[RFC PATCH] spi: spi-nor: Add dual flash support in spi-nor framework
Ashok Reddy Soma
ashokred at xilinx.com
Wed Dec 4 16:11:17 CET 2019
Hi Vignesh, Jagan,
Thanks for the review comments.
I will split in to patches and revert.
Vignesh,
I will do some cleanup and try to address your comments wherever applicable in the next drop.
Thanks,
Ashok
> -----Original Message-----
> From: Vignesh Raghavendra <vigneshr at ti.com>
> Sent: Wednesday, December 4, 2019 3:36 PM
> To: Ashok Reddy Soma <ashokred at xilinx.com>; u-boot at lists.denx.de;
> jagan at amarulasolutions.com
> Cc: Michal Simek <michals at xilinx.com>
> Subject: Re: [RFC PATCH] spi: spi-nor: Add dual flash support in spi-nor
> framework
>
> Hi,
>
> On 26/11/19 12:39 pm, Michal Simek wrote:
> > On 19. 11. 19 15:20, Ashok Reddy Soma wrote:
> >> Add dual parallel and dual stacked support in spi-nor framework.
> >> Add dual flash support for nor-scan, read and write.
> >>
>
> How does the DT representation of these flashes look like?
> Is it in alignment with Linux proposal here:
> https://patchwork.ozlabs.org/patch/1201508/?
>
> Also could you provide a concise summary of what changes in programming
> sequence to support dual flashes? This will ease process of review.
>
> >> Signed-off-by: Ashok Reddy Soma <ashok.reddy.soma at xilinx.com>
> >> ---
> [...]
>
> >> diff --git a/drivers/mtd/spi/sf_internal.h
> >> b/drivers/mtd/spi/sf_internal.h index 5c643034c6..d836261f01 100644
> >> --- a/drivers/mtd/spi/sf_internal.h
> >> +++ b/drivers/mtd/spi/sf_internal.h
> >> @@ -15,6 +15,15 @@
> >> #define SPI_NOR_MAX_ID_LEN 6
> >> #define SPI_NOR_MAX_ADDR_WIDTH 4
> >>
> >> +/* Dual SPI flash memories - see SPI_COMM_DUAL_... */ enum
> >> +spi_dual_flash {
> >> + SF_SINGLE_FLASH = 0,
> >> + SF_DUAL_STACKED_FLASH = BIT(0),
> >> + SF_DUAL_PARALLEL_FLASH = BIT(1),
> >> +};
> >> +
> >> +#define SPI_FLASH_16MB_BOUN 0x1000000
> >> +
>
> Do we need this constant? Can we just use SZ_16M instead?
>
> >> struct flash_info {
> >> #if !CONFIG_IS_ENABLED(SPI_FLASH_TINY)
> >> char *name;
> >> diff --git a/drivers/mtd/spi/spi-nor-core.c
> >> b/drivers/mtd/spi/spi-nor-core.c index 5a8c084255..a53c78ff22 100644
> >> --- a/drivers/mtd/spi/spi-nor-core.c
> >> +++ b/drivers/mtd/spi/spi-nor-core.c
> >> @@ -21,6 +21,7 @@
> >> #include <spi-mem.h>
> >> #include <spi.h>
> >>
> >> +#include <linux/mtd/spi-nor.h>
> >> #include "sf_internal.h"
> >>
> >> /* Define max times to check status register before we give up. */
> >> @@ -34,6 +35,11 @@
> >>
> >> #define DEFAULT_READY_WAIT_JIFFIES (40UL * HZ)
> >>
> >> +#define SPI_MASTER_QUAD_MODE BIT(6) /* support quad
> mode */
>
> Why can't SPI_RX_QUAD/SPI_TX_QUAD be used?
>
> >> +#define SPI_MASTER_DATA_STRIPE BIT(7) /* support data stripe
> */
>
> What does this mean? In particular what is "data stripe"?
>
> >> +#define SPI_MASTER_BOTH_CS BIT(8) /* assert both chip selects
> */
>
> I don't see any users of above flags.
>
> >> +#define SPI_MASTER_U_PAGE BIT(9) /* select upper flash */
> >> +
>
> These macros need to be moved to include/spi.h where rest of the flag fields
> are defined.
