[PATCH v3] elf: Test dlopen (NULL, RTLD_LAZY) from an ELF constructor

[Florian Weimer]

* Sam James:

>> diff --git a/elf/tst-dlopen-constructor-null.c b/elf/tst-dlopen-constructor-null.c
>> new file mode 100644
>> index 0000000000..58c4660c7d
>> --- /dev/null
>> +++ b/elf/tst-dlopen-constructor-null.c
>> @@ -0,0 +1,38 @@
>> +/* Verify that dlopen (NULL, RTLD_LAZY) does not complete initialization.
>> +   Copyright (C) 2024 Free Software Foundation, Inc.
>> +   This file is part of the GNU C Library.
>> +
>> +   The GNU C Library is free software; you can redistribute it and/or
>> +   modify it under the terms of the GNU Lesser General Public
>> +   License as published by the Free Software Foundation; either
>> +   version 2.1 of the License, or (at your option) any later version.
>> +
>> +   The GNU C Library is distributed in the hope that it will be useful,
>> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
>> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
>> +   Lesser General Public License for more details.
>> +
>> +   You should have received a copy of the GNU Lesser General Public
>> +   License along with the GNU C Library; if not, see
>> +   <https://www.gnu.org/licenses/>.  */
>> +
>> +/* This test mimics what the glvndSetupPthreads function in libglvnd
>> +   does. */
>
> Have you confirmed that reverting the fixes breaks the test?

If I cherry-pick the changes from commit 9897ced8e78db5d813166a7ccccfd
(minus the elf/Makefile update for the tests, due to conflicts), I get
this:

FAIL: elf/tst-dlopen-constructor-null
original exit status 1
info: tst-dlopen-constructor-null-mod1.so constructor
info: tst-dlopen-constructor-null-mod2.so constructor (mod1_status=0)error: mod1_status == 1 && mod2_status == 0 expected

So I assume the test does exercise the bug.

>> +#include <stdlib.h>
>> +#include <support/check.h>
>> +
>> +/* Defined an initialized in the shared objects.  */
>
> "an initialized"?

“and initialized”.  Fixed locally, thanks.

Florian