- use c89-style comments

  Closes issue #5194

25 files changed, 854 insertions, 853 deletions

diff --git a/libc/string/bfin/memchr.S b/libc/string/bfin/memchr.S
index 88e46bef6..26d419f7c 100644
--- a/libc/string/bfin/memchr.S
+++ b/libc/string/bfin/memchr.S
@@ -25,8 +25,8 @@
 
 .weak _memchr
 ENTRY(_memchr)
-	P0 = R0;             // P0 = address
-	P2 = R2;             // P2 = count
+	P0 = R0;             /* P0 = address */
+	P2 = R2;             /* P2 = count */
 	R1 = R1.B(Z);
 	CC = R2 == 0;
 	IF CC JUMP .Lfailed;
diff --git a/libc/string/bfin/strcmp.S b/libc/string/bfin/strcmp.S
index 12e8c53c6..ef23aa9ab 100644
--- a/libc/string/bfin/strcmp.S
+++ b/libc/string/bfin/strcmp.S
@@ -29,66 +29,66 @@ ENTRY(_strcmp)
 	p1 = r0;
 	p2 = r1;
 
-	p0 = -1;	// (need for loop counter init)
+	p0 = -1;	/* (need for loop counter init) */
 
-	  // check if byte aligned
-	r0 = r0 | r1;	// check both pointers at same time
-	r0 <<= 30;	// dump all but last 2 bits
-	cc = az;	// are they zero?
-	if !cc jump .Lunaligned;	// no; use unaligned code.
-			// fall-thru for aligned case..
+	  /* check if byte aligned */
+	r0 = r0 | r1;	/* check both pointers at same time */
+	r0 <<= 30;	/* dump all but last 2 bits */
+	cc = az;	/* are they zero? */
+	if !cc jump .Lunaligned;	/* no; use unaligned code. */
+			/* fall-thru for aligned case.. */
 
-	  // note that r0 is zero from the previous...
-	  //           p0 set to -1
+	  /* note that r0 is zero from the previous... */
+	  /*           p0 set to -1 */
 
 	LSETUP (.Lbeginloop, .Lendloop) lc0=p0;
-	  // pick up first words
+	  /* pick up first words */
 	r1 = [p1++];
 	r2 = [p2++];
-	  // make up mask:  0FF0FF
+	  /* make up mask:  0FF0FF */
 	r7 = 0xFF;
 	r7.h = 0xFF;
-		// loop : 9 cycles to check 4 characters
+		/* loop : 9 cycles to check 4 characters */
 	cc = r1 == r2;
 .Lbeginloop:
-	if !cc jump .Lnotequal4;	// compare failure, exit loop
+	if !cc jump .Lnotequal4;	/* compare failure, exit loop */
 
-	  // starting with   44332211
-	  // see if char 3 or char 1 is 0
-	r3 = r1 & r7;		// form 00330011
-	  // add to zero, and (r2 is free, reload)
+	  /* starting with   44332211 */
+	  /* see if char 3 or char 1 is 0 */
+	r3 = r1 & r7;		/* form 00330011 */
+	  /* add to zero, and (r2 is free, reload) */
 	r6 = r3 +|+ r0 || r2 = [p2++] || nop;
-	cc = az;	// true if either is zero
-	r3 = r1 ^ r3;	        // form 44002200 (4321^0301 => 4020)
-				// (trick, saves having another mask)
-	// add to zero,  and  (r1 is free, reload)
+	cc = az;	/* true if either is zero */
+	r3 = r1 ^ r3;	        /* form 44002200 (4321^0301 => 4020) */
+				/* (trick, saves having another mask) */
+	/* add to zero,  and  (r1 is free, reload) */
 	r6 = r3 +|+ r0 || r1 = [p1++] || nop;
-	cc |= az;	// true if either is zero
-	if cc jump .Lzero4;	// leave if a zero somewhere
+	cc |= az;	/* true if either is zero */
+	if cc jump .Lzero4;	/* leave if a zero somewhere */
 .Lendloop:
 	cc = r1 == r2;
 
- // loop exits
-.Lnotequal4:		// compare failure on 4-char compare
-			// address pointers are one word ahead;
-			// faster to use zero4 exit code
+ /* loop exits */
+.Lnotequal4:		/* compare failure on 4-char compare */
+			/* address pointers are one word ahead; */
+			/* faster to use zero4 exit code */
 	p1 += 4;
 	p2 += 4;
 
-.Lzero4:			// one of the bytes in word 1 is zero
-			// but we've already fetched the next word; so
-			// backup two to look at failing word again
+.Lzero4:			/* one of the bytes in word 1 is zero */
+			/* but we've already fetched the next word; so */
+			/* backup two to look at failing word again */
 	p1 += -8;
 	p2 += -8;
 
 
 
-		// here when pointers are unaligned: checks one
-		// character at a time.  Also use at the end of
-		// the word-check algorithm to figure out what happened
+		/* here when pointers are unaligned: checks one */
+		/* character at a time.  Also use at the end of */
+		/* the word-check algorithm to figure out what happened */
 .Lunaligned:
-	  //	R0 is non-zero from before.
-	  //           p0 set to -1
+	  /*	R0 is non-zero from before. */
+	  /*           p0 set to -1 */
 
 	r0 = 0 (Z);
 	r1 = B[p1++] (Z);
@@ -96,18 +96,18 @@ ENTRY(_strcmp)
 	LSETUP (.Lbeginloop1, .Lendloop1) lc0=p0;
 
 .Lbeginloop1:
-	cc = r1;	// first char must be non-zero
-	// chars must be the same
+	cc = r1;	/* first char must be non-zero */
+	/* chars must be the same */
 	r3 = r2 - r1 (NS) || r1 = B[p1++] (Z) || nop;
 	cc &= az;
-	r3 = r0 - r2;	// second char must be non-zero
+	r3 = r0 - r2;	/* second char must be non-zero */
 	cc &= an;
 	if !cc jump .Lexitloop1;
 .Lendloop1:
 	r2 = B[p2++] (Z);
 
-.Lexitloop1: // here means we found a zero or a difference.
-	   // we have r2(N), p2(N), r1(N+1), p1(N+2)
+.Lexitloop1: /* here means we found a zero or a difference. */
+	   /* we have r2(N), p2(N), r1(N+1), p1(N+2) */
 	r1=B[p1+ -2] (Z);
 	r0 = r1 - r2;
 	(r7:4) = [sp++];
diff --git a/libc/string/generic/strtok_r.c b/libc/string/generic/strtok_r.c
index d082d226e..7648212f7 100644
--- a/libc/string/generic/strtok_r.c
+++ b/libc/string/generic/strtok_r.c
@@ -29,17 +29,17 @@
 # define __rawmemchr strchr
 /* Experimentally off - libc_hidden_proto(strchr) */
 #endif
-
-/* Parse S into tokens separated by characters in DELIM.
+#if 0
+   Parse S into tokens separated by characters in DELIM.
    If S is NULL, the saved pointer in SAVE_PTR is used as
    the next starting point.  For example:
 	char s[] = "-abc-=-def";
 	char *sp;
-	x = strtok_r(s, "-", &sp);	// x = "abc", sp = "=-def"
-	x = strtok_r(NULL, "-=", &sp);	// x = "def", sp = NULL
-	x = strtok_r(NULL, "=", &sp);	// x = NULL
-		// s = "abc\0-def\0"
-*/
+	x = strtok_r(s, "-", &sp);	/* x = "abc", sp = "=-def" */
+	x = strtok_r(NULL, "-=", &sp);	/* x = "def", sp = NULL */
+	x = strtok_r(NULL, "=", &sp);	/* x = NULL */
+		/* s = "abc\0-def\0" */
+#endif
 char *strtok_r (char *s, const char *delim, char **save_ptr)
 {
   char *token;
diff --git a/libc/string/ia64/bzero.S b/libc/string/ia64/bzero.S
index d390838a6..1f0f8b7ac 100644
--- a/libc/string/ia64/bzero.S
+++ b/libc/string/ia64/bzero.S
@@ -47,13 +47,13 @@
 #define ptr1		r28
 #define ptr2		r27
 #define ptr3		r26
-#define ptr9 		r24
+#define ptr9		r24
 #define	loopcnt		r23
 #define linecnt		r22
 #define bytecnt		r21
 
-// This routine uses only scratch predicate registers (p6 - p15)
-#define p_scr		p6	// default register for same-cycle branches
+/* This routine uses only scratch predicate registers (p6 - p15) */
+#define p_scr		p6	/* default register for same-cycle branches */
 #define p_unalgn	p9
 #define p_y		p11
 #define p_n		p12
@@ -65,7 +65,7 @@
 #define MIN1		15
 #define MIN1P1HALF	8
 #define LINE_SIZE	128
-#define LSIZE_SH        7			// shift amount
+#define LSIZE_SH        7			/* shift amount */
 #define PREF_AHEAD	8
 
 #define USE_FLP
@@ -87,49 +87,49 @@ ENTRY(bzero)
 	movi0	save_lc = ar.lc
 } { .mmi
 	.body
-	mov	ret0 = dest		// return value
+	mov	ret0 = dest		/* return value */
 	nop.m	0
 	cmp.eq	p_scr, p0 = cnt, r0
 ;; }
 { .mmi
-	and	ptr2 = -(MIN1+1), dest	// aligned address
-	and	tmp = MIN1, dest	// prepare to check for alignment
-	tbit.nz p_y, p_n = dest, 0	// Do we have an odd address? (M_B_U)
+	and	ptr2 = -(MIN1+1), dest	/* aligned address */
+	and	tmp = MIN1, dest	/* prepare to check for alignment */
+	tbit.nz p_y, p_n = dest, 0	/* Do we have an odd address? (M_B_U) */
 } { .mib
 	mov	ptr1 = dest
 	nop.i	0
-(p_scr)	br.ret.dpnt.many rp		// return immediately if count = 0
+(p_scr)	br.ret.dpnt.many rp		/* return immediately if count = 0 */
 ;; }
 { .mib
 	cmp.ne	p_unalgn, p0 = tmp, r0
-} { .mib					// NB: # of bytes to move is 1
-	sub	bytecnt = (MIN1+1), tmp		//     higher than loopcnt
-	cmp.gt	p_scr, p0 = 16, cnt		// is it a minimalistic task?
-(p_scr)	br.cond.dptk.many .move_bytes_unaligned	// go move just a few (M_B_U)
+} { .mib					/* NB: # of bytes to move is 1 */
+	sub	bytecnt = (MIN1+1), tmp		/*     higher than loopcnt */
+	cmp.gt	p_scr, p0 = 16, cnt		/* is it a minimalistic task? */
+(p_scr)	br.cond.dptk.many .move_bytes_unaligned	/* go move just a few (M_B_U) */
 ;; }
 { .mmi
-(p_unalgn) add	ptr1 = (MIN1+1), ptr2		// after alignment
-(p_unalgn) add	ptr2 = MIN1P1HALF, ptr2		// after alignment
-(p_unalgn) tbit.nz.unc p_y, p_n = bytecnt, 3	// should we do a st8 ?
+(p_unalgn) add	ptr1 = (MIN1+1), ptr2		/* after alignment */
+(p_unalgn) add	ptr2 = MIN1P1HALF, ptr2		/* after alignment */
+(p_unalgn) tbit.nz.unc p_y, p_n = bytecnt, 3	/* should we do a st8 ? */
 ;; }
 { .mib
 (p_y)	add	cnt = -8, cnt
-(p_unalgn) tbit.nz.unc p_yy, p_nn = bytecnt, 2	// should we do a st4 ?
+(p_unalgn) tbit.nz.unc p_yy, p_nn = bytecnt, 2	/* should we do a st4 ? */
 } { .mib
 (p_y)	st8	[ptr2] = r0,-4
 (p_n)	add	ptr2 = 4, ptr2
 ;; }
 { .mib
 (p_yy)	add	cnt = -4, cnt
-(p_unalgn) tbit.nz.unc p_y, p_n = bytecnt, 1	// should we do a st2 ?
+(p_unalgn) tbit.nz.unc p_y, p_n = bytecnt, 1	/* should we do a st2 ? */
 } { .mib
 (p_yy)	st4	[ptr2] = r0,-2
 (p_nn)	add	ptr2 = 2, ptr2
 ;; }
 { .mmi
-	mov	tmp = LINE_SIZE+1		// for compare
+	mov	tmp = LINE_SIZE+1		/* for compare */
 (p_y)	add	cnt = -2, cnt
-(p_unalgn) tbit.nz.unc p_yy, p_nn = bytecnt, 0	// should we do a st1 ?
+(p_unalgn) tbit.nz.unc p_yy, p_nn = bytecnt, 0	/* should we do a st1 ? */
 } { .mmi
 	nop.m	0
 (p_y)	st2	[ptr2] = r0,-1
@@ -138,44 +138,44 @@ ENTRY(bzero)
 
 { .mmi
 (p_yy)	st1	[ptr2] = r0
-  	cmp.gt	p_scr, p0 = tmp, cnt		// is it a minimalistic task?
+	cmp.gt	p_scr, p0 = tmp, cnt		/* is it a minimalistic task? */
 } { .mbb
 (p_yy)	add	cnt = -1, cnt
-(p_scr)	br.cond.dpnt.many .fraction_of_line	// go move just a few
+(p_scr)	br.cond.dpnt.many .fraction_of_line	/* go move just a few */
 ;; }
 { .mib
-	nop.m 	0
+	nop.m	0
 	shr.u	linecnt = cnt, LSIZE_SH
 	nop.b	0
 ;; }
 
 	.align 32
-.l1b:	// ------------------//  L1B: store ahead into cache lines; fill later
+.l1b:	/* ------------------  L1B: store ahead into cache lines; fill later */
 { .mmi
-	and	tmp = -(LINE_SIZE), cnt		// compute end of range
-	mov	ptr9 = ptr1			// used for prefetching
-	and	cnt = (LINE_SIZE-1), cnt	// remainder
+	and	tmp = -(LINE_SIZE), cnt		/* compute end of range */
+	mov	ptr9 = ptr1			/* used for prefetching */
+	and	cnt = (LINE_SIZE-1), cnt	/* remainder */
 } { .mmi
-	mov	loopcnt = PREF_AHEAD-1		// default prefetch loop
-	cmp.gt	p_scr, p0 = PREF_AHEAD, linecnt	// check against actual value
+	mov	loopcnt = PREF_AHEAD-1		/* default prefetch loop */
+	cmp.gt	p_scr, p0 = PREF_AHEAD, linecnt	/* check against actual value */
 ;; }
 { .mmi
 (p_scr)	add	loopcnt = -1, linecnt
-	add	ptr2 = 16, ptr1	// start of stores (beyond prefetch stores)
-	add	ptr1 = tmp, ptr1	// first address beyond total range
+	add	ptr2 = 16, ptr1	/* start of stores (beyond prefetch stores) */
+	add	ptr1 = tmp, ptr1	/* first address beyond total range */
 ;; }
 { .mmi
-	add	tmp = -1, linecnt	// next loop count
+	add	tmp = -1, linecnt	/* next loop count */
 	movi0	ar.lc = loopcnt
 ;; }
 .pref_l1b:
 { .mib
-	stf.spill [ptr9] = f0, 128	// Do stores one cache line apart
+	stf.spill [ptr9] = f0, 128	/* Do stores one cache line apart */
 	nop.i   0
 	br.cloop.dptk.few .pref_l1b
 ;; }
 { .mmi
-	add	ptr0 = 16, ptr2		// Two stores in parallel
+	add	ptr0 = 16, ptr2		/* Two stores in parallel */
 	movi0	ar.lc = tmp
 ;; }
 .l1bx:
@@ -190,7 +190,7 @@ ENTRY(bzero)
  { .mmi
 	stf.spill [ptr2] = f0, 32
 	stf.spill [ptr0] = f0, 64
- 	cmp.lt	p_scr, p0 = ptr9, ptr1	// do we need more prefetching?
+	cmp.lt	p_scr, p0 = ptr9, ptr1	/* do we need more prefetching? */
  ;; }
 { .mmb
 	stf.spill [ptr2] = f0, 32
@@ -198,14 +198,14 @@ ENTRY(bzero)
 	br.cloop.dptk.few .l1bx
 ;; }
 { .mib
-	cmp.gt  p_scr, p0 = 8, cnt	// just a few bytes left ?
+	cmp.gt  p_scr, p0 = 8, cnt	/* just a few bytes left ? */
 (p_scr)	br.cond.dpnt.many  .move_bytes_from_alignment
 ;; }
 
 .fraction_of_line:
 { .mib
 	add	ptr2 = 16, ptr1
-	shr.u	loopcnt = cnt, 5   	// loopcnt = cnt / 32
+	shr.u	loopcnt = cnt, 5	/* loopcnt = cnt / 32 */
 ;; }
 { .mib
 	cmp.eq	p_scr, p0 = loopcnt, r0
@@ -213,11 +213,11 @@ ENTRY(bzero)
 (p_scr)	br.cond.dpnt.many .store_words
 ;; }
 { .mib
-	and	cnt = 0x1f, cnt		// compute the remaining cnt
+	and	cnt = 0x1f, cnt		/* compute the remaining cnt */
 	movi0   ar.lc = loopcnt
 ;; }
 	.align 32
-.l2:	// -----------------------------//  L2A:  store 32B in 2 cycles
+.l2:	/* -----------------------------  L2A:  store 32B in 2 cycles */
 { .mmb
 	store	[ptr1] = myval, 8
 	store	[ptr2] = myval, 8
@@ -228,38 +228,38 @@ ENTRY(bzero)
 ;; }
 .store_words:
 { .mib
-	cmp.gt	p_scr, p0 = 8, cnt	// just a few bytes left ?
-(p_scr)	br.cond.dpnt.many .move_bytes_from_alignment	// Branch
+	cmp.gt	p_scr, p0 = 8, cnt	/* just a few bytes left ? */
+(p_scr)	br.cond.dpnt.many .move_bytes_from_alignment	/* Branch */
 ;; }
 
 { .mmi
-	store	[ptr1] = myval, 8	// store
-	cmp.le	p_y, p_n = 16, cnt	//
-	add	cnt = -8, cnt		// subtract
+	store	[ptr1] = myval, 8	/* store */
+	cmp.le	p_y, p_n = 16, cnt	/* */
+	add	cnt = -8, cnt		/* subtract */
 ;; }
 { .mmi
-(p_y)	store	[ptr1] = myval, 8	// store
+(p_y)	store	[ptr1] = myval, 8	/* store */
 (p_y)	cmp.le.unc p_yy, p_nn = 16, cnt
-(p_y)	add	cnt = -8, cnt		// subtract
+(p_y)	add	cnt = -8, cnt		/* subtract */
 ;; }
-{ .mmi					// store
+{ .mmi					/* store */
 (p_yy)	store	[ptr1] = myval, 8
-(p_yy)	add	cnt = -8, cnt		// subtract
+(p_yy)	add	cnt = -8, cnt		/* subtract */
 ;; }
 
 .move_bytes_from_alignment:
 { .mib
 	cmp.eq	p_scr, p0 = cnt, r0
-	tbit.nz.unc p_y, p0 = cnt, 2	// should we terminate with a st4 ?
+	tbit.nz.unc p_y, p0 = cnt, 2	/* should we terminate with a st4 ? */
 (p_scr)	br.cond.dpnt.few .restore_and_exit
 ;; }
 { .mib
 (p_y)	st4	[ptr1] = r0,4
-	tbit.nz.unc p_yy, p0 = cnt, 1	// should we terminate with a st2 ?
+	tbit.nz.unc p_yy, p0 = cnt, 1	/* should we terminate with a st2 ? */
 ;; }
 { .mib
 (p_yy)	st2	[ptr1] = r0,2
-	tbit.nz.unc p_y, p0 = cnt, 0	// should we terminate with a st1 ?
+	tbit.nz.unc p_y, p0 = cnt, 0	/* should we terminate with a st1 ? */
 ;; }
 