> I see this flag is being set in nor->spi->flags, but who uses it? Its very difficult
> to review patches when there are no users. Without any users, this comes a
> dead code which will be dropped while cleaning up.
>
>
> >> static int spi_nor_read_write_reg(struct spi_nor *nor, struct
> spi_mem_op
> >> *op, void *buf)
> >> {
> >> @@ -146,15 +152,24 @@ static ssize_t spi_nor_write_data(struct
> >> spi_nor *nor, loff_t to, size_t len, static int read_sr(struct
> >> spi_nor *nor) {
> >> int ret;
> >> - u8 val;
> >> + u8 val[2];
> >>
> >> - ret = nor->read_reg(nor, SPINOR_OP_RDSR, &val, 1);
> >> - if (ret < 0) {
> >> + if (nor->isparallel) {
> >> + ret = nor->read_reg(nor, SPINOR_OP_RDSR, &val[0], 2);
> >> + if (ret < 0) {
> >> + pr_debug("error %d reading SR\n", (int)ret);
> >> + return ret;
> >> + }
> >> + val[0] |= val[1];
> >> + } else {
> >> + ret = nor->read_reg(nor, SPINOR_OP_RDSR, &val[0], 1);
> >> + if (ret < 0) {
> >> pr_debug("error %d reading SR\n", (int)ret);
> >> return ret;
>
>
> Above bit of code is repeated more than once in this patch.
> You should refractor and add a wrapper for nor->read_reg() that takes care
> of these manipulations for dual parallel flash. Can it be extend further to
> handle stacked as well?
>
> Individual helpers (such as read_fsr() below) to read/write registers should
> not have to bother about flash geometry but just use above helper.
>
> >> + }
> >> }
> >>
> >> - return val;
> >> + return val[0];
> >> }
> >>
> >> /*
> >> @@ -165,15 +180,24 @@ static int read_sr(struct spi_nor *nor) static
> >> int read_fsr(struct spi_nor *nor) {
> >> int ret;
> >> - u8 val;
> >> + u8 val[2];
> >>
> >> - ret = nor->read_reg(nor, SPINOR_OP_RDFSR, &val, 1);
> >> - if (ret < 0) {
> >> + if (nor->isparallel) {
> >> + ret = nor->read_reg(nor, SPINOR_OP_RDFSR, &val[0], 2);
> >> + if (ret < 0) {
> >> + pr_debug("error %d reading SR\n", (int)ret);
> >> + return ret;
> >> + }
> >> + val[0] &= val[1];
> >> + } else {
> >> + ret = nor->read_reg(nor, SPINOR_OP_RDFSR, &val[0], 1);
> >> + if (ret < 0) {
> >> pr_debug("error %d reading FSR\n", ret);
> >> return ret;
> >> + }
> >> }
> >>
> >> - return val;
> >> + return val[0];
> >> }
> >>
> >> /*
> >> @@ -288,12 +312,17 @@ static u8 spi_nor_convert_3to4_erase(u8
> opcode)
> >> static void spi_nor_set_4byte_opcodes(struct spi_nor *nor,
> >> const struct flash_info *info) {
> >> + bool shift = 0;
> >> +
> >> + if (nor->isparallel)
> >> + shift = 1;
> >> +
> >> /* Do some manufacturer fixups first */
> >> switch (JEDEC_MFR(info)) {
> >> case SNOR_MFR_SPANSION:
> >> /* No small sector erase for 4-byte command set */
> >> nor->erase_opcode = SPINOR_OP_SE;
> >> - nor->mtd.erasesize = info->sector_size;
> >> + nor->mtd.erasesize = info->sector_size << shift;
> >> break;
> >>
> >> default:
> >> @@ -465,12 +494,19 @@ static int clean_bar(struct spi_nor *nor)
> >>
> >> static int write_bar(struct spi_nor *nor, u32 offset) {
> >> - u8 cmd, bank_sel;
> >> + u8 cmd, bank_sel, upage_curr;
> >> int ret;
> >>
> >> - bank_sel = offset / SZ_16M;
> >> - if (bank_sel == nor->bank_curr)
> >> - goto bar_end;
> >> + bank_sel = offset / (SZ_16M << nor->shift);
> >> +
> >> + upage_curr = nor->spi->flags & SPI_XFER_U_PAGE;
> >> +
> >> + if ((!nor->isstacked) || (nor->upage_prev == upage_curr)) {
> >> + if (bank_sel == nor->bank_curr)
> >> + goto bar_end;
> >> + } else {
> >> + nor->upage_prev = upage_curr;
> >> + }
> >>
> >> cmd = nor->bank_write_cmd;
> >> write_enable(nor);
> >> @@ -488,8 +524,12 @@ bar_end:
> >> static int read_bar(struct spi_nor *nor, const struct flash_info
> >> *info) {
> >> u8 curr_bank = 0;
> >> + u8 curr_bank_up = 0;
> >> int ret;
> >>
> >> + if (nor->size <= SPI_FLASH_16MB_BOUN)
> >> + goto bar_end;
> >> +
>
> Hmm, is read_bar() called when flash size is <=16M?
