On 12/05/2013 10:25 AM, Paul Eggert wrote:
> On 12/04/2013 11:51 AM, Paul Eggert wrote:
>> or we can change the test case to not tickle the bug.
>
> I did the latter, as follows. Maybe at some point this
> should be filed as a bug against glibc, since the bug occurs
> even with _FORTIFY_SOURCE de
On 12/05/2013 05:25 PM, Paul Eggert wrote:
> -static int
> +static int __always_inline
> test_open (int (*func) (char const *, int, ...), bool print)
This fails `make check` in coreutils...
$ make V=1 TESTS= check
...
test-open.h:34:1: error: '__inline' is not at beginning of declaration
[-Werr
On 12/04/2013 11:51 AM, Paul Eggert wrote:
> or we can change the test case to not tickle the bug.
I did the latter, as follows. Maybe at some point this
should be filed as a bug against glibc, since the bug occurs
even with _FORTIFY_SOURCE defined to 1 and that isn't supposed
to break conforming
On 12/04/2013 09:45 AM, Eric Blake wrote:
> Has
> this issue been fixed in the meantime by Paul's improvements to inline
> handling?
No, it hasn't. It's a bug in glibc, clearly.
We can either change gnulib to fix the bug for gnulib-using programs,
or we can change the test case to not tickle the
On 06/20/2013 09:45 PM, Daiki Ueno wrote:
> I got the following error when compiling test-open.c with
> -D_FORTIFY_SOURCE=2.
>
> $ gcc -DHAVE_CONFIG_H -I. -DGNULIB_STRICT_CHECKING=1 -I. -I. -I.. -I./..
> -I../gllib -I./../gllib -Wall -O1 -D_FORTIFY_SOURCE=2 -g -MT test-open.o
> -MD -MP -MF
b01e858b395 Mon Sep 17 00:00:00 2001
From: Daiki Ueno
Date: Fri, 21 Jun 2013 12:37:20 +0900
Subject: [PATCH] test-open: avoid compilation error with -D_FORTIFY_SOURCE=2
* tests/test-open.c (do_open): New wrapper around the system open.
---
ChangeLog | 5 +
tests/test-open.c | 24 +