 { .mib
@@ -281,38 +281,38 @@ ENTRY(bzero)
 (p_n)	add	ptr2 = 2, ptr1
 } { .mmi
 (p_y)	add	ptr2 = 3, ptr1
-(p_y)	st1	[ptr1] = r0, 1		// fill 1 (odd-aligned) byte
-(p_y)	add	cnt = -1, cnt		// [15, 14 (or less) left]
+(p_y)	st1	[ptr1] = r0, 1		/* fill 1 (odd-aligned) byte */
+(p_y)	add	cnt = -1, cnt		/* [15, 14 (or less) left] */
 ;; }
 { .mmi
 (p_yy)	cmp.le.unc p_y, p0 = 8, cnt
-	add	ptr3 = ptr1, cnt	// prepare last store
+	add	ptr3 = ptr1, cnt	/* prepare last store */
 	movi0	ar.lc = save_lc
 } { .mmi
-(p_yy)	st2	[ptr1] = r0, 4		// fill 2 (aligned) bytes
-(p_yy)	st2	[ptr2] = r0, 4		// fill 2 (aligned) bytes
-(p_yy)	add	cnt = -4, cnt		// [11, 10 (o less) left]
+(p_yy)	st2	[ptr1] = r0, 4		/* fill 2 (aligned) bytes */
+(p_yy)	st2	[ptr2] = r0, 4		/* fill 2 (aligned) bytes */
+(p_yy)	add	cnt = -4, cnt		/* [11, 10 (o less) left] */
 ;; }
 { .mmi
 (p_y)	cmp.le.unc p_yy, p0 = 8, cnt
-	add	ptr3 = -1, ptr3		// last store
-	tbit.nz p_scr, p0 = cnt, 1	// will there be a st2 at the end ?
+	add	ptr3 = -1, ptr3		/* last store */
+	tbit.nz p_scr, p0 = cnt, 1	/* will there be a st2 at the end ? */
 } { .mmi
-(p_y)	st2	[ptr1] = r0, 4		// fill 2 (aligned) bytes
-(p_y)	st2	[ptr2] = r0, 4		// fill 2 (aligned) bytes
-(p_y)	add	cnt = -4, cnt		// [7, 6 (or less) left]
+(p_y)	st2	[ptr1] = r0, 4		/* fill 2 (aligned) bytes */
+(p_y)	st2	[ptr2] = r0, 4		/* fill 2 (aligned) bytes */
+(p_y)	add	cnt = -4, cnt		/* [7, 6 (or less) left] */
 ;; }
 { .mmi
-(p_yy)	st2	[ptr1] = r0, 4		// fill 2 (aligned) bytes
-(p_yy)	st2	[ptr2] = r0, 4		// fill 2 (aligned) bytes
-					// [3, 2 (or less) left]
-	tbit.nz p_y, p0 = cnt, 0	// will there be a st1 at the end ?
+(p_yy)	st2	[ptr1] = r0, 4		/* fill 2 (aligned) bytes */
+(p_yy)	st2	[ptr2] = r0, 4		/* fill 2 (aligned) bytes */
+					/* [3, 2 (or less) left] */
+	tbit.nz p_y, p0 = cnt, 0	/* will there be a st1 at the end ? */
 } { .mmi
 (p_yy)	add	cnt = -4, cnt
 ;; }
 { .mmb
-(p_scr)	st2	[ptr1] = r0		// fill 2 (aligned) bytes
-(p_y)	st1	[ptr3] = r0		// fill last byte (using ptr3)
+(p_scr)	st2	[ptr1] = r0		/* fill 2 (aligned) bytes */
+(p_y)	st1	[ptr3] = r0		/* fill last byte (using ptr3) */
 	br.ret.sptk.many rp
 ;; }
 END(bzero)
diff --git a/libc/string/ia64/memccpy.S b/libc/string/ia64/memccpy.S
index 1afba3637..259d680bc 100644
--- a/libc/string/ia64/memccpy.S
+++ b/libc/string/ia64/memccpy.S
@@ -23,7 +23,7 @@
    Inputs:
         in0:    dest
         in1:    src
-  	in2:	char
+	in2:	char
         in3:    byte count
 
    This implementation assumes little endian mode (UM.be = 0).
@@ -69,75 +69,75 @@ ENTRY(memccpy)
 	.rotr	r[MEMLAT + 7], tmp1[4], tmp2[4], val[4], tmp3[2], pos0[2]
 	.rotp	p[MEMLAT + 6 + 1]
 
-	mov	ret0 = r0		// return NULL if no match
+	mov	ret0 = r0		/* return NULL if no match */
 	.save pr, saved_pr
-	mov	saved_pr = pr		// save the predicate registers
-	mov 	dest = in0		// dest
+	mov	saved_pr = pr		/* save the predicate registers */
+	mov 	dest = in0		/* dest */
 	.save ar.lc, saved_lc
-        mov 	saved_lc = ar.lc	// save the loop counter
-        mov 	saved_ec = ar.ec	// save the loop counter
+        mov 	saved_lc = ar.lc	/* save the loop counter */
+        mov 	saved_ec = ar.ec	/* save the loop counter */
 	.body
-	mov 	src = in1		// src
-	extr.u	char = in2, 0, 8	// char
-	mov	len = in3		// len
-	sub	tmp = r0, in0		// tmp = -dest
-	cmp.ne	p7, p0 = r0, r0		// clear p7
+	mov 	src = in1		/* src */
+	extr.u	char = in2, 0, 8	/* char */
+	mov	len = in3		/* len */
+	sub	tmp = r0, in0		/* tmp = -dest */
+	cmp.ne	p7, p0 = r0, r0		/* clear p7 */
 	;;
-	and	loopcnt = 7, tmp	// loopcnt = -dest % 8
-	cmp.ge	p6, p0 = OP_T_THRES, len	// is len <= OP_T_THRES
-	mov	ar.ec = 0		// ec not guaranteed zero on entry
-(p6)	br.cond.spnt	.cpyfew		// copy byte by byte
+	and	loopcnt = 7, tmp	/* loopcnt = -dest % 8 */
+	cmp.ge	p6, p0 = OP_T_THRES, len	/* is len <= OP_T_THRES */
+	mov	ar.ec = 0		/* ec not guaranteed zero on entry */
+(p6)	br.cond.spnt	.cpyfew		/* copy byte by byte */
 	;;
 	cmp.eq	p6, p0 = loopcnt, r0
 	mux1	charx8 = char, @brcst
 (p6)	br.cond.sptk .dest_aligned
-	sub	len = len, loopcnt	// len -= -dest % 8
-	adds	loopcnt = -1, loopcnt	// --loopcnt
+	sub	len = len, loopcnt	/* len -= -dest % 8 */
+	adds	loopcnt = -1, loopcnt	/* --loopcnt */
 	;;
 	mov	ar.lc = loopcnt
-.l1:					// copy -dest % 8 bytes
-	ld1	value = [src], 1	// value = *src++
+.l1:					/* copy -dest % 8 bytes */
+	ld1	value = [src], 1	/* value = *src++ */
 	;;
-	st1	[dest] = value, 1	// *dest++ = value
+	st1	[dest] = value, 1	/* *dest++ = value */
 	cmp.eq	p6, p0 = value, char
 (p6)	br.cond.spnt .foundit
 	br.cloop.dptk .l1
 .dest_aligned:
-	and	sh1 = 7, src 		// sh1 = src % 8
-	and	tmp = -8, len   	// tmp = len & -OPSIZ
-	and	asrc = -8, src		// asrc = src & -OPSIZ  -- align src
-	shr.u	loopcnt = len, 3	// loopcnt = len / 8
-	and	len = 7, len ;;		// len = len % 8
-	shl	sh1 = sh1, 3		// sh1 = 8 * (src % 8)
-	adds	loopcnt = -1, loopcnt	// --loopcnt
-	mov     pr.rot = 1 << 16 ;;	// set rotating predicates
-	sub	sh2 = 64, sh1		// sh2 = 64 - sh1
-	mov	ar.lc = loopcnt		// set LC
-	cmp.eq  p6, p0 = sh1, r0 	// is the src aligned?
+	and	sh1 = 7, src 		/* sh1 = src % 8 */
+	and	tmp = -8, len   	/* tmp = len & -OPSIZ */
+	and	asrc = -8, src		/* asrc = src & -OPSIZ  -- align src */
+	shr.u	loopcnt = len, 3	/* loopcnt = len / 8 */
+	and	len = 7, len ;;		/* len = len % 8 */
+	shl	sh1 = sh1, 3		/* sh1 = 8 * (src % 8) */
+	adds	loopcnt = -1, loopcnt	/* --loopcnt */
+	mov     pr.rot = 1 << 16 ;;	/* set rotating predicates */
+	sub	sh2 = 64, sh1		/* sh2 = 64 - sh1 */
+	mov	ar.lc = loopcnt		/* set LC */
+	cmp.eq  p6, p0 = sh1, r0 	/* is the src aligned? */
 (p6)    br.cond.sptk .src_aligned ;;
-	add	src = src, tmp		// src += len & -OPSIZ
-	mov	ar.ec = MEMLAT + 6 + 1 	// six more passes needed
-	ld8	r[1] = [asrc], 8 	// r[1] = w0
-	cmp.ne	p6, p0 = r0, r0	;;	// clear p6
+	add	src = src, tmp		/* src += len & -OPSIZ */
+	mov	ar.ec = MEMLAT + 6 + 1 	/* six more passes needed */
+	ld8	r[1] = [asrc], 8 	/* r[1] = w0 */
+	cmp.ne	p6, p0 = r0, r0	;;	/* clear p6 */
 	ALIGN(32)
 .l2:
-(p[0])		ld8.s	r[0] = [asrc], 8		// r[0] = w1
-(p[MEMLAT])	shr.u	tmp1[0] = r[1 + MEMLAT], sh1	// tmp1 = w0 >> sh1
-(p[MEMLAT])	shl	tmp2[0] = r[0 + MEMLAT], sh2  	// tmp2 = w1 << sh2
+(p[0])		ld8.s	r[0] = [asrc], 8		/* r[0] = w1 */
+(p[MEMLAT])	shr.u	tmp1[0] = r[1 + MEMLAT], sh1	/* tmp1 = w0 >> sh1 */
+(p[MEMLAT])	shl	tmp2[0] = r[0 + MEMLAT], sh2  	/* tmp2 = w1 << sh2 */
 (p[MEMLAT+4])	xor	tmp3[0] = val[1], charx8
 (p[MEMLAT+5])	czx1.r	pos0[0] = tmp3[1]
-(p[MEMLAT+6])	chk.s	r[6 + MEMLAT], .recovery1	// our data isn't
-							// valid - rollback!
+(p[MEMLAT+6])	chk.s	r[6 + MEMLAT], .recovery1	/* our data isn't */
+							/* valid - rollback! */
 (p[MEMLAT+6])	cmp.ne	p6, p0 = 8, pos0[1]
 (p6)		br.cond.spnt	.gotit
-(p[MEMLAT+6])	st8	[dest] = val[3], 8		// store val to dest
-(p[MEMLAT+3])	or	val[0] = tmp1[3], tmp2[3] 	// val = tmp1 | tmp2
+(p[MEMLAT+6])	st8	[dest] = val[3], 8		/* store val to dest */
+(p[MEMLAT+3])	or	val[0] = tmp1[3], tmp2[3] 	/* val = tmp1 | tmp2 */
 		br.ctop.sptk    .l2
 		br.cond.sptk .cpyfew
 
 .src_aligned:
-		cmp.ne  p6, p0 = r0, r0			// clear p6
-		mov     ar.ec = MEMLAT + 2 + 1 ;;	// set EC
+		cmp.ne  p6, p0 = r0, r0			/* clear p6 */
+		mov     ar.ec = MEMLAT + 2 + 1 ;;	/* set EC */
 .l3:
 (p[0])		ld8.s	r[0] = [src], 8
 (p[MEMLAT])	xor	tmp3[0] = r[MEMLAT], charx8
@@ -149,8 +149,8 @@ ENTRY(memccpy)
 (p[MEMLAT+2])	st8	[dest] = r[MEMLAT+2], 8
 		br.ctop.dptk .l3
 .cpyfew:
-	cmp.eq	p6, p0 = len, r0	// is len == 0 ?
-	adds	len = -1, len		// --len;
+	cmp.eq	p6, p0 = len, r0	/* is len == 0 ? */
+	adds	len = -1, len		/* --len; */
 (p6)	br.cond.spnt	.restore_and_exit ;;
 	mov	ar.lc = len
 .l4:
@@ -163,14 +163,14 @@ ENTRY(memccpy)
 .foundit:
 (p6)	mov	ret0 = dest
 .restore_and_exit:
-	mov     pr = saved_pr, -1    	// restore the predicate registers
-	mov 	ar.lc = saved_lc	// restore the loop counter
-	mov 	ar.ec = saved_ec ;;	// restore the epilog counter
+	mov     pr = saved_pr, -1    	/* restore the predicate registers */
+	mov 	ar.lc = saved_lc	/* restore the loop counter */
+	mov 	ar.ec = saved_ec ;;	/* restore the epilog counter */
 	br.ret.sptk.many b0
 .gotit:
 	.pred.rel "mutex" p6, p7
-(p6)	mov	value = val[3]		// if coming from l2
-(p7)	mov	value = r[MEMLAT+2]	// if coming from l3
+(p6)	mov	value = val[3]		/* if coming from l2 */
+(p7)	mov	value = r[MEMLAT+2]	/* if coming from l3 */
 	mov	ar.lc = pos0[1] ;;
 .l5:
 	extr.u	tmp = value, 0, 8 ;;
diff --git a/libc/string/ia64/memchr.S b/libc/string/ia64/memchr.S
index 2bf078fe6..0246b5997 100644
--- a/libc/string/ia64/memchr.S
+++ b/libc/string/ia64/memchr.S
@@ -62,18 +62,18 @@ ENTRY(__memchr)
 	.rotr	value[MEMLAT+1], addr[MEMLAT+3], aux[2], poschr[2]
 	.rotp	p[MEMLAT+3]
 	.save ar.lc, saved_lc
-        mov 	saved_lc = ar.lc 	// save the loop counter
+        mov 	saved_lc = ar.lc 	/* save the loop counter */
 	.save pr, saved_pr
-	mov	saved_pr = pr		// save the predicates
+	mov	saved_pr = pr		/* save the predicates */
 	.body
 	mov 	ret0 = str
-	and 	tmp = 7, str		// tmp = str % 8
-	cmp.ne	p7, p0 = r0, r0		// clear p7
-	extr.u	chr = in1, 0, 8		// chr = (unsigned char) in1
+	and 	tmp = 7, str		/* tmp = str % 8 */
+	cmp.ne	p7, p0 = r0, r0		/* clear p7 */
+	extr.u	chr = in1, 0, 8		/* chr = (unsigned char) in1 */
 	mov	len = in2
-	cmp.gtu	p6, p0 = 16, in2	// use a simple loop for short
-(p6)	br.cond.spnt .srchfew ;;	// searches
-	sub	loopcnt = 8, tmp	// loopcnt = 8 - tmp
+	cmp.gtu	p6, p0 = 16, in2	/* use a simple loop for short */
+(p6)	br.cond.spnt .srchfew ;;	/* searches */
+	sub	loopcnt = 8, tmp	/* loopcnt = 8 - tmp */
 	cmp.eq	p6, p0 = tmp, r0
 (p6)	br.cond.sptk	.str_aligned;;
 	sub	len = len, loopcnt
@@ -86,12 +86,12 @@ ENTRY(__memchr)
 (p6)	br.cond.spnt	.foundit
 	br.cloop.sptk	.l1 ;;
 .str_aligned:
-	cmp.ne	p6, p0 = r0, r0		// clear p6
-	shr.u	loopcnt = len, 3	// loopcnt = len / 8
-	and 	len = 7, len ;;		// remaining len = len & 7
+	cmp.ne	p6, p0 = r0, r0		/* clear p6 */
+	shr.u	loopcnt = len, 3	/* loopcnt = len / 8 */
+	and 	len = 7, len ;;		/* remaining len = len & 7 */
 	adds	loopcnt = -1, loopcnt
 	mov	ar.ec = MEMLAT + 3
-	mux1	chrx8 = chr, @brcst ;;	// get a word full of chr
+	mux1	chrx8 = chr, @brcst ;;	/* get a word full of chr */
 	mov	ar.lc = loopcnt
 	mov	pr.rot = 1 << 16 ;;
 .l2:
@@ -114,12 +114,12 @@ ENTRY(__memchr)
 (p6)	br.cond.dpnt	.foundit
 	br.cloop.sptk	.l3 ;;
 .notfound:
-	cmp.ne	p6, p0 = r0, r0	// clear p6 (p7 was already 0 when we got here)
-	mov	ret0 = r0 ;;	// return NULL
+	cmp.ne	p6, p0 = r0, r0	/* clear p6 (p7 was already 0 when we got here) */
+	mov	ret0 = r0 ;;	/* return NULL */
 .foundit:
 	.pred.rel "mutex" p6, p7
-(p6)	adds	ret0 = -1, ret0 		   // if we got here from l1 or l3
-(p7)	add	ret0 = addr[MEMLAT+2], poschr[1]   // if we got here from l2
+(p6)	adds	ret0 = -1, ret0 		   /* if we got here from l1 or l3 */
+(p7)	add	ret0 = addr[MEMLAT+2], poschr[1]   /* if we got here from l2 */
 	mov	pr = saved_pr, -1
 	mov	ar.lc = saved_lc
 	br.ret.sptk.many b0
diff --git a/libc/string/ia64/memcmp.S b/libc/string/ia64/memcmp.S
index 8b0c096ce..adb1a20de 100644
--- a/libc/string/ia64/memcmp.S
+++ b/libc/string/ia64/memcmp.S
@@ -28,7 +28,7 @@
    In this form, it assumes little endian mode.  For big endian mode, the
    the two shifts in .l2 must be inverted:
 
-	shl   	tmp1[0] = r[1 + MEMLAT], sh1   // tmp1 = w0 << sh1
+	shl	tmp1[0] = r[1 + MEMLAT], sh1   // tmp1 = w0 << sh1
 	shr.u   tmp2[0] = r[0 + MEMLAT], sh2   // tmp2 = w1 >> sh2
 
    and all the mux1 instructions should be replaced by plain mov's.  */
@@ -36,8 +36,8 @@
 #include "sysdep.h"
 #undef ret
 
-#define OP_T_THRES 	16
-#define OPSIZ 		8
+#define OP_T_THRES	16
+#define OPSIZ		8
 #define MEMLAT		2
 
 #define start		r15
@@ -56,85 +56,85 @@
 
 ENTRY(memcmp)
 	.prologue
-	alloc 	r2 = ar.pfs, 3, 37, 0, 40
+	alloc	r2 = ar.pfs, 3, 37, 0, 40
 
 	.rotr	r[MEMLAT + 2], q[MEMLAT + 5], tmp1[4], tmp2[4], val[2]
 	.rotp	p[MEMLAT + 4 + 1]
 
-	mov	ret0 = r0		// by default return value = 0
+	mov	ret0 = r0		/* by default return value = 0 */
 	.save pr, saved_pr
-	mov	saved_pr = pr		// save the predicate registers
+	mov	saved_pr = pr		/* save the predicate registers */
 	.save ar.lc, saved_lc
-        mov 	saved_lc = ar.lc	// save the loop counter
+        mov	saved_lc = ar.lc	/* save the loop counter */
 	.body
-	mov 	dest = in0		// dest
-	mov 	src = in1		// src
-	mov	len = in2		// len
-	sub	tmp = r0, in0		// tmp = -dest
+	mov	dest = in0		/* dest */
+	mov	src = in1		/* src */
+	mov	len = in2		/* len */
+	sub	tmp = r0, in0		/* tmp = -dest */
 	;;
-	and	loopcnt = 7, tmp		// loopcnt = -dest % 8
-	cmp.ge	p6, p0 = OP_T_THRES, len	// is len <= OP_T_THRES
-(p6)	br.cond.spnt	.cmpfew			// compare byte by byte
+	and	loopcnt = 7, tmp		/* loopcnt = -dest % 8 */
+	cmp.ge	p6, p0 = OP_T_THRES, len	/* is len <= OP_T_THRES */
+(p6)	br.cond.spnt	.cmpfew			/* compare byte by byte */
 	;;
 	cmp.eq	p6, p0 = loopcnt, r0
 (p6)	br.cond.sptk .dest_aligned
-	sub	len = len, loopcnt	// len -= -dest % 8
-	adds	loopcnt = -1, loopcnt	// --loopcnt
+	sub	len = len, loopcnt	/* len -= -dest % 8 */
+	adds	loopcnt = -1, loopcnt	/* --loopcnt */
 	;;
 	mov	ar.lc = loopcnt
-.l1:					// copy -dest % 8 bytes
-	ld1	value1 = [src], 1	// value = *src++
+.l1:					/* copy -dest % 8 bytes */
+	ld1	value1 = [src], 1	/* value = *src++ */
 	ld1	value2 = [dest], 1
 	;;
 	cmp.ne	p6, p0 = value1, value2
 (p6)	br.cond.spnt .done
 	br.cloop.dptk .l1
 .dest_aligned:
-	and	sh1 = 7, src 		// sh1 = src % 8
-	and	tmp = -8, len   	// tmp = len & -OPSIZ
-	and	asrc = -8, src		// asrc = src & -OPSIZ  -- align src
-	shr.u	loopcnt = len, 3	// loopcnt = len / 8
-	and	len = 7, len ;;		// len = len % 8
-	shl	sh1 = sh1, 3		// sh1 = 8 * (src % 8)
-	adds	loopcnt = -1, loopcnt	// --loopcnt
-	mov     pr.rot = 1 << 16 ;;	// set rotating predicates
-	sub	sh2 = 64, sh1		// sh2 = 64 - sh1
-	mov	ar.lc = loopcnt		// set LC
-	cmp.eq  p6, p0 = sh1, r0 	// is the src aligned?
+	and	sh1 = 7, src		/* sh1 = src % 8 */
+	and	tmp = -8, len		/* tmp = len & -OPSIZ */
+	and	asrc = -8, src		/* asrc = src & -OPSIZ  -- align src */
+	shr.u	loopcnt = len, 3	/* loopcnt = len / 8 */
+	and	len = 7, len ;;		/* len = len % 8 */
+	shl	sh1 = sh1, 3		/* sh1 = 8 * (src % 8) */
+	adds	loopcnt = -1, loopcnt	/* --loopcnt */
+	mov     pr.rot = 1 << 16 ;;	/* set rotating predicates */
+	sub	sh2 = 64, sh1		/* sh2 = 64 - sh1 */
+	mov	ar.lc = loopcnt		/* set LC */
+	cmp.eq  p6, p0 = sh1, r0	/* is the src aligned? */
 (p6)    br.cond.sptk .src_aligned
-	add	src = src, tmp		// src += len & -OPSIZ
-	mov	ar.ec = MEMLAT + 4 + 1 	// four more passes needed
-	ld8	r[1] = [asrc], 8 ;;	// r[1] = w0
+	add	src = src, tmp		/* src += len & -OPSIZ */
+	mov	ar.ec = MEMLAT + 4 + 1	/* four more passes needed */
+	ld8	r[1] = [asrc], 8 ;;	/* r[1] = w0 */
 	.align	32
 
