[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20181107184435.GA168339@google.com>
Date: Wed, 7 Nov 2018 10:44:37 -0800
From: Brian Norris <briannorris@...omium.org>
To: lists@...dbynature.de
Cc: sky@...ki.is, Guenter Roeck <linux@...ck-us.net>,
yamada.masahiro@...ionext.com,
Linux Kernel <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] Revert "scripts/setlocalversion: git: Make -dirty check
more robust"
On Tue, Nov 06, 2018 at 08:00:36PM -0800, Brian Norris wrote:
> On a different tangent: how about the --no-optional-locks (see
> git(1))? Will this get you your "up-to-date" result without writing to
> the .git directory? I've only read the documentation, but not tested
> it.
I think I can add a little to this: that option definitely avoids
writing to the git index (thus, avoiding the problem that Guenter and I
are trying to resolve) and the documentation suggests it should still
get up-to-date index information (thus, avoiding the problem that Genki
was trying to fix). So something like this would work (on top of the
revert):
diff --git a/scripts/setlocalversion b/scripts/setlocalversion
index 71f39410691b..7dfe45badcca 100755
--- a/scripts/setlocalversion
+++ b/scripts/setlocalversion
@@ -74,7 +74,7 @@ scm_version()
fi
# Check for uncommitted changes
- if git diff-index --name-only HEAD | grep -qv "^scripts/package"; then
+ if git --no-optional-locks status -uno --porcelain | grep -qv '^.. scripts/package'; then
printf '%s' -dirty
fi
Unfortunately, --no-optional-locks is new as of git 2.14. Dunno how new
of a git we expect people to use.
Brian
Powered by blists - more mailing lists