core(lantern): allow non-XHRs to depend on CPU Nodes (#11767)

GoogleChrome · Dec 11, 2020 · 7915708 · 7915708
1 parent 2aa9845
commit 7915708
Show file tree

Hide file tree

Showing 20 changed files with 162 additions and 149 deletions.
diff --git a/lighthouse-core/computed/page-dependency-graph.js b/lighthouse-core/computed/page-dependency-graph.js
@@ -186,15 +186,26 @@ class PageDependencyGraph {
    * @param {Array<CPUNode>} cpuNodes
    */
   static linkCPUNodes(rootNode, networkNodeOutput, cpuNodes) {
+    /** @type {Set<LH.Crdp.Network.ResourceType|undefined>} */
+    const linkableResourceTypes = new Set([
+      NetworkRequest.TYPES.XHR, NetworkRequest.TYPES.Fetch, NetworkRequest.TYPES.Script,
+    ]);
+
     /** @param {CPUNode} cpuNode @param {string} reqId */
     function addDependentNetworkRequest(cpuNode, reqId) {
       const networkNode = networkNodeOutput.idToNodeMap.get(reqId);
       if (!networkNode ||
-          // Ignore all non-XHRs
-          networkNode.record.resourceType !== NetworkRequest.TYPES.XHR ||
           // Ignore all network nodes that started before this CPU task started
           // A network request that started earlier could not possibly have been started by this task
           networkNode.startTime <= cpuNode.startTime) return;
+      const {record} = networkNode;
+      const resourceType = record.resourceType ||
+        record.redirectDestination && record.redirectDestination.resourceType;
+      if (!linkableResourceTypes.has(resourceType)) {
+        // We only link some resources to CPU nodes because we observe LCP simulation
+        // regressions when including images, etc.
+        return;
+      }
       cpuNode.addDependent(networkNode);
     }
 

diff --git a/lighthouse-core/test/audits/__snapshots__/metrics-test.js.snap b/lighthouse-core/test/audits/__snapshots__/metrics-test.js.snap
@@ -59,13 +59,13 @@ Object {
   "cumulativeLayoutShiftAllFrames": 0,
   "estimatedInputLatency": 543,
   "estimatedInputLatencyTs": undefined,
-  "firstCPUIdle": 3677,
+  "firstCPUIdle": 3790,
   "firstCPUIdleTs": undefined,
   "firstContentfulPaint": 2289,
   "firstContentfulPaintTs": undefined,
   "firstMeaningfulPaint": 2758,
   "firstMeaningfulPaintTs": undefined,
-  "interactive": 4462,
+  "interactive": 4671,
   "interactiveTs": undefined,
   "largestContentfulPaint": 2758,
   "largestContentfulPaintAllFrames": undefined,
@@ -102,7 +102,7 @@ Object {
   "observedTraceEndTs": 713044439102,
   "speedIndex": 3681,
   "speedIndexTs": undefined,
-  "totalBlockingTime": 1167,
+  "totalBlockingTime": 1205,
 }
 `;
 
@@ -163,15 +163,15 @@ exports[`Performance: metrics evaluates valid input correctly 1`] = `
 Object {
   "cumulativeLayoutShift": 0,
   "cumulativeLayoutShiftAllFrames": 0,
-  "estimatedInputLatency": 81,
+  "estimatedInputLatency": 78,
   "estimatedInputLatencyTs": undefined,
-  "firstCPUIdle": 3351,
+  "firstCPUIdle": 4313,
   "firstCPUIdleTs": undefined,
   "firstContentfulPaint": 1337,
   "firstContentfulPaintTs": undefined,
   "firstMeaningfulPaint": 1553,
   "firstMeaningfulPaintTs": undefined,
-  "interactive": 3427,
+  "interactive": 4348,
   "interactiveTs": undefined,
   "largestContentfulPaint": undefined,
   "largestContentfulPaintAllFrames": undefined,

