[tor-talk] Tor and AES-NI acceleration , and Tor profiling

[Moritz Bartl]

>> Hmm.  On examination it looks like there might be some uses of
>> OpenSSL's AES_encrypt function left around in your profile.  Try
>> changing the beginning of Tor's aes.c so that the line that now says:
> Curious whether you had a chance to try this.  From   the latest
> profile, your biggest timesink is _x86_64_AES_encrypt_compact.  But
> that's not the AESNI version.  I think that using EVP there would get
> the right result, if I understand how engines work correctly.

Sorry for the delay. To be honest I didn't find the time to compile it 
on my own using the debian ruleset, and Sebastian told me not to use the 
standard compiles to replace the packaged version.

The node has been lifted to 0.2.3.8 now and oprofile resetted.

opreport -g -l /usr/sbin/tor
http://axigy1.torservers.net/opreport.txt

opcontrol --dump && opreport -l
http://axigy1.torservers.net/opdump.txt

-- 
Moritz Bartl
https://www.torservers.net/