vendor: update buildkit to 1e6032c

Signed-off-by: CrazyMax <crazy-max@users.noreply.github.com>

vendor/go.opentelemetry.io/otel/sdk/trace/attributesmap.go

@@ -1,91 +0,0 @@
-// Copyright The OpenTelemetry Authors
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
-
-package trace // import "go.opentelemetry.io/otel/sdk/trace"
-
-import (
-	"container/list"
-
-	"go.opentelemetry.io/otel/attribute"
-)
-
-// attributesMap is a capped map of attributes, holding the most recent attributes.
-// Eviction is done via a LRU method, the oldest entry is removed to create room for a new entry.
-// Updates are allowed and they refresh the usage of the key.
-//
-// This is based from https://github.com/hashicorp/golang-lru/blob/master/simplelru/lru.go
-// With a subset of the its operations and specific for holding attribute.KeyValue
-type attributesMap struct {
-	attributes   map[attribute.Key]*list.Element
-	evictList    *list.List
-	droppedCount int
-	capacity     int
-}
-
-func newAttributesMap(capacity int) *attributesMap {
-	lm := &attributesMap{
-		attributes: make(map[attribute.Key]*list.Element),
-		evictList:  list.New(),
-		capacity:   capacity,
-	}
-	return lm
-}
-
-func (am *attributesMap) add(kv attribute.KeyValue) {
-	// Check for existing item
-	if ent, ok := am.attributes[kv.Key]; ok {
-		am.evictList.MoveToFront(ent)
-		ent.Value = &kv
-		return
-	}
-
-	// Add new item
-	entry := am.evictList.PushFront(&kv)
-	am.attributes[kv.Key] = entry
-
-	// Verify size not exceeded
-	if am.evictList.Len() > am.capacity {
-		am.removeOldest()
-		am.droppedCount++
-	}
-}
-
-// toKeyValue copies the attributesMap into a slice of attribute.KeyValue and
-// returns it. If the map is empty, a nil is returned.
-// TODO: Is it more efficient to return a pointer to the slice?
-func (am *attributesMap) toKeyValue() []attribute.KeyValue {
-	len := am.evictList.Len()
-	if len == 0 {
-		return nil
-	}
-
-	attributes := make([]attribute.KeyValue, 0, len)
-	for ent := am.evictList.Back(); ent != nil; ent = ent.Prev() {
-		if value, ok := ent.Value.(*attribute.KeyValue); ok {
-			attributes = append(attributes, *value)
-		}
-	}
-
-	return attributes
-}
-
-// removeOldest removes the oldest item from the cache.
-func (am *attributesMap) removeOldest() {
-	ent := am.evictList.Back()
-	if ent != nil {
-		am.evictList.Remove(ent)
-		kv := ent.Value.(*attribute.KeyValue)
-		delete(am.attributes, kv.Key)
-	}
-}

vendor/go.opentelemetry.io/otel/sdk/trace/batch_span_processor.go

@@ -23,14 +23,15 @@ import (
 
 	"go.opentelemetry.io/otel"
 	"go.opentelemetry.io/otel/internal/global"
+	"go.opentelemetry.io/otel/sdk/internal/env"
 	"go.opentelemetry.io/otel/trace"
 )
 
 // Defaults for BatchSpanProcessorOptions.
 const (
 	DefaultMaxQueueSize       = 2048
-	DefaultBatchTimeout       = 5000 * time.Millisecond
-	DefaultExportTimeout      = 30000 * time.Millisecond
+	DefaultScheduleDelay      = 5000
+	DefaultExportTimeout      = 30000
 	DefaultMaxExportBatchSize = 512
 )
 