-// We enter this loop with p6 cleared by the above comparison
+/* We enter this loop with p6 cleared by the above comparison */
 
 .l2:
-(p[0])		ld8	r[0] = [asrc], 8		// r[0] = w1
+(p[0])		ld8	r[0] = [asrc], 8		/* r[0] = w1 */
 (p[0])		ld8	q[0] = [dest], 8
-(p[MEMLAT])	shr.u	tmp1[0] = r[1 + MEMLAT], sh1	// tmp1 = w0 >> sh1
-(p[MEMLAT])	shl	tmp2[0] = r[0 + MEMLAT], sh2  	// tmp2 = w1 << sh2
+(p[MEMLAT])	shr.u	tmp1[0] = r[1 + MEMLAT], sh1	/* tmp1 = w0 >> sh1 */
+(p[MEMLAT])	shl	tmp2[0] = r[0 + MEMLAT], sh2	/* tmp2 = w1 << sh2 */
 (p[MEMLAT+4])	cmp.ne	p6, p0 = q[MEMLAT + 4], val[1]
-(p[MEMLAT+3])	or	val[0] = tmp1[3], tmp2[3] 	// val = tmp1 | tmp2
+(p[MEMLAT+3])	or	val[0] = tmp1[3], tmp2[3]	/* val = tmp1 | tmp2 */
 (p6)		br.cond.spnt .l2exit
 		br.ctop.sptk    .l2
 		br.cond.sptk .cmpfew
 .l3exit:
 	mux1	value1 = r[MEMLAT], @rev
 	mux1	value2 = q[MEMLAT], @rev
-	cmp.ne	p6, p0 = r0, r0	;;	// clear p6
+	cmp.ne	p6, p0 = r0, r0	;;	/* clear p6 */
 .l2exit:
 (p6)	mux1	value1 = val[1], @rev
 (p6)	mux1	value2 = q[MEMLAT + 4], @rev ;;
 	cmp.ltu	p6, p7 = value2, value1 ;;
 (p6)	mov	ret0 = -1
 (p7)	mov	ret0 = 1
-	mov     pr = saved_pr, -1    	// restore the predicate registers
-	mov 	ar.lc = saved_lc	// restore the loop counter
+	mov     pr = saved_pr, -1	/* restore the predicate registers */
+	mov	ar.lc = saved_lc	/* restore the loop counter */
 	br.ret.sptk.many b0
 .src_aligned:
-	cmp.ne	p6, p0 = r0, r0		// clear p6
-	mov     ar.ec = MEMLAT + 1 ;;	// set EC
+	cmp.ne	p6, p0 = r0, r0		/* clear p6 */
+	mov     ar.ec = MEMLAT + 1 ;;	/* set EC */
 .l3:
 (p[0])		ld8	r[0] = [src], 8
 (p[0])		ld8	q[0] = [dest], 8
@@ -142,8 +142,8 @@ ENTRY(memcmp)
 (p6)		br.cond.spnt .l3exit
 		br.ctop.dptk .l3 ;;
 .cmpfew:
-	cmp.eq	p6, p0 = len, r0	// is len == 0 ?
-	adds	len = -1, len		// --len;
+	cmp.eq	p6, p0 = len, r0	/* is len == 0 ? */
+	adds	len = -1, len		/* --len; */
 (p6)	br.cond.spnt	.restore_and_exit ;;
 	mov	ar.lc = len
 .l4:
@@ -154,10 +154,10 @@ ENTRY(memcmp)
 (p6)	br.cond.spnt	.done
 	br.cloop.dptk	.l4 ;;
 .done:
-(p6)	sub	ret0 = value2, value1	// don't execute it if falling thru
+(p6)	sub	ret0 = value2, value1	/* don't execute it if falling thru */
 .restore_and_exit:
-	mov     pr = saved_pr, -1    	// restore the predicate registers
-	mov 	ar.lc = saved_lc	// restore the loop counter
+	mov     pr = saved_pr, -1	/* restore the predicate registers */
+	mov	ar.lc = saved_lc	/* restore the loop counter */
 	br.ret.sptk.many b0
 END(memcmp)
 libc_hidden_def (memcmp)
diff --git a/libc/string/ia64/memcpy.S b/libc/string/ia64/memcpy.S
index 810eb0c0e..6c48a72d9 100644
--- a/libc/string/ia64/memcpy.S
+++ b/libc/string/ia64/memcpy.S
@@ -42,8 +42,8 @@
 
 #define LFETCH_DIST     500
 
-#define ALIGN_UNROLL_no   4 // no. of elements
-#define ALIGN_UNROLL_sh	  2 // (shift amount)
+#define ALIGN_UNROLL_no   4 /* no. of elements */
+#define ALIGN_UNROLL_sh	  2 /* (shift amount) */
 
 #define MEMLAT	8
 #define Nrot	((4*(MEMLAT+2) + 7) & ~7)
@@ -168,76 +168,76 @@ ENTRY(memcpy)
 	.rotr	r[MEMLAT+1], s[MEMLAT+2], q[MEMLAT+1], t[MEMLAT+1]
 	.rotp	p[MEMLAT+2]
 	.rotf	fr[MEMLAT+1], fq[MEMLAT+1], fs[MEMLAT+1], ft[MEMLAT+1]
-	mov	ret0 = in0		// return tmp2 = dest
+	mov	ret0 = in0		/* return tmp2 = dest */
 	.save   pr, saved_pr
-	movi0	saved_pr = pr		// save the predicate registers
+	movi0	saved_pr = pr		/* save the predicate registers */
 } { .mmi
-	and	tmp4 = 7, in0 		// check if destination is aligned
-	mov 	dest = in0		// dest
-	mov 	src = in1		// src
+	and	tmp4 = 7, in0 		/* check if destination is aligned */
+	mov 	dest = in0		/* dest */
+	mov 	src = in1		/* src */
 ;; }
 { .mii
-	cmp.eq	p_scr, p0 = in2, r0	// if (len == 0)
+	cmp.eq	p_scr, p0 = in2, r0	/* if (len == 0) */
 	.save   ar.lc, saved_lc
-        movi0 	saved_lc = ar.lc	// save the loop counter
+        movi0 	saved_lc = ar.lc	/* save the loop counter */
 	.body
-	cmp.ge	p_few, p0 = OP_T_THRES, in2 // is len <= OP_T_THRESH
+	cmp.ge	p_few, p0 = OP_T_THRES, in2 /* is len <= OP_T_THRESH */
 } { .mbb
-	mov	len = in2		// len
-(p_scr)	br.cond.dpnt.few .restore_and_exit // 	Branch no. 1: return dest
-(p_few) br.cond.dpnt.many .copy_bytes	// Branch no. 2: copy byte by byte
+	mov	len = in2		/* len */
+(p_scr)	br.cond.dpnt.few .restore_and_exit /* 	Branch no. 1: return dest */
+(p_few) br.cond.dpnt.many .copy_bytes	/* Branch no. 2: copy byte by byte */
 ;; }
 { .mmi
 #if defined(USE_LFETCH)
-	lfetch.nt1 [dest]		//
-	lfetch.nt1 [src]		//
+	lfetch.nt1 [dest]		/* */
+	lfetch.nt1 [src]		/* */
 #endif
-	shr.u	elemcnt = len, 3	// elemcnt = len / 8
+	shr.u	elemcnt = len, 3	/* elemcnt = len / 8 */
 } { .mib
-	cmp.eq	p_scr, p0 = tmp4, r0	// is destination aligned?
-	sub	loopcnt = 7, tmp4	//
+	cmp.eq	p_scr, p0 = tmp4, r0	/* is destination aligned? */
+	sub	loopcnt = 7, tmp4	/* */
 (p_scr) br.cond.dptk.many .dest_aligned
 ;; }
 { .mmi
-	ld1	tmp2 = [src], 1		//
-	sub	len = len, loopcnt, 1	// reduce len
-	movi0	ar.lc = loopcnt		//
+	ld1	tmp2 = [src], 1		/* */
+	sub	len = len, loopcnt, 1	/* reduce len */
+	movi0	ar.lc = loopcnt		/* */
 } { .mib
-	cmp.ne  p_scr, p0 = 0, loopcnt	// avoid loading beyond end-point
+	cmp.ne  p_scr, p0 = 0, loopcnt	/* avoid loading beyond end-point */
 ;; }
 
-.l0:	// ---------------------------- // L0: Align src on 8-byte boundary
+.l0:	/* ---------------------------- L0: Align src on 8-byte boundary */
 { .mmi
-	st1	[dest] = tmp2, 1	//
-(p_scr)	ld1	tmp2 = [src], 1		//
+	st1	[dest] = tmp2, 1	/* */
+(p_scr)	ld1	tmp2 = [src], 1		/* */
 } { .mib
-	cmp.lt	p_scr, p0 = 1, loopcnt	// avoid load beyond end-point
+	cmp.lt	p_scr, p0 = 1, loopcnt	/* avoid load beyond end-point */
 	add	loopcnt = -1, loopcnt
-	br.cloop.dptk.few .l0		//
+	br.cloop.dptk.few .l0		/* */
 ;; }
 
 .dest_aligned:
 { .mmi
-	and	tmp4 = 7, src		// ready for alignment check
-	shr.u	elemcnt = len, 3	// elemcnt = len / 8
+	and	tmp4 = 7, src		/* ready for alignment check */
+	shr.u	elemcnt = len, 3	/* elemcnt = len / 8 */
 ;; }
 { .mib
-	cmp.ne	p_scr, p0 = tmp4, r0	// is source also aligned
-	tbit.nz p_xtr, p_nxtr = src, 3	// prepare a separate move if src
-} { .mib				// is not 16B aligned
-	add	ptr2 = LFETCH_DIST, dest	// prefetch address
+	cmp.ne	p_scr, p0 = tmp4, r0	/* is source also aligned */
+	tbit.nz p_xtr, p_nxtr = src, 3	/* prepare a separate move if src */
+} { .mib				/* is not 16B aligned */
+	add	ptr2 = LFETCH_DIST, dest	/* prefetch address */
 	add	ptr1 = LFETCH_DIST, src
 (p_scr) br.cond.dptk.many .src_not_aligned
 ;; }
 
-// The optimal case, when dest, and src are aligned
+/* The optimal case, when dest, and src are aligned */
 
 .both_aligned:
 { .mmi
 	.pred.rel "mutex",p_xtr,p_nxtr
-(p_xtr)	cmp.gt  p_scr, p0 = ALIGN_UNROLL_no+1, elemcnt // Need N + 1 to qualify
-(p_nxtr) cmp.gt p_scr, p0 = ALIGN_UNROLL_no, elemcnt  // Need only N to qualify
-	movi0	pr.rot = 1 << 16	// set rotating predicates
+(p_xtr)	cmp.gt  p_scr, p0 = ALIGN_UNROLL_no+1, elemcnt /* Need N + 1 to qualify */
+(p_nxtr) cmp.gt p_scr, p0 = ALIGN_UNROLL_no, elemcnt  /* Need only N to qualify */
+	movi0	pr.rot = 1 << 16	/* set rotating predicates */
 } { .mib
 (p_scr) br.cond.dpnt.many .copy_full_words
 ;; }
@@ -245,21 +245,21 @@ ENTRY(memcpy)
 { .mmi
 (p_xtr)	load	tempreg = [src], 8
 (p_xtr) add 	elemcnt = -1, elemcnt
-	movi0	ar.ec = MEMLAT + 1	// set the epilog counter
+	movi0	ar.ec = MEMLAT + 1	/* set the epilog counter */
 ;; }
 { .mmi
-(p_xtr) add	len = -8, len		//
-	add 	asrc = 16, src 		// one bank apart (for USE_INT)
-	shr.u	loopcnt = elemcnt, ALIGN_UNROLL_sh  // cater for unrolling
+(p_xtr) add	len = -8, len		/* */
+	add 	asrc = 16, src 		/* one bank apart (for USE_INT) */
+	shr.u	loopcnt = elemcnt, ALIGN_UNROLL_sh  /* cater for unrolling */
 ;;}
 { .mmi
 	add	loopcnt = -1, loopcnt
-(p_xtr)	store	[dest] = tempreg, 8	// copy the "extra" word
+(p_xtr)	store	[dest] = tempreg, 8	/* copy the "extra" word */
 	nop.i	0
 ;; }
 { .mib
 	add	adest = 16, dest
-	movi0	ar.lc = loopcnt 	// set the loop counter
+	movi0	ar.lc = loopcnt 	/* set the loop counter */
 ;; }
 
 #ifdef  GAS_ALIGN_BREAKS_UNWIND_INFO
@@ -268,7 +268,7 @@ ENTRY(memcpy)
 	.align	32
 #endif
 #if defined(USE_FLP)
-.l1: // ------------------------------- // L1: Everything a multiple of 8
+.l1: /* ------------------------------- L1: Everything a multiple of 8 */
 { .mmi
 #if defined(USE_LFETCH)
 (p[0])	lfetch.nt1 [ptr2],32
@@ -290,7 +290,7 @@ ENTRY(memcpy)
 	br.ctop.dptk.many .l1
 ;; }
 #elif defined(USE_INT)
-.l1: // ------------------------------- // L1: Everything a multiple of 8
+.l1: /* ------------------------------- L1: Everything a multiple of 8 */
 { .mmi
 (p[0])	load	the_r[0] = [src], 8
 (p[0])	load	the_q[0] = [asrc], 8
@@ -317,58 +317,58 @@ ENTRY(memcpy)
 
 .copy_full_words:
 { .mib
-	cmp.gt	p_scr, p0 = 8, len	//
-	shr.u	elemcnt = len, 3	//
+	cmp.gt	p_scr, p0 = 8, len	/* */
+	shr.u	elemcnt = len, 3	/* */
 (p_scr) br.cond.dpnt.many .copy_bytes
 ;; }
 { .mii
 	load	tempreg = [src], 8
-	add	loopcnt = -1, elemcnt	//
+	add	loopcnt = -1, elemcnt	/* */
 ;; }
 { .mii
-	cmp.ne	p_scr, p0 = 0, loopcnt	//
-	mov	ar.lc = loopcnt		//
+	cmp.ne	p_scr, p0 = 0, loopcnt	/* */
+	mov	ar.lc = loopcnt		/* */
 ;; }
 
-.l2: // ------------------------------- // L2: Max 4 words copied separately
+.l2: /* ------------------------------- L2: Max 4 words copied separately */
 { .mmi
 	store	[dest] = tempreg, 8
-(p_scr)	load	tempreg = [src], 8	//
+(p_scr)	load	tempreg = [src], 8	/* */
 	add	len = -8, len
 } { .mib
-	cmp.lt	p_scr, p0 = 1, loopcnt	// avoid load beyond end-point
+	cmp.lt	p_scr, p0 = 1, loopcnt	/* avoid load beyond end-point */
 	add	loopcnt = -1, loopcnt
 	br.cloop.dptk.few  .l2
 ;; }
 
 .copy_bytes:
 { .mib
-	cmp.eq	p_scr, p0 = len, r0	// is len == 0 ?
-	add	loopcnt = -1, len	// len--;
+	cmp.eq	p_scr, p0 = len, r0	/* is len == 0 ? */
+	add	loopcnt = -1, len	/* len--; */
 (p_scr)	br.cond.spnt	.restore_and_exit
 ;; }
 { .mii
 	ld1	tmp2 = [src], 1
 	movi0	ar.lc = loopcnt
-	cmp.ne	p_scr, p0 = 0, loopcnt	// avoid load beyond end-point
+	cmp.ne	p_scr, p0 = 0, loopcnt	/* avoid load beyond end-point */
 ;; }
 
-.l3: // ------------------------------- // L3: Final byte move
+.l3: /* ------------------------------- L3: Final byte move */
 { .mmi
 	st1	[dest] = tmp2, 1
 (p_scr)	ld1	tmp2 = [src], 1
 } { .mib
-	cmp.lt	p_scr, p0 = 1, loopcnt	// avoid load beyond end-point
+	cmp.lt	p_scr, p0 = 1, loopcnt	/* avoid load beyond end-point */
 	add	loopcnt = -1, loopcnt
 	br.cloop.dptk.few  .l3
 ;; }
 
 .restore_and_exit:
 { .mmi
-	movi0	pr = saved_pr, -1	// restore the predicate registers
+	movi0	pr = saved_pr, -1	/* restore the predicate registers */
 ;; }
 { .mib
-	movi0	ar.lc = saved_lc	// restore the loop counter
+	movi0	ar.lc = saved_lc	/* restore the loop counter */
 	br.ret.sptk.many b0
 ;; }
 
@@ -376,41 +376,41 @@ ENTRY(memcpy)
 .src_not_aligned:
 { .mmi
 	cmp.gt	p_scr, p0 = 16, len
-	and	sh1 = 7, src 		// sh1 = src % 8
-	shr.u	loopcnt = len, 4	// element-cnt = len / 16
+	and	sh1 = 7, src 		/* sh1 = src % 8 */
+	shr.u	loopcnt = len, 4	/* element-cnt = len / 16 */
 } { .mib
 	add	tmp4 = @ltoff(.table), gp
 	add 	tmp3 = @ltoff(.loop56), gp
-(p_scr)	br.cond.dpnt.many .copy_bytes	// do byte by byte if too few
+(p_scr)	br.cond.dpnt.many .copy_bytes	/* do byte by byte if too few */
 ;; }
 { .mmi
-	and	asrc = -8, src		// asrc = (-8) -- align src for loop
-	add 	loopcnt = -1, loopcnt	// loopcnt--
-	shl	sh1 = sh1, 3		// sh1 = 8 * (src % 8)
+	and	asrc = -8, src		/* asrc = (-8) -- align src for loop */
+	add 	loopcnt = -1, loopcnt	/* loopcnt-- */
+	shl	sh1 = sh1, 3		/* sh1 = 8 * (src % 8) */
 } { .mmi
-	ld8	ptable = [tmp4]		// ptable = &table
-	ld8	ploop56 = [tmp3]	// ploop56 = &loop56
-	and	tmp2 = -16, len		// tmp2 = len & -OPSIZ
+	ld8	ptable = [tmp4]		/* ptable = &table */
+	ld8	ploop56 = [tmp3]	/* ploop56 = &loop56 */
+	and	tmp2 = -16, len		/* tmp2 = len & -OPSIZ */
 ;; }
 { .mmi
-	add	tmp3 = ptable, sh1	// tmp3 = &table + sh1
-	add	src = src, tmp2		// src += len & (-16)
-	movi0	ar.lc = loopcnt		// set LC
+	add	tmp3 = ptable, sh1	/* tmp3 = &table + sh1 */
+	add	src = src, tmp2		/* src += len & (-16) */
+	movi0	ar.lc = loopcnt		/* set LC */
 ;; }
 { .mmi
-	ld8	tmp4 = [tmp3]		// tmp4 = loop offset
-	sub	len = len, tmp2		// len -= len & (-16)
-	movi0	ar.ec = MEMLAT + 2 	// one more pass needed
+	ld8	tmp4 = [tmp3]		/* tmp4 = loop offset */
+	sub	len = len, tmp2		/* len -= len & (-16) */
+	movi0	ar.ec = MEMLAT + 2 	/* one more pass needed */
 ;; }
 { .mmi
-	ld8	s[1] = [asrc], 8	// preload
-	sub	loopaddr = ploop56,tmp4	// loopadd = &loop56 - loop offset
-	movi0   pr.rot = 1 << 16	// set rotating predicates
+	ld8	s[1] = [asrc], 8	/* preload */
+	sub	loopaddr = ploop56,tmp4	/* loopadd = &loop56 - loop offset */
+	movi0   pr.rot = 1 << 16	/* set rotating predicates */
 ;; }
 { .mib
 	nop.m	0
 	movi0	b6 = loopaddr
-	br	b6			// jump to the appropriate loop
+	br	b6			/* jump to the appropriate loop */
 ;; }
 
