[PATCH] feature-removal: arch code lacking CONFIG_HAVE_ARCH_TRACEHOOK

[PATCH] feature-removal: arch code lacking CONFIG_HAVE_ARCH_TRACEHOOK

[Roland McGrath]

This documents the intent for generic code to stop catering to old
arch code that has not been updated to the various new interfaces
that are indicated by 'select HAVE_ARCH_TRACEHOOK' in Kconfig.  The
most-used arch's have already converted and some others are starting
to do so.  Help for arch maintainers is available from me and Oleg
if they ask.  (I already submitted 95% of the work for arch/arm,
though it has not gotten any feedback from those arch maintainers.)

The ptrace maintainers and people working on other new features for
user task debugging and tracing need the freedom to reorganize the
generic code and make it depend on the modern set of arch interfaces.
With HAVE_ARCH_TRACEHOOK, such work won't need arch-specific changes.

Signed-off-by: Roland McGrath <roland@redhat.com>
---
 Documentation/feature-removal-schedule.txt |   14 ++++++++++++++
 1 files changed, 14 insertions(+), 0 deletions(-)

diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt
index de491a3..7a52b94 100644
--- a/Documentation/feature-removal-schedule.txt
+++ b/Documentation/feature-removal-schedule.txt
@@ -437,3 +437,17 @@ Why:	Superseded by tdfxfb. I2C/DDC support used to live in a separate
 	driver but this caused driver conflicts.
 Who:	Jean Delvare <khali@linux-fr.org>
 	Krzysztof Helt <krzysztof.h1@wp.pl>
+
+---------------------------
+
+What:	arch ptrace code not converted to CONFIG_HAVE_ARCH_TRACEHOOK
+When:	2.6.32
+Why:	Crufty old arch code holds up work on generic ptrace code,
+	clean-ups, and new features in the user debugging/tracing area.
+	After the cutoff, generic ptrace code will not try to cater to
+	old arch code, and your arch builds might break if not updated.
+	See arch/Kconfig comments for what code each arch should supply
+	to enable HAVE_ARCH_TRACEHOOK.  Assistance with updating arch
+	code is available for the asking.
+Who:	Roland McGrath <roland@redhat.com>
+	Oleg Nesterov <oleg@redhat.com>

Re: [PATCH] feature-removal: arch code lacking CONFIG_HAVE_ARCH_TRACEHOOK

[Matthew Wilcox]

On Thu, May 14, 2009 at 12:13:56PM -0700, Roland McGrath wrote:
> This documents the intent for generic code to stop catering to old
> arch code that has not been updated to the various new interfaces
> that are indicated by 'select HAVE_ARCH_TRACEHOOK' in Kconfig.  The
> most-used arch's have already converted and some others are starting
> to do so.  Help for arch maintainers is available from me and Oleg
> if they ask.  (I already submitted 95% of the work for arch/arm,
> though it has not gotten any feedback from those arch maintainers.)
> 
> The ptrace maintainers and people working on other new features for
> user task debugging and tracing need the freedom to reorganize the
> generic code and make it depend on the modern set of arch interfaces.
> With HAVE_ARCH_TRACEHOOK, such work won't need arch-specific changes.

This really isn't how feature-removal-schedule has been used in the past.

Let's just take a look at this:

Converted
~~~~~~~~~

ia64
powerpc
s390
sh
sparc
x86

Not Converted (in mainline)
~~~~~~~~~~~~~~~~~~~~~~~~~~~

alpha
arm (you mention above you're waiting for the maintainer)
avr32
blackfin
cris
frv
h8300
m32r
m68k
m68knommu
microblaze
mips
mn10300
parisc
um
xtensa

so 6 converted, 16 not.  I don't think you're at the stage where you
can bully people into converting yet.

-- 
Matthew Wilcox				Intel Open Source Technology Centre
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours.  We can't possibly take such
a retrograde step."