@@ -89,11 +90,22 @@ var _ SpanProcessor = (*batchSpanProcessor)(nil)
 //
 // If the exporter is nil, the span processor will preform no action.
 func NewBatchSpanProcessor(exporter SpanExporter, options ...BatchSpanProcessorOption) SpanProcessor {
+	maxQueueSize := env.BatchSpanProcessorMaxQueueSize(DefaultMaxQueueSize)
+	maxExportBatchSize := env.BatchSpanProcessorMaxExportBatchSize(DefaultMaxExportBatchSize)
+
+	if maxExportBatchSize > maxQueueSize {
+		if DefaultMaxExportBatchSize > maxQueueSize {
+			maxExportBatchSize = maxQueueSize
+		} else {
+			maxExportBatchSize = DefaultMaxExportBatchSize
+		}
+	}
+
 	o := BatchSpanProcessorOptions{
-		BatchTimeout:       DefaultBatchTimeout,
-		ExportTimeout:      DefaultExportTimeout,
-		MaxQueueSize:       DefaultMaxQueueSize,
-		MaxExportBatchSize: DefaultMaxExportBatchSize,
+		BatchTimeout:       time.Duration(env.BatchSpanProcessorScheduleDelay(DefaultScheduleDelay)) * time.Millisecond,
+		ExportTimeout:      time.Duration(env.BatchSpanProcessorExportTimeout(DefaultExportTimeout)) * time.Millisecond,
+		MaxQueueSize:       maxQueueSize,
+		MaxExportBatchSize: maxExportBatchSize,
 	}
 	for _, opt := range options {
 		opt(&o)
@@ -238,7 +250,7 @@ func (bsp *batchSpanProcessor) exportSpans(ctx context.Context) error {
 	}
 
 	if l := len(bsp.batch); l > 0 {
-		global.Debug("exporting spans", "count", len(bsp.batch))
+		global.Debug("exporting spans", "count", len(bsp.batch), "dropped", bsp.dropped)
 		err := bsp.e.ExportSpans(ctx, bsp.batch)
 
 		// A new batch is always created after exporting, even if the batch failed to be exported.
@@ -369,3 +381,16 @@ func (bsp *batchSpanProcessor) enqueueBlockOnQueueFull(ctx context.Context, sd R
 	}
 	return false
 }
+
+// MarshalLog is the marshaling function used by the logging system to represent this exporter.
+func (bsp *batchSpanProcessor) MarshalLog() interface{} {
+	return struct {
+		Type         string
+		SpanExporter SpanExporter
+		Config       BatchSpanProcessorOptions
+	}{
+		Type:         "BatchSpanProcessor",
+		SpanExporter: bsp.e,
+		Config:       bsp.o,
+	}
+}

vendor/go.opentelemetry.io/otel/sdk/trace/evictedqueue.go

@@ -14,24 +14,25 @@
 
 package trace // import "go.opentelemetry.io/otel/sdk/trace"
 
+// evictedQueue is a FIFO queue with a configurable capacity.
 type evictedQueue struct {
 	queue        []interface{}
 	capacity     int
 	droppedCount int
 }
 
-func newEvictedQueue(capacity int) *evictedQueue {
-	eq := &evictedQueue{
-		capacity: capacity,
-		queue:    make([]interface{}, 0),
-	}
-
-	return eq
+func newEvictedQueue(capacity int) evictedQueue {
+	// Do not pre-allocate queue, do this lazily.
+	return evictedQueue{capacity: capacity}
 }
 
+// add adds value to the evictedQueue eq. If eq is at capacity, the oldest
+// queued value will be discarded and the drop count incremented.
 func (eq *evictedQueue) add(value interface{}) {
 	if len(eq.queue) == eq.capacity {
-		eq.queue = eq.queue[1:]
+		// Drop first-in while avoiding allocating more capacity to eq.queue.
+		copy(eq.queue[:eq.capacity-1], eq.queue[1:])
+		eq.queue = eq.queue[:eq.capacity-1]
 		eq.droppedCount++
 	}
 	eq.queue = append(eq.queue, value)

vendor/go.opentelemetry.io/otel/sdk/trace/provider.go

@@ -21,6 +21,7 @@ import (
 	"sync/atomic"
 
