On Tue, Jul 18, 2023 at 11:47 AM Tudor Ambarus tudor.ambarus@linaro.org wrote:
Hi, Linus,
On 7/18/23 09:32, Michael Walle wrote:
FWIW, I'm fine with the removed no_sfdp_flags if INFO(, 0, 0, 0).
We'll need a v5 where you test again the flash with mtd_utils, as we want to get rid of n_sectors and sectors_size and instead determine them from SFDP. We agreed that the flash entry should be defined with the following params:
{ "w25q128", INFO(0xef4018, 0, 0, 0) PARSE_SFDP FLAGS(SPI_NOR_HAS_LOCK | SPI_NOR_HAS_TB) },
Roger that!
I sent a v5 which does this, tested by taking the reported size in dmesg and divide by reported (correct) eraseblock size and it results in the same number of sectors as well.
I put the details into the commit message.
Yours, Linus Walleij