Skip to contentSkip to conversation

Fixes needed befor KS 1.0

PATCH In review Yordan Karadzhov 2 participants 5 messages Jun 14, 2019 – Jun 15, 2019

lore
Share & tools
lore
#0 lore raw

v2 cover Fixes needed befor KS 1.0

The important one is the one fixing the visualization model aborting on 32 bit system [PATCH 2].

Changes in v2:
- [PATCH 2] includes a fix for void ksmodel_jump_to()
- [PATCH 3] is new.

Yordan Karadzhov (3):
kernel-shark: Fix a bug when plotting the last trace record
kernel-shark: Always use 64 bit variables for timestamps.
kernel-shark: Fix all warnings when building on 32 bit systems

 kernel-shark/examples/datahisto.c  |  4 ++--
 kernel-shark/src/KsTraceGraph.cpp  | 10 +++++-----
 kernel-shark/src/libkshark-model.c | 27 +++++++++++++++++++--------
 kernel-shark/src/libkshark-model.h |  2 +-
 kernel-shark/src/libkshark.c       |  4 ++--
 5 files changed, 29 insertions(+), 18 deletions(-)
Signature · 1 line
2.20.1
#1 lore raw

PATCH v2 1/3 kernel-shark: Fix a bug when plotting the last trace record

On a first glance this patch may looks like reverting commit 9336dd6bcd38 (kernel-shark: Fix a bug in ksmodel_set_next_bin_edge())

The point is that for the last bin we want to increment its upper edge used when checking if the bin is empty, but we do not want to touch the lower edge time used by kshark_find_entry_by_time().

Fixes 9336dd6bcd38 (kernel-shark: Fix a bug in ksmodel_set_next_bin_edge())
---
 kernel-shark/src/libkshark-model.c | 22 ++++++++++++++++------
 1 file changed, 16 insertions(+), 6 deletions(-)
kernel-shark/src/libkshark-model.c +16 -6
index 978cd70..0cac924 100644
@@ -260,20 +260,30 @@ static size_t ksmodel_set_upper_edge(struct kshark_trace_histo *histo)
 static void ksmodel_set_next_bin_edge(struct kshark_trace_histo *histo,
 				      size_t bin, size_t last_row)
 {
-	size_t time, next_bin = bin + 1;
+	size_t time_min, time_max, next_bin = bin + 1;
 	ssize_t row;
 
-	/* Calculate the beginning of the next bin. */
-	time = histo->min + next_bin * histo->bin_size;
+	/* Calculate the beginning and the end of the next bin. */
+	time_min = histo->min + next_bin * histo->bin_size;
+	time_max = time_min + histo->bin_size;
+	/*
+	 * The timestamp of the very last entry of the dataset can be exactly
+	 * equal to the value of the upper edge of the range. This is very
+	 * likely to happen when we use ksmodel_set_in_range_bining(). In this
+	 * case we have to increase the size of the very last bin in order to
+	 * make sure that the last entry of the dataset will fall into it.
+	 */
+	if (next_bin == histo->n_bins - 1)
+		++time_max;
 
 	/*
 	 * Find the index of the first entry inside
-	 * the next bin (timestamp > time).
+	 * the next bin (timestamp > time_min).
 	 */
-	row = kshark_find_entry_by_time(time, histo->data, last_row,
+	row = kshark_find_entry_by_time(time_min, histo->data, last_row,
 					histo->data_size - 1);
 
-	if (row < 0 || histo->data[row]->ts >= time + histo->bin_size) {
+	if (row < 0 || histo->data[row]->ts >= time_max) {
 		/* The bin is empty. */
 		histo->map[next_bin] = KS_EMPTY_BIN;
 		return;
Signature · 1 line
2.20.1
#2 lore raw

PATCH v2 2/3 kernel-shark: Always use 64 bit variables for timestamps.

In reply to Yordan Karadzhov

All time stamps of the trace records are coded with 64 bits, however on some systems the size_t type can be 32 bits.