 	"go.opentelemetry.io/otel"
+	"go.opentelemetry.io/otel/internal/global"
 	"go.opentelemetry.io/otel/sdk/instrumentation"
 	"go.opentelemetry.io/otel/sdk/resource"
 	"go.opentelemetry.io/otel/trace"
@@ -52,14 +53,34 @@ type tracerProviderConfig struct {
 	resource *resource.Resource
 }
 
+// MarshalLog is the marshaling function used by the logging system to represent this exporter.
+func (cfg tracerProviderConfig) MarshalLog() interface{} {
+	return struct {
+		SpanProcessors  []SpanProcessor
+		SamplerType     string
+		IDGeneratorType string
+		SpanLimits      SpanLimits
+		Resource        *resource.Resource
+	}{
+		SpanProcessors:  cfg.processors,
+		SamplerType:     fmt.Sprintf("%T", cfg.sampler),
+		IDGeneratorType: fmt.Sprintf("%T", cfg.idGenerator),
+		SpanLimits:      cfg.spanLimits,
+		Resource:        cfg.resource,
+	}
+}
+
 type TracerProvider struct {
 	mu             sync.Mutex
 	namedTracer    map[instrumentation.Library]*tracer
 	spanProcessors atomic.Value
-	sampler        Sampler
-	idGenerator    IDGenerator
-	spanLimits     SpanLimits
-	resource       *resource.Resource
+
+	// These fields are not protected by the lock mu. They are assumed to be
+	// immutable after creation of the TracerProvider.
+	sampler     Sampler
+	idGenerator IDGenerator
+	spanLimits  SpanLimits
+	resource    *resource.Resource
 }
 
 var _ trace.TracerProvider = &TracerProvider{}
@@ -75,13 +96,13 @@ var _ trace.TracerProvider = &TracerProvider{}
 // The passed opts are used to override these default values and configure the
 // returned TracerProvider appropriately.
 func NewTracerProvider(opts ...TracerProviderOption) *TracerProvider {
-	o := &tracerProviderConfig{}
+	o := tracerProviderConfig{}
 
 	for _, opt := range opts {
-		opt.apply(o)
+		o = opt.apply(o)
 	}
 
-	ensureValidTracerProviderConfig(o)
+	o = ensureValidTracerProviderConfig(o)
 
 	tp := &TracerProvider{
 		namedTracer: make(map[instrumentation.Library]*tracer),
@@ -91,6 +112,8 @@ func NewTracerProvider(opts ...TracerProviderOption) *TracerProvider {
 		resource:    o.resource,
 	}
 
+	global.Info("TracerProvider created", "config", o)
+
 	for _, sp := range o.processors {
 		tp.RegisterSpanProcessor(sp)
 	}
@@ -125,6 +148,7 @@ func (p *TracerProvider) Tracer(name string, opts ...trace.TracerOption) trace.T
 			instrumentationLibrary: il,
 		}
 		p.namedTracer[il] = t
+		global.Info("Tracer created", "name", name, "version", c.InstrumentationVersion(), "schemaURL", c.SchemaURL())
 	}
 	return t
 }
@@ -235,13 +259,13 @@ func (p *TracerProvider) Shutdown(ctx context.Context) error {
 }
 
 type TracerProviderOption interface {
-	apply(*tracerProviderConfig)
+	apply(tracerProviderConfig) tracerProviderConfig
 }
 
-type traceProviderOptionFunc func(*tracerProviderConfig)
+type traceProviderOptionFunc func(tracerProviderConfig) tracerProviderConfig
 
-func (fn traceProviderOptionFunc) apply(cfg *tracerProviderConfig) {
-	fn(cfg)
+func (fn traceProviderOptionFunc) apply(cfg tracerProviderConfig) tracerProviderConfig {
+	return fn(cfg)
 }
 
 // WithSyncer registers the exporter with the TracerProvider using a
@@ -264,8 +288,9 @@ func WithBatcher(e SpanExporter, opts ...BatchSpanProcessorOption) TracerProvide
 
