All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=734171

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgtvOtswkM


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

Android Builder

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8976603520385658608

Based on the graph, the regression is entirely in native code.

Output from tools/binary_size/diagnose_bloat.py eb2c92dd6e900e0159c7c23f81749ead03d9faa1 --cloud:

4 symbols added (+), 4 changed (~), 1 removed (-), 486534 unchanged (not shown)
Number of unique symbols 376828 -> 376831 (+3)
1 paths added, 0 removed, 4 changed

Showing 486,543 symbols (aliases not grouped for diffs) with total pss: 48 bytes
.text=48 bytes   .rodata=0 bytes    .data*=0 bytes    .bss=0 bytes    total=48 bytes
Number of unique paths: 15620

Index | Running Total | Section@Address | Δ PSS (Δ size_without_padding) | Path
------------------------------------------------------------
- 0)       -528 (-1100.0%) t@0x0        -528 (524->0)      third_party/WebKit/Source/core/css/properties/CSSShorthandPropertyAPIFlex.cpp
               blink::CSSShorthandPropertyAPIFlex::parseShorthand
+ 1)         -4 (-8.3%) t@0x1cc83b8  +524 (0->524)      third_party/WebKit/Source/core/css/properties/CSSShorthandPropertyAPIFlex.cpp
               blink::CSSShorthandPropertyAPIFlex::parseShorthand
+ 2)        168 (350.0%) t@0x1cc830c  +172 (0->172)      third_party/WebKit/Source/core/css/properties/CSSShorthandPropertyAPIBorderRadius.cpp
               blink::CSSShorthandPropertyAPIBorderRadius::parseShorthand
~ 3)         44 (91.7%) t@0x1cb9a78  -124 (1828->1704)  third_party/WebKit/Source/core/css/parser/CSSPropertyParser.cpp
               blink::CSSPropertyParser::ParseShorthand
+ 4)         52 (108.3%) t@0x0        +8 (0->0.0)        {{no path}}
               ** aggregate padding of diff'ed symbols
+ 5)         44 (91.7%) d@0x0        -8 (0->0.0)        {{no path}}
               ** aggregate padding of diff'ed symbols
~ 6)         36 (75.0%) t@0x1cb8ec8  -8 (2848->2840)    third_party/WebKit/Source/core/css/parser/CSSPropertyParser.cpp
               blink::CSSPropertyParser::ParseSingleValue
~ 7)         44 (91.7%) d@0x2d355c0  +8 (920->928)      blink/core/css/properties/CSSPropertyDescriptor.cpp
               blink::cssPropertyDescriptors
~ 8)         48 (100.0%) t@Group      +4 (0->0)          {{no path}}
               ** symbol gaps (count=9)

Forgot the top of the output:

Section Sizes (Total=24.1kb (24632 bytes)):
    .bss: 0 bytes (0 bytes) (not included in totals)
    .data: 0 bytes (0 bytes) (0.0%)
    .rel.dyn: 24.0kb (24584 bytes) (99.8%)
    .rodata: 0 bytes (0 bytes) (0.0%)
    .text: 48 bytes (48 bytes) (0.2%)

Looks like the increase is entirely from added relocations.

I'd guess we'll just have to write this off as "Compilers, how do they work ¯\_(ツ)_/¯"

+pcc,thakis in case they have any insights, but closing as won't fix as well.

=== BISECT JOB RESULTS ===
NO Perf regression found, tests failed to produce values

Bisect Details
  Configuration: android_nexus7_perf_bisect
  Benchmark    : resource_sizes
  Metric       : MonochromePublic.apk_Specifics/normalized apk size


To Run This Test
  src/build/android/resource_sizes.py --chromium-output-directory {CHROMIUM_OUTPUT_DIR} --chartjson {CHROMIUM_OUTPUT_DIR}/apks/MonochromePublic.apk

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8976603520385658608

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5243533620936704


| O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq
|  X  | for more information addressing perf regression bugs. For feedback,
| / \ | file a bug with component Speed>Bisection.  Thank you!

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8976600432529250000

=== BISECT JOB RESULTS ===
NO Perf regression found, tests failed to produce values

Bisect Details
  Configuration: android_nexus7_perf_bisect
  Benchmark    : resource_sizes
  Metric       : MonochromePublic.apk_Specifics/normalized apk size


To Run This Test
  src/build/android/resource_sizes.py --chromium-output-directory {CHROMIUM_OUTPUT_DIR} --chartjson {CHROMIUM_OUTPUT_DIR}/apks/MonochromePublic.apk

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8976600432529250000

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5243533620936704


| O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq
|  X  | for more information addressing perf regression bugs. For feedback,
| / \ | file a bug with component Speed>Bisection.  Thank you!