WTF?

[Michael Matz]

Hi,

On Wed, 25 Nov 2009, Richard Kenner wrote:

> > And local patches.  Basically _no_ patch will apply anymore as HJ changed 
> > every single file.  
> 
> That's an exaggeration since only a few lines in each file were change. 

The top 12 from the 2.2 MB patch:

 tree-vect-loop.c           |  386 ++++++------
 tree-vect-slp.c            |  394 ++++++------
 ipa-type-escape.c          |  432 +++++++-------
 tree-vect-data-refs.c      |  462 +++++++--------
 df-problems.c              |  552 +++++++++---------
 tree-data-ref.c            |  558 +++++++++---------
 tree-scalar-evolution.c    |  576 +++++++++---------
 df-scan.c                  |  580 +++++++++---------
 tree-vect-stmts.c          |  618 ++++++++++----------
 ipa-struct-reorg.c         |  706 +++++++++++------------
 sel-sched-ir.c             |  732 +++++++++++------------
 sel-sched.c                | 1376 ++++++++++++++++++++++---------------------

> The vast majority of outstanding patches won't be affected.

Do you offer fixing my patches in case I'm becoming too annoyed to fix the 
conflicts?  I guess not, but I have some queued for stage 1, which don't 
apply anymore.

As for us having allowed such massive changes in the past: e.g.
  http://gcc.gnu.org/ml/gcc/2007-03/msg00417.html (and the surrounding 
thread) or http://gcc.gnu.org/ml/gcc-patches/2009-06/msg00561.html say 
something different.

In my mind it's very simple: trailing whitespace poses exactly _no_ 
problem (except of being against the coding standard), hence removing them 
should be weighed fairly aggressively against any disadvantage.  Basically 
if even just one patch in the wild could possibly be broken by a 
whitespace change, that wouldn't also be broken by non-whitespace, then it 
simply should not be done.

But I don't want to discuss whitespace policies actually, I want that HJ 
follows patch submission policies, and I want this checkin be reverted or 
removed.


Ciao,
Michael.