remove -lea

code/jasmin/mlkem_avx2/Makefile

@@ -6,7 +6,7 @@
 CC     ?= /usr/bin/gcc
 GFLAGS ?=
 CFLAGS := -Wall -Wextra -g -Ofast -fomit-frame-pointer
-JFLAGS := -lea ${JADDFLAGS}
+JFLAGS := ${JADDFLAGS}
 OS     := $(shell uname -s)
 
 .SECONDARY: jpoly.s jpolyvec.s jfips202.s jindcpa.s jindcpa.o jkem.s

code/jasmin/mlkem_avx2/extraction/jkem_avx2.ec

@@ -5089,8 +5089,8 @@ module M(SC:Syscall_t) = {
     skp);
     buf <- Array64.init
            (fun i_0 => if 0 <= i_0 < 0 + 32 then aux.[i_0-0] else buf.[i_0]);
-    hp <- (skp + (W64.of_int 32));
-    hp <- (hp + (W64.of_int (((24 * 3) * 256) `|>>` 3)));
+    hp <- skp;
+    hp <- (hp + (W64.of_int (32 + (((24 * 3) * 256) `|>>` 3))));
     aux_0 <- (32 %/ 8);
     i <- 0;
     while (i < aux_0) {

code/jasmin/mlkem_avx2/kem.jinc

@@ -106,8 +106,8 @@ fn __crypto_kem_dec_jazz(reg u64 shkp, reg u64 ctp, reg u64 skp)
 
   buf[0:MLKEM_INDCPA_MSGBYTES] = __indcpa_dec_1(buf[0:MLKEM_INDCPA_MSGBYTES], ctp, skp);
 
-  hp = skp + 32;
-  hp += 24 * MLKEM_K * MLKEM_N>>3;
+  hp = skp;
+  hp += 32 + (24 * MLKEM_K * MLKEM_N>>3);
 
   /* fixme: should loads be 256-bits long? */
   for i=0 to MLKEM_SYMBYTES/8

code/jasmin/mlkem_ref/Makefile

@@ -4,7 +4,7 @@
 
 CC     ?= /usr/bin/gcc
 CFLAGS := -Wall -Wextra -g -O3 -fomit-frame-pointer
-JFLAGS := -lea ${JADDFLAGS}
+JFLAGS := ${JADDFLAGS}
 OS     := $(shell uname -s)
 
 .SECONDARY: jpoly.s jpolyvec.s jfips203.s jindcpa.s jkem.s

code/jasmin/mlkem_ref/extraction/jkem.ec

@@ -1114,21 +1114,25 @@ module M(SC:Syscall_t) = {
         zeta_0 <- zetasp.[(W64.to_uint zetasctr)];
         zetasctr <- (zetasctr + (W64.of_int 1));
         j <- start;
-        cmp <- (start + len);
+        cmp <- start;
+        cmp <- (cmp + len);
 
         while ((j \ult cmp)) {
-          offset <- (j + len);
+          offset <- j;
+          offset <- (offset + len);
           s <- rp.[(W64.to_uint offset)];
           t <- rp.[(W64.to_uint j)];
-          m <- (s + t);
+          m <- s;
+          m <- (m + t);
           m <@ __barrett_reduce (m);
           rp.[(W64.to_uint j)] <- m;
           t <- (t - s);
           t <@ __fqmul (t, zeta_0);
           rp.[(W64.to_uint offset)] <- t;
           j <- (j + (W64.of_int 1));
         }
-        start <- (j + len);
+        start <- j;
+        start <- (start + len);
       }
       len <- (len `<<` (W8.of_int 1));
     }
@@ -1169,10 +1173,12 @@ module M(SC:Syscall_t) = {
         zetasctr <- (zetasctr + (W64.of_int 1));
         zeta_0 <- zetasp.[(W64.to_uint zetasctr)];
         j <- start;
-        cmp <- (start + len);
+        cmp <- start;
+        cmp <- (cmp + len);
 
         while ((j \ult cmp)) {
-          offset <- (j + len);
+          offset <- j;
+          offset <- (offset + len);
           t <- rp.[(W64.to_uint offset)];
           t <@ __fqmul (t, zeta_0);
           s <- rp.[(W64.to_uint j)];
@@ -1183,7 +1189,8 @@ module M(SC:Syscall_t) = {
           rp.[(W64.to_uint j)] <- t;
           j <- (j + (W64.of_int 1));
         }
-        start <- (j + len);
+        start <- j;
+        start <- (start + len);
       }
       len <- (len `>>` (W8.of_int 1));
     }
@@ -2308,8 +2315,8 @@ module M(SC:Syscall_t) = {
     skp);
     buf <- Array64.init
            (fun i_0 => if 0 <= i_0 < 0 + 32 then aux.[i_0-0] else buf.[i_0]);
-    hp <- (skp + (W64.of_int 32));
-    hp <- (hp + (W64.of_int (((24 * 3) * 256) `|>>` 3)));
+    hp <- skp;
+    hp <- (hp + (W64.of_int (32 + (((24 * 3) * 256) `|>>` 3))));
     aux_0 <- (32 %/ 8);
     i <- 0;
     while (i < aux_0) {

code/jasmin/mlkem_ref/kem.jinc

@@ -107,8 +107,8 @@ fn __crypto_kem_dec_jazz(reg u64 shkp, reg u64 ctp, reg u64 skp)
 
   buf[0:MLKEM_MSGBYTES] = __indcpa_dec(buf[0:MLKEM_MSGBYTES], ctp, skp);
 
-  hp = skp + 32;
-  hp += 24 * MLKEM_K * MLKEM_N>>3;
+  hp = skp;
+  hp += 32 + (24 * MLKEM_K * MLKEM_N>>3);
 
   for i=0 to MLKEM_SYMBYTES/8
   {

code/jasmin/mlkem_ref/poly.jinc

@@ -505,21 +505,21 @@ fn _poly_invntt(reg ptr u16[MLKEM_N] rp) -> reg ptr u16[MLKEM_N]
       zetasctr += 1;
 
       j = start;
-      cmp = start + len;
+      cmp = start; cmp += len;
       while (j < cmp)
       {
-        offset = j + len;
+        offset = j; offset += len;
         s = rp[(int)offset];
         t = rp[(int)j];
-        m = s + t;
+        m = s; m += t;
         m = __barrett_reduce(m);
         rp[(int)j] = m;
         t -= s;
         t = __fqmul(t, zeta);
         rp[(int)offset] = t;
         j += 1;
       }
-      start = j + len;
+      start = j; start += len;
     }
     len <<= 1;
   }
@@ -563,10 +563,10 @@ fn _poly_ntt(reg ptr u16[MLKEM_N] rp) -> reg ptr u16[MLKEM_N]
       zetasctr += 1;
       zeta = zetasp[(int)zetasctr];
       j = start;
-      cmp = start + len;
+      cmp = start; cmp += len;
       while (j < cmp)
       {
-        offset = j + len;
+        offset = j; offset += len;
         t = rp[(int)offset];
         t = __fqmul(t, zeta);
         s = rp[(int)j];
@@ -577,7 +577,7 @@ fn _poly_ntt(reg ptr u16[MLKEM_N] rp) -> reg ptr u16[MLKEM_N]
         rp[(int)j] = t;
         j += 1;
       }
-      start = j + len;
+      start = j; start += len;
     }
     len >>= 1;
   }