diff --git a/lighthouse-core/test/audits/__snapshots__/predictive-perf-test.js.snap b/lighthouse-core/test/audits/__snapshots__/predictive-perf-test.js.snap
@@ -7,21 +7,21 @@ Object {
   "optimisticFMP": 2289,
   "optimisticLCP": 2289,
   "optimisticSI": 1393,
-  "optimisticTTFCPUI": 3677,
-  "optimisticTTI": 3677,
+  "optimisticTTFCPUI": 3790,
+  "optimisticTTI": 3790,
   "pessimisticEIL": 904,
   "pessimisticFCP": 2289,
   "pessimisticFMP": 3228,
   "pessimisticLCP": 3228,
   "pessimisticSI": 3047,
-  "pessimisticTTFCPUI": 5248,
-  "pessimisticTTI": 5248,
+  "pessimisticTTFCPUI": 5551,
+  "pessimisticTTI": 5551,
   "roughEstimateOfEIL": 543,
   "roughEstimateOfFCP": 2289,
   "roughEstimateOfFMP": 2758,
   "roughEstimateOfLCP": 2758,
   "roughEstimateOfSI": 3681,
-  "roughEstimateOfTTFCPUI": 3677,
-  "roughEstimateOfTTI": 4462,
+  "roughEstimateOfTTFCPUI": 3790,
+  "roughEstimateOfTTI": 4671,
 }
 `;
diff --git a/lighthouse-core/test/audits/byte-efficiency/__snapshots__/byte-efficiency-audit-test.js.snap b/lighthouse-core/test/audits/byte-efficiency/__snapshots__/byte-efficiency-audit-test.js.snap
@@ -2,11 +2,11 @@
 
 exports[`Byte efficiency base audit should allow overriding of computeWasteWithTTIGraph 1`] = `
 Object {
-  "default": 1330,
-  "justTTI": 800,
+  "default": 560,
+  "justTTI": 504,
 }
 `;
 
-exports[`Byte efficiency base audit should create load simulator with the specified settings 1`] = `1330`;
+exports[`Byte efficiency base audit should create load simulator with the specified settings 1`] = `960`;
 
-exports[`Byte efficiency base audit should create load simulator with the specified settings 2`] = `22520`;
+exports[`Byte efficiency base audit should create load simulator with the specified settings 2`] = `21500`;
diff --git a/lighthouse-core/test/audits/byte-efficiency/byte-efficiency-audit-test.js b/lighthouse-core/test/audits/byte-efficiency/byte-efficiency-audit-test.js
@@ -242,15 +242,17 @@ describe('Byte efficiency base audit', () => {
     class MockAudit extends ByteEfficiencyAudit {
       static audit_(artifacts, records) {
         return {
-          items: records.map(record => ({url: record.url, wastedBytes: record.transferSize})),
+          items: records.map(record => ({url: record.url, wastedBytes: record.transferSize * 0.5})),
           headings: [],
         };
       }
     }
 
     class MockJustTTIAudit extends MockAudit {
       static computeWasteWithTTIGraph(results, graph, simulator) {
-        return ByteEfficiencyAudit.computeWasteWithTTIGraph(results, graph, simulator,
+        // TODO: Pass in a graph that organically has a lower TTI result rather than forcing it
+        // to be scaled down.
+        return 0.9 * ByteEfficiencyAudit.computeWasteWithTTIGraph(results, graph, simulator,
           {includeLoad: false});
       }
     }

diff --git a/lighthouse-core/test/audits/dobetterweb/uses-http2-test.js b/lighthouse-core/test/audits/dobetterweb/uses-http2-test.js
@@ -44,8 +44,8 @@ describe('Resources are fetched over http/2', () => {
     // make sure we flag all the rest
     expect(result.details.items).toHaveLength(60);
     // make sure we report savings
-    expect(result.numericValue).toMatchInlineSnapshot(`1310`);
-    expect(result.details.overallSavingsMs).toMatchInlineSnapshot(`1310`);
+    expect(result.numericValue).toMatchInlineSnapshot(`1340`);
+    expect(result.details.overallSavingsMs).toMatchInlineSnapshot(`1340`);
     // make sure we have a failing score
     expect(result.score).toBeLessThan(0.5);
   });
@@ -70,7 +70,7 @@ describe('Resources are fetched over http/2', () => {
     expect(urls).not.toContain(records[30].url);
     expect(result.details.items).toHaveLength(30);
     // make sure we report less savings
-    expect(result.numericValue).toMatchInlineSnapshot(`850`);
-    expect(result.details.overallSavingsMs).toMatchInlineSnapshot(`850`);
+    expect(result.numericValue).toMatchInlineSnapshot(`500`);
+    expect(result.details.overallSavingsMs).toMatchInlineSnapshot(`500`);
   });
 });
diff --git a/lighthouse-core/test/computed/metrics/__snapshots__/estimated-input-latency-test.js.snap b/lighthouse-core/test/computed/metrics/__snapshots__/estimated-input-latency-test.js.snap
@@ -2,8 +2,8 @@
 