 	LOOP(8)
@@ -426,7 +426,7 @@ libc_hidden_def (memcpy)
 	.rodata
 	.align 8
 .table:
-	data8	0			// dummy entry
+	data8	0			/* dummy entry */
 	data8 	.loop56 - .loop8
 	data8 	.loop56 - .loop16
 	data8 	.loop56 - .loop24
diff --git a/libc/string/ia64/memmove.S b/libc/string/ia64/memmove.S
index 00342d8e0..beaada6fc 100644
--- a/libc/string/ia64/memmove.S
+++ b/libc/string/ia64/memmove.S
@@ -81,48 +81,48 @@ ENTRY(memmove)
 	alloc 	r2 = ar.pfs, 3, Nrot - 3, 0, Nrot
 	.rotr	r[MEMLAT + 2], q[MEMLAT + 1]
 	.rotp	p[MEMLAT + 2]
-	mov	ret0 = in0		// return value = dest
+	mov	ret0 = in0		/* return value = dest */
 	.save pr, saved_pr
-	mov	saved_pr = pr		// save the predicate registers
+	mov	saved_pr = pr		/* save the predicate registers */
 	.save ar.lc, saved_lc
-        mov 	saved_lc = ar.lc	// save the loop counter
+        mov 	saved_lc = ar.lc	/* save the loop counter */
 	.body
-	or	tmp3 = in0, in1 ;;	// tmp3 = dest | src
-	or	tmp3 = tmp3, in2	// tmp3 = dest | src | len
-	mov 	dest = in0		// dest
-	mov 	src = in1		// src
-	mov	len = in2		// len
-	sub	tmp2 = r0, in0		// tmp2 = -dest
-	cmp.eq	p6, p0 = in2, r0	// if (len == 0)
-(p6)	br.cond.spnt .restore_and_exit;;// 	return dest;
-	and	tmp4 = 7, tmp3 		// tmp4 = (dest | src | len) & 7
-	cmp.le	p6, p0 = dest, src	// if dest <= src it's always safe
-(p6)	br.cond.spnt .forward		// to copy forward
+	or	tmp3 = in0, in1 ;;	/* tmp3 = dest | src */
+	or	tmp3 = tmp3, in2	/* tmp3 = dest | src | len */
+	mov 	dest = in0		/* dest */
+	mov 	src = in1		/* src */
+	mov	len = in2		/* len */
+	sub	tmp2 = r0, in0		/* tmp2 = -dest */
+	cmp.eq	p6, p0 = in2, r0	/* if (len == 0) */
+(p6)	br.cond.spnt .restore_and_exit;;/* 	return dest; */
+	and	tmp4 = 7, tmp3 		/* tmp4 = (dest | src | len) & 7 */
+	cmp.le	p6, p0 = dest, src	/* if dest <= src it's always safe */
+(p6)	br.cond.spnt .forward		/* to copy forward */
 	add	tmp3 = src, len;;
-	cmp.lt	p6, p0 = dest, tmp3	// if dest > src && dest < src + len
-(p6)	br.cond.spnt .backward		// we have to copy backward
+	cmp.lt	p6, p0 = dest, tmp3	/* if dest > src && dest < src + len */
+(p6)	br.cond.spnt .backward		/* we have to copy backward */
 
 .forward:
-	shr.u	loopcnt = len, 4 ;;	// loopcnt = len / 16
-	cmp.ne	p6, p0 = tmp4, r0	// if ((dest | src | len) & 7 != 0)
-(p6)	br.cond.sptk .next		//	goto next;
+	shr.u	loopcnt = len, 4 ;;	/* loopcnt = len / 16 */
+	cmp.ne	p6, p0 = tmp4, r0	/* if ((dest | src | len) & 7 != 0) */
+(p6)	br.cond.sptk .next		/*	goto next; */
 
-// The optimal case, when dest, src and len are all multiples of 8
+/* The optimal case, when dest, src and len are all multiples of 8 */
 
 	and	tmp3 = 0xf, len
-	mov	pr.rot = 1 << 16	// set rotating predicates
-	mov	ar.ec = MEMLAT + 1 ;;	// set the epilog counter
-	cmp.ne	p6, p0 = tmp3, r0	// do we have to copy an extra word?
-	adds	loopcnt = -1, loopcnt;;	// --loopcnt
+	mov	pr.rot = 1 << 16	/* set rotating predicates */
+	mov	ar.ec = MEMLAT + 1 ;;	/* set the epilog counter */
+	cmp.ne	p6, p0 = tmp3, r0	/* do we have to copy an extra word? */
+	adds	loopcnt = -1, loopcnt;;	/* --loopcnt */
 (p6)	ld8	value = [src], 8;;
-(p6)	st8	[dest] = value, 8	// copy the "odd" word
-	mov	ar.lc = loopcnt 	// set the loop counter
+(p6)	st8	[dest] = value, 8	/* copy the "odd" word */
+	mov	ar.lc = loopcnt 	/* set the loop counter */
 	cmp.eq	p6, p0 = 8, len
-(p6)	br.cond.spnt .restore_and_exit;;// the one-word special case
-	adds	adest = 8, dest		// set adest one word ahead of dest
-	adds	asrc = 8, src ;;	// set asrc one word ahead of src
-	nop.b	0			// get the "golden" alignment for
-	nop.b	0			// the next loop
+(p6)	br.cond.spnt .restore_and_exit;;/* the one-word special case */
+	adds	adest = 8, dest		/* set adest one word ahead of dest */
+	adds	asrc = 8, src ;;	/* set asrc one word ahead of src */
+	nop.b	0			/* get the "golden" alignment for */
+	nop.b	0			/* the next loop */
 .l0:
 (p[0])		ld8	r[0] = [src], 16
 (p[0])		ld8	q[0] = [asrc], 16
@@ -130,50 +130,50 @@ ENTRY(memmove)
 (p[MEMLAT])	st8	[adest] = q[MEMLAT], 16
 		br.ctop.dptk .l0 ;;
 
-	mov	pr = saved_pr, -1	// restore the predicate registers
-	mov	ar.lc = saved_lc	// restore the loop counter
+	mov	pr = saved_pr, -1	/* restore the predicate registers */
+	mov	ar.lc = saved_lc	/* restore the loop counter */
 	br.ret.sptk.many b0
 .next:
-	cmp.ge	p6, p0 = OP_T_THRES, len	// is len <= OP_T_THRES
-	and	loopcnt = 7, tmp2 		// loopcnt = -dest % 8
-(p6)	br.cond.spnt	.cpyfew			// copy byte by byte
+	cmp.ge	p6, p0 = OP_T_THRES, len	/* is len <= OP_T_THRES */
+	and	loopcnt = 7, tmp2 		/* loopcnt = -dest % 8 */
+(p6)	br.cond.spnt	.cpyfew			/* copy byte by byte */
 	;;
 	cmp.eq	p6, p0 = loopcnt, r0
 (p6)	br.cond.sptk	.dest_aligned
-	sub	len = len, loopcnt	// len -= -dest % 8
-	adds	loopcnt = -1, loopcnt	// --loopcnt
+	sub	len = len, loopcnt	/* len -= -dest % 8 */
+	adds	loopcnt = -1, loopcnt	/* --loopcnt */
 	;;
 	mov	ar.lc = loopcnt
-.l1:					// copy -dest % 8 bytes
-	ld1	value = [src], 1	// value = *src++
+.l1:					/* copy -dest % 8 bytes */
+	ld1	value = [src], 1	/* value = *src++ */
 	;;
-	st1	[dest] = value, 1	// *dest++ = value
+	st1	[dest] = value, 1	/* *dest++ = value */
 	br.cloop.dptk .l1
 .dest_aligned:
-	and	sh1 = 7, src 		// sh1 = src % 8
-	and	tmp2 = -8, len   	// tmp2 = len & -OPSIZ
-	and	asrc = -8, src		// asrc = src & -OPSIZ  -- align src
-	shr.u	loopcnt = len, 3	// loopcnt = len / 8
-	and	len = 7, len;;		// len = len % 8
-	adds	loopcnt = -1, loopcnt	// --loopcnt
+	and	sh1 = 7, src 		/* sh1 = src % 8 */
+	and	tmp2 = -8, len   	/* tmp2 = len & -OPSIZ */
+	and	asrc = -8, src		/* asrc = src & -OPSIZ  -- align src */
+	shr.u	loopcnt = len, 3	/* loopcnt = len / 8 */
+	and	len = 7, len;;		/* len = len % 8 */
+	adds	loopcnt = -1, loopcnt	/* --loopcnt */
 	addl	tmp4 = @ltoff(.table), gp
 	addl	tmp3 = @ltoff(.loop56), gp
-	mov     ar.ec = MEMLAT + 1	// set EC
-	mov     pr.rot = 1 << 16;;	// set rotating predicates
-	mov	ar.lc = loopcnt		// set LC
-	cmp.eq  p6, p0 = sh1, r0 	// is the src aligned?
+	mov     ar.ec = MEMLAT + 1	/* set EC */
+	mov     pr.rot = 1 << 16;;	/* set rotating predicates */
+	mov	ar.lc = loopcnt		/* set LC */
+	cmp.eq  p6, p0 = sh1, r0 	/* is the src aligned? */
 (p6)    br.cond.sptk .src_aligned
-	add	src = src, tmp2		// src += len & -OPSIZ
-	shl	sh1 = sh1, 3		// sh1 = 8 * (src % 8)
-	ld8	ploop56 = [tmp3]	// ploop56 = &loop56
-	ld8	ptable = [tmp4];;	// ptable = &table
-	add	tmp3 = ptable, sh1;;	// tmp3 = &table + sh1
-	mov	ar.ec = MEMLAT + 1 + 1 // one more pass needed
-	ld8	tmp4 = [tmp3];;		// tmp4 = loop offset
-	sub	loopaddr = ploop56,tmp4	// loopadd = &loop56 - loop offset
-	ld8	r[1] = [asrc], 8;;	// w0
+	add	src = src, tmp2		/* src += len & -OPSIZ */
+	shl	sh1 = sh1, 3		/* sh1 = 8 * (src % 8) */
+	ld8	ploop56 = [tmp3]	/* ploop56 = &loop56 */
+	ld8	ptable = [tmp4];;	/* ptable = &table */
+	add	tmp3 = ptable, sh1;;	/* tmp3 = &table + sh1 */
+	mov	ar.ec = MEMLAT + 1 + 1 /* one more pass needed */
+	ld8	tmp4 = [tmp3];;		/* tmp4 = loop offset */
+	sub	loopaddr = ploop56,tmp4	/* loopadd = &loop56 - loop offset */
+	ld8	r[1] = [asrc], 8;;	/* w0 */
 	mov	b6 = loopaddr;;
-	br	b6			// jump to the appropriate loop
+	br	b6			/* jump to the appropriate loop */
 
 	LOOP(8)
 	LOOP(16)
@@ -189,8 +189,8 @@ ENTRY(memmove)
 (p[MEMLAT])	st8	[dest] = r[MEMLAT], 8
 		br.ctop.dptk .l3
 .cpyfew:
-	cmp.eq	p6, p0 = len, r0	// is len == 0 ?
-	adds	len = -1, len		// --len;
+	cmp.eq	p6, p0 = len, r0	/* is len == 0 ? */
+	adds	len = -1, len		/* --len; */
 (p6)	br.cond.spnt	.restore_and_exit ;;
 	mov	ar.lc = len
 .l4:
@@ -199,36 +199,36 @@ ENTRY(memmove)
 	st1	[dest] = value, 1
 	br.cloop.dptk	.l4 ;;
 .restore_and_exit:
-	mov     pr = saved_pr, -1    	// restore the predicate registers
-	mov 	ar.lc = saved_lc	// restore the loop counter
+	mov     pr = saved_pr, -1    	/* restore the predicate registers */
+	mov 	ar.lc = saved_lc	/* restore the loop counter */
 	br.ret.sptk.many b0
 
-// In the case of a backward copy, optimise only the case when everything
-// is a multiple of 8, otherwise copy byte by byte.  The backward copy is
-// used only when the blocks are overlapping and dest > src.
-
+/* In the case of a backward copy, optimise only the case when everything
+   is a multiple of 8, otherwise copy byte by byte.  The backward copy is
+   used only when the blocks are overlapping and dest > src.
+*/
 .backward:
-	shr.u	loopcnt = len, 3	// loopcnt = len / 8
-	add	src = src, len		// src points one byte past the end
-	add	dest = dest, len ;; 	// dest points one byte past the end
-	mov	ar.ec = MEMLAT + 1	// set the epilog counter
-	mov	pr.rot = 1 << 16	// set rotating predicates
-	adds	loopcnt = -1, loopcnt	// --loopcnt
-	cmp.ne	p6, p0 = tmp4, r0	// if ((dest | src | len) & 7 != 0)
-(p6)	br.cond.sptk .bytecopy ;;	// copy byte by byte backward
-	adds	src = -8, src		// src points to the last word
-	adds	dest = -8, dest 	// dest points to the last word
-	mov	ar.lc = loopcnt;;	// set the loop counter
+	shr.u	loopcnt = len, 3	/* loopcnt = len / 8 */
+	add	src = src, len		/* src points one byte past the end */
+	add	dest = dest, len ;; 	/* dest points one byte past the end */
+	mov	ar.ec = MEMLAT + 1	/* set the epilog counter */
+	mov	pr.rot = 1 << 16	/* set rotating predicates */
+	adds	loopcnt = -1, loopcnt	/* --loopcnt */
+	cmp.ne	p6, p0 = tmp4, r0	/* if ((dest | src | len) & 7 != 0) */
+(p6)	br.cond.sptk .bytecopy ;;	/* copy byte by byte backward */
+	adds	src = -8, src		/* src points to the last word */
+	adds	dest = -8, dest 	/* dest points to the last word */
+	mov	ar.lc = loopcnt;;	/* set the loop counter */
 .l5:
 (p[0])		ld8	r[0] = [src], -8
 (p[MEMLAT])	st8	[dest] = r[MEMLAT], -8
 		br.ctop.dptk .l5
 		br.cond.sptk .restore_and_exit
 .bytecopy:
-	adds	src = -1, src		// src points to the last byte
-	adds	dest = -1, dest		// dest points to the last byte
-	adds	loopcnt = -1, len;;	// loopcnt = len - 1
-	mov	ar.lc = loopcnt;;	// set the loop counter
+	adds	src = -1, src		/* src points to the last byte */
+	adds	dest = -1, dest		/* dest points to the last byte */
+	adds	loopcnt = -1, len;;	/* loopcnt = len - 1 */
+	mov	ar.lc = loopcnt;;	/* set the loop counter */
 .l6:
 (p[0])		ld1	r[0] = [src], -1
 (p[MEMLAT])	st1	[dest] = r[MEMLAT], -1
@@ -239,7 +239,7 @@ END(memmove)
 	.rodata
 	.align 8
 .table:
-	data8	0			// dummy entry
+	data8	0			/* dummy entry */
 	data8 	.loop56 - .loop8
 	data8 	.loop56 - .loop16
 	data8 	.loop56 - .loop24
diff --git a/libc/string/ia64/memset.S b/libc/string/ia64/memset.S
index ed27f3f31..45df5838e 100644
--- a/libc/string/ia64/memset.S
+++ b/libc/string/ia64/memset.S
@@ -46,15 +46,15 @@
 #define ptr1		r28
 #define ptr2		r27
 #define ptr3		r26
-#define ptr9 		r24
+#define ptr9		r24
 #define	loopcnt		r23
 #define linecnt		r22
 #define bytecnt		r21
 
 #define fvalue		f6
 
-// This routine uses only scratch predicate registers (p6 - p15)
-#define p_scr		p6			// default register for same-cycle branches
+/* This routine uses only scratch predicate registers (p6 - p15) */
+#define p_scr		p6	/* default register for same-cycle branches */
 #define p_nz		p7
 #define p_zr		p8
 #define p_unalgn	p9
@@ -68,7 +68,7 @@
 #define MIN1		15
 #define MIN1P1HALF	8
 #define LINE_SIZE	128
-#define LSIZE_SH        7			// shift amount
+#define LSIZE_SH        7			/* shift amount */
 #define PREF_AHEAD	8
 
 #define USE_FLP
@@ -90,97 +90,97 @@ ENTRY(memset)
 	movi0	save_lc = ar.lc
 } { .mmi
 	.body
-	mov	ret0 = dest		// return value
-	cmp.ne	p_nz, p_zr = value, r0	// use stf.spill if value is zero
+	mov	ret0 = dest		/* return value */
+	cmp.ne	p_nz, p_zr = value, r0	/* use stf.spill if value is zero */
 	cmp.eq	p_scr, p0 = cnt, r0
 ;; }
 { .mmi
-	and	ptr2 = -(MIN1+1), dest	// aligned address
-	and	tmp = MIN1, dest	// prepare to check for alignment
-	tbit.nz p_y, p_n = dest, 0	// Do we have an odd address? (M_B_U)
+	and	ptr2 = -(MIN1+1), dest	/* aligned address */
+	and	tmp = MIN1, dest	/* prepare to check for alignment */
+	tbit.nz p_y, p_n = dest, 0	/* Do we have an odd address? (M_B_U) */
 } { .mib
 	mov	ptr1 = dest
-	mux1	value = value, @brcst	// create 8 identical bytes in word
-(p_scr)	br.ret.dpnt.many rp		// return immediately if count = 0
+	mux1	value = value, @brcst	/* create 8 identical bytes in word */
+(p_scr)	br.ret.dpnt.many rp		/* return immediately if count = 0 */
 ;; }
 { .mib
 	cmp.ne	p_unalgn, p0 = tmp, r0
-} { .mib				// NB: # of bytes to move is 1 higher
-	sub	bytecnt = (MIN1+1), tmp	//     than loopcnt
-	cmp.gt	p_scr, p0 = 16, cnt		// is it a minimalistic task?
-(p_scr)	br.cond.dptk.many .move_bytes_unaligned	// go move just a few (M_B_U)
+} { .mib				/* NB: # of bytes to move is 1 higher */
+	sub	bytecnt = (MIN1+1), tmp	/*     than loopcnt */
+	cmp.gt	p_scr, p0 = 16, cnt		/* is it a minimalistic task? */
+(p_scr)	br.cond.dptk.many .move_bytes_unaligned	/* go move just a few (M_B_U) */
 ;; }
 { .mmi
-(p_unalgn) add	ptr1 = (MIN1+1), ptr2		// after alignment
-(p_unalgn) add	ptr2 = MIN1P1HALF, ptr2		// after alignment
-(p_unalgn) tbit.nz.unc p_y, p_n = bytecnt, 3	// should we do a st8 ?
+(p_unalgn) add	ptr1 = (MIN1+1), ptr2		/* after alignment */
+(p_unalgn) add	ptr2 = MIN1P1HALF, ptr2		/* after alignment */
+(p_unalgn) tbit.nz.unc p_y, p_n = bytecnt, 3	/* should we do a st8 ? */
 ;; }
 { .mib
 (p_y)	add	cnt = -8, cnt
-(p_unalgn) tbit.nz.unc p_yy, p_nn = bytecnt, 2	// should we do a st4 ?
+(p_unalgn) tbit.nz.unc p_yy, p_nn = bytecnt, 2	/* should we do a st4 ? */
 } { .mib
 (p_y)	st8	[ptr2] = value, -4
 (p_n)	add	ptr2 = 4, ptr2
 ;; }
 { .mib
 (p_yy)	add	cnt = -4, cnt
-(p_unalgn) tbit.nz.unc p_y, p_n = bytecnt, 1	// should we do a st2 ?
+(p_unalgn) tbit.nz.unc p_y, p_n = bytecnt, 1	/* should we do a st2 ? */
 } { .mib
 (p_yy)	st4	[ptr2] = value, -2
 (p_nn)	add	ptr2 = 2, ptr2
 ;; }
 { .mmi
-	mov	tmp = LINE_SIZE+1		// for compare
+	mov	tmp = LINE_SIZE+1		/* for compare */
 (p_y)	add	cnt = -2, cnt
-(p_unalgn) tbit.nz.unc p_yy, p_nn = bytecnt, 0	// should we do a st1 ?
+(p_unalgn) tbit.nz.unc p_yy, p_nn = bytecnt, 0	/* should we do a st1 ? */
 } { .mmi
-	setf.sig fvalue=value			// transfer value to FLP side
+	setf.sig fvalue=value			/* transfer value to FLP side */
 (p_y)	st2	[ptr2] = value, -1
 (p_n)	add	ptr2 = 1, ptr2
 ;; }
 
 { .mmi
 (p_yy)	st1	[ptr2] = value
-  	cmp.gt	p_scr, p0 = tmp, cnt		// is it a minimalistic task?
+	cmp.gt	p_scr, p0 = tmp, cnt		/* is it a minimalistic task? */
 } { .mbb
 (p_yy)	add	cnt = -1, cnt
-(p_scr)	br.cond.dpnt.many .fraction_of_line	// go move just a few
+(p_scr)	br.cond.dpnt.many .fraction_of_line	/* go move just a few */
 ;; }
 
 { .mib
 	nop.m 0
 	shr.u	linecnt = cnt, LSIZE_SH
-(p_zr)	br.cond.dptk.many .l1b			// Jump to use stf.spill
+(p_zr)	br.cond.dptk.many .l1b			/* Jump to use stf.spill */
 ;; }
 