 // WithSpanProcessor registers the SpanProcessor with a TracerProvider.
 func WithSpanProcessor(sp SpanProcessor) TracerProviderOption {
-	return traceProviderOptionFunc(func(cfg *tracerProviderConfig) {
+	return traceProviderOptionFunc(func(cfg tracerProviderConfig) tracerProviderConfig {
 		cfg.processors = append(cfg.processors, sp)
+		return cfg
 	})
 }
 
@@ -277,12 +302,13 @@ func WithSpanProcessor(sp SpanProcessor) TracerProviderOption {
 // If this option is not used, the TracerProvider will use the
 // resource.Default() Resource by default.
 func WithResource(r *resource.Resource) TracerProviderOption {
-	return traceProviderOptionFunc(func(cfg *tracerProviderConfig) {
+	return traceProviderOptionFunc(func(cfg tracerProviderConfig) tracerProviderConfig {
 		var err error
 		cfg.resource, err = resource.Merge(resource.Environment(), r)
 		if err != nil {
 			otel.Handle(err)
 		}
+		return cfg
 	})
 }
 
@@ -294,10 +320,11 @@ func WithResource(r *resource.Resource) TracerProviderOption {
 // If this option is not used, the TracerProvider will use a random number
 // IDGenerator by default.
 func WithIDGenerator(g IDGenerator) TracerProviderOption {
-	return traceProviderOptionFunc(func(cfg *tracerProviderConfig) {
+	return traceProviderOptionFunc(func(cfg tracerProviderConfig) tracerProviderConfig {
 		if g != nil {
 			cfg.idGenerator = g
 		}
+		return cfg
 	})
 }
 
@@ -309,10 +336,11 @@ func WithIDGenerator(g IDGenerator) TracerProviderOption {
 // If this option is not used, the TracerProvider will use a
 // ParentBased(AlwaysSample) Sampler by default.
 func WithSampler(s Sampler) TracerProviderOption {
-	return traceProviderOptionFunc(func(cfg *tracerProviderConfig) {
+	return traceProviderOptionFunc(func(cfg tracerProviderConfig) tracerProviderConfig {
 		if s != nil {
 			cfg.sampler = s
 		}
+		return cfg
 	})
 }
 
@@ -324,13 +352,14 @@ func WithSampler(s Sampler) TracerProviderOption {
 // If this option is not used, the TracerProvider will use the default
 // SpanLimits.
 func WithSpanLimits(sl SpanLimits) TracerProviderOption {
-	return traceProviderOptionFunc(func(cfg *tracerProviderConfig) {
+	return traceProviderOptionFunc(func(cfg tracerProviderConfig) tracerProviderConfig {
 		cfg.spanLimits = sl
+		return cfg
 	})
 }
 
 // ensureValidTracerProviderConfig ensures that given TracerProviderConfig is valid.
-func ensureValidTracerProviderConfig(cfg *tracerProviderConfig) {
+func ensureValidTracerProviderConfig(cfg tracerProviderConfig) tracerProviderConfig {
 	if cfg.sampler == nil {
 		cfg.sampler = ParentBased(AlwaysSample())
 	}
@@ -341,4 +370,5 @@ func ensureValidTracerProviderConfig(cfg *tracerProviderConfig) {
 	if cfg.resource == nil {
 		cfg.resource = resource.Default()
 	}
+	return cfg
 }

vendor/go.opentelemetry.io/otel/sdk/trace/sampling.go

@@ -187,7 +187,7 @@ func configureSamplersForParentBased(samplers []ParentBasedSamplerOption) sample
 	}
 
 	for _, so := range samplers {
-		so.apply(&c)
+		c = so.apply(c)
 	}
 
 	return c
@@ -201,7 +201,7 @@ type samplerConfig struct {
 
 // ParentBasedSamplerOption configures the sampler for a particular sampling case.
 type ParentBasedSamplerOption interface {
-	apply(*samplerConfig)
+	apply(samplerConfig) samplerConfig
 }
 
 // WithRemoteParentSampled sets the sampler for the case of sampled remote parent.
@@ -213,8 +213,9 @@ type remoteParentSampledOption struct {
 	s Sampler
 }
 
