[PATCH] omap3: nand: bch ecc support added

Vimal Singh vimal.newwork at gmail.com
Thu Jan 20 10:17:25 EST 2011


Hi Ghorai,

I am the initial author of this patch. And I guess this patch still
uses most of my work.


Thanks and Regards,
Vimal

On Thu, Jan 20, 2011 at 3:48 PM, Sukumar Ghorai <s-ghorai at ti.com> wrote:
> bch error correction (t=4 and t=8) for 512 bytes support added.
> Tested in omap-3630 es-1.1 silicon.
>
> Need to select the bch-ecc from board file. E.g.
> arch/arm/mach-omap2/board-flash.c: board_nand_init()
> board_nand_data.ecc_opt = OMAP_ECC_BCH4_CODE_HW
>
> This patch has dependency on -
> http://www.mail-archive.com/linux-omap@vger.kernel.org/msg42658.html
>
> Signed-off-by: Sukumar Ghorai <s-ghorai at ti.com>
> ---
>  arch/arm/mach-omap2/gpmc.c             |  126 ++++++++---
>  arch/arm/plat-omap/include/plat/gpmc.h |    6 +-
>  drivers/mtd/nand/Makefile              |    1 +
>  drivers/mtd/nand/omap2.c               |  119 ++++++++--
>  drivers/mtd/nand/omap_bch_decoder.c    |  393 ++++++++++++++++++++++++++++++++
>  5 files changed, 583 insertions(+), 62 deletions(-)
>  create mode 100644 drivers/mtd/nand/omap_bch_decoder.c
>
> diff --git a/arch/arm/mach-omap2/gpmc.c b/arch/arm/mach-omap2/gpmc.c
> index 29c9732..91cfdca 100644
> --- a/arch/arm/mach-omap2/gpmc.c
> +++ b/arch/arm/mach-omap2/gpmc.c
> @@ -48,6 +48,7 @@
>  #define GPMC_ECC_CONTROL       0x1f8
>  #define GPMC_ECC_SIZE_CONFIG   0x1fc
>  #define GPMC_ECC1_RESULT        0x200
> +#define GPMC_ECC_BCH_RESULT_0  0x240
>
>  #define GPMC_CS0_OFFSET                0x60
>  #define GPMC_CS_SIZE           0x30
> @@ -94,7 +95,6 @@ static struct resource        gpmc_mem_root;
>  static struct resource gpmc_cs_mem[GPMC_CS_NUM];
>  static DEFINE_SPINLOCK(gpmc_mem_lock);
>  static unsigned int gpmc_cs_map;       /* flag for cs which are initialized */
> -static int gpmc_ecc_used = -EINVAL;    /* cs using ecc engine */
>
>  static void __iomem *gpmc_base;
>
> @@ -832,52 +832,77 @@ void omap3_gpmc_restore_context(void)
>
>  /**
>  * gpmc_enable_hwecc - enable hardware ecc functionality
> + * @ecc_type: ecc type e.g. Hamming, BCH
>  * @cs: chip select number
>  * @mode: read/write mode
>  * @dev_width: device bus width(1 for x16, 0 for x8)
>  * @ecc_size: bytes for which ECC will be generated
>  */
> -int gpmc_enable_hwecc(int cs, int mode, int dev_width, int ecc_size)
> +int gpmc_enable_hwecc(int ecc_type, int cs, int mode,
> +                       int dev_width, int ecc_size)
>  {
> -       unsigned int val;
> -
> -       /* check if ecc module is in used */
> -       if (gpmc_ecc_used != -EINVAL)
> -               return -EINVAL;
> -
> -       gpmc_ecc_used = cs;
> -
> -       /* clear ecc and enable bits */
> -       val = ((0x00000001<<8) | 0x00000001);
> -       gpmc_write_reg(GPMC_ECC_CONTROL, val);
> -
> -       /* program ecc and result sizes */
> -       val = ((((ecc_size >> 1) - 1) << 22) | (0x0000000F));
> -       gpmc_write_reg(GPMC_ECC_SIZE_CONFIG, val);
> +       unsigned int bch_mod = 0, bch_wrapmode = 0, eccsize1 = 0, eccsize0 = 0;
> +       unsigned int ecc_conf_val = 0, ecc_size_conf_val = 0;
>
>        switch (mode) {
>        case GPMC_ECC_READ:
> -               gpmc_write_reg(GPMC_ECC_CONTROL, 0x101);
> +               if (ecc_type == OMAP_ECC_BCH4_CODE_HW) {
> +                       eccsize1 = 0xD; eccsize0 = 0x48;
> +                       bch_mod = 0;
> +                       bch_wrapmode = 0x09;
> +               } else if (ecc_type == OMAP_ECC_BCH8_CODE_HW) {
> +                       eccsize1 = 0x1A; eccsize0 = 0x18;
> +                       bch_mod = 1;
> +                       bch_wrapmode = 0x04;
> +               } else
> +                       eccsize1 = ((ecc_size >> 1) - 1) << 22;
>                break;
> +
>        case GPMC_ECC_READSYN:
> -                gpmc_write_reg(GPMC_ECC_CONTROL, 0x100);
>                break;
> +
>        case GPMC_ECC_WRITE:
> -               gpmc_write_reg(GPMC_ECC_CONTROL, 0x101);
> +               if (ecc_type == OMAP_ECC_BCH4_CODE_HW) {
> +                       eccsize1 = 0x20; eccsize0 = 0x00;
> +                       bch_mod = 0;
> +                       bch_wrapmode = 0x06;
> +               } else if (ecc_type == OMAP_ECC_BCH8_CODE_HW) {
> +                       eccsize1 = 0x20; eccsize0 = 0x00;