---
 kernel-shark/src/libkshark-model.c | 7 ++++---
 kernel-shark/src/libkshark-model.h | 2 +-
 2 files changed, 5 insertions(+), 4 deletions(-)
kernel-shark/src/libkshark-model.c +4 -3
index 0cac924..18f9c69 100644
@@ -260,7 +260,8 @@ static size_t ksmodel_set_upper_edge(struct kshark_trace_histo *histo)
 static void ksmodel_set_next_bin_edge(struct kshark_trace_histo *histo,
 				      size_t bin, size_t last_row)
 {
-	size_t time_min, time_max, next_bin = bin + 1;
+	uint64_t time_min, time_max;
+	size_t next_bin = bin + 1;
 	ssize_t row;
 
 	/* Calculate the beginning and the end of the next bin. */
@@ -601,9 +602,9 @@ void ksmodel_shift_backward(struct kshark_trace_histo *histo, size_t n)
  * @param histo: Input location for the model descriptor.
  * @param ts: position in time to be visualized.
  */
-void ksmodel_jump_to(struct kshark_trace_histo *histo, size_t ts)
+void ksmodel_jump_to(struct kshark_trace_histo *histo, uint64_t ts)
 {
-	size_t min, max, range_min;
+	uint64_t min, max, range_min;
 
 	if (ts > histo->min && ts < histo->max) {
 		/*
kernel-shark/src/libkshark-model.h +1 -1
index 95c30b6..47793b1 100644
@@ -89,7 +89,7 @@ void ksmodel_shift_forward(struct kshark_trace_histo *histo, size_t n);
 
 void ksmodel_shift_backward(struct kshark_trace_histo *histo, size_t n);
 
-void ksmodel_jump_to(struct kshark_trace_histo *histo, size_t ts);
+void ksmodel_jump_to(struct kshark_trace_histo *histo, uint64_t ts);
 
 void ksmodel_zoom_out(struct kshark_trace_histo *histo,
 		      double r, int mark);
Signature · 1 line
2.20.1
#3 lore raw

PATCH v2 3/3 kernel-shark: Fix all warnings when building on 32 bit systems

In reply to Yordan Karadzhov

Fixing warnings coming from:
../examples/datahisto.c
../src/KsTraceGraph.cpp
../src/libkshark.c

---
 kernel-shark/examples/datahisto.c |  4 ++--
 kernel-shark/src/KsTraceGraph.cpp | 10 +++++-----
 kernel-shark/src/libkshark.c      |  4 ++--
 3 files changed, 9 insertions(+), 9 deletions(-)
kernel-shark/examples/datahisto.c +2 -2
index 02c6285..b177b08 100644
@@ -56,11 +56,11 @@ void dump_bin(struct kshark_trace_histo *histo, int bin,
 		puts ("EMPTY BIN");
 	} else {
 		entry_str = kshark_dump_entry(e_front);
-		printf("%li -> %s\n", i_front, entry_str);
+		printf("%zd -> %s\n", i_front, entry_str);
 		free(entry_str);
 
 		entry_str = kshark_dump_entry(e_back);
-		printf("%li -> %s\n", i_back, entry_str);
+		printf("%zd -> %s\n", i_back, entry_str);
 		free(entry_str);
 	}
 
kernel-shark/src/KsTraceGraph.cpp +5 -5
index da2c6aa..324f36e 100644
@@ -277,7 +277,7 @@ void KsTraceGraph::_resetPointer(uint64_t ts, int cpu, int pid)
 	QString pointer;
 
 	kshark_convert_nano(ts, &sec, &usec);
-	pointer.sprintf("%lu.%06lu", sec, usec);
+	pointer.sprintf("%" PRIu64 ".%06" PRIu64 "", sec, usec);
 	_labelP2.setText(pointer);
 
 	if (pid > 0 && cpu >= 0) {
@@ -313,7 +313,7 @@ void KsTraceGraph::_setPointerInfo(size_t i)
 	uint64_t sec, usec;
 
