Armv8.1-M: Remove unused r3 and r12 from push/pop in Keccak x4

Neither r3 nor r12 is callee-saved according to AAPCS.

Signed-off-by: Matthias J. Kannwischer <matthias@kannwischer.eu>

Author: mkannwischer

dev/fips202/armv81m/src/keccak_f1600_x4_mve.S

@@ -28,8 +28,8 @@
       c_parameter: const uint32_t *rc
       description: Keccak round constants in bit-interleaved form (24 pairs of 32-bit words)
   Stack:
-    bytes: 236
-    description: register preservation (44) + SIMD registers (64) + temporary storage (128)
+    bytes: 228
+    description: register preservation (36) + SIMD registers (64) + temporary storage (128)
 */
 
 // ---------------------------------------------------------------------------
@@ -460,7 +460,7 @@ qA20_l .req q2
 .global MLK_ASM_NAMESPACE(keccak_f1600_x4_mve_asm)
 MLK_ASM_FN_SYMBOL(keccak_f1600_x4_mve_asm)
 
-    push {r3,r4,r5,r6,r7,r8,r9,r10,r11,r12,lr}
+    push {r4,r5,r6,r7,r8,r9,r10,r11,lr}
     vpush {d8-d15}
     sub sp, #8*16
 
@@ -1626,7 +1626,7 @@ keccak_f1600_x4_mve_asm_roundend:
     add sp, #8*16
 
     vpop {d8-d15}
-    ldmia.w sp!, {r3,r4,r5,r6,r7,r8,r9,r10,r11,r12, pc}
+    ldmia.w sp!, {r4,r5,r6,r7,r8,r9,r10,r11, pc}
 
 /****************** REGISTER DEALLOCATIONS *******************/
     .unreq qA00_h

mlkem/src/fips202/native/armv81m/src/keccak_f1600_x4_mve.S

@@ -28,8 +28,8 @@
       c_parameter: const uint32_t *rc
       description: Keccak round constants in bit-interleaved form (24 pairs of 32-bit words)
   Stack:
-    bytes: 236
-    description: register preservation (44) + SIMD registers (64) + temporary storage (128)
+    bytes: 228
+    description: register preservation (36) + SIMD registers (64) + temporary storage (128)
 */
 
 // ---------------------------------------------------------------------------
@@ -79,7 +79,7 @@
 .global MLK_ASM_NAMESPACE(keccak_f1600_x4_mve_asm)
 MLK_ASM_FN_SYMBOL(keccak_f1600_x4_mve_asm)
 
-        push.w {r3, r4, r5, r6, r7, r8, r9, r10, r11, r12, lr}
+        push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr}
         vpush {d8, d9, d10, d11, d12, d13, d14, d15}
         sub sp, #0x80
         mov r6, r2
@@ -659,7 +659,7 @@ keccak_f1600_x4_mve_asm_roundend_pre:
 keccak_f1600_x4_mve_asm_roundend:
         add sp, #0x80
         vpop {d8, d9, d10, d11, d12, d13, d14, d15}
-        pop.w {r3, r4, r5, r6, r7, r8, r9, r10, r11, r12, pc}
+        pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc}
         nop
 
 MLK_ASM_FN_SIZE(keccak_f1600_x4_mve_asm)