> +                       bch_mod = 1;
> +                       bch_wrapmode = 0x06;
> +               } else
> +                       eccsize1 = ((ecc_size >> 1) - 1) << 22;
>                break;
> +
>        default:
>                printk(KERN_INFO "Error: Unrecognized Mode[%d]!\n", mode);
>                break;
>        }
>
> -       /* (ECC 16 or 8 bit col) | ( CS  )  | ECC Enable */
> -       val = (dev_width << 7) | (cs << 1) | (0x1);
> -       gpmc_write_reg(GPMC_ECC_CONFIG, val);
> +       /* clear ecc and enable bits */
> +       if ((ecc_type == OMAP_ECC_BCH4_CODE_HW) ||
> +               (ecc_type == OMAP_ECC_BCH8_CODE_HW)) {
> +               gpmc_write_reg(GPMC_ECC_CONTROL, 0x00000001);
> +               ecc_size_conf_val = (eccsize1 << 22) | (eccsize0 << 12);
> +               ecc_conf_val = ((0x01 << 16) | (bch_mod << 12)
> +                       | (bch_wrapmode << 8) | (dev_width << 7)
> +                       | (0x03 << 4) | (cs << 1) | (0x1));
> +       } else {
> +               gpmc_write_reg(GPMC_ECC_CONTROL, 0x00000101);
> +               ecc_size_conf_val = (eccsize1 << 22) | 0x0000000F;
> +               ecc_conf_val = (dev_width << 7) | (cs << 1) | (0x1);
> +       }
> +
> +       gpmc_write_reg(GPMC_ECC_SIZE_CONFIG, ecc_size_conf_val);
> +       gpmc_write_reg(GPMC_ECC_CONFIG, ecc_conf_val);
> +       gpmc_write_reg(GPMC_ECC_CONTROL, 0x00000101);
> +
>        return 0;
>  }
>
>  /**
>  * gpmc_calculate_ecc - generate non-inverted ecc bytes
> + * @ecc_type: ecc type e.g. Hamming, BCH
>  * @cs: chip select number
>  * @dat: data pointer over which ecc is computed
>  * @ecc_code: ecc code buffer
> @@ -888,20 +913,51 @@ int gpmc_enable_hwecc(int cs, int mode, int dev_width, int ecc_size)
>  * an erased page will produce an ECC mismatch between generated and read
>  * ECC bytes that has to be dealt with separately.
>  */
> -int gpmc_calculate_ecc(int cs, const u_char *dat, u_char *ecc_code)
> +int gpmc_calculate_ecc(int ecc_type, int cs,
> +                       const u_char *dat, u_char *ecc_code)
>  {
> -       unsigned int val = 0x0;
> -
> -       if (gpmc_ecc_used != cs)
> -               return -EINVAL;
> +       unsigned int reg;
> +       unsigned int val1 = 0x0, val2 = 0x0;
> +       unsigned int val3 = 0x0, val4 = 0x0;
> +       int i;
>
> -       /* read ecc result */
> -       val = gpmc_read_reg(GPMC_ECC1_RESULT);
> -       *ecc_code++ = val;          /* P128e, ..., P1e */
> -       *ecc_code++ = val >> 16;    /* P128o, ..., P1o */
> -       /* P2048o, P1024o, P512o, P256o, P2048e, P1024e, P512e, P256e */
> -       *ecc_code++ = ((val >> 8) & 0x0f) | ((val >> 20) & 0xf0);
> +       if ((ecc_type == OMAP_ECC_BCH4_CODE_HW) ||
> +               (ecc_type == OMAP_ECC_BCH8_CODE_HW)) {
> +               for (i = 0; i < 4; i++) {
> +                       /*
> +                        * Reading HW ECC_BCH_Results
> +                        * 0x240-0x24C, 0x250-0x25C, 0x260-0x26C, 0x270-0x27C
> +                        */
> +                       reg =  GPMC_ECC_BCH_RESULT_0 + (0x10 * i);
> +                       val1 = gpmc_read_reg(reg);
> +                       val2 = gpmc_read_reg(reg + 4);
> +                       if (ecc_type == OMAP_ECC_BCH8_CODE_HW) {
> +                               val3 = gpmc_read_reg(reg + 8);
> +                               val4 = gpmc_read_reg(reg + 12);
> +
> +                               *ecc_code++ = (val4 & 0xFF);
> +                               *ecc_code++ = ((val3 >> 24) & 0xFF);
> +                               *ecc_code++ = ((val3 >> 16) & 0xFF);
> +                               *ecc_code++ = ((val3 >> 8) & 0xFF);
> +                               *ecc_code++ = (val3 & 0xFF);
> +                               *ecc_code++ = ((val2 >> 24) & 0xFF);
> +                       }
> +                       *ecc_code++ = ((val2 >> 16) & 0xFF);
> +                       *ecc_code++ = ((val2 >> 8) & 0xFF);
> +                       *ecc_code++ = (val2 & 0xFF);
> +                       *ecc_code++ = ((val1 >> 24) & 0xFF);
> +                       *ecc_code++ = ((val1 >> 16) & 0xFF);
> +                       *ecc_code++ = ((val1 >> 8) & 0xFF);
> +                       *ecc_code++ = (val1 & 0xFF);
> +               }
> +       } else {
