[En-Nut-Discussion] NutGpio Api and Multidrive for AVR

[Uwe Bonnes]

>>>>> "Ole" == Ole Reinhardt <ole.reinhardt at embedded-it.de> writes:

    Ole> Hi Uwe, unfortunately I did not really understand the functionality
    Ole> of your enhanced GPIO API functions "NutGpioPinDrive()" and
    Ole> "NutGpioPinRelease()"

    Ole> Could you explaint them a little more in detail?

    >> The last NutGpioPinSetHig makes sure the AVR pullup turns on the
    >> internal pullup. For non-AVR architectures, NutGpioPinSetLow(...)
    >> and NutGpioPinDrive(...) are empty function in most cases

    Ole> Is this just a typo or do you really mean that NutGpioPinSetLow()
    Ole> is an empty function in most cases? If so, why?

    Ole> I just have concerns that we might have different understanding of
    Ole> what the function shall do on different architectures and will just
    Ole> prevent to get some incompatibilities there.

As explained, for AVR GPIO port setup and usage are not seperated in the
registers. So there is no (easy(*)) way on AVR to make the sequence

GpioPinConfigSet(PORT, PIN, GPIO_CFG_OUTPUT)
GpioPinSetLow(PORT, PIN)
GpioPinSetHigh(PORT, PIN)

do sbi(DDR(PORT), PIN), cbi(PORT, PIN), sbi(PORT, PIN)

and 

GpioPinConfigSet(PORT, PIN, GPIO_CFG_OUTPUT|GPIO_CFG_MULTIDRIVE|GPIO_CFG_PULLUP)
GpioPinSetLow(PORT, PIN)
GpioPinSetHigh(PORT, PIN)

do sbi(DDR(PORT), PIN), cbi(PORT, PIN), cbi(DDR(PORT), PIN) and eventual 
additional sbi( PORT, PIN) to activate the PULLUP again.

So I require now to write the sequence like
GpioPinConfigSet(PORT, PIN, GPIO_CFG_OUTPUT|GPIO_CFG_MULTIDRIVE)
GpioPinSetLow(PORT, PIN)
GpioPinSetDrive(PORT, PIN)
GpioPinSetRelease(PORT, PIN)
GpioPinSetHigh(PORT, PIN)
if the code should be portable to AVR too, like in the bitbanging one-wire
bus driver dev/owibus_bbif.c

For any sensible architecture the new commands are defined in the header as
#define GpioPinSetDrive(PORT, PIN)
#define GpioPinSetRelease(PORT, PIN)
so they evaluate empty.

(*) We could carry shadow register to hold the configuration we set up, but
I think that is incredible overhead and memory hog. So the two new commands
are a fair price we pay for compatibility on AVR

Bye

-- 
Uwe Bonnes                bon at elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------