-func (o remoteParentSampledOption) apply(config *samplerConfig) {
+func (o remoteParentSampledOption) apply(config samplerConfig) samplerConfig {
 	config.remoteParentSampled = o.s
+	return config
 }
 
 // WithRemoteParentNotSampled sets the sampler for the case of remote parent
@@ -227,8 +228,9 @@ type remoteParentNotSampledOption struct {
 	s Sampler
 }
 
-func (o remoteParentNotSampledOption) apply(config *samplerConfig) {
+func (o remoteParentNotSampledOption) apply(config samplerConfig) samplerConfig {
 	config.remoteParentNotSampled = o.s
+	return config
 }
 
 // WithLocalParentSampled sets the sampler for the case of sampled local parent.
@@ -240,8 +242,9 @@ type localParentSampledOption struct {
 	s Sampler
 }
 
-func (o localParentSampledOption) apply(config *samplerConfig) {
+func (o localParentSampledOption) apply(config samplerConfig) samplerConfig {
 	config.localParentSampled = o.s
+	return config
 }
 
 // WithLocalParentNotSampled sets the sampler for the case of local parent
@@ -254,8 +257,9 @@ type localParentNotSampledOption struct {
 	s Sampler
 }
 
-func (o localParentNotSampledOption) apply(config *samplerConfig) {
+func (o localParentNotSampledOption) apply(config samplerConfig) samplerConfig {
 	config.localParentNotSampled = o.s
+	return config
 }
 
 func (pb parentBased) ShouldSample(p SamplingParameters) SamplingResult {

vendor/go.opentelemetry.io/otel/sdk/trace/simple_span_processor.go

@@ -115,3 +115,14 @@ func (ssp *simpleSpanProcessor) Shutdown(ctx context.Context) error {
 func (ssp *simpleSpanProcessor) ForceFlush(context.Context) error {
 	return nil
 }
+
+// MarshalLog is the marshaling function used by the logging system to represent this exporter.
+func (ssp *simpleSpanProcessor) MarshalLog() interface{} {
+	return struct {
+		Type     string
+		Exporter SpanExporter
+	}{
+		Type:     "SimpleSpanProcessor",
+		Exporter: ssp.exporter,
+	}
+}

vendor/go.opentelemetry.io/otel/sdk/trace/span.go

@@ -54,6 +54,7 @@ type ReadOnlySpan interface {
 	// the span has not ended.
 	EndTime() time.Time
 	// Attributes returns the defining attributes of the span.
+	// The order of the returned attributes is not guaranteed to be stable across invocations.
 	Attributes() []attribute.KeyValue
 	// Links returns all the links the span has to other spans.
 	Links() []Link
@@ -126,35 +127,29 @@ type recordingSpan struct {
 	// childSpanCount holds the number of child spans created for this span.
 	childSpanCount int
 
-	// resource contains attributes representing an entity that produced this
-	// span.
-	resource *resource.Resource
-
-	// instrumentationLibrary defines the instrumentation library used to
-	// provide instrumentation.
-	instrumentationLibrary instrumentation.Library
-
 	// spanContext holds the SpanContext of this span.
 	spanContext trace.SpanContext
 
-	// attributes are capped at configured limit. When the capacity is reached
-	// an oldest entry is removed to create room for a new entry.
-	attributes *attributesMap
+	// attributes is a collection of user provided key/values. The collection
+	// is constrained by a configurable maximum held by the parent
+	// TracerProvider. When additional attributes are added after this maximum
+	// is reached these attributes the user is attempting to add are dropped.
+	// This dropped number of attributes is tracked and reported in the
+	// ReadOnlySpan exported when the span ends.
+	attributes        []attribute.KeyValue
+	droppedAttributes int
 
 	// events are stored in FIFO queue capped by configured limit.
-	events *evictedQueue
+	events evictedQueue
 
 	// links are stored in FIFO queue capped by configured limit.
-	links *evictedQueue
+	links evictedQueue
 
