From e6ec6773c4ad10d6b4f260b3a2086547f2153ef1 Mon Sep 17 00:00:00 2001
From: Matt Caswell <matt@openssl.org>
Date: Thu, 25 May 2017 15:18:22 +0100
Subject: [PATCH] Document that HMAC() with a NULL md is not thread safe

Fixes #3541

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3554)
---
 doc/crypto/HMAC.pod | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/doc/crypto/HMAC.pod b/doc/crypto/HMAC.pod
index 87f7e33086..219c9ba208 100644
--- a/doc/crypto/HMAC.pod
+++ b/doc/crypto/HMAC.pod
@@ -57,7 +57,8 @@ B<key_len> bytes long.
 It places the result in B<md> (which must have space for the output of
 the hash function, which is no more than B<EVP_MAX_MD_SIZE> bytes).
 If B<md> is NULL, the digest is placed in a static array.  The size of
-the output is placed in B<md_len>, unless it is B<NULL>.
+the output is placed in B<md_len>, unless it is B<NULL>. Note: passing a NULL
+value for B<md>  to use the static array is not thread safe.
 
 B<evp_md> can be EVP_sha1(), EVP_ripemd160() etc.
 
-- 
2.25.1