[tor-commits] [tor/master] Merge branch 'more_aes_hackery_rebased'

nickm at torproject.org nickm at torproject.org
Fri Nov 25 15:41:44 UTC 2011


commit 25c9e3aab9237b9d01229eb9bdb73450cc4ccfcb
Merge: fcf5c92 fbec45c
Author: Nick Mathewson <nickm at torproject.org>
Date:   Fri Nov 25 10:36:13 2011 -0500

    Merge branch 'more_aes_hackery_rebased'
    
    Conflicts:
    	changes/aes_hackery

 changes/aes_hackery    |   10 +++
 src/common/aes.c       |  208 +++++++++++++++++++++++++++++++++---------------
 src/common/aes.h       |    2 +
 src/common/crypto.c    |    3 +
 src/test/test_crypto.c |   33 +++++++-
 5 files changed, 186 insertions(+), 70 deletions(-)

diff --cc changes/aes_hackery
index 0000000,b22cefe..8b2ef1a
mode 000000,100644..100644
--- a/changes/aes_hackery
+++ b/changes/aes_hackery
@@@ -1,0 -1,16 +1,10 @@@
 -  o Code simplification and refactoring:
 -    - Unconditionally use OpenSSL's AES implementation instead of our
 -      old built-in one. OpenSSL's AES has been better for a while, and
 -      relatively few servers should still be on any version of OpenSSL
 -      that doesn't have good optimized assembly AES.
 -
+   o Major features (AES performance):
 -    - Use OpenSSL's EVP interface for AES encryption, so that all
 -      AES operations can use hardware acceleration (if present).
 -      Resolves issue #4442.
 -    - But only use the EVP interface when AES acceleration is enabled,
 -      to avoid a performance regression.  Resolves issue #4525.
+     - When using OpenSSL 1.0.0 or later, use OpenSSL's counter mode
+       implementation; it makes AES_CTR about 7% faster than our old one
+       (which was about 10% faster than the one OpenSSL used to provide).
+       Resolves issue #4526.
++
++  o Major bugfixes (AES performance):
++    - Only use the EVP interface when AES acceleration is enabled,
++      to avoid a 5-7% performance regression.  Resolves issue #4525,
++      bugfix on 0.2.3.8-alpha.



More information about the tor-commits mailing list