This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] PR middle-end/67220: GCC fails to properly handle libcall symbol visibility of built functions


On Wed, Oct 14, 2015 at 5:51 PM, H.J. Lu <hjl.tools@gmail.com> wrote:
> On Wed, Oct 14, 2015 at 9:46 AM, Ramana Radhakrishnan
> <ramana.gcc@googlemail.com> wrote:
>> On Wed, Oct 14, 2015 at 5:21 PM, H.J. Lu <hongjiu.lu@intel.com> wrote:
>>
>>> ---
>>> gcc/c/
>>>
>>>         PR middle-end/67220
>>>         * c-decl.c (diagnose_mismatched_decls): Copy explicit visibility
>>>         to builtin function.
>>>
>>> gcc/
>>>
>>>         PR middle-end/67220
>>>         * expr.c (init_block_move_fn): Copy visibility from the builtin
>>>         memcpy.
>>>         (init_block_clear_fn): Copy visibility from the builtin memset.
>>>
>>> gcc/testsuite/
>>>
>>>         PR middle-end/67220
>>>         * gcc.target/i386/pr67220-1.c: New test.
>>>         * gcc.target/i386/pr67220-2.c: Likewise.
>>>         * gcc.target/i386/pr67220-3.c: Likewise.
>>>         * gcc.target/i386/pr67220-4.c: Likewise.
>>>         * gcc.target/i386/pr67220-5.c: Likewise.
>>>         * gcc.target/i386/pr67220-6.c: Likewise.
>>
>> Why aren't these tests in gcc.dg ?  The problem affects all targets
>> not just x86.
>>
>
> If I move tests to gcc.dg, would you mind updating them to verify
> that they pass on arm?


It's not just a question of ARM. This affects all targets that support
symbol visibility in shared libraries ... please do the math as to how
many targets are affected. Now reading the test even more, it appears
that you also need a dg-require-visibility so that this test is run on
all targets that support symbol visibility.

The test as written uses target_fpic - target-supports.exp does not
have anything ARM specific about handling this - there is some m68k ,
so it should just work on any target that supports symbol visibility.
If it doesn't, that target has an issue and then folks interested in
that target will do something about it. Even if you don't fix the
issue on every target, please have the courtesy of letting them find
it in some sort of automatic manner rather than auditing every single
commit in every gcc.target directory.


regards
Ramana


>
>
> --
> H.J.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]