From: Suzuki K Poulose <[email protected]>
The pid of the task could be traced as VMID when the kernel is
running at EL2. Teach the decoder to look for vmid when the
context_id is invalid but we have a valid VMID.
Cc: Mike Leach <[email protected]>
Cc: Mathieu Poirier <[email protected]>
Cc: Al Grant <[email protected]>
Co-developed-by: Leo Yan <[email protected]>
Signed-off-by: Suzuki K Poulose <[email protected]>
Signed-off-by: Leo Yan <[email protected]>
---
.../perf/util/cs-etm-decoder/cs-etm-decoder.c | 32 ++++++++++++++++---
1 file changed, 28 insertions(+), 4 deletions(-)
diff --git a/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c b/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c
index cd007cc9c283..9e81169dfa76 100644
--- a/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c
+++ b/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c
@@ -6,6 +6,7 @@
* Author: Mathieu Poirier <[email protected]>
*/
+#include <linux/coresight-pmu.h>
#include <linux/err.h>
#include <linux/list.h>
#include <linux/zalloc.h>
@@ -500,13 +501,36 @@ cs_etm_decoder__set_tid(struct cs_etm_queue *etmq,
const ocsd_generic_trace_elem *elem,
const uint8_t trace_chan_id)
{
- pid_t tid;
+ pid_t tid = -1;
+ u64 pid_fmt;
+ int ret;
- /* Ignore PE_CONTEXT packets that don't have a valid contextID */
- if (!elem->context.ctxt_id_valid)
+ ret = cs_etm__get_pid_fmt(trace_chan_id, &pid_fmt);
+ if (ret)
+ return OCSD_RESP_FATAL_SYS_ERR;
+
+ /*
+ * Process the PE_CONTEXT packets if we have a valid
+ * contextID or VMID.
+ * If the kernel is running at EL2, the PID is traced
+ * in contextidr_el2 as VMID.
+ */
+ switch (pid_fmt) {
+ case BIT(ETM_OPT_CTXTID):
+ if (elem->context.ctxt_id_valid)
+ tid = elem->context.context_id;
+ break;
+ case BIT(ETM_OPT_CTXTID_IN_VMID):
+ if (elem->context.vmid_valid)
+ tid = elem->context.vmid;
+ break;
+ default:
+ break;
+ }
+
+ if (tid == -1)
return OCSD_RESP_CONT;
- tid = elem->context.context_id;
if (cs_etm__etmq_set_tid(etmq, tid, trace_chan_id))
return OCSD_RESP_FATAL_SYS_ERR;
--
2.25.1
Hi Leo
On 1/9/21 7:44 AM, Leo Yan wrote:
> From: Suzuki K Poulose <[email protected]>
>
> The pid of the task could be traced as VMID when the kernel is
> running at EL2. Teach the decoder to look for vmid when the
> context_id is invalid but we have a valid VMID.
Thank you again for cleaning up this ! Please see one comment
below.
>
> Cc: Mike Leach <[email protected]>
> Cc: Mathieu Poirier <[email protected]>
> Cc: Al Grant <[email protected]>
> Co-developed-by: Leo Yan <[email protected]>
> Signed-off-by: Suzuki K Poulose <[email protected]>
> Signed-off-by: Leo Yan <[email protected]>
> ---
> .../perf/util/cs-etm-decoder/cs-etm-decoder.c | 32 ++++++++++++++++---
> 1 file changed, 28 insertions(+), 4 deletions(-)
>
> diff --git a/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c b/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c
> index cd007cc9c283..9e81169dfa76 100644
> --- a/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c
> +++ b/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c
> @@ -6,6 +6,7 @@
> * Author: Mathieu Poirier <[email protected]>
> */
>
> +#include <linux/coresight-pmu.h>
> #include <linux/err.h>
> #include <linux/list.h>
> #include <linux/zalloc.h>
> @@ -500,13 +501,36 @@ cs_etm_decoder__set_tid(struct cs_etm_queue *etmq,
> const ocsd_generic_trace_elem *elem,
> const uint8_t trace_chan_id)
> {
> - pid_t tid;
> + pid_t tid = -1;
> + u64 pid_fmt;
> + int ret;
>
> - /* Ignore PE_CONTEXT packets that don't have a valid contextID */
> - if (!elem->context.ctxt_id_valid)
> + ret = cs_etm__get_pid_fmt(trace_chan_id, &pid_fmt);
> + if (ret)
> + return OCSD_RESP_FATAL_SYS_ERR;
The patch looks fine to me. I am wondering if this can be cached somewhere in
the etmq to avoid doing the search everytime we hit a CID ? Surely for a session,
this woudn't change and thus for the decoder life time.
Cheers
Suzuki
On Mon, Jan 11, 2021 at 10:07:03AM +0000, Suzuki Kuruppassery Poulose wrote:
> Hi Leo
>
> On 1/9/21 7:44 AM, Leo Yan wrote:
> > From: Suzuki K Poulose <[email protected]>
> >
> > The pid of the task could be traced as VMID when the kernel is
> > running at EL2. Teach the decoder to look for vmid when the
> > context_id is invalid but we have a valid VMID.
>
> Thank you again for cleaning up this ! Please see one comment
> below.
Welcome!
> > Cc: Mike Leach <[email protected]>
> > Cc: Mathieu Poirier <[email protected]>
> > Cc: Al Grant <[email protected]>
> > Co-developed-by: Leo Yan <[email protected]>
> > Signed-off-by: Suzuki K Poulose <[email protected]>
> > Signed-off-by: Leo Yan <[email protected]>
> > ---
> > .../perf/util/cs-etm-decoder/cs-etm-decoder.c | 32 ++++++++++++++++---
> > 1 file changed, 28 insertions(+), 4 deletions(-)
> >
> > diff --git a/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c b/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c
> > index cd007cc9c283..9e81169dfa76 100644
> > --- a/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c
> > +++ b/tools/perf/util/cs-etm-decoder/cs-etm-decoder.c
> > @@ -6,6 +6,7 @@
> > * Author: Mathieu Poirier <[email protected]>
> > */
> > +#include <linux/coresight-pmu.h>
> > #include <linux/err.h>
> > #include <linux/list.h>
> > #include <linux/zalloc.h>
> > @@ -500,13 +501,36 @@ cs_etm_decoder__set_tid(struct cs_etm_queue *etmq,
> > const ocsd_generic_trace_elem *elem,
> > const uint8_t trace_chan_id)
> > {
> > - pid_t tid;
> > + pid_t tid = -1;
> > + u64 pid_fmt;
> > + int ret;
> > - /* Ignore PE_CONTEXT packets that don't have a valid contextID */
> > - if (!elem->context.ctxt_id_valid)
> > + ret = cs_etm__get_pid_fmt(trace_chan_id, &pid_fmt);
> > + if (ret)
> > + return OCSD_RESP_FATAL_SYS_ERR;
>
> The patch looks fine to me. I am wondering if this can be cached somewhere in
> the etmq to avoid doing the search everytime we hit a CID ? Surely for a session,
> this woudn't change and thus for the decoder life time.
Good suggestion, will refine for this in next version.
Thanks,
Leo