Add show_from profile filter. (google#372)

doc/README.md

@@ -94,6 +94,8 @@ Some common pprof options are:
   *regex*.
 * **-ignore= _regex_:** Do not include samples that include a report entry matching
   *regex*.
+* **-show\_from= _regex_:** Do not show entries above the first one that 
+  matches *regex*.
 * **-show= _regex_:** Only show entries that match *regex*.
 * **-hide= _regex_:** Do not show entries that match *regex*.
 

internal/driver/commands.go

@@ -186,6 +186,10 @@ var pprofVariables = variables{
 		"Only show nodes matching regexp",
 		"If set, only show nodes that match this location.",
 		"Matching includes the function name, filename or object name.")},
+	"show_from": &variable{stringKind, "", "", helpText(
+		"Drops functions above the highest matched frame.",
+		"If set, all frames above the highest match are dropped from every sample.",
+		"Matching includes the function name, filename or object name.")},
 	"tagfocus": &variable{stringKind, "", "", helpText(
 		"Restricts to samples with tags in range or matched by regexp",
 		"Use name=value syntax to limit the matching to a specific tag.",

internal/driver/driver.go

@@ -150,11 +150,11 @@ func generateReport(p *profile.Profile, cmd []string, vars variables, o *plugin.
 }
 
 func applyCommandOverrides(cmd []string, v variables) variables {
-	trim, focus, tagfocus, hide := v["trim"].boolValue(), true, true, true
+	trim, tagfilter, filter := v["trim"].boolValue(), true, true
 
 	switch cmd[0] {
 	case "proto", "raw":
-		trim, focus, tagfocus, hide = false, false, false, false
+		trim, tagfilter, filter = false, false, false
 		v.set("addresses", "t")
 	case "callgrind", "kcachegrind":
 		trim = false
@@ -163,7 +163,7 @@ func applyCommandOverrides(cmd []string, v variables) variables {
 		trim = false
 		v.set("addressnoinlines", "t")
 	case "peek":
-		trim, focus, hide = false, false, false
+		trim, filter = false, false
 	case "list":
 		v.set("nodecount", "0")
 		v.set("lines", "t")
@@ -181,17 +181,16 @@ func applyCommandOverrides(cmd []string, v variables) variables {
 		v.set("nodefraction", "0")
 		v.set("edgefraction", "0")
 	}
-	if !focus {
-		v.set("focus", "")
-		v.set("ignore", "")
-	}
-	if !tagfocus {
+	if !tagfilter {
 		v.set("tagfocus", "")
 		v.set("tagignore", "")
 	}
-	if !hide {
+	if !filter {
+		v.set("focus", "")
+		v.set("ignore", "")
 		v.set("hide", "")
 		v.set("show", "")
+		v.set("show_from", "")
 	}
 	return v
 }
@@ -242,7 +241,7 @@ func reportOptions(p *profile.Profile, numLabelUnits map[string]string, vars var
 	}
 
 	var filters []string
-	for _, k := range []string{"focus", "ignore", "hide", "show", "tagfocus", "tagignore", "tagshow", "taghide"} {
+	for _, k := range []string{"focus", "ignore", "hide", "show", "show_from", "tagfocus", "tagignore", "tagshow", "taghide"} {
 		v := vars[k].value
 		if v != "" {
 			filters = append(filters, k+"="+v)

internal/driver/driver_focus.go

@@ -33,6 +33,7 @@ func applyFocus(prof *profile.Profile, numLabelUnits map[string]string, v variab
 	ignore, err := compileRegexOption("ignore", v["ignore"].value, err)
 	hide, err := compileRegexOption("hide", v["hide"].value, err)
 	show, err := compileRegexOption("show", v["show"].value, err)
+	showfrom, err := compileRegexOption("show_from", v["show_from"].value, err)
 	tagfocus, err := compileTagFilter("tagfocus", v["tagfocus"].value, numLabelUnits, ui, err)
 	tagignore, err := compileTagFilter("tagignore", v["tagignore"].value, numLabelUnits, ui, err)
 	prunefrom, err := compileRegexOption("prune_from", v["prune_from"].value, err)
@@ -46,6 +47,9 @@ func applyFocus(prof *profile.Profile, numLabelUnits map[string]string, v variab
 	warnNoMatches(hide == nil || hm, "Hide", ui)
 	warnNoMatches(show == nil || hnm, "Show", ui)
 