>
> >> switch (JEDEC_MFR(info)) {
> >> case SNOR_MFR_SPANSION:
> >> nor->bank_read_cmd = SPINOR_OP_BRRD; @@ -500,12
> +540,31 @@ static
> >> int read_bar(struct spi_nor *nor, const struct flash_info *info)
> >> nor->bank_write_cmd = SPINOR_OP_WREAR;
> >> }
> >>
> >> - ret = nor->read_reg(nor, nor->bank_read_cmd,
> >> + if (nor->isparallel) {
> >> + nor->spi->flags |= SPI_XFER_LOWER;
> >> + ret = nor->read_reg(nor, nor->bank_read_cmd,
> >> &curr_bank, 1);
> >> - if (ret) {
> >> - debug("SF: fail to read bank addr register\n");
> >> - return ret;
> >> + if (ret)
> >> + return ret;
> >> +
> >> + nor->spi->flags |= SPI_XFER_UPPER;
> >> + ret = nor->read_reg(nor, nor->bank_read_cmd,
> >> + &curr_bank_up, 1);
> >> + if (ret)
> >> + return ret;
> >> + if (curr_bank != curr_bank_up) {
> >> + printf("Incorrect Bank selections Dual parallel\n");
> >> + return -EINVAL;
> >> + }
> >> + } else {
> >> + ret = nor->read_reg(nor, nor->bank_read_cmd,
> >> + &curr_bank, 1);
> >> + if (ret) {
> >> + debug("SF: fail to read bank addr register\n");
> >> + return ret;
> >> + }
> >> }
> >> +bar_end:
> >> nor->bank_curr = curr_bank;
> >>
> >> return 0;
> >> @@ -540,7 +599,7 @@ static int spi_nor_erase_sector(struct spi_nor
> >> *nor, u32 addr) static int spi_nor_erase(struct mtd_info *mtd,
> >> struct erase_info *instr) {
> >> struct spi_nor *nor = mtd_to_spi_nor(mtd);
> >> - u32 addr, len, rem;
> >> + u32 addr, len, rem, offset;
> >> int ret;
> >>
> >> dev_dbg(nor->dev, "at 0x%llx, len %lld\n", (long long)instr->addr,
> >> @@ -554,14 +613,37 @@ static int spi_nor_erase(struct mtd_info *mtd,
> struct erase_info *instr)
> >> len = instr->len;
> >>
> >> while (len) {
> >> + write_enable(nor);
> >> + offset = addr;
> >> + if (nor->isparallel) {
> >> + offset /= 2;
> >> + nor->spi->flags |= SPI_XFER_STRIPE;
> >> + }
> >> +
> >> + if (nor->isstacked) {
> >> + if (offset >= (mtd->size / 2)) {
> >> + offset = offset - (mtd->size / 2);
> >> + nor->spi->flags |= SPI_MASTER_U_PAGE;
> >> + } else {
> >> + nor->spi->flags &= ~SPI_MASTER_U_PAGE;
> >> + }
> >> + }
> >> + if (nor->addr_width == 3) {
> >> #ifdef CONFIG_SPI_FLASH_BAR
> >> - ret = write_bar(nor, addr);
> >> - if (ret < 0)
> >> - return ret;
> >> + /* Update Extended Address Register */
> >> + ret = write_bar(nor, offset);
> >> + if (ret)
> >> + goto erase_err;
> >> #endif
> >> + }
> >> +
> >> + ret = spi_nor_wait_till_ready(nor);
> >> + if (ret)
> >> + goto erase_err;
> >> +
> >> write_enable(nor);
> >>
> >> - ret = spi_nor_erase_sector(nor, addr);
> >> + ret = spi_nor_erase_sector(nor, offset);
> >> if (ret)
> >> goto erase_err;
> >>
> >> @@ -861,12 +943,33 @@ static int stm_unlock(struct spi_nor *nor,
> >> loff_t ofs, uint64_t len) static int stm_is_locked(struct spi_nor
> >> *nor, loff_t ofs, uint64_t len) {
> >> int status;
> >> + u8 sr;
> >> + int status_up;