> +               /* read ecc result */
> +               val1 = gpmc_read_reg(GPMC_ECC1_RESULT);
> +               *ecc_code++ = val1;          /* P128e, ..., P1e */
> +               *ecc_code++ = val1 >> 16;    /* P128o, ..., P1o */
> +               /* P2048o, P1024o, P512o, P256o, P2048e, P1024e, P512e, P256e */
> +               *ecc_code++ = ((val1 >> 8) & 0x0f) | ((val1 >> 20) & 0xf0);
> +       }
>
> -       gpmc_ecc_used = -EINVAL;
>        return 0;
>  }
> diff --git a/arch/arm/plat-omap/include/plat/gpmc.h b/arch/arm/plat-omap/include/plat/gpmc.h
> index 49aea09..838c185 100644
> --- a/arch/arm/plat-omap/include/plat/gpmc.h
> +++ b/arch/arm/plat-omap/include/plat/gpmc.h
> @@ -92,6 +92,8 @@ enum omap_ecc {
>        OMAP_ECC_HAMMING_CODE_HW, /* gpmc to detect the error */
>                /* 1-bit ecc: stored at begining of spare area as romcode */
>        OMAP_ECC_HAMMING_CODE_HW_ROMCODE, /* gpmc method & romcode layout */
> +       OMAP_ECC_BCH4_CODE_HW, /* gpmc bch detection & s/w method correction */
> +       OMAP_ECC_BCH8_CODE_HW, /* gpmc bch detection & s/w method correction */
>  };
>
>  /*
> @@ -156,6 +158,6 @@ extern int gpmc_cs_configure(int cs, int cmd, int wval);
>  extern int gpmc_nand_read(int cs, int cmd);
>  extern int gpmc_nand_write(int cs, int cmd, int wval);
>
> -int gpmc_enable_hwecc(int cs, int mode, int dev_width, int ecc_size);
> -int gpmc_calculate_ecc(int cs, const u_char *dat, u_char *ecc_code);
> +int gpmc_enable_hwecc(int ecc, int cs, int mode, int dev_width, int ecc_size);
> +int gpmc_calculate_ecc(int ecc, int cs, const u_char *dat, u_char *ecc_code);
>  #endif
> diff --git a/drivers/mtd/nand/Makefile b/drivers/mtd/nand/Makefile
> index 8ad6fae..ae02711 100644
> --- a/drivers/mtd/nand/Makefile
> +++ b/drivers/mtd/nand/Makefile
> @@ -29,6 +29,7 @@ obj-$(CONFIG_MTD_NAND_NDFC)           += ndfc.o
>  obj-$(CONFIG_MTD_NAND_ATMEL)           += atmel_nand.o
>  obj-$(CONFIG_MTD_NAND_GPIO)            += gpio.o
>  obj-$(CONFIG_MTD_NAND_OMAP2)           += omap2.o
> +obj-$(CONFIG_MTD_NAND_OMAP2)           += omap_bch_decoder.o
>  obj-$(CONFIG_MTD_NAND_CM_X270)         += cmx270_nand.o
>  obj-$(CONFIG_MTD_NAND_PXA3xx)          += pxa3xx_nand.o
>  obj-$(CONFIG_MTD_NAND_TMIO)            += tmio_nand.o
> diff --git a/drivers/mtd/nand/omap2.c b/drivers/mtd/nand/omap2.c
> index 4e33972..14c7dfe 100644
> --- a/drivers/mtd/nand/omap2.c
> +++ b/drivers/mtd/nand/omap2.c
> @@ -98,6 +98,8 @@
>  static const char *part_probes[] = { "cmdlinepart", NULL };
>  #endif
>
> +int decode_bch(int select_4_8, unsigned char *ecc, unsigned int *err_loc);
> +
>  /* oob info generated runtime depending on ecc algorithm and layout selected */
>  static struct nand_ecclayout omap_oobinfo;
>  /* Define some generic bad / good block scan pattern which are used
> @@ -130,7 +132,8 @@ struct omap_nand_info {
>                OMAP_NAND_IO_WRITE,     /* write */
>        } iomode;
>        u_char                          *buf;
> -       int                                     buf_len;
> +       int                             buf_len;
> +       int                             ecc_opt;
>  };
>
>  /**
> @@ -529,7 +532,6 @@ static void omap_read_buf_irq_pref(struct mtd_info *mtd, u_char *buf, int len)
>        struct omap_nand_info *info = container_of(mtd,
>                                                struct omap_nand_info, mtd);
>        int ret = 0;
> -
>        if (len <= mtd->oobsize) {
>                omap_read_buf_pref(mtd, buf, len);
>                return;
> @@ -803,6 +805,8 @@ static int omap_correct_data(struct mtd_info *mtd, u_char *dat,
>        struct omap_nand_info *info = container_of(mtd, struct omap_nand_info,
>                                                        mtd);
>        int blockCnt = 0, i = 0, ret = 0;
> +       int j, eccsize, eccflag, count;
> +       unsigned int err_loc[8];
>
>        /* Ex NAND_ECC_HW12_2048 */
>        if ((info->nand.ecc.mode == NAND_ECC_HW) &&
> @@ -811,16 +815,57 @@ static int omap_correct_data(struct mtd_info *mtd, u_char *dat,
>        else
>                blockCnt = 1;
>
> -       for (i = 0; i < blockCnt; i++) {
> -               if (memcmp(read_ecc, calc_ecc, 3) != 0) {
> -                       ret = omap_compare_ecc(read_ecc, calc_ecc, dat);
> -                       if (ret < 0)
