This fixes a regression from 450c358881dbc6dcfb2692b8f4debb7db8fdc54b; after testing for -Wl,-z,defs, CFLAGS was restored from an incorrect variable, leaving CFLAGS essentially empty.
This would break builds for e.g. arm with Clang, where -mthumb would be essential for being able to build some inline assembly snippets correctly.
This also had the effect of dropping the default "-g -O2" arguments from CFLAGS.
Signed-off-by: Martin Storsjö martin@martin.st
From: Martin Storsjö martin@martin.st
This fixes a regression from 450c358881dbc6dcfb2692b8f4debb7db8fdc54b; after testing for -Wl,-z,defs, CFLAGS was restored from an incorrect variable, leaving CFLAGS essentially empty.
This would break builds for e.g. arm with Clang, where -mthumb would be essential for being able to build some inline assembly snippets correctly.
This also had the effect of dropping the default "-g -O2" arguments from CFLAGS.
Signed-off-by: Martin Storsjö martin@martin.st --- configure.ac | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/configure.ac b/configure.ac index d288c544dfd..4803b7abea4 100644 --- a/configure.ac +++ b/configure.ac @@ -771,7 +771,7 @@ case $host_os in dnl On FreeBSD, shared libraries using environ fail to link with -Wl,-z,defs AC_LINK_IFELSE([AC_LANG_SOURCE([[extern char **environ; char **envp; void myfunc(void) { envp = environ; }]])], [ac_cv_wl_z_defs=yes],[ac_cv_wl_z_defs=no]) - CFLAGS=$ac_save_cflags]) + CFLAGS=$ac_save_CFLAGS]) test $ac_cv_wl_z_defs != yes || AS_VAR_APPEND([UNIXLDFLAGS],[" -Wl,-z,defs"])
WINE_TRY_CFLAGS([-Wl,--export-dynamic],[WINELOADER_LDFLAGS="-Wl,--export-dynamic"])
also for configure file ( line 10125 ) https://gitlab.winehq.org/wine/wine/-/commit/450c358881dbc6dcfb2692b8f4debb7...
On Wed Oct 4 12:51:40 2023 +0000, aur-r wrote:
also for configure file ( line 10125 ) https://gitlab.winehq.org/wine/wine/-/commit/450c358881dbc6dcfb2692b8f4debb7...
I leave regenerating the configure script up to @julliard - I believe that's the usual procedure for such things.
On Wed Oct 4 12:53:05 2023 +0000, Martin Storsjö wrote:
I leave regenerating the configure script up to @julliard - I believe that's the usual procedure for such things.
Thanks, didn't know.