[chromium-bugs] Issue 440500 in chromium: Get Chrome compiling on VS2015

[c***@googlecode.com]

Status: Started
Owner: ----
CC: ***@chromium.org, ***@chromium.org, ***@chromium.org,
***@chromium.org
Labels: Type-Feature Pri-2 OS-Windows

New issue 440500 by ***@chromium.org: Get Chrome compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

- Improved C++11/14 support
- Security improvements:
http://blogs.msdn.com/b/vcblog/archive/2014/12/08/visual-studio-2015-preview-work-in-progress-security-feature.aspx
- Some /analyze checked rolled into main compile
- Possibly some improved support for "cloud" builds.


Initial test on Preview spews a lot of warnings, see attached.

d:\src\cr2\src>set GYP
GYP_DEFINES=branding=Chrome
GYP_MSVS_VERSION=2015

d:\src\cr2\src>ninja -C out\Release chrome -k0 > log_vs2015.txt

Many are "argument hiding class variable" (C4458). There's quite a lot of
them, and they sort of seem like something that could cause a bug, though
with perhaps too high of a false positive.

It also has local shadowing now (C4456) which seems much more useful and we
probably want to keep.

See also https://code.google.com/p/chromium/issues/detail?id=360799 from a
while back.

Attachments:
log_vs2015.txt 2.1 MB
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #1 on issue 440500 by ***@chromium.org: Get Chrome compiling on
VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

(I just realized I built against 8.0 SDK not 8.1 which is probably the
cause of some of the windows headers warnings)
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #3 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

In discussions last month about variable shadowing there was a strong
desire to allow:

Foo::Foo(int x) : x(x) {}

I find that pattern rather obtuse, but its popularity does suggest that
4458 should be disabled at least initially.

I'd actually recommend disabling 4456 as well. The churn on variable
shadowing warnings is significant and resolving them should be treated as a
separate task rather than as part of the 2015 port. Enabling 4456 is also
something that should be done on a per-project basis rather than globally,
enabling 4456 and -WShadow simultaneously. Otherwise the warnings will come
back rather quickly.

BoringSSL just enable -WShadow after /analyze pointed out some confusing
(but not incorrect) code.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #4 on issue 440500 by ***@chromium.org: Get Chrome compiling on
VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Example of 4456, 4457, 4458 as there doesn't appear to be any docs on msdn
yet.


d:\src>type x.cc
class C {
C(int a) : a(a) {}

void F() {
int b;
{
int b;
}
}

void G(int c) {
{
int c;
}
}

void H(int a) {
}

void I() {
int a;
}

int a;
};

d:\src>cl /Bv /W4 /wd4101 /wd4100 x.cc /c
Microsoft (R) C/C++ Optimizing Compiler Version 19.00.22310.1 for x86
Copyright (C) Microsoft Corporation. All rights reserved.

Compiler Passes:
C:\Program Files (x86)\Microsoft Visual Studio
14.0\VC\BIN\amd64_x86\cl.exe: Version 19.00.22310.1
C:\Program Files (x86)\Microsoft Visual Studio
14.0\VC\BIN\amd64_x86\c1.dll: Version 19.00.22310.1
C:\Program Files (x86)\Microsoft Visual Studio
14.0\VC\BIN\amd64_x86\c1xx.dll: Version 19.00.22310.1
C:\Program Files (x86)\Microsoft Visual Studio
14.0\VC\BIN\amd64_x86\c2.dll: Version 19.00.22310.1
C:\Program Files (x86)\Microsoft Visual Studio
14.0\VC\BIN\amd64_x86\link.exe: Version 14.00.22310.1
C:\Program Files (x86)\Microsoft Visual Studio
14.0\VC\BIN\amd64\mspdb140.dll: Version 14.00.22310.1
C:\Program Files (x86)\Microsoft Visual Studio
14.0\VC\BIN\amd64_x86\1033\clui.dll: Version 19.00.22310.1

x.cc
x.cc(7): warning C4456: declaration of 'b' hides previous local declaration
x.cc(5): note: see declaration of 'b'
x.cc(13): warning C4457: declaration of 'c' hides function parameter
x.cc(11): note: see declaration of 'c'
x.cc(21): warning C4458: declaration of 'a' hides class member
x.cc(24): note: see declaration of 'C::a'
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #5 on issue 440500 by ***@chromium.org: Get Chrome compiling on
VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Bruce: they won't come back because we use /WX everywhere. :)

4458 doesn't seem to warn on the constructor, rather it's like function 'I'
above. I'm not sure if it's useful or not, I guess we'll have to see how
frequently it occurs and if it finds any real bugs.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Updates:
Cc: ***@chromium.org

Comment #7 on issue 440500 by ***@chromium.org: Get Chrome compiling on
VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Some notable changes from 2013->2015 to work through:

- hash_map spews a deprecation warning to use unordered_map instead.
- there's actually 4 separate shadowing warnings (one for globals too in
addition to the 3 above). We trip a lot of these.
- timespec is now in time.h which breaks on various compat shims in
third_party
- it warns on
int bitfield:1;
...
bitfield = true;
because there's overflow, the valid range is [-1..0].
- it warns on dbghelp.h in 8.1 because there's useless typedefs like this
typedef enum {
hdBase = 0, // root directory for dbghelp
hdSym, // where symbols are stored
hdSrc, // where source is stored
hdMax // end marker
};
- there's a bunch problems with redefs of INT8_MIN, et al. We've had this
problem before but I don't immediately see why it's different than previous
toolchains.
- there's a real snprintf (!= _snprintf) which again messes with a lot of
third_party compat shims. They even went as far as adding
#ifdef snprintf
#error Macro definition of snprintf conflicts with Standard Library
function declaration
#endif
to stdio.h because it's been not-available for so long I guess. As Bruce
pointed out to me last week, these
https://code.google.com/p/chromium/codesearch#search/&q=%22%23define%20snprintf%20_snprintf%22&sq=package:chromium&type=cs
are probably all somewhere between incorrect and security holes.
- time_t is quite unloved: (that's #error, not warning or deprecated)
c:\program files (x86)\microsoft visual studio
14.0\vc\include\yvals.h(40): fatal error C1189: #error: 32-bit time_t is
strongly deprecated and can trigger crashes in the C++ Standard Library.
- There's a bunch of complaints about "narrowing conversions", e.g.
d:\src\cr3\src\printing\emf_win.cc(548): warning C4838: conversion
from 'int' to 'FLOAT' requires a narrowing conversion

That's not all of them, but it's a decent sample. The summary of which is:
it'll be a while before we know whether we can complete a with-CFG build,
and what the impact on the binary size, performance, etc. is.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #8 on issue 440500 by ***@chromium.org: Get Chrome compiling on
VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Re time_t: Is there some define you can set to get a 64-bit time_t?
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #9 on issue 440500 by ***@chromium.org: Get Chrome compiling on
VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Ah. It would seem it defaults the other way, but libjingle haxed it with
this gem:
https://code.google.com/p/chromium/codesearch#chromium/src/third_party/libjingle/libjingle.gyp&l=217
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #10 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

I was wondering where you got the warning about the one-bit bitfield. I
investigated because I wanted to know why /analyze didn't find the bug, and
that's because nobody ever compares allocated_ against 1, so in fact the
code behaves fine. But, confusing. The simplest fix is to change the type
to unsigned, along with memberNameIsStatic_, so that 1 can be stored and
retrieved.

It was good of them to break the variable shadowing warnings up into four
categories. That's quite helpful.

The narrowing conversion warnings are new in C++11 and are a bit of a pain.
g++ gives a warning when you assign a float/double to an int, regardless of
whether it will fit. Okay, I can deal with that.

Going the other way, g++ gives no warnings for assigning an int to a float,
even if it won't convert without truncation. clang++ handles this
perfectly, only giving warnings if the conversion is inexact. In short, g++
and clang++, compiling with "-c -std=c++11 -g main.cpp", compile this code
with no warnings:

struct Foo
{
float f;
int i;
};

Foo g_foo = {0, 0};

With a larger constant assigned to 'f' clang++ says:
main.cpp:9:14: error: constant expression evaluates to 154153153 which
cannot be narrowed to type 'float' [-Wc++11-narrowing]

If VC++ gives a warning or error on compiling the code above then we should
file a bug.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #15 on issue 440500 by ***@google.com: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

I have an XP VM so I could test whether VS 2015 binaries run there, but I
haven't installed VS 2015 -- it makes me slightly nervous. If Scott wants
to share out binaries (simple project that statically links with CRT) I can
test.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #17 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Hi,

Just FYI: I've tried to make VS2015 preview work on goma server, and
currently it's working.

# I've spent one week to run VS2014-CTP work on goma server, and its patch
still looks effective :-)
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #18 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Should I push VS2015-preview to goma servers? If you want, I'll do it.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #20 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

==========================

d:\src\x>"c:\Program Files (x86)\Microsoft Visual Studio 14.0"\vc\vcvarsall
amd64_x86

d:\src\x>type x.cc
#include <stdio.h>
#include <windows.h>
int CALLBACK WinMain(HINSTANCE hInstance, HINSTANCE hPrevInstance,
LPSTR lpCmdLine, int nCmdShow) {
printf("wee!\n");
MessageBox(0, "hi vs2015", NULL, 0);
}


d:\src\x>cl /MT /Zi /nologo x.cc /link /out:x.exe user32.lib
/subsystem:windows,5.01
x.cc

d:\src\x>dumpbin /headers x.exe | grep version
14.00 linker version
5.01 operating system version
0.00 image version
5.01 subsystem version
0 Win32 version

==========================

The binary ran OK on an XP SP3 VM. So I guess at least the basics of the
CRT still work on XP.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #21 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Maybe try some STL code to see how that behaves on XP -- create a few
containers, insert, iterate, spindle, and mutilate.

You should also test this:

double pi_d = 3.14159265358979323846;
printf(“pi = %.33f\n + %.33f\n”, pi_d, sin(pi_d));

just because I like the fact that VS 2015 handles %.33f so much better.

