From: Roger Quadros <rogerq@kernel.org>
To: Michael Nazzareno Trimarchi <michael@amarulasolutions.com>
Cc: dario.binacchi@amarulasolutions.com, trini@konsulko.com,
u-boot@lists.denx.de
Subject: Re: [u-boot][PATCH 04/14] mtd: rawnand: omap_gpmc: Optimize NAND reads
Date: Sat, 15 Oct 2022 16:29:15 +0300 [thread overview]
Message-ID: <32f321a3-35ad-fed4-12d8-a98877286cd9@kernel.org> (raw)
In-Reply-To: <CAOf5uwk4isTG4pCts_8_Ry3PzWGtoLTo2p7W3VHCPFeJiUy-7g@mail.gmail.com>
Hi Michael,
On 15/10/2022 10:24, Michael Nazzareno Trimarchi wrote:
> Hi
>
> On Tue, Oct 11, 2022 at 1:50 PM Roger Quadros <rogerq@kernel.org> wrote:
>>
>> Rename omap_nand_read() to omap_nand_read_buf() to reflect
>> actual behaviour.
>>
>> Use FIFO read address instead of raw read address for reads.
>>
>> The GPMC automatically converts 32-bit/16-bit reads to NAND
>> device specific reads (8/16 bit). Use the largest possible
>> read granularity size for more efficient reads.
>>
>> Signed-off-by: Roger Quadros <rogerq@kernel.org>
>> ---
>> drivers/mtd/nand/raw/omap_gpmc.c | 49 ++++++++++++++++++--------------
>> 1 file changed, 28 insertions(+), 21 deletions(-)
>>
>> diff --git a/drivers/mtd/nand/raw/omap_gpmc.c b/drivers/mtd/nand/raw/omap_gpmc.c
>> index d62c3e6fce..b36fe762b3 100644
>> --- a/drivers/mtd/nand/raw/omap_gpmc.c
>> +++ b/drivers/mtd/nand/raw/omap_gpmc.c
>> @@ -55,6 +55,7 @@ struct omap_nand_info {
>> enum omap_ecc ecc_scheme;
>> uint8_t cs;
>> uint8_t ws; /* wait status pin (0,1) */
>> + void __iomem *fifo;
>> };
>>
>> /* We are wasting a bit of memory but al least we are safe */
>> @@ -350,6 +351,20 @@ static int omap_calculate_ecc(struct mtd_info *mtd, const uint8_t *dat,
>> return 0;
>> }
>>
>> +static inline void omap_nand_read_buf(struct mtd_info *mtd, uint8_t *buf, int len)
>> +{
>> + struct nand_chip *chip = mtd_to_nand(mtd);
>> + struct omap_nand_info *info = nand_get_controller_data(chip);
>> + u32 alignment = ((uintptr_t)buf | len) & 3;
>> +
>> + if (alignment & 1)
>> + readsb(info->fifo, buf, len);
>> + else if (alignment & 3)
>> + readsw(info->fifo, buf, len >> 1);
>> + else
>> + readsl(info->fifo, buf, len >> 2);
>> +}
>> +
>> #ifdef CONFIG_NAND_OMAP_GPMC_PREFETCH
>>
>> #define PREFETCH_CONFIG1_CS_SHIFT 24
>> @@ -415,7 +430,7 @@ static int __read_prefetch_aligned(struct nand_chip *chip, uint32_t *buf, int le
>> cnt = PREFETCH_STATUS_FIFO_CNT(cnt);
>>
>> for (i = 0; i < cnt / 4; i++) {
>> - *buf++ = readl(CONFIG_SYS_NAND_BASE);
>> + *buf++ = readl(info->fifo);
>> len -= 4;
>> }
>> } while (len);
>> @@ -425,16 +440,6 @@ static int __read_prefetch_aligned(struct nand_chip *chip, uint32_t *buf, int le
>> return 0;
>> }
>>
>> -static inline void omap_nand_read(struct mtd_info *mtd, uint8_t *buf, int len)
