Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(504)

Issue 1493893005: allocator cleanup: remove deprecated windows tc-malloc bits (Closed)

Created:
5 years ago by Primiano Tucci (use gerrit)
Modified:
5 years ago
CC:
chromium-reviews, wfh+watch_chromium.org, Dai Mikurube (NOT FULLTIME), vmpstr+watch_chromium.org, chrisha, ssid, brettw
Base URL:
https://chromium.googlesource.com/chromium/src.git@allocator_gyp
Target Ref:
refs/pending/heads/master
Project:
chromium
Visibility:
Public.

Description

allocator cleanup: remove deprecated windows tc-malloc bits use_allocator=tcmalloc is deprecated and bitrotten on Windows. Cleaning up all the build files that refer to that. I tested the CL on the following configurations and analyzed the produced ninja output: OS=win component=shared_library OS=win GYP_MSVS_VERSION=2015 component=shared_library 0 ninja diff. This is expected, as in component builds allocator was (and is) never pulled in. OS=win Diff: http://pastebin.com/PKaHJ3i2 The diffs here are: 1. The -DPERFTOOLS_DLL_DECL flag disappearing from all the win targets that depend on allocator. This is an expected consequence of having removed the define here from the direct_dependent_settings. 2. dynamic_annotations dep going away. that was just for tc-malloc. OS=win GYP_MSVS_VERSION=2015 Diff: http://pastebin.com/nNDzETTV The diffs here are: 1. The allocator target becomes a no-op. This is intended because, according to common.gypi, the win_use_allocator_shim=0 for MSVS2015. The difference here is that previously the allocator target was built for MSVS2015 (because it was guarded only by !shared_library) but not depended on (because the depending guards looked at win_use_allocator_shim which is more restrictive than !shared_library). 2. Some changes to allocator_unittest, which are irrelevant as allocator_unittest is being moved to base_unittest in crrev.com/1500733002. OS=win clang=1 asan=1 Diff: http://pastebin.com/GkH10947 This is similar to the previous (OS=win GYP_MSVS_VERSION=2015) case. In essence, when asan=1 the shim is force-disabled, so no targets end up depending on allocator.gyp as they check for win_use_allocator_shim. However, before this CL the allocator target was still being built, because we were just looking at the more relaxed !shared_library condition. BUG=564618 Committed: https://crrev.com/33a9814719fb7b0789a11310d96edc6a8fd941a1 Cr-Commit-Position: refs/heads/master@{#363459}

Patch Set 1 : #

Patch Set 2 : rebase #

Total comments: 2

Patch Set 3 : Rebase after brettw changes in crrev.com/1499773002 #

Unified diffs Side-by-side diffs Delta from patch set Stats (+1 line, -51 lines) Patch
M base/allocator/BUILD.gn View 1 2 2 chunks +0 lines, -32 lines 0 comments Download
M base/allocator/allocator.gyp View 1 2 4 chunks +1 line, -19 lines 0 comments Download

Dependent Patchsets:

Messages

Total messages: 30 (19 generated)
Primiano Tucci (use gerrit)
PTAL, small follow up cleanup on windows for tcmalloc. The ninja diff make sense to ...
5 years ago (2015-12-04 10:47:32 UTC) #9
Ruud van Asseldonk
https://codereview.chromium.org/1493893005/diff/80001/base/allocator/BUILD.gn File base/allocator/BUILD.gn (left): https://codereview.chromium.org/1493893005/diff/80001/base/allocator/BUILD.gn#oldcode262 base/allocator/BUILD.gn:262: deps += [ ":prep_libc" ] Why can this be ...
5 years ago (2015-12-04 17:25:37 UTC) #11
Primiano Tucci (use gerrit)
https://codereview.chromium.org/1493893005/diff/80001/base/allocator/BUILD.gn File base/allocator/BUILD.gn (left): https://codereview.chromium.org/1493893005/diff/80001/base/allocator/BUILD.gn#oldcode262 base/allocator/BUILD.gn:262: deps += [ ":prep_libc" ] On 2015/12/04 17:25:37, Ruud ...
5 years ago (2015-12-04 17:40:21 UTC) #12
Primiano Tucci (use gerrit)
> becuse tcmalloc is not supported on android anymore. so this is_win inside And when ...
5 years ago (2015-12-04 17:44:06 UTC) #13
Will Harris
lgtm this change looks benign considering we have no tcmalloc on windows What I'd really ...
5 years ago (2015-12-05 03:36:07 UTC) #15
Nico
lgtm2 fwiw (typo "diffrence" in cl description)
5 years ago (2015-12-05 03:43:03 UTC) #16
commit-bot: I haz the power
CQ is trying da patch. Follow status at https://chromium-cq-status.appspot.com/patch-status/1493893005/100001 View timeline at https://chromium-cq-status.appspot.com/patch-timeline/1493893005/100001
5 years ago (2015-12-07 10:09:07 UTC) #22
commit-bot: I haz the power
Try jobs failed on following builders: linux_chromium_rel_ng on tryserver.chromium.linux (JOB_FAILED, http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_rel_ng/builds/151303)
5 years ago (2015-12-07 11:04:45 UTC) #24
commit-bot: I haz the power
CQ is trying da patch. Follow status at https://chromium-cq-status.appspot.com/patch-status/1493893005/100001 View timeline at https://chromium-cq-status.appspot.com/patch-timeline/1493893005/100001
5 years ago (2015-12-07 11:08:50 UTC) #26
commit-bot: I haz the power
Committed patchset #3 (id:100001)
5 years ago (2015-12-07 12:40:42 UTC) #28
commit-bot: I haz the power
5 years ago (2015-12-07 12:41:25 UTC) #30
Message was sent while issue was closed.
Patchset 3 (id:??) landed as
https://crrev.com/33a9814719fb7b0789a11310d96edc6a8fd941a1
Cr-Commit-Position: refs/heads/master@{#363459}

Powered by Google App Engine
This is Rietveld 408576698