| Message ID | 20200501100729.1237040-1-boris.brezillon@collabora.com |
|---|---|
| Headers | show |
| Series | mtd: rawnand: davinci: Convert to exec_op() | expand |
pt., 1 maj 2020 o 12:07 Boris Brezillon <boris.brezillon@collabora.com> napisał(a): > > Hello, > > A bit of context to explain the motivation behind those conversions > I've been sending for the last couple of weeks. The raw NAND subsystem > carries a lot of history which makes any rework not only painful, but > also subject to regressions which we only detect when someone dares to > update its kernel on one of those ancient HW. While carrying drivers > for old HW is not a problem per se, carrying ancient and unmaintained > drivers that are not converted to new APIs is a maintenance burden, > hence this massive conversion attempt I'm conducting here. > > So here is a series converting the Davinci NAND controller driver to > exec_op(), plus a bunch of minor improvements done along the way. > > Regards, > > Boris > > Boris Brezillon (5): > mtd: rawnand: davinci: Inherit from nand_controller > mtd: rawnand: davinci: Stop using nand_chip.legacy.IO_ADDR_{R,W} > mtd: rawnand: davinci: Implement exec_op() > mtd: rawnand: davinci: Get rid of the legacy interface implementation > mtd: rawnand: davinci: Change the {read,write}_buf prototypes > > drivers/mtd/nand/raw/davinci_nand.c | 161 +++++++++++++++------------- > 1 file changed, 85 insertions(+), 76 deletions(-) > > -- > 2.25.3 > Hi Boris, Thanks for doing this. Unfortunately this breaks NAND on da850-lcdk with the following error message: nand: No NAND device found I'm super busy this week and so I don't have the time to investigate further, I can get back to it next week hopefully. Bart
On Mon, 4 May 2020 18:34:55 +0200 Bartosz Golaszewski <bgolaszewski@baylibre.com> wrote: > pt., 1 maj 2020 o 12:07 Boris Brezillon > <boris.brezillon@collabora.com> napisał(a): > > > > Hello, > > > > A bit of context to explain the motivation behind those conversions > > I've been sending for the last couple of weeks. The raw NAND subsystem > > carries a lot of history which makes any rework not only painful, but > > also subject to regressions which we only detect when someone dares to > > update its kernel on one of those ancient HW. While carrying drivers > > for old HW is not a problem per se, carrying ancient and unmaintained > > drivers that are not converted to new APIs is a maintenance burden, > > hence this massive conversion attempt I'm conducting here. > > > > So here is a series converting the Davinci NAND controller driver to > > exec_op(), plus a bunch of minor improvements done along the way. > > > > Regards, > > > > Boris > > > > Boris Brezillon (5): > > mtd: rawnand: davinci: Inherit from nand_controller > > mtd: rawnand: davinci: Stop using nand_chip.legacy.IO_ADDR_{R,W} > > mtd: rawnand: davinci: Implement exec_op() > > mtd: rawnand: davinci: Get rid of the legacy interface implementation > > mtd: rawnand: davinci: Change the {read,write}_buf prototypes > > > > drivers/mtd/nand/raw/davinci_nand.c | 161 +++++++++++++++------------- > > 1 file changed, 85 insertions(+), 76 deletions(-) > > > > -- > > 2.25.3 > > > > Hi Boris, > > Thanks for doing this. Unfortunately this breaks NAND on da850-lcdk > with the following error message: > > nand: No NAND device found > > I'm super busy this week and so I don't have the time to investigate > further, I can get back to it next week hopefully. No worries. And let me know if you need any help to debug that. Thanks! Boris
On Mon, 4 May 2020 18:34:55 +0200 Bartosz Golaszewski <bgolaszewski@baylibre.com> wrote: > pt., 1 maj 2020 o 12:07 Boris Brezillon > <boris.brezillon@collabora.com> napisał(a): > > > > Hello, > > > > A bit of context to explain the motivation behind those conversions > > I've been sending for the last couple of weeks. The raw NAND subsystem > > carries a lot of history which makes any rework not only painful, but > > also subject to regressions which we only detect when someone dares to > > update its kernel on one of those ancient HW. While carrying drivers > > for old HW is not a problem per se, carrying ancient and unmaintained > > drivers that are not converted to new APIs is a maintenance burden, > > hence this massive conversion attempt I'm conducting here. > > > > So here is a series converting the Davinci NAND controller driver to > > exec_op(), plus a bunch of minor improvements done along the way. > > > > Regards, > > > > Boris > > > > Boris Brezillon (5): > > mtd: rawnand: davinci: Inherit from nand_controller > > mtd: rawnand: davinci: Stop using nand_chip.legacy.IO_ADDR_{R,W} > > mtd: rawnand: davinci: Implement