 exports[`Metrics: EIL should compute a simulated value 1`] = `
 Object {
-  "optimistic": 101,
+  "optimistic": 93,
   "pessimistic": 101,
-  "timing": 81,
+  "timing": 78,
 }
 `;
diff --git a/lighthouse-core/test/computed/metrics/__snapshots__/first-cpu-idle-test.js.snap b/lighthouse-core/test/computed/metrics/__snapshots__/first-cpu-idle-test.js.snap
@@ -2,8 +2,8 @@
 
 exports[`FirstInteractive computed artifact: should simulate when settings specify 1`] = `
 Object {
-  "optimistic": 3351,
-  "pessimistic": 3502,
-  "timing": 3351,
+  "optimistic": 4313,
+  "pessimistic": 4383,
+  "timing": 4313,
 }
 `;
diff --git a/lighthouse-core/test/computed/metrics/__snapshots__/interactive-test.js.snap b/lighthouse-core/test/computed/metrics/__snapshots__/interactive-test.js.snap
@@ -2,8 +2,8 @@
 
 exports[`Metrics: TTI should compute a simulated value 1`] = `
 Object {
-  "optimistic": 3351,
-  "pessimistic": 3502,
-  "timing": 3427,
+  "optimistic": 4313,
+  "pessimistic": 4383,
+  "timing": 4348,
 }
 `;
diff --git a/...use-core/test/computed/metrics/__snapshots__/lantern-estimated-input-latency-test.js.snap b/...use-core/test/computed/metrics/__snapshots__/lantern-estimated-input-latency-test.js.snap
@@ -2,8 +2,8 @@
 
 exports[`Metrics: Lantern EIL should compute a simulated value 1`] = `
 Object {
-  "optimistic": 101,
+  "optimistic": 93,
   "pessimistic": 101,
-  "timing": 81,
+  "timing": 78,
 }
 `;
diff --git a/lighthouse-core/test/computed/metrics/__snapshots__/lantern-first-cpu-idle-test.js.snap b/lighthouse-core/test/computed/metrics/__snapshots__/lantern-first-cpu-idle-test.js.snap
@@ -2,8 +2,8 @@
 
 exports[`Metrics: Lantern TTFCPUI should compute predicted value 1`] = `
 Object {
-  "optimistic": 3351,
-  "pessimistic": 3502,
-  "timing": 3351,
+  "optimistic": 4313,
+  "pessimistic": 4383,
+  "timing": 4313,
 }
 `;
diff --git a/lighthouse-core/test/computed/metrics/__snapshots__/lantern-interactive-test.js.snap b/lighthouse-core/test/computed/metrics/__snapshots__/lantern-interactive-test.js.snap
@@ -2,9 +2,9 @@
 
 exports[`Metrics: Lantern TTI should compute predicted value 1`] = `
 Object {
-  "optimistic": 3351,
-  "pessimistic": 3502,
-  "timing": 3427,
+  "optimistic": 4313,
+  "pessimistic": 4383,
+  "timing": 4348,
 }
 `;
 

diff --git a/lighthouse-core/test/computed/metrics/first-cpu-idle-test.js b/lighthouse-core/test/computed/metrics/first-cpu-idle-test.js
@@ -65,7 +65,7 @@ describe('FirstInteractive computed artifact:', () => {
       optimistic: Math.round(result.optimisticEstimate.timeInMs),
       pessimistic: Math.round(result.pessimisticEstimate.timeInMs),
     }).toMatchSnapshot();
-    assert.equal(result.optimisticEstimate.nodeTimings.size, 19);
+    assert.equal(result.optimisticEstimate.nodeTimings.size, 20);
     assert.equal(result.pessimisticEstimate.nodeTimings.size, 80);
     assert.ok(result.optimisticGraph, 'should have created optimistic graph');
     assert.ok(result.pessimisticGraph, 'should have created pessimistic graph');

diff --git a/lighthouse-core/test/computed/metrics/interactive-test.js b/lighthouse-core/test/computed/metrics/interactive-test.js
@@ -38,7 +38,7 @@ describe('Metrics: TTI', () => {
       optimistic: Math.round(result.optimisticEstimate.timeInMs),
       pessimistic: Math.round(result.pessimisticEstimate.timeInMs),
     }).toMatchSnapshot();
-    assert.equal(result.optimisticEstimate.nodeTimings.size, 19);
+    assert.equal(result.optimisticEstimate.nodeTimings.size, 20);
     assert.equal(result.pessimisticEstimate.nodeTimings.size, 80);
     assert.ok(result.optimisticGraph, 'should have created optimistic graph');
     assert.ok(result.pessimisticGraph, 'should have created pessimistic graph');