 	// executionTracerTaskEnd ends the execution tracer span.
 	executionTracerTaskEnd func()
 
 	// tracer is the SDK tracer that created this span.
 	tracer *tracer
-
-	// spanLimits holds the limits to this span.
-	spanLimits SpanLimits
 }
 
 var _ ReadWriteSpan = (*recordingSpan)(nil)
@@ -205,11 +200,80 @@ func (s *recordingSpan) SetStatus(code codes.Code, description string) {
 // will be overwritten with the value contained in attributes.
 //
 // If this span is not being recorded than this method does nothing.
+//
+// If adding attributes to the span would exceed the maximum amount of
+// attributes the span is configured to have, the last added attributes will
+// be dropped.
 func (s *recordingSpan) SetAttributes(attributes ...attribute.KeyValue) {
 	if !s.IsRecording() {
 		return
 	}
-	s.copyToCappedAttributes(attributes...)
+
+	s.mu.Lock()
+	defer s.mu.Unlock()
+
+	// If adding these attributes could exceed the capacity of s perform a
+	// de-duplication and truncation while adding to avoid over allocation.
+	if len(s.attributes)+len(attributes) > s.tracer.provider.spanLimits.AttributeCountLimit {
+		s.addOverCapAttrs(attributes)
+		return
+	}
+
+	// Otherwise, add without deduplication. When attributes are read they
+	// will be deduplicated, optimizing the operation.
+	for _, a := range attributes {
+		if !a.Valid() {
+			// Drop all invalid attributes.
+			s.droppedAttributes++
+			continue
+		}
+		s.attributes = append(s.attributes, a)
+	}
+}
+
+// addOverCapAttrs adds the attributes attrs to the span s while
+// de-duplicating the attributes of s and attrs and dropping attributes that
+// exceed the capacity of s.
+//
+// This method assumes s.mu.Lock is held by the caller.
+//
+// This method should only be called when there is a possibility that adding
+// attrs to s will exceed the capacity of s. Otherwise, attrs should be added
+// to s without checking for duplicates and all retrieval methods of the
+// attributes for s will de-duplicate as needed.
+func (s *recordingSpan) addOverCapAttrs(attrs []attribute.KeyValue) {
+	// In order to not allocate more capacity to s.attributes than needed,
+	// prune and truncate this addition of attributes while adding.
+
+	// Do not set a capacity when creating this map. Benchmark testing has
+	// showed this to only add unused memory allocations in general use.
+	exists := make(map[attribute.Key]int)
+	s.dedupeAttrsFromRecord(&exists)
+
+	// Now that s.attributes is deduplicated, adding unique attributes up to
+	// the capacity of s will not over allocate s.attributes.
+	for _, a := range attrs {
+		if !a.Valid() {
+			// Drop all invalid attributes.
+			s.droppedAttributes++
+			continue
+		}
+
+		if idx, ok := exists[a.Key]; ok {
+			// Perform all updates before dropping, even when at capacity.
+			s.attributes[idx] = a
+			continue
+		}
+
+		if len(s.attributes) >= s.tracer.provider.spanLimits.AttributeCountLimit {
+			// Do not just drop all of the remaining attributes, make sure
+			// updates are checked and performed.
+			s.droppedAttributes++
+		} else {
+			s.attributes = append(s.attributes, a)
+			exists[a.Key] = len(s.attributes) - 1
+		}
+	}
 }
 
 // End ends the span. This method does nothing if the span is already ended or
@@ -336,9 +400,9 @@ func (s *recordingSpan) addEvent(name string, o ...trace.EventOption) {
 	// Discard over limited attributes
 	attributes := c.Attributes()
 	var discarded int
-	if len(attributes) > s.spanLimits.AttributePerEventCountLimit {
-		discarded = len(attributes) - s.spanLimits.AttributePerEventCountLimit
-		attributes = attributes[:s.spanLimits.AttributePerEventCountLimit]
+	if len(attributes) > s.tracer.provider.spanLimits.AttributePerEventCountLimit {
+		discarded = len(attributes) - s.tracer.provider.spanLimits.AttributePerEventCountLimit
+		attributes = attributes[:s.tracer.provider.spanLimits.AttributePerEventCountLimit]
 	}
 	s.mu.Lock()
 	defer s.mu.Unlock()
@@ -399,13 +463,45 @@ func (s *recordingSpan) EndTime() time.Time {
 }
 
