From defd2e8128d0caace4622286f4157646d121bc46 Mon Sep 17 00:00:00 2001
From: Joel Rosdahl <joel@rosdahl.net>
Date: Thu, 11 Apr 2019 22:10:43 +0200
Subject: [PATCH] Rewrite mdfour routines to not modify state when computing
 the result
MIME-Version: 1.0
Content-Type: text/plain; charset=utf8
Content-Transfer-Encoding: 8bit

This makes the interface much more intuitive, at the minor expense of
having to copy 32-96 extra bytes (unnecessary if the hash state won’t be
fed with more bytes) when fetching the result.

This change is also partly motivated by the code that handles several
-arch options – it calls get_object_name_from_cpp multiple times to
compute a combined hash, but that function also computes the hash result
each time.
---
 src/hash.c           |  9 ++-------
 src/mdfour.c         | 29 ++++++++++++++++-------------
 src/mdfour.h         |  3 +--
 unittest/test_hash.c | 12 +++++++++++-
 4 files changed, 30 insertions(+), 23 deletions(-)

diff --git a/src/hash.c b/src/hash.c
index ae89e85fe..bf067620a 100644
--- a/src/hash.c
+++ b/src/hash.c
@@ -32,10 +32,6 @@ do_hash_buffer(struct hash *hash, const void *s, size_t len)
 {
 	assert(s);
 
-	// The hash state cannot be updated after the result has been fetched via
-	// hash_result/hash_result_as_bytes.
-	assert(!hash->md.finalized);
-
 	mdfour_update(&hash->md, (const unsigned char *)s, len);
 	if (len > 0 && hash->debug_binary) {
 		(void) fwrite(s, 1, len, hash->debug_binary);
@@ -90,7 +86,7 @@ void hash_enable_debug(
 size_t
 hash_input_size(struct hash *hash)
 {
-	return hash->md.totalN;
+	return hash->md.totalN + hash->md.tail_len;
 }
 
 void
@@ -106,13 +102,12 @@ hash_result(struct hash *hash)
 	unsigned char sum[16];
 
 	hash_result_as_bytes(hash, sum);
-	return format_hash_as_string(sum, (unsigned) hash->md.totalN);
+	return format_hash_as_string(sum, hash_input_size(hash));
 }
 
 void
 hash_result_as_bytes(struct hash *hash, unsigned char *out)
 {
-	mdfour_update(&hash->md, NULL, 0);
 	mdfour_result(&hash->md, out);
 }
 
diff --git a/src/mdfour.c b/src/mdfour.c
index 11e0fba1e..c0aec6bc2 100644
--- a/src/mdfour.c
+++ b/src/mdfour.c
@@ -1,5 +1,5 @@
 // Copyright (C) 1997-1998 Andrew Tridgell
-// Copyright (C) 2009-2018 Joel Rosdahl
+// Copyright (C) 2009-2019 Joel Rosdahl
 //
 // This program is free software; you can redistribute it and/or modify it
 // under the terms of the GNU General Public License as published by the Free
@@ -129,7 +129,6 @@ mdfour_begin(struct mdfour *md)
 	md->D = 0x10325476;
 	md->totalN = 0;
 	md->tail_len = 0;
-	md->finalized = 0;
 }
 
 static
@@ -160,13 +159,7 @@ void mdfour_tail(struct mdfour *md, const unsigned char *in, size_t n)
 void
 mdfour_update(struct mdfour *md, const unsigned char *in, size_t n)
 {
-	if (!in) {
-		if (!md->finalized) {
-			mdfour_tail(md, md->tail, md->tail_len);
-			md->finalized = 1;
-		}
-		return;
-	}
+	assert(in);
 
 	uint32_t M[16];
 	if (md->tail_len) {
@@ -203,8 +196,18 @@ mdfour_update(struct mdfour *md, const unsigned char *in, size_t n)
 void
 mdfour_result(struct mdfour *md, unsigned char *out)
 {
-	copy4(out, md->A);
-	copy4(out+4, md->B);
-	copy4(out+8, md->C);
-	copy4(out+12, md->D);
+	struct mdfour result;
+	result.A = md->A;
+	result.B = md->B;
+	result.C = md->C;
+	result.D = md->D;
+	result.totalN = md->totalN;
+	result.tail_len = md->tail_len;
+	memcpy(result.tail, md->tail, result.tail_len);
+
+	mdfour_tail(&result, result.tail, result.tail_len);
+	copy4(out, result.A);
+	copy4(out+4, result.B);
+	copy4(out+8, result.C);
+	copy4(out+12, result.D);
 }
diff --git a/src/mdfour.h b/src/mdfour.h
index c196a09e9..761a19ee5 100644
--- a/src/mdfour.h
+++ b/src/mdfour.h
@@ -7,9 +7,8 @@
 struct mdfour {
 	uint32_t A, B, C, D;
 	size_t totalN;
-	unsigned char tail[64];
 	size_t tail_len;
-	int finalized;
+	unsigned char tail[64];
 };
 
 void mdfour_begin(struct mdfour *md);
diff --git a/unittest/test_hash.c b/unittest/test_hash.c
index 5278560e2..2e9502dfd 100644
--- a/unittest/test_hash.c
+++ b/unittest/test_hash.c
@@ -1,4 +1,4 @@
-// Copyright (C) 2010-2018 Joel Rosdahl
+// Copyright (C) 2010-2019 Joel Rosdahl
 //
 // This program is free software; you can redistribute it and/or modify it
 // under the terms of the GNU General Public License as published by the Free
@@ -56,6 +56,16 @@ TEST(test_vectors_from_rfc_1320_should_be_correct)
 	}
 }
 
+TEST(hash_result_should_not_alter_state)
+{
+	struct hash *h = hash_init();
+	hash_string(h, "message");
+	free(hash_result(h));
+	hash_string(h, " digest");
+	CHECK_STR_EQ_FREE2("d9130a8164549fe818874806e1c7014b-14", hash_result(h));
+	hash_free(h);
+}
+
 TEST(hash_result_should_be_idempotent)
 {
 	struct hash *h = hash_init();
-- 
2.47.2