diff --git a/lighthouse-core/test/computed/metrics/lantern-first-cpu-idle-test.js b/lighthouse-core/test/computed/metrics/lantern-first-cpu-idle-test.js
@@ -23,7 +23,7 @@ describe('Metrics: Lantern TTFCPUI', () => {
       optimistic: Math.round(result.optimisticEstimate.timeInMs),
       pessimistic: Math.round(result.pessimisticEstimate.timeInMs),
     }).toMatchSnapshot();
-    assert.equal(result.optimisticEstimate.nodeTimings.size, 19);
+    assert.equal(result.optimisticEstimate.nodeTimings.size, 20);
     assert.equal(result.pessimisticEstimate.nodeTimings.size, 80);
     assert.ok(result.optimisticGraph, 'should have created optimistic graph');
     assert.ok(result.pessimisticGraph, 'should have created pessimistic graph');

diff --git a/lighthouse-core/test/computed/metrics/lantern-interactive-test.js b/lighthouse-core/test/computed/metrics/lantern-interactive-test.js
@@ -27,7 +27,7 @@ describe('Metrics: Lantern TTI', () => {
       optimistic: Math.round(result.optimisticEstimate.timeInMs),
       pessimistic: Math.round(result.pessimisticEstimate.timeInMs),
     }).toMatchSnapshot();
-    assert.equal(result.optimisticEstimate.nodeTimings.size, 19);
+    assert.equal(result.optimisticEstimate.nodeTimings.size, 20);
     assert.equal(result.pessimisticEstimate.nodeTimings.size, 80);
     assert.ok(result.optimisticGraph, 'should have created optimistic graph');
     assert.ok(result.pessimisticGraph, 'should have created pessimistic graph');

diff --git a/lighthouse-core/test/computed/metrics/speed-index-test.js b/lighthouse-core/test/computed/metrics/speed-index-test.js
@@ -61,8 +61,8 @@ describe('Metrics: Speed Index', () => {
     }).toMatchInlineSnapshot(`
       Object {
         "optimistic": 575,
-        "pessimistic": 563,
-        "timing": 599,
+        "pessimistic": 633,
+        "timing": 635,
       }
     `);
   });

diff --git a/lighthouse-core/test/fixtures/lantern-master-accuracy.json b/lighthouse-core/test/fixtures/lantern-master-accuracy.json
@@ -1,32 +1,32 @@
 {
   "roughEstimateOfFCP": {
-    "p50": 0.292391744233104,
-    "p90": 0.501397081651661,
-    "p95": 0.5795269168026101
+    "p50": 0.28115015974440893,
+    "p90": 0.49063322495742373,
+    "p95": 0.5191288993525603
   },
   "roughEstimateOfFMP": {
-    "p50": 0.30206175397060275,
+    "p50": 0.301651376146789,
     "p90": 0.5385564466378778,
-    "p95": 0.6641924269472893
+    "p95": 0.625285684604197
   },
   "roughEstimateOfSI": {
-    "p50": 0.274947094253622,
-    "p90": 0.701696317749276,
-    "p95": 1.0080896079651525
+    "p50": 0.260989010989011,
+    "p90": 0.6901117087298304,
+    "p95": 0.9592408214063473
   },
   "roughEstimateOfTTFCPUI": {
-    "p50": 0.28465238303454304,
-    "p90": 0.6509379230322955,
-    "p95": 0.8098063599218334
+    "p50": 0.2543563799887577,
+    "p90": 0.6242117117117117,
+    "p95": 0.8210511483725518
   },
   "roughEstimateOfTTI": {
-    "p50": 0.3055258632709021,
-    "p90": 0.6577081000373274,
+    "p50": 0.24779470729751404,
+    "p90": 0.6419348364220245,
     "p95": 0.743897672766033
   },
   "roughEstimateOfLCP": {
     "p50": 0.203757225433526,
-    "p90": 0.6464365256124721,
-    "p95": 0.8768898488120951
+    "p90": 0.6962974535228871,
+    "p95": 0.9262573279851898
   }
 }