Hello Dominic,
Sunday, July 15, 2007, 11:44:10 PM, you wrote:
forget it if you're interested in performance near a C implementation
such as GNU sha1sum.
I don't think it's unreasonable to think we could get near to C performance
and
we've been getting closer.
btw, if someone interested
Malte Milatz malte at gmx-topmail.de writes:
See the following link for a purely functional and straight-forward
implementation of SHA1. Disclaimer: Please be kind to me, I haven't
done much Haskell (yet). And I know nothing about SHA1 except its
specification.
Dominic Steinitz:
Malte Milatz:
http://hpaste.org/1695#a2
It performs better than the SHA1 algorithm in Crypto: It is faster by a
factor of approximately e. It is also competitive (regarding time)
with the »unsafe« SHA1 implementation posted here some days ago,
Great. Can you
I wrote:
Before I started tweaking (with the help of the IRC channel), I had read
the input in as a ByteString, then unpacked it to [Word8] for further
processing.
I forgot to mention that I still used ByteString.length here. That may
be a very important factor.
Malte
this is my current one
http://hpaste.org/1704
doing a sum with yours (2nd one on http://hpaste.org/1695#a2) on a 4mb mp3 i got
real0m33.147s
user0m13.129s
sys 0m0.724s
with mine i got
real0m1.262s
user0m1.096s
sys 0m0.120s
On 7/15/07, Malte Milatz [EMAIL PROTECTED]
you probably got crappy timings with mine because it was compiled with
-prof, and without all the optimization options. you should try some
of them on your implementation. -O2 -fexcess-precision
-funbox-strict-fields gave me a significant boost.
On 7/15/07, Anatoly Yakovenko [EMAIL PROTECTED]
Anatoly Yakovenko:
you probably got crappy timings with mine because it was compiled with
-prof, and without all the optimization options.
Alright, with your current version and without -prof I get
real0m1.176s
user0m1.104s
sys 0m0.056s
while mine doesn't get better than (thanks