASH3LAYOUT0_ECC0) +#define BV_BCH_FLASH3LAYOUT0_ECC0__NONE 0x0 +#define BV_BCH_FLASH3LAYOUT0_ECC0__ECC2 0x1 +#define BV_BCH_FLASH3LAYOUT0_ECC0__ECC4 0x2 +#define BV_BCH_FLASH3LAYOUT0_ECC0__ECC6 0x3 +#define BV_BCH_FLASH3LAYOUT0_ECC0__ECC8 0x4 +#define BV_BCH_FLASH3LAYOUT0_ECC0__ECC10 0x5 +#define BV_BCH_FLASH3LAYOUT0_ECC0__ECC12 0x6 +#define BV_BCH_FLASH3LAYOUT0_ECC0__ECC14 0x7 +#define BV_BCH_FLASH3LAYOUT0_ECC0__ECC16 0x8 +#define BV_BCH_FLASH3LAYOUT0_ECC0__ECC18 0x9 +#define BV_BCH_FLASH3LAYOUT0_ECC0__ECC20 0xA +#define BP_BCH_FLASH3LAYOUT0_DATA0_SIZE 0 +#define BM_BCH_FLASH3LAYOUT0_DATA0_SIZE 0x00000FFF +#define BF_BCH_FLASH3LAYOUT0_DATA0_SIZE(v) \ + (((v) << 0) & BM_BCH_FLASH3LAYOUT0_DATA0_SIZE) + +/*============================================================================*/ + +#define HW_BCH_FLASH3LAYOUT1 (0x000000f0) + +#define BP_BCH_FLASH3LAYOUT1_PAGE_SIZE 16 +#define BM_BCH_FLASH3LAYOUT1_PAGE_SIZE 0xFFFF0000 +#define BF_BCH_FLASH3LAYOUT1_PAGE_SIZE(v) \ + (((v) << 16) & BM_BCH_FLASH3LAYOUT1_PAGE_SIZE) +#define BP_BCH_FLASH3LAYOUT1_ECCN 12 +#define BM_BCH_FLASH3LAYOUT1_ECCN 0x0000F000 +#define BF_BCH_FLASH3LAYOUT1_ECCN(v) \ + (((v) << 12) & BM_BCH_FLASH3LAYOUT1_ECCN) +#define BV_BCH_FLASH3LAYOUT1_ECCN__NONE 0x0 +#define BV_BCH_FLASH3LAYOUT1_ECCN__ECC2 0x1 +#define BV_BCH_FLASH3LAYOUT1_ECCN__ECC4 0x2 +#define BV_BCH_FLASH3LAYOUT1_ECCN__ECC6 0x3 +#define BV_BCH_FLASH3LAYOUT1_ECCN__ECC8 0x4 +#define BV_BCH_FLASH3LAYOUT1_ECCN__ECC10 0x5 +#define BV_BCH_FLASH3LAYOUT1_ECCN__ECC12 0x6 +#define BV_BCH_FLASH3LAYOUT1_ECCN__ECC14 0x7 +#define BV_BCH_FLASH3LAYOUT1_ECCN__ECC16 0x8 +#define BV_BCH_FLASH3LAYOUT1_ECCN__ECC18 0x9 +#define BV_BCH_FLASH3LAYOUT1_ECCN__ECC20 0xA +#define BP_BCH_FLASH3LAYOUT1_DATAN_SIZE 0 +#define BM_BCH_FLASH3LAYOUT1_DATAN_SIZE 0x00000FFF +#define BF_BCH_FLASH3LAYOUT1_DATAN_SIZE(v) \ + (((v) << 0) & BM_BCH_FLASH3LAYOUT1_DATAN_SIZE) + +/*============================================================================*/ + +#define HW_BCH_DEBUG0 (0x00000100) +#define HW_BCH_DEBUG0_SET (0x00000104) +#define HW_BCH_DEBUG0_CLR (0x00000108) +#define HW_BCH_DEBUG0_TOG (0x0000010c) + +#define BP_BCH_DEBUG0_RSVD1 27 +#define BM_BCH_DEBUG0_RSVD1 0xF8000000 +#define BF_BCH_DEBUG0_RSVD1(v) \ + (((v) << 27) & BM_BCH_DEBUG0_RSVD1) +#define BM_BCH_DEBUG0_ROM_BIST_ENABLE 0x04000000 +#define BM_BCH_DEBUG0_ROM_BIST_COMPLETE 0x02000000 +#define BP_BCH_DEBUG0_KES_DEBUG_SYNDROME_SYMBOL 16 +#define BM_BCH_DEBUG0_KES_DEBUG_SYNDROME_SYMBOL 0x01FF0000 +#define BF_BCH_DEBUG0_KES_DEBUG_SYNDROME_SYMBOL(v) \ + (((v) << 16) & BM_BCH_DEBUG0_KES_DEBUG_SYNDROME_SYMBOL) +#define BV_BCH_DEBUG0_KES_DEBUG_SYNDROME_SYMBOL__NORMAL 0x0 +#define BV_BCH_DEBUG0_KES_DEBUG_SYNDROME_SYMBOL__TEST_MODE 0x1 +#define BM_BCH_DEBUG0_KES_DEBUG_SHIFT_SYND 0x00008000 +#define BM_BCH_DEBUG0_KES_DEBUG_PAYLOAD_FLAG 0x00004000 +#define BV_BCH_DEBUG0_KES_DEBUG_PAYLOAD_FLAG__DATA 0x1 +#define BV_BCH_DEBUG0_KES_DEBUG_PAYLOAD_FLAG__AUX 0x1 +#define BM_BCH_DEBUG0_KES_DEBUG_MODE4K 0x00002000 +#define BV_BCH_DEBUG0_KES_DEBUG_MODE4K__4k 0x1 +#define BV_BCH_DEBUG0_KES_DEBUG_MODE4K__2k 0x1 +#define BM_BCH_DEBUG0_KES_DEBUG_KICK 0x00001000 +#define BM_BCH_DEBUG0_KES_STANDALONE 0x00000800 +#define BV_BCH_DEBUG0_KES_STANDALONE__NORMAL 0x0 +#define BV_BCH_DEBUG0_KES_STANDALONE__TEST_MODE 0x1 +#define BM_BCH_DEBUG0_KES_DEBUG_STEP 0x00000400 +#define BM_BCH_DEBUG0_KES_DEBUG_STALL 0x00000200 +#define BV_BCH_DEBUG0_KES_DEBUG_STALL__NORMAL 0x0 +#define BV_BCH_DEBUG0_KES_DEBUG_STALL__WAIT 0x1 +#define BM_BCH_DEBUG0_BM_KES_TEST_BYPASS 0x00000100 +#define BV_BCH_DEBUG0_BM_KES_TEST_BYPASS__NORMAL 0x0 +#define BV_BCH_DEBUG0_BM_KES_TEST_BYPASS__TEST_MODE 0x1 +#define BP_BCH_DEBUG0_RSVD0 6 +#define BM_BCH_DEBUG0_RSVD0 0x000000C0 +#define BF_BCH_DEBUG0_RSVD0(v) \ + (((v) << 6) & BM_BCH_DEBUG0_RSVD0) +#define BP_BCH_DEBUG0_DEBUG_REG_SELECT 0 +#define BM_BCH_DEBUG0_DEBUG_REG_SELECT 0x0000003F +#define BF_BCH_DEBUG0_DEBUG_REG_SELECT(v) \ + (((v) << 0) & BM_BCH_DEBUG0_DEBUG_REG_SELECT) + +/*============================================================================*/ + +#define HW_BCH_DBGKESREAD (0x00000110) + +#define BP_BCH_DBGKESREAD_VALUES 0 +#define BM_BCH_DBGKESREAD_VALUES 0xFFFFFFFF +#define BF_BCH_DBGKESREAD_VALUES(v) (v) + +/*============================================================================*/ + +#define HW_BCH_DBGCSFEREAD (0x00000120) + +#define BP_BCH_DBGCSFEREAD_VALUES 0 +#define BM_BCH_DBGCSFEREAD_VALUES 0xFFFFFFFF +#define BF_BCH_DBGCSFEREAD_VALUES(v) (v) + +/*============================================================================*/ + +#define HW_BCH_DBGSYNDGENREAD (0x00000130) + +#define BP_BCH_DBGSYNDGENREAD_VALUES 0 +#define BM_BCH_DBGSYNDGENREAD_VALUES 0xFFFFFFFF +#define BF_BCH_DBGSYNDGENREAD_VALUES(v) (v) + +/*============================================================================*/ + +#define HW_BCH_DBGAHBMREAD (0x00000140) + +#define BP_BCH_DBGAHBMREAD_VALUES 0 +#define BM_BCH_DBGAHBMREAD_VALUES 0xFFFFFFFF +#define BF_BCH_DBGAHBMREAD_VALUES(v) (v) + +/*============================================================================*/ + +#define HW_BCH_BLOCKNAME (0x00000150) + +#define BP_BCH_BLOCKNAME_NAME 0 +#define BM_BCH_BLOCKNAME_NAME 0xFFFFFFFF +#define BF_BCH_BLOCKNAME_NAME(v) (v) + +/*============================================================================*/ + +#define