> -                               return ret;
> +       switch (info->ecc_opt) {
> +       case OMAP_ECC_HAMMING_CODE_HW:
> +       case OMAP_ECC_HAMMING_CODE_HW_ROMCODE:
> +               for (i = 0; i < blockCnt; i++) {
> +                       if (memcmp(read_ecc, calc_ecc, 3) != 0) {
> +                               ret = omap_compare_ecc(read_ecc, calc_ecc, dat);
> +                               if (ret < 0)
> +                                       return ret;
> +                       }
> +                       read_ecc += 3;
> +                       calc_ecc += 3;
> +                       dat      += 512;
>                }
> -               read_ecc += 3;
> -               calc_ecc += 3;
> -               dat      += 512;
> +               break;
> +
> +       case OMAP_ECC_BCH4_CODE_HW:
> +               eccsize = 7;
> +               gpmc_calculate_ecc(info->ecc_opt, info->gpmc_cs, dat, calc_ecc);
> +               for (i = 0; i < blockCnt; i++) {
> +                       /* check if any ecc error */
> +                       eccflag = 0;
> +                       for (j = 0; (j < eccsize) && (eccflag == 0); j++)
> +                               if (calc_ecc[j] != 0)
> +                                       eccflag = 1;
> +
> +                       if (eccflag == 1) {
> +                               eccflag = 0;
> +                               for (j = 0; (j < eccsize) &&
> +                                               (eccflag == 0); j++)
> +                                       if (read_ecc[j] != 0xFF)
> +                                               eccflag = 1;
> +                       }
> +
> +                       count = 0;
> +                       if (eccflag == 1)
> +                               count = decode_bch(0, calc_ecc, err_loc);
> +
> +                       for (j = 0; j < count; j++) {
> +                               if (err_loc[j] < 4096)
> +                                       dat[err_loc[j] >> 3] ^=
> +                                                       1 << (err_loc[j] & 7);
> +                               /* else, not interested to correct ecc */
> +                       }
> +
> +                       calc_ecc = calc_ecc + eccsize;
> +                       read_ecc = read_ecc + eccsize;
> +                       dat += 512;
> +               }
> +               break;
>        }
> +
>        return 0;
>  }
>
> @@ -841,7 +886,7 @@ static int omap_calculate_ecc(struct mtd_info *mtd, const u_char *dat,
>  {
>        struct omap_nand_info *info = container_of(mtd, struct omap_nand_info,
>                                                        mtd);
> -       return gpmc_calculate_ecc(info->gpmc_cs, dat, ecc_code);
> +       return gpmc_calculate_ecc(info->ecc_opt, info->gpmc_cs, dat, ecc_code);
>  }
>
>  /**
> @@ -856,7 +901,8 @@ static void omap_enable_hwecc(struct mtd_info *mtd, int mode)
>        struct nand_chip *chip = mtd->priv;
>        unsigned int dev_width = (chip->options & NAND_BUSWIDTH_16) ? 1 : 0;
>
> -       gpmc_enable_hwecc(info->gpmc_cs, mode, dev_width, info->nand.ecc.size);
> +       gpmc_enable_hwecc(info->ecc_opt, info->gpmc_cs, mode,
> +                               dev_width, info->nand.ecc.size);
>  }
>
>  /**
> @@ -953,6 +999,7 @@ static int __devinit omap_nand_probe(struct platform_device *pdev)
>        info->mtd.priv          = &info->nand;
>        info->mtd.name          = dev_name(&pdev->dev);
>        info->mtd.owner         = THIS_MODULE;
> +       info->ecc_opt           = pdata->ecc_opt;
>
>        info->nand.options      = pdata->devsize;
>        info->nand.options      |= NAND_SKIP_BBTSCAN;
> @@ -991,7 +1038,6 @@ static int __devinit omap_nand_probe(struct platform_device *pdev)
>                info->nand.waitfunc = omap_wait;
>                info->nand.chip_delay = 50;
>        }
> -
>        switch (pdata->xfer_type) {
>        case NAND_OMAP_PREFETCH_POLLED:
>                info->nand.read_buf   = omap_read_buf_pref;
> @@ -1052,10 +1098,17 @@ static int __devinit omap_nand_probe(struct platform_device *pdev)
>        /* selsect the ecc type */
>        if (pdata->ecc_opt == OMAP_ECC_HAMMING_CODE_DEFAULT)
>                info->nand.ecc.mode = NAND_ECC_SOFT;
> -       else if ((pdata->ecc_opt == OMAP_ECC_HAMMING_CODE_HW) ||
