Skip to contentSkip to conversation

Appending '+' to version, since 2.6.35-rc2

PATCH Mark Hills 3 participants 4 messages Jun 13, 2010 – Jun 15, 2010 Kbuild

lore
Share & tools
lore

Mark Hills

author
#0 lore raw

Commit 85a256d adds a new feature:

[...] When CONFIG_LOCALVERSION_AUTO is not enabled, a `+' is appended to the kernel version to represent that the kernel has been revised since the last release unless "make LOCALVERSION=" was used to uniquely identify the build.

I'm finding this inconvenient for 'casual' kernel development.

For example, my usual workflow goes something like:

1) find something I suspect is a bug
2) upgrade to the latest stable or -rc kernel, to confirm
3) experiment with patch to fix it

Since 2.6.35-rc2, going from step 2 to 3 causes the version of the kernel to change: forcing time-consuming rebuilds, installing modules, and manual changes associated with a new version number; eg. updating the lilo/grub config, or the distribution's initrd.

Whilst I can understand the motivation for the '+' being mandatory, the added stages above unnecessarily raise the bar for casual developers, or other people who need to 'dip' into the kernel and test patches. These people are least likely to be aware of using LOCALVERSION to override it too.

May I suggest that this particular part of the commit be reverted (example below)? In favour of a user, if they wish, explicitly setting LOCALVERSION or using CONFIG_LOCALVERSION_AUTO.

Signature · 12 lines
Mark


From: Mark Hills <[email protected]>
Date: Sun, 13 Jun 2010 19:16:55 +0100
Subject: [PATCH] kbuild: Do not append mandatory '+' to kernel release


Signed-off-by: Mark Hills <[email protected]>
---
 Makefile |   14 --------------
 1 files changed, 0 insertions(+), 14 deletions(-)
Makefile +0 -14
index d49d96c..6870f6f 100644
@@ -909,14 +909,6 @@ $(vmlinux-dirs): prepare scripts
 #	    $(scm-identifier)		(unique SCM tag, if one exists)
 #	      ./scripts/setlocalversion	(only with CONFIG_LOCALVERSION_AUTO)
 #	      .scmversion		(only with CONFIG_LOCALVERSION_AUTO)
-#	    +				(only without CONFIG_LOCALVERSION_AUTO
-#					 and without LOCALVERSION= and
-#					 repository is at non-tagged commit)
-#
-# For kernels without CONFIG_LOCALVERSION_AUTO compiled from an SCM that has
-# been revised beyond a tagged commit, `+' is appended to the version string
-# when not overridden by using "make LOCALVERSION=".  This indicates that the
-# kernel is not a vanilla release version and has been modified.
 
 pattern = ".*/localversion[^~]*"
 string  = $(shell cat /dev/null \
@@ -942,12 +934,6 @@ endif
 
 ifdef CONFIG_LOCALVERSION_AUTO
 	localver-extra = $(scm-identifier)
-else
-	ifneq ($(scm-identifier),)
-		ifeq ($(LOCALVERSION),)
-			localver-extra = +
-		endif
-	endif
 endif
 
 localver-full = $(localver)$(LOCALVERSION)$(localver-extra)
Signature · 1 line
1.7.1
#1 lore raw
Quoted 9 lines · Mark Hills

Commit 85a256d adds a new feature:

[...] When CONFIG_LOCALVERSION_AUTO is not enabled, a `+' is appended to the kernel version to represent that the kernel has been revised since the last release unless "make LOCALVERSION=" was used to uniquely identify the build.

I'm finding this inconvenient for 'casual' kernel development.

Quoted 8 lines

For example, my usual workflow goes something like:

1) find something I suspect is a bug
2) upgrade to the latest stable or -rc kernel, to confirm
3) experiment with patch to fix it

Since 2.6.35-rc2, going from step 2 to 3 causes the version of the kernel to change: forcing time-consuming rebuilds,

init/version.c
kernel/configs.c
+ whatever patch has changed
+ vmlinux 2 or 3 times

That's not time consuming.

Quoted 3 lines

installing modules, and manual
changes associated with a new version number; eg. updating the lilo/grub config, or the distribution's initrd.

author
#2 lore raw
Quoted 19 lines · Alexey Dobriyan
Quoted 9 lines · Mark Hills

Commit 85a256d adds a new feature:

[...] When CONFIG_LOCALVERSION_AUTO is not enabled, a `+' is appended to the kernel version to represent that the kernel has been revised since the last release unless "make LOCALVERSION=" was used to uniquely identify the build.

I'm finding this inconvenient for 'casual' kernel development.

Quoted 8 lines

For example, my usual workflow goes something like:

1) find something I suspect is a bug
2) upgrade to the latest stable or -rc kernel, to confirm
3) experiment with patch to fix it

Since 2.6.35-rc2, going from step 2 to 3 causes the version of the kernel to change: forcing time-consuming rebuilds,

Quoted 6 lines

init/version.c
kernel/configs.c
+ whatever patch has changed
+ vmlinux 2 or 3 times

That's not time consuming.

There is also the modpost step, which is most time consuming.

Quoted 5 lines
Quoted 4 lines

installing modules, and manual
changes associated with a new version number; eg. updating the lilo/grub config, or the distribution's initrd.

Signature · 1 line
Mark
#3 lore raw
In reply to Mark Hills
Quoted 11 lines · Mark Hills

Commit 85a256d adds a new feature:

[...] When CONFIG_LOCALVERSION_AUTO is not enabled, a `+' is appended to the kernel version to represent that the kernel has been revised since the last release unless "make LOCALVERSION=" was used to uniquely identify the build.

I'm finding this inconvenient for 'casual' kernel development.

For example, my usual workflow goes something like:

FWIW, I agree with this.
If I wanted a version suffix of this sort I'd just use CONFIG_LOCALVERSION_AUTO.

Benny

Quoted 19 lines

1) find something I suspect is a bug
2) upgrade to the latest stable or -rc kernel, to confirm
3) experiment with patch to fix it

Since 2.6.35-rc2, going from step 2 to 3 causes the version of the kernel to change: forcing time-consuming rebuilds, installing modules, and manual changes associated with a new version number; eg. updating the lilo/grub config, or the distribution's initrd.

Whilst I can understand the motivation for the '+' being mandatory, the added stages above unnecessarily raise the bar for casual developers, or other people who need to 'dip' into the kernel and test patches. These people are least likely to be aware of using LOCALVERSION to override it too.

May I suggest that this particular part of the commit be reverted (example below)? In favour of a user, if they wish, explicitly setting LOCALVERSION or using CONFIG_LOCALVERSION_AUTO.