diff --git a/src/bn256/fr.rs b/src/bn256/fr.rs
index a376566..1708ccf 100644
--- a/src/bn256/fr.rs
+++ b/src/bn256/fr.rs
@@ -388,7 +388,6 @@ impl Fr {
         let mut r5: u64;
         let mut r6: u64;
         let mut r7: u64;
-        let rhs = self.clone();
         unsafe {
             asm!(
                 // schoolbook multiplication
@@ -399,79 +398,82 @@ impl Fr {
                 //    b3   | b3 * a0 | b3 * a1 | b3 * a2 | b3 * a3
 
                 // init registers
-                "xor r13, r13",
-                "xor r14, r14",
-                "xor r15, r15",
+                "mov r13, qword ptr [{a_ptr} + 0]",
+                "mov r14, qword ptr [{a_ptr} + 8]",
+                "mov r15, qword ptr [{a_ptr} + 16]",
 
                 // `a0`
-                "mov rdx, qword ptr [{a_ptr} + 0]",
+                "mov rdx, r13",
 
                 // a0 * b0
-                "mulx r9, r8, qword ptr [{b_ptr} + 0]",
+                "mulx r9, r8, r13",
 
                 // a0 * b1
-                "mulx r10, rax, qword ptr [{b_ptr} + 8]",
+                "mulx r10, rax, r14",
                 "add r9, rax",
 
                 // a0 * b2
-                "mulx r11, rax, qword ptr [{b_ptr} + 16]",
+                "mulx r11, rax, r15",
                 "adcx r10, rax",
 
                 // a0 * b3
-                "mulx r12, rax, qword ptr [{b_ptr} + 24]",
+                "mulx r12, rax, qword ptr [{a_ptr} + 24]",
                 "adcx r11, rax",
                 "adc r12, 0",
 
                 // `a1`
-                "mov rdx, [{a_ptr} + 8]",
+                "mov rdx, r14",
 
                 // a1 * b0
-                "mulx rcx, rax, qword ptr [{b_ptr} + 0]",
+                "mulx rcx, rax, r13",
                 "add r9, rax",
                 "adcx r10, rcx",
                 "adc r11, 0",
 
                 // a1 * b1
-                "mulx rcx, rax, qword ptr [{b_ptr} + 8]",
+                "mulx rcx, rax, r14",
                 "add r10, rax",
                 "adcx r11, rcx",
                 "adc r12, 0",
+                "xor r13, r13",
 
                 // a1 * b2
-                "mulx rcx, rax, qword ptr [{b_ptr} + 16]",
+                "mulx rcx, rax, r15",
                 "add r11, rax",
                 "adcx r12, rcx",
                 "adc r13, 0",
+                "xor r14, r14",
 
                 // a1 * b3
-                "mulx rcx, rax, qword ptr [{b_ptr} + 24]",
+                "mulx rcx, rax, qword ptr [{a_ptr} + 24]",
                 "add r12, rax",
                 "adcx r13, rcx",
                 "adc r14, 0",
 
                 // `a2`
-                "mov rdx, [{a_ptr} + 16]",
+                "mov rdx, r15",
 
                 // a2 * b0
-                "mulx rcx, rax, qword ptr [{b_ptr} + 0]",
+                "mulx rcx, rax, qword ptr [{a_ptr} + 0]",
                 "add r10, rax",
                 "adcx r11, rcx",
                 "adc r12, 0",
 
                 // a2 * b1
-                "mulx rcx, rax, qword ptr [{b_ptr} + 8]",
+                "mulx rcx, rax, qword ptr [{a_ptr} + 8]",
                 "add r11, rax",
                 "adcx r12, rcx",
                 "adc r13, 0",
 
                 // a2 * b2
-                "mulx rcx, rax, qword ptr [{b_ptr} + 16]",
+                "mulx rcx, rax, r15",
                 "add r12, rax",
                 "adcx r13, rcx",
                 "adc r14, 0",
+                "xor r15, r15",
 
                 // a2 * b3
-                "mulx rcx, rax, qword ptr [{b_ptr} + 24]",
+                "mulx rcx, rax, qword ptr [{a_ptr} + 24]",
                 "adcx r13, rax",
                 "adcx r14, rcx",
                 "adc r15, 0",
@@ -480,30 +482,29 @@ impl Fr {
                 "mov rdx, [{a_ptr} + 24]",
 
                 // a3 * b0
-                "mulx rcx, rax, qword ptr [{b_ptr} + 0]",
+                "mulx rcx, rax, qword ptr [{a_ptr} + 0]",
                 "add r11, rax",
                 "adcx r12, rcx",
                 "adc r13, 0",
 
                 // a3 * b1
-                "mulx rcx, rax, qword ptr [{b_ptr} + 8]",
+                "mulx rcx, rax, qword ptr [{a_ptr} + 8]",
                 "adcx r12, rax",
                 "adcx r13, rcx",
                 "adc r14, 0",
 
                 // a3 * b2
-                "mulx rcx, rax, qword ptr [{b_ptr} + 16]",
+                "mulx rcx, rax, qword ptr [{a_ptr} + 16]",
                 "adcx r13, rax",
                 "adcx r14, rcx",
                 "adc r15, 0",
 
                 // a3 * b3
-                "mulx rcx, rax, qword ptr [{b_ptr} + 24]",
+                "mulx rcx, rax, qword ptr [{a_ptr} + 24]",
                 "adcx r14, rax",
                 "adc r15, rcx",
 
                 a_ptr = in(reg) self.0.as_ptr(),
-                b_ptr = in(reg) rhs.0.as_ptr(),
                 out("rax") _,
                 out("rcx") _,
                 out("rdx") _,
@@ -1015,11 +1016,6 @@ impl Fr {
                 options(pure, readonly, nostack)
             )
         }
-        if self.0 == [0, 0, 0, 0] {
-            println!("{:?}", MODULUS.0);
-            println!("{:?}", [r0, r1, r2, r3]);
-            println!("{:?}", r4);
-        }
         Self([r0, r1, r2, r3])
     }
 }
@@ -1274,7 +1270,6 @@ mod fr_tests {
         let b = a * a;
         assert!(b != Fr::one());
         let c = b * a;
-        println!("{:?}", c);
         assert!(c == Fr::one());
     }