HW_BCH_VERSION (0x00000160) + +#define BP_BCH_VERSION_MAJOR 24 +#define BM_BCH_VERSION_MAJOR 0xFF000000 +#define BF_BCH_VERSION_MAJOR(v) \ + (((v) << 24) & BM_BCH_VERSION_MAJOR) +#define BP_BCH_VERSION_MINOR 16 +#define BM_BCH_VERSION_MINOR 0x00FF0000 +#define BF_BCH_VERSION_MINOR(v) \ + (((v) << 16) & BM_BCH_VERSION_MINOR) +#define BP_BCH_VERSION_STEP 0 +#define BM_BCH_VERSION_STEP 0x0000FFFF +#define BF_BCH_VERSION_STEP(v) \ + (((v) << 0) & BM_BCH_VERSION_STEP) + +/*============================================================================*/ + +#endif diff --git a/drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx28.h b/drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx28.h new file mode 100644 index 0000000..dcb3b7d --- /dev/null +++ b/drivers/mtd/nand/gpmi-nfc/gpmi-regs-imx28.h @@ -0,0 +1,421 @@ +/* + * Freescale GPMI NFC NAND Flash Driver + * + * Copyright 2008-2010 Freescale Semiconductor, Inc. All Rights Reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * Xml Revision: 2.2 + * Template revision: 26195 + */ + +#ifndef __GPMI_NFC_GPMI_REGS_H +#define __GPMI_NFC_GPMI_REGS_H + +/*============================================================================*/ + +#define HW_GPMI_CTRL0 (0x00000000) +#define HW_GPMI_CTRL0_SET (0x00000004) +#define HW_GPMI_CTRL0_CLR (0x00000008) +#define HW_GPMI_CTRL0_TOG (0x0000000c) + +#define BM_GPMI_CTRL0_SFTRST 0x80000000 +#define BV_GPMI_CTRL0_SFTRST__RUN 0x0 +#define BV_GPMI_CTRL0_SFTRST__RESET 0x1 +#define BM_GPMI_CTRL0_CLKGATE 0x40000000 +#define BV_GPMI_CTRL0_CLKGATE__RUN 0x0 +#define BV_GPMI_CTRL0_CLKGATE__NO_CLKS 0x1 +#define BM_GPMI_CTRL0_RUN 0x20000000 +#define BV_GPMI_CTRL0_RUN__IDLE 0x0 +#define BV_GPMI_CTRL0_RUN__BUSY 0x1 +#define BM_GPMI_CTRL0_DEV_IRQ_EN 0x10000000 +#define BM_GPMI_CTRL0_LOCK_CS 0x08000000 +#define BV_GPMI_CTRL0_LOCK_CS__DISABLED 0x0 +#define BV_GPMI_CTRL0_LOCK_CS__ENABLED 0x1 +#define BM_GPMI_CTRL0_UDMA 0x04000000 +#define BV_GPMI_CTRL0_UDMA__DISABLED 0x0 +#define BV_GPMI_CTRL0_UDMA__ENABLED 0x1 +#define BP_GPMI_CTRL0_COMMAND_MODE 24 +#define BM_GPMI_CTRL0_COMMAND_MODE 0x03000000 +#define BF_GPMI_CTRL0_COMMAND_MODE(v) \ + (((v) << 24) & BM_GPMI_CTRL0_COMMAND_MODE) +#define BV_GPMI_CTRL0_COMMAND_MODE__WRITE 0x0 +#define BV_GPMI_CTRL0_COMMAND_MODE__READ 0x1 +#define BV_GPMI_CTRL0_COMMAND_MODE__READ_AND_COMPARE 0x2 +#define BV_GPMI_CTRL0_COMMAND_MODE__WAIT_FOR_READY 0x3 +#define BM_GPMI_CTRL0_WORD_LENGTH 0x00800000 +#define BV_GPMI_CTRL0_WORD_LENGTH__16_BIT 0x0 +#define BV_GPMI_CTRL0_WORD_LENGTH__8_BIT 0x1 +#define BP_GPMI_CTRL0_CS 20 +#define BM_GPMI_CTRL0_CS 0x00700000 +#define BF_GPMI_CTRL0_CS(v) \ + (((v) << 20) & BM_GPMI_CTRL0_CS) +#define BP_GPMI_CTRL0_ADDRESS 17 +#define BM_GPMI_CTRL0_ADDRESS 0x000E0000 +#define BF_GPMI_CTRL0_ADDRESS(v) \ + (((v) << 17) & BM_GPMI_CTRL0_ADDRESS) +#define BV_GPMI_CTRL0_ADDRESS__NAND_DATA 0x0 +#define BV_GPMI_CTRL0_ADDRESS__NAND_CLE 0x1 +#define BV_GPMI_CTRL0_ADDRESS__NAND_ALE 0x2 +#define BM_GPMI_CTRL0_ADDRESS_INCREMENT 0x00010000 +#define BV_GPMI_CTRL0_ADDRESS_INCREMENT__DISABLED 0x0 +#define BV_GPMI_CTRL0_ADDRESS_INCREMENT__ENABLED 0x1 +#define BP_GPMI_CTRL0_XFER_COUNT 0 +#define BM_GPMI_CTRL0_XFER_COUNT 0x0000FFFF +#define BF_GPMI_CTRL0_XFER_COUNT(v) \ + (((v) << 0) & BM_GPMI_CTRL0_XFER_COUNT) + +/*============================================================================*/ + +#define HW_GPMI_COMPARE (0x00000010) + +#define BP_GPMI_COMPARE_MASK 16 +#define BM_GPMI_COMPARE_MASK 0xFFFF0000 +#define BF_GPMI_COMPARE_MASK(v) \ + (((v) << 16) & BM_GPMI_COMPARE_MASK) +#define BP_GPMI_COMPARE_REFERENCE 0 +#define BM_GPMI_COMPARE_REFERENCE 0x0000FFFF +#define BF_GPMI_COMPARE_REFERENCE(v) \ + (((v) << 0) & BM_GPMI_COMPARE_REFERENCE) + +/*============================================================================*/ + +#define HW_GPMI_ECCCTRL (0x00000020) +#define HW_GPMI_ECCCTRL_SET (0x00000024) +#define HW_GPMI_ECCCTRL_CLR (0x00000028) +#define HW_GPMI_ECCCTRL_TOG (0x0000002c) + +#define BP_GPMI_ECCCTRL_HANDLE 16 +#define BM_GPMI_ECCCTRL_HANDLE 0xFFFF0000 +#define BF_GPMI_ECCCTRL_HANDLE(v) \ + (((v) << 16) & BM_GPMI_ECCCTRL_HANDLE) +#define BM_GPMI_ECCCTRL_RSVD2 0x00008000 +#define BP_GPMI_ECCCTRL_ECC_CMD 13 +#define BM_GPMI_ECCCTRL_ECC_CMD 0x00006000 +#define BF_GPMI_ECCCTRL_ECC_CMD(v) \ + (((v) << 13) & BM_GPMI_ECCCTRL_ECC_CMD) +#define BV_GPMI_ECCCTRL_ECC_CMD__DECODE 0x0 +#define BV_GPMI_ECCCTRL_ECC_CMD__ENCODE 0x1 +#define BV_GPMI_ECCCTRL_ECC_CMD__RESERVE2 0x2 +#define BV_GPMI_ECCCTRL_ECC_CMD__RESERVE3 0x3 +#define BM_GPMI_ECCCTRL_ENABLE_ECC 0x00001000 +#define BV_GPMI_ECCCTRL_ENABLE_ECC__ENABLE 0x1 +#define BV_GPMI_ECCCTRL_ENABLE_ECC__DISABLE 0x0 +#define BP_GPMI_ECCCTRL_RSVD1 9 +#define BM_GPMI_ECCCTRL_RSVD1 0x00000E00 +#define BF_GPMI_ECCCTRL_RSVD1(v) \ + (((v) << 9) & BM_GPMI_ECCCTRL_RSVD1) +#define BP_GPMI_ECCCTRL_BUFFER_MASK 0 +#define BM_GPMI_ECCCTRL_BUFFER_MASK 0x000001FF +#define BF_GPMI_ECCCTRL_BUFFER_MASK(v) \ + (((v) << 0) & BM_GPMI_ECCCTRL_BUFFER_MASK) +#define BV_GPMI_ECCCTRL_BUFFER_MASK__BCH_AUXONLY 0x100 +#define BV_GPMI_ECCCTRL_BUFFER_MASK__BCH_PAGE 0x1FF + +/*============================================================================*/ + +#define HW_GPMI_ECCCOUNT (0x00000030) + +#define BP_GPMI_ECCCOUNT_RSVD2 16 +#define BM_GPMI_ECCCOUNT_RSVD2 0xFFFF0000 +#define BF_GPMI_ECCCOUNT_RSVD2(v) \ + (((v) << 