[dev.cc] crypto/md5, crypto/sha1: restore a few SP references

Applying my post-submit comments from CL 5120.
The rewrite there changed the code from writing to the stack
frame to writing below the stack frame.

Change-Id: Ie7e0563c0c1731fede2bcefeaf3c9d88a0cf4063
Reviewed-on: https://go-review.googlesource.com/5470Reviewed-by: David Crawshaw <crawshaw@golang.org>

src/crypto/md5/md5block_arm.s

@@ -29,31 +29,31 @@
 //12(FP) is p.cap
 //
 // Stack frame
-#define p_end	-4	// -4(R13==SP) pointer to the end of data
-#define p_data	-8	// -8(R13) current data pointer
-#define buf	(-8-4*16)	//-72(R13) 16 words temporary buffer
+#define p_end	end-4(SP)	// pointer to the end of data
+#define p_data	data-8(SP)	// current data pointer
+#define buf	buffer-(8+4*16)(SP)	//16 words temporary buffer
 		// 3 words at 4..12(R13) for called routine parameters
 
 TEXT	·block(SB), NOSPLIT, $84-16
 	MOVW	p+4(FP), Rdata	// pointer to the data
 	MOVW	p_len+8(FP), Rt0	// number of bytes
 	ADD	Rdata, Rt0
-	MOVW	Rt0, p_end(R13)	// pointer to end of data
+	MOVW	Rt0, p_end	// pointer to end of data
 
 loop:
-	MOVW	Rdata, p_data(R13)	// Save Rdata
+	MOVW	Rdata, p_data	// Save Rdata
 	AND.S	$3, Rdata, Rt0	// TST $3, Rdata not working see issue 5921
 	BEQ	aligned			// aligned detected - skip copy
 
 	// Copy the unaligned source data into the aligned temporary buffer
 	// memove(to=4(R13), from=8(R13), n=12(R13)) - Corrupts all registers
-	MOVW	$buf(R13), Rtable	// to
+	MOVW	$buf, Rtable	// to
 	MOVW	$64, Rc0		// n
 	MOVM.IB	[Rtable,Rdata,Rc0], (R13)
 	BL	runtime·memmove(SB)
 
 	// Point to the local aligned copy of the data
-	MOVW	$buf(R13), Rdata
+	MOVW	$buf, Rdata
 
 aligned:
 	// Point to the table of constants
@@ -217,8 +217,8 @@ aligned:
 
 	MOVM.IA [Ra,Rb,Rc,Rd], (Rt0)
 
-	MOVW	p_data(R13), Rdata
-	MOVW	p_end(R13), Rt0
+	MOVW	p_data, Rdata
+	MOVW	p_end, Rt0
 	ADD	$64, Rdata
 	CMP	Rt0, Rdata
 	BLO	loop

src/crypto/sha1/sha1block_arm.s

@@ -45,10 +45,10 @@
 //12(FP) is p.cap
 //
 // Stack frame
-#define p_end	-4		// -4(SP) pointer to the end of data
-#define p_data	(p_end - 4)	// -8(SP) current data pointer
-#define w_buf	(p_data - 4*80)	// -328(SP) 80 words temporary buffer w uint32[80]
-#define saved	(w_buf - 4*5)	// -348(SP) saved sha1 registers a,b,c,d,e - these must be last
+#define p_end	end-4(SP)		// pointer to the end of data
+#define p_data	data-8(SP)	// current data pointer (unused?)
+#define w_buf	buf-(8+4*80)(SP)	//80 words temporary buffer w uint32[80]
+#define saved	abcde-(8+4*80+4*5)(SP)	// saved sha1 registers a,b,c,d,e - these must be last (unused?)
 // Total size +4 for saved LR is 352
 
 	// w[i] = p[j]<<24 | p[j+1]<<16 | p[j+2]<<8 | p[j+3]
@@ -141,7 +141,7 @@ TEXT	·block(SB), 0, $352-16
 	MOVW	p+4(FP), Rdata	// pointer to the data
 	MOVW	p_len+8(FP), Rt0	// number of bytes
 	ADD	Rdata, Rt0
-	MOVW	Rt0, p_end(R13)	// pointer to end of data
+	MOVW	Rt0, p_end	// pointer to end of data
 
 	// Load up initial SHA1 accumulator
 	MOVW	dig+0(FP), Rt0
@@ -151,7 +151,7 @@ loop:
 	// Save registers at SP+4 onwards
 	MOVM.IB [Ra,Rb,Rc,Rd,Re], (R13)
 
-	MOVW	$w_buf(R13), Rw
+	MOVW	$w_buf, Rw
 	MOVW	$0x5A827999, Rconst
 	MOVW	$3, Rctr
 loop1:	ROUND1(Ra, Rb, Rc, Rd, Re)
@@ -206,7 +206,7 @@ loop4:	ROUND4(Ra, Rb, Rc, Rd, Re)
 	ADD	Rctr, Rd
 	ADD	Rw, Re
 
-	MOVW	p_end(R13), Rt0
+	MOVW	p_end, Rt0
 	CMP	Rt0, Rdata
 	BLO	loop