[patch] Lno branch merge part 12 -- induction variable optimizations

[Andrew Pinski]

On Sep 2, 2004, at 7:50 PM, Mark Mitchell wrote:

> I'm not sure what you're saying.
>
> I would like to approve Zdenek's patch, but I'm not nearly as expert 
> in this area of the compiler as other people.  So, I want to give them 
> a chance to speak up; it's hard to take things out after they go in.  
> Are you saying that you think I have an obligation to approve the 
> patch, even though I think there are other people who might have 
> useful information to add?

What I am trying to say is that you should not have conditionally 
approved
it based on the date which is right after the start of Stage3.  By 
doing that
you make anyone else who would have approved it earlier as it is already
approved.

Also if you were so worry about timing of stage3 and number of projects 
going
in for 3.5.0, you should have asked for Projects to be submitted in 
stage1
while things were in development.  I should note that when Apple was 
doing
strees, you asked for a developmental plan[1] but you did not do the 
same for
the LNO, the tree-ssa or the tree-profiling folks.  I think you are 
being
one sided here which is wrong as a release manager (or maybe it was 
because
effected "your" front-end unlike all other improvements to GCC which is 
still
all one sidedness).

Really I think the situational here just sucks and could have been 
planed
way in advanced of stage3 instead of three weeks.  Maybe the SC should
reconsider you as the release manager as you are not doing your job 
except
when it comes to the end of stage2. (Yes I am know that you are part of 
the
SC but this is just wrong as it is conflict of interest).

Thanks,
Andrew Pinski
Not speaking as an employee of Apple but someone who came here to help

[1] http://gcc.gnu.org/ml/gcc-patches/2004-03/msg00756.html