+	sfm := prof.ShowFrom(showfrom)
+	warnNoMatches(showfrom == nil || sfm, "ShowFrom", ui)
+
 	tfm, tim := prof.FilterSamplesByTag(tagfocus, tagignore)
 	warnNoMatches(tagfocus == nil || tfm, "TagFocus", ui)
 	warnNoMatches(tagignore == nil || tim, "TagIgnore", ui)

internal/driver/driver_test.go

@@ -65,6 +65,7 @@ func TestParse(t *testing.T) {
 		{"topproto,lines,cum,hide=mangled[X3]0", "cpu"},
 		{"tree,lines,cum,focus=[24]00", "heap"},
 		{"tree,relative_percentages,cum,focus=[24]00", "heap"},
+		{"tree,lines,cum,show_from=line2", "cpu"},
 		{"callgrind", "cpu"},
 		{"callgrind,call_tree", "cpu"},
 		{"callgrind", "heap"},
@@ -261,6 +262,9 @@ func solutionFilename(source string, f *testFlags) string {
 	if f.strings["ignore"] != "" || f.strings["tagignore"] != "" {
 		name = append(name, "ignore")
 	}
+	if f.strings["show_from"] != "" {
+		name = append(name, "show_from")
+	}
 	name = addString(name, f, []string{"hide", "show"})
 	if f.strings["unit"] != "minimum" {
 		name = addString(name, f, []string{"unit"})

internal/driver/testdata/pprof.cpu.cum.lines.tree.show_from

@@ -0,0 +1,16 @@
+Active filters:
+   show_from=line2
+Showing nodes accounting for 1.01s, 90.18% of 1.12s total
+----------------------------------------------------------+-------------
+      flat  flat%   sum%        cum   cum%   calls calls% + context 	 	 
+----------------------------------------------------------+-------------
+         0     0%     0%      1.01s 90.18%                | line2000 testdata/file2000.src:4
+                                             1.01s   100% |   line2001 testdata/file2000.src:9 (inline)
+----------------------------------------------------------+-------------
+                                             1.01s   100% |   line2000 testdata/file2000.src:4 (inline)
+     0.01s  0.89%  0.89%      1.01s 90.18%                | line2001 testdata/file2000.src:9
+                                                1s 99.01% |   line1000 testdata/file1000.src:1
+----------------------------------------------------------+-------------
+                                                1s   100% |   line2001 testdata/file2000.src:9
+        1s 89.29% 90.18%         1s 89.29%                | line1000 testdata/file1000.src:1
+----------------------------------------------------------+-------------

profile/filter.go

@@ -74,6 +74,71 @@ func (p *Profile) FilterSamplesByName(focus, ignore, hide, show *regexp.Regexp)
 	return
 }
 