>> -{
>> - struct nand_chip *chip = mtd_to_nand(mtd);
>> -
>> - if (chip->options & NAND_BUSWIDTH_16)
>> - nand_read_buf16(mtd, buf, len);
>> - else
>> - nand_read_buf(mtd, buf, len);
>> -}
>> -
>> static void omap_nand_read_prefetch(struct mtd_info *mtd, uint8_t *buf, int len)
>> {
>> int ret;
>> @@ -447,7 +452,7 @@ static void omap_nand_read_prefetch(struct mtd_info *mtd, uint8_t *buf, int len)
>> */
>> head = ((uintptr_t)buf) % 4;
>> if (head) {
>> - omap_nand_read(mtd, buf, head);
>> + omap_nand_read_buf(mtd, buf, head);
>> buf += head;
>> len -= head;
>> }
>> @@ -461,10 +466,10 @@ static void omap_nand_read_prefetch(struct mtd_info *mtd, uint8_t *buf, int len)
>> ret = __read_prefetch_aligned(chip, (uint32_t *)buf, len - tail);
>> if (ret < 0) {
>> /* fallback in case the prefetch engine is busy */
>> - omap_nand_read(mtd, buf, len);
>> + omap_nand_read_buf(mtd, buf, len);
>> } else if (tail) {
>> buf += len - tail;
>> - omap_nand_read(mtd, buf, tail);
>> + omap_nand_read_buf(mtd, buf, tail);
>> }
>> }
>> #endif /* CONFIG_NAND_OMAP_GPMC_PREFETCH */
>> @@ -1001,6 +1006,8 @@ int board_nand_init(struct nand_chip *nand)
>> int32_t gpmc_config = 0;
>> int cs = cs_next++;
>> int err = 0;
>> + struct omap_nand_info *info;
>> +
>> /*
>> * xloader/Uboot's gpmc configuration would have configured GPMC for
>> * nand type of memory. The following logic scans and latches on to the
>> @@ -1029,9 +1036,12 @@ int board_nand_init(struct nand_chip *nand)
>>
>> nand->IO_ADDR_R = (void __iomem *)&gpmc_cfg->cs[cs].nand_dat;
>> nand->IO_ADDR_W = (void __iomem *)&gpmc_cfg->cs[cs].nand_cmd;
>> - omap_nand_info[cs].control = NULL;
>> - omap_nand_info[cs].cs = cs;
>> - omap_nand_info[cs].ws = wscfg[cs];
>> +
>> + info = &omap_nand_info[cs];
>> + info->control = NULL;
>> + info->cs = cs;
>> + info->ws = wscfg[cs];
>> + info->fifo = (void __iomem *)CONFIG_SYS_NAND_BASE;
>> nand_set_controller_data(nand, &omap_nand_info[cs]);
>> nand->cmd_ctrl = omap_nand_hwcontrol;
>> nand->options |= NAND_NO_PADDING | NAND_CACHEPRG;
>> @@ -1062,10 +1072,7 @@ int board_nand_init(struct nand_chip *nand)
>> #ifdef CONFIG_NAND_OMAP_GPMC_PREFETCH
>> nand->read_buf = omap_nand_read_prefetch;
>> #else
>> - if (nand->options & NAND_BUSWIDTH_16)
>> - nand->read_buf = nand_read_buf16;
>> - else
>> - nand->read_buf = nand_read_buf;
>> + nand->read_buf = omap_nand_read_buf;
>> #endif
>>
>> nand->dev_ready = omap_dev_ready;
>> --
>> 2.17.1
>>
>
> Is possible for you to split in use of fifo and optimize read/alignment
Sure, I'll remember to do that when I re-spin this series.
>
> Otherwise
>
> Reviewed-by: Michael Trimarchi <michael@amarulasolutions.com>
>
>
Thanks!
