[FFmpeg-devel] [PATCH] configure: Add workaround for qtkit detection using ICC, compiler.

[Timothy Gu]

On Sun, Apr 6, 2014 at 9:53 AM, Thilo Borgmann <thilo.borgmann at mail.de> wrote:
> Am 06.04.14 15:52, schrieb Michael Niedermayer:
>> On Sun, Apr 06, 2014 at 03:02:16PM +0200, Thilo Borgmann wrote:
>>> Like Reimar suggested on log, this works for me for proper detection of qtkit using:
>>>
>>> Mac OS X 10.9 with clang
>>> Linux with gcc
>>> Linux with icc
>>>
>>> -Thilo
>>
>>>  configure |    2 +-
>>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>> 96f5a6cf4915a1b8139050b48e08e1cc673871f4  0001-configure-Add-workaround-for-qtkit-detection-using-I.patch
>>> From d9a1e4177d2bc55410c747576f8bc5820d4938ab Mon Sep 17 00:00:00 2001
>>> From: Thilo Borgmann <thilo.borgmann at mail.de>
>>> Date: Sun, 6 Apr 2014 14:57:24 +0200
>>> Subject: [PATCH] configure: Add workaround for qtkit detection using ICC
>>>  compiler.
>>>
>>> ---
>>>  configure | 2 +-
>>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/configure b/configure
>>> index f238276..945c847 100755
>>> --- a/configure
>>> +++ b/configure
>>> @@ -4671,7 +4671,7 @@ enabled openssl           && { check_lib openssl/ssl.h SSL_library_init -lssl -l
>>>                                 check_lib openssl/ssl.h SSL_library_init -lssl32 -leay32 ||
>>>                                 check_lib openssl/ssl.h SSL_library_init -lssl -lcrypto -lws2_32 -lgdi32 ||
>>>                                 die "ERROR: openssl not found"; }
>>> -enabled qtkit_indev      && { check_header_oc QTKit/QTKit.h || disable qtkit_indev; }
>>> +enabled qtkit_indev      && { { check_header QTKit/QTKit.h && check_header_oc QTKit/QTKit.h; } || disable qtkit_indev; }
>>
>> applied
>>
>> though if someone can fix check_header_oc that would probably be
>> better
>>
>> maybe by checking if a output file has been created or something like
>> that (dont forget to make sure the output file doesnt exist before
>> the test)
>>
>> but for now this patch is better than nothing, build should not be
>> broken while the best solution is discussed
>
> Unfortunately, testing "stat $TMPO" is not sufficient, since this ICC call even
> generates a file...... not of size 0....
>
> part of config.log after adding applying the attached patch:
>
> check_header_oc QTKit/QTKit.h
> check_oc
> BEGIN /tmp/ffconf.DCnCE0fZ.m
>     1   #include <QTKit/QTKit.h>
>     2   int main(void) { return 0; }
> END /tmp/ffconf.DCnCE0fZ.m
> /opt/intel/composerxe-2011.4.191/bin/intel64/icc -D_ISOC99_SOURCE
> -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -D_POSIX_C_SOURCE=200112
> -D_XOPEN_SOURCE=600 -std=c99 -fomit-frame-pointer -pthread -c -o
> /tmp/ffconf.XKczCL4F.o /tmp/ffconf.DCnCE0fZ.m
> icc: warning #10147: no action performed for specified file(s)
> check_stat /tmp/ffconf.XKczCL4F.o
>   File: `/tmp/ffconf.XKczCL4F.o'
>   Size: 1464            Blocks: 8          IO Block: 4096   regular file
> Device: 801h/2049d      Inode: 16002333    Links: 1
> Access: (0644/-rw-r--r--)  Uid: ( 1000/    fate)   Gid: ( 1000/    fate)
> Access: 2014-04-06 18:37:33.036520001 +0200
> Modify: 2014-04-06 18:37:32.956520001 +0200
> Change: 2014-04-06 18:37:32.956520001 +0200

Damn icc. Why don't we just probe for "warning #10147" assuming that
the number don't change with updates.