Why pic18fxxx microcontrollers is better for C language

Question

Pic18fxxx family has a lot advantages and improvements from the pic16fxxx family of microcontrollers. What is the particular feature that make them better for C programming than the pic16fxxx devices?

• Adam Lawrence: What makes you think they're ideal for C? I would argue that any PIC24 device is far more 'ideal' than any PIC16/18 device from an architectural standpoint (more registers, more RAM...), and any PIC32 is far more 'ideal' than a PIC24 by nature of it's MIPS architecture...
• xkrpz: Sorry, i write it wrong. Ideal compared to pic14f devices not to others.
• user17592: PIC14F? That family doesn't exist, does it?
• xkrpz: I mean pic16fxxx. Mistake upon mistake!
• Olin Lathrop: Actually there was a PIC 14 once, but now along with the PIC 17 it "never existed".

Answer 1

I don't know about "ideal for". That's a marketing term that has no place in a learned discussion.

However, one big advantage of the PIC 18 architecture versus the original PIC 16 architecture, especially related to compilers, is that it is possible to implement a software data stack on a PIC 18 with single instructions for PUSH and POP. The PIC 18 also has a deeper call stack, 32 versus 8, which can help when a compiler implicitly calls subroutines. The PLUSW indirect addressing mode is probably something a compiler would make more frequent use of than a human programmer.

The PIC 18 also has other advantages that are useful both for compilers and human programmers, like 3 hardware pointers instead of one, auto inc/dec indirect addressing modes, a 8x8 into 16 hardware multiplier, add with carry, and subtract with borrow, and a few other niceties.

• xkrpz: You are right, i changed to 'better' than 'ideal'.

Answer 2

Some C compilers store local variables on a hardware or software stack. When code enters a subroutine, the size of its local variables is added to or subtracted from a register which is used as a stack/frame pointer. When the "enhanced instruction set" is enabled, PIC18Fxx instructions can access the first ~96 bytes following the frame-pointer address directly. This is ALMOST a really good feature. It would have been a really good feature if e.g. the processor had made 15 bytes of the unbanked address space map to addresses 1-15 bytes above FSR2, and 7 bytes map to addresses 1-7 bytes above the each of the other FSR registers, while leaving most of the unbanked space as unbanked RAM. Unfortunately, notwithstanding the fact that most functions won't even have eight bytes of local storage (much less 16), giving up 96 addresses to handle offsets 0-95 above FSR2 is absurdly wasteful.

Note that some other C compiler systems assign fixed addresses for local variables; such systems determine at link time which functions cannot be "live" simultaneously, and can allocate the variables for such functions at the same addresses. Provided code does is not called in re-entrant fashion, this approach can be just as efficient (sometimes more so) than using a stack; it does, however, not work with re-entrant routines.

Answer 3

You can ask google well about the lot of comparison between PIC16F vs PIC18F.

Coming to you question Why pic18fxxx microcontrollers are better for C language?

PIC18F or any other micro-controller that can be assumed better for C language are those micro controllers that supports such assembly instructions by default that can help C compilers to make efficient and optimized hex files.

One major example of these kind of instructions can be MATH related instructions

XC8 compiler supports extended instruction set which is specially helps C compilers to generate we optimized binary file.

• Olin Lathrop: This is a non-answer. The OP already knows that a PIC 18 supports C better, but asked how specifically it does that.
• Abdul Rehman: I mentioned already that by supporting special power full instruction set that pic16f does not have
• Olin Lathrop: Right, but that is obvious and was already known. The OP is already aware that the PIC 18 has "a lot advantages and improvements from the pic16fxxx". Without explaining what specifically, like particular instructions and the like, you think aids compilers, this is no answer at all.