From f9efc3aedee5069c8a96fa12c79b94ca2860b5d8 Mon Sep 17 00:00:00 2001
From: Dmitry Vagner <typecat@Dmitrys-MacBook-Pro.local>
Date: Tue, 9 Aug 2022 15:11:18 -0700
Subject: [PATCH] draft implementation

---
 .../asm/ripemd/{iterate.asm => hashloop.asm}  |  0
 evm/src/cpu/kernel/asm/ripemd/subroutines.asm | 84 +++++++++++++++++--
 2 files changed, 79 insertions(+), 5 deletions(-)
 rename evm/src/cpu/kernel/asm/ripemd/{iterate.asm => hashloop.asm} (100%)

diff --git a/evm/src/cpu/kernel/asm/ripemd/iterate.asm b/evm/src/cpu/kernel/asm/ripemd/hashloop.asm
similarity index 100%
rename from evm/src/cpu/kernel/asm/ripemd/iterate.asm
rename to evm/src/cpu/kernel/asm/ripemd/hashloop.asm
diff --git a/evm/src/cpu/kernel/asm/ripemd/subroutines.asm b/evm/src/cpu/kernel/asm/ripemd/subroutines.asm
index 05cf0c15..1590e515 100644
--- a/evm/src/cpu/kernel/asm/ripemd/subroutines.asm
+++ b/evm/src/cpu/kernel/asm/ripemd/subroutines.asm
@@ -1,3 +1,62 @@
+/// Note that Fj, Kj last for 16 iterations, but sj, rj update each call
+///
+/// def R(a, b, c, d, e, Fj, Kj, sj, rj, X):
+///     a = u32(ROL(sj, u32(Fj(b, c, d) + a + X[rj] + Kj)) + e)
+///     c = ROL(10, c)
+///     return e, a, b, c, d
+
+global R:
+    jumpdest
+    // stack: a, b, c, d, e, Fj, Kj, retdest
+    dup4
+    // stack: d, a, b, c, d, e, Fj, Kj, retdest
+    dup4
+    // stack: c, d, a, b, c, d, e, Fj, Kj, retdest
+    dup4 
+    // stack: b, c, d, a, b, c, d, e, Fj, Kj, retdest
+    dup9
+    // stack: Fj, b, c, d, a, b, c, d, e, Fj, Kj, retdest
+    jump---------------------------------------------------------------------------TODO
+    // stack: Fj(b, c, d), a, b, c, d, e, Fj, Kj, retdest
+    add
+    // stack: Fj(b, c, d) + a, b, c, d, e, Fj, Kj, retdest
+    push X[rj]---------------------------------------------------------------------TODO
+    // stack: X[rj], Fj(b, c, d) + a, b, c, d, e, Fj, Kj, retdest
+    add
+    // stack: X[rj] + Fj(b, c, d) + a, b, c, d, e, Fj, Kj, retdest
+    dup7
+    // stack: Kj, X[rj] + Fj(b, c, d) + a, b, c, d, e, Fj, Kj, retdest
+    add
+    // stack: Kj + X[rj] + Fj(b, c, d) + a, b, c, d, e, Fj, Kj, retdest
+    %u32
+    // stack: Kj + X[rj] + Fj(b, c, d) + a, b, c, d, e, Fj, Kj, retdest
+    push sj------------------------------------------------------------------------TODO
+    // stack: sj, Kj + X[rj] + Fj(b, c, d) + a, b, c, d, e, Fj, Kj, retdest
+    push ROL
+    // stack: ROL, sj, Kj + X[rj] + Fj(b, c, d) + a, b, c, d, e, Fj, Kj, retdest
+    jump---------------------------------------------------------------------------TODO
+    // stack: ROL(sj, Kj + X[rj] + Fj(b, c, d) + a), b, c, d, e, Fj, Kj, retdest
+    dup5
+    // stack: e, ROL(sj, Kj + X[rj] + Fj(b, c, d) + a), b, c, d, e, Fj, Kj, retdest
+    add
+    // stack: e + ROL(sj, Kj + X[rj] + Fj(b, c, d) + a), b, c, d, e, Fj, Kj, retdest
+    %u32
+    // stack: e + ROL(sj, Kj + X[rj] + Fj(b, c, d) + a), b, c, d, e, Fj, Kj, retdest
+    swap2
+    // stack: c, b, e + ROL(sj, Kj + X[rj] + Fj(b, c, d) + a), d, e, Fj, Kj, retdest
+    push 10
+    // stack: 10, c, b, e + ROL(sj, Kj + X[rj] + Fj(b, c, d) + a), d, e, Fj, Kj, retdest
+    push ROL
+    // stack: ROL, 10, c, b, e + ROL(sj, Kj + X[rj] + Fj(b, c, d) + a), d, e, Fj, Kj, retdest
+    jump---------------------------------------------------------------------------TODO
+    // stack: ROL(10, c), b, e + ROL(sj, Kj + X[rj] + Fj(b, c, d) + a), d, e, Fj, Kj, retdest
+    %stack (c, b, a, d, e) -> (e, a, b, c, d)
+    // stack: e, e + ROL(sj, Kj + X[rj] + Fj(b, c, d) + a), b, ROL(10, c), d, e, Fj, Kj, retdest
+
+
+/// def ROL(n, x):
+///     return (u32(x << n)) | (x >> (32 - n))
+
 global Rol:
     jumpdest
     // stack: n, x, retdest
@@ -28,6 +87,9 @@ global Rol:
     jump
 
 
+/// def F0(x, y, z):
+///     return x ^ y ^ z
+
 global F0:
     jumpdest
     // stack: x, y, z, retdest
@@ -40,6 +102,9 @@ global F0:
     jump
 
 
+/// def F1(x, y, z):
+///     return (x & y) | (u32(~x) & z)
+
 global F1:
     jumpdest
     // stack: x, y, z, retdest
@@ -53,7 +118,7 @@ global F1:
     // stack: z, x, y & x, retdest
     swap1
     // stack: x, z, y & x, retdest
-    %not_u32
+    %not_32
     // stack: ~x, z, y & x, retdest
     and
     // stack: ~x & z, y & x, retdest
@@ -64,22 +129,28 @@ global F1:
     jump
 
 
+/// def F2(x, y, z):
+///     return (x | u32(~y)) ^ z
+
 global F2:
     jumpdest
     // stack: x, y, z, retdest
     swap1
     // stack: y, x, z, retdest
-    %not_u32
+    %not_32
     // stack: ~y, x, z, retdest
     or
     // stack: ~y | x, z, retdest
     xor
-     // stack: (~y | x) ^ z, retdest
+    // stack: (~y | x) ^ z, retdest
     swap1
     // stack: retdest, (~y | x) ^ z
     jump
 
 
+/// def F3(x, y, z):
+///     return (x & z) | (u32(~z) & y)
+
 global F3:
     jumpdest
     // stack: x, y, z, retdest
@@ -89,7 +160,7 @@ global F3:
     // stack: z & x, y, z, retdest
     swap2
     // stack: z, y, z & x, retdest
-    %not_u32
+    %not_32
     // stack: ~z, y, z & x, retdest
     and
     // stack: ~z & y, z & x, retdest
@@ -100,12 +171,15 @@ global F3:
     jump 
 
 
+/// def F4(x, y, z):
+///     return x ^ (y | u32(~z))
+
 global F4:
     jumpdest 
     // stack: x, y, z, retdest
     swap2
     // stack: z, y, x, retdest
-    %not_u32
+    %not_32
     // stack: ~z, y, x, retdest
     or
     // stack: ~z | y, x, retdest