New issue
Advanced search Search tips

Issue 922535 link

Starred by 1 user

Issue metadata

Status: WontFix
Owner:
Closed: Jan 17
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

0.1%-0.1% regression in nacl_helper at 623215:623224

Project Member Reported by 42576172...@developer.gserviceaccount.com, Jan 16 (6 days ago)

Issue description

See the link to graphs below.
 
Project Member

Comment 1 by 42576172...@developer.gserviceaccount.com, Jan 16 (6 days ago)

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=922535

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=36957226ec10939678fb0f862cf04855da252f70d395070172cd20eb40c7151b


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

Google Chrome Linux x64

Comment 2 by sullivan@chromium.org, Jan 16 (6 days ago)

Owner: g...@chromium.org
gbiv: the ADFO roll from 73.0.3672.0_rc-r1 to 73.0.3673.0_rc-r1 is in the range, could that have caused the size increase?

Comment 3 by g...@chromium.org, Jan 16 (6 days ago)

Certainly possible; we apply the AFDO profile to nacl_helper (which I might try to figure out how to avoid, because issue 916130 is a thing...)

Comment 4 by g...@chromium.org, Jan 17 (5 days ago)

Cc: g...@chromium.org
Owner: guidou@chromium.org
Status: WontFix (was: Untriaged)
Appears to be r623218; local builds:

r623222: 388015664      ./NaclHelper/nacl_helper
r623221: 388015664      ./NaclHelper/nacl_helper
r623220: 388015664      ./NaclHelper/nacl_helper
r623219: 388015664      ./NaclHelper/nacl_helper
r623218: 387993368      ./NaclHelper/nacl_helper
r623217: 387507448      ./NaclHelper/nacl_helper

(AFDO was r623219)

Bloaty says:

     VM SIZE                      FILE SIZE
 --------------                --------------
  [ = ]       0 .debug_info     +201Ki  +0.1%
  [ = ]       0 .debug_str      +176Ki  +0.2%
  [ = ]       0 .gdb_index     +84.5Ki  +0.2%
  [ = ]       0 .debug_line    +7.25Ki  +0.1%
  [ = ]       0 .debug_abbrev  +4.30Ki  +0.1%
  [ = ]       0 .strtab            +83  +0.0%
  [ = ]       0 .symtab            +72  +0.0%
  [ = ]       0 [Unmapped]         +10  +0.3%
  [ = ]       0 .debug_macinfo      +3  +0.0%
  [ = ]       0 TOTAL           +474Ki  +0.1%

...Which looks like a pretty similar story to the resolution of  issue 922504 . (read: all debuginfo changes)

Assigning to the author of r623218 if they want to have a look, but as noted in  issue 922504 , debuginfo deltas like this are apparently not very actionable. I also doubt we ship nonstripped nacl_helpers to our users, so I'll throw in a WontFix for free. As before, please anyone feel free to reopen if you disagree. :)

Sign in to add a comment