 // Attributes returns the attributes of this span.
+//
+// The order of the returned attributes is not guaranteed to be stable.
 func (s *recordingSpan) Attributes() []attribute.KeyValue {
 	s.mu.Lock()
 	defer s.mu.Unlock()
-	if s.attributes.evictList.Len() == 0 {
-		return []attribute.KeyValue{}
+	s.dedupeAttrs()
+	return s.attributes
+}
+
+// dedupeAttrs deduplicates the attributes of s to fit capacity.
+//
+// This method assumes s.mu.Lock is held by the caller.
+func (s *recordingSpan) dedupeAttrs() {
+	// Do not set a capacity when creating this map. Benchmark testing has
+	// showed this to only add unused memory allocations in general use.
+	exists := make(map[attribute.Key]int)
+	s.dedupeAttrsFromRecord(&exists)
+}
+
+// dedupeAttrsFromRecord deduplicates the attributes of s to fit capacity
+// using record as the record of unique attribute keys to their index.
+//
+// This method assumes s.mu.Lock is held by the caller.
+func (s *recordingSpan) dedupeAttrsFromRecord(record *map[attribute.Key]int) {
+	// Use the fact that slices share the same backing array.
+	unique := s.attributes[:0]
+	for _, a := range s.attributes {
+		if idx, ok := (*record)[a.Key]; ok {
+			unique[idx] = a
+		} else {
+			unique = append(unique, a)
+			(*record)[a.Key] = len(unique) - 1
+		}
 	}
-	return s.attributes.toKeyValue()
+	// s.attributes have element types of attribute.KeyValue. These types are
+	// not pointers and they themselves do not contain pointer fields,
+	// therefore the duplicate values do not need to be zeroed for them to be
+	// garbage collected.
+	s.attributes = unique
 }
 
 // Links returns the links of this span.
@@ -440,7 +536,7 @@ func (s *recordingSpan) Status() Status {
 func (s *recordingSpan) InstrumentationLibrary() instrumentation.Library {
 	s.mu.Lock()
 	defer s.mu.Unlock()
-	return s.instrumentationLibrary
+	return s.tracer.instrumentationLibrary
 }
 
 // Resource returns the Resource associated with the Tracer that created this
@@ -448,7 +544,7 @@ func (s *recordingSpan) InstrumentationLibrary() instrumentation.Library {
 func (s *recordingSpan) Resource() *resource.Resource {
 	s.mu.Lock()
 	defer s.mu.Unlock()
-	return s.resource
+	return s.tracer.provider.resource
 }
 
 func (s *recordingSpan) addLink(link trace.Link) {
@@ -461,9 +557,9 @@ func (s *recordingSpan) addLink(link trace.Link) {
 	var droppedAttributeCount int
 
 	// Discard over limited attributes
-	if len(link.Attributes) > s.spanLimits.AttributePerLinkCountLimit {
-		droppedAttributeCount = len(link.Attributes) - s.spanLimits.AttributePerLinkCountLimit
-		link.Attributes = link.Attributes[:s.spanLimits.AttributePerLinkCountLimit]
+	if len(link.Attributes) > s.tracer.provider.spanLimits.AttributePerLinkCountLimit {
+		droppedAttributeCount = len(link.Attributes) - s.tracer.provider.spanLimits.AttributePerLinkCountLimit
+		link.Attributes = link.Attributes[:s.tracer.provider.spanLimits.AttributePerLinkCountLimit]
 	}
 
 	s.links.add(Link{link.SpanContext, link.Attributes, droppedAttributeCount})
@@ -474,7 +570,7 @@ func (s *recordingSpan) addLink(link trace.Link) {
 func (s *recordingSpan) DroppedAttributes() int {
 	s.mu.Lock()
 	defer s.mu.Unlock()
-	return s.attributes.droppedCount
+	return s.droppedAttributes
 }
 