> >> + u8 sr_up;
> >> +
> >> + if (nor->isparallel) {
> >> + nor->spi->flags |= SPI_XFER_LOWER;
> >> + sr = read_sr(nor);
> >> + if (sr < 0)
> >> + return 0;
> >> + status = stm_is_locked_sr(nor, ofs, len, sr);
> >> +
> >> + nor->spi->flags |= SPI_XFER_UPPER;
> >> + sr_up = read_sr(nor);
> >> + if (sr_up < 0)
> >> + return sr_up;
> >> + status_up = stm_is_locked_sr(nor, ofs, len, sr_up);
> >> + status = status && status_up;
> >> +
> >> + return status;
> >> + } else {
> >>
> >> status = read_sr(nor);
> >> if (status < 0)
> >> return status;
> >>
> >> return stm_is_locked_sr(nor, ofs, len, status);
> >> + }
> >> }
> >> #endif /* CONFIG_SPI_FLASH_STMICRO */
> >>
> >> @@ -876,6 +979,9 @@ static const struct flash_info
> *spi_nor_read_id(struct spi_nor *nor)
> >> u8 id[SPI_NOR_MAX_ID_LEN];
> >> const struct flash_info *info;
> >>
> >> + if (nor->isparallel)
> >> + nor->spi->flags |= SPI_XFER_LOWER;
> >> +
> >> tmp = nor->read_reg(nor, SPINOR_OP_RDID, id,
> SPI_NOR_MAX_ID_LEN);
> >> if (tmp < 0) {
> >> dev_dbg(nor->dev, "error %d reading JEDEC ID\n", tmp);
> @@ -900,28
> >> +1006,98 @@ static int spi_nor_read(struct mtd_info *mtd, loff_t
> >> from, size_t len, {
> >> struct spi_nor *nor = mtd_to_spi_nor(mtd);
> >> int ret;
> >> + u32 offset = from;
> >> + u32 stack_shift = 0;
> >> + u32 read_len = 0;
> >> + u32 rem_bank_len = 0;
> >> + u8 bank;
> >> + u8 is_ofst_odd = 0;
> >> + u8 cur_bank;
> >> + u8 nxt_bank;
> >> + u32 bank_size;
> >> +
> >> +#define OFFSET_16_MB 0x1000000
> >>
> >> dev_dbg(nor->dev, "from 0x%08x, len %zd\n", (u32)from, len);
> >>
> >> + if ((nor->isparallel) && (offset & 1)) {
> >> + /* We can hit this case when we use file system like ubifs */
> >> + from = (loff_t)(from - 1);
> >> + len = (size_t)(len + 1);
> >> + is_ofst_odd = 1;
> >> + }
> >> +
> >> while (len) {
> >> - loff_t addr = from;
> >> - size_t read_len = len;
> >> + if (nor->addr_width == 3) {
> >> + bank = (u32)from / (OFFSET_16_MB << nor->shift);
> >> + rem_bank_len = ((OFFSET_16_MB << nor->shift) *
> >> + (bank + 1)) - from;
> >> + }
> >> + offset = from;
> >>
> >> -#ifdef CONFIG_SPI_FLASH_BAR
> >> - u32 remain_len;
> >> + if (nor->isparallel) {
> >> + offset /= 2;
> >> + nor->spi->flags = SPI_XFER_STRIPE;
> >> + }
> >>
> >> - ret = write_bar(nor, addr);
> >> - if (ret < 0)
> >> - return log_ret(ret);
> >> - remain_len = (SZ_16M * (nor->bank_curr + 1)) - addr;
> >> + if (nor->isstacked) {
> >> + stack_shift = 1;
> >> + if (offset >= (mtd->size / 2)) {
> >> + offset = offset - (mtd->size / 2);
> >> + nor->spi->flags |= SPI_MASTER_U_PAGE;
> >> + } else {
> >> + nor->spi->flags &= ~SPI_MASTER_U_PAGE;
> >> + }
> >> + }
> >>
> >> - if (len < remain_len)
> >> - read_len = len;
> >> - else
> >> - read_len = remain_len;
> >> + if (nor->addr_width == 4) {
> >> + /*
> >> + * Some flash devices like N25Q512 have multiple dies
