On Fri, 26 Oct 2018 01:32:09 +0900 Tokunori Ikegami ikegami@allied-telesis.co.jp wrote:
In OpenWrt Project the flash write error caused on some products.
It's okay to mention that the issue was discovered by the OpenWRT team, but I'd rephrase it differently.
"As reported by the OpenWRT team, write requests sometimes fail on some platforms".
Also the issue can be fixed by using chip_good() instead of chip_ready(). The chip_ready() just checks the value from flash memory twice. And the chip_good() checks the value with the expected value. Probably the issue can be fixed as checked correctly by the chip_good(). So change to use chip_good() instead of chip_ready().
Well, that's not really explaining why you think chip_good() should be used instead of chip_ready(). So I went on and looked at the chip_good(), chip_ready() and do_write_oneword() implementation, and also looked at users of do_write_oneword(). It seems this function is used to write data to the flash, and apparently the "one bit should toggle to reflect a busy state" does not apply when writing things to the memory array (it's probably working for other CFI commands, but I guess it takes more time to actually change the level of a NOR cell, hence the result of 2 identical reads does not mean that the write is done).
Also, it seems that cmdset_0001 is not implementing chip_ready() the same way, and I wonder if cmdset_0002 implementation is correct to start with. Or maybe I don't get what chip_ready() is for.
Anyway, this is the sort of clarification I'd like to have.
Signed-off-by: Tokunori Ikegami ikegami@allied-telesis.co.jp Signed-off-by: Hauke Mehrtens hauke@hauke-m.de Signed-off-by: Koen Vandeputte koen.vandeputte@ncentric.com Signed-off-by: Fabio Bettoni fbettoni@gmail.com
Has the patch really gone through all those people? SoB is used when you apply a patch in your tree or when you're the original author.
Co-Developed-by: Hauke Mehrtens hauke@hauke-m.de Co-Developed-by: Koen Vandeputte koen.vandeputte@ncentric.com Co-Developed-by: Fabio Bettoni fbettoni@gmail.com
Not sure we want to add new undocumented tags, but you can mention that all those people helped you find/debug the issue. They can also add their Reviewed-by/Tested-by if they like.
Reported-by: Fabio Bettoni fbettoni@gmail.com Cc: Chris Packham chris.packham@alliedtelesis.co.nz Cc: Joakim Tjernlund Joakim.Tjernlund@infinera.com Cc: Boris Brezillon boris.brezillon@free-electrons.com Cc: linux-mtd@lists.infradead.org Cc: stable@vger.kernel.org
Changes since v2:
- Just update the commit message for the comment.
Changes since v1:
- Just update the commit message.
Background: This is required for OpenWrt Project to result the flash write issue as below patche. https://git.openwrt.org/?p=openwrt/openwrt.git;a=commitdiff;h=ddc11c3932c7b7b7df7d5fbd48f207e77619eaa7
Also the original patch in OpenWRT is below. https://github.com/openwrt/openwrt/blob/v18.06.0/target/linux/ar71xx/patches-4.9/403-mtd_fix_cfi_cmdset_0002_status_check.patch
The reason to use chip_good() is that just actually fix the issue. And also in the past I had fixed the erase function also as same way by the patch below. https://patchwork.ozlabs.org/patch/922656/ Note: The reason for the patch for erase is same.
In my understanding the chip_ready() is just checked the value twice from flash. So I think that sometimes incorrect value is read twice and it is depended on the flash device behavior but not sure..
So change to use chip_good() instead of chip_ready().
drivers/mtd/chips/cfi_cmdset_0002.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-)
diff --git a/drivers/mtd/chips/cfi_cmdset_0002.c b/drivers/mtd/chips/cfi_cmdset_0002.c index 72428b6bfc47..251c9e1675bd 100644 --- a/drivers/mtd/chips/cfi_cmdset_0002.c +++ b/drivers/mtd/chips/cfi_cmdset_0002.c @@ -1627,31 +1627,37 @@ static int __xipram do_write_oneword(struct map_info *map, struct flchip *chip, continue; }
if (time_after(jiffies, timeo) && !chip_ready(map, adr)){
if (chip_good(map, adr, datum))break;if (time_after(jiffies, timeo)){ xip_enable(map, chip, adr); printk(KERN_WARNING "MTD %s(): software timeout\n", __func__); xip_disable(map, chip, adr);ret = -EIO; break;
if (chip_ready(map, adr))break;- /* Latency issues. Drop the lock, wait a while and retry */ UDELAY(map, chip, adr, 1); }
- /* Did we succeed? */
- if (!chip_good(map, adr, datum)) {
- if (ret) { /* reset on all failures. */ map_write(map, CMD(0xF0), chip->start); /* FIXME - should have reset delay before continuing */
if (++retry_cnt <= MAX_RETRIES)
if (++retry_cnt <= MAX_RETRIES) {ret = 0; goto retry;}ret = -EIO; }
- xip_enable(map, chip, adr);
Not a big deal, but I'd prefer to not have coding style changes mixed with functional changes (in other words, you can drop the addition of blanks lines around xip_enable()).
op_done: if (mode == FL_OTP_WRITE) otp_exit(map, chip, adr, map_bankwidth(map));