David Jander wrote:
Hi Stefano,
Hi David,
I just checked in the reference manual of the i.MX31, and there the meaning of this bit has the same polarity as on the i.MX51, so you'll need to fix this also at the end of the spi_setup_slave() function, in the #else path of the #ifdef CONFIG_MX51 directive.
if (!(mode & SPI_CPOL)) ctrl_reg |= MXC_CSPICTRL_POL;
should be:
if (mode & SPI_CPOL) ctrl_reg |= MXC_CSPICTRL_POL;
Would be nice if someone with a MX31 board could verify this.
I can test myself on a qong board - I hope someone else can test on other MX.31 boards.
I agree, and I guess you can better include it in your patch-set, otherwise I'd have to wait for your patches and then provide my own patch on top of that.... too complicated :-)
I will do it - I think I could send in a short time the whole patchset for review :-).
I am also adding support for S25FL032P chips to the spansion driver. Will post a patch later.
I have seen. However, it should be better if you send the patch also to the maintainer for the MTD subsystem (Stefan Roese, I set his address in CC).
Right now I have correct detection of the chip, but the environment is not saved and read back correctly. Still investigating... maybe some chip configuration prolem in the spansion driver?
Probably. I have got some issues with the ST that do not depend from the modifications in SPI driver.
Best regards, Stefano Babic