> >> + * in it. Read operation in these devices is bounded
> >> + * by its die segment. In a continuous read, across
> >> + * multiple dies, when the last byte of the selected
> >> + * die segment is read, the next byte read is the
> >> + * first byte of the same die segment. This is Die
> >> + * cross over issue. So to handle this issue, split
> >> + * a read transaction, that spans across multiple
> >> + * banks, into one read per bank. Bank size is 16MB
> >> + * for single and dual stacked mode and 32MB for
> dual
> >> + * parallel mode.
> >> + */
> >> + if (nor->spi && nor->spi->multi_die) {
>
> Who sets nor->spi->multi_die?
>
> I have stopped reviewing at this point... This series is incomplete and difficult
> to review.
> Please split this into two patch, one adding support for dual parallel flash and
> one adding dual stacked. Also, include all dependent controller changes and
> other changes as separate patches.
>
> BTW Is there a datasheet for dual flashes or an appnote?
>
> Regards
> Vignesh
>
> >> + bank_size = (OFFSET_16_MB << nor->shift);
> >> + cur_bank = offset / bank_size;
> >> + nxt_bank = (offset + len) / bank_size;
> >> + if (cur_bank != nxt_bank)
> >> + rem_bank_len = (bank_size *
> >> + (cur_bank + 1)) -
> >> + offset;
> >> + else
> >> + rem_bank_len = (mtd->size >>
> >> + stack_shift) -
> >> + (offset << nor->shift);
> >> + } else {
> >> + rem_bank_len = (mtd->size >> stack_shift) -
> >> + (offset << nor->shift);
> >> + }
> >> + }
> >> +
> >> + if (nor->addr_width == 3) {
> >> +#ifdef CONFIG_SPI_FLASH_BAR
> >> + write_bar(nor, offset);
> >> #endif
> >> + }
> >> +
> >> + if (len < rem_bank_len)
> >> + read_len = len;
> >> + else
> >> + read_len = rem_bank_len;
> >>
> >> - ret = nor->read(nor, addr, read_len, buf);
> >> + ret = spi_nor_wait_till_ready(nor);
> >> + if (ret)
> >> + goto read_err;
> >> +
> >> + ret = nor->read(nor, offset, read_len, buf);
> >> if (ret == 0) {
> >> /* We shouldn't see 0-length reads */
> >> ret = -EIO;
> >> @@ -930,7 +1106,12 @@ static int spi_nor_read(struct mtd_info *mtd,
> loff_t from, size_t len,
> >> if (ret < 0)
> >> goto read_err;
> >>
> >> - *retlen += ret;
> >> + if (is_ofst_odd == 1) {
> >> + memcpy(buf, (buf + 1), (len - 1));
> >> + *retlen += (ret - 1);
> >> + } else {
> >> + *retlen += ret;
> >> + }
> >> buf += ret;
> >> from += ret;
> >> len -= ret;
> >> @@ -1223,12 +1404,41 @@ static int spi_nor_write(struct mtd_info *mtd,
> loff_t to, size_t len,
> >> struct spi_nor *nor = mtd_to_spi_nor(mtd);
> >> size_t page_offset, page_remain, i;
> >> ssize_t ret;
> >> + ssize_t written;
> >> + loff_t addr;
> >> + u8 bank = 0;
> >> + u32 offset, stack_shift = 0;
> >> + u32 rem_bank_len = 0;
> >>
> >> dev_dbg(nor->dev, "to 0x%08x, len %zd\n", (u32)to, len);
> >>
> >> + /*
> >> + * Cannot write to odd offset in parallel mode,