Also, I assume you've seen these?
http://blogs.msdn.com/b/vcblog/archive/2014/06/10/the-great-crt-refactoring.aspx
http://blogs.msdn.com/b/vcblog/archive/2014/06/18/crt-features-fixes-and-breaking-changes-in-visual-studio-14-ctp1.aspx

The change in %S behavior is something we'll have to watch for.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #26 on issue 440500 by ***@chromium.org: Get Chrome compiling on
VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Are those asserts only on xp?
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #27 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Yes, out of my sample set of { xp-sp3-vm, 81-dev-machine }.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #28 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

I'm fairly certain the concurrency related stuff depends on ConcRT in VS
2013 too.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #31 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

I'm not entirely certain.
http://msdn.microsoft.com/en-us/library/jj851139.aspx appears to claim
support, but it also mentions all of VS2013, VS2012, and VS2010 as if they
were current versions. So I would definitely want to do a round of testing
before taking advantage of any library features.

Also, from MS DevRel, the latest word we have is "The current plan is to be
able to target XP as a runtime platform using the VC++ 2015 bits." though
the also say that they require XP SP3, and we say we support SP2
https://support.google.com/chrome/answer/95346?hl=en .
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #32 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

At least some of the <atomic> features work fine on Windows XP. For
instance, it is quite clear that atomic acquire/release fences will work
just fine on XP because they have no run-time support requirements -- they
inject no code at all. I wouldn't be surprised if many other atomic
features depend either on nothing or just on core OS features.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #33 on issue 440500 by ***@chromium.org: Get Chrome compiling on
VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Some headers are documented as using ConcRT. For example, <mutex>:

http://msdn.microsoft.com/en-us/library/hh921467.aspx

ConcRT is allegedly supported on Windows XP SP3 and later.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #34 on issue 440500 by ***@chromium.org: Get Chrome compiling on
VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Do we know how many users we have on XP SP2 vs XP SP3?
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #36 on issue 440500 by ***@hotmail.com: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

And I don't think we are going to move immediately to VS2015, right?
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #37 on issue 440500 by ***@hotmail.com: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

As a side note, the usual marketshare sources claims that total XP
marketshare is finally falling. Does the Chrome stats confirm this?
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #39 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c39

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/6c2204658de1f7ed8350c43d522c592272aaa821

commit 6c2204658de1f7ed8350c43d522c592272aaa821
Author: scottmg <***@chromium.org>
Date: Tue Jan 27 20:20:40 2015

Fix variable shadowing in TLS code on VS2015

d:\src\cr3\src\base\threading\thread_local_storage.cc(143): error C2220:
warning treated as error - no 'object' file generated
d:\src\cr3\src\base\threading\thread_local_storage.cc(143): warning C4457:
declaration of 'value' hides function parameter
d:\src\cr3\src\base\threading\thread_local_storage.cc(110): note: see
declaration of 'value'

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/880023002

Cr-Commit-Position: refs/heads/master@{#313342}

[modify]
http://crrev.com/6c2204658de1f7ed8350c43d522c592272aaa821/base/threading/thread_local_storage.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #40 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c40

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/3c598102874c539a4b5e7f9bc51f624f245827dc

commit 3c598102874c539a4b5e7f9bc51f624f245827dc
Author: scottmg <***@chromium.org>
Date: Tue Jan 27 20:21:31 2015

Fix variable shadowing warning in safe_sprintf in VS2015

d:\src\cr3\src\base\strings\safe_sprintf.cc(513): error C2220: warning
treated as error - no 'object' file generated
d:\src\cr3\src\base\strings\safe_sprintf.cc(513): warning C4456:
declaration of 'ch' hides previous local declaration
d:\src\cr3\src\base\strings\safe_sprintf.cc(450): note: see declaration
of 'ch'

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/868023005

Cr-Commit-Position: refs/heads/master@{#313343}

[modify]
http://crrev.com/3c598102874c539a4b5e7f9bc51f624f245827dc/base/strings/safe_sprintf.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #41 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Unfortunately new instances of variable shadowing keep appearing. According
to the /analyze builder, these are the new warnings from the 22nd to 26th:

storage\browser\blob\blob_storage_context.cc(190) : warning C6246: Local
declaration of 'entry' hides declaration of the same name in outer scope.
third_party\webkit\source\core\html\formdatalist.cpp(106) : warning C6246:
Local declaration of 'key' hides declaration of the same name in outer
scope.
third_party\webkit\source\core\html\formdatalist.cpp(151) : warning C6246:
Local declaration of 'key' hides declaration of the same name in outer
scope.
third_party\webkit\source\core\html\formdatalist.cpp(76) : warning C6246:
Local declaration of 'key' hides declaration of the same name in outer
scope.
v8\src\compiler\ast-graph-builder.cc(1073) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\ast-graph-builder.cc(1085) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\ast-graph-builder.cc(1113) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\ast-graph-builder.cc(1148) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\ast-graph-builder.cc(1155) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\ast-graph-builder.cc(1162) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\ast-graph-builder.cc(1169) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\ast-graph-builder.cc(1180) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\ast-graph-builder.cc(1996) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\ast-graph-builder.cc(2184) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\ast-graph-builder.cc(2228) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\ast-graph-builder.cc(2242) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\ast-graph-builder.cc(2250) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\ast-graph-builder.cc(2259) : warning C6246: Local
declaration of 'op' hides declaration of the same name in outer scope.
v8\src\compiler\register-allocator-verifier.cc(566) : warning C6246: Local
declaration of 'incoming' hides declaration of the same name in outer scope.

The first one looked dodgy enough that I commented on the code review. The
others I am ignoring.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #42 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Loading Image...

Yeah, at some point we might want to just suppress the warning(s). The ones
I looked at so far seemed unnecessarily confusing so I think it's worth
just fixing them. I just did 'base'; the churn there should be fairly low,
I think.

[ And... I have a cold, so this matches my brain's skillz at the moment. :)
]
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #46 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c46

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/fc5b7077362f170c9205dc976dbd144c6f6bd017

commit fc5b7077362f170c9205dc976dbd144c6f6bd017
Author: scottmg <***@chromium.org>
Date: Tue Jan 27 21:46:28 2015

Fix variable shadowing warning in logging on VS2015

d:\src\cr3\src\base\logging.cc(167): error C2220: warning treated as error
- no 'object' file generated
d:\src\cr3\src\base\logging.cc(167): warning C4459: declaration
of 'log_file' hides global declaration
d:\src\cr3\src\base\logging.cc(103): note: see declaration
of 'logging::`anonymous-namespace'::log_file'

There was 3 different things called log_file in this file: a global
FileHandle,
a local PathString, and a PathChar* member variable in LoggingSettings. :/

(Should probably rename all the globals to g_ prefixed in a followup.)

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/883853003

Cr-Commit-Position: refs/heads/master@{#313364}

[modify]
http://crrev.com/fc5b7077362f170c9205dc976dbd144c6f6bd017/base/logging.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #51 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c51

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/3a1c346e8d17824ec8732333ac167206c5dad322

commit 3a1c346e8d17824ec8732333ac167206c5dad322
Author: scottmg <***@chromium.org>
Date: Tue Jan 27 22:57:30 2015

Fix variable shadowing warnings in UDPSocketWin on VS2015

d:\src\cr3\src\net\udp\udp_socket_win.cc(330): error C2220: warning treated
as error - no 'object' file generated
d:\src\cr3\src\net\udp\udp_socket_win.cc(330): warning C4457: declaration
of 'address' hides function parameter
d:\src\cr3\src\net\udp\udp_socket_win.cc(319): note: see declaration
of 'address'
d:\src\cr3\src\net\udp\udp_socket_win.cc(351): warning C4457: declaration
of 'address' hides function parameter
d:\src\cr3\src\net\udp\udp_socket_win.cc(340): note: see declaration
of 'address'

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/882733003

Cr-Commit-Position: refs/heads/master@{#313390}

[modify]
http://crrev.com/3a1c346e8d17824ec8732333ac167206c5dad322/net/udp/udp_socket_win.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #53 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c53

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/61da3cc6e760e2b93126df78081838cd74a95fb8

commit 61da3cc6e760e2b93126df78081838cd74a95fb8
Author: scottmg <***@chromium.org>
Date: Wed Jan 28 00:02:37 2015

Fix variable shadowing warning in CRLSet on VS2015

d:\src\cr3\src\net\cert\crl_set.cc(53): error C2220: warning treated as
error - no 'object' file generated
d:\src\cr3\src\net\cert\crl_set.cc(53): warning C4456: declaration of 'i'
hides previous local declaration
d:\src\cr3\src\net\cert\crl_set.cc(47): note: see declaration of 'i'

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/882833002

Cr-Commit-Position: refs/heads/master@{#313410}

[modify]
http://crrev.com/61da3cc6e760e2b93126df78081838cd74a95fb8/net/cert/crl_set.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #54 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c54

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/d722c2ce033811a1dcb33d7c2fb7d1cfc7041e78

commit d722c2ce033811a1dcb33d7c2fb7d1cfc7041e78
Author: cpu <***@chromium.org>
Date: Wed Jan 28 04:13:31 2015

fix autofill shadowing vars warning

found by the vc /analyze bot

BUG=440500

Review URL: https://codereview.chromium.org/876343004

Cr-Commit-Position: refs/heads/master@{#313433}

[modify]
http://crrev.com/d722c2ce033811a1dcb33d7c2fb7d1cfc7041e78/chrome/browser/ui/autofill/autofill_dialog_controller_unittest.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #55 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c55

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/85fd8b5480d2eb4a2f85f8bc6365969a6de259c3

commit 85fd8b5480d2eb4a2f85f8bc6365969a6de259c3
Author: scottmg <***@chromium.org>
Date: Wed Jan 28 18:25:41 2015

Fix truncation warning in PEImage on VS2015

d:\src\cr3\src\base\win\pe_image.h(250): error C2220: warning treated as
error - no 'object' file generated
d:\src\cr3\src\base\win\pe_image.h(250): warning C4302: 'reinterpret_cast':
truncation from 'LPCSTR' to 'WORD'

Separate ptr->int from truncation to WORD.

Also, while we're here, fix a cast in IsOrdinal.

