All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=881363

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=151ab718268042509ac1a6033bd85248914e496446077e804b073d348767d487


Bot(s) for this bug's original alert(s):

mac-10_12_laptop_low_end-perf

rendering.desktop - Benchmark documentation link:
  https://bit.ly/rendering-benchmarks

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/1189fcd5640000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/1189fcd5640000

turn off use-zoom-for-dsf on Android by eirage@chromium.org
https://chromium.googlesource.com/chromium/src/+/14d5c4330c08ac9bcdcda7960561e3e92c135da6
1.258 → 1.036 (-0.2224)

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions

Benchmark documentation link:
  https://bit.ly/rendering-benchmarks

My change is turning off a flag on Android, should have affect on mac. Un-assign myself

s/should have/should have no/

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/1421958b640000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/1421958b640000

Managing the context was dropped. Adding back in. by herb@google.com
https://skia.googlesource.com/skia/+/aca2b9eebdaf5ec45ba87a25c60510988fc54756
0.8655 → 1.266 (+0.4002)

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions

Benchmark documentation link:
  https://bit.ly/rendering-benchmarks

Hey Khushal,

In some ways it looks like having the scaler context have a longer lifetime was a pretty good performance win for about 5-10% memory use. Is this something we want to explore? 

Here is a corresponding memory increase  bug 883755 .

This is a dupe of issue 879321.