 	kshark_convert_nano(e->ts, &sec, &usec);
-	pointer.sprintf("%lu.%06lu", sec, usec);
+	pointer.sprintf("%" PRIu64 ".%06" PRIu64 "", sec, usec);
 	_labelP2.setText(pointer);
 
 	comm.append("-");
@@ -601,17 +601,17 @@ void KsTraceGraph::_updateTimeLegends()
 	QString tMin, tMid, tMax;
 
 	kshark_convert_nano(_glWindow.model()->histo()->min, &sec, &usec);
-	tMin.sprintf("%lu.%06lu", sec, usec);
+	tMin.sprintf("%" PRIu64 ".%06" PRIu64 "", sec, usec);
 	_labelXMin.setText(tMin);
 
 	tsMid = (_glWindow.model()->histo()->min +
 		 _glWindow.model()->histo()->max) / 2;
 	kshark_convert_nano(tsMid, &sec, &usec);
-	tMid.sprintf("%lu.%06lu", sec, usec);
+	tMid.sprintf("%" PRIu64 ".%06" PRIu64 "", sec, usec);
 	_labelXMid.setText(tMid);
 
 	kshark_convert_nano(_glWindow.model()->histo()->max, &sec, &usec);
-	tMax.sprintf("%lu.%06lu", sec, usec);
+	tMax.sprintf("%" PRIu64 ".%06" PRIu64 "", sec, usec);
 	_labelXMax.setText(tMax);
 }
 
kernel-shark/src/libkshark.c +2 -2
index 0f0a1ba..9aaf1b7 100644
@@ -1423,7 +1423,7 @@ char* kshark_dump_custom_entry(struct kshark_context *kshark_ctx,
 	event_name = info_func(kshark_ctx, entry, false);
 	info = info_func(kshark_ctx, entry, true);
 
-	size = asprintf(&entry_str, "%li; %s-%i; CPU %i; ; %s; %s",
+	size = asprintf(&entry_str, "%" PRIu64 "; %s-%i; CPU %i; ; %s; %s",
 			entry->ts,
 			task,
 			entry->pid,
@@ -1472,7 +1472,7 @@ char* kshark_dump_entry(const struct kshark_entry *entry)
 		event_name = event? event->name : "[UNKNOWN EVENT]";
 		lat = kshark_get_latency(kshark_ctx->pevent, data);
 
-		size = asprintf(&temp_str, "%li; %s-%i; CPU %i; %s;",
+		size = asprintf(&temp_str, "%" PRIu64 "; %s-%i; CPU %i; %s;",
 				entry->ts,
 				task,
 				entry->pid,
Signature · 1 line
2.20.1
#4 lore raw

Re: [PATCH v2 3/3] kernel-shark: Fix all warnings when building on 32 bit systems

Quoted 98 lines · Yordan Karadzhov

Fixing warnings coming from:
../examples/datahisto.c
../src/KsTraceGraph.cpp
../src/libkshark.c

---
 kernel-shark/examples/datahisto.c |  4 ++--
 kernel-shark/src/KsTraceGraph.cpp | 10 +++++-----
 kernel-shark/src/libkshark.c      |  4 ++--
 3 files changed, 9 insertions(+), 9 deletions(-)
kernel-shark/examples/datahisto.c +2 -2
index 02c6285..b177b08 100644
@@ -56,11 +56,11 @@ void dump_bin(struct kshark_trace_histo *histo, int bin,
                puts ("EMPTY BIN");
        } else {
                entry_str = kshark_dump_entry(e_front);
-               printf("%li -> %s\n", i_front, entry_str);
+               printf("%zd -> %s\n", i_front, entry_str);
                free(entry_str);
 
                entry_str = kshark_dump_entry(e_back);
-               printf("%li -> %s\n", i_back, entry_str);
+               printf("%zd -> %s\n", i_back, entry_str);
                free(entry_str);
        }
 
