Re: [PATCH] libpthread NOLOAD removal

[manu%netbsd.org@localhost (Emmanuel Dreyfus)]

Thor Lancelot Simon <tls%panix.com@localhost> wrote:

> "I picked a test that I knew wouldn't actually provoke any problems one
> might expect to arise, and since it didn't provoke them, COMMIT YESTERDAY!"
> 
> I'm sorry if that's not what you meant, but it sure sounds like it from
> the above.

I have the feeling you did not read my entire message. Please note three
points:

- There is no impact on a I/O bound operation like build.sh. Indeed we
expected it, but it is good to have a confirmation. 

- As I said, I also tested a billion loops around gmtime(), which is not
I/O bound, and got a significant performance hit of 1%.

- I don't know where you saw I was going to commit ASAP. I am just
posting patches for review and comments and I am quite suprised by your
reaction here. If you think the performance tests I ran are irrelevant,
please propose better ones.

-- 
Emmanuel Dreyfus
http://hcpnet.free.fr/pubz
manu%netbsd.org@localhost