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/883873002

Cr-Commit-Position: refs/heads/master@{#313543}

[modify]
http://crrev.com/85fd8b5480d2eb4a2f85f8bc6365969a6de259c3/base/win/pe_image.h
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Issue 440500: Get Chrome compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

This issue is now blocking issue chromium:358267.
See https://code.google.com/p/chromium/issues/detail?id=358267

--
You received this message because you are listed in the owner
or CC fields of this issue, or because you starred this issue.
You may adjust your issue notification preferences at:
http://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #57 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c57

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/356a110050a17a19bf9168fadc0a1acb828c582b

commit 356a110050a17a19bf9168fadc0a1acb828c582b
Author: scottmg <***@chromium.org>
Date: Fri Feb 20 19:29:47 2015

win vs2015: fix variable shadowing warnings in net/quic

d:\src\cr2\src\net\quic\crypto\cert_compressor.cc(198): error C2220:
warning treated as error - no 'object' file generated
d:\src\cr2\src\net\quic\crypto\cert_compressor.cc(198): warning C4456:
declaration of 'i' hides previous local declaration
d:\src\cr2\src\net\quic\crypto\cert_compressor.cc(189): note: see
declaration of 'i'

d:\src\cr2\src\net\quic\crypto\quic_crypto_server_config.cc(668): error
C2220: warning treated as error - no 'object' file generated
d:\src\cr2\src\net\quic\crypto\quic_crypto_server_config.cc(668): warning
C4456: declaration of 'client_hello_serialized' hides previous local
declaration
d:\src\cr2\src\net\quic\crypto\quic_crypto_server_config.cc(652): note: see
declaration of 'client_hello_serialized'
d:\src\cr2\src\net\quic\crypto\quic_crypto_server_config.cc(1336): warning
C4456: declaration of 'i' hides previous local declaration
d:\src\cr2\src\net\quic\crypto\quic_crypto_server_config.cc(1292): note:
see declaration of 'i'
d:\src\cr2\src\net\quic\crypto\quic_crypto_server_config.cc(1481): warning
C4457: declaration of 'token' hides function parameter
d:\src\cr2\src\net\quic\crypto\quic_crypto_server_config.cc(1472): note:
see declaration of 'token'
d:\src\cr2\src\net\quic\crypto\quic_crypto_server_config.cc(1493): warning
C4457: declaration of 'token' hides function parameter
d:\src\cr2\src\net\quic\crypto\quic_crypto_server_config.cc(1472): note:
see declaration of 'token'

d:\src\cr2\src\net\quic\crypto\strike_register.cc(498): error C2220:
warning treated as error - no 'object' file generated
d:\src\cr2\src\net\quic\crypto\strike_register.cc(498): warning C4456:
declaration of 'i' hides previous local declaration
d:\src\cr2\src\net\quic\crypto\strike_register.cc(453): note: see
declaration of 'i'
c:\program files (x86)\microsoft visual studio
14.0\vc\include\utility(151): note: while compiling class template member
function 'std::pair<uint32,uint32>::pair<uint32,uint32,void>(std::pair<uint32,uint32>
&&) noexcept'
d:\src\cr2\src\net\quic\crypto\strike_register.cc(502): warning C4456:
declaration of 'bit' hides previous local declaration
d:\src\cr2\src\net\quic\crypto\strike_register.cc(454): note: see
declaration of 'bit'
ninja: build stopped: subcommand failed.

d:\src\cr2\src\net\quic\quic_framer.cc(1350): error C2220: warning treated
as error - no 'object' file generated
d:\src\cr2\src\net\quic\quic_framer.cc(1350): warning C4456: declaration
of 'i' hides previous local declaration
d:\src\cr2\src\net\quic\quic_framer.cc(1338): note: see declaration of 'i'
d:\src\cr2\src\net\quic\quic_framer.cc(2102): warning C4456: declaration
of 'time_delta_us' hides previous local declaration
d:\src\cr2\src\net\quic\quic_framer.cc(2080): note: see declaration
of 'time_delta_us'
ninja: build stopped: subcommand failed.

BUG=440500
R=***@chromium.org

Review URL: https://codereview.chromium.org/941933002

Cr-Commit-Position: refs/heads/master@{#317370}

[modify]
http://crrev.com/356a110050a17a19bf9168fadc0a1acb828c582b/net/quic/crypto/cert_compressor.cc
[modify]
http://crrev.com/356a110050a17a19bf9168fadc0a1acb828c582b/net/quic/crypto/quic_crypto_server_config.cc
[modify]
http://crrev.com/356a110050a17a19bf9168fadc0a1acb828c582b/net/quic/crypto/strike_register.cc
[modify]
http://crrev.com/356a110050a17a19bf9168fadc0a1acb828c582b/net/quic/quic_framer.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #59 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c59

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/cb6a123e432e617052721ef43823fd60b002284e

commit cb6a123e432e617052721ef43823fd60b002284e
Author: scottmg <***@chromium.org>
Date: Fri Feb 20 20:59:34 2015

win vs2015: avoid some variable shadowing warnings in crypto/p224

e.g.

d:\src\cr2\src\crypto\p224.cc(478): error C2220: warning treated as error -
no 'object' file generated
d:\src\cr2\src\crypto\p224.cc(478): warning C4456: declaration of 'j' hides
previous local declaration
d:\src\cr2\src\crypto\p224.cc(447): note: see declaration of 'j'
d:\src\cr2\src\crypto\p224.cc(498): warning C4456: declaration of 'i' hides
previous local declaration
d:\src\cr2\src\crypto\p224.cc(447): note: see declaration of 'i'
d:\src\cr2\src\crypto\p224.cc(516): warning C4456: declaration of 'i' hides
previous local declaration
d:\src\cr2\src\crypto\p224.cc(447): note: see declaration of 'i'
d:\src\cr2\src\crypto\p224.cc(526): warning C4456: declaration of 'i' hides
previous local declaration
d:\src\cr2\src\crypto\p224.cc(447): note: see declaration of 'i'
d:\src\cr2\src\crypto\p224.cc(694): warning C4458: declaration of 'x' hides
class member
d:\src\cr2\src\crypto\p224.h(35): note: see declaration
of 'crypto::p224::Point::x'
d:\src\cr2\src\crypto\p224.cc(694): warning C4458: declaration of 'y' hides
class member
d:\src\cr2\src\crypto\p224.h(35): note: see declaration
of 'crypto::p224::Point::y'

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/945633003

Cr-Commit-Position: refs/heads/master@{#317398}

[modify]
http://crrev.com/cb6a123e432e617052721ef43823fd60b002284e/crypto/p224.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #60 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c60

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/345451878a79d06faa191176bffa7bd65a70f7dc

commit 345451878a79d06faa191176bffa7bd65a70f7dc
Author: scottmg <***@chromium.org>
Date: Fri Feb 20 22:37:37 2015

win 2015: followup to https://codereview.chromium.org/941933002

Previous CL was (potentially?) incorrect here.

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/946003003

Cr-Commit-Position: refs/heads/master@{#317429}

[modify]
http://crrev.com/345451878a79d06faa191176bffa7bd65a70f7dc/net/quic/crypto/strike_register.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #61 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c61

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/df75e1de1238c157580ec10d6f9478d7222b4d69

commit df75e1de1238c157580ec10d6f9478d7222b4d69
Author: scottmg <***@chromium.org>
Date: Fri Feb 20 22:42:35 2015

win vs2015: avoid variable shadowing warning in url_canon.h

d:\src\cr2\src\url\url_canon.h(737): error C2220: warning treated as error
- no 'object' file generated
d:\src\cr2\src\url\url_canon.h(737): warning C4459: declaration
of 'empty_string' hides global declaration
d:\src\cr2\src\url\gurl.cc(22): note: see declaration
of '`anonymous-namespace'::empty_string'

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/937473003

Cr-Commit-Position: refs/heads/master@{#317431}

[modify]
http://crrev.com/df75e1de1238c157580ec10d6f9478d7222b4d69/url/url_canon.h
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #63 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c63

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/31f54fb62bd9696392f1a767014a8abf5e33edf1

commit 31f54fb62bd9696392f1a767014a8abf5e33edf1
Author: scottmg <***@chromium.org>
Date: Thu Mar 19 01:17:02 2015

win vs2015: disable some warnings in non-chromium code, change magic
maxilksize

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/943893003

Cr-Commit-Position: refs/heads/master@{#321261}

[modify]
http://crrev.com/31f54fb62bd9696392f1a767014a8abf5e33edf1/build/common.gypi
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #71 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c71

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/2622e27d880e88b21d8eceb98b6863cdf6dc9afd

commit 2622e27d880e88b21d8eceb98b6863cdf6dc9afd
Author: brucedawson <***@chromium.org>
Date: Wed Apr 08 00:13:54 2015

Don't #define snprintf in VS 2015 builds.

The function snprintf is not available in Visual Studio prior to VS
2015. In some places this omission is dealt with with a #define to map
snprintf to _snprintf. This was never a good idea (_snprintf doesn't
guarantee null-termination) and with VS 2015 it is not even legal.

This change conditionally disables one instance of this remapping.

The build errors that occur without this change are:

stdio.h(1926): warning C4005: 'snprintf': macro redefinition
stdio.h(1926): note: command-line arguments: see previous
definition of 'snprintf'
stdio.h(1929): fatal error C1189: #error: Macro definition of
snprintf conflicts with Standard Library function declaration

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1069433002

Cr-Commit-Position: refs/heads/master@{#324157}

[modify]
http://crrev.com/2622e27d880e88b21d8eceb98b6863cdf6dc9afd/content/content_shell.gypi
[modify]
http://crrev.com/2622e27d880e88b21d8eceb98b6863cdf6dc9afd/third_party/libexif/libexif.gyp
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #73 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c73

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/51428030f98124b4ec5d7c16d9b99e13970c2621

commit 51428030f98124b4ec5d7c16d9b99e13970c2621
Author: brucedawson <***@chromium.org>
Date: Thu Apr 09 02:12:41 2015

Fix illegal use of const in container contents to unbreak VS 2015.

For details see the STL Breaking Changes (VS14 CTP1) section of:
http://blogs.msdn.com/b/vcblog/archive/2014/06/06/
c-14-stl-features-fixes-and-breaking-changes-in-
visual-studio-14-ctp1.aspx

In particular:
"The Standard has always forbidden containers of const elements...
...elements must be Assignable, which const T isn't."

Error message is:
error C2338: The C++ Standard forbids containers of const
elements because allocator<const T> is ill-formed.

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1066213002

Cr-Commit-Position: refs/heads/master@{#324339}

[modify]
http://crrev.com/51428030f98124b4ec5d7c16d9b99e13970c2621/sandbox/win/src/handle_closer.h
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #74 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Update: The pdfium standalone project builds, and runs pdfium_test.exe.
This requires applying https://codereview.chromium.org/1076933002 (some
temporary build settings changes for VS 2015) and
https://codereview.chromium.org/1059023003/ (get pdfium to stop using a
5-month old version of v8).