> -               (pdata->ecc_opt == OMAP_ECC_HAMMING_CODE_HW_ROMCODE)) {
> -               info->nand.ecc.bytes            = 3;
> -               info->nand.ecc.size             = 512;
> +       else {
> +               if (pdata->ecc_opt == OMAP_ECC_BCH4_CODE_HW) {
> +                       info->nand.ecc.bytes    = 4*7;
> +                       info->nand.ecc.size     = 4*512;
> +               } else if (pdata->ecc_opt == OMAP_ECC_BCH8_CODE_HW) {
> +                       info->nand.ecc.bytes    = 13;
> +                       info->nand.ecc.size     = 4*512;
> +               } else {
> +                       info->nand.ecc.bytes    = 3;
> +                       info->nand.ecc.size     = 512;
> +               }
>                info->nand.ecc.calculate        = omap_calculate_ecc;
>                info->nand.ecc.hwctl            = omap_enable_hwecc;
>                info->nand.ecc.correct          = omap_correct_data;
> @@ -1073,8 +1126,8 @@ static int __devinit omap_nand_probe(struct platform_device *pdev)
>                }
>        }
>
> -       /* rom code layout */
> -       if (pdata->ecc_opt == OMAP_ECC_HAMMING_CODE_HW_ROMCODE) {
> +       /* select ecc lyout */
> +       if (info->nand.ecc.mode != NAND_ECC_SOFT) {
>
>                if (info->nand.options & NAND_BUSWIDTH_16)
>                        offset = 2;
> @@ -1082,15 +1135,31 @@ static int __devinit omap_nand_probe(struct platform_device *pdev)
>                        offset = 1;
>                        info->nand.badblock_pattern = &bb_descrip_flashbased;
>                }
> -               omap_oobinfo.eccbytes = 3 * (info->mtd.oobsize/16);
> -               for (i = 0; i < omap_oobinfo.eccbytes; i++)
> -                       omap_oobinfo.eccpos[i] = i+offset;
>
> -               omap_oobinfo.oobfree->offset = offset + omap_oobinfo.eccbytes;
> -               omap_oobinfo.oobfree->length = info->mtd.oobsize -
> -                                       (offset + omap_oobinfo.eccbytes);
> +               if (info->mtd.oobsize == 64)
> +                       omap_oobinfo.eccbytes = info->nand.ecc.bytes *
> +                                               2048/info->nand.ecc.size;
> +               else
> +                       omap_oobinfo.eccbytes = info->nand.ecc.bytes;
> +
> +               if (pdata->ecc_opt == OMAP_ECC_HAMMING_CODE_HW_ROMCODE) {
> +                       for (i = 0; i < omap_oobinfo.eccbytes; i++)
> +                               omap_oobinfo.eccpos[i] = i + offset;
> +                       omap_oobinfo.oobfree->offset =
> +                                               offset + omap_oobinfo.eccbytes;
> +                       omap_oobinfo.oobfree->length = info->mtd.oobsize -
> +                                               offset - omap_oobinfo.eccbytes;
> +               } else {
>
> +                       omap_oobinfo.oobfree->offset = offset;
> +                       omap_oobinfo.oobfree->length = info->mtd.oobsize -
> +                                               offset - omap_oobinfo.eccbytes;
> +                       offset = info->mtd.oobsize - omap_oobinfo.eccbytes;
> +                       for (i = 0; i < omap_oobinfo.eccbytes; i++)
> +                               omap_oobinfo.eccpos[i] = i + offset;
> +               }
>                info->nand.ecc.layout = &omap_oobinfo;
> +
>        }
>
>  #ifdef CONFIG_MTD_PARTITIONS
> diff --git a/drivers/mtd/nand/omap_bch_decoder.c b/drivers/mtd/nand/omap_bch_decoder.c
> new file mode 100644
> index 0000000..da42bda
> --- /dev/null
> +++ b/drivers/mtd/nand/omap_bch_decoder.c
> @@ -0,0 +1,393 @@
> +/*
> + * drivers/mtd/nand/omap_omap_bch_decoder.c
> + *
> + * Whole BCH ECC Decoder (Post hardware generated syndrome decoding)
> + *
> + * Copyright (c) 2007 Texas Instruments
> + *
> + * Author: Sukumar Ghorai <s-ghorai at ti.com
> + *                Michael Fillinger <m-fillinger at ti.com>
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + */
> +#undef DEBUG
> +
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +
> +#define mm             13
> +#define kk_shorten     4096
> +#define nn             8191    /* Length of codeword, n = 2**mm - 1 */
> +
> +#define PPP    0x201B  /* Primary Polynomial : x^13 + x^4 + x^3 + x + 1 */
