[FFmpeg-devel] [PATCH] configure: link to libatomic when it's present

[Brad Smith]

On 1/22/2022 4:00 AM, Hendrik Leppkes wrote:

> On Sat, Jan 22, 2022 at 7:43 AM Brad Smith
> <brad-at-comstyle.com at ffmpeg.org>  wrote:
>> On 1/19/2022 10:23 AM, James Almer wrote:
>>
>>> On 1/19/2022 10:48 AM, Anton Khirnov wrote:
>>>> C11 atomics in some configurations (e.g. 64bit operations on ppc64 with
>>>> GCC) require linking to libatomic.
>>>>
>>>> Fixes #9275
>>>> ---
>>>>    configure | 25 ++++++++++++++++---------
>>>>    1 file changed, 16 insertions(+), 9 deletions(-)
>>>>
>>>> diff --git a/configure b/configure
>>>> index 1413122d87..3059c154df 100755
>>>> --- a/configure
>>>> +++ b/configure
>>>> @@ -3794,20 +3794,20 @@ cws2fws_extralibs="zlib_extralibs"
>>>>      # libraries, in any order
>>>>    avcodec_deps="avutil"
>>>> -avcodec_suggest="libm"
>>>> +avcodec_suggest="libm stdatomic"
>>>>    avdevice_deps="avformat avcodec avutil"
>>>> -avdevice_suggest="libm"
>>>> +avdevice_suggest="libm stdatomic"
>>>>    avfilter_deps="avutil"
>>>> -avfilter_suggest="libm"
>>>> +avfilter_suggest="libm stdatomic"
>>>>    avformat_deps="avcodec avutil"
>>>> -avformat_suggest="libm network zlib"
>>>> -avutil_suggest="clock_gettime ffnvcodec libm libdrm libmfx opencl
>>>> user32 vaapi vulkan videotoolbox corefoundation corevideo coremedia
>>>> bcrypt"
>>>> +avformat_suggest="libm network zlib stdatomic"
>>>> +avutil_suggest="clock_gettime ffnvcodec libm libdrm libmfx opencl
>>>> user32 vaapi vulkan videotoolbox corefoundation corevideo coremedia
>>>> bcrypt stdatomic"
>>>>    postproc_deps="avutil gpl"
>>>> -postproc_suggest="libm"
>>>> +postproc_suggest="libm stdatomic"
>>>>    swresample_deps="avutil"
>>>> -swresample_suggest="libm libsoxr"
>>>> +swresample_suggest="libm libsoxr stdatomic"
>>>>    swscale_deps="avutil"
>>>> -swscale_suggest="libm"
>>>> +swscale_suggest="libm stdatomic"
>>>>      avcodec_extralibs="pthreads_extralibs iconv_extralibs
>>>> dxva2_extralibs"
>>>>    avfilter_extralibs="pthreads_extralibs"
>>>> @@ -6324,7 +6324,14 @@ check_headers asm/types.h
>>>>    # it seems there are versions of clang in some distros that try to
>>>> use the
>>>>    # gcc headers, which explodes for stdatomic
>>>>    # so we also check that atomics actually work here
>>>> -check_builtin stdatomic stdatomic.h "atomic_int foo, bar =
>>>> ATOMIC_VAR_INIT(-1); atomic_store(&foo, 0); foo += bar"
>>>> +#
>>>> +# some configurations also require linking to libatomic, so try
>>>> +# both with -latomic and without
>>>> +for LATOMIC in "-latomic" ""; do
>>>> +    check_builtin stdatomic
>>>> stdatomic.h                                                 \
>>>> +        "atomic_int foo, bar = ATOMIC_VAR_INIT(-1);
>>>> atomic_store(&foo, 0); foo += bar"  \
>>>> +        $LATOMIC && eval stdatomic_extralibs="\$LATOMIC" && break
>>>> +done
>>> LGTM now, thanks.
>>>
>>>>      check_lib advapi32 "windows.h"            RegCloseKey
>>>> -ladvapi32
>>>>    check_lib bcrypt   "windows.h bcrypt.h"   BCryptGenRandom -lbcrypt &&
>> Wait, this should be checking without first then with, if the first test
>> without fails.
>>
> This was covered earlier in the thread for the reason it is not - its
> deliberate, because exhaustive functionality checks would be very
> complicated.

Testing this commit out it does as I had suspected and even with --as-needed
causes a false positive on OpenBSD / FreeBSD.  Now erroneously tries to link
against libatomic and unlike the other project (haproxy) I ran across an 
overly
simplistic test (doesn't even involve linking, just checking compiler 
predefined
macros) I don't see any options to disable the broken test either.