Pdfium has 385 unique warnings of 6 different types. All but six of those
warnings (one type) are variable shadowing. The six warnings are about
narrowing conversions.

https://codereview.chromium.org/1075673002 and
https://codereview.chromium.org/1078513002 are landed (pdfium doesn't
update bugs when changes are landed).
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #78 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c78

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/d636ce1d4130a6887d2fdc0aaea889cb5c80924b

commit d636ce1d4130a6887d2fdc0aaea889cb5c80924b
Author: scottmg <***@chromium.org>
Date: Tue Apr 14 00:38:08 2015

win: fix narrowing warning in native_theme_win

d:\src\cr3\src\ui\native_theme\native_theme_win.cc(1517): warning C4838:
conversion from 'int' to 'DWORD' requires a narrowing conversion

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1082093002

Cr-Commit-Position: refs/heads/master@{#324960}

[modify]
http://crrev.com/d636ce1d4130a6887d2fdc0aaea889cb5c80924b/ui/native_theme/native_theme_win.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #79 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c79

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/54732d89d4fc5bf21936d029d9505a3795714026

commit 54732d89d4fc5bf21936d029d9505a3795714026
Author: scottmg <***@chromium.org>
Date: Tue Apr 14 00:55:07 2015

win: fix narrowing warning on vs2015

d:\src\cr3\src\ui\platform_window\win\win_window.cc(119): warning C4838:
conversion from 'LONG' to 'DWORD' requires a narrowing conversion

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1083833002

Cr-Commit-Position: refs/heads/master@{#324970}

[modify]
http://crrev.com/54732d89d4fc5bf21936d029d9505a3795714026/ui/platform_window/win/win_window.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #80 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c80

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/cbe384e1542d4f7bd6482caca6aa4122ff240e3e

commit cbe384e1542d4f7bd6482caca6aa4122ff240e3e
Author: scottmg <***@chromium.org>
Date: Tue Apr 14 01:27:05 2015

Cherry-pick some protobuf changes to get building on VS2015

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1064723003

Cr-Commit-Position: refs/heads/master@{#324977}

[modify]
http://crrev.com/cbe384e1542d4f7bd6482caca6aa4122ff240e3e/third_party/protobuf/README.chromium
[modify]
http://crrev.com/cbe384e1542d4f7bd6482caca6aa4122ff240e3e/third_party/protobuf/src/google/protobuf/stubs/hash.h
[add]
http://crrev.com/cbe384e1542d4f7bd6482caca6aa4122ff240e3e/third_party/protobuf/src/google/protobuf/stubs/pbconfig.h
[modify]
http://crrev.com/cbe384e1542d4f7bd6482caca6aa4122ff240e3e/third_party/protobuf/vsprojects/config.h
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #86 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c86

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/548cc22b487981ecd15a2e0b869a15309446e6f4

commit 548cc22b487981ecd15a2e0b869a15309446e6f4
Author: brucedawson <***@chromium.org>
Date: Mon Apr 20 12:09:19 2015

Fix errors caused by illegal #define of snprintf.

Because VS has, for a long time, not had snprintf some developers have
fixed this with #define snprintf _snprintf. This is scary because the
two functions behave differently, and it also breaks with VS 2015 which
does implement snprintf. So, the #defines all need a version check.

The errors are:

stdio.h(1926): warning C4005: 'snprintf': macro redefinition
libxml\win32\config.h(98): note: see previous definition of 'snprintf'
libxml\src\include\win32config.h(98): note: see previous def of 'snprintf'
stdio.h(1929): fatal error C1189: #error: Macro definition of snprintf
conflicts with Standard Library function declaration

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1072653003

Cr-Commit-Position: refs/heads/master@{#325832}

[modify]
http://crrev.com/548cc22b487981ecd15a2e0b869a15309446e6f4/third_party/libxml/README.chromium
[add]
http://crrev.com/548cc22b487981ecd15a2e0b869a15309446e6f4/third_party/libxml/patches/snprintf_config
[add]
http://crrev.com/548cc22b487981ecd15a2e0b869a15309446e6f4/third_party/libxml/patches/snprintf_win32config
[modify]
http://crrev.com/548cc22b487981ecd15a2e0b869a15309446e6f4/third_party/libxml/src/include/win32config.h
[modify]
http://crrev.com/548cc22b487981ecd15a2e0b869a15309446e6f4/third_party/libxml/win32/config.h
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #87 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c87

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/76ef7c9b28521b1f5c27d2677196028f01d0a576

commit 76ef7c9b28521b1f5c27d2677196028f01d0a576
Author: thestig <***@chromium.org>
Date: Mon Apr 20 22:27:41 2015

Roll DEPS for PDFium to 19ae17578f99621100a26dac3e2c7c3dbf7c7cd1

19ae175 Remove release method from CBA_AnnotItetator.
bc580c6 Remove Release() from CPDF_PageObject
0fff5ef Remove unused FX_NEW definition.
4cb07c9 Fix all remaining instances of FX_NEW.
4860684 Replace FX_NEW with new, remove tests from fpdfsdk
9162283 Set m_FontType in CPDF_Font() constructor.
29bc87d Add nonstd::unique_ptr as a workaround until std::unique_ptr is
allowed.
ee2fe42 Make CFX_StockFontArray more robust.
5c357a5 Replace FX_NEW with new, remove tests from fpdfapi
7f6b667 Replace FX_NEW with new, remove tests from fpdftext
cfc1a65 Replace FX_NEW with new, remote tests from fxcodec
357d16f Replace FX_NEW with new, remove tests in fpdfdoc.
269aaa2 Supress the two noisiest warnings on linux standalone builds.
0e2e6ce Fix pdfium standalone to compile with VS 2015.
3f754d4 Remove checks in fxge/{apple,win32,skia,dib} now that FX_NEW cant
return 0
0266232 Remove checks in fxge/ge now that FX_NEW can't return 0.
6c0d01b Remove checks in fxcrt now that FX_NEW can't return 0.
ae4256f Fix offset outside bounds of constant string warnings
b330016 Don't use NEON optimized code in lpng_v163
ff05471 Make FX_NEW throw on OOM.
b48704f Fix two new / free mismatches
4926900 Kill CFX_Object.

BUG=380476,440500,477162

Review URL: https://codereview.chromium.org/1100613002

Cr-Commit-Position: refs/heads/master@{#325925}

[modify] http://crrev.com/76ef7c9b28521b1f5c27d2677196028f01d0a576/DEPS
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #91 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c91

The following revision refers to this bug:
http://src.chromium.org/viewvc/blink?view=rev&rev=194559

------------------------------------------------------------------
r194559 | ***@chromium.org | 2015-04-28T01:45:03.156350Z

Changed paths:
M
http://src.chromium.org/viewvc/blink/trunk/Source/platform/heap/StackFrameDepth.h?r1=194559&r2=194558&pathrev=194559

vs2015: Suppress warning C4172: returning address of local variable or
temporary: dummy

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1107263002
-----------------------------------------------------------------
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #95 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

I got (most of) the rest staged locally and linking seems exceedingly slow.

Filed https://code.google.com/p/chromium/issues/detail?id=482238.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #98 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c98

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/72ef46b64a7b8b6fef8ef81a4cf2959c119c9f4c

commit 72ef46b64a7b8b6fef8ef81a4cf2959c119c9f4c
Author: scottmg <***@chromium.org>
Date: Wed Apr 29 05:59:39 2015

vs2015: Don't have posix error codes defined during libxml build

FAILED: ninja -t msvc -e environment.x86 -- "C:\Program Files
(x86)\Microsoft Visual Studio 14.0\VC\bin\cl.exe" /nologo /showIncludes /FC
@obj\third_party\libxml\src\libxml.nanohttp.obj.rsp
/c ..\..\third_party\libxml\src\nanohttp.c
/Foobj\third_party\libxml\src\libxml.nanohttp.obj
/Fdobj\third_party\libxml\libxml.c.pdb
d:\src\cr2\src\third_party\libxml\src\include\wsockcompat.h(41): error
C2220: warning treated as error - no 'object' file generated
d:\src\cr2\src\third_party\libxml\src\include\wsockcompat.h(41): warning
C4005: 'EWOULDBLOCK': macro redefinition
... etc.

(I don't know why we have nanoftp.c and nanohttp.c being built, but that
seems wrong too.)

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1104333003

Cr-Commit-Position: refs/heads/master@{#327440}

[modify]
http://crrev.com/72ef46b64a7b8b6fef8ef81a4cf2959c119c9f4c/third_party/libxml/README.chromium
[add]
http://crrev.com/72ef46b64a7b8b6fef8ef81a4cf2959c119c9f4c/third_party/libxml/patches/win32-no-posix-error-codes
[modify]
http://crrev.com/72ef46b64a7b8b6fef8ef81a4cf2959c119c9f4c/third_party/libxml/src/libxml.h
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #103 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c103

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/d8fb9c32b6513c005e6791eaff60989d49acf1db

commit d8fb9c32b6513c005e6791eaff60989d49acf1db
Author: scottmg <***@chromium.org>
Date: Thu Apr 30 19:14:51 2015

vs2015: remove unused and deprecated hash_set #include

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1118003002

Cr-Commit-Position: refs/heads/master@{#327760}

[modify]
http://crrev.com/d8fb9c32b6513c005e6791eaff60989d49acf1db/device/bluetooth/bluetooth_adapter_win.cc
[modify]
http://crrev.com/d8fb9c32b6513c005e6791eaff60989d49acf1db/device/bluetooth/bluetooth_adapter_win.h
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #104 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c104

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/76cb78651198ca8901524a7d2db70bc2a6521004

commit 76cb78651198ca8901524a7d2db70bc2a6521004
Author: scottmg <***@chromium.org>
Date: Thu Apr 30 20:30:17 2015

vs2015: avoid int->float warning in printing/emf_win.cc

d:\src\cr2\src\printing\emf_win.cc(548): warning C4838: conversion
from 'int' to 'FLOAT' requires a narrowing conversion
d:\src\cr2\src\printing\emf_win.cc(569): warning C4838: conversion
from 'int' to 'FLOAT' requires a narrowing conversion