> +#define P      0x001B  /* With omitted x^13 */
> +#define POLY   12      /* degree of the primary Polynomial less one */
> +
> +/**
> + * mpy_mod_gf - GALOIS field multiplier
> + * Input  : A(x), B(x)
> + * Output : A(x)*B(x) mod P(x)
> + */
> +static unsigned int mpy_mod_gf(unsigned int a, unsigned int b)
> +{
> +       unsigned int R = 0;
> +       unsigned int R1 = 0;
> +       unsigned int k = 0;
> +
> +       for (k = 0; k < mm; k++) {
> +
> +               R = (R << 1) & 0x1FFE;
> +               if (R1 == 1)
> +                       R ^= P;
> +
> +               if (((a >> (POLY - k)) & 1) == 1)
> +                       R ^= b;
> +
> +               if (k < POLY)
> +                       R1 = (R >> POLY) & 1;
> +       }
> +       return R;
> +}
> +
> +/**
> + * chien - CHIEN search
> + *
> + * @location - Error location vector pointer
> + *
> + * Inputs  : ELP(z)
> + *          No. of found errors
> + *          Size of input codeword
> + * Outputs : Up to 8 locations
> + *          No. of errors
> + */
> +static int chien(unsigned int select_4_8, int err_nums,
> +                               unsigned int err[], unsigned int *location)
> +{
> +       int i, count; /* Number of dectected errors */
> +       /* Contains accumulation of evaluation at x^i (i:1->8) */
> +       unsigned int gammas[8] = {0};
> +       unsigned int alpha;
> +       unsigned int bit, ecc_bits;
> +       unsigned int elp_sum;
> +
> +       ecc_bits = (select_4_8 == 0) ? 52 : 104;
> +
> +       /* Start evaluation at Alpha**8192 and decreasing */
> +       for (i = 0; i < 8; i++)
> +               gammas[i] = err[i];
> +
> +       count = 0;
> +       for (i = 1; (i <= nn) && (count < err_nums); i++) {
> +
> +               /* Result of evaluation at root */
> +               elp_sum = 1 ^ gammas[0] ^ gammas[1] ^
> +                               gammas[2] ^ gammas[3] ^
> +                               gammas[4] ^ gammas[5] ^
> +                               gammas[6] ^ gammas[7];
> +
> +               alpha = PPP >> 1;
> +               gammas[0] = mpy_mod_gf(gammas[0], alpha);
> +               alpha = mpy_mod_gf(alpha, (PPP >> 1));  /* x alphha^-2 */
> +               gammas[1] = mpy_mod_gf(gammas[1], alpha);
> +               alpha = mpy_mod_gf(alpha, (PPP >> 1));  /* x alphha^-2 */
> +               gammas[2] = mpy_mod_gf(gammas[2], alpha);
> +               alpha = mpy_mod_gf(alpha, (PPP >> 1));  /* x alphha^-3 */
> +               gammas[3] = mpy_mod_gf(gammas[3], alpha);
> +               alpha = mpy_mod_gf(alpha, (PPP >> 1));  /* x alphha^-4 */
> +               gammas[4] = mpy_mod_gf(gammas[4], alpha);
> +               alpha = mpy_mod_gf(alpha, (PPP >> 1));  /* x alphha^-5 */
> +               gammas[5] = mpy_mod_gf(gammas[5], alpha);
> +               alpha = mpy_mod_gf(alpha, (PPP >> 1));  /* x alphha^-6 */
> +               gammas[6] = mpy_mod_gf(gammas[6], alpha);
> +               alpha = mpy_mod_gf(alpha, (PPP >> 1));  /* x alphha^-7 */
> +               gammas[7] = mpy_mod_gf(gammas[7], alpha);
> +
> +               if (elp_sum == 0) {
> +                       /* calculate bit position in main data area */
> +                       bit = ((i-1) & ~7)|(7-((i-1) & 7));
> +                       if (i >= 2 * ecc_bits)
> +                               location[count++] =
> +                                       kk_shorten - (bit - 2 * ecc_bits) - 1;
> +               }
> +       }
> +
> +       /* Failure: No. of detected errors != No. or corrected errors */
> +       if (count != err_nums) {
> +               count = -1;
> +               printk(KERN_ERR "BCH decoding failed\n");
> +       }
> +       for (i = 0; i < count; i++)
> +               pr_debug("%d ", location[i]);
> +
> +       return count;
> +}
> +
> +/* synd : 16 Syndromes
> + * return: gamaas - Coefficients to the error polynomial
> + * return: : Number of detected errors
> +*/
> +static unsigned int berlekamp(unsigned int select_4_8,
> +                       unsigned int synd[], unsigned int err[])
> +{
> +       int loop, iteration;
> +       unsigned int LL = 0;            /* Detected errors */
> +       unsigned int d = 0;     /* Distance between Syndromes and ELP[n](z) */
> +       unsigned int invd = 0;          /* Inverse of d */
> +       /* Intermediate ELP[n](z).
