On 02/07/2018 07:18 AM, Steven Rostedt wrote: > On Sun, 4 Feb 2018 11:20:13 +0800 > szts...@gmail.com wrote: > >> From: Zamir SUN <szts...@gmail.com> >> >> The current detection for swig will cause output to be >> /usr/bin/swig >> y >> So this will never be equal to y. With this patch, the swig path is >> removed from output, so the detection can work as expected. >> >> Fixes 3bf187a43b7e6302592552ecbc294e5820249687 > > Hi Zamir, > > Actually, a fix was already sent to me: > > http://lkml.kernel.org/r/20171220180748.17273-18-vladislav.valtc...@gmail.com > > I'm working on add that one. >
Thanks for the heads-up! I did not see that before. While just a note for this, in Fedora 27: $ if command -v swig; then echo 1; else echo 0; fi /usr/bin/swig 1 Actually this test has the same problem as of `which swig` - When swig exists, the var SWIG_DEFINED will not be 1 (as shown above). I see Vladislav is already on this thread. I hope this can be fixed before merging the patch set. > Thanks! > > -- Steve > > >> >> Signed-off-by: Zamir SUN (Red Hat) <szts...@gmail.com> >> --- >> Makefile | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/Makefile b/Makefile >> index a5d2c38..7c0d1a6 100644 >> --- a/Makefile >> +++ b/Makefile >> @@ -105,7 +105,7 @@ PYTHON_GUI := ctracecmd.so ctracecmdgui.so >> PYTHON_VERS ?= python >> >> # Can build python? >> -ifeq ($(shell sh -c "pkg-config --cflags $(PYTHON_VERS) > /dev/null 2>&1 && >> which swig && echo y"), y) >> +ifeq ($(shell sh -c "pkg-config --cflags $(PYTHON_VERS) > /dev/null 2>&1 && >> which swig > /dev/null && echo y"), y) >> PYTHON_PLUGINS := plugin_python.so >> BUILD_PYTHON := $(PYTHON) $(PYTHON_PLUGINS) >> PYTHON_SO_INSTALL := ctracecmd.install > -- Zamir SUN Fedora user GPG : 1D86 6D4A 49CE 4BBD 72CF FCF5 D856 6E11 F2A0 525E