in the name of zero

ain’t she the cutest!!!!!!?????? i’m soo… like… smitten! but geez! i can’t believe i’m in love with a damn drawing!

i’m actually using gcc4 and noticed that there are minor deviations
from the usual prolog/epilog functions since gcc3 when doing intel. i
have to read up on this.

so let’s try to follow a simple “hello world” program to see how gcc makes ppc
programs. it’s compiled with no optimizations.

int procedure1(const char *string)
{
        printf(string);
        return 0;
}
	
int main(int argc, char *argv[])
{
        return procedure1(\"hello world\n\");
}

the V4 stack frame.

	SP---->	+---------------------------------------+
		| back chain to caller			| 0
		+---------------------------------------+
		| saved LR				| 4
		+---------------------------------------+
		| Parameter save area (P)		| 8
		+---------------------------------------+
		| Alloca space (A)			| 8+P
		+---------------------------------------+
		| Local variable space (L)		| 8+P+A
		+---------------------------------------+
		| saved CR (C)				| 8+P+A+L
		+---------------------------------------+
		| Save area for GP registers (G)	| 8+P+A+L+C
		+---------------------------------------+
		| Save area for FP registers (F)	| 8+P+A+L+C+G
		+---------------------------------------+
	old SP->| back chain to caller's caller		|
		+---------------------------------------+

the V4 Registers

r0 volatile, may be used by function linkage
r1 stack pointer
r2 reserved for system
r3 .. r4 volatile, pass 1st - 2nd int args, return 1st - 2nd ints
r5 .. r10 volatile, pass 3rd - 8th int args
r11 .. r12 volatile, may be used by function linkage
r13 small data area pointer
r14 .. r31 saved
f0 volatile
f1 volatile, pass 1st float arg, return 1st float
f2 .. f8 volatile, pass 2nd - 8th float args
f9 .. f13 volatile
f14 .. f30 saved
f31 saved, static chain if needed.
lr volatile, return address
ctr volatile
xer volatile
fpscr volatile*
cr0 volatile
cr1 volatile**
cr2 .. cr4 saved
cr5 .. cr7 volatile

* The VE, OE, UE, ZE, XE, NI, and RN (rounding mode) bits of the FPSCR may be
changed only by a called function such as fpsetround that has the documented
effect of changing them, the rest of the FPSCR is volatile.

** Bit 6 of the CR (CR1 floating point invalid exception bit) is set to 1 if a
variable argument function is passed floating point arguments in registers.

the PPC architecture does not have a push/pop instruction that implicitly
operates on the stack. as such, stack management could be a little more “hands
on” on PPC in comparison to intel. the stack frame convention above is defined
to support parameter passing, reserved registers (nonvolatile) preservation and
local variables. each function which either calls another function or modifies
saved register must create a stack frame from memory set aside for use as a
stack defined by the r1 register (r1 = sp).

the stwu instruction assures that stack frame allocation is atomic. stack
space is allocated and the sp (r1) is updated in just one instruction.

parameter passing. ppc passes function parameters thru registers rather than
pushing them all in the stack.

<procedure1>
stwu    r1,-32(r1)  ; save former and allocate a new stack frame
mflr    r0          ; return addr of procedure1()
stw     r31,28(r1)  ; save r31
stw     r0,36(r1)   ; save r0
mr      r31,r1      ; mirror sp
stw     r3,8(r31)   ; save char * to stack
lwz     r3,8(r31)   ; argument to printf
bl      10010a00

li      r0,0        ; r0 = 0
mr      r3,r0       ; r3 = 0 = return value
lwz     r11,0(r1)   ; access back chain
lwz     r0,4(r11)   ; access return address
mtlr    r0          ; update link register with return addr of procedure1()
lwz     r31,-4(r11) ; restore r31
mr      r1,r11      ; restore caller’s stack
blr                 ; return to caller
	
<main>
stwu    r1,-32(r1)      ; save former frame and allocate a new stack frame
mflr    r0              ; return address of main()
stw     r31,28(r1)      ; save r31
stw     r0,36(r1)       ; save link register
mr      r31,r1          ; mirror sp
stw     r3,8(r31)       ; save r3
stw     r4,12(r31)      ; save r4
lis     r9,4096         ; lower 16 bits of string addr
addi    r3,r9,2176      ; higher 16 bits of string addr
bl      10000434
 ; procedure1(char *)
mr      r0,r3           ; r0 = return val of procedure1()
mr      r3,r0           ; r3 = 0 (retval)
lwz     r11,0(r1)       ; access back chain
lwz     r0,4(r11)       ; access return address
mtlr    r0              ; update link register
lwz     r31,-4(r11)     ; restore r31
mr      r1,r11          ; restore stack frame
blr                     ; leave main()

next time, we’ll delve more deeper and see how to access system calls directly
and how these system calls are implemented.