 #ifndef GAS_ALIGN_BREAKS_UNWIND_INFO
-	.align 32 // -------- //  L1A: store ahead into cache lines; fill later
+	.align 32 /* --------  L1A: store ahead into cache lines; fill later */
 #endif
 { .mmi
-	and	tmp = -(LINE_SIZE), cnt		// compute end of range
-	mov	ptr9 = ptr1			// used for prefetching
-	and	cnt = (LINE_SIZE-1), cnt	// remainder
+	and	tmp = -(LINE_SIZE), cnt		/* compute end of range */
+	mov	ptr9 = ptr1			/* used for prefetching */
+	and	cnt = (LINE_SIZE-1), cnt	/* remainder */
 } { .mmi
-	mov	loopcnt = PREF_AHEAD-1		// default prefetch loop
-	cmp.gt	p_scr, p0 = PREF_AHEAD, linecnt	// check against actual value
+	mov	loopcnt = PREF_AHEAD-1		/* default prefetch loop */
+	cmp.gt	p_scr, p0 = PREF_AHEAD, linecnt	/* check against actual value */
 ;; }
 { .mmi
-(p_scr)	add	loopcnt = -1, linecnt		// start of stores
-	add	ptr2 = 8, ptr1			// (beyond prefetch stores)
-	add	ptr1 = tmp, ptr1		// first address beyond total
-;; }						// range
+(p_scr)	add	loopcnt = -1, linecnt		/* start of stores */
+	add	ptr2 = 8, ptr1			/* (beyond prefetch stores) */
+	add	ptr1 = tmp, ptr1		/* first address beyond total */
+;; }						/* range */
 { .mmi
-	add	tmp = -1, linecnt		// next loop count
+	add	tmp = -1, linecnt		/* next loop count */
 	movi0	ar.lc = loopcnt
 ;; }
 .pref_l1a:
 { .mib
-	store [ptr9] = myval, 128	// Do stores one cache line apart
+	store [ptr9] = myval, 128	/* Do stores one cache line apart */
 	nop.i	0
 	br.cloop.dptk.few .pref_l1a
 ;; }
 { .mmi
-	add	ptr0 = 16, ptr2		// Two stores in parallel
+	add	ptr0 = 16, ptr2		/* Two stores in parallel */
 	movi0	ar.lc = tmp
 ;; }
 .l1ax:
@@ -211,7 +211,7 @@ ENTRY(memset)
  { .mmi
 	store [ptr2] = myval, 8
 	store [ptr0] = myval, 32
- 	cmp.lt	p_scr, p0 = ptr9, ptr1		// do we need more prefetching?
+	cmp.lt	p_scr, p0 = ptr9, ptr1	/* do we need more prefetching? */
  ;; }
 { .mmb
 	store [ptr2] = myval, 24
@@ -219,9 +219,9 @@ ENTRY(memset)
 	br.cloop.dptk.few .l1ax
 ;; }
 { .mbb
-	cmp.le  p_scr, p0 = 8, cnt		// just a few bytes left ?
-(p_scr) br.cond.dpnt.many  .fraction_of_line	// Branch no. 2
-	br.cond.dpnt.many  .move_bytes_from_alignment	// Branch no. 3
+	cmp.le  p_scr, p0 = 8, cnt		/* just a few bytes left ? */
+(p_scr) br.cond.dpnt.many  .fraction_of_line	/* Branch no. 2 */
+	br.cond.dpnt.many  .move_bytes_from_alignment	/* Branch no. 3 */
 ;; }
 
 #ifdef GAS_ALIGN_BREAKS_UNWIND_INFO
@@ -229,32 +229,32 @@ ENTRY(memset)
 #else
 	.align 32
 #endif
-.l1b:	// ------------------ //  L1B: store ahead into cache lines; fill later
+.l1b:	/* ------------------  L1B: store ahead into cache lines; fill later */
 { .mmi
-	and	tmp = -(LINE_SIZE), cnt		// compute end of range
-	mov	ptr9 = ptr1			// used for prefetching
-	and	cnt = (LINE_SIZE-1), cnt	// remainder
+	and	tmp = -(LINE_SIZE), cnt		/* compute end of range */
+	mov	ptr9 = ptr1			/* used for prefetching */
+	and	cnt = (LINE_SIZE-1), cnt	/* remainder */
 } { .mmi
-	mov	loopcnt = PREF_AHEAD-1		// default prefetch loop
-	cmp.gt	p_scr, p0 = PREF_AHEAD, linecnt	// check against actual value
+	mov	loopcnt = PREF_AHEAD-1		/* default prefetch loop */
+	cmp.gt	p_scr, p0 = PREF_AHEAD, linecnt	/* check against actual value */
 ;; }
 { .mmi
 (p_scr)	add	loopcnt = -1, linecnt
-	add	ptr2 = 16, ptr1	// start of stores (beyond prefetch stores)
-	add	ptr1 = tmp, ptr1	// first address beyond total range
+	add	ptr2 = 16, ptr1	/* start of stores (beyond prefetch stores) */
+	add	ptr1 = tmp, ptr1	/* first address beyond total range */
 ;; }
 { .mmi
-	add	tmp = -1, linecnt	// next loop count
+	add	tmp = -1, linecnt	/* next loop count */
 	movi0	ar.lc = loopcnt
 ;; }
 .pref_l1b:
 { .mib
-	stf.spill [ptr9] = f0, 128	// Do stores one cache line apart
+	stf.spill [ptr9] = f0, 128	/* Do stores one cache line apart */
 	nop.i   0
 	br.cloop.dptk.few .pref_l1b
 ;; }
 { .mmi
-	add	ptr0 = 16, ptr2		// Two stores in parallel
+	add	ptr0 = 16, ptr2		/* Two stores in parallel */
 	movi0	ar.lc = tmp
 ;; }
 .l1bx:
@@ -269,7 +269,7 @@ ENTRY(memset)
  { .mmi
 	stf.spill [ptr2] = f0, 32
 	stf.spill [ptr0] = f0, 64
- 	cmp.lt	p_scr, p0 = ptr9, ptr1	// do we need more prefetching?
+	cmp.lt	p_scr, p0 = ptr9, ptr1	/* do we need more prefetching? */
  ;; }
 { .mmb
 	stf.spill [ptr2] = f0, 32
@@ -277,14 +277,14 @@ ENTRY(memset)
 	br.cloop.dptk.few .l1bx
 ;; }
 { .mib
-	cmp.gt  p_scr, p0 = 8, cnt	// just a few bytes left ?
+	cmp.gt  p_scr, p0 = 8, cnt	/* just a few bytes left ? */
 (p_scr)	br.cond.dpnt.many  .move_bytes_from_alignment
 ;; }
 
 .fraction_of_line:
 { .mib
 	add	ptr2 = 16, ptr1
-	shr.u	loopcnt = cnt, 5   	// loopcnt = cnt / 32
+	shr.u	loopcnt = cnt, 5	/* loopcnt = cnt / 32 */
 ;; }
 { .mib
 	cmp.eq	p_scr, p0 = loopcnt, r0
@@ -292,13 +292,13 @@ ENTRY(memset)
 (p_scr)	br.cond.dpnt.many store_words
 ;; }
 { .mib
-	and	cnt = 0x1f, cnt		// compute the remaining cnt
+	and	cnt = 0x1f, cnt		/* compute the remaining cnt */
 	movi0   ar.lc = loopcnt
 ;; }
 #ifndef GAS_ALIGN_BREAKS_UNWIND_INFO
 	.align 32
 #endif
-.l2:	// ---------------------------- //  L2A:  store 32B in 2 cycles
+.l2:	/* ----------------------------  L2A:  store 32B in 2 cycles */
 { .mmb
 	store	[ptr1] = myval, 8
 	store	[ptr2] = myval, 8
@@ -309,34 +309,34 @@ ENTRY(memset)
 ;; }
 store_words:
 { .mib
-	cmp.gt	p_scr, p0 = 8, cnt		// just a few bytes left ?
-(p_scr)	br.cond.dpnt.many .move_bytes_from_alignment	// Branch
+	cmp.gt	p_scr, p0 = 8, cnt		/* just a few bytes left ? */
+(p_scr)	br.cond.dpnt.many .move_bytes_from_alignment	/* Branch */
 ;; }
 
 { .mmi
-	store	[ptr1] = myval, 8		// store
-	cmp.le	p_y, p_n = 16, cnt		//
-	add	cnt = -8, cnt			// subtract
+	store	[ptr1] = myval, 8		/* store */
+	cmp.le	p_y, p_n = 16, cnt		/* */
+	add	cnt = -8, cnt			/* subtract */
 ;; }
 { .mmi
-(p_y)	store	[ptr1] = myval, 8		// store
-(p_y)	cmp.le.unc p_yy, p_nn = 16, cnt		//
-(p_y)	add	cnt = -8, cnt			// subtract
+(p_y)	store	[ptr1] = myval, 8		/* store */
+(p_y)	cmp.le.unc p_yy, p_nn = 16, cnt		/* */
+(p_y)	add	cnt = -8, cnt			/* subtract */
 ;; }
-{ .mmi						// store
-(p_yy)	store	[ptr1] = myval, 8		//
-(p_yy)	add	cnt = -8, cnt			// subtract
+{ .mmi						/* store */
+(p_yy)	store	[ptr1] = myval, 8		/* */
+(p_yy)	add	cnt = -8, cnt			/* subtract */
 ;; }
 
 .move_bytes_from_alignment:
 { .mib
 	cmp.eq	p_scr, p0 = cnt, r0
-	tbit.nz.unc p_y, p0 = cnt, 2	// should we terminate with a st4 ?
+	tbit.nz.unc p_y, p0 = cnt, 2	/* should we terminate with a st4 ? */
 (p_scr)	br.cond.dpnt.few .restore_and_exit
 ;; }
 { .mib
 (p_y)	st4	[ptr1] = value, 4
-	tbit.nz.unc p_yy, p0 = cnt, 1	// should we terminate with a st2 ?
+	tbit.nz.unc p_yy, p0 = cnt, 1	/* should we terminate with a st2 ? */
 ;; }
 { .mib
 (p_yy)	st2	[ptr1] = value, 2
@@ -362,38 +362,38 @@ store_words:
 (p_n)	add	ptr2 = 2, ptr1
 } { .mmi
 (p_y)	add	ptr2 = 3, ptr1
-(p_y)	st1	[ptr1] = value, 1	// fill 1 (odd-aligned) byte
-(p_y)	add	cnt = -1, cnt		// [15, 14 (or less) left]
+(p_y)	st1	[ptr1] = value, 1	/* fill 1 (odd-aligned) byte */
+(p_y)	add	cnt = -1, cnt		/* [15, 14 (or less) left] */
 ;; }
 { .mmi
 (p_yy)	cmp.le.unc p_y, p0 = 8, cnt
-	add	ptr3 = ptr1, cnt	// prepare last store
+	add	ptr3 = ptr1, cnt	/* prepare last store */
 	movi0	ar.lc = save_lc
 } { .mmi
-(p_yy)	st2	[ptr1] = value, 4	// fill 2 (aligned) bytes
-(p_yy)	st2	[ptr2] = value, 4	// fill 2 (aligned) bytes
-(p_yy)	add	cnt = -4, cnt		// [11, 10 (o less) left]
+(p_yy)	st2	[ptr1] = value, 4	/* fill 2 (aligned) bytes */
+(p_yy)	st2	[ptr2] = value, 4	/* fill 2 (aligned) bytes */
+(p_yy)	add	cnt = -4, cnt		/* [11, 10 (o less) left] */
 ;; }
 { .mmi
 (p_y)	cmp.le.unc p_yy, p0 = 8, cnt
-	add	ptr3 = -1, ptr3		// last store
-	tbit.nz p_scr, p0 = cnt, 1	// will there be a st2 at the end ?
+	add	ptr3 = -1, ptr3		/* last store */
+	tbit.nz p_scr, p0 = cnt, 1	/* will there be a st2 at the end ? */
 } { .mmi
-(p_y)	st2	[ptr1] = value, 4	// fill 2 (aligned) bytes
-(p_y)	st2	[ptr2] = value, 4	// fill 2 (aligned) bytes
-(p_y)	add	cnt = -4, cnt		// [7, 6 (or less) left]
+(p_y)	st2	[ptr1] = value, 4	/* fill 2 (aligned) bytes */
+(p_y)	st2	[ptr2] = value, 4	/* fill 2 (aligned) bytes */
+(p_y)	add	cnt = -4, cnt		/* [7, 6 (or less) left] */
 ;; }
 { .mmi
-(p_yy)	st2	[ptr1] = value, 4	// fill 2 (aligned) bytes
-(p_yy)	st2	[ptr2] = value, 4	// fill 2 (aligned) bytes
-					// [3, 2 (or less) left]
-	tbit.nz p_y, p0 = cnt, 0	// will there be a st1 at the end ?
+(p_yy)	st2	[ptr1] = value, 4	/* fill 2 (aligned) bytes */
+(p_yy)	st2	[ptr2] = value, 4	/* fill 2 (aligned) bytes */
+					/* [3, 2 (or less) left] */
+	tbit.nz p_y, p0 = cnt, 0	/* will there be a st1 at the end ? */
 } { .mmi
 (p_yy)	add	cnt = -4, cnt
 ;; }
 { .mmb
-(p_scr)	st2	[ptr1] = value		// fill 2 (aligned) bytes
-(p_y)	st1	[ptr3] = value		// fill last byte (using ptr3)
+(p_scr)	st2	[ptr1] = value		/* fill 2 (aligned) bytes */
+(p_y)	st1	[ptr3] = value		/* fill last byte (using ptr3) */
 	br.ret.sptk.many rp
 ;; }
 END(memset)
diff --git a/libc/string/ia64/strchr.S b/libc/string/ia64/strchr.S
index 401a07941..66703f26d 100644
--- a/libc/string/ia64/strchr.S
+++ b/libc/string/ia64/strchr.S
@@ -49,15 +49,15 @@ ENTRY(strchr)
 	.prologue
 	alloc r2 = ar.pfs, 2, 0, 0, 0
 	.save ar.lc, saved_lc
-        mov 	saved_lc = ar.lc 	// save the loop counter
+        mov 	saved_lc = ar.lc 	/* save the loop counter */
 	.body
 	mov 	ret0 = str	
-	and 	tmp = 7, str		// tmp = str % 8
+	and 	tmp = 7, str		/* tmp = str % 8 */
 	mux1	chrx8 = chr, @brcst
-	extr.u	chr = chr, 0, 8		// retain only the last byte
-	cmp.ne	p8, p0 = r0, r0		// clear p8
+	extr.u	chr = chr, 0, 8		/* retain only the last byte */
+	cmp.ne	p8, p0 = r0, r0		/* clear p8 */
 	;;
-	sub	loopcnt = 8, tmp	// loopcnt = 8 - tmp
+	sub	loopcnt = 8, tmp	/* loopcnt = 8 - tmp */
 	cmp.eq	p6, p0 = tmp, r0
 (p6)	br.cond.sptk	.str_aligned;;
 	adds	loopcnt = -1, loopcnt;;
@@ -75,10 +75,10 @@ ENTRY(strchr)
 	nop.b	0
 	nop.b 	0
 .l2:	
-	ld8.s	val2 = [ret0], 8	// don't bomb out here
+	ld8.s	val2 = [ret0], 8	/* don't bomb out here */
 	czx1.r	pos0 = val1	
-	xor	tmp = val1, chrx8	// if val1 contains chr, tmp will
-	;;				// contain a zero in its position
+	xor	tmp = val1, chrx8	/* if val1 contains chr, tmp will */
+	;;				/* contain a zero in its position */
 	czx1.r	poschr = tmp
 	cmp.ne	p6, p0 = 8, pos0
 	;;
@@ -90,21 +90,21 @@ ENTRY(strchr)
 	mov	val1 = val2	
 	br.cond.dptk .l2
 .foundit:
-(p6)	cmp.lt	p8, p0 = pos0, poschr	// we found chr and null in the word
-(p8)	br.cond.spnt .notfound		// null was found before chr
+(p6)	cmp.lt	p8, p0 = pos0, poschr	/* we found chr and null in the word */
+(p8)	br.cond.spnt .notfound		/* null was found before chr */
 	add	ret0 = ret0, poschr ;;
-	adds	ret0 = -15, ret0 ;;	// should be -16, but we decrement
-.restore_and_exit:			// ret0 in the next instruction
-	adds	ret0 = -1, ret0		// ret0 was pointing 1 char too far
-	mov 	ar.lc = saved_lc	// restore the loop counter
+	adds	ret0 = -15, ret0 ;;	/* should be -16, but we decrement */
+.restore_and_exit:			/* ret0 in the next instruction */
+	adds	ret0 = -1, ret0		/* ret0 was pointing 1 char too far */
+	mov 	ar.lc = saved_lc	/* restore the loop counter */
 	br.ret.sptk.many b0
 .notfound:
-	mov	ret0 = r0		// return NULL if null was found
+	mov	ret0 = r0		/* return NULL if null was found */
 	mov 	ar.lc = saved_lc
 	br.ret.sptk.many b0
 .recovery:
 	adds	ret0 = -8, ret0;;
-	ld8	val2 = [ret0], 8	// bomb out here
+	ld8	val2 = [ret0], 8	/* bomb out here */
 	br.cond.sptk	.back
 END(strchr)
 libc_hidden_def (strchr)
diff --git a/libc/string/ia64/strcmp.S b/libc/string/ia64/strcmp.S
index d3b41e642..4da72fa10 100644
--- a/libc/string/ia64/strcmp.S
+++ b/libc/string/ia64/strcmp.S
@@ -42,7 +42,7 @@ ENTRY(strcmp)
 .loop:
 	ld1	val1 = [s1], 1
 	ld1	val2 = [s2], 1
-	cmp.eq	p6, p0 = r0, r0		// set p6
+	cmp.eq	p6, p0 = r0, r0		/* set p6 */
 	;;
 	cmp.ne.and p6, p0 = val1, r0
 	cmp.ne.and p6, p0 = val2, r0
diff --git a/libc/string/ia64/strcpy.S b/libc/string/ia64/strcpy.S
index e4a9915ca..7b002f661 100644
--- a/libc/string/ia64/strcpy.S
+++ b/libc/string/ia64/strcpy.S
@@ -27,8 +27,8 @@
    In this form, it assumes little endian mode.  For big endian mode, the
    the two shifts in .l2 must be inverted:
 
-	shl   	value = r[1], sh1   	// value = w0 << sh1
-	shr.u   tmp = r[0], sh2   	// tmp = w1 >> sh2
+	shl	value = r[1], sh1	// value = w0 << sh1
+	shr.u   tmp = r[0], sh2		// tmp = w1 >> sh2
  */
 
 #include "sysdep.h"
@@ -53,62 +53,62 @@
 
 ENTRY(strcpy)
 	.prologue
-	alloc 	r2 = ar.pfs, 2, 0, 30, 32
+	alloc	r2 = ar.pfs, 2, 0, 30, 32
 
 #define MEMLAT 2
 	.rotr	r[MEMLAT + 2]
 	.rotp	p[MEMLAT + 1]
 
-	mov	ret0 = in0		// return value = dest
+	mov	ret0 = in0		/* return value = dest */
 	.save pr, saved_pr
-	mov	saved_pr = pr           // save the predicate registers
+	mov	saved_pr = pr           /* save the predicate registers */
 	.save ar.lc, saved_lc
-        mov 	saved_lc = ar.lc	// save the loop counter
+        mov	saved_lc = ar.lc	/* save the loop counter */
 	.body
-	sub	tmp = r0, in0 ;;	// tmp = -dest
-	mov 	dest = in0		// dest
-	mov 	src = in1		// src
-	and	loopcnt = 7, tmp ;;	// loopcnt = -dest % 8
+	sub	tmp = r0, in0 ;;	/* tmp = -dest */
+	mov	dest = in0		/* dest */
+	mov	src = in1		/* src */
+	and	loopcnt = 7, tmp ;;	/* loopcnt = -dest % 8 */
 	cmp.eq	p6, p0 = loopcnt, r0
-	adds	loopcnt = -1, loopcnt	// --loopcnt
+	adds	loopcnt = -1, loopcnt	/* --loopcnt */
 (p6)	br.cond.sptk .dest_aligned ;;
 	mov	ar.lc = loopcnt
-.l1:					// copy -dest % 8 bytes
-	ld1	c = [src], 1		// c = *src++
+.l1:					/* copy -dest % 8 bytes */
+	ld1	c = [src], 1		/* c = *src++ */
 	;;
-	st1	[dest] = c, 1		// *dest++ = c
+	st1	[dest] = c, 1		/* *dest++ = c */
 	cmp.eq	p6, p0 = c, r0
 (p6)	br.cond.dpnt .restore_and_exit
 	br.cloop.dptk .l1 ;;
 .dest_aligned:
-	and	sh1 = 7, src 		// sh1 = src % 8
-	mov	ar.lc = -1		// "infinite" loop
-	and	asrc = -8, src ;;	// asrc = src & -OPSIZ  -- align src
+	and	sh1 = 7, src		/* sh1 = src % 8 */
+	mov	ar.lc = -1		/* "infinite" loop */
+	and	asrc = -8, src ;;	/* asrc = src & -OPSIZ  -- align src */
 	sub	thresh = 8, sh1
-	mov	pr.rot = 1 << 16	// set rotating predicates
-	cmp.ne	p7, p0 = r0, r0		// clear p7
-	shl	sh1 = sh1, 3 ;;		// sh1 = 8 * (src % 8)
-	sub	sh2 = 64, sh1		// sh2 = 64 - sh1
-	cmp.eq  p6, p0 = sh1, r0 	// is the src aligned?
+	mov	pr.rot = 1 << 16	/* set rotating predicates */
+	cmp.ne	p7, p0 = r0, r0		/* clear p7 */
+	shl	sh1 = sh1, 3 ;;		/* sh1 = 8 * (src % 8) */
+	sub	sh2 = 64, sh1		/* sh2 = 64 - sh1 */
+	cmp.eq  p6, p0 = sh1, r0	/* is the src aligned? */
 (p6)    br.cond.sptk .src_aligned ;;
 	ld8	r[1] = [asrc],8 ;;
 
