4.4-stable review patch. If anyone has any objections, please let me know. ------------------ From: Adrian Hunter <adrian.hunter@xxxxxxxxx> commit 63d8e38f6ae6c36dd5b5ba0e8c112e8861532ea2 upstream. sync_switch is a facility to synchronize decoding more closely with the point in the kernel when the context actually switched. The flag when sync_switch is enabled was global to the decoding, whereas it is really specific to the CPU. The trace data for different CPUs is put on different queues, so add sync_switch to the intel_pt_queue structure and use that in preference to the global setting in the intel_pt structure. That fixes problems decoding one CPU's trace because sync_switch was disabled on a different CPU's queue. Signed-off-by: Adrian Hunter <adrian.hunter@xxxxxxxxx> Cc: Jiri Olsa <jolsa@xxxxxxxxxx> Cc: stable@xxxxxxxxxxxxxxx Link: http://lkml.kernel.org/r/1520431349-30689-3-git-send-email-adrian.hunter@xxxxxxxxx Signed-off-by: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> --- tools/perf/util/intel-pt.c | 32 +++++++++++++++++++++++++------- 1 file changed, 25 insertions(+), 7 deletions(-) --- a/tools/perf/util/intel-pt.c +++ b/tools/perf/util/intel-pt.c @@ -125,6 +125,7 @@ struct intel_pt_queue { bool stop; bool step_through_buffers; bool use_buffer_pid_tid; + bool sync_switch; pid_t pid, tid; int cpu; int switch_state; @@ -852,10 +853,12 @@ static int intel_pt_setup_queue(struct i if (pt->timeless_decoding || !pt->have_sched_switch) ptq->use_buffer_pid_tid = true; } + + ptq->sync_switch = pt->sync_switch; } if (!ptq->on_heap && - (!pt->sync_switch || + (!ptq->sync_switch || ptq->switch_state != INTEL_PT_SS_EXPECTING_SWITCH_EVENT)) { const struct intel_pt_state *state; int ret; @@ -1238,7 +1241,7 @@ static int intel_pt_sample(struct intel_ if (pt->synth_opts.last_branch) intel_pt_update_last_branch_rb(ptq); - if (!pt->sync_switch) + if (!ptq->sync_switch) return 0; if (intel_pt_is_switch_ip(ptq, state->to_ip)) { @@ -1319,6 +1322,21 @@ static u64 intel_pt_switch_ip(struct int return switch_ip; } +static void intel_pt_enable_sync_switch(struct intel_pt *pt) +{ + unsigned int i; + + pt->sync_switch = true; + + for (i = 0; i < pt->queues.nr_queues; i++) { + struct auxtrace_queue *queue = &pt->queues.queue_array[i]; + struct intel_pt_queue *ptq = queue->priv; + + if (ptq) + ptq->sync_switch = true; + } +} + static int intel_pt_run_decoder(struct intel_pt_queue *ptq, u64 *timestamp) { const struct intel_pt_state *state = ptq->state; @@ -1335,7 +1353,7 @@ static int intel_pt_run_decoder(struct i if (pt->switch_ip) { intel_pt_log("switch_ip: %"PRIx64" ptss_ip: %"PRIx64"\n", pt->switch_ip, pt->ptss_ip); - pt->sync_switch = true; + intel_pt_enable_sync_switch(pt); } } } @@ -1351,9 +1369,9 @@ static int intel_pt_run_decoder(struct i if (state->err) { if (state->err == INTEL_PT_ERR_NODATA) return 1; - if (pt->sync_switch && + if (ptq->sync_switch && state->from_ip >= pt->kernel_start) { - pt->sync_switch = false; + ptq->sync_switch = false; intel_pt_next_tid(pt, ptq); } if (pt->synth_opts.errors) { @@ -1379,7 +1397,7 @@ static int intel_pt_run_decoder(struct i state->timestamp, state->est_timestamp); ptq->timestamp = state->est_timestamp; /* Use estimated TSC in unknown switch state */ - } else if (pt->sync_switch && + } else if (ptq->sync_switch && ptq->switch_state == INTEL_PT_SS_UNKNOWN && intel_pt_is_switch_ip(ptq, state->to_ip) && ptq->next_tid == -1) { @@ -1526,7 +1544,7 @@ static int intel_pt_sync_switch(struct i return 1; ptq = intel_pt_cpu_to_ptq(pt, cpu); - if (!ptq) + if (!ptq || !ptq->sync_switch) return 1; switch (ptq->switch_state) {