> +        * Final ELP[n](z) is Error Location Polynomial
> +        */
> +       unsigned int gammas[16] = {0};
> +       /* Intermediate normalized ELP[n](z) : D[n](z) */
> +       unsigned int D[16] = {0};
> +       /* Temporary value that holds an ELP[n](z) coefficient */
> +       unsigned int next_gamma = 0;
> +
> +       int e = 0;
> +       unsigned int sign = 0;
> +       unsigned int u = 0;
> +       unsigned int v = 0;
> +       unsigned int C1 = 0, C2 = 0;
> +       unsigned int ss = 0;
> +       unsigned int tmp_v = 0, tmp_s = 0;
> +       unsigned int tmp_poly;
> +
> +       /*-------------- Step 0 ------------------*/
> +       for (loop = 0; loop < 16; loop++)
> +               gammas[loop] = 0;
> +       gammas[0] = 1;
> +       D[1] = 1;
> +
> +       iteration = 0;
> +       LL = 0;
> +       while ((iteration < ((select_4_8+1)*2*4)) &&
> +                       (LL <= ((select_4_8+1)*4))) {
> +
> +               pr_debug("\nIteration.............%d\n", iteration);
> +               d = 0;
> +               /* Step: 0 */
> +               for (loop = 0; loop <= LL; loop++) {
> +                       tmp_poly = mpy_mod_gf(
> +                                       gammas[loop], synd[iteration - loop]);
> +                       d ^= tmp_poly;
> +                       pr_debug("%02d. s=0 LL=%x poly %x\n",
> +                                       loop, LL, tmp_poly);
> +               }
> +
> +               /* Step 1: 1 cycle only to perform inversion */
> +               v = d << 1;
> +               e = -1;
> +               sign = 1;
> +               ss = 0x2000;
> +               invd = 0;
> +               u = PPP;
> +               for (loop = 0; (d != 0) && (loop <= (2 * POLY)); loop++) {
> +                       pr_debug("%02d. s=1 LL=%x poly NULL\n",
> +                                               loop, LL);
> +                       C1 = (v >> 13) & 1;
> +                       C2 = C1 & sign;
> +
> +                       sign ^= C2 ^ (e == 0);
> +
> +                       tmp_v = v;
> +                       tmp_s = ss;
> +
> +                       if (C1 == 1) {
> +                               v ^= u;
> +                               ss ^= invd;
> +                       }
> +                       v = (v << 1) & 0x3FFF;
> +                       if (C2 == 1) {
> +                               u = tmp_v;
> +                               invd = tmp_s;
> +                               e = -e;
> +                       }
> +                       invd >>= 1;
> +                       e--;
> +               }
> +
> +               for (loop = 0; (d != 0) && (loop <= (iteration + 1)); loop++) {
> +                       /* Step 2
> +                        * Interleaved with Step 3, if L<(n-k)
> +                        * invd: Update of ELP[n](z) = ELP[n-1](z) - d.D[n-1](z)
> +                        */
> +
> +                       /* Holds value of ELP coefficient until precedent
> +                        * value does not have to be used anymore
> +                        */
> +                       tmp_poly = mpy_mod_gf(d, D[loop]);
> +                       pr_debug("%02d. s=2 LL=%x poly %x\n",
> +                                               loop, LL, tmp_poly);
> +
> +                       next_gamma = gammas[loop] ^ tmp_poly;
> +                       if ((2 * LL) < (iteration + 1)) {
> +                               /* Interleaving with Step 3
> +                                * for parallelized update of ELP(z) and D(z)
> +                                */
> +                       } else {
> +                               /* Update of ELP(z) only -> stay in Step 2 */
> +                               gammas[loop] = next_gamma;
> +                               if (loop == (iteration + 1)) {
> +                                       /* to step 4 */
> +                                       break;
> +                               }
> +                       }
> +
> +                       /* Step 3
> +                        * Always interleaved with Step 2 (case when L<(n-k))
> +                        * Update of D[n-1](z) = ELP[n-1](z)/d
> +                        */
> +                       D[loop] = mpy_mod_gf(gammas[loop], invd);
> +                       pr_debug("%02d. s=3 LL=%x poly %x\n",
> +                                       loop, LL, D[loop]);
> +
> +                       /* Can safely update ELP[n](z) */
> +                       gammas[loop] = next_gamma;
> +
> +                       if (loop == (iteration + 1)) {
> +                               /* If update finished */
> +                               LL = iteration - LL + 1;
> +                               /* to step 4 */
> +                               break;
> +                       }
> +                       /* Else, interleaving to step 2*/
> +               }
> +
> +               /* Step 4: Update D(z): i:0->L */
> +               /* Final update of D[n](z) = D[n](z).z*/
> +               for (loop = 0; loop < 15; loop++) /* Left Shift */
> +                       D[15 - loop] = D[14 - loop];
> +
> +               D[0] = 0;
> +
> +               iteration++;
> +       } /* while */
> +
> +       /* Processing finished, copy ELP to final registers : 0->2t-1*/
> +       for (loop = 0; loop < 8; loop++)
> +               err[loop] = gammas[loop+1];
> +
> +       pr_debug("\n Err poly:");
> +       for (loop = 0; loop < 8; loop++)
> +               pr_debug("0x%x ", err[loop]);
> +
> +       return LL;
> +}
> +
> +/*
> + * syndrome - Generate syndrome components from hw generate syndrome
> + * r(x) = c(x) + e(x)
> + * s(x) = c(x) mod g(x) + e(x) mod g(x) =  e(x) mod g(x)
> + * so receiver checks if the syndrome s(x) = r(x) mod g(x) is equal to zero.