--
cheers,
-roger
next prev parent reply other threads:[~2022-10-15 13:29 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-11 11:49 [u-boot][PATCH 00/14] rawnand: omap_gpmc: driver model support Roger Quadros
2022-10-11 11:49 ` [u-boot][PATCH 01/14] mtd: rawnand: omap_gpmc: Deprecate asm/arch/mem.h Roger Quadros
2022-10-12 10:01 ` Michael Nazzareno Trimarchi
2022-10-11 11:50 ` [u-boot][PATCH 02/14] mtd: rawnand: omap_gpmc: Enable build for K2/K3 platforms Roger Quadros
2022-10-12 11:49 ` Michael Nazzareno Trimarchi
2022-10-11 11:50 ` [u-boot][PATCH 03/14] mtd: rawnand: omap_gpmc: Fix build warning on 64-bit platforms Roger Quadros
2022-10-15 5:53 ` Michael Nazzareno Trimarchi
2022-10-11 11:50 ` [u-boot][PATCH 04/14] mtd: rawnand: omap_gpmc: Optimize NAND reads Roger Quadros
2022-10-15 7:24 ` Michael Nazzareno Trimarchi
2022-10-15 13:29 ` Roger Quadros [this message]
2022-10-17 6:39 ` Michael Nazzareno Trimarchi
2022-10-17 8:14 ` Roger Quadros
2022-10-11 11:50 ` [u-boot][PATCH 05/14] mtd: rawnand: omap_gpmc: Fix BCH6/16 HW based correction Roger Quadros
2022-11-28 14:03 ` Michael Nazzareno Trimarchi
2022-11-29 15:25 ` Dario Binacchi
2022-11-29 16:17 ` Tom Rini
2022-11-30 8:11 ` Dario Binacchi
2022-12-01 11:42 ` Roger Quadros
2022-10-11 11:50 ` [u-boot][PATCH 06/14] mtd: rawnand: nand_base: Allow base driver to be used in SPL without nand_bbt Roger Quadros
2022-11-28 14:27 ` Michael Nazzareno Trimarchi
2022-11-29 13:04 ` Roger Quadros
2022-10-11 11:50 ` [u-boot][PATCH 07/14] mtd: rawnand: nand_spl_loaders: Fix cast type build warning Roger Quadros
2022-11-06 19:50 ` Michael Nazzareno Trimarchi
2022-10-11 11:50 ` [u-boot][PATCH 08/14] mtd: rawnand: omap_gpmc: Reduce .bss usage Roger Quadros
2022-11-28 14:11 ` Michael Nazzareno Trimarchi
2022-10-11 11:50 ` [u-boot][PATCH 09/14] dt-bindings: mtd: Add ti, gpmc-nand DT binding documentation Roger Quadros
2022-10-11 11:50 ` [u-boot][PATCH 10/14] mtd: rawnand: omap_gpmc: support u-boot driver model Roger Quadros
2022-10-11 15:01 ` Adam Ford
2022-10-12 6:22 ` Roger Quadros
2022-10-12 11:42 ` Adam Ford
2022-10-12 11:57 ` Ladislav Michl
2022-10-12 12:02 ` Adam Ford
2022-10-13 11:17 ` Adam Ford
2022-10-13 19:42 ` Roger Quadros
2022-10-11 11:50 ` [u-boot][PATCH 11/14] mtd: rawnand: omap_gpmc: Add SPL NAND support Roger Quadros
2022-10-11 11:50 ` [u-boot][PATCH 12/14] mtd: rawnand: omap_gpmc: Enable SYS_NAND_PAGE_COUNT for OMAP_GPMC Roger Quadros
2022-10-11 11:50 ` [u-boot][PATCH 13/14] dt-bindings: mtd: Add ti, elm DT binding documentation Roger Quadros
2022-10-11 11:50 ` [u-boot][PATCH 14/14] mtd: rawnand: omap_elm: u-boot driver model support Roger Quadros
2022-11-04 13:27 ` [u-boot][PATCH 00/14] rawnand: omap_gpmc: " Roger Quadros
2022-11-08 9:26 ` Michael Nazzareno Trimarchi
2022-11-25 12:38 ` Roger Quadros
2022-12-11 13:56 ` Dario Binacchi
2022-12-12 9:12 ` Roger Quadros
2022-12-12 9:27 ` Dario Binacchi
2022-12-12 9:39 ` Michael Nazzareno Trimarchi
2022-12-17 13:00 ` Roger Quadros
2022-12-17 13:38 ` Michael Nazzareno Trimarchi
2022-12-17 13:43 ` Roger Quadros
2022-12-17 13:46 ` Michael Nazzareno Trimarchi
2022-12-17 13:51 ` Michael Nazzareno Trimarchi
2022-12-17 13:59 ` Roger Quadros
2022-12-17 14:09 ` Michael Nazzareno Trimarchi
2022-12-17 14:48 ` Michael Nazzareno Trimarchi
2022-12-12 13:58 ` Tom Rini
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=32f321a3-35ad-fed4-12d8-a98877286cd9@kernel.org \
--to=rogerq@kernel.org \
--cc=dario.binacchi@amarulasolutions.com \
--cc=michael@amarulasolutions.com \
--cc=trini@konsulko.com \
--cc=u-boot@lists.denx.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox