Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(46)

Issue 2624163003: Remainder of V8 removal (Closed)

Created:
3 years, 11 months ago by scottmg
Modified:
3 years, 11 months ago
CC:
chromium-reviews, cbentzel+watch_chromium.org, qsr+mojo_chromium.org, viettrungluu+watch_chromium.org, jam, yzshen+watch_chromium.org, abarth-chromium, Aaron Boodman, darin-cc_chromium.org, asvitkine+watch_chromium.org, darin (slow to review)
Target Ref:
refs/pending/heads/master
Project:
chromium
Visibility:
Public.

Description

Remainder of V8 removal chrome.dll: before: 48,394,752 after: 34,550,272 == -13,844,480 bytes is_debug = false is_chrome_branded = true is_component_build = false symbol_level = 2 is_official_build = true target_cpu = "x86" BUG=581766

Patch Set 1 #

Patch Set 2 : . #

Patch Set 3 : . #

Unified diffs Side-by-side diffs Delta from patch set Stats (+21 lines, -13 lines) Patch
M chrome/BUILD.gn View 1 2 1 chunk +5 lines, -4 lines 0 comments Download
M chrome/DEPS View 1 chunk +1 line, -2 lines 0 comments Download
M chrome/browser/DEPS View 1 chunk +0 lines, -1 line 0 comments Download
M chrome/browser/about_flags.cc View 1 chunk +2 lines, -0 lines 0 comments Download
M chrome/browser/net/proxy_service_factory.cc View 2 chunks +2 lines, -0 lines 0 comments Download
M chrome/browser/task_manager/providers/browser_process_task.cc View 1 chunk +2 lines, -0 lines 0 comments Download
M chrome/child/DEPS View 1 1 chunk +1 line, -1 line 0 comments Download
M chrome/renderer/DEPS View 1 1 chunk +1 line, -0 lines 0 comments Download
M chrome/utility/DEPS View 1 1 chunk +1 line, -0 lines 0 comments Download
M content/app/content_main_runner.cc View 1 chunk +2 lines, -0 lines 0 comments Download
M content/browser/BUILD.gn View 1 chunk +0 lines, -1 line 0 comments Download
M content/browser/DEPS View 1 chunk +2 lines, -0 lines 0 comments Download
M content/browser/child_process_launcher.cc View 1 chunk +1 line, -1 line 0 comments Download
M net/BUILD.gn View 1 chunk +1 line, -1 line 0 comments Download
M net/proxy/proxy_service_mojo.cc View 1 chunk +0 lines, -1 line 0 comments Download
M ppapi/proxy/BUILD.gn View 1 chunk +0 lines, -1 line 0 comments Download

Depends on Patchset:

Messages

Total messages: 12 (5 generated)
brucedawson
I'm confused by the sizes that you're seeing. On today's x86 canary chrome.dll is 40,665,944 ...
3 years, 11 months ago (2017-01-12 23:28:00 UTC) #5
scottmg
On 2017/01/12 23:28:00, brucedawson wrote: > I'm confused by the sizes that you're seeing. On ...
3 years, 11 months ago (2017-01-12 23:30:55 UTC) #6
scottmg
On 2017/01/12 23:30:55, scottmg wrote: > On 2017/01/12 23:28:00, brucedawson wrote: > > I'm confused ...
3 years, 11 months ago (2017-01-12 23:32:06 UTC) #7
scottmg
On 2017/01/12 23:28:00, brucedawson wrote: > I'm confused by the sizes that you're seeing. On ...
3 years, 11 months ago (2017-01-12 23:50:40 UTC) #9
Sébastien Marchand
I've opened https://bugs.chromium.org/p/chromium/issues/detail?id=680795 to fix this full_wpo_on_official madness, I really want to get rid of ...
3 years, 11 months ago (2017-01-13 00:07:29 UTC) #10
scottmg
On 2017/01/12 23:30:55, scottmg wrote: > On 2017/01/12 23:28:00, brucedawson wrote: > > I'm confused ...
3 years, 11 months ago (2017-01-13 05:24:33 UTC) #11
scottmg
3 years, 11 months ago (2017-01-13 05:25:48 UTC) #12
On 2017/01/13 05:24:33, scottmg wrote:
> On 2017/01/12 23:30:55, scottmg wrote:
> > On 2017/01/12 23:28:00, brucedawson wrote:
> > > I'm confused by the sizes that you're seeing. On today's x86 canary
> chrome.dll
> > > is 40,665,944 bytes. The one setting that you're missing (oh the land
mines
> > that
> > > we leave - this one needs fixing) is full_wpo_on_official, although I have
a
> > > hard time believing that that could make an almost 8 MB difference. Here
are
> > my
> > > settings which give me a canary-comparable binary size:
> > > 
> > > is_chrome_branded = true
> > > is_component_build = false
> > > is_debug = false
> > > target_cpu = "x86"
> > > is_official_build = true
> > > full_wpo_on_official = true
> > 
> > Hmm, that could be it I guess. It's possible it manages to drop much of
mostly
> > unused copy of blink, etc when full_wpo_on_official is on?
> 
> Yeah, it looks like full_wpo_on_official completely changes things.
> 
> chrome.dll official non-full_wpo:
> before: 48,394,752
> after: 34,550,272
> 
> chrome.dll official full_wpo:
> before:  40,519,168
> after: 40,512,000

Er, those numbers are just breaking the Blink dep, not v8, so not completely
comparable. But the two "before"s are roughly comparable.

Powered by Google App Engine
This is Rietveld 408576698