> + * unsigned int s[16]; - Syndromes
> + */
> +static void syndrome(unsigned int select_4_8,
> +                                       unsigned char *ecc, unsigned int syn[])
> +{
> +       unsigned int k, l, t;
> +       unsigned int alpha_bit, R_bit;
> +       int ecc_pos, ecc_min;
> +
> +       /* 2t-1 = 15 (for t=8) minimal polynomials of the first 15 powers of a
> +        * primitive elemmants of GF(m); Even powers minimal polynomials are
> +        * duplicate of odd powers' minimal polynomials.
> +        * Odd powers of alpha (1 to 15)
> +        */
> +       unsigned int pow_alpha[8] = {0x0002, 0x0008, 0x0020, 0x0080,
> +                                0x0200, 0x0800, 0x001B, 0x006C};
> +
> +       pr_debug("\n ECC[0..n]: ");
> +       for (k = 0; k < 13; k++)
> +               pr_debug("0x%x ", ecc[k]);
> +
> +       if (select_4_8 == 0) {
> +               t = 4;
> +               ecc_pos = 55; /* bits(52-bits): 55->4 */
> +               ecc_min = 4;
> +       } else {
> +               t = 8;
> +               ecc_pos = 103; /* bits: 103->0 */
> +               ecc_min = 0;
> +       }
> +
> +       /* total numbber of syndrom to be used is 2t */
> +       /* Step1: calculate the odd syndrome(s) */
> +       R_bit = ((ecc[ecc_pos/8] >> (7 - ecc_pos%8)) & 1);
> +       ecc_pos--;
> +       for (k = 0; k < t; k++)
> +               syn[2 * k] = R_bit;
> +
> +       while (ecc_pos >= ecc_min) {
> +               R_bit = ((ecc[ecc_pos/8] >> (7 - ecc_pos%8)) & 1);
> +               ecc_pos--;
> +
> +               for (k = 0; k < t; k++) {
> +                       /* Accumulate value of x^i at alpha^(2k+1) */
> +                       if (R_bit == 1)
> +                               syn[2*k] ^= pow_alpha[k];
> +
> +                       /* Compute a**(2k+1), using LSFR */
> +                       for (l = 0; l < (2 * k + 1); l++) {
> +                               alpha_bit = (pow_alpha[k] >> POLY) & 1;
> +                               pow_alpha[k] = (pow_alpha[k] << 1) & 0x1FFF;
> +                               if (alpha_bit == 1)
> +                                       pow_alpha[k] ^= P;
> +                       }
> +               }
> +       }
> +
> +       /* Step2: calculate the even syndrome(s)
> +        * Compute S(a), where a is an even power of alpha
> +        * Evenry even power of primitive element has the same minimal
> +        * polynomial as some odd power of elemets.
> +        * And based on S(a^2) = S^2(a)
> +        */
> +       for (k = 0; k < t; k++)
> +               syn[2*k+1] = mpy_mod_gf(syn[k], syn[k]);
> +
> +       pr_debug("\n Syndromes: ");
> +       for (k = 0; k < 16; k++)
> +               pr_debug("0x%x ", syn[k]);
> +}
> +
> +/**
> + * decode_bch - BCH decoder for 4- and 8-bit error correction
> + *
> + * @ecc - ECC syndrome generated by hw BCH engine
> + * @err_loc - pointer to error location array
> + *
> + * This function does post sydrome generation (hw generated) decoding
> + * for:-
> + * Dimension of Galoise Field: m = 13
> + * Length of codeword: n = 2**m - 1
> + * Number of errors that can be corrected: 4- or 8-bits
> + * Length of information bit: kk = nn - rr
> + */
> +int decode_bch(int select_4_8, unsigned char *ecc, unsigned int *err_loc)
> +{
> +       int no_of_err;
> +       unsigned int syn[16] = {0,};    /* 16 Syndromes */
> +       unsigned int err_poly[8] = {0,};
> +       /* Coefficients to the error polynomial
> +        * ELP(x) = 1 + err0.x + err1.x^2 + ... + err7.x^8
> +        */
> +
> +       /* Decoting involes three steps
> +        * 1. Compute the syndrom from teh received codeword,
> +        * 2. Find the error location polynomial from a set of equations
> +        *     derived from the syndrome,
> +        * 3. Use the error location polynomial to identify errants bits,
> +        *
> +        * And correcttion done by bit flips using error locaiton and expected
> +        * to be outseide of this implementation.
> +        */
> +       syndrome(select_4_8, ecc, syn);
> +       no_of_err = berlekamp(select_4_8, syn, err_poly);
> +       if (no_of_err <= (4 << select_4_8))
> +               no_of_err = chien(select_4_8, no_of_err, err_poly, err_loc);
> +
> +       return no_of_err;
> +}
> +EXPORT_SYMBOL(decode_bch);
> +
> --
> 1.7.0.4
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>



-- 
Regards,
Vimal Singh



More information about the linux-mtd mailing list