Fabio,
On 3 November 2015 at 20:29, Fabio Estevam festevam@gmail.com wrote:
Jagan,
On Tue, Nov 3, 2015 at 12:51 PM, Jagan Teki jteki@openedev.com wrote:
if (flash->is_locked(flash, offset, len) > 0) {printf("offset 0x%x is protected and cannot be written\n", offset);return -EINVAL;}This flash lock check related to non-dm lock ops so it's look not good to me as we assigned lock ops for both dm and non-dm cases.
This is getting totally ridiculous: you have been seeing this patch series so many times and now you say this is not OK and you don't give me any clue as to what you would like me to change.
Nothing is worse than before or now - as spi-flash code is configured both the dm and non-dm cases and it's been thinking job as I have planning for proper design[1] to go ahead. I understand you sent multiple versions but I already told that "please don't send the next version" until the current version discussion finishes.
[1] http://git.denx.de/?p=u-boot/u-boot-spi.git;a=shortlog;h=refs/heads/next-spi...
That's why I proposed earlier that you take the non-SPI patches first and then we can rework the SPI flash pieces.
I will come back for this patch changes what I thought, probably lock ops shouldn't be separate for dm and non-dm as we have different ops based on the flash itself.
I wish we can make progress on this. Not sure how to proceed though.
thanks!