16) & BM_GPMI_ECCCOUNT_RSVD2) +#define BP_GPMI_ECCCOUNT_COUNT 0 +#define BM_GPMI_ECCCOUNT_COUNT 0x0000FFFF +#define BF_GPMI_ECCCOUNT_COUNT(v) \ + (((v) << 0) & BM_GPMI_ECCCOUNT_COUNT) + +/*============================================================================*/ + +#define HW_GPMI_PAYLOAD (0x00000040) + +#define BP_GPMI_PAYLOAD_ADDRESS 2 +#define BM_GPMI_PAYLOAD_ADDRESS 0xFFFFFFFC +#define BF_GPMI_PAYLOAD_ADDRESS(v) \ + (((v) << 2) & BM_GPMI_PAYLOAD_ADDRESS) +#define BP_GPMI_PAYLOAD_RSVD0 0 +#define BM_GPMI_PAYLOAD_RSVD0 0x00000003 +#define BF_GPMI_PAYLOAD_RSVD0(v) \ + (((v) << 0) & BM_GPMI_PAYLOAD_RSVD0) + +/*============================================================================*/ + +#define HW_GPMI_AUXILIARY (0x00000050) + +#define BP_GPMI_AUXILIARY_ADDRESS 2 +#define BM_GPMI_AUXILIARY_ADDRESS 0xFFFFFFFC +#define BF_GPMI_AUXILIARY_ADDRESS(v) \ + (((v) << 2) & BM_GPMI_AUXILIARY_ADDRESS) +#define BP_GPMI_AUXILIARY_RSVD0 0 +#define BM_GPMI_AUXILIARY_RSVD0 0x00000003 +#define BF_GPMI_AUXILIARY_RSVD0(v) \ + (((v) << 0) & BM_GPMI_AUXILIARY_RSVD0) + +/*============================================================================*/ + +#define HW_GPMI_CTRL1 (0x00000060) +#define HW_GPMI_CTRL1_SET (0x00000064) +#define HW_GPMI_CTRL1_CLR (0x00000068) +#define HW_GPMI_CTRL1_TOG (0x0000006c) + +#define BP_GPMI_CTRL1_RSVD2 25 +#define BM_GPMI_CTRL1_RSVD2 0xFE000000 +#define BF_GPMI_CTRL1_RSVD2(v) \ + (((v) << 25) & BM_GPMI_CTRL1_RSVD2) +#define BM_GPMI_CTRL1_DECOUPLE_CS 0x01000000 +#define BP_GPMI_CTRL1_WRN_DLY_SEL 22 +#define BM_GPMI_CTRL1_WRN_DLY_SEL 0x00C00000 +#define BF_GPMI_CTRL1_WRN_DLY_SEL(v) \ + (((v) << 22) & BM_GPMI_CTRL1_WRN_DLY_SEL) +#define BM_GPMI_CTRL1_RSVD1 0x00200000 +#define BM_GPMI_CTRL1_TIMEOUT_IRQ_EN 0x00100000 +#define BM_GPMI_CTRL1_GANGED_RDYBUSY 0x00080000 +#define BM_GPMI_CTRL1_BCH_MODE 0x00040000 +#define BP_GPMI_CTRL1_DLL_ENABLE 17 +#define BM_GPMI_CTRL1_DLL_ENABLE 0x00020000 +#define BP_GPMI_CTRL1_HALF_PERIOD 16 +#define BM_GPMI_CTRL1_HALF_PERIOD 0x00010000 +#define BP_GPMI_CTRL1_RDN_DELAY 12 +#define BM_GPMI_CTRL1_RDN_DELAY 0x0000F000 +#define BF_GPMI_CTRL1_RDN_DELAY(v) \ + (((v) << 12) & BM_GPMI_CTRL1_RDN_DELAY) +#define BM_GPMI_CTRL1_DMA2ECC_MODE 0x00000800 +#define BM_GPMI_CTRL1_DEV_IRQ 0x00000400 +#define BM_GPMI_CTRL1_TIMEOUT_IRQ 0x00000200 +#define BM_GPMI_CTRL1_BURST_EN 0x00000100 +#define BM_GPMI_CTRL1_ABORT_WAIT_REQUEST 0x00000080 +#define BP_GPMI_CTRL1_ABORT_WAIT_FOR_READY_CHANNEL 4 +#define BM_GPMI_CTRL1_ABORT_WAIT_FOR_READY_CHANNEL 0x00000070 +#define BF_GPMI_CTRL1_ABORT_WAIT_FOR_READY_CHANNEL(v) \ + (((v) << 4) & BM_GPMI_CTRL1_ABORT_WAIT_FOR_READY_CHANNEL) +#define BM_GPMI_CTRL1_DEV_RESET 0x00000008 +#define BV_GPMI_CTRL1_DEV_RESET__ENABLED 0x0 +#define BV_GPMI_CTRL1_DEV_RESET__DISABLED 0x1 +#define BM_GPMI_CTRL1_ATA_IRQRDY_POLARITY 0x00000004 +#define BV_GPMI_CTRL1_ATA_IRQRDY_POLARITY__ACTIVELOW 0x0 +#define BV_GPMI_CTRL1_ATA_IRQRDY_POLARITY__ACTIVEHIGH 0x1 +#define BM_GPMI_CTRL1_CAMERA_MODE 0x00000002 +#define BM_GPMI_CTRL1_GPMI_MODE 0x00000001 +#define BV_GPMI_CTRL1_GPMI_MODE__NAND 0x0 +#define BV_GPMI_CTRL1_GPMI_MODE__ATA 0x1 + +/*============================================================================*/ + +#define HW_GPMI_TIMING0 (0x00000070) + +#define BP_GPMI_TIMING0_RSVD1 24 +#define BM_GPMI_TIMING0_RSVD1 0xFF000000 +#define BF_GPMI_TIMING0_RSVD1(v) \ + (((v) << 24) & BM_GPMI_TIMING0_RSVD1) +#define BP_GPMI_TIMING0_ADDRESS_SETUP 16 +#define BM_GPMI_TIMING0_ADDRESS_SETUP 0x00FF0000 +#define BF_GPMI_TIMING0_ADDRESS_SETUP(v) \ + (((v) << 16) & BM_GPMI_TIMING0_ADDRESS_SETUP) +#define BP_GPMI_TIMING0_DATA_HOLD 8 +#define BM_GPMI_TIMING0_DATA_HOLD 0x0000FF00 +#define BF_GPMI_TIMING0_DATA_HOLD(v) \ + (((v) << 8) & BM_GPMI_TIMING0_DATA_HOLD) +#define BP_GPMI_TIMING0_DATA_SETUP 0 +#define BM_GPMI_TIMING0_DATA_SETUP 0x000000FF +#define BF_GPMI_TIMING0_DATA_SETUP(v) \ + (((v) << 0) & BM_GPMI_TIMING0_DATA_SETUP) + +/*============================================================================*/ + +#define HW_GPMI_TIMING1 (0x00000080) + +#define BP_GPMI_TIMING1_DEVICE_BUSY_TIMEOUT 16 +#define BM_GPMI_TIMING1_DEVICE_BUSY_TIMEOUT 0xFFFF0000 +#define BF_GPMI_TIMING1_DEVICE_BUSY_TIMEOUT(v) \ + (((v) << 16) & BM_GPMI_TIMING1_DEVICE_BUSY_TIMEOUT) +#define BP_GPMI_TIMING1_RSVD1 0 +#define BM_GPMI_TIMING1_RSVD1 0x0000FFFF +#define BF_GPMI_TIMING1_RSVD1(v) \ + (((v) << 0) & BM_GPMI_TIMING1_RSVD1) + +/*============================================================================*/ + +#define HW_GPMI_TIMING2 (0x00000090) + +#define BP_GPMI_TIMING2_UDMA_TRP 24 +#define BM_GPMI_TIMING2_UDMA_TRP 0xFF000000 +#define BF_GPMI_TIMING2_UDMA_TRP(v) \ + (((v) << 24) & BM_GPMI_TIMING2_UDMA_TRP) +#define BP_GPMI_TIMING2_UDMA_ENV 16 +#define BM_GPMI_TIMING2_UDMA_ENV 0x00FF0000 +#define BF_GPMI_TIMING2_UDMA_ENV(v) \ + (((v) << 16) & BM_GPMI_TIMING2_UDMA_ENV) +#define BP_GPMI_TIMING2_UDMA_HOLD 8 +#define BM_GPMI_TIMING2_UDMA_HOLD 0x0000FF00 +#define BF_GPMI_TIMING2_UDMA_HOLD(v) \ + (((v) << 8) & BM_GPMI_TIMING2_UDMA_HOLD) +#define BP_GPMI_TIMING2_UDMA_SETUP 0 +#define BM_GPMI_TIMING2_UDMA_SETUP 0x000000FF +#define BF_GPMI_TIMING2_UDMA_SETUP(v) \ + (((v) << 0) & BM_GPMI_TIMING2_UDMA_SETUP) + +/*============================================================================*/ + +#define HW_GPMI_DATA (0x000000a0) + +#define