> >> + * so write 2 bytes first
> >> + */
> >> + if ((nor->isparallel) && (to & 1)) {
> >> +
> >> + u8 two[2] = {0xff, buf[0]};
> >> + size_t local_retlen;
> >> +
> >> + ret = spi_nor_write(mtd, to & ~1, 2, &local_retlen, two);
> >> + if (ret < 0)
> >> + return ret;
> >> +
> >> + *retlen += 1; /* We've written only one actual byte */
> >> + ++buf;
> >> + --len;
> >> + ++to;
> >> + }
> >> +
> >> for (i = 0; i < len; ) {
> >> - ssize_t written;
> >> - loff_t addr = to + i;
> >> + addr = to + i;
> >> +
> >> + if (nor->addr_width == 3) {
> >> + bank = (u32)to / (OFFSET_16_MB << nor->shift);
> >> + rem_bank_len = ((OFFSET_16_MB << nor->shift) *
> >> + (bank + 1)) - to;
> >> + }
> >>
> >> /*
> >> * If page_size is a power of two, the offset can be quickly
> @@
> >> -1245,17 +1455,51 @@ static int spi_nor_write(struct mtd_info *mtd,
> >> loff_t to, size_t len,
> >>
> >> page_offset = do_div(aux, nor->page_size);
> >> }
> >> - /* the size of data remaining on the first page */
> >> - page_remain = min_t(size_t,
> >> - nor->page_size - page_offset, len - i);
> >>
> >> + offset = (to + i);
> >> + if (nor->isparallel) {
> >> + offset /= 2;
> >> + nor->spi->flags |= SPI_XFER_STRIPE;
> >> + }
> >> +
> >> + if (nor->isstacked) {
> >> + stack_shift = 1;
> >> + if (offset >= (mtd->size / 2)) {
> >> + offset = offset - (mtd->size / 2);
> >> + nor->spi->flags |= SPI_MASTER_U_PAGE;
> >> + } else {
> >> + nor->spi->flags &= ~SPI_MASTER_U_PAGE;
> >> + }
> >> + }
> >> +
> >> + /* Die cross over issue is not handled */
> >> + if (nor->addr_width == 4)
> >> + rem_bank_len = (mtd->size >> stack_shift) - offset;
> >> + if (nor->addr_width == 3) {
> >> #ifdef CONFIG_SPI_FLASH_BAR
> >> - ret = write_bar(nor, addr);
> >> - if (ret < 0)
> >> - return ret;
> >> + write_bar(nor, offset);
> >> #endif
> >> + }
> >> + if (nor->isstacked) {
> >> + if (len <= rem_bank_len) {
> >> + page_remain = min_t(size_t, nor->page_size
> -
> >> + page_offset, len - i);
> >> + } else {
> >> + /* size of data remaining on the first page */
> >> + page_remain = rem_bank_len;
> >> + }
> >> + } else {
> >> + page_remain = min_t(size_t, nor->page_size -
> >> + page_offset, len - i);
> >> + }
> >> +
> >> + ret = spi_nor_wait_till_ready(nor);
> >> + if (ret)
> >> + goto write_err;
> >> +
> >> +
> >> write_enable(nor);
> >> - ret = nor->write(nor, addr, page_remain, buf + i);
> >> + ret = nor->write(nor, offset, page_remain, buf + i);
> >> if (ret < 0)
> >> goto write_err;
> >> written = ret;
> >> @@ -1265,6 +1509,13 @@ static int spi_nor_write(struct mtd_info *mtd,
> loff_t to, size_t len,
> >> goto write_err;
> >> *retlen += written;
> >> i += written;
> >> + if (written != page_remain) {
> >> + dev_err(nor->dev,
> >> + "While writing %zu bytes written %zd
> bytes\n",
> >> + page_remain, written);
> >> + ret = -EIO;
> >> + goto write_err;
