Markus Armbruster <arm...@redhat.com> writes: > Stefan Hajnoczi <stefa...@gmail.com> writes: > >> On Sun, Aug 16, 2020 at 12:46:10PM -0500, dubo...@gmail.com wrote: >>> From: Josh DuBois <j...@joshdubois.com> >>> >>> Tracing can be enabled at the command line or via the >>> monitor. Command-line trace options are recorded during >>> trace_opt_parse(), but tracing is not enabled until the various >>> front-ends later call trace_init_file(). If the user passes a trace >>> option on the command-line, remember that and enable tracing during >>> trace_init_file(). Otherwise, trace_init_file() should record the >>> trace file specified by the frontend and avoid enabling traces >>> until the user requests them via the monitor. >>> >>> This fixes 1b7157be3a8c4300fc8044d40f4b2e64a152a1b4 and also >>> db25d56c014aa1a96319c663e0a60346a223b31e, by allowing the user >>> to enable traces on the command line and also avoiding >>> unwanted trace-<pid> files when the user has not asked for them. >>> >>> Fixes: 1b7157be3a8c4300fc8044d40f4b2e64a152a1b4 >>> Signed-off-by: Josh DuBois <j...@joshdubois.com> >>> --- >>> trace/control.c | 6 +++++- >>> 1 file changed, 5 insertions(+), 1 deletion(-) >> >> Thanks, applied to my tracing-next tree: >> https://github.com/stefanha/qemu/commits/tracing-next >> >> Stefan > > Pull request?
Pretty-please?
