All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=868992

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=89d2f08229b64be5a8044069bb2c7275a0cebbb21fcc3114bc2888013a7505bb


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

linux-perf

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/16190e4ba40000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/16190e4ba40000

Remove SK_SUPPORT_LEGACY_TEXT_BLOB by fmalita@chromium.org
https://chromium.googlesource.com/chromium/src/+/1f3c291349fb2cc932858a3ea6ff41e691fa05df
2.442e+05 → 3.028e+05 (+5.858e+04)

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

Herb, did we expect this switch to increase memory pressure?

Yes. But probably not this much. How can I learn more about these tests? It
seems like about 100K increase in all of them

+perezjs who owns the benchmark according to the sheet: https://docs.google.com/spreadsheets/d/1xaAo0_SU3iDfGdqDJZX_jRV0QtkufwHUKH3kQKF3YQs/edit#gid=0

More info on the system health tests: https://chromium.googlesource.com/chromium/src/+/master/docs/speed/benchmark/harnesses/system_health.md

Some more info on memory specific side of things:
https://chromium.googlesource.com/chromium/src/+/master/docs/memory-infra/memory_benchmarks.md

Also it's usually a good idea to have a look and compare a couple of traces before/after your change. These are from the perf dashboard on the larger regression:

Before:
https://console.developers.google.com/m/cloudstorage/b/chrome-telemetry-output/o/load_news_cnn_2018-07-27_19-48-34_72073.html

After:
https://console.developers.google.com/m/cloudstorage/b/chrome-telemetry-output/o/load_news_cnn_2018-07-28_00-17-27_99545.html

In this example the growth seems to come from some skia > sk_glyph_cache > CNN_Sans_Display_* fonts.

Deleted: skia.png 167 KB

	skia.png 167 KB View Download

 Issue 871256  has been merged into this issue.

 Issue 871368  has been merged into this issue.

 Issue 871855  has been merged into this issue.

This is working as intended. It is part of a larger change to make the glyph cache multi-threaded.