> >> + }
> >> }
> >>
> >> write_err:
> >> @@ -2094,6 +2345,12 @@ static int spi_nor_init_params(struct spi_nor
> *nor,
> >> params->size = info->sector_size * info->n_sectors;
> >> params->page_size = info->page_size;
> >>
> >> + if (nor->isparallel)
> >> + params->page_size <<= nor->shift;
> >> +
> >> + if (nor->isparallel || nor->isstacked)
> >> + params->size <<= nor->shift;
> >> +
> >> /* (Fast) Read settings. */
> >> params->hwcaps.mask |= SNOR_HWCAPS_READ;
> >> spi_nor_set_read_settings(¶ms->reads[SNOR_CMD_READ],
> >> @@ -2280,24 +2537,27 @@ static int spi_nor_select_erase(struct spi_nor
> *nor,
> >> const struct flash_info *info)
> >> {
> >> struct mtd_info *mtd = &nor->mtd;
> >> + bool shift = 0;
> >>
> >> /* Do nothing if already configured from SFDP. */
> >> if (mtd->erasesize)
> >> return 0;
> >>
> >> + if (nor->isparallel)
> >> + shift = 1;
> >> #ifdef CONFIG_SPI_FLASH_USE_4K_SECTORS
> >> /* prefer "small sector" erase if possible */
> >> if (info->flags & SECT_4K) {
> >> nor->erase_opcode = SPINOR_OP_BE_4K;
> >> - mtd->erasesize = 4096;
> >> + mtd->erasesize = 4096 << shift;
> >> } else if (info->flags & SECT_4K_PMC) {
> >> nor->erase_opcode = SPINOR_OP_BE_4K_PMC;
> >> - mtd->erasesize = 4096;
> >> + mtd->erasesize = 4096 << shift;
> >> } else
> >> #endif
> >> {
> >> nor->erase_opcode = SPINOR_OP_SE;
> >> - mtd->erasesize = info->sector_size;
> >> + mtd->erasesize = info->sector_size << shift;
> >> }
> >> return 0;
> >> }
> >> @@ -2442,9 +2702,14 @@ int spi_nor_scan(struct spi_nor *nor)
> >> hwcaps.mask |= SNOR_HWCAPS_READ_1_2_2;
> >> }
> >>
> >> - info = spi_nor_read_id(nor);
> >> + nor->isparallel = (spi->option == SF_DUAL_PARALLEL_FLASH) ? 1 : 0;
> >> + nor->isstacked = (spi->option == SF_DUAL_STACKED_FLASH) ? 1 : 0;
> >> + nor->shift = (nor->isparallel || nor->isstacked) ? 1 : 0;
> >> +
> >> + info = (struct flash_info *)spi_nor_read_id(nor);
> >> if (IS_ERR_OR_NULL(info))
> >> return -ENOENT;
> >> +
> >> /* Parse the Serial Flash Discoverable Parameters table. */
> >> ret = spi_nor_init_params(nor, info, ¶ms);
> >> if (ret)
> >> diff --git a/drivers/spi/spi-mem.c b/drivers/spi/spi-mem.c index
> >> 7788ab9953..2471132f41 100644
> >> --- a/drivers/spi/spi-mem.c
> >> +++ b/drivers/spi/spi-mem.c
> >> @@ -202,7 +202,7 @@ int spi_mem_exec_op(struct spi_slave *slave,
> const struct spi_mem_op *op)
> >> const u8 *tx_buf = NULL;
> >> u8 *rx_buf = NULL;
> >> int op_len;
> >> - u32 flag;
> >> + u32 flag = 0;
> >> int ret;
> >> int i;
> >>
> >> @@ -362,24 +362,35 @@ int spi_mem_exec_op(struct spi_slave *slave,
> const struct spi_mem_op *op)
> >> if (op->dummy.nbytes)
> >> memset(op_buf + pos, 0xff, op->dummy.nbytes);
> >>
> >> + if (slave->flags & SPI_XFER_U_PAGE)
> >> + flag |= SPI_XFER_U_PAGE;
> >> +
> >> + if (slave->flags & SPI_XFER_LOWER)
> >> + flag |= SPI_XFER_LOWER;
