Kbuild: Ignore GREP_OPTIONS env variable

Commit Message

Borislav Petkov Oct. 30, 2013, 4:06 p.m. UTC

Hi,

this is just a general poll of what people think about the issue here,
just so that at least we've talked about it.

So I had defined GREP_OPTIONS=--color=always on one of my boxes and had
forgotten about it and the kernel build started failing because we use
grep quite a while in the tree and it started issuing shell color markup
which generated garbage files, like the syscall headers on x86, for
example.

I have a fix below which seems to take care of it but what is the
general opinion: Do we want to be more robust against the environment we
find on a machine before building the kernel or let the user figure it
out himself that he should be using

GREP_OPTIONS=--color=auto

in the first place and it is his own moronic fault if he does 'always'?

Opinions, comments?

Thanks.

--
From: Borislav Petkov <[e-mail]>
Subject: [PATCH] Kbuild: Ignore GREP_OPTIONS env variable
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

When building the kernel in a shell which defines GREP_OPTIONS so that
grep behavior is modified, we can break the generation of the syscalls
table like so:

__SYSCALL_COMMON(0, sys_read, sys_read)
__SYSCALL_COMMON(1, sys_write, sys_write)
__SYSCALL_COMMON(10, sys_mprotect, sys_mprotect) ...

This is just the initial breakage, later we barf when generating
modules.

In this case, GREP_OPTIONS contains "--color=always" which adds the shell
colors markup and completely fudges the headers under ...generated/asm/.

Fix that by ignoring the GREP_OPTIONS variable for the whole kernel
build as we tend to use grep at a bunch of places.

Signed-off-by: Borislav Petkov <[e-mail]>
---
 Makefile | 3 +++
 1 file changed, 3 insertions(+)

Comments

Michal Marek Nov. 6, 2013, 9:42 p.m. UTC | #1

Dne 30.10.2013 17:06, Borislav Petkov napsal(a):
> So I had defined GREP_OPTIONS=--color=always on one of my boxes and had
> forgotten about it and the kernel build started failing because we use
> grep quite a while in the tree and it started issuing shell color markup
> which generated garbage files, like the syscall headers on x86, for
> example.
> 
> I have a fix below which seems to take care of it but what is the
> general opinion: Do we want to be more robust against the environment we
> find on a machine before building the kernel or let the user figure it
> out himself that he should be using
> 
> GREP_OPTIONS=--color=auto
> 
> in the first place and it is his own moronic fault if he does 'always'?

I think that on a scale from one to ten, building with
GREP_OPTIONS=--color=always and building with LC_COLLATE=<language with
non-ASCII sorting rules> score about the same. And are already
sanitizing the LC_* variables.


> +GREP_OPTIONS=
> +export GREP_OPTIONS

Just remove it from the enviromnet completely, like the next statement does:

> +
>  # Avoid funny character set dependencies
>  unexport LC_ALL
>  LC_COLLATE=C

Michal

--
To unsubscribe from this list: send the line "unsubscribe linux-kbuild" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch

diff --git a/Makefile b/Makefile
index 868c0eb67b08..7a2acd9a5d1c 100644
--- a/Makefile
+++ b/Makefile
@@ -16,6 +16,9 @@  NAME = One Giant Leap for Frogkind
 # o  print "Entering directory ...";
 MAKEFLAGS += -rR --no-print-directory
 
+GREP_OPTIONS=
+export GREP_OPTIONS
+
 # Avoid funny character set dependencies
 unexport LC_ALL
 LC_COLLATE=C