 // DroppedLinks returns the number of links dropped by the span due to limits
@@ -514,20 +610,21 @@ func (s *recordingSpan) snapshot() ReadOnlySpan {
 	defer s.mu.Unlock()
 
 	sd.endTime = s.endTime
-	sd.instrumentationLibrary = s.instrumentationLibrary
+	sd.instrumentationLibrary = s.tracer.instrumentationLibrary
 	sd.name = s.name
 	sd.parent = s.parent
-	sd.resource = s.resource
+	sd.resource = s.tracer.provider.resource
 	sd.spanContext = s.spanContext
 	sd.spanKind = s.spanKind
 	sd.startTime = s.startTime
 	sd.status = s.status
 	sd.childSpanCount = s.childSpanCount
 
-	if s.attributes.evictList.Len() > 0 {
-		sd.attributes = s.attributes.toKeyValue()
-		sd.droppedAttributeCount = s.attributes.droppedCount
+	if len(s.attributes) > 0 {
+		s.dedupeAttrs()
+		sd.attributes = s.attributes
 	}
+	sd.droppedAttributeCount = s.droppedAttributes
 	if len(s.events.queue) > 0 {
 		sd.events = s.interfaceArrayToEventArray()
 		sd.droppedEventCount = s.events.droppedCount
@@ -555,18 +652,6 @@ func (s *recordingSpan) interfaceArrayToEventArray() []Event {
 	return eventArr
 }
 
-func (s *recordingSpan) copyToCappedAttributes(attributes ...attribute.KeyValue) {
-	s.mu.Lock()
-	defer s.mu.Unlock()
-	for _, a := range attributes {
-		// Ensure attributes conform to the specification:
-		// https://github.com/open-telemetry/opentelemetry-specification/blob/v1.0.1/specification/common/common.md#attributes
-		if a.Valid() {
-			s.attributes.add(a)
-		}
-	}
-}
-
 func (s *recordingSpan) addChild() {
 	if !s.IsRecording() {
 		return

vendor/go.opentelemetry.io/otel/sdk/trace/tracer.go

@@ -122,18 +122,22 @@ func (tr *tracer) newRecordingSpan(psc, sc trace.SpanContext, name string, sr Sa
 	}
 
 	s := &recordingSpan{
-		parent:                 psc,
-		spanContext:            sc,
-		spanKind:               trace.ValidateSpanKind(config.SpanKind()),
-		name:                   name,
-		startTime:              startTime,
-		attributes:             newAttributesMap(tr.provider.spanLimits.AttributeCountLimit),
-		events:                 newEvictedQueue(tr.provider.spanLimits.EventCountLimit),
-		links:                  newEvictedQueue(tr.provider.spanLimits.LinkCountLimit),
-		tracer:                 tr,
-		spanLimits:             tr.provider.spanLimits,
-		resource:               tr.provider.resource,
-		instrumentationLibrary: tr.instrumentationLibrary,
+		// Do not pre-allocate the attributes slice here! Doing so will
+		// allocate memory that is likely never going to be used, or if used,
+		// will be over-sized. The default Go compiler has been tested to
+		// dynamically allocate needed space very well. Benchmarking has shown
+		// it to be more performant than what we can predetermine here,
+		// especially for the common use case of few to no added
+		// attributes.
+
+		parent:      psc,
+		spanContext: sc,
+		spanKind:    trace.ValidateSpanKind(config.SpanKind()),
+		name:        name,
+		startTime:   startTime,
+		events:      newEvictedQueue(tr.provider.spanLimits.EventCountLimit),
+		links:       newEvictedQueue(tr.provider.spanLimits.LinkCountLimit),
+		tracer:      tr,
 	}
 
 	for _, l := range config.Links() {