Lorenzo Lutti wrote: > From: "Kevin Hilman" <[EMAIL PROTECTED]> > >>> 1) In arch/arm/mach-davinci/clock.c, the GPIO clock is >>> set at 27 MHz (you >>> can see it also in /proc/davinci_clocks) while the >>> maximum GPIO clock speed >>> is 10 MHz. >> Good catch. > > However this doesn't seem to affect the GPIO operation in any way, I suppose > it has just an informative purpose. GPIO stage is fed with SYSCLK5, and if > I'm not wrong, with the board defaults that's 94.5 MHz. The "maximum > operation speed" of 10 MHz depicted in SPRUE25 is probably just a > consequence of the actual delay propagation of the GPIO stages (see the > DM6446 datasheet, page 108). > >>> 2) In arch/arm/mach-davinci/gpio.c, the function >>> gpio_set_direction() is >>> declared as __init. This is quite restrictive, because >>> there are a lot of >>> situations where you may need to change a the direction >>> of a GPIO >>> dynamically (e.g. bidirectional data lines). >> Removal of __init is OK, but your patch also renames the >> version in >> gpio.c (to something that is never called) and moves the >> function to >> gpio.h. What's the reason for that? > > Yes, you are right, it doesn't make sense. I just wanted to reproduce the > way the other functions were implemented (an inline "stub" in gpio.h that > calls when needed the real function in gpio.c), but then I've mistakenly > copied all the function in the .h, and this doesn't help at all. :) > > I just removed __init, the new patch (still against the original code) is > attached.
OK, this patch looks pretty good. One comment and formalities for patch submission that make it easier for me to automate the patch process. When you add the gpiorate, you commented out the comment rate. Just remove it. The patch clearly shows what was removed and what was added. Can you submit your patches inline with an email subject of the form: [PATCH] ARM: DaVinci: <brief one-line description> Then in the email a more detail description (if necessary) and then a Signed-off-by: Your Name <[EMAIL PROTECTED]> Followed by the patch. Thanks a lot for you contributions. I don't mean to make your life more difficult, but this is the linux-kernel way for public submission and review of patches. Following this makes it easier to review, comment and then merge incoming patches. Thanks, Kevin _______________________________________________ Davinci-linux-open-source mailing list [email protected] http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