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1113113002

Cr-Commit-Position: refs/heads/master@{#327781}

[modify]
http://crrev.com/76cb78651198ca8901524a7d2db70bc2a6521004/printing/emf_win.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #106 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c106

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/c03e665e34982fe98f197a4a0b589f62192087ed

commit c03e665e34982fe98f197a4a0b589f62192087ed
Author: scottmg <***@chromium.org>
Date: Fri May 01 02:48:51 2015

vs2015: avoid warning about narrowing pointer to int in base/win

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1113043003

Cr-Commit-Position: refs/heads/master@{#327866}

[modify]
http://crrev.com/c03e665e34982fe98f197a4a0b589f62192087ed/base/win/shortcut.cc
[modify]
http://crrev.com/c03e665e34982fe98f197a4a0b589f62192087ed/base/win/win_util.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #107 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c107

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/359cb535d023f8c83f7b6a796bc43a30f4594ad7

commit 359cb535d023f8c83f7b6a796bc43a30f4594ad7
Author: scottmg <***@chromium.org>
Date: Fri May 01 02:50:04 2015

vs2015: fix narrowing warnings in sandbox/win/

d:\src\cr2\src\sandbox\win\src\broker_services.cc(282): warning
C4311: 'reinterpret_cast': pointer truncation from 'LPOVERLAPPED' to 'DWORD'
d:\src\cr2\src\sandbox\win\src\broker_services.cc(282): warning
C4302: 'reinterpret_cast': truncation from 'LPOVERLAPPED' to 'DWORD'
d:\src\cr2\src\sandbox\win\src\broker_services.cc(312): warning
C4311: 'reinterpret_cast': pointer truncation from 'LPOVERLAPPED' to 'DWORD'
d:\src\cr2\src\sandbox\win\src\broker_services.cc(312): warning
C4302: 'reinterpret_cast': truncation from 'LPOVERLAPPED' to 'DWORD'
d:\src\cr2\src\sandbox\win\src\broker_services.cc(546): warning
C4312: 'reinterpret_cast': conversion from 'DWORD' to 'LPOVERLAPPED' of
greater size

d:\src\cr2\src\sandbox\win\src\eat_resolver.cc(51): warning
C4302: 'reinterpret_cast': truncation from 'void *' to 'DWORD'
d:\src\cr2\src\sandbox\win\src\eat_resolver.cc(52): warning
C4302: 'reinterpret_cast': truncation from 'const void *' to 'DWORD'

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1119783002

Cr-Commit-Position: refs/heads/master@{#327868}

[modify]
http://crrev.com/359cb535d023f8c83f7b6a796bc43a30f4594ad7/sandbox/win/src/broker_services.cc
[modify]
http://crrev.com/359cb535d023f8c83f7b6a796bc43a30f4594ad7/sandbox/win/src/eat_resolver.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #108 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c108

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/37bb04f1d7e827fe8e1b79996a9937d502834061

commit 37bb04f1d7e827fe8e1b79996a9937d502834061
Author: scottmg <***@chromium.org>
Date: Fri May 01 04:16:13 2015

vs2015: avoid narrowing warning in ui/base/ime/win

d:\src\cr2\src\ui\base\ime\win\imm32_manager.cc(137): warning
C4302: 'reinterpret_cast': truncation from 'HKL' to 'LANGID'
d:\src\cr2\src\ui\base\ime\win\imm32_manager.cc(557): warning C4302: 'type
cast': truncation from 'HKL__ *' to 'WORD'

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1114183002

Cr-Commit-Position: refs/heads/master@{#327877}

[modify]
http://crrev.com/37bb04f1d7e827fe8e1b79996a9937d502834061/ui/base/ime/win/imm32_manager.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #109 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c109

The following revision refers to this bug:
http://src.chromium.org/viewvc/blink?view=rev&rev=194837

------------------------------------------------------------------
r194837 | ***@chromium.org | 2015-05-02T03:56:04.431379Z

Changed paths:
M
http://src.chromium.org/viewvc/blink/trunk/Source/core/dom/DOMMatrixReadOnly.cpp?r1=194837&r2=194836&pathrev=194837
M
http://src.chromium.org/viewvc/blink/trunk/Source/core/html/canvas/WebGLRenderingContextBase.h?r1=194837&r2=194836&pathrev=194837
M
http://src.chromium.org/viewvc/blink/trunk/Source/core/animation/DeferredLegacyStyleInterpolation.cpp?r1=194837&r2=194836&pathrev=194837
M
http://src.chromium.org/viewvc/blink/trunk/Source/core/css/TreeBoundaryCrossingRules.cpp?r1=194837&r2=194836&pathrev=194837

vs2015: warning/compile fixes

It's a little more particular about having not-yet-defined types
wrapped in oilpan ptrs, and warns on the float->int narrowings.

BUG=440500

Review URL: https://codereview.chromium.org/1119103002
-----------------------------------------------------------------
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Updates:
Blockedon: chromium:491983

Comment #118 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

(No comment was entered for this change.)
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Updates:
Blockedon: webrtc:4521

Comment #121 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Here's an update on the remaining bits to get "chrome" compiling and
linking.

I haven't tried a branding=Chrome build yet, nor any unit test binaries.

Many recent "regressions", so probably should at least get an FYI bot going
soon to be able to catch future ones while they're fresh.

-----------------------------------------------------------------------

Windows SDK path is messed up for non-depot_tools toolchain and has been
for a while. Not really related to vs2015, but needs to be fixed to test
with system-installed toolchain.
https://codereview.chromium.org/1152003003/

-----------------------------------------------------------------------

Still some narrowing warnings. Nico mentioned that that's disabled on
POSIX/clang anyway, so probably might as well just disable on Windows too.
I added a bunch of static_cast to avoid the warnings already, not sure if
it's worth backing those out too.
https://codereview.chromium.org/1153633008/

-----------------------------------------------------------------------

d:\src\cr2\src\third_party\webrtc\base\httpcommon.cc(390): error
C2065: 'timezone': undeclared identifier

Logged, but not fixed upstream (I don’t think)
https://code.google.com/p/webrtc/issues/detail?id=4521

Posted https://webrtc-codereview.appspot.com/50089004/ as a fix.

-----------------------------------------------------------------------

d:\src\cr2\src\third_party\webkit\source\core\css\cssprimitivevalue.cpp(832):
error
C2065: '_TWO_DIGIT_EXPONENT': undeclared identifier
d:\src\cr2\src\third_party\webkit\source\core\css\cssprimitivevalue.cpp(832):
error
C3861: '_set_output_format': identifier not found
d:\src\cr2\src\third_party\webkit\source\core\css\cssprimitivevalue.cpp(836):
error
C3861: '_set_output_format': identifier not found

It’s still documented in VS2015, but doesn’t appear to be in any header
(supposed to be in stdio.h).

Filed upstream:
https://connect.microsoft.com/VisualStudio/feedback/details/1368280

https://code.google.com/p/chromium/issues/detail?id=491979 for Chromium.

-----------------------------------------------------------------------

Thread safe statics CRITICAL_SECTION initialization failing inside the
sandbox at startup

cpu@ fix (untested, but should be OK):
https://codereview.chromium.org/1138463005/

-----------------------------------------------------------------------

template deduction problem in devtools_embedder_message_dispatcher.cc

Might be compiler problem? Or might be invalid code, not sure.

https://code.google.com/p/chromium/issues/detail?id=491973

-----------------------------------------------------------------------

Behaviour change/bug fix (?) in the preprocessor breaks
libjingle/source/talk/session/media/srtpfilter.cc

https://code.google.com/p/chromium/issues/detail?id=491982

-----------------------------------------------------------------------

Fail to compile in MultiColumnFragmentainerGroup, only in
component=shared_library.

https://code.google.com/p/chromium/issues/detail?id=491988

-----------------------------------------------------------------------

Link failing due to const char Name_[] in mojom generated code
https://code.google.com/p/chromium/issues/detail?id=491983

-----------------------------------------------------------------------

Link failing due to confused dllimport in Supplementable (oilpan)

https://code.google.com/p/chromium/issues/detail?id=491995
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #126 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c126

The following revision refers to this bug:
http://src.chromium.org/viewvc/blink?view=rev&rev=196266

------------------------------------------------------------------
r196266 | ***@chromium.org | 2015-06-02T01:11:06.246826Z

Changed paths:
M
http://src.chromium.org/viewvc/blink/trunk/Source/core/css/CSSPrimitiveValue.cpp?r1=196266&r2=196265&pathrev=196266

win: _set_ouput_format is no longer available or required on VS2015

The default is now a standards-conforming two-digit exponent mode.

R=***@chromium.org
BUG=440500, 491979,
https://connect.microsoft.com/VisualStudio/feedback/details/1368280

Review URL: https://codereview.chromium.org/1161773006
-----------------------------------------------------------------
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #127 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c127

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/176353a4bceff74ca34d6ebd0708b8894cda6d73

commit 176353a4bceff74ca34d6ebd0708b8894cda6d73
Author: scottmg <***@chromium.org>
Date: Tue Jun 02 03:32:23 2015

Update vs2015 toolchain hash

Missed some files in the last version. (Corresponds to
ps#8 in https://codereview.chromium.org/1165563003/ )

TBR=***@chromium.org
BUG=492774,440500

Review URL: https://codereview.chromium.org/1168433003

Cr-Commit-Position: refs/heads/master@{#332339}

[modify]
http://crrev.com/176353a4bceff74ca34d6ebd0708b8894cda6d73/build/vs_toolchain.py
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #129 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c129

The following revision refers to this bug:

https://chromium.googlesource.com/external/webrtc.git/+/84f81d8fa10d65349ae3132dc98fa8353907a225

commit 84f81d8fa10d65349ae3132dc98fa8353907a225
Author: scottmg <***@chromium.org>
Date: Tue Jun 02 09:41:07 2015

Fix implicit size_t to uint16_t warning on VS2015.

Reviewed as https://review.webrtc.org/54609004/, but uploaded against
wrong Base URL.

R=***@webrtc.org
TBR=
BUG=chromium:440500

Review URL: https://webrtc-codereview.appspot.com/54649004

Patch from scottmg <***@chromium.org>.

