Re: [PATCH RFC] Simplify the Linux kernel by reducing its state space

On Sat, 31 Mar 2012, Russell King - ARM Linux wrote:

On Sun, Apr 01, 2012 at 12:33:21AM +0800, Paul E. McKenney wrote:

quoted

Although there have been numerous complaints about the complexity of
parallel programming (especially over the past 5-10 years), the plain
truth is that the incremental complexity of parallel programming over
that of sequential programming is not as large as is commonly believed.
Despite that you might have heard, the mind-numbing complexity of modern
computer systems is not due so much to there being multiple CPUs, but
rather to there being any CPUs at all.  In short, for the ultimate in
computer-system simplicity, the optimal choice is NR_CPUS=0.

This commit therefore limits kernel builds to zero CPUs.  This change
has the beneficial side effect of rendering all kernel bugs harmless.
Furthermore, this commit enables additional beneficial changes, for
example, the removal of those parts of the kernel that are not needed
when there are zero CPUs.

Signed-off-by: Paul E. McKenney <redacted>
Reviewed-by: Thomas Gleixner <redacted>

Great work, but I don't think you've gone far enough with this.

What would really help is if you could consolidate all these NR_CPUS
definitions into one place so we don't have essentially the same thing
scattered across all these architectures.  We're already doing this on
ARM across our platforms, and its about time such an approach was taken
across the entire kernel tree.

It looks like the MIPS solution would be the best one to pick.
Could you rework your patch to do this please?

While you're at it, you might like to consider that having zero CPUs
makes all this architecture support redundant, so maybe you've missed
a trick there - according to my count, we could get rid of almost 3
million lines of code from arch.  We could replace all that with a
single standard implementation.

For a first step we can deprecated arch/ and make it depend on
CONFIG_STAGING. That way we can have it around a bit for sentimental
reasons w/o having a lot of churn.

Suggested-by: Russell King <redacted>
Signed-off-by: Thomas Gleixner <redacted>

Index: tip/Makefile
===================================================================

--- tip.orig/Makefile
+++ tip/Makefile

@@ -564,7 +564,9 @@ else
 KBUILD_CFLAGS	+= -O2
 endif
 
+ifdef CONFIG_ARCH_SUPPORT
 include $(srctree)/arch/$(SRCARCH)/Makefile
+endif
 
 ifneq ($(CONFIG_FRAME_WARN),0)
 KBUILD_CFLAGS += $(call cc-option,-Wframe-larger-than=${CONFIG_FRAME_WARN})

Index: tip/drivers/staging/Kconfig
===================================================================

--- tip.orig/drivers/staging/Kconfig
+++ tip/drivers/staging/Kconfig

@@ -1,6 +1,10 @@
+config ARCH_SUPPORT
+	bool
+
 menuconfig STAGING
 	bool "Staging drivers"
 	default n
+	select ARCH_SUPPORT
 	---help---
 	  This option allows you to select a number of drivers that are
 	  not of the "normal" Linux kernel quality level.  These drivers

Index: tip/Documentation/feature-removal-schedule.txt
===================================================================

--- tip.orig/Documentation/feature-removal-schedule.txt
+++ tip/Documentation/feature-removal-schedule.txt

@@ -537,3 +537,13 @@ When:	3.6
 Why:	setitimer is not returning -EFAULT if user pointer is NULL. This
 	violates the spec.
 Who:	Sasikantha Babu <sasikanth.v19@gmail.com>
+
+-----------------------------
+
+What:	Remove arch
+When:	April 1st 2013
+Why:    NR_CPUS=0 made arch/ obsolete. Keep it around a bit for
+	sentimental reasons.
+Who:	paulmck,tglx.rmk
+
+