> >> + if (slave->flags & SPI_XFER_UPPER)
> >> + flag |= SPI_XFER_UPPER;
> >> + if (slave->flags & SPI_XFER_STRIPE)
> >> + flag |= SPI_XFER_STRIPE;
> >> +
> >> /* 1st transfer: opcode + address + dummy cycles */
> >> - flag = SPI_XFER_BEGIN;
> >> /* Make sure to set END bit if no tx or rx data messages follow */
> >> if (!tx_buf && !rx_buf)
> >> flag |= SPI_XFER_END;
> >>
> >> - ret = spi_xfer(slave, op_len * 8, op_buf, NULL, flag);
> >> + ret = spi_xfer(slave, op_len * 8, op_buf, NULL, flag |
> >> +SPI_XFER_BEGIN);
> >> if (ret)
> >> return ret;
> >>
> >> /* 2nd transfer: rx or tx data path */
> >> if (tx_buf || rx_buf) {
> >> ret = spi_xfer(slave, op->data.nbytes * 8, tx_buf,
> >> - rx_buf, SPI_XFER_END);
> >> + rx_buf, flag | SPI_XFER_END);
> >> if (ret)
> >> return ret;
> >> }
> >>
> >> + slave->flags &= ~SPI_XFER_MASK;
> >> +
> >> spi_release_bus(slave);
> >>
> >> for (i = 0; i < pos; i++)
> >> diff --git a/include/linux/mtd/spi-nor.h
> >> b/include/linux/mtd/spi-nor.h index f9964a7664..c7475e7a12 100644
> >> --- a/include/linux/mtd/spi-nor.h
> >> +++ b/include/linux/mtd/spi-nor.h
> >> @@ -309,11 +309,15 @@ struct spi_nor {
> >> u8 bank_read_cmd;
> >> u8 bank_write_cmd;
> >> u8 bank_curr;
> >> + u8 upage_prev;
> >> #endif
> >> enum spi_nor_protocol read_proto;
> >> enum spi_nor_protocol write_proto;
> >> enum spi_nor_protocol reg_proto;
> >> bool sst_write_second;
> >> + bool shift;
> >> + bool isparallel;
> >> + bool isstacked;
> >> u32 flags;
> >> u8 cmd_buf[SPI_NOR_MAX_CMD_SIZE];
> >>
> >> diff --git a/include/spi.h b/include/spi.h index
> >> 6fbb4336ce..69f69e590c 100644
> >> --- a/include/spi.h
> >> +++ b/include/spi.h
> >> @@ -31,6 +31,12 @@
> >> #define SPI_RX_DUAL BIT(12) /* receive with 2
> wires */
> >> #define SPI_RX_QUAD BIT(13) /* receive with 4
> wires */
> >>
> >> +/* SPI transfer flags */
> >> +#define SPI_XFER_STRIPE (1 << 6)
> >> +#define SPI_XFER_MASK (3 << 8)
> >> +#define SPI_XFER_LOWER (1 << 8)
> >> +#define SPI_XFER_UPPER (2 << 8)
> >> +
> >> /* Header byte that marks the start of the message */
> >> #define SPI_PREAMBLE_END_BYTE 0xec
> >>
> >> @@ -108,13 +114,16 @@ struct spi_slave {
> >> unsigned int max_read_size;
> >> unsigned int max_write_size;
> >> void *memory_map;
> >> + u8 option;
> >>
> >> u8 flags;
> >> + bool multi_die; /* flash with multiple dies*/
> >> #define SPI_XFER_BEGIN BIT(0) /* Assert CS before transfer
> */
> >> #define SPI_XFER_END BIT(1) /* Deassert CS after transfer
> */
> >> #define SPI_XFER_ONCE (SPI_XFER_BEGIN | SPI_XFER_END)
> >> #define SPI_XFER_MMAP BIT(2) /* Memory Mapped start */
> >> #define SPI_XFER_MMAP_END BIT(3) /* Memory Mapped End */
> >> +#define SPI_XFER_U_PAGE BIT(4)
> >> };
> >>
> >> /**
> >>
> >
> > Jagan: any comment?
> >
> > M
> >
>
> --
> Regards
> Vignesh
More information about the U-Boot
mailing list