 	.align	32
 .l2:
 	ld8.s	r[0] = [asrc], 8
-	shr.u	value = r[1], sh1 ;; 	// value = w0 >> sh1
-	czx1.r	pos = value ;;		// do we have an "early" zero
-	cmp.lt	p7, p0 = pos, thresh	// in w0 >> sh1?
+	shr.u	value = r[1], sh1 ;;	/* value = w0 >> sh1 */
+	czx1.r	pos = value ;;		/* do we have an "early" zero */
+	cmp.lt	p7, p0 = pos, thresh	/* in w0 >> sh1? */
 (p7)	br.cond.dpnt .found0
-	chk.s	r[0], .recovery2	// it is safe to do that only
-.back2:					// after the previous test
-	shl	tmp = r[0], sh2  	// tmp = w1 << sh2
+	chk.s	r[0], .recovery2	/* it is safe to do that only */
+.back2:					/* after the previous test */
+	shl	tmp = r[0], sh2		/* tmp = w1 << sh2 */
 	;;
-	or	value = value, tmp ;;	// value |= tmp
+	or	value = value, tmp ;;	/* value |= tmp */
 	czx1.r	pos = value ;;
 	cmp.ne	p7, p0 = 8, pos
 (p7)	br.cond.dpnt .found0
-	st8	[dest] = value, 8	// store val to dest
+	st8	[dest] = value, 8	/* store val to dest */
 	br.ctop.dptk    .l2 ;;
 .src_aligned:
 .l3:
@@ -124,14 +124,14 @@ ENTRY(strcpy)
 .found0:
 	mov	ar.lc = pos
 .l4:
-	extr.u	c = value, 0, 8		// c = value & 0xff
+	extr.u	c = value, 0, 8		/* c = value & 0xff */
 	shr.u	value = value, 8
 	;;
 	st1	[dest] = c, 1
 	br.cloop.dptk	.l4 ;;
 .restore_and_exit:
-	mov 	ar.lc = saved_lc	// restore the loop counter
-	mov	pr = saved_pr, -1	// restore the predicate registers
+	mov	ar.lc = saved_lc	/* restore the loop counter */
+	mov	pr = saved_pr, -1	/* restore the predicate registers */
 	br.ret.sptk.many b0
 .recovery2:
 	add	tmp = -8, asrc ;;
diff --git a/libc/string/ia64/strlen.S b/libc/string/ia64/strlen.S
index 9b27a2d1b..edbe84359 100644
--- a/libc/string/ia64/strlen.S
+++ b/libc/string/ia64/strlen.S
@@ -50,13 +50,13 @@ ENTRY(strlen)
 	.prologue
 	alloc r2 = ar.pfs, 1, 0, 0, 0
 	.save ar.lc, saved_lc
-        mov 	saved_lc = ar.lc 	// save the loop counter
+        mov 	saved_lc = ar.lc 	/* save the loop counter */
 	.body
 	mov 	str = in0	
-	mov 	len = r0		// len = 0
-	and 	tmp = 7, in0		// tmp = str % 8
+	mov 	len = r0		/* len = 0 */
+	and 	tmp = 7, in0		/* tmp = str % 8 */
 	;;
-	sub	loopcnt = 8, tmp	// loopcnt = 8 - tmp
+	sub	loopcnt = 8, tmp	/* loopcnt = 8 - tmp */
 	cmp.eq	p6, p0 = tmp, r0
 (p6)	br.cond.sptk	.str_aligned;;
 	adds	loopcnt = -1, loopcnt;;
@@ -69,11 +69,11 @@ ENTRY(strlen)
 	adds	len = 1, len
 	br.cloop.dptk	.l1
 .str_aligned:
-	mov	origadd = str		// origadd = orig
+	mov	origadd = str		/* origadd = orig */
 	ld8	val1 = [str], 8;;
 	nop.b	0
 	nop.b 	0
-.l2:	ld8.s	val2 = [str], 8		// don't bomb out here
+.l2:	ld8.s	val2 = [str], 8		/* don't bomb out here */
 	czx1.r	pos0 = val1	
 	;;
 	cmp.ne	p6, p0 = 8, pos0
@@ -83,16 +83,16 @@ ENTRY(strlen)
 	mov	val1 = val2	
 	br.cond.dptk	.l2
 .foundit:
-	sub	tmp = str, origadd	// tmp = crt address - orig
+	sub	tmp = str, origadd	/* tmp = crt address - orig */
 	add	len = len, pos0;;
 	add	len = len, tmp;;
 	adds	len = -16, len
 .restore_and_exit:
-	mov ar.lc = saved_lc		// restore the loop counter
+	mov ar.lc = saved_lc		/* restore the loop counter */
 	br.ret.sptk.many b0
 .recovery:
 	adds	str = -8, str;;
-	ld8	val2 = [str], 8		// bomb out here
+	ld8	val2 = [str], 8		/* bomb out here */
 	br.cond.sptk	.back
 END(strlen)
 libc_hidden_def (strlen)
diff --git a/libc/string/ia64/strncmp.S b/libc/string/ia64/strncmp.S
index 8e0373c7f..e31f8fbd9 100644
--- a/libc/string/ia64/strncmp.S
+++ b/libc/string/ia64/strncmp.S
@@ -23,7 +23,7 @@
    Inputs:
         in0:    s1
         in1:    s2
-  	in2:	n
+	in2:	n
 
    Unlike memcmp(), this function is optimized for mismatches within the
    first few characters.  */
@@ -42,13 +42,13 @@
 ENTRY(strncmp)
 	alloc	r2 = ar.pfs, 3, 0, 0, 0
 	mov	ret0 = r0
-	cmp.eq  p6, p0 = r0, r0		// set p6
-	cmp.eq	p7, p0 = n, r0		// return immediately if n == 0
+	cmp.eq  p6, p0 = r0, r0		/* set p6 */
+	cmp.eq	p7, p0 = n, r0		/* return immediately if n == 0 */
 (p7)	br.cond.spnt .restore_and_exit ;;
 .loop:
 	ld1	val1 = [s1], 1
 	ld1	val2 = [s2], 1
-	adds	n = -1, n		// n--
+	adds	n = -1, n		/* n-- */
 	;;
 	cmp.ne.and p6, p0 = val1, r0
 	cmp.ne.and p6, p0 = val2, r0
@@ -58,5 +58,5 @@ ENTRY(strncmp)
 	sub	ret0 = val1, val2
 .restore_and_exit:
 	br.ret.sptk.many b0
-END(strncmp)	
+END(strncmp)
 libc_hidden_weak (strncmp)
diff --git a/libc/string/ia64/strncpy.S b/libc/string/ia64/strncpy.S
index 4f1129350..3f29bbd52 100644
--- a/libc/string/ia64/strncpy.S
+++ b/libc/string/ia64/strncpy.S
@@ -58,64 +58,64 @@ ENTRY(strncpy)
 	.rotr	r[MEMLAT + 2]
 	.rotp	p[MEMLAT + 1]
 
-	mov	ret0 = in0		// return value = dest
+	mov	ret0 = in0		/* return value = dest */
 	.save pr, saved_pr
-	mov	saved_pr = pr           // save the predicate registers
+	mov	saved_pr = pr           /* save the predicate registers */
 	.save ar.lc, saved_lc
-	mov 	saved_lc = ar.lc	// save the loop counter
-	mov	ar.ec = 0		// ec is not guaranteed to
-					// be zero upon function entry
+	mov 	saved_lc = ar.lc	/* save the loop counter */
+	mov	ar.ec = 0		/* ec is not guaranteed to */
+					/* be zero upon function entry */
 	.body
 	cmp.geu p6, p5 = 24, in2
 (p6)	br.cond.spnt .short_len
-	sub	tmp = r0, in0 ;;	// tmp = -dest
-	mov	len = in2		// len
-	mov 	dest = in0		// dest
-	mov 	src = in1		// src
-	and	tmp = 7, tmp ;;		// loopcnt = -dest % 8
+	sub	tmp = r0, in0 ;;	/* tmp = -dest */
+	mov	len = in2		/* len */
+	mov 	dest = in0		/* dest */
+	mov 	src = in1		/* src */
+	and	tmp = 7, tmp ;;		/* loopcnt = -dest % 8 */
 	cmp.eq	p6, p7 = tmp, r0
-	adds	loopcnt = -1, tmp	// --loopcnt
+	adds	loopcnt = -1, tmp	/* --loopcnt */
 (p6)	br.cond.sptk .dest_aligned ;;
-	sub	len = len, tmp		// len -= -dest % 8
+	sub	len = len, tmp		/* len -= -dest % 8 */
 	mov	ar.lc = loopcnt
-.l1:					// copy -dest % 8 bytes
-(p5)	ld1	c = [src], 1		// c = *src++
+.l1:					/* copy -dest % 8 bytes */
+(p5)	ld1	c = [src], 1		/* c = *src++ */
 	;;
-	st1	[dest] = c, 1		// *dest++ = c
+	st1	[dest] = c, 1		/* *dest++ = c */
 	cmp.ne	p5, p7 = c, r0
 	br.cloop.dptk .l1 ;;
 (p7)	br.cond.dpnt	.found0_align
 
-.dest_aligned:				// p7 should be cleared here
-	shr.u	c = len, 3		// c = len / 8
-	and	sh1 = 7, src 		// sh1 = src % 8
-	and	asrc = -8, src ;;	// asrc = src & -OPSIZ  -- align src
-	adds	c = (MEMLAT-1), c	// c = (len / 8) + MEMLAT - 1
+.dest_aligned:				/* p7 should be cleared here */
+	shr.u	c = len, 3		/* c = len / 8 */
+	and	sh1 = 7, src 		/* sh1 = src % 8 */
+	and	asrc = -8, src ;;	/* asrc = src & -OPSIZ  -- align src */
+	adds	c = (MEMLAT-1), c	/* c = (len / 8) + MEMLAT - 1 */
 	sub	thresh = 8, sh1
-	mov	pr.rot = 1 << 16	// set rotating predicates
-	shl	sh1 = sh1, 3 ;;		// sh1 = 8 * (src % 8)
-	mov	ar.lc = c		// "infinite" loop
-	sub	sh2 = 64, sh1		// sh2 = 64 - sh1
-	cmp.eq  p6, p0 = sh1, r0 	// is the src aligned?
+	mov	pr.rot = 1 << 16	/* set rotating predicates */
+	shl	sh1 = sh1, 3 ;;		/* sh1 = 8 * (src % 8) */
+	mov	ar.lc = c		/* "infinite" loop */
+	sub	sh2 = 64, sh1		/* sh2 = 64 - sh1 */
+	cmp.eq  p6, p0 = sh1, r0 	/* is the src aligned? */
 (p6)    br.cond.sptk .src_aligned
-	adds	c = -(MEMLAT-1), c ;;	// c = (len / 8)
+	adds	c = -(MEMLAT-1), c ;;	/* c = (len / 8) */
 	ld8	r[1] = [asrc],8
 	mov	ar.lc = c ;;
 
 	.align	32
 .l2:
-(p6)	st8	[dest] = value, 8	// store val to dest
+(p6)	st8	[dest] = value, 8	/* store val to dest */
 	ld8.s	r[0] = [asrc], 8
-	shr.u	value = r[1], sh1 ;; 	// value = w0 >> sh1
-	czx1.r	pos = value ;;		// do we have an "early" zero
-	cmp.lt	p7, p0 = pos, thresh	// in w0 >> sh1?
-	adds	len = -8, len		// len -= 8
+	shr.u	value = r[1], sh1 ;; 	/* value = w0 >> sh1 */
+	czx1.r	pos = value ;;		/* do we have an "early" zero */
+	cmp.lt	p7, p0 = pos, thresh	/* in w0 >> sh1? */
+	adds	len = -8, len		/* len -= 8 */
 (p7)	br.cond.dpnt .nonalign_found0
-	chk.s	r[0], .recovery2	// it is safe to do that only
-.back2:					// after the previous test
-	shl	tmp = r[0], sh2  	// tmp = w1 << sh2
+	chk.s	r[0], .recovery2	/* it is safe to do that only */
+.back2:					/* after the previous test */
+	shl	tmp = r[0], sh2  	/* tmp = w1 << sh2 */
 	;;
-	or	value = value, tmp ;;	// value |= tmp
+	or	value = value, tmp ;;	/* value |= tmp */
 	czx1.r	pos = value ;;
 	cmp.ne	p7, p6 = 8, pos
 (p7)	br.cond.dpnt .nonalign_found0
@@ -137,7 +137,7 @@ ENTRY(strncpy)
 (p[MEMLAT])	mov	value = r[MEMLAT]
 (p[MEMLAT])	czx1.r	pos = r[MEMLAT] ;;
 (p[MEMLAT])	cmp.ne	p7, p0 = 8, pos
-(p[MEMLAT])	adds	len = -8, len	// len -= 8
+(p[MEMLAT])	adds	len = -8, len	/* len -= 8 */
 (p7)		br.cond.dpnt .found0
 (p[MEMLAT])	st8	[dest] = r[MEMLAT], 8
 		br.ctop.dptk .l3 ;;
@@ -152,7 +152,7 @@ ENTRY(strncpy)
 (p5)	br.cond.dptk	.restore_and_exit ;;
 	mov	ar.lc = len
 .l4:
-(p6)	extr.u	c = value, 0, 8		// c = value & 0xff
+(p6)	extr.u	c = value, 0, 8		/* c = value & 0xff */
 (p6)	shr.u	value = value, 8 ;;
 	st1	[dest] = c, 1
 	cmp.ne	p6, p0 = c, r0
@@ -165,7 +165,7 @@ ENTRY(strncpy)
 	mov	value = 0 ;;
 .found0:
 	shl	tmp = pos, 3
-	shr.u	loopcnt = len, 4	// loopcnt = len / 16
+	shr.u	loopcnt = len, 4	/* loopcnt = len / 16 */
 	mov	c = -1 ;;
 	cmp.eq	p6, p0 = loopcnt, r0
 	adds	loopcnt = -1, loopcnt
@@ -192,24 +192,24 @@ ENTRY(strncpy)
 	st1	[dest] = r0, 1
 	br.cloop.dptk	.l7 ;;
 .restore_and_exit:
-	mov 	ar.lc = saved_lc	// restore the loop counter
-	mov	pr = saved_pr, -1	// restore the predicate registers
+	mov 	ar.lc = saved_lc	/* restore the loop counter */
+	mov	pr = saved_pr, -1	/* restore the predicate registers */
 	br.ret.sptk.many b0
 
 .short_len:
 	cmp.eq	p5, p0 = in2, r0
 	adds	loopcnt = -1, in2
 (p5)	br.cond.spnt .restore_and_exit ;;
-	mov	ar.lc = loopcnt		// p6 should be set when we get here
+	mov	ar.lc = loopcnt		/* p6 should be set when we get here */
 .l8:
-(p6)	ld1	c = [in1], 1		// c = *src++
+(p6)	ld1	c = [in1], 1		/* c = *src++ */
 	;;
-	st1	[in0] = c, 1		// *dest++ = c
+	st1	[in0] = c, 1		/* *dest++ = c */
 (p6)	cmp.ne	p6, p0 = c, r0
 	br.cloop.dptk .l8
 	;;
-	mov 	ar.lc = saved_lc	// restore the loop counter
-	mov	pr = saved_pr, -1	// restore the predicate registers
+	mov 	ar.lc = saved_lc	/* restore the loop counter */
+	mov	pr = saved_pr, -1	/* restore the predicate registers */
 	br.ret.sptk.many b0
 .recovery2:
 	add	c = 8, len
diff --git a/libc/string/sh64/memcpy.S b/libc/string/sh64/memcpy.S
index 3c0ea0c0d..470784ecd 100644
--- a/libc/string/sh64/memcpy.S
+++ b/libc/string/sh64/memcpy.S
@@ -151,7 +151,7 @@ Large:
 	add r2, r4, r5
 	ldlo.q r3, 0, r0
 	addi r5, -16, r5
-	movi 64+8, r27 // could subtract r7 from that.
+	movi 64+8, r27 /* could subtract r7 from that. */
 	stlo.q r2, 0, r0
 	sthi.q r2, 7, r0
 	ldx.q r22, r6, r0
diff --git a/libc/string/sh64/memset.S b/libc/string/sh64/memset.S
index f588323f0..1b8812cd6 100644
--- a/libc/string/sh64/memset.S
+++ b/libc/string/sh64/memset.S
@@ -32,12 +32,12 @@ memset:
 	ptabs r18, tr2
 	mshflo.b r3,r3,r3
 	add r4, r22, r23
-	mperm.w r3, r63, r3	// Fill pattern now in every byte of r3
+	mperm.w r3, r63, r3	/* Fill pattern now in every byte of r3 */
 
 	movi 8, r9
-	bgtu/u r23, r9, tr0 // multiquad
+	bgtu/u r23, r9, tr0 /* multiquad */
 
-	beqi/u r4, 0, tr2       // Return with size 0 - ensures no mem accesses
+	beqi/u r4, 0, tr2       /* Return with size 0 - ensures no mem accesses */
 	ldlo.q r2, 0, r7
 	shlli r4, 2, r4
 	movi -1, r8
@@ -52,20 +52,21 @@ multiquad:
 	stlo.q r2, 0, r3
 	shlri r23, 3, r24
 	add r2, r4, r5
-	beqi/u r24, 1, tr0 // lastquad
+	beqi/u r24, 1, tr0 /* lastquad */
 	pta/l loop, tr1
 	sub r2, r22, r25
-	andi r5, -8, r20   // calculate end address and
-	addi r20, -7*8, r8 // loop end address; This might overflow, so we need
-	                   // to use a different test before we start the loop
-	bge/u r24, r9, tr1 // loop
+	andi r5, -8, r20   /* calculate end address and */
+	addi r20, -7*8, r8 /* loop end address; This might overflow, so we need
+	                      to use a different test before we start the loop
+	                    */
+	bge/u r24, r9, tr1 /* loop */
 	st.q r25, 8, r3
 	st.q r20, -8, r3
 	shlri r24, 1, r24
-	beqi/u r24, 1, tr0 // lastquad
+	beqi/u r24, 1, tr0 /* lastquad */
 	st.q r25, 16, r3
 	st.q r20, -16, r3
-	beqi/u r24, 2, tr0 // lastquad
+	beqi/u r24, 2, tr0 /* lastquad */
 	st.q r25, 24, r3
 	st.q r20, -24, r3
 lastquad:
@@ -73,15 +74,15 @@ lastquad:
 	blink tr2,r63
 
 loop:
-!!!	alloco r25, 32	// QQQ comment out for short-term fix to SHUK #3895.
-			// QQQ commenting out is locically correct, but sub-optimal
-			// QQQ Sean McGoogan - 4th April 2003.
+!!!	alloco r25, 32	/* QQQ comment out for short-term fix to SHUK #3895.
+			   QQQ commenting out is locically correct, but sub-optimal
+			   QQQ Sean McGoogan - 4th April 2003. */
 	st.q r25, 8, r3
 	st.q r25, 16, r3
 	st.q r25, 24, r3
 	st.q r25, 32, r3
 	addi r25, 32, r25
-	bgeu/l r8, r25, tr1 // loop
+	bgeu/l r8, r25, tr1 /* loop */
 
 	st.q r20, -40, r3
 	st.q r20, -32, r3
diff --git a/libc/string/sh64/strcpy.S b/libc/string/sh64/strcpy.S
index da79d5143..f317707b7 100644
--- a/libc/string/sh64/strcpy.S
+++ b/libc/string/sh64/strcpy.S
@@ -31,7 +31,7 @@ strcpy:
 	addi r2, 8, r0
 	mcmpeq.b r4,r63,r6
 	SHHI r6,r7,r6
-	bnei/u r6,0,tr1 // shortstring
+	bnei/u r6,0,tr1 /* shortstring */
 	pta/l no_lddst, tr2
 	ori r3,-8,r23
 	sub r2, r23, r0
@@ -41,28 +41,28 @@ strcpy:
 	pta/l loop, tr0
 	ori r2,-8,r22
 	mcmpeq.b r5, r63, r6
-	bgt/u r22, r23, tr2 // no_lddst
+	bgt/u r22, r23, tr2 /* no_lddst */
 