+// ShowFrom drops all stack frames above the highest matching frame and returns
+// whether a match was found. If showFrom is nil it returns false and does not
+// modify the profile.
+//
+// Example: consider a sample with frames [A, B, C, B], where A is the root.
+// ShowFrom(nil) returns false and has frames [A, B, C, B].
+// ShowFrom(A) returns true and has frames [A, B, C, B].
+// ShowFrom(B) returns true and has frames [B, C, B].
+// ShowFrom(C) returns true and has frames [C, B].
+// ShowFrom(D) returns false and drops the sample because no frames remain.
+func (p *Profile) ShowFrom(showFrom *regexp.Regexp) (matched bool) {
+	if showFrom == nil {
+		return false
+	}
+	// showFromLocs stores location IDs that matched ShowFrom.
+	showFromLocs := make(map[uint64]bool)
+	// Apply to locations.
+	for _, loc := range p.Location {
+		if filterShowFromLocation(loc, showFrom) {
+			showFromLocs[loc.ID] = true
+			matched = true
+		}
+	}
+	// For all samples, strip locations after the highest matching one.
+	s := make([]*Sample, 0, len(p.Sample))
+	for _, sample := range p.Sample {
+		for i := len(sample.Location) - 1; i >= 0; i-- {
+			if showFromLocs[sample.Location[i].ID] {
+				sample.Location = sample.Location[:i+1]
+				s = append(s, sample)
+				break
+			}
+		}
+	}
+	p.Sample = s
+	return matched
+}
+
+// filterShowFromLocation tests a showFrom regex against a location, removes
+// lines after the last match and returns whether a match was found. If the
+// mapping is matched, then all lines are kept.
+func filterShowFromLocation(loc *Location, showFrom *regexp.Regexp) bool {
+	if m := loc.Mapping; m != nil && showFrom.MatchString(m.File) {
+		return true
+	}
+	if i := loc.lastMatchedLineIndex(showFrom); i >= 0 {
+		loc.Line = loc.Line[:i+1]
+		return true
+	}
+	return false
+}
+
+// lastMatchedLineIndex returns the index of the last line that matches a regex,
+// or -1 if no match is found.
+func (loc *Location) lastMatchedLineIndex(re *regexp.Regexp) int {
+	for i := len(loc.Line) - 1; i >= 0; i-- {
+		if fn := loc.Line[i].Function; fn != nil {
+			if re.MatchString(fn.Name) || re.MatchString(fn.Filename) {
+				return i
+			}
+		}
+	}
+	return -1
+}
+
 // FilterTagsByName filters the tags in a profile and only keeps
 // tags that match show and not hide.
 func (p *Profile) FilterTagsByName(show, hide *regexp.Regexp) (sm, hm bool) {

profile/filter_test.go

@@ -101,7 +101,29 @@ var inlinesProfile = &Profile{
 	},
 }
 
-func TestFilter(t *testing.T) {
+var emptyLinesLocs = []*Location{
+	{ID: 1, Mapping: mappings[0], Address: 0x1000, Line: []Line{{Function: functions[0], Line: 1}, {Function: functions[1], Line: 1}}},
+	{ID: 2, Mapping: mappings[0], Address: 0x2000, Line: []Line{}},
+	{ID: 3, Mapping: mappings[1], Address: 0x2000, Line: []Line{}},
+}
+
+var emptyLinesProfile = &Profile{
+	TimeNanos:     10000,
+	PeriodType:    &ValueType{Type: "cpu", Unit: "milliseconds"},
+	Period:        1,
+	DurationNanos: 10e9,
+	SampleType:    []*ValueType{{Type: "samples", Unit: "count"}},
+	Mapping:       mappings,
+	Function:      functions,
+	Location:      emptyLinesLocs,
+	Sample: []*Sample{
+		{Value: []int64{1}, Location: []*Location{emptyLinesLocs[0], emptyLinesLocs[1]}},
+		{Value: []int64{2}, Location: []*Location{emptyLinesLocs[2]}},
+		{Value: []int64{3}, Location: []*Location{}},
+	},
+}
+
+func TestFilterSamplesByName(t *testing.T) {
 	for _, tc := range []struct {
 		// name is the name of the test case.
 		name string
@@ -392,6 +414,130 @@ func TestFilter(t *testing.T) {
 	}
 }
 