exec_op() > > mtd: rawnand: davinci: Get rid of the legacy interface implementation > > mtd: rawnand: davinci: Change the {read,write}_buf prototypes > > > > drivers/mtd/nand/raw/davinci_nand.c | 161 +++++++++++++++------------- > > 1 file changed, 85 insertions(+), 76 deletions(-) > > > > -- > > 2.25.3 > > > > Hi Boris, > > Thanks for doing this. Unfortunately this breaks NAND on da850-lcdk > with the following error message: > > nand: No NAND device found I had a second look and the below diff should fix the detection (you can also find those changes in my exec-op-conversion branch [1]). [1]https://github.com/bbrezillon/linux/tree/nand/exec-op-conversion --->8--- diff --git a/drivers/mtd/nand/raw/davinci_nand.c b/drivers/mtd/nand/raw/davinci_nand.c index 296047884c6a..0eeb30c7fc4e 100644 --- a/drivers/mtd/nand/raw/davinci_nand.c +++ b/drivers/mtd/nand/raw/davinci_nand.c @@ -678,6 +678,33 @@ static int davinci_nand_attach_chip(struct nand_chip *chip) return ret; } +static void nand_davinci_data_in(struct davinci_nand_info *info, void *buf, + unsigned int len, bool force_8bit) +{ + u32 alignment = ((uintptr_t)buf | len) & 3; + + if (force_8bit || (alignment & 1)) + ioread8_rep(info->current_cs, buf, len); + else if (alignment & 3) + ioread16_rep(info->current_cs, buf, len >> 1); + else + ioread32_rep(info->current_cs, buf, len >> 2); +} + +static void nand_davinci_data_out(struct davinci_nand_info *info, + const void *buf, unsigned int len, + bool force_8bit) +{ + u32 alignment = ((uintptr_t)buf | len) & 3; + + if (force_8bit || (alignment & 1)) + iowrite8_rep(info->current_cs, buf, len); + else if (alignment & 3) + iowrite16_rep(info->current_cs, buf, len >> 1); + else + iowrite32_rep(info->current_cs, buf, len >> 2); +} + static int davinci_nand_exec_instr(struct davinci_nand_info *info, const struct nand_op_instr *instr) { @@ -699,13 +726,15 @@ static int davinci_nand_exec_instr(struct davinci_nand_info *info, break; case NAND_OP_DATA_IN_INSTR: - nand_davinci_read_buf(&info->chip, instr->ctx.data.buf.in, - instr->ctx.data.len); + nand_davinci_data_in(info, instr->ctx.data.buf.in, + instr->ctx.data.len, + instr->ctx.data.force_8bit); break; case NAND_OP_DATA_OUT_INSTR: - nand_davinci_write_buf(&info->chip, instr->ctx.data.buf.out, - instr->ctx.data.len); + nand_davinci_data_out(info, instr->ctx.data.buf.out, + instr->ctx.data.len, + instr->ctx.data.force_8bit); break; case NAND_OP_WAITRDY_INSTR: @@ -731,20 +760,21 @@ static int davinci_nand_exec_op(struct nand_chip *chip, { struct davinci_nand_info *info = to_davinci_nand(nand_to_mtd(chip)); unsigned int i; - int ret; if (check_only) - return true; + return 0; info->current_cs = info->vaddr + (op->cs * info->mask_chipsel); for (i = 0; i < op->ninstrs; i++) { + int ret; + ret = davinci_nand_exec_instr(info, &op->instrs[i]); if (ret) - break; + return ret; } - return ret; + return 0; } static const struct nand_controller_ops davinci_nand_controller_ops = {
sob., 9 maj 2020 o 16:13 Boris Brezillon <boris.brezillon@collabora.com> napisał(a): > > [snip] > > > > > Hi Boris, > > > > Thanks for doing this. Unfortunately this breaks NAND on da850-lcdk > > with the following error message: > > > > nand: No NAND device found > > I had a second look and the below diff should fix the detection (you > can also find those changes in my exec-op-conversion branch [1]). > > [1]https://github.com/bbrezillon/linux/tree/nand/exec-op-conversion > Hi Boris, Yes, this branch works now. Do you want to send out a v2? I'm happy to add a Tested-by to it. Bart
Hello, A bit of context to explain the motivation behind those conversions I've been sending for the last couple of weeks. The raw NAND subsystem carries a lot of history which makes any rework not only painful, but also subject to regressions which we only detect when someone dares to update its kernel on one of those ancient HW. While carrying drivers for old HW is not a problem per se, carrying ancient and unmaintained drivers that are not converted to new APIs is a maintenance burden, hence this massive conversion attempt I'm conducting here. So here is a series converting the Davinci NAND controller driver to exec_op(), plus a bunch of minor improvements done along the way. Regards, Boris Boris Brezillon (5): mtd: rawnand: davinci: Inherit from nand_controller mtd: rawnand: davinci: Stop using nand_chip.legacy.IO_ADDR_{R,W} mtd: rawnand: davinci: Implement exec_op() mtd: rawnand: davinci: Get rid of the legacy interface implementation mtd: rawnand: davinci: Change the {read,write}_buf prototypes drivers/mtd/nand/raw/davinci_nand.c | 161 +++++++++++++++------------- 1 file changed, 85 insertions(+), 76 deletions(-)