-	// r22 < r23 :  Need to do a load from the destination.
-	// r22 == r23 : Doesn't actually need to load from destination,
-	//              but still can be handled here.
+	/* r22 < r23 :  Need to do a load from the destination. */
+	/* r22 == r23 : Doesn't actually need to load from destination, */
+	/*              but still can be handled here. */
 	ldlo.q r2, 0, r9
 	movi -1, r8
 	SHLO r8, r7, r8
 	mcmv r4, r8, r9
 	stlo.q r2, 0, r9
-	beqi/l r6, 0, tr0 // loop
+	beqi/l r6, 0, tr0 /* loop */
 
 	add r5, r63, r4
 	addi r0, 8, r0
-	blink tr1, r63 // shortstring
+	blink tr1, r63 /* shortstring */
 no_lddst:
-	// r22 > r23: note that for r22 == r23 the sthi.q would clobber
-	//            bytes before the destination region.
+	/* r22 > r23: note that for r22 == r23 the sthi.q would clobber */
+	/*            bytes before the destination region. */
 	stlo.q r2, 0, r4
 	SHHI r4, r7, r4
 	sthi.q r0, -1, r4
-	beqi/l r6, 0, tr0 // loop
+	beqi/l r6, 0, tr0 /* loop */
 
 	add r5, r63, r4
 	addi r0, 8, r0
@@ -77,7 +77,7 @@ shortstring2:
 	shlri r4,8,r4
 	addi r0,1,r0
 	bnei/l r5,0,tr1
-	blink tr4,r63 // return
+	blink tr4,r63 /* return */
 	
 	.balign 8
 loop:
@@ -86,16 +86,16 @@ loop:
 	addi r0, 16, r0
 	sthi.q r0, -9, r5
 	mcmpeq.b r4, r63, r6
-	bnei/u r6, 0, tr1 // shortstring
+	bnei/u r6, 0, tr1 /* shortstring */
 	ldx.q r0, r21, r5
 	stlo.q r0, -8, r4
 	sthi.q r0, -1, r4
 	mcmpeq.b r5, r63, r6
-	beqi/l r6, 0, tr0 // loop
+	beqi/l r6, 0, tr0 /* loop */
 
 	add r5, r63, r4
 	addi r0, 8, r0
-	blink tr1, r63 // shortstring
+	blink tr1, r63 /* shortstring */
 
 	.size	strcpy,.-strcpy
 
diff --git a/libc/string/xtensa/memcpy.S b/libc/string/xtensa/memcpy.S
index 19f3a6818..fc04c023e 100644
--- a/libc/string/xtensa/memcpy.S
+++ b/libc/string/xtensa/memcpy.S
@@ -83,7 +83,7 @@ __memcpy_aux:
 	loopnez	a4, 2f
 #else
 	beqz	a4, 2f
-	add	a7, a3, a4	// a7 = end address for source
+	add	a7, a3, a4	/* a7 = end address for source */
 #endif
 1:	l8ui	a6, a3, 0
 	addi	a3, a3, 1
@@ -98,7 +98,7 @@ __memcpy_aux:
 /* Destination is unaligned.  */
 
 	.align	4
-.Ldst1mod2: // dst is only byte aligned
+.Ldst1mod2: /* dst is only byte aligned */
 
 	/* Do short copies byte-by-byte.  */
 	_bltui	a4, 7, .Lbytecopy
@@ -113,7 +113,7 @@ __memcpy_aux:
 	/* Return to main algorithm if dst is now aligned.  */
 	_bbci.l	a5, 1, .Ldstaligned
 
-.Ldst2mod4: // dst has 16-bit alignment
+.Ldst2mod4: /* dst has 16-bit alignment */
 
 	/* Do short copies byte-by-byte.  */
 	_bltui	a4, 6, .Lbytecopy
@@ -134,7 +134,7 @@ __memcpy_aux:
 ENTRY (memcpy)
 	/* a2 = dst, a3 = src, a4 = len */
 
-	mov	a5, a2		// copy dst so that a2 is return value
+	mov	a5, a2		/* copy dst so that a2 is return value */
 	_bbsi.l	a2, 0, .Ldst1mod2
 	_bbsi.l	a2, 1, .Ldst2mod4
 .Ldstaligned:
@@ -152,7 +152,7 @@ ENTRY (memcpy)
 #else
 	beqz	a7, 2f
 	slli	a8, a7, 4
-	add	a8, a8, a3	// a8 = end of last 16B source chunk
+	add	a8, a8, a3	/* a8 = end of last 16B source chunk */
 #endif
 1:	l32i	a6, a3, 0
 	l32i	a7, a3, 4
@@ -218,18 +218,18 @@ ENTRY (memcpy)
 
 	/* Copy 16 bytes per iteration for word-aligned dst and
 	   unaligned src.  */
-	ssa8	a3		// set shift amount from byte offset
+	ssa8	a3		/* set shift amount from byte offset */
 #if UNALIGNED_ADDRESSES_CHECKED
-	and	a11, a3, a8	// save unalignment offset for below
-	sub	a3, a3, a11	// align a3
+	and	a11, a3, a8	/* save unalignment offset for below */
+	sub	a3, a3, a11	/* align a3 */
 #endif
-	l32i	a6, a3, 0	// load first word
+	l32i	a6, a3, 0	/* load first word */
 #if XCHAL_HAVE_LOOPS
 	loopnez	a7, 2f
 #else
 	beqz	a7, 2f
 	slli	a10, a7, 4
-	add	a10, a10, a3	// a10 = end of last 16B source chunk
+	add	a10, a10, a3	/* a10 = end of last 16B source chunk */
 #endif
 1:	l32i	a7, a3, 4
 	l32i	a8, a3, 8
@@ -273,7 +273,7 @@ ENTRY (memcpy)
 	mov	a6, a7
 4:
 #if UNALIGNED_ADDRESSES_CHECKED
-	add	a3, a3, a11	// readjust a3 with correct misalignment
+	add	a3, a3, a11	/* readjust a3 with correct misalignment */
 #endif
 	bbsi.l	a4, 1, 5f
 	bbsi.l	a4, 0, 6f
diff --git a/libc/string/xtensa/memset.S b/libc/string/xtensa/memset.S
index c0928825d..076b8f001 100644
--- a/libc/string/xtensa/memset.S
+++ b/libc/string/xtensa/memset.S
@@ -29,7 +29,7 @@
    The algorithm is as follows:
 
    Create a word with c in all byte positions.
-	
+
    If the destination is aligned, set 16B chunks with a loop, and then
    finish up with 8B, 4B, 2B, and 1B stores conditional on the length.
 
@@ -57,7 +57,7 @@ __memset_aux:
 	loopnez	a4, 2f
 #else
 	beqz	a4, 2f
-	add	a6, a5, a4	// a6 = ending address
+	add	a6, a5, a4	/* a6 = ending address */
 #endif
 1:	s8i	a3, a5, 0
 	addi	a5, a5, 1
@@ -71,7 +71,7 @@ __memset_aux:
 
 	.align	4
 
-.Ldst1mod2: // dst is only byte aligned
+.Ldst1mod2: /* dst is only byte aligned */
 
 	/* Do short sizes byte-by-byte.  */
 	bltui	a4, 8, .Lbyteset
@@ -84,7 +84,7 @@ __memset_aux:
 	/* Now retest if dst is aligned.  */
 	_bbci.l	a5, 1, .Ldstaligned
 
-.Ldst2mod4: // dst has 16-bit alignment
+.Ldst2mod4: /* dst has 16-bit alignment */
 
 	/* Do short sizes byte-by-byte.  */
 	bltui	a4, 8, .Lbyteset
@@ -108,7 +108,7 @@ ENTRY (memset)
 	slli	a7, a3, 16
 	or	a3, a3, a7
 
-	mov	a5, a2		// copy dst so that a2 is return value
+	mov	a5, a2		/* copy dst so that a2 is return value */
 
 	/* Check if dst is unaligned.  */
 	_bbsi.l	a2, 0, .Ldst1mod2
@@ -124,7 +124,7 @@ ENTRY (memset)
 #else
 	beqz	a7, 2f
 	slli	a6, a7, 4
-	add	a6, a6, a5	// a6 = end of last 16B chunk
+	add	a6, a6, a5	/* a6 = end of last 16B chunk */
 #endif
 	/* Set 16 bytes per iteration.  */
 1:	s32i	a3, a5, 0
diff --git a/libc/string/xtensa/strcmp.S b/libc/string/xtensa/strcmp.S
index 622bb27ed..ac058a2bf 100644
--- a/libc/string/xtensa/strcmp.S
+++ b/libc/string/xtensa/strcmp.S
@@ -45,35 +45,35 @@
 ENTRY (strcmp)
 	/* a2 = s1, a3 = s2 */
 
-	l8ui	a8, a2, 0	// byte 0 from s1
-	l8ui	a9, a3, 0	// byte 0 from s2
-	movi	a10, 3		// mask
+	l8ui	a8, a2, 0	/* byte 0 from s1 */
+	l8ui	a9, a3, 0	/* byte 0 from s2 */
+	movi	a10, 3		/* mask */
 	bne	a8, a9, .Lretdiff
 
 	or	a11, a2, a3
 	bnone	a11, a10, .Laligned
 
-	xor	a11, a2, a3	// compare low two bits of s1 and s2
-	bany	a11, a10, .Lunaligned	// if they have different alignment
+	xor	a11, a2, a3	/* compare low two bits of s1 and s2 */
+	bany	a11, a10, .Lunaligned	/* if they have different alignment */
 
 	/* s1/s2 are not word-aligned.  */
-	addi	a2, a2, 1	// advance s1
-	beqz	a8, .Leq	// bytes equal, if zero, strings are equal
-	addi	a3, a3, 1	// advance s2
-	bnone	a2, a10, .Laligned // if s1/s2 now aligned
-	l8ui	a8, a2, 0	// byte 1 from s1
-	l8ui	a9, a3, 0	// byte 1 from s2
-	addi	a2, a2, 1	// advance s1
-	bne	a8, a9, .Lretdiff // if different, return difference
-	beqz	a8, .Leq	// bytes equal, if zero, strings are equal
-	addi	a3, a3, 1	// advance s2
-	bnone	a2, a10, .Laligned // if s1/s2 now aligned
-	l8ui	a8, a2, 0	// byte 2 from s1
-	l8ui	a9, a3, 0	// byte 2 from s2
-	addi	a2, a2, 1	// advance s1
-	bne	a8, a9, .Lretdiff // if different, return difference
-	beqz	a8, .Leq	// bytes equal, if zero, strings are equal
-	addi	a3, a3, 1	// advance s2
+	addi	a2, a2, 1	/* advance s1 */
+	beqz	a8, .Leq	/* bytes equal, if zero, strings are equal */
+	addi	a3, a3, 1	/* advance s2 */
+	bnone	a2, a10, .Laligned /* if s1/s2 now aligned */
+	l8ui	a8, a2, 0	/* byte 1 from s1 */
+	l8ui	a9, a3, 0	/* byte 1 from s2 */
+	addi	a2, a2, 1	/* advance s1 */
+	bne	a8, a9, .Lretdiff /* if different, return difference */
+	beqz	a8, .Leq	/* bytes equal, if zero, strings are equal */
+	addi	a3, a3, 1	/* advance s2 */
+	bnone	a2, a10, .Laligned /* if s1/s2 now aligned */
+	l8ui	a8, a2, 0	/* byte 2 from s1 */
+	l8ui	a9, a3, 0	/* byte 2 from s2 */
+	addi	a2, a2, 1	/* advance s1 */
+	bne	a8, a9, .Lretdiff /* if different, return difference */
+	beqz	a8, .Leq	/* bytes equal, if zero, strings are equal */
+	addi	a3, a3, 1	/* advance s2 */
 	j	.Laligned
 
 /* s1 and s2 have different alignment.
@@ -92,8 +92,8 @@ ENTRY (strcmp)
 	/* (2 mod 4) alignment for loop instruction */
 .Lunaligned:
 #if XCHAL_HAVE_LOOPS
-	_movi.n	a8, 0		// set up for the maximum loop count
-	loop	a8, .Lretdiff	// loop forever (almost anyway)
+	_movi.n	a8, 0		/* set up for the maximum loop count */
+	loop	a8, .Lretdiff	/* loop forever (almost anyway) */
 #endif
 .Lnextbyte:
 	l8ui	a8, a2, 0
@@ -131,32 +131,32 @@ ENTRY (strcmp)
 #if XCHAL_HAVE_LOOPS
 .Laligned:
 	.begin	no-transform
-	l32r	a4, .Lmask0	// mask for byte 0
+	l32r	a4, .Lmask0	/* mask for byte 0 */
 	l32r	a7, .Lmask4
 	/* Loop forever.  (a4 is more than than the maximum number
 	   of iterations) */
 	loop	a4, .Laligned_done
 
 	/* First unrolled loop body.  */
-	l32i	a8, a2, 0	// get word from s1
-	l32i	a9, a3, 0	// get word from s2
+	l32i	a8, a2, 0	/* get word from s1 */
+	l32i	a9, a3, 0	/* get word from s2 */
 	slli	a5, a8, 1
 	bne	a8, a9, .Lwne2
 	or	a9, a8, a5
 	bnall	a9, a7, .Lprobeq
 
 	/* Second unrolled loop body.  */
-	l32i	a8, a2, 4	// get word from s1+4
-	l32i	a9, a3, 4	// get word from s2+4
+	l32i	a8, a2, 4	/* get word from s1+4 */
+	l32i	a9, a3, 4	/* get word from s2+4 */
 	slli	a5, a8, 1
 	bne	a8, a9, .Lwne2
 	or	a9, a8, a5
 	bnall	a9, a7, .Lprobeq2
 
-	addi	a2, a2, 8	// advance s1 pointer
-	addi	a3, a3, 8	// advance s2 pointer
+	addi	a2, a2, 8	/* advance s1 pointer */
+	addi	a3, a3, 8	/* advance s2 pointer */
 .Laligned_done:
-	or	a1, a1, a1	// nop
+	or	a1, a1, a1	/* nop */
 
 .Lprobeq2:
 	/* Adjust pointers to account for the loop unrolling.  */
@@ -166,15 +166,15 @@ ENTRY (strcmp)
 #else /* !XCHAL_HAVE_LOOPS */
 
 .Laligned:
-	movi	a4, MASK0	// mask for byte 0
+	movi	a4, MASK0	/* mask for byte 0 */
 	movi	a7, MASK4
 	j	.Lfirstword
 .Lnextword:
-	addi	a2, a2, 4	// advance s1 pointer
-	addi	a3, a3, 4	// advance s2 pointer
+	addi	a2, a2, 4	/* advance s1 pointer */
+	addi	a3, a3, 4	/* advance s2 pointer */
 .Lfirstword:
-	l32i	a8, a2, 0	// get word from s1
-	l32i	a9, a3, 0	// get word from s2
+	l32i	a8, a2, 0	/* get word from s1 */
+	l32i	a9, a3, 0	/* get word from s2 */
 	slli	a5, a8, 1
 	bne	a8, a9, .Lwne2
 	or	a9, a8, a5
@@ -186,49 +186,49 @@ ENTRY (strcmp)
 	/* Words are probably equal, but check for sure.
 	   If not, loop over the rest of string using normal algorithm.  */
 
-	bnone	a8, a4, .Leq	// if byte 0 is zero
-	l32r	a5, .Lmask1	// mask for byte 1
-	l32r	a6, .Lmask2	// mask for byte 2
-	bnone	a8, a5, .Leq	// if byte 1 is zero
-	l32r	a7, .Lmask3	// mask for byte 3
-	bnone	a8, a6, .Leq	// if byte 2 is zero
-	bnone	a8, a7, .Leq	// if byte 3 is zero
-	addi.n	a2, a2, 4	// advance s1 pointer
-	addi.n	a3, a3, 4	// advance s2 pointer
+	bnone	a8, a4, .Leq	/* if byte 0 is zero */
+	l32r	a5, .Lmask1	/* mask for byte 1 */
+	l32r	a6, .Lmask2	/* mask for byte 2 */
+	bnone	a8, a5, .Leq	/* if byte 1 is zero */
+	l32r	a7, .Lmask3	/* mask for byte 3 */
+	bnone	a8, a6, .Leq	/* if byte 2 is zero */
+	bnone	a8, a7, .Leq	/* if byte 3 is zero */
+	addi.n	a2, a2, 4	/* advance s1 pointer */
+	addi.n	a3, a3, 4	/* advance s2 pointer */
 #if XCHAL_HAVE_LOOPS
 
 	/* align (1 mod 4) */
-	loop	a4, .Leq	// loop forever (a4 is bigger than max iters)
+	loop	a4, .Leq	/* loop forever (a4 is bigger than max iters) */
 	.end	no-transform
 
-	l32i	a8, a2, 0	// get word from s1
-	l32i	a9, a3, 0	// get word from s2
-	addi	a2, a2, 4	// advance s1 pointer
+	l32i	a8, a2, 0	/* get word from s1 */
+	l32i	a9, a3, 0	/* get word from s2 */
+	addi	a2, a2, 4	/* advance s1 pointer */
 	bne	a8, a9, .Lwne
-	bnone	a8, a4, .Leq	// if byte 0 is zero
-	bnone	a8, a5, .Leq	// if byte 1 is zero
-	bnone	a8, a6, .Leq	// if byte 2 is zero
-	bnone	a8, a7, .Leq	// if byte 3 is zero
-	addi	a3, a3, 4	// advance s2 pointer
+	bnone	a8, a4, .Leq	/* if byte 0 is zero */
+	bnone	a8, a5, .Leq	/* if byte 1 is zero */
+	bnone	a8, a6, .Leq	/* if byte 2 is zero */
+	bnone	a8, a7, .Leq	/* if byte 3 is zero */
+	addi	a3, a3, 4	/* advance s2 pointer */
 
 #else /* !XCHAL_HAVE_LOOPS */
 
 	j	.Lfirstword2
 .Lnextword2:
-	addi	a3, a3, 4	// advance s2 pointer
+	addi	a3, a3, 4	/* advance s2 pointer */
 .Lfirstword2:
-	l32i	a8, a2, 0	// get word from s1
-	l32i	a9, a3, 0	// get word from s2
-	addi	a2, a2, 4	// advance s1 pointer
+	l32i	a8, a2, 0	/* get word from s1 */
+	l32i	a9, a3, 0	/* get word from s2 */
+	addi	a2, a2, 4	/* advance s1 pointer */
 	bne	a8, a9, .Lwne
-	bnone	a8, a4, .Leq	// if byte 0 is zero
-	bnone	a8, a5, .Leq	// if byte 1 is zero
-	bnone	a8, a6, .Leq	// if byte 2 is zero
-	bany	a8, a7, .Lnextword2	// if byte 3 is zero
+	bnone	a8, a4, .Leq	/* if byte 0 is zero */
+	bnone	a8, a5, .Leq	/* if byte 1 is zero */
+	bnone	a8, a6, .Leq	/* if byte 2 is zero */
+	bany	a8, a7, .Lnextword2	/* if byte 3 is zero */
 #endif /* !XCHAL_HAVE_LOOPS */
 
 	/* Words are equal; some byte is zero.  */
-.Leq:	movi	a2, 0		// return equal
+.Leq:	movi	a2, 0		/* return equal */
 	retw
 
 .Lwne2:	/* Words are not equal.  On big-endian processors, if none of the
@@ -243,18 +243,18 @@ ENTRY (strcmp)
 .Lposreturn:
 	movi	a2, 1
 	retw
-.Lsomezero:	// There is probably some zero byte.
+.Lsomezero:	/* There is probably some zero byte. */
 #endif /* __XTENSA_EB__ */
 .Lwne:	/* Words are not equal.  */
-	xor	a2, a8, a9	// get word with nonzero in byte that differs
-	bany	a2, a4, .Ldiff0	// if byte 0 differs
-	movi	a5, MASK1	// mask for byte 1
-	bnone	a8, a4, .Leq	// if byte 0 is zero
-	bany	a2, a5, .Ldiff1	// if byte 1 differs
-	movi	a6, MASK2	// mask for byte 2
-	bnone	a8, a5, .Leq	// if byte 1 is zero
-	bany	a2, a6, .Ldiff2	// if byte 2 differs
-	bnone	a8, a6, .Leq	// if byte 2 is zero
+	xor	a2, a8, a9	/* get word with nonzero in byte that differs */
+	bany	a2, a4, .Ldiff0	/* if byte 0 differs */
+	movi	a5, MASK1	/* mask for byte 1 */
+	bnone	a8, a4, .Leq	/* if byte 0 is zero */
+	bany	a2, a5, .Ldiff1	/* if byte 1 differs */
+	movi	a6, MASK2	/* mask for byte 2 */
+	bnone	a8, a5, .Leq	/* if byte 1 is zero */
+	bany	a2, a6, .Ldiff2	/* if byte 2 differs */
+	bnone	a8, a6, .Leq	/* if byte 2 is zero */
 #ifdef __XTENSA_EB__
 .Ldiff3:
 .Ldiff2:
diff --git a/libc/string/xtensa/strcpy.S b/libc/string/xtensa/strcpy.S
index 108070384..dc0a15175 100644
--- a/libc/string/xtensa/strcpy.S
+++ b/libc/string/xtensa/strcpy.S
@@ -36,7 +36,7 @@
 ENTRY (strcpy)
 	/* a2 = dst, a3 = src */
 