kernel-shark/src/KsTraceGraph.cpp +5 -5
index da2c6aa..324f36e 100644
@@ -277,7 +277,7 @@ void KsTraceGraph::_resetPointer(uint64_t ts, int cpu, int pid)
        QString pointer;
 
        kshark_convert_nano(ts, &sec, &usec);
-       pointer.sprintf("%lu.%06lu", sec, usec);
+       pointer.sprintf("%" PRIu64 ".%06" PRIu64 "", sec, usec);
        _labelP2.setText(pointer);
 
        if (pid > 0 && cpu >= 0) {
@@ -313,7 +313,7 @@ void KsTraceGraph::_setPointerInfo(size_t i)
        uint64_t sec, usec;
 
        kshark_convert_nano(e->ts, &sec, &usec);
-       pointer.sprintf("%lu.%06lu", sec, usec);
+       pointer.sprintf("%" PRIu64 ".%06" PRIu64 "", sec, usec);
        _labelP2.setText(pointer);
 
        comm.append("-");
@@ -601,17 +601,17 @@ void KsTraceGraph::_updateTimeLegends()
        QString tMin, tMid, tMax;
 
        kshark_convert_nano(_glWindow.model()->histo()->min, &sec, &usec);
-       tMin.sprintf("%lu.%06lu", sec, usec);
+       tMin.sprintf("%" PRIu64 ".%06" PRIu64 "", sec, usec);
        _labelXMin.setText(tMin);
 
        tsMid = (_glWindow.model()->histo()->min +
                 _glWindow.model()->histo()->max) / 2;
        kshark_convert_nano(tsMid, &sec, &usec);
-       tMid.sprintf("%lu.%06lu", sec, usec);
+       tMid.sprintf("%" PRIu64 ".%06" PRIu64 "", sec, usec);
        _labelXMid.setText(tMid);
 
        kshark_convert_nano(_glWindow.model()->histo()->max, &sec, &usec);
-       tMax.sprintf("%lu.%06lu", sec, usec);
+       tMax.sprintf("%" PRIu64 ".%06" PRIu64 "", sec, usec);
        _labelXMax.setText(tMax);
 }
 
kernel-shark/src/libkshark.c +2 -3
index 0f0a1ba..9aaf1b7 100644
@@ -1423,7 +1423,7 @@ char* kshark_dump_custom_entry(struct kshark_context *kshark_ctx,
        event_name = info_func(kshark_ctx, entry, false);
        info = info_func(kshark_ctx, entry, true);
 
-       size = asprintf(&entry_str, "%li; %s-%i; CPU %i; ; %s; %s",
+       size = asprintf(&entry_str, "%" PRIu64 "; %s-%i; CPU %i; ; %s; %s",
                        entry->ts,
                        task,
                        entry->pid,
@@ -1472,7 +1472,7 @@ char* kshark_dump_entry(const struct kshark_entry *entry)
                event_name = event? event->name : "[UNKNOWN EVENT]";
                lat = kshark_get_latency(kshark_ctx->pevent, data);
 
-               size = asprintf(&temp_str, "%li; %s-%i; CPU %i; %s;",
+               size = asprintf(&temp_str, "%" PRIu64 "; %s-%i; CPU %i; %s;",
                                entry->ts,
                                task,
                                entry->pid,
--
Quoted 2 lines

2.20.1

Hi Yordan,

I had to manually apply your two patches from 2019-06-14 manually on top of your first patch from 2019-06-12. I was able to build KernelShark on my Raspberry Pi 3
model B+ and observe that your combined changes resolved the abort issue as well as compiler warnings. I also observed the same good results on my ODROID-XU3
from hardkernel.com which also runs a 32-bit armv7l kernel as well as my 96Boards
ROCK960 model C which runs a 64-bit aarch64 kernel.

Please see bugzilla for an attached patch file which shows the combined changes I
manually applied as your patches intended.

Regards,
Alan