On Mon, Dec 19, 2022 at 12:21:47PM +0000, Christophe Leroy wrote: > > > Le 19/12/2022 à 13:18, Greg KH a écrit : > > On Fri, Dec 16, 2022 at 12:38:12PM +0100, Christophe Leroy wrote: > >> From: Tony Jones <tonyj@xxxxxxx> > >> > >> [Upstream commit d72eadbc1d2866fc047edd4535ffb0298fe240be] > >> > >> tests/attr.c invokes attr.py via an explicit invocation of Python > >> ($PYTHON) so there is therefore no need for an explicit shebang. > >> > >> Also most distros follow pep-0394 which recommends that /usr/bin/python > >> refer only to v2 and so may not exist on the system (if PYTHON=python3). > >> > >> Signed-off-by: Tony Jones <tonyj@xxxxxxx> > >> Acked-by: Jiri Olsa <jolsa@xxxxxxxxxx> > >> Cc: Jonathan Corbet <corbet@xxxxxxx> > >> Cc: Ravi Bangoria <ravi.bangoria@xxxxxxxxxxxxx> > >> Cc: Seeteena Thoufeek <s1seetee@xxxxxxxxxxxxxxxxxx> > >> Link: http://lkml.kernel.org/r/20190124005229.16146-5-tonyj@xxxxxxx > >> Signed-off-by: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx> > >> Signed-off-by: Christophe Leroy <christophe.leroy@xxxxxxxxxx> > >> --- > >> tools/perf/tests/attr.py | 1 - > >> 1 file changed, 1 deletion(-) > > > > Why only 4.14? What about 4.19? > > > > > I submitted that backport because I encountered the problem while > building perf for 4.14. > > I didn't look at other kernel versions. In the future, remember that we can not add fixes to an older version and not a newer one, otherwise you would have a regression moving to a newer kernel. thanks, greg k-h