-	mov	a10, a2		// leave dst in return value register
+	mov	a10, a2		/* leave dst in return value register */
 	movi	a4, MASK0
 	movi	a5, MASK1
 	movi	a6, MASK2
@@ -51,23 +51,23 @@ ENTRY (strcpy)
 
 	j	.Ldstunaligned
 
-.Lsrc1mod2: // src address is odd
-	l8ui	a8, a3, 0	// get byte 0
-	addi	a3, a3, 1	// advance src pointer
-	s8i	a8, a10, 0	// store byte 0
-	beqz	a8, 1f		// if byte 0 is zero
-	addi	a10, a10, 1	// advance dst pointer
-	bbci.l	a3, 1, .Lsrcaligned // if src is now word-aligned
+.Lsrc1mod2: /* src address is odd */
+	l8ui	a8, a3, 0	/* get byte 0 */
+	addi	a3, a3, 1	/* advance src pointer */
+	s8i	a8, a10, 0	/* store byte 0 */
+	beqz	a8, 1f		/* if byte 0 is zero */
+	addi	a10, a10, 1	/* advance dst pointer */
+	bbci.l	a3, 1, .Lsrcaligned /* if src is now word-aligned */
 
-.Lsrc2mod4: // src address is 2 mod 4
-	l8ui	a8, a3, 0	// get byte 0
+.Lsrc2mod4: /* src address is 2 mod 4 */
+	l8ui	a8, a3, 0	/* get byte 0 */
 	/* 1-cycle interlock */
-	s8i	a8, a10, 0	// store byte 0
-	beqz	a8, 1f		// if byte 0 is zero
-	l8ui	a8, a3, 1	// get byte 0
-	addi	a3, a3, 2	// advance src pointer
-	s8i	a8, a10, 1	// store byte 0
-	addi	a10, a10, 2	// advance dst pointer
+	s8i	a8, a10, 0	/* store byte 0 */
+	beqz	a8, 1f		/* if byte 0 is zero */
+	l8ui	a8, a3, 1	/* get byte 0 */
+	addi	a3, a3, 2	/* advance src pointer */
+	s8i	a8, a10, 1	/* store byte 0 */
+	addi	a10, a10, 2	/* advance dst pointer */
 	bnez	a8, .Lsrcaligned
 1:	retw
 
@@ -78,28 +78,28 @@ ENTRY (strcpy)
 #if XCHAL_HAVE_LOOPS
 	/* (2 mod 4) alignment for loop instruction */
 .Laligned:
-	_movi.n	a8, 0		// set up for the maximum loop count
-	loop	a8, .Lz3	// loop forever (almost anyway)
-	l32i	a8, a3, 0	// get word from src
-	addi	a3, a3, 4	// advance src pointer
-	bnone	a8, a4, .Lz0	// if byte 0 is zero
-	bnone	a8, a5, .Lz1	// if byte 1 is zero
-	bnone	a8, a6, .Lz2	// if byte 2 is zero
-	s32i	a8, a10, 0	// store word to dst
-	bnone	a8, a7, .Lz3	// if byte 3 is zero
-	addi	a10, a10, 4	// advance dst pointer
+	_movi.n	a8, 0		/* set up for the maximum loop count */
+	loop	a8, .Lz3	/* loop forever (almost anyway) */
+	l32i	a8, a3, 0	/* get word from src */
+	addi	a3, a3, 4	/* advance src pointer */
+	bnone	a8, a4, .Lz0	/* if byte 0 is zero */
+	bnone	a8, a5, .Lz1	/* if byte 1 is zero */
+	bnone	a8, a6, .Lz2	/* if byte 2 is zero */
+	s32i	a8, a10, 0	/* store word to dst */
+	bnone	a8, a7, .Lz3	/* if byte 3 is zero */
+	addi	a10, a10, 4	/* advance dst pointer */
 
 #else /* !XCHAL_HAVE_LOOPS */
 
-1:	addi	a10, a10, 4	// advance dst pointer
+1:	addi	a10, a10, 4	/* advance dst pointer */
 .Laligned:
-	l32i	a8, a3, 0	// get word from src
-	addi	a3, a3, 4	// advance src pointer
-	bnone	a8, a4, .Lz0	// if byte 0 is zero
-	bnone	a8, a5, .Lz1	// if byte 1 is zero
-	bnone	a8, a6, .Lz2	// if byte 2 is zero
-	s32i	a8, a10, 0	// store word to dst
-	bany	a8, a7, 1b	// if byte 3 is zero
+	l32i	a8, a3, 0	/* get word from src */
+	addi	a3, a3, 4	/* advance src pointer */
+	bnone	a8, a4, .Lz0	/* if byte 0 is zero */
+	bnone	a8, a5, .Lz1	/* if byte 1 is zero */
+	bnone	a8, a6, .Lz2	/* if byte 2 is zero */
+	s32i	a8, a10, 0	/* store word to dst */
+	bany	a8, a7, 1b	/* if byte 3 is zero */
 #endif /* !XCHAL_HAVE_LOOPS */
 
 .Lz3:	/* Byte 3 is zero.  */
@@ -133,8 +133,8 @@ ENTRY (strcpy)
 .Ldstunaligned:
 
 #if XCHAL_HAVE_LOOPS
-	_movi.n	a8, 0		// set up for the maximum loop count
-	loop	a8, 2f		// loop forever (almost anyway)
+	_movi.n	a8, 0		/* set up for the maximum loop count */
+	loop	a8, 2f		/* loop forever (almost anyway) */
 #endif
 1:	l8ui	a8, a3, 0
 	addi	a3, a3, 1
diff --git a/libc/string/xtensa/strlen.S b/libc/string/xtensa/strlen.S
index dd72c16fa..9ee4995f4 100644
--- a/libc/string/xtensa/strlen.S
+++ b/libc/string/xtensa/strlen.S
@@ -36,7 +36,7 @@
 ENTRY (strlen)
 	/* a2 = s */
 
-	addi	a3, a2, -4	// because we overincrement at the end
+	addi	a3, a2, -4	/* because we overincrement at the end */
 	movi	a4, MASK0
 	movi	a5, MASK1
 	movi	a6, MASK2
@@ -45,21 +45,21 @@ ENTRY (strlen)
 	bbsi.l	a2, 1, .L2mod4
 	j	.Laligned
 
-.L1mod2: // address is odd
-	l8ui	a8, a3, 4	// get byte 0
-	addi	a3, a3, 1	// advance string pointer
-	beqz	a8, .Lz3	// if byte 0 is zero
-	bbci.l	a3, 1, .Laligned // if string pointer is now word-aligned
+.L1mod2: /* address is odd */
+	l8ui	a8, a3, 4	/* get byte 0 */
+	addi	a3, a3, 1	/* advance string pointer */
+	beqz	a8, .Lz3	/* if byte 0 is zero */
+	bbci.l	a3, 1, .Laligned /* if string pointer is now word-aligned */
 
-.L2mod4: // address is 2 mod 4
-	addi	a3, a3, 2	// advance ptr for aligned access
-	l32i	a8, a3, 0	// get word with first two bytes of string
-	bnone	a8, a6, .Lz2	// if byte 2 (of word, not string) is zero
-	bany	a8, a7, .Laligned // if byte 3 (of word, not string) is nonzero
+.L2mod4: /* address is 2 mod 4 */
+	addi	a3, a3, 2	/* advance ptr for aligned access */
+	l32i	a8, a3, 0	/* get word with first two bytes of string */
+	bnone	a8, a6, .Lz2	/* if byte 2 (of word, not string) is zero */
+	bany	a8, a7, .Laligned /* if byte 3 (of word, not string) is nonzero */
 
 	/* Byte 3 is zero.  */
-	addi	a3, a3, 3	// point to zero byte
-	sub	a2, a3, a2	// subtract to get length
+	addi	a3, a3, 3	/* point to zero byte */
+	sub	a2, a3, a2	/* subtract to get length */
 	retw
 
 
@@ -69,36 +69,36 @@ ENTRY (strlen)
 	/* (2 mod 4) alignment for loop instruction */
 .Laligned:
 #if XCHAL_HAVE_LOOPS
-	_movi.n	a8, 0		// set up for the maximum loop count
-	loop	a8, .Lz3	// loop forever (almost anyway)
+	_movi.n	a8, 0		/* set up for the maximum loop count */
+	loop	a8, .Lz3	/* loop forever (almost anyway) */
 #endif
-1:	l32i	a8, a3, 4	// get next word of string
-	addi	a3, a3, 4	// advance string pointer
-	bnone	a8, a4, .Lz0	// if byte 0 is zero
-	bnone	a8, a5, .Lz1	// if byte 1 is zero
-	bnone	a8, a6, .Lz2	// if byte 2 is zero
+1:	l32i	a8, a3, 4	/* get next word of string */
+	addi	a3, a3, 4	/* advance string pointer */
+	bnone	a8, a4, .Lz0	/* if byte 0 is zero */
+	bnone	a8, a5, .Lz1	/* if byte 1 is zero */
+	bnone	a8, a6, .Lz2	/* if byte 2 is zero */
 #if XCHAL_HAVE_LOOPS
-	bnone	a8, a7, .Lz3	// if byte 3 is zero
+	bnone	a8, a7, .Lz3	/* if byte 3 is zero */
 #else
-	bany	a8, a7, 1b	// repeat if byte 3 is non-zero
+	bany	a8, a7, 1b	/* repeat if byte 3 is non-zero */
 #endif
 
 .Lz3:	/* Byte 3 is zero.  */
-	addi	a3, a3, 3	// point to zero byte
+	addi	a3, a3, 3	/* point to zero byte */
 	/* Fall through....  */
 
 .Lz0:	/* Byte 0 is zero.  */
-	sub	a2, a3, a2	// subtract to get length
+	sub	a2, a3, a2	/* subtract to get length */
 	retw
 
 .Lz1:	/* Byte 1 is zero.  */
-	addi	a3, a3, 1	// point to zero byte
-	sub	a2, a3, a2	// subtract to get length
+	addi	a3, a3, 1	/* point to zero byte */
+	sub	a2, a3, a2	/* subtract to get length */
 	retw
 
 .Lz2:	/* Byte 2 is zero.  */
-	addi	a3, a3, 2	// point to zero byte
-	sub	a2, a3, a2	// subtract to get length
+	addi	a3, a3, 2	/* point to zero byte */
+	sub	a2, a3, a2	/* subtract to get length */
 	retw
 
 libc_hidden_def (strlen)
diff --git a/libc/string/xtensa/strncpy.S b/libc/string/xtensa/strncpy.S
index 7ba2ef77d..fe3ec894c 100644
--- a/libc/string/xtensa/strncpy.S
+++ b/libc/string/xtensa/strncpy.S
@@ -41,29 +41,29 @@
 	.literal_position
 __strncpy_aux:
 
-.Lsrc1mod2: // src address is odd
-	l8ui	a8, a3, 0	// get byte 0
-	addi	a3, a3, 1	// advance src pointer
-	s8i	a8, a10, 0	// store byte 0
-	addi	a4, a4, -1	// decrement n
-	beqz    a4, .Lret       // if n is zero
-	addi	a10, a10, 1	// advance dst pointer
-	beqz	a8, .Lfill	// if byte 0 is zero
-	bbci.l	a3, 1, .Lsrcaligned // if src is now word-aligned
-
-.Lsrc2mod4: // src address is 2 mod 4
-	l8ui	a8, a3, 0	// get byte 0
-	addi	a4, a4, -1	// decrement n
-	s8i	a8, a10, 0	// store byte 0
-	beqz    a4, .Lret       // if n is zero
-	addi	a10, a10, 1	// advance dst pointer
-	beqz	a8, .Lfill	// if byte 0 is zero
-	l8ui	a8, a3, 1	// get byte 0
-	addi	a3, a3, 2	// advance src pointer
-	s8i	a8, a10, 0	// store byte 0
-	addi	a4, a4, -1	// decrement n
-	beqz    a4, .Lret       // if n is zero
-	addi	a10, a10, 1	// advance dst pointer
+.Lsrc1mod2: /* src address is odd */
+	l8ui	a8, a3, 0	/* get byte 0 */
+	addi	a3, a3, 1	/* advance src pointer */
+	s8i	a8, a10, 0	/* store byte 0 */
+	addi	a4, a4, -1	/* decrement n */
+	beqz    a4, .Lret       /* if n is zero */
+	addi	a10, a10, 1	/* advance dst pointer */
+	beqz	a8, .Lfill	/* if byte 0 is zero */
+	bbci.l	a3, 1, .Lsrcaligned /* if src is now word-aligned */
+
+.Lsrc2mod4: /* src address is 2 mod 4 */
+	l8ui	a8, a3, 0	/* get byte 0 */
+	addi	a4, a4, -1	/* decrement n */
+	s8i	a8, a10, 0	/* store byte 0 */
+	beqz    a4, .Lret       /* if n is zero */
+	addi	a10, a10, 1	/* advance dst pointer */
+	beqz	a8, .Lfill	/* if byte 0 is zero */
+	l8ui	a8, a3, 1	/* get byte 0 */
+	addi	a3, a3, 2	/* advance src pointer */
+	s8i	a8, a10, 0	/* store byte 0 */
+	addi	a4, a4, -1	/* decrement n */
+	beqz    a4, .Lret       /* if n is zero */
+	addi	a10, a10, 1	/* advance dst pointer */
 	bnez	a8, .Lsrcaligned
 	j	.Lfill
 
@@ -74,8 +74,8 @@ __strncpy_aux:
 ENTRY (strncpy)
 	/* a2 = dst, a3 = src */
 
-	mov	a10, a2		// leave dst in return value register
-	beqz    a4, .Lret       // if n is zero
+	mov	a10, a2		/* leave dst in return value register */
+	beqz    a4, .Lret       /* if n is zero */
 
 	movi	a11, MASK0
 	movi	a5, MASK1
@@ -125,28 +125,28 @@ ENTRY (strncpy)
 
 .Lfillcleanup:
 	/* Fill leftover (1 to 3) bytes with zero.  */
-	s8i	a9, a10, 0	// store byte 0
-	addi	a4, a4, -1	// decrement n
+	s8i	a9, a10, 0	/* store byte 0 */
+	addi	a4, a4, -1	/* decrement n */
 	addi	a10, a10, 1
-	bnez    a4, .Lfillcleanup 
+	bnez    a4, .Lfillcleanup
 
 2:	retw
-	
-.Lfill1mod2: // dst address is odd
-	s8i	a9, a10, 0	// store byte 0
-	addi	a4, a4, -1	// decrement n
-	beqz    a4, 2b		// if n is zero
-	addi    a10, a10, 1	// advance dst pointer
-	bbci.l	a10, 1, .Lfillaligned // if dst is now word-aligned
-
-.Lfill2mod4: // dst address is 2 mod 4
-	s8i	a9, a10, 0	// store byte 0
-	addi	a4, a4, -1	// decrement n
-	beqz    a4, 2b		// if n is zero
-	s8i	a9, a10, 1	// store byte 1
-	addi	a4, a4, -1	// decrement n
-	beqz    a4, 2b		// if n is zero
-	addi    a10, a10, 2	// advance dst pointer
+
+.Lfill1mod2: /* dst address is odd */
+	s8i	a9, a10, 0	/* store byte 0 */
+	addi	a4, a4, -1	/* decrement n */
+	beqz    a4, 2b		/* if n is zero */
+	addi    a10, a10, 1	/* advance dst pointer */
+	bbci.l	a10, 1, .Lfillaligned /* if dst is now word-aligned */
+
+.Lfill2mod4: /* dst address is 2 mod 4 */
+	s8i	a9, a10, 0	/* store byte 0 */
+	addi	a4, a4, -1	/* decrement n */
+	beqz    a4, 2b		/* if n is zero */
+	s8i	a9, a10, 1	/* store byte 1 */
+	addi	a4, a4, -1	/* decrement n */
+	beqz    a4, 2b		/* if n is zero */
+	addi    a10, a10, 2	/* advance dst pointer */
 	j	.Lfillaligned
 
 
@@ -156,32 +156,32 @@ ENTRY (strncpy)
 	/* (2 mod 4) alignment for loop instruction */
 .Laligned:
 #if XCHAL_HAVE_LOOPS
-	_movi.n	a8, 0		// set up for the maximum loop count
-	loop	a8, 1f		// loop forever (almost anyway)
-	blti	a4, 5, .Ldstunaligned // n is near limit; do one at a time
-	l32i	a8, a3, 0	// get word from src
-	addi	a3, a3, 4	// advance src pointer
-	bnone	a8, a11, .Lz0	// if byte 0 is zero
-	bnone	a8, a5, .Lz1	// if byte 1 is zero
-	bnone	a8, a6, .Lz2	// if byte 2 is zero
-	s32i	a8, a10, 0	// store word to dst
-	addi	a4, a4, -4	// decrement n
-	addi	a10, a10, 4	// advance dst pointer
-	bnone	a8, a7, .Lfill	// if byte 3 is zero
-1:	
+	_movi.n	a8, 0		/* set up for the maximum loop count */
+	loop	a8, 1f		/* loop forever (almost anyway) */
+	blti	a4, 5, .Ldstunaligned /* n is near limit; do one at a time */
+	l32i	a8, a3, 0	/* get word from src */
+	addi	a3, a3, 4	/* advance src pointer */
+	bnone	a8, a11, .Lz0	/* if byte 0 is zero */
+	bnone	a8, a5, .Lz1	/* if byte 1 is zero */
+	bnone	a8, a6, .Lz2	/* if byte 2 is zero */
+	s32i	a8, a10, 0	/* store word to dst */
+	addi	a4, a4, -4	/* decrement n */
+	addi	a10, a10, 4	/* advance dst pointer */
+	bnone	a8, a7, .Lfill	/* if byte 3 is zero */
+1:
 
 #else /* !XCHAL_HAVE_LOOPS */
 
-1:	blti	a4, 5, .Ldstunaligned // n is near limit; do one at a time
-	l32i	a8, a3, 0	// get word from src
-	addi	a3, a3, 4	// advance src pointer
-	bnone	a8, a11, .Lz0	// if byte 0 is zero
-	bnone	a8, a5, .Lz1	// if byte 1 is zero
-	bnone	a8, a6, .Lz2	// if byte 2 is zero
-	s32i	a8, a10, 0	// store word to dst
-	addi	a4, a4, -4	// decrement n
-	addi	a10, a10, 4	// advance dst pointer
-	bany	a8, a7, 1b	// no zeroes
+1:	blti	a4, 5, .Ldstunaligned /* n is near limit; do one at a time */
+	l32i	a8, a3, 0	/* get word from src */
+	addi	a3, a3, 4	/* advance src pointer */
+	bnone	a8, a11, .Lz0	/* if byte 0 is zero */
+	bnone	a8, a5, .Lz1	/* if byte 1 is zero */
+	bnone	a8, a6, .Lz2	/* if byte 2 is zero */
+	s32i	a8, a10, 0	/* store word to dst */
+	addi	a4, a4, -4	/* decrement n */
+	addi	a10, a10, 4	/* advance dst pointer */
+	bany	a8, a7, 1b	/* no zeroes */
 #endif /* !XCHAL_HAVE_LOOPS */
 
 	j	.Lfill
@@ -191,8 +191,8 @@ ENTRY (strncpy)
 	movi	a8, 0
 #endif
 	s8i	a8, a10, 0
-	addi	a4, a4, -1	// decrement n
-	addi	a10, a10, 1	// advance dst pointer
+	addi	a4, a4, -1	/* decrement n */
+	addi	a10, a10, 1	/* advance dst pointer */
 	j	.Lfill
 
 .Lz1:	/* Byte 1 is zero.  */
@@ -200,8 +200,8 @@ ENTRY (strncpy)
         extui   a8, a8, 16, 16
 #endif
 	s16i	a8, a10, 0
-	addi	a4, a4, -2	// decrement n
-	addi	a10, a10, 2	// advance dst pointer
+	addi	a4, a4, -2	/* decrement n */
+	addi	a10, a10, 2	/* advance dst pointer */
 	j	.Lfill
 
 .Lz2:	/* Byte 2 is zero.  */
@@ -211,8 +211,8 @@ ENTRY (strncpy)
 	s16i	a8, a10, 0
 	movi	a8, 0
 	s8i	a8, a10, 2
-	addi	a4, a4, -3	// decrement n
-	addi	a10, a10, 3	// advance dst pointer
+	addi	a4, a4, -3	/* decrement n */
+	addi	a10, a10, 3	/* advance dst pointer */
 	j	.Lfill
 
 	.align	4
@@ -220,8 +220,8 @@ ENTRY (strncpy)
 .Ldstunaligned:
 
 #if XCHAL_HAVE_LOOPS
-	_movi.n	a8, 0		// set up for the maximum loop count
-	loop	a8, 2f		// loop forever (almost anyway)
+	_movi.n	a8, 0		/* set up for the maximum loop count */
+	loop	a8, 2f		/* loop forever (almost anyway) */
 #endif
 1:	l8ui	a8, a3, 0
 	addi	a3, a3, 1