Cr-Commit-Position: refs/heads/master@{#9351}

[modify]
http://crrev.com/84f81d8fa10d65349ae3132dc98fa8353907a225/webrtc/video/video_send_stream.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #134 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c134

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/tools/depot_tools.git/+/3f722d1f6358963ad61824e78b97335fb63b1868

commit 3f722d1f6358963ad61824e78b97335fb63b1868
Author: ***@chromium.org <***@chromium.org>
Date: Tue Jun 02 02:54:51 2015

Rework/update toolchain script for vs2015

This is the other side of https://codereview.chromium.org/1163723003/

The changes here are to remove the use of 'vs2013_files' and 'win8sdk'
(as those will be different numbers soon enough) but still maintain
behaviour for the old "style" while in transition.

Secondarily, to remove the dependence of these two scripts on
'toolchain2013.py' as most of the script is now unused.

R=***@chromium.org
BUG=440500, 492774

Review URL: https://codereview.chromium.org/1165563003

git-svn-id: svn://svn.chromium.org/chrome/trunk/tools/***@295485
0039d316-1c4b-4281-b951-d872f2087c98

[modify]
http://crrev.com/3f722d1f6358963ad61824e78b97335fb63b1868/win_toolchain/get_toolchain_if_necessary.py
[modify]
http://crrev.com/3f722d1f6358963ad61824e78b97335fb63b1868/win_toolchain/package_from_installed.py
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #138 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c138

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/ec4a2e23d70045832b3ffc12095cab4a89cd629e

commit ec4a2e23d70045832b3ffc12095cab4a89cd629e
Author: scottmg <***@chromium.org>
Date: Wed Jun 10 02:22:52 2015

Set -fmsc-version appropriately for clang on VS2015

Mostly turns on char16_t and char32_t to not error out immediately.

R=***@chromium.org
BUG=82385,440500,498544

Review URL: https://codereview.chromium.org/1172903002

Cr-Commit-Position: refs/heads/master@{#333657}

[modify]
http://crrev.com/ec4a2e23d70045832b3ffc12095cab4a89cd629e/build/common.gypi
[modify]
http://crrev.com/ec4a2e23d70045832b3ffc12095cab4a89cd629e/build/config/compiler/BUILD.gn
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #139 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c139

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/4248771f4f3482c059dff8248de585c733b8b69f

commit 4248771f4f3482c059dff8248de585c733b8b69f
Author: hans <***@chromium.org>
Date: Wed Jun 10 20:28:41 2015

Try to fix the Win-Clang gn build

Builds were failing with:

ERROR at //build/config/compiler/BUILD.gn:116:11: Undefined identifier
if (visual_studio_version == "2013") {
^--------------------

BUG=82385, 440500

Review URL: https://codereview.chromium.org/1178603002

Cr-Commit-Position: refs/heads/master@{#333787}

[modify]
http://crrev.com/4248771f4f3482c059dff8248de585c733b8b69f/build/config/compiler/BUILD.gn
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #146 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c146

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/9b3d2bb5ac548c8063823a2aefde76d6ddd0ad46

commit 9b3d2bb5ac548c8063823a2aefde76d6ddd0ad46
Author: scottmg <***@chromium.org>
Date: Tue Jun 16 22:40:50 2015

vs2015: fix float->int warning in gfx/rect_unittests

c:\b\build\slave\chromium-rel-builder\build\src\ui\gfx\geometry\rect_unittest.cc(490):
error
C2220: warning treated as error - no 'object' file generated
c:\b\build\slave\chromium-rel-builder\build\src\ui\gfx\geometry\rect_unittest.cc(490):
warning
C4305: 'initializing': truncation from 'const int' to 'float'
c:\b\build\slave\chromium-rel-builder\build\src\ui\gfx\geometry\rect_unittest.cc(526):
warning
C4305: 'initializing': truncation from 'const int' to 'float'

After https://codereview.chromium.org/1186133003.

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1186243003

Cr-Commit-Position: refs/heads/master@{#334723}

[modify]
http://crrev.com/9b3d2bb5ac548c8063823a2aefde76d6ddd0ad46/ui/gfx/geometry/rect_unittest.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #147 on issue 440500 by ***@gmail.com: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Hello Scott,
I can't create right package from installed toolchain, here are my steps:
1. Configure a virtual machine in the VirtualBox for the Windows 7 (64 bit)
installation, disable a network card.
2. Install the Windows 7 from
en_windows_7_ultimate_with_sp1_x64_dvd_u_677332.iso (SHA-1:
36AE90DEFBAD9D9539E649B193AE573B77A71C83).
3. Install latest offline update for root certificates from rootsupd.exe
(SHA-1: 9232833004CF02135DBA887D2037C4487FADD42D,
https://code.google.com/p/chromium/issues/detail?id=402543#c15 )
4. Install the VS2015 RC from
en_visual_studio_professional_2015_rc_x86_dvd_6649994.iso (SHA-1:
2A1EE16D566900AF40BA3C2B387EB9CB6D62F099). Select only the MFC.
5. Nothing
OR
Create offline installation of the Windows 8.1 SDK from sdksetup.exe
(SHA-1: 4540251F0CA62DA5FD8E68CD4E3129930E7C5AE9) via command "sdksetup
/quiet /layout C:\SDK", copy and install it. Select only the Windows SDK
and the Debugging Tools.
6. Copy updated depot_tools folder and run
command "C:\depot_tools\python276_bin\python.exe
C:\depot_tools\win_toolchain\package_from_installed.py 2015"
7. Result: 912ad2070707ec67f1c5c667e2184c4743f2ee1b.zip
OR
1bbd9d1d8648f81dc6f1120bfef961671bbab0c3.zip


Your SHA-1 hash is 40721575c85171cea5d7afe5ec17bd108a94796e (
https://chromium.googlesource.com/chromium/src.git/+/master/build/vs_toolchain.py
). I'm doing something wrong either way, again :( Attached files contain
CRC32 hash of every single file in my extracted toolchains created via
Total Commander. Please at least upload your hashlist.

Attachments:
912ad2070707ec67f1c5c667e2184c4743f2ee1b.sfv 380 KB
1bbd9d1d8648f81dc6f1120bfef961671bbab0c3.sfv 406 KB
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #148 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

#147, Hi, it probably doesn't matter much until RTM since we won't actually
be switching to 2015 before that.

But I've attached an sfv of mine, sort and diffing vs. 1bbd9d....sfv I get
this

C:\Users\Scott\Downloads>diff
1bbd9d1d8648f81dc6f1120bfef961671bbab0c3.sfv ../Desktop/vs_files.sfv
1,2c1,2
< ; Generated by WIN-SFV32 v1.0
< ; (Compatible: Total Commander 8.51a)
---

; Generated by WIN-SFV32 v1.1a (QuickSFV v2.36 Compatibility Mode) on
2015-06-19 at 14:08:02
; http://www.QuickSFV.org

2168a2169

VC\include\CodeCoverage\CodeCoverage.h 3DC306D9

3797a3799,3800

win_sdk\Catalogs\cat036a669bbe9ef85a13c02fcfc37c2fa7.cat 923A9642
win_sdk\Catalogs\cat043c74735c25e27b33dd0c6d76be7d6d.cat 6B368FA4

3841a3845

win_sdk\Catalogs\cat7989b22fad032f64c282e2b62486231f.cat E7BCF8CC

3853a3858

win_sdk\Catalogs\catac1c5b183f159f7d2aebc1d745eb8e9a.cat EB8D16CD

7756c7761
< win_sdk\References\CommonConfiguration\Neutral\Annotated\Windows.winmd
F665D0E6
---

win_sdk\References\CommonConfiguration\Neutral\Annotated\Windows.winmd
7D55072E

which is a pretty odd set of minor differences. Hmm...

I don't remember how I installed the sdk now, I think I probably just
installed it "online".

I'm not sure why you wouldn't have VC/include/CodeCoverage/CodeCoverage.h
though.

Attachments:
vs_files.sfv 406 KB
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #149 on issue 440500 by ***@gmail.com: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

#148, Hi Scott, thank you for comparison and the sfv. Unlike
en_visual_studio_professional_2015_rc_x86_dvd_6649994.iso,
en_visual_studio_enterprise_2015_rc_x86_dvd_6649702.iso (SHA-1:
596DF9E404E2F120815C37B39D3AA64B4189073)contains
VC\include\CodeCoverage\CodeCoverage.h.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #151 on issue 440500 by ***@gmail.com: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

#150, Unfortunately no idea, I installed the SDK online and result is same.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #153 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

I've accidentally continued building pdfium with VS 2015 and I found a
nasty code-gen bug in VS 2015 RC because of this. It only manifests in
debug builds but that's because the problematic code is #ifdefed out in
release builds - the fundamental compiler issue occurs across a wide range
of compiler and optimizer settings.

The bug is
https://connect.microsoft.com/VisualStudio/feedback/details/1531423. No
word yet on whether it is fixed in VS 2015 RTM.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #154 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c154

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/d5b9a8565675b8969dd34b6aa765adcd443b5527

commit d5b9a8565675b8969dd34b6aa765adcd443b5527
Author: brucedawson <***@chromium.org>
Date: Thu Jul 16 00:43:45 2015

Disable VS 2015 warning that fires on compiler-generated constructors.

C4589 fires when a derived class illegally calls the constructor for
its virtual base class. Unfortunately it fires inappropriately.

If Base has a pure-virtual function and Derived is dllexported and
derives virtually from the base class then the compiler-generated
copy constructor will illegally initialize the virtual base class.
The warning is:
warning C4589 : Constructor of abstract class 'Derived' ignores
initializer for virtual base class 'Base'
note : This diagnostic occurred in the compiler generated function
'Derived::Derived(const Derived &)'

While this warning can be potentially useful (it found bad code in
crbug.com/510497) it mostly fires inappropriately. Here is repro code:

class __declspec(dllexport) Base {
virtual void SomeFunction() = 0;
};

class __declspec(dllexport) Derived : virtual public Base {
// Deleting the assignment operator avoids the warning
//Derived(const Derived&) = delete;

// Overriding SomeFunction avoids the warning
//void SomeFunction() override {};
};

The bug has been reported.

