lkml.org 
[lkml]   [2008]   [Mar]   [3]   [last100]   RSS Feed
Views: [more markup]  [less markup]  [headers]  [forward] 
 
Messages in this thread
/
DateMon, 3 Mar 2008 13:50:27 +0100
FromIngo Molnar <>
SubjectRe: [2.6 patch] acpi/battery.c: make 2 functions static
* Adrian Bunk <bunk@kernel.org> wrote:

> I can only repeat that I did state several times on linux-kernel that 
> it never worked.
> 
> If you consider it my fault that noone reads my emails then you are 
> right that it's my fault...

well, i'm trying to assume the best, so please explain the following 
sequence of events to me:

1) as you said you knew about this bug - which bug causes more inlining
   overhead than hundreds of your uninlining patches combined. The bug
   was introduced ~2 years ago in -mm - before the feature hit mainline
   in v2.6.16.

2) the fix was really trivial and the intention of the feature was well 
   understood - but the feature stayed as a NOP in the upstream kernel 
   for 2 years.

still, while you clearly had interest in this general area of the kernel 
(for example you wrote hundreds of tiny uninlining patches that work 
towards a similar goal), but strangely at the same time you neither 
fixed, nor properly escallated this _far_ bigger bug that causes +2.3% 
of text bloat on x86 [more than 120K of kernel text]. In fact:

- you created bugzillas for far smaller bugs in the past, but you never
  created a bugzilla for this that i'm aware of.

- you never directly raised this issue with us: "look guys, this thing
  really is broken - please reply to me with a fix".

- you never said "this is a regression that should be fixed" to any of
  the regression lists.

in other words: for about two years you knew about a bug that should 
have been fixed the day after it got introduced.

i obviously cannot know what your intentions were with this conduct, so 
i'm eagerly awaiting your explanation for it. Thanks,

	Ingo


\
 
 \ /
  Last update: 2008-03-03 13:53    [from the cache]
©2003-2008