From d89fa814c17d6b957f2e296cd57c4941839049b3 Mon Sep 17 00:00:00 2001
From: Yann Collet <cyan@fb.com>
Date: Tue, 28 Aug 2018 18:19:19 -0700
Subject: [PATCH] added a README

for documentation
---
 contrib/largeNbDicts/README.md | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)
 create mode 100644 contrib/largeNbDicts/README.md

diff --git a/contrib/largeNbDicts/README.md b/contrib/largeNbDicts/README.md
new file mode 100644
index 00000000..7eba229a
--- /dev/null
+++ b/contrib/largeNbDicts/README.md
@@ -0,0 +1,19 @@
+largeNbDicts
+=====================
+
+`largeNbDicts` is a benchmark test tool
+dedicated to the specific scenario of
+dictionary decompression using a very large number of dictionaries,
+which suffers from increased latency due to cache misses.
+It's created in a bid to investigate performance for this scenario,
+and experiment mitigation techniques.
+
+Command line :
+```
+$ largeNbDicts filename [Options]
+Options :
+--clevel=#     : use compression level # (default: 3)
+--blockSize=#  : cut input into blocks of size # (default: 4096)
+--dictionary=# : use # as a dictionary (default: create one)
+--nbDicts=#    : set nb of dictionaries to # (default: one per block)
+```