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1234253003

Cr-Commit-Position: refs/heads/master@{#338964}

[modify]
http://crrev.com/d5b9a8565675b8969dd34b6aa765adcd443b5527/build/common.gypi
[modify]
http://crrev.com/d5b9a8565675b8969dd34b6aa765adcd443b5527/build/config/compiler/BUILD.gn
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #155 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

For my own future reference, the VS 2015 buildbots are:

http://build.chromium.org/p/chromium.fyi/builders/Chromium%20Builder
http://build.chromium.org/p/chromium.fyi/builders/Chromium%20Builder%20%28dbg%29

Breakage on the dbg builder is a known RC compiler bug
(https://connect.microsoft.com/VisualStudio/feedback/details/1531423) that
is fixed in RTM.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #156 on issue 440500 by ***@google.com: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

V8 tracking bug here: https://code.google.com/p/v8/issues/detail?id=4326
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Updates:
Cc: ***@chromium.org

Comment #157 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

(No comment was entered for this change.)
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #159 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Change https://codereview.chromium.org/1265503002/, plus installing the VS
2015 CRT redistributables has got the VS 2015 debug builder
(http://build.chromium.org/p/chromium.fyi/builders/Chromium%20Builder%20%28dbg%29)
to make it a bit further. Instead of failing due to a compiler bug it now
fails because it is looking for the VS 2013 debug CRT DLLs. This also
happens to me with local component builds. Here is the error message from

Failed to find an input file: Input file
C:\b\build\slave\chromium-dbg-builder\build\src\out\Debug\msvcp120d.dll
doesn't exist

--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #160 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

It should be pretty easy to fix, it's just that base.isolate explicitly
list msvcp120.dll as a dependency, I'll make it use msvcp140.dll for the
VS2015 builds
(https://code.google.com/p/chromium/codesearch#chromium/src/base/base.isolate)
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #162 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

The C1090 error brings back nightmares
(https://code.google.com/p/chromium/issues/detail?id=339215)

The VS2015 builder is currently broken (something in Blink), I'm looking at
this.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #163 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c163

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/ad86c6fcef0fbdf26da9165ae8bbbd3748201afc

commit ad86c6fcef0fbdf26da9165ae8bbbd3748201afc
Author: brucedawson <***@chromium.org>
Date: Tue Aug 11 21:05:14 2015

Reduce DefaultConcurrentLinks from phys/4GB to phys/5GB.

This is copied from the gyp repo, ninja.py:
https://codereview.chromium.org/1255813005

VS 2015 uses about 20% more memory when linking. This caused a
severe page-fault storm on my development machine and subsequent tests
showed that linking could use ~63 GB of RAM (15 parallel links on a
64 GB machine). This suggests that VS 2013 would also be using a lot
of RAM (52.5 GB?) such that very little was left for other
applications or disk cache.

Linker working sets vary wildly but with VS 2013 there are six links
that use over 6 GB of working set and with VS 2015 there are seven.

This change reduces parallel links by 20% (increase RAM-per-linker by
25%). The value can be overridden by setting GYP_LINK_CONCURRENCY.

Tests with VS 2013 showed no statistically significant change in
build times when link concurrency was reduced from 15 to 10 or 12.
Further reductions may be prudent when we switch to VS 2015.

BUG=440500
R=***@chromium.org

Review URL: https://codereview.chromium.org/1261073004

Cr-Commit-Position: refs/heads/master@{#342892}

[modify]
http://crrev.com/ad86c6fcef0fbdf26da9165ae8bbbd3748201afc/build/toolchain/get_concurrent_links.py
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #168 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

I'm fixing the issue mentioned in
https://code.google.com/p/chromium/issues/detail?id=440500#c167. It's a
case of missing 'const', easily corrected.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #169 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c169

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/7526953339ccb21e1225bc1d849ddbd109c3c2fb

commit 7526953339ccb21e1225bc1d849ddbd109c3c2fb
Author: brucedawson <***@chromium.org>
Date: Tue Aug 18 05:19:10 2015

Added const to DownloadItemIdCompare parameters

The DownloadItemIdCompare function was originally written to take
non-const pointers. This was presumably unintentional. This omission
means that const objects cannot be compared, and it breaks VS 2015
whose STL implementation prototypes the comparison path as taking const
objects.

The error is:

error C3848: expression having type
'const DownloadGroupNotification::DownloadItemIdCompare' would lose some
const-volatile qualifiers in order to call
'bool DownloadGroupNotification::DownloadItemIdCompare::operator ()'

The fix is simple and safe and appropriate and will unblock the VS
2015 builder.

The bug was introduced in https://codereview.chromium.org/1287803003

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1298003002

Cr-Commit-Position: refs/heads/master@{#343816}

[modify]
http://crrev.com/7526953339ccb21e1225bc1d849ddbd109c3c2fb/chrome/browser/download/notification/download_group_notification.h
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #170 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Release builds are green again. Debug builds are still red, now with:

unresolved external symbol "public: void __thiscall
policy::PolicyMap::Set(...)

That's new as of build 3914:
http://build.chromium.org/p/chromium.fyi/builders/Chromium%20Builder%20%28dbg%29/builds/3914

Investigating.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #171 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

It looks like out\Debug\policy_component.dll.lib is not getting linked into
chrome.exe, and that is why the unresolved symbol is occurring. I don't
know why this is different from VS 2013 builds and I have to move onto some
other projects for now.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #173 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Regarding comment #172, win_fastlink=1 is working fine for me.

When I set win_fastlink=1 and build with VS 2015 I see that the PDBs are
significantly smaller. I'm using gyp, specifically "python

set gyp_defines

GYP_DEFINES=component=shared_library disable_nacl=1

dir out\Release\base.dll.pdb | find /i ".pdb"

08/20/2015 05:53 PM 19,214,336 base.dll.pdb


C:\src\analyze_chromium\src>set gyp_defines
GYP_DEFINES=component=shared_library disable_nacl=1 win_fastlink=1
C:\src\analyze_chromium\src>dir out\Release\base.dll.pdb | find /i ".pdb"
08/20/2015 05:44 PM 7,983,104 base.dll.pdb
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #174 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

@brucedawson: Sorry for the confusion. On closer look I see that
win_fastlink support was only added for Release builds (added to
Release_Base) and I was creating a Debug build. Would it make sense to
support this flag with Debug builds as well?
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #175 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

@#174: I patched common.gypi locally to add win_fastlink support in
Debug_Base and ran a Debug build of my Chromium-based project with
/DEBUG:FASTLINK. Some numbers compared to a VS2015 Debug build without
/DEBUG:FASTLINK:

compile time = ~6% slower (3706.89s vs 3511.5s)
link time = ~47% faster (141.84s vs 269.40s)
peak link time memory usage (total system) = ~29% less (5.5GB vs 7.8GB*)
out/Debug size = ~5% smaller (16.4GB vs 17.3GB)

* there may be some paging going on since my system has only 8GB of RAM.

Interesting side note: even without /DEBUG:FASTLINK the VS2015 Debug build
linked ~18% faster than the VS2013 Debug build (269.40s vs 327.35s) in my
testing.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #176 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

It would make sense to support fastlink for debug builds as well. You can
put together a patch if you want. If you don't then I will.

- The compile time slowdown you report must be spurious because fastlink
only affects the linker. Or it is a side-channel affect where faster
linking stole CPU time or memory from the compiler.
- The most reliable way to measure the impact of fastlink (not sensitive to
distortions from build contention and paging) is looking at the size of
the .pdb file. That is where the out/Debug reduction is coming from. Delete
the .pdb before tests because VC++ never shrinks .pdb files. Then you can
be sure it is working.

I talked about this here:
https://randomascii.wordpress.com/2015/07/27/programming-is-puzzles/
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #178 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Regarding the linker errors on VS 2015 debug builds... (see comment #171).
I believe it is a compiler/linker bug and I have filed a connect bug
regarding it:
https://connect.microsoft.com/VisualStudio/feedback/details/1737411/compiler-creates-references-to-global-delete-which-are-satisfied-by-pseudo-random-object-files

I'm waiting to here back from Microsoft on whether they have any suggested
workarounds.

After comparing VS 2013 and VS 2015 verbose links I confirmed that
policy::PolicyMap::Set is not supposed to be linked into chrome.exe. The
linker looks for it because policy.cloud_policy_generated.obj is pulled in
erroneously. Searching for that .obj in the verbose linker output revealed
this line:

Searching obj\components\policy.lib:
Found "void __cdecl __global_delete(void *,unsigned int)"
(?__global_delete@@***@Z)
Referenced in msvcprtd.lib(locale0_implib.obj)
Loaded policy.lib(policy.cloud_policy_generated.obj)

So, the linker is pulling in policy.cloud_policy_generated.obj in order to
get a definition for __global_delete. This is a compiler generated function
that is inserted into any object file that does a DLL export of a class
with a virtual function. When msvcprtd.lib is pulled in the linker then
chooses a pseudo-randomly selected object file that exports the symbol to
be linked in, leading to the error:

Investigative steps, for future reference:

Getting the verbose linker output was done by doing a verbose build and
retaining the .rsp files:
ninja -v -d keeprsp -C out\Debug chrome
Then I added the compiler and linker to the path with:
call "%VS140COMNTOOLS%..\..\VC\vcvarsall.bat" amd64_x86
Then I could manually run the link with the /verbose flag:
link.exe /verbose /nologo /OUT:initialexe\chrome.exe
@initialexe\chrome.exe.rsp
Then I searched the verbose output for policy.cloud_policy_generated.obj. I
knew this .obj file was not supposed to be linked in because it used a
symbol from policy_component.dll.lib and I proved that that wasn't needed
by renaming it. Then I reduced cloud_policy_generated.cc while repeatedly
running "dumpbin policy.cloud_policy_generated.obj /symbols | find
/i "global_delete"" until I determined what triggered the problematic
export.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #181 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c181

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/84f30fe06877193d39bb2cb9fe345c2e12e695b3

commit 84f30fe06877193d39bb2cb9fe345c2e12e695b3
Author: brucedawson <***@chromium.org>
Date: Fri Sep 11 03:08:29 2015

Get GenericScopedHandle::Set to preserve LastError code

Code like this is elegant and clean and common in Chrome:

event_.Set(CreateEvent(NULL, FALSE, FALSE, wname.c_str()));
if (event_.Get() && GetLastError() != ERROR_ALREADY_EXISTS) {

However it will behave incorrectly if event_.Set() zeroes out the
Windows LastError code, which VC++ 2015 frequently does. This change
avoids that.