+func TestShowFrom(t *testing.T) {
+	for _, tc := range []struct {
+		name     string
+		profile  *Profile
+		showFrom *regexp.Regexp
+		// wantMatch is the expected return value.
+		wantMatch bool
+		// wantSampleFuncs contains expected stack functions and sample value after
+		// filtering, in the same order as in the profile. The format is as
+		// returned by sampleFuncs function below, which is "callee caller: <num>".
+		wantSampleFuncs []string
+	}{
+		{
+			name:            "nil showFrom keeps all frames",
+			profile:         noInlinesProfile,
+			wantMatch:       false,
+			wantSampleFuncs: allNoInlinesSampleFuncs,
+		},
+		{
+			name:      "showFrom with no matches drops all samples",
+			profile:   noInlinesProfile,
+			showFrom:  regexp.MustCompile("unknown"),
+			wantMatch: false,
+		},
+		{
+			name:      "showFrom matches function names",
+			profile:   noInlinesProfile,
+			showFrom:  regexp.MustCompile("fun1"),
+			wantMatch: true,
+			wantSampleFuncs: []string{
+				"fun0 fun1: 1",
+				"fun4 fun5 fun1: 2",
+				"fun9 fun4 fun10: 4",
+			},
+		},
+		{
+			name:      "showFrom matches file names",
+			profile:   noInlinesProfile,
+			showFrom:  regexp.MustCompile("file1"),
+			wantMatch: true,
+			wantSampleFuncs: []string{
+				"fun0 fun1: 1",
+				"fun4 fun5 fun1: 2",
+				"fun9 fun4 fun10: 4",
+			},
+		},
+		{
+			name:      "showFrom matches mapping names",
+			profile:   noInlinesProfile,
+			showFrom:  regexp.MustCompile("map1"),
+			wantMatch: true,
+			wantSampleFuncs: []string{
+				"fun9 fun4 fun10: 4",
+			},
+		},
+		{
+			name:      "showFrom drops frames above highest of multiple matches",
+			profile:   noInlinesProfile,
+			showFrom:  regexp.MustCompile("fun[12]"),
+			wantMatch: true,
+			wantSampleFuncs: []string{
+				"fun0 fun1 fun2: 1",
+				"fun4 fun5 fun1: 2",
+				"fun9 fun4 fun10: 4",
+			},
+		},
+		{
+			name:      "showFrom matches inline functions",
+			profile:   inlinesProfile,
+			showFrom:  regexp.MustCompile("fun0|fun5"),
+			wantMatch: true,
+			wantSampleFuncs: []string{
+				"fun0: 1",
+				"fun4 fun5: 2",
+			},
+		},
+		{
+			name:      "showFrom drops frames above highest of multiple inline matches",
+			profile:   inlinesProfile,
+			showFrom:  regexp.MustCompile("fun[1245]"),
+			wantMatch: true,
+			wantSampleFuncs: []string{
+				"fun0 fun1 fun2: 1",
+				"fun4 fun5: 2",
+			},
+		},
+		{
+			name:      "showFrom keeps all lines when matching mapping and function",
+			profile:   inlinesProfile,
+			showFrom:  regexp.MustCompile("map0|fun5"),
+			wantMatch: true,
+			wantSampleFuncs: []string{
+				"fun0 fun1 fun2 fun3: 1",
+				"fun4 fun5 fun6: 2",
+			},
+		},
+		{
+			name:      "showFrom matches location with empty lines",
+			profile:   emptyLinesProfile,
+			showFrom:  regexp.MustCompile("map1"),
+			wantMatch: true,
+			wantSampleFuncs: []string{
+				": 2",
+			},
+		},
+	} {
+		t.Run(tc.name, func(t *testing.T) {
+			p := tc.profile.Copy()
+
+			if gotMatch := p.ShowFrom(tc.showFrom); gotMatch != tc.wantMatch {
+				t.Errorf("match got %+v, want %+v", gotMatch, tc.wantMatch)
+			}
+
+			if got, want := strings.Join(sampleFuncs(p), "\n")+"\n", strings.Join(tc.wantSampleFuncs, "\n")+"\n"; got != want {
+				diff, err := proftest.Diff([]byte(want), []byte(got))
+				if err != nil {
+					t.Fatalf("failed to get diff: %v", err)
+				}
+				t.Errorf("profile samples got diff(want->got):\n%s", diff)
+			}
+		})
+	}
+}
+
 // sampleFuncs returns a slice of strings where each string represents one
 // profile sample in the format "<fun1> <fun2> <fun3>: <value>". This allows
 // the expected values for test cases to be specifed in human-readable strings.