BP_GPMI_DATA_DATA 0 +#define BM_GPMI_DATA_DATA 0xFFFFFFFF +#define BF_GPMI_DATA_DATA(v) (v) + +#define HW_GPMI_STAT (0x000000b0) + +#define BP_GPMI_STAT_READY_BUSY 24 +#define BM_GPMI_STAT_READY_BUSY 0xFF000000 +#define BF_GPMI_STAT_READY_BUSY(v) \ + (((v) << 24) & BM_GPMI_STAT_READY_BUSY) +#define BP_GPMI_STAT_RDY_TIMEOUT 16 +#define BM_GPMI_STAT_RDY_TIMEOUT 0x00FF0000 +#define BF_GPMI_STAT_RDY_TIMEOUT(v) \ + (((v) << 16) & BM_GPMI_STAT_RDY_TIMEOUT) +#define BM_GPMI_STAT_DEV7_ERROR 0x00008000 +#define BM_GPMI_STAT_DEV6_ERROR 0x00004000 +#define BM_GPMI_STAT_DEV5_ERROR 0x00002000 +#define BM_GPMI_STAT_DEV4_ERROR 0x00001000 +#define BM_GPMI_STAT_DEV3_ERROR 0x00000800 +#define BM_GPMI_STAT_DEV2_ERROR 0x00000400 +#define BM_GPMI_STAT_DEERROR 0x00000200 +#define BM_GPMI_STAT_DEV0_ERROR 0x00000100 +#define BP_GPMI_STAT_RSVD1 5 +#define BM_GPMI_STAT_RSVD1 0x000000E0 +#define BF_GPMI_STAT_RSVD1(v) \ + (((v) << 5) & BM_GPMI_STAT_RSVD1) +#define BM_GPMI_STAT_ATA_IRQ 0x00000010 +#define BM_GPMI_STAT_INVALID_BUFFER_MASK 0x00000008 +#define BM_GPMI_STAT_FIFO_EMPTY 0x00000004 +#define BV_GPMI_STAT_FIFO_EMPTY__NOT_EMPTY 0x0 +#define BV_GPMI_STAT_FIFO_EMPTY__EMPTY 0x1 +#define BM_GPMI_STAT_FIFO_FULL 0x00000002 +#define BV_GPMI_STAT_FIFO_FULL__NOT_FULL 0x0 +#define BV_GPMI_STAT_FIFO_FULL__FULL 0x1 +#define BM_GPMI_STAT_PRESENT 0x00000001 +#define BV_GPMI_STAT_PRESENT__UNAVAILABLE 0x0 +#define BV_GPMI_STAT_PRESENT__AVAILABLE 0x1 + +/*============================================================================*/ + +#define HW_GPMI_DEBUG (0x000000c0) + +#define BP_GPMI_DEBUG_WAIT_FOR_READY_END 24 +#define BM_GPMI_DEBUG_WAIT_FOR_READY_END 0xFF000000 +#define BF_GPMI_DEBUG_WAIT_FOR_READY_END(v) \ + (((v) << 24) & BM_GPMI_DEBUG_WAIT_FOR_READY_END) +#define BP_GPMI_DEBUG_DMA_SENSE 16 +#define BM_GPMI_DEBUG_DMA_SENSE 0x00FF0000 +#define BF_GPMI_DEBUG_DMA_SENSE(v) \ + (((v) << 16) & BM_GPMI_DEBUG_DMA_SENSE) +#define BP_GPMI_DEBUG_DMAREQ 8 +#define BM_GPMI_DEBUG_DMAREQ 0x0000FF00 +#define BF_GPMI_DEBUG_DMAREQ(v) \ + (((v) << 8) & BM_GPMI_DEBUG_DMAREQ) +#define BP_GPMI_DEBUG_CMD_END 0 +#define BM_GPMI_DEBUG_CMD_END 0x000000FF +#define BF_GPMI_DEBUG_CMD_END(v) \ + (((v) << 0) & BM_GPMI_DEBUG_CMD_END) + +/*============================================================================*/ + +#define HW_GPMI_VERSION (0x000000d0) + +#define BP_GPMI_VERSION_MAJOR 24 +#define BM_GPMI_VERSION_MAJOR 0xFF000000 +#define BF_GPMI_VERSION_MAJOR(v) \ + (((v) << 24) & BM_GPMI_VERSION_MAJOR) +#define BP_GPMI_VERSION_MINOR 16 +#define BM_GPMI_VERSION_MINOR 0x00FF0000 +#define BF_GPMI_VERSION_MINOR(v) \ + (((v) << 16) & BM_GPMI_VERSION_MINOR) +#define BP_GPMI_VERSION_STEP 0 +#define BM_GPMI_VERSION_STEP 0x0000FFFF +#define BF_GPMI_VERSION_STEP(v) \ + (((v) << 0) & BM_GPMI_VERSION_STEP) + +/*============================================================================*/ + +#define HW_GPMI_DEBUG2 (0x000000e0) + +#define BP_GPMI_DEBUG2_RSVD1 28 +#define BM_GPMI_DEBUG2_RSVD1 0xF0000000 +#define BF_GPMI_DEBUG2_RSVD1(v) \ + (((v) << 28) & BM_GPMI_DEBUG2_RSVD1) +#define BP_GPMI_DEBUG2_UDMA_STATE 24 +#define BM_GPMI_DEBUG2_UDMA_STATE 0x0F000000 +#define BF_GPMI_DEBUG2_UDMA_STATE(v) \ + (((v) << 24) & BM_GPMI_DEBUG2_UDMA_STATE) +#define BM_GPMI_DEBUG2_BUSY 0x00800000 +#define BV_GPMI_DEBUG2_BUSY__DISABLED 0x0 +#define BV_GPMI_DEBUG2_BUSY__ENABLED 0x1 +#define BP_GPMI_DEBUG2_PIN_STATE 20 +#define BM_GPMI_DEBUG2_PIN_STATE 0x00700000 +#define BF_GPMI_DEBUG2_PIN_STATE(v) \ + (((v) << 20) & BM_GPMI_DEBUG2_PIN_STATE) +#define BV_GPMI_DEBUG2_PIN_STATE__PSM_IDLE 0x0 +#define BV_GPMI_DEBUG2_PIN_STATE__PSM_BYTCNT 0x1 +#define BV_GPMI_DEBUG2_PIN_STATE__PSM_ADDR 0x2 +#define BV_GPMI_DEBUG2_PIN_STATE__PSM_STALL 0x3 +#define BV_GPMI_DEBUG2_PIN_STATE__PSM_STROBE 0x4 +#define BV_GPMI_DEBUG2_PIN_STATE__PSM_ATARDY 0x5 +#define BV_GPMI_DEBUG2_PIN_STATE__PSM_DHOLD 0x6 +#define BV_GPMI_DEBUG2_PIN_STATE__PSM_DONE 0x7 +#define BP_GPMI_DEBUG2_MAIN_STATE 16 +#define BM_GPMI_DEBUG2_MAIN_STATE 0x000F0000 +#define BF_GPMI_DEBUG2_MAIN_STATE(v) \ + (((v) << 16) & BM_GPMI_DEBUG2_MAIN_STATE) +#define BV_GPMI_DEBUG2_MAIN_STATE__MSM_IDLE 0x0 +#define BV_GPMI_DEBUG2_MAIN_STATE__MSM_BYTCNT 0x1 +#define BV_GPMI_DEBUG2_MAIN_STATE__MSM_WAITFE 0x2 +#define BV_GPMI_DEBUG2_MAIN_STATE__MSM_WAITFR 0x3 +#define BV_GPMI_DEBUG2_MAIN_STATE__MSM_DMAREQ 0x4 +#define BV_GPMI_DEBUG2_MAIN_STATE__MSM_DMAACK 0x5 +#define BV_GPMI_DEBUG2_MAIN_STATE__MSM_WAITFF 0x6 +#define BV_GPMI_DEBUG2_MAIN_STATE__MSM_LDFIFO 0x7 +#define BV_GPMI_DEBUG2_MAIN_STATE__MSM_LDDMAR 0x8 +#define BV_GPMI_DEBUG2_MAIN_STATE__MSM_RDCMP 0x9 +#define BV_GPMI_DEBUG2_MAIN_STATE__MSM_DONE 0xA +#define BP_GPMI_DEBUG2_SYND2GPMI_BE 12 +#define BM_GPMI_DEBUG2_SYND2GPMI_BE 0x0000F000 +#define BF_GPMI_DEBUG2_SYND2GPMI_BE(v) \ + (((v) << 12) & BM_GPMI_DEBUG2_SYND2GPMI_BE) +#define BM_GPMI_DEBUG2_GPMI2SYND_VALID 0x00000800 +#define BM_GPMI_DEBUG2_GPMI2SYND_READY 0x00000400 +#define BM_GPMI_DEBUG2_SYND2GPMI_VALID 0x00000200 +#define BM_GPMI_DEBUG2_SYND2GPMI_READY 0x00000100 +#define BM_GPMI_DEBUG2_VIEW_DELAYED_RDN 0x00000080 +#define BM_GPMI_DEBUG2_UPDATE_WINDOW 0x00000040 +#define BP_GPMI_DEBUG2_RDN_TAP 0 +#define BM_GPMI_DEBUG2_RDN_TAP 0x0000003F +#define BF_GPMI_DEBUG2_RDN_TAP(v) \ + (((v) << 0) & BM_GPMI_DEBUG2_RDN_TAP) + +/*============================================================================*/ + +#define HW_GPMI_DEBUG3 (0x000000f0) + +#define BP_GPMI_DEBUG3_APB_WORD_CNTR 16 +#define BM_GPMI_DEBUG3_APB_WORD_CNTR 0xFFFF0000 +#define BF_GPMI_DEBUG3_APB_WORD_CNTR(v) \ + (((v) << 16) & BM_GPMI_DEBUG3_APB_WORD_CNTR) +#define BP_GPMI_DEBUG3_DEV_WORD_CNTR 0 +#define BM_GPMI_DEBUG3_DEV_WORD_CNTR 0x0000FFFF +#define BF_GPMI_DEBUG3_DEV_WORD_CNTR(v) \ + (((v) << 0) & BM_GPMI_DEBUG3_DEV_WORD_CNTR) + +/*============================================================================*/ + +#endif diff --git a/drivers/mtd/nand/gpmi-nfc/hal-imx28.c b/drivers/mtd/nand/gpmi-nfc/hal-imx28.c new file mode 100644 index 0000000..ff87d7f --- /dev/null +++ b/drivers/mtd/nand/gpmi-nfc/hal-imx28.c @@ -0,0 +1,503 @@ +/* + * Freescale GPMI NFC NAND Flash Driver + * + * Copyright (C) 2011 Freescale Semiconductor, Inc. + * Copyright (C) 2008 Embedded Alley Solutions, Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + */ +#include "gpmi-nfc.h" +#include "gpmi-regs-imx28.h" +#include "bch-regs-imx28.h" + +static int init_hal_imx28(struct gpmi_nfc_data *this) +{ + struct resources *resources = &this->resources; + + /* Enable the clock. */ + clk_enable(resources->clock); + + /* Reset the GPMI block. */ + mxs_reset_block(resources->gpmi_regs); + + /* Choose NAND mode. */ + __raw_writel(BM_GPMI_CTRL1_GPMI_MODE, + resources->gpmi_regs + HW_GPMI_CTRL1_CLR); + + /* Set the IRQ polarity. */ + __raw_writel(BM_GPMI_CTRL1_ATA_IRQRDY_POLARITY, + resources->gpmi_regs + HW_GPMI_CTRL1_SET); + + /* Disable write protection. */ + __raw_writel(BM_GPMI_CTRL1_DEV_RESET, + resources->gpmi_regs + HW_GPMI_CTRL1_SET); + + /* Select BCH ECC. */ + __raw_writel(BM_GPMI_CTRL1_BCH_MODE, + resources->gpmi_regs + HW_GPMI_CTRL1_SET); + + /* Disable the clock. */ + clk_disable(resources->clock); + + return 0; +} + +/* Configures the NFC geometry for BCH. */ +static int set_geometry(struct gpmi_nfc_data *this) +{ + struct resources *resources = &this->resources; + struct nfc_geometry *nfc = &this->nfc_geometry; + unsigned int block_count; + unsigned int block_size; + unsigned int metadata_size; + unsigned int ecc_strength; + unsigned int page_size; + + /* We make the abstract choices in a common function. */ + if (common_nfc_set_geometry(this)) + return !0; + + /* Translate the abstract choices into register fields. */ + block_count = nfc->ecc_chunk_count - 1; + block_size = nfc->ecc_chunk_size_in_bytes; + metadata_size = nfc->metadata_size_in_bytes; + ecc_strength = nfc->ecc_strength >> 1; + page_size = nfc->page_size_in_bytes; + + /* Enable the clock. */ + clk_enable(resources->clock); + + /* + * Reset the BCH block. Notice that we pass in true for the just_enable + * flag. This is because the soft reset for the version 0 BCH block + * doesn't work and the version 1 BCH block is similar enough that we + * suspect the same (though this has not been officially tested). If you + * try to soft reset a version 0 BCH block, it becomes unusable until + * the next hard reset. + */ + mxs_reset_block(resources->bch_regs); + + /* Configure layout 0. */ + __raw_writel( + BF_BCH_FLASH0LAYOUT0_NBLOCKS(block_count) | + BF_BCH_FLASH0LAYOUT0_META_SIZE(metadata_size) | + BF_BCH_FLASH0LAYOUT0_ECC0(ecc_strength) | + BF_BCH_FLASH0LAYOUT0_DATA0_SIZE(block_size) , + resources->bch_regs + HW_BCH_FLASH0LAYOUT0); + + __raw_writel( + BF_BCH_FLASH0LAYOUT1_PAGE_SIZE(page_size) | + BF_BCH_FLASH0LAYOUT1_ECCN(ecc_strength) | + BF_BCH_FLASH0LAYOUT1_DATAN_SIZE(block_size) , + resources->bch_regs + HW_BCH_FLASH0LAYOUT1); + + /* Set *all* chip selects to use layout 0. */ + __raw_writel(0, resources->bch_regs + HW_BCH_LAYOUTSELECT); + + /* Enable interrupts. */ + __raw_writel(BM_BCH_CTRL_COMPLETE_IRQ_EN, + resources->bch_regs + HW_BCH_CTRL_SET); + + /* Disable the clock. */ + clk_disable(resources->clock); + + return 0; +} + +static int set_timing(struct gpmi_nfc_data *this, + const struct nand_timing *timing) +{ + struct nfc_hal *nfc = this->nfc; + + nfc->timing = *timing; + return 0; +} + +/** + * get_timing() - Retrieves the NFC hardware timing. + * + * @this: Per-device data. + * @clock_frequency_in_hz: The clock frequency, in Hz, during the current + * I/O transaction. If no I/O transaction is in + * progress, this is the clock frequency during the + * most recent I/O transaction. + * @hardware_timing: The hardware timing configuration in effect during + * the current I/O transaction. If no I/O transaction + * is in progress, this is the hardware timing + * configuration during the most recent I/O + * transaction. + */ +static void get_timing(struct gpmi_nfc_data *this, + unsigned long *clock_frequency_in_hz, + struct gpmi_nfc_hardware_timing *hardware_timing) +{ + struct resources *resources = &this->resources; + struct nfc_hal *nfc = this->nfc; + unsigned char *gpmi_regs = resources->gpmi_regs; + uint32_t register_image; + + /* Return the clock frequency. */ + *clock_frequency_in_hz = nfc->clock_frequency_in_hz; + + /* We'll be reading the hardware, so let's enable the clock. */ + clk_enable(resources->clock); + + /* Retrieve the hardware timing. */ + register_image = __raw_readl(gpmi_regs + HW_GPMI_TIMING0); + + hardware_timing->data_setup_in_cycles = + (register_image & BM_GPMI_TIMING0_DATA_SETUP) >> + BP_GPMI_TIMING0_DATA_SETUP; + + hardware_timing->data_hold_in_cycles = + (register_image & BM_GPMI_TIMING0_DATA_HOLD) >> + BP_GPMI_TIMING0_DATA_HOLD; + + hardware_timing->address_setup_in_cycles = + (register_image & BM_GPMI_TIMING0_ADDRESS_SETUP) >> + BP_GPMI_TIMING0_ADDRESS_SETUP; + + register_image = __raw_readl(gpmi_regs + HW_GPMI_CTRL1); + + hardware_timing->use_half_periods = + (register_image & BM_GPMI_CTRL1_HALF_PERIOD) >> + BP_GPMI_CTRL1_HALF_PERIOD; + + hardware_timing->sample_delay_factor = + (register_image & BM_GPMI_CTRL1_RDN_DELAY) >> + BP_GPMI_CTRL1_RDN_DELAY; + + /* We're done reading the hardware, so disable the clock. */ + clk_disable(resources->clock); +} + +static void exit(struct gpmi_nfc_data *this) +{ +} + +static void begin(struct gpmi_nfc_data *this) +{ + struct resources *resources = &this->resources; + + /* Enable the clock. */ + clk_enable(resources->clock); +} + +static void end(struct gpmi_nfc_data *this) +{ + struct resources *resources = &this->resources; + clk_disable(resources->clock); +} + +/* Clears a BCH interrupt. */ +static void clear_bch(struct gpmi_nfc_data *this) +{ + struct resources *r = &this->resources; + + __raw_writel(BM_BCH_CTRL_COMPLETE_IRQ, r->bch_regs + HW_BCH_CTRL_CLR); +} + +/* Returns the Ready/Busy status of the given chip. */ +static int is_ready(struct gpmi_nfc_data *this, unsigned chip) +{ + struct resources *resources = &this->resources; + uint32_t mask; + uint32_t register_image; + + /* Extract and return the status. */ + mask = BF_GPMI_STAT_READY_BUSY(1 << chip); + register_image = __raw_readl(resources->gpmi_regs + HW_GPMI_STAT); + return !!(register_image & mask); +} + +/* Sends a command and associated addresses. */ +static int send_command(struct gpmi_nfc_data *this) +{ + struct dma_chan *channel = get_dma_chan(this); + struct mil *mil = &this->mil; + struct dma_async_tx_descriptor *desc; + struct scatterlist *sgl; + u32 pio[3]; + + /* [1] send out the PIO words */ + pio[0] = BF_GPMI_CTRL0_COMMAND_MODE(BV_GPMI_CTRL0_COMMAND_MODE__WRITE) + | BM_GPMI_CTRL0_WORD_LENGTH + | BF_GPMI_CTRL0_CS(mil->current_chip) + | BF_GPMI_CTRL0_ADDRESS(BV_GPMI_CTRL0_ADDRESS__NAND_CLE) + | BM_GPMI_CTRL0_ADDRESS_INCREMENT + | BF_GPMI_CTRL0_XFER_COUNT(mil->command_length); + pio[1] = pio[2] = 0; + desc = channel->device->device_prep_slave_sg(channel, + (struct scatterlist *)pio, + ARRAY_SIZE(pio), DMA_NONE, 0); + if (!desc) { + log("step 1 error"); + return -1; + } + + /* [2] send out the COMMAND + ADDRESS string stored in @buffer */ + sgl = &mil->cmd_sgl; + + dma_map_sg(this->dev, sgl, 1, DMA_TO_DEVICE); + sgl->length = mil->command_length; + desc = channel->device->device_prep_slave_sg(channel, + sgl, 1, DMA_TO_DEVICE, 1); + if (!desc) { + log("error"); + return -1; + } + + /* [3] submit the DMA */ + this->dma_type = DMA_FOR_COMMAND; + start_dma_without_bch_irq(this, desc); + return 0; +} + +static int send_data(struct gpmi_nfc_data *this) +{ + struct dma_async_tx_descriptor *desc; + struct dma_chan *channel = get_dma_chan(this); + struct mil *mil = &this->mil; + uint32_t command_mode; + uint32_t address; + u32 pio[2]; + + /* [1] PIO */ + command_mode = BV_GPMI_CTRL0_COMMAND_MODE__WRITE; + address = BV_GPMI_CTRL0_ADDRESS__NAND_DATA; + + pio[0] = + BF_GPMI_CTRL0_COMMAND_MODE(command_mode) | + BM_GPMI_CTRL0_WORD_LENGTH | + BF_GPMI_CTRL0_CS(mil->current_chip) | + BF_GPMI_CTRL0_ADDRESS(address) | + BF_GPMI_CTRL0_XFER_COUNT(mil->upper_len); + pio[1] = 0; + desc = channel->device->device_prep_slave_sg(channel, + (struct scatterlist *)pio, + ARRAY_SIZE(pio), DMA_NONE, 0); + if (!desc) { + log("step 1 error"); + return -1; + } + + /* [2] : send DMA request */ + prepare_data_dma(this, DMA_TO_DEVICE); + desc = channel->device->device_prep_slave_sg(channel, &mil->data_sgl, + 1, DMA_TO_DEVICE, 1); + if (!desc) { + log("step 2 error"); + return -1; + } + /* [3] submit the DMA */ + this->dma_type = DMA_FOR_WRITE_DATA; + start_dma_without_bch_irq(this, desc); + return 0; +} + +static int read_data(struct gpmi_nfc_data *this) +{ + struct dma_async_tx_descriptor *desc; + struct dma_chan *channel = get_dma_chan(this); + struct mil *mil = &this->mil; + u32 pio[2]; + + /* [1] : send PIO */ + pio[0] = BF_GPMI_CTRL0_COMMAND_MODE(BV_GPMI_CTRL0_COMMAND_MODE__READ) + | BM_GPMI_CTRL0_WORD_LENGTH + | BF_GPMI_CTRL0_CS(mil->current_chip) + | BF_GPMI_CTRL0_ADDRESS(BV_GPMI_CTRL0_ADDRESS__NAND_DATA) + | BF_GPMI_CTRL0_XFER_COUNT(mil->upper_len); + pio[1] = 0; + desc = channel->device->device_prep_slave_sg(channel, + (struct scatterlist *)pio, + ARRAY_SIZE(pio), DMA_NONE, 0); + if (!desc) { + log("step 1 error"); + return -1; + } + + /* [2] : setup DMA buffer */ + prepare_data_dma(this, DMA_FROM_DEVICE); + desc = channel->device->device_prep_slave_sg(channel, &mil->data_sgl, + 1, DMA_FROM_DEVICE, 1); + if (!desc) { + log("step 2 error"); + return -1; + } + + /* [3] : submit the DMA */ + this->dma_type = DMA_FOR_READ_DATA; + start_dma_without_bch_irq(this, desc); + return 0; +} + +static int send_page(struct gpmi_nfc_data *this, + dma_addr_t payload, dma_addr_t auxiliary) +{ + struct nfc_geometry *geo = &this->nfc_geometry; + uint32_t command_mode; + uint32_t address; + uint32_t ecc_command; + uint32_t buffer_mask; + struct dma_async_tx_descriptor *desc; + struct dma_chan *channel = get_dma_chan(this); + struct mil *mil = &this->mil; + int chip = mil->current_chip; + u32 pio[6]; + + /* A DMA descriptor that does an ECC page read. */ + command_mode = BV_GPMI_CTRL0_COMMAND_MODE__WRITE; + address = BV_GPMI_CTRL0_ADDRESS__NAND_DATA; + ecc_command = BV_GPMI_ECCCTRL_ECC_CMD__ENCODE; + buffer_mask = BV_GPMI_ECCCTRL_BUFFER_MASK__BCH_PAGE | + BV_GPMI_ECCCTRL_BUFFER_MASK__BCH_AUXONLY; + + pio[0] = + BF_GPMI_CTRL0_COMMAND_MODE(command_mode) | + BM_GPMI_CTRL0_WORD_LENGTH | + BF_GPMI_CTRL0_CS(chip) | + BF_GPMI_CTRL0_ADDRESS(address) | + BF_GPMI_CTRL0_XFER_COUNT(0) ; + pio[1] = 0; + pio[2] = + BM_GPMI_ECCCTRL_ENABLE_ECC | + BF_GPMI_ECCCTRL_ECC_CMD(ecc_command) | + BF_GPMI_ECCCTRL_BUFFER_MASK(buffer_mask) ; + pio[3] = geo->page_size_in_bytes; + pio[4] = payload; + pio[5] = auxiliary; + + desc = channel->device->device_prep_slave_sg(channel, + (struct scatterlist *)pio, + ARRAY_SIZE(pio), DMA_NONE, 0); + if (!desc) { + log("step 2 error"); + return -1; + } + this->dma_type = DMA_FOR_WRITE_ECC_PAGE; + return start_dma_with_bch_irq(this, desc); +} + +static int read_page(struct gpmi_nfc_data *this, + dma_addr_t payload, dma_addr_t auxiliary) +{ + struct nfc_geometry *geo = &this->nfc_geometry; + uint32_t command_mode; + uint32_t address; + uint32_t ecc_command; + uint32_t buffer_mask; + + struct dma_async_tx_descriptor *desc; + struct dma_chan *channel = get_dma_chan(this); + struct mil *mil = &this->mil; + int chip = mil->current_chip; + u32 pio[6]; + + /* [1] Wait for the chip to report ready. */ + command_mode = BV_GPMI_CTRL0_COMMAND_MODE__WAIT_FOR_READY; + address = BV_GPMI_CTRL0_ADDRESS__NAND_DATA; + + pio[0] = BF_GPMI_CTRL0_COMMAND_MODE(command_mode) + | BM_GPMI_CTRL0_WORD_LENGTH + | BF_GPMI_CTRL0_CS(chip) + | BF_GPMI_CTRL0_ADDRESS(address) + | BF_GPMI_CTRL0_XFER_COUNT(0); + pio[1] = 0; + desc = channel->device->device_prep_slave_sg(channel, + (struct scatterlist *)pio, 2, DMA_NONE, 0); + if (!desc) { + log("step 1 error"); + return -1; + } + + /* [2] Enable the BCH block and read. */ + command_mode = BV_GPMI_CTRL0_COMMAND_MODE__READ; + address = BV_GPMI_CTRL0_ADDRESS__NAND_DATA; + ecc_command = BV_GPMI_ECCCTRL_ECC_CMD__DECODE; + buffer_mask = BV_GPMI_ECCCTRL_BUFFER_MASK__BCH_PAGE + | BV_GPMI_ECCCTRL_BUFFER_MASK__BCH_AUXONLY; + + pio[0] = BF_GPMI_CTRL0_COMMAND_MODE(command_mode) + | BM_GPMI_CTRL0_WORD_LENGTH + | BF_GPMI_CTRL0_CS(chip) + | BF_GPMI_CTRL0_ADDRESS(address) + | BF_GPMI_CTRL0_XFER_COUNT(geo->page_size_in_bytes); + + pio[1] = 0; + pio[2] = BM_GPMI_ECCCTRL_ENABLE_ECC + | BF_GPMI_ECCCTRL_ECC_CMD(ecc_command) + | BF_GPMI_ECCCTRL_BUFFER_MASK(buffer_mask); + pio[3] = geo->page_size_in_bytes; + pio[4] = payload; + pio[5] = auxiliary; + desc = channel->device->device_prep_slave_sg(channel, + (struct scatterlist *)pio, + ARRAY_SIZE(pio), DMA_NONE, 1); + if (!desc) { + log("step 2 error"); + return -1; + } + + /* [3] Disable the BCH block */ + command_mode = BV_GPMI_CTRL0_COMMAND_MODE__WAIT_FOR_READY; + address = BV_GPMI_CTRL0_ADDRESS__NAND_DATA; + + pio[0] = + BF_GPMI_CTRL0_COMMAND_MODE(command_mode) | + BM_GPMI_CTRL0_WORD_LENGTH | + BF_GPMI_CTRL0_CS(chip) | + BF_GPMI_CTRL0_ADDRESS(address) | + BF_GPMI_CTRL0_XFER_COUNT(geo->page_size_in_bytes) ; + pio[1] = 0; + desc = channel->device->device_prep_slave_sg(channel, + (struct scatterlist *)pio, 2, DMA_NONE, 1); + if (!desc) { + log("step 3 error"); + return -1; + } + + /* [4] submit the DMA */ + this->dma_type = DMA_FOR_READ_ECC_PAGE; + return start_dma_with_bch_irq(this, desc); +} + +/* This structure represents the NFC HAL for this version of the hardware. */ +struct nfc_hal gpmi_nfc_hal_imx28 = { + .version = 1, + .description = "8-chip GPMI and BCH", + .max_chip_count = 8, + .max_data_setup_cycles = (BM_GPMI_TIMING0_DATA_SETUP >> + BP_GPMI_TIMING0_DATA_SETUP), + .internal_data_setup_in_ns = 0, + .max_sample_delay_factor = (BM_GPMI_CTRL1_RDN_DELAY >> + BP_GPMI_CTRL1_RDN_DELAY), + .max_dll_clock_period_in_ns = 32, + .max_dll_delay_in_ns = 16, + .init = init_hal_imx28, + .set_geometry = set_geometry, + .set_timing = set_timing, + .get_timing = get_timing, + .exit = exit, + .begin = begin, + .end = end, + .clear_bch = clear_bch, + .is_ready = is_ready, + .send_command = send_command, + .read_data = read_data, + .send_data = send_data, + .send_page = send_page, + .read_page = read_page, +}; diff --git a/drivers/mtd/nand/gpmi-nfc/rom-imx28.c b/drivers/mtd/nand/gpmi-nfc/rom-imx28.c new file mode 100644 index 0000000..03be07f --- /dev/null +++ b/drivers/mtd/nand/gpmi-nfc/rom-imx28.c @@ -0,0 +1,66 @@ +/* + * Freescale GPMI NFC NAND Flash Driver + * + * Copyright (C) 2010-2011 Freescale Semiconductor, Inc. + * Copyright (C) 2008 Embedded Alley Solutions, Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + */ +#include "gpmi-nfc.h" + +/* Sets geometry for the Boot ROM Helper. */ +static int set_geometry(struct gpmi_nfc_data *this) +{ + struct gpmi_nfc_platform_data *pdata = this->pdata; + struct boot_rom_geometry *geometry = &this->rom_geometry; + int error; + + /* Version-independent geometry. */ + error = gpmi_nfc_rom_helper_set_geometry(this); + if (error) + return error; + + /* + * Check if the platform data indicates we are to protect the boot area. + */ + if (!pdata->boot_area_size_in_bytes) { + geometry->boot_area_count = 0; + geometry->boot_area_size_in_bytes = 0; + return 0; + } + + /* + * If control arrives here, we are supposed to set up partitions to + * protect the boot areas. In this version of the ROM, we support only + * one boot area. + */ + geometry->boot_area_count = 1; + + /* + * Use the platform's boot area size. + */ + geometry->boot_area_size_in_bytes = pdata->boot_area_size_in_bytes; + + return 0; +} + +/* This structure represents the Boot ROM Helper for this version. */ +struct boot_rom_helper gpmi_nfc_boot_rom_imx28 = { + .version = 1, + .description = "Single-chip boot area, " + "block mark swapping supported", + .swap_block_mark = true, + .set_geometry = set_geometry, +}; -- 1.7.0.4 ^ permalink raw reply related [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-16 1:55 ` [PATCH 5/7] add GPMI support for imx28 Huang Shijie @ 2011-03-31 9:47 ` Artem Bityutskiy 2011-03-31 10:09 ` Huang Shijie 2011-03-31 9:49 ` Artem Bityutskiy 2011-03-31 9:49 ` Artem Bityutskiy 2 siblings, 1 reply; 21+ messages in thread From: Artem Bityutskiy @ 2011-03-31 9:47 UTC (permalink / raw) To: Huang Shijie; +Cc: linux-mtd, linux, dwmw2, linux-arm-kernel, David.Woodhouse On Wed, 2011-03-16 at 09:55 +0800, Huang Shijie wrote: > +#define BF_BCH_CTRL_RSVD5(v) \ > + (((v) << 23) & BM_BCH_CTRL_RSVD5) Many lines where you used "\" fit 80 lines. Just do not use tabs between name and value (which is ugly, imo), use a space and then it fits 80 lines. This neater than splitting things on 2 lines. -- Best Regards, Artem Bityutskiy (Артём Битюцкий) ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-31 9:47 ` Artem Bityutskiy @ 2011-03-31 10:09 ` Huang Shijie 0 siblings, 0 replies; 21+ messages in thread From: Huang Shijie @ 2011-03-31 10:09 UTC (permalink / raw) To: dedekind1; +Cc: linux-mtd, linux, dwmw2, linux-arm-kernel, David.Woodhouse Hi: > On Wed, 2011-03-16 at 09:55 +0800, Huang Shijie wrote: >> +#define BF_BCH_CTRL_RSVD5(v) \ >> + (((v)<< 23)& BM_BCH_CTRL_RSVD5) > Many lines where you used "\" fit 80 lines. Just do not use tabs between > name and value (which is ugly, imo), use a space and then it fits 80 > lines. This neater than splitting things on 2 lines. > Please check the latest version of the driver. The code has been remove in new version. Best Regards Huang Shijie ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-16 1:55 ` [PATCH 5/7] add GPMI support for imx28 Huang Shijie 2011-03-31 9:47 ` Artem Bityutskiy @ 2011-03-31 9:49 ` Artem Bityutskiy 2011-03-31 9:49 ` Artem Bityutskiy 2 siblings, 0 replies; 21+ messages in thread From: Artem Bityutskiy @ 2011-03-31 9:49 UTC (permalink / raw) To: Huang Shijie; +Cc: linux-mtd, linux, dwmw2, linux-arm-kernel, David.Woodhouse On Wed, 2011-03-16 at 09:55 +0800, Huang Shijie wrote: > + pio[0] = > + BF_GPMI_CTRL0_COMMAND_MODE(command_mode) | > + BM_GPMI_CTRL0_WORD_LENGTH | > + BF_GPMI_CTRL0_CS(mil->current_chip) | > + BF_GPMI_CTRL0_ADDRESS(address) | > + BF_GPMI_CTRL0_XFER_COUNT(mil->upper_len); Strange way to format the code ... -- Best Regards, Artem Bityutskiy (Артём Битюцкий) ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCH 5/7] add GPMI support for imx28 2011-03-16 1:55 ` [PATCH 5/7] add GPMI support for imx28 Huang Shijie 2011-03-31 9:47 ` Artem Bityutskiy 2011-03-31 9:49 ` Artem Bityutskiy @ 2011-03-31 9:49 ` Artem Bityutskiy 2 siblings, 0 replies; 21+ messages in thread From: Artem Bityutskiy @ 2011-03-31 9:49 UTC (permalink / raw) To: Huang Shijie; +Cc: linux-mtd, linux, dwmw2, linux-arm-kernel, David.Woodhouse On Wed, 2011-03-16 at 09:55 +0800, Huang Shijie wrote: > + .description = "Single-chip boot area, " > + "block mark swapping > supported", Because you love tabs so much, you have to do this strange splitting of strings. Just use normal spaces and you'll fit many things in one line. -- Best Regards, Artem Bityutskiy (Артём Битюцкий) ^ permalink raw reply [flat|nested] 21+ messages in thread
end of thread, other threads:[~2011-03-31 10:09 UTC | newest]
Thread overview: 21+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <1300239773-4222-1-git-send-email-b32955@freescale.com>
[not found] ` <1300239773-4222-2-git-send-email-b32955@freescale.com>
[not found] ` <19840.36163.967469.53941@ipc1.ka-ro>
[not found] ` <4D809453.4090603@freescale.com>
[not found] ` <19840.43943.592336.854865@ipc1.ka-ro>
2011-03-17 2:19 ` [PATCH 1/7] ARM: add GPMI support for imx23/imx28 Huang Shijie
2011-03-17 10:36 ` Lothar Waßmann
2011-03-18 2:06 ` Huang Shijie
[not found] ` <1300239773-4222-6-git-send-email-b32955@freescale.com>
[not found] ` <19848.39439.380861.195051@ipc1.ka-ro>
2011-03-23 3:11 ` [PATCH 5/7] add GPMI support for imx28 Huang Shijie
2011-03-23 14:56 ` Lothar Waßmann
2011-03-23 15:16 ` Florian Fainelli
2011-03-24 3:07 ` Huang Shijie
2011-03-24 3:03 ` Huang Shijie
2011-03-24 7:34 ` Lothar Waßmann
2011-03-24 8:26 ` Jason Liu
2011-03-24 8:32 ` Wolfram Sang
2011-03-24 8:33 ` Lothar Waßmann
2011-03-24 8:51 ` Huang Shijie
2011-03-24 13:54 ` Florian Fainelli
2011-03-25 2:50 ` Huang Shijie
[not found] ` <1300239773-4222-3-git-send-email-b32955@freescale.com>
[not found] ` <19848.38860.624803.339369@ipc1.ka-ro>
2011-03-23 3:41 ` [PATCH 2/7] add the common code for GPMI driver Huang Shijie
2011-03-16 1:55 [PATCH 0/7] add the GPMI controller driver for IMX23/IMX28 Huang Shijie
2011-03-16 1:55 ` [PATCH 5/7] add GPMI support for imx28 Huang Shijie
2011-03-31 9:47 ` Artem Bityutskiy
2011-03-31 10:09 ` Huang Shijie
2011-03-31 9:49 ` Artem Bityutskiy
2011-03-31 9:49 ` Artem Bityutskiy
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).