R=***@chromium.org
BUG=528394,529981,440500

Review URL: https://codereview.chromium.org/1320673016

Cr-Commit-Position: refs/heads/master@{#348324}

[modify]
http://crrev.com/84f30fe06877193d39bb2cb9fe345c2e12e695b3/base/BUILD.gn
[modify]
http://crrev.com/84f30fe06877193d39bb2cb9fe345c2e12e695b3/base/base.gyp
[modify]
http://crrev.com/84f30fe06877193d39bb2cb9fe345c2e12e695b3/base/win/scoped_handle.h
[add]
http://crrev.com/84f30fe06877193d39bb2cb9fe345c2e12e695b3/base/win/scoped_handle_unittest.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Updates:
Blockedon: chromium:533002

Comment #182 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

(No comment was entered for this change.)
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #185 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

I was seeing a compile warning/error in Release/x64, possibly related to
2015?

src\chromium\src\third_party\zlib\crc_folding.c(286): warning C4311: 'type
cast': pointer truncation from 'const unsigned char *' to 'unsigned long'
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #192 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Hi,

Today I've confirmed chrome builds with VS2015 (depot_tools package,
GYP_MSVS_VERSION=2015) + goma without any error. I had a few issues last
week (e.g OLEAUT32.dll is missing), but they have been all fixed.

So, now, anytime ok to enable goma with VS2015.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #197 on issue 440500 by ***@chromium.org: Get Chrome compiling
on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

It's a good news that goma is working correctly with VS2015.
If you found goma related error, I'm happy to fix it.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #198 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c198

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/93a8d6c6a00e690fc4fcdc55d016d0ac8c6fab7c

commit 93a8d6c6a00e690fc4fcdc55d016d0ac8c6fab7c
Author: brucedawson <***@chromium.org>
Date: Tue Oct 13 15:51:25 2015

Remove crc_folding.c from google.patch

The recent change to crc_folding.c was added to google.patch but this
was inappropriate because crc_folding.c is not part of the original zlib
distribution. This was noticed by scottmg@

R=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1397813003

Cr-Commit-Position: refs/heads/master@{#353760}

[modify]
http://crrev.com/93a8d6c6a00e690fc4fcdc55d016d0ac8c6fab7c/third_party/zlib/google.patch
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #200 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c200

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/6fc7bbe2e81f3b9e0a37e4fcb37896a18f08afde

commit 6fc7bbe2e81f3b9e0a37e4fcb37896a18f08afde
Author: brucedawson <***@chromium.org>
Date: Wed Oct 28 01:56:26 2015

Add missing include of <iterator>

Some VS 2015 builds fail if you use back_inserter without explicitly
including <iterator>. This has broken the VS 2015 build machines for a
while (unclear exact time due to a separate build break).

I cannot reproduce the compile failure locally but I believe that this
fix is appropriate and will probably fix the build machines.

See http://en.cppreference.com/w/cpp/iterator/back_inserter.

BUG=440500

Review URL: https://codereview.chromium.org/1424473005

Cr-Commit-Position: refs/heads/master@{#356486}

[modify]
http://crrev.com/6fc7bbe2e81f3b9e0a37e4fcb37896a18f08afde/base/trace_event/memory_profiler_heap_dump_writer.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #201 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c201

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/365355d3edb4c491a6d5a6573c14509ea6b7260c

commit 365355d3edb4c491a6d5a6573c14509ea6b7260c
Author: brucedawson <***@chromium.org>
Date: Wed Oct 28 04:52:22 2015

Avoid double-to-float truncation warning

29.97 is a double-precision constant. VC++ 2015 complains if it is used
to initialize a float member of a struct, which is entirely reasonable.
29.97 and 29.97f have different values, and C++ 11 is supposed to
prohibit rounding when initializing aggregates.

TBR=***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1407083005

Cr-Commit-Position: refs/heads/master@{#356514}

[modify]
http://crrev.com/365355d3edb4c491a6d5a6573c14509ea6b7260c/media/capture/video/fake_video_capture_device_unittest.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #203 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c203

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/0693ed787f6ffd2b248ef8cf2c0dc050a4e9d8b2

commit 0693ed787f6ffd2b248ef8cf2c0dc050a4e9d8b2
Author: brucedawson <***@chromium.org>
Date: Fri Oct 30 01:28:16 2015

Get VS 2015 to use const int definitions

VC++ up to VS 2015 RTM does not require explicit storage allocation for
static const integers declared in classes. VS 2015 Update 1 requires
these storage definitions in some cases. It's unclear exactly what
cases - simple tests work with and without the explicit storage
allocation.

Many previous versions of VC++ have theoretically *allowed* a
definition to supply storage, but tests on VC++ 2013 show that this
doesn't actually work correctly - it leads to duplicate definition
errors. So, the change is scoped to VS 2015 only.

This changed was landed in the upstream protobuf repo:
https://github.com/google/protobuf/commit/a74e912a8be1274

With this change the chrome target builds with the latest VS 2015.

R=***@chromium.org, ***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1422453005

Cr-Commit-Position: refs/heads/master@{#357013}

[modify]
http://crrev.com/0693ed787f6ffd2b248ef8cf2c0dc050a4e9d8b2/third_party/WebKit/Source/platform/graphics/Color.cpp
[modify]
http://crrev.com/0693ed787f6ffd2b248ef8cf2c0dc050a4e9d8b2/third_party/protobuf/README.chromium
[modify]
http://crrev.com/0693ed787f6ffd2b248ef8cf2c0dc050a4e9d8b2/third_party/protobuf/src/google/protobuf/compiler/cpp/cpp_enum.cc
[modify]
http://crrev.com/0693ed787f6ffd2b248ef8cf2c0dc050a4e9d8b2/third_party/protobuf/src/google/protobuf/compiler/cpp/cpp_extension.cc
[modify]
http://crrev.com/0693ed787f6ffd2b248ef8cf2c0dc050a4e9d8b2/third_party/protobuf/src/google/protobuf/compiler/cpp/cpp_message.cc
[modify]
http://crrev.com/0693ed787f6ffd2b248ef8cf2c0dc050a4e9d8b2/third_party/protobuf/src/google/protobuf/compiler/plugin.pb.cc
[modify]
http://crrev.com/0693ed787f6ffd2b248ef8cf2c0dc050a4e9d8b2/third_party/protobuf/src/google/protobuf/descriptor.cc
[modify]
http://crrev.com/0693ed787f6ffd2b248ef8cf2c0dc050a4e9d8b2/third_party/protobuf/src/google/protobuf/descriptor.pb.cc
[modify]
http://crrev.com/0693ed787f6ffd2b248ef8cf2c0dc050a4e9d8b2/third_party/protobuf/src/google/protobuf/wire_format_lite.cc
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #204 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

VS 2015 Update 1 Release Candidate was released yesterday:
https://www.visualstudio.com/news/vs2015-update1-vs

It includes one (undocumented) breaking change which was partially fixed
with
crrev.com/1428003002 (gets chrome building) and will be completely fixed
with crrev.com/1422453005 (gets all targets building).

I'm examining test failures (bug 533002) and malloc hooking (481611).
Linker crashes (bug 482671) should be fixed, and Ctrl+C crashes (bug
486849) are not really blocking.

Compiler packaging (bug 495944) is still an open issue.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Updates:
Owner: ***@chromium.org

Comment #205 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500

Assigning to myself.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #206 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c206

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/599857c9a6a07024a6047080f91f71f801deb4f8

commit 599857c9a6a07024a6047080f91f71f801deb4f8
Author: brucedawson <***@chromium.org>
Date: Wed Nov 04 00:48:59 2015

Avoid const int definition problems

VC++ up to VS 2015 RTM does not require explicit storage allocation for
static const integers declared in classes. VS 2015 Update 1 requires
these storage definitions in some cases. In order to avoid #ifs
around the storage definitions this change switches the problematic
consts to enums, for maximum portability.

Where needed the enums have types specified.

Many previous versions of VC++ have theoretically *allowed* a
definition to supply storage, but tests on VC++ 2013 show that this
doesn't actually work correctly - it leads to duplicate definition
errors. So, enums are the only #if option.

With this change all targets build with the latest VS 2015.

See also 1422453005.

R=***@chromium.org,***@chromium.org
BUG=440500

Review URL: https://codereview.chromium.org/1428003002

Cr-Commit-Position: refs/heads/master@{#357688}

[modify]
http://crrev.com/599857c9a6a07024a6047080f91f71f801deb4f8/base/time/time.h
[modify]
http://crrev.com/599857c9a6a07024a6047080f91f71f801deb4f8/gpu/command_buffer/client/fenced_allocator.cc
[modify]
http://crrev.com/599857c9a6a07024a6047080f91f71f801deb4f8/gpu/command_buffer/client/fenced_allocator.h
[modify]
http://crrev.com/599857c9a6a07024a6047080f91f71f801deb4f8/ui/gfx/icon_util.h
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #208 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c208

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/23e5ea254fea6ad3618588f177406640e7104d38

commit 23e5ea254fea6ad3618588f177406640e7104d38
Author: brucedawson <***@chromium.org>
Date: Thu Nov 05 23:04:37 2015

Add VC++ 2015 file to .gitignore

When using VC++ 2015 Update 1 RC the IDE drops *.VC.opendb files which
then clutter the git status update like this:

Untracked files:
(use "git add <file>..." to include in what will be committed)

chrome_dev.VC.opendb

This change tells git to ignore them.

BUG=440500

Review URL: https://codereview.chromium.org/1413153015

Cr-Commit-Position: refs/heads/master@{#358163}

[modify]
http://crrev.com/23e5ea254fea6ad3618588f177406640e7104d38/.gitignore
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[c***@googlecode.com]

Comment #209 on issue 440500 by ***@chromium.org: Get Chrome
compiling on VS2015
https://code.google.com/p/chromium/issues/detail?id=440500#c209

The following revision refers to this bug:

https://chromium.googlesource.com/chromium/src.git/+/8f81220727fea5fc0c920984d62791880b33a908

commit 8f81220727fea5fc0c920984d62791880b33a908
Author: brucedawson <***@chromium.org>
Date: Sat Nov 07 00:55:23 2015

Ignore VS 2015 .pyproj files and .vs directories

BUG=440500

Review URL: https://codereview.chromium.org/1415713006

Cr-Commit-Position: refs/heads/master@{#358476}

[modify]
http://crrev.com/8f81220727fea5fc0c920984d62791880b33a908/.gitignore
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings