Re: Stack Smash Protection disabled

[Jean-Yves Migeon <jeanyves.migeon%free.fr@localhost>]

Martin Husemann wrote:
> There are two mostly orthogonal questions:
>
>  - would we be willing to spend 5% performance for the level of 
additional
>    security we gain?
>
> Let's wait for core's vote, no need to discuss it again and again.
>
>  - the 5% number is way higher than expected - we need to examine why!
>
> Any ideas and more detailed benchmarks welcome.

FYI, I am planing to make a run for some benchmarks tomorrow evening, 
with and w/o SSP compiled in kernel. Probably libmicro and sysbench, at 
first glance.

Unless someone beats me to it, *shrug*

--
Jean-Yves Migeon
jeanyves.migeon%free.fr@localhost