Fix tests for check-ipc Clang plugin to pass on Windows.

Fix tests for check-ipc Clang plugin to pass on Windows.

Add -fno-delayed-template-parsing to disable MSVC-like behavior (check-ipc
will only be used on Android, the only platform where we need 32/64-bit
processes to talk to each other), remove test that relies on size_t being
unsigned long.

BUG=581409
Committed: https://crrev.com/b9d266457c4b0b619304c99a28fce037be7b0507
Cr-Commit-Position: refs/heads/master@{#380271}

[Dmitry Skiba, 2016-03-09 23:03:47]

dskiba@google.com changed reviewers:
+ dcheng@chromium.org, hans@chromium.org, thakis@chromium.org

[hans, 2016-03-09 23:15:07]

lgtm

Please add a motivation to the CL description why -fno-delayed-template-parsing
is OK.

[Dmitry Skiba, 2016-03-09 23:20:05]

Description was changed from

==========
Fix tests for check-ipc Clang plugin to pass on Windows.

Add -fno-delayed-template-parsing to disable MSVC-like behaviour,
remove test that relies on size_t being unsigned long.

BUG=581409
==========

to

==========
Fix tests for check-ipc Clang plugin to pass on Windows.

Add -fno-delayed-template-parsing to disable MSVC-like behavior (check-ipc
will only be used on Android, the only platform where we need 32/64-bit
processes to talk to each other), remove test that relies on size_t being
unsigned long.

BUG=581409
==========

[Dmitry Skiba, 2016-03-09 23:20:41]

On 2016/03/09 23:15:07, hans wrote:
> lgtm
> 
> Please add a motivation to the CL description why
-fno-delayed-template-parsing
> is OK.

Done.

[Dmitry Skiba, 2016-03-09 23:20:57]

The CQ bit was checked by dskiba@google.com

[commit-bot: I haz the power, 2016-03-09 23:21:32]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1785523002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1785523002/20001

[Nico, 2016-03-09 23:22:33]

this is cool to get the tests passing, but in real life clang-cl builds with
-fno-delayed-template-parsing and the test noticed a real problem. it's ok to
land this to unblock the roll, but i think you need to figure out how to get
things working with -fno-delayed-template-parsing (see e.g.
https://bugs.chromium.org/p/chromium/issues/detail?id=486571#c25)

[Dmitry Skiba, 2016-03-09 23:29:51]

On 2016/03/09 23:22:33, Nico wrote:
> this is cool to get the tests passing, but in real life clang-cl builds with
> -fno-delayed-template-parsing and the test noticed a real problem. it's ok to
> land this to unblock the roll, but i think you need to figure out how to get
> things working with -fno-delayed-template-parsing (see e.g.
> https://bugs.chromium.org/p/chromium/issues/detail?id=486571#c25)

I'll check that out, thanks!

[commit-bot: I haz the power, 2016-03-10 00:25:25]

Description was changed from

==========
Fix tests for check-ipc Clang plugin to pass on Windows.

Add -fno-delayed-template-parsing to disable MSVC-like behavior (check-ipc
will only be used on Android, the only platform where we need 32/64-bit
processes to talk to each other), remove test that relies on size_t being
unsigned long.

BUG=581409
==========

to

==========
Fix tests for check-ipc Clang plugin to pass on Windows.

Add -fno-delayed-template-parsing to disable MSVC-like behavior (check-ipc
will only be used on Android, the only platform where we need 32/64-bit
processes to talk to each other), remove test that relies on size_t being
unsigned long.

BUG=581409
==========

[commit-bot: I haz the power, 2016-03-10 00:25:25]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-03-10 00:26:41]

Description was changed from

==========
Fix tests for check-ipc Clang plugin to pass on Windows.

Add -fno-delayed-template-parsing to disable MSVC-like behavior (check-ipc
will only be used on Android, the only platform where we need 32/64-bit
processes to talk to each other), remove test that relies on size_t being
unsigned long.

BUG=581409
==========

to

==========
Fix tests for check-ipc Clang plugin to pass on Windows.

Add -fno-delayed-template-parsing to disable MSVC-like behavior (check-ipc
will only be used on Android, the only platform where we need 32/64-bit
processes to talk to each other), remove test that relies on size_t being
unsigned long.

BUG=581409

Committed: https://crrev.com/b9d266457c4b0b619304c99a28fce037be7b0507
Cr-Commit-Position: refs/heads/master@{#380271}
==========

[commit-bot: I haz the power, 2016-03-10 00:26:42]

Patchset 2 (id:??) landed as
https://crrev.com/b9d266457c4b0b619304c99a28fce037be7b0507
Cr-Commit-Position: refs/heads/master@{#380271}

[Nico, 2016-03-10 15:10:59]

All the ToT bots are still red, this didn't help. See
https://build.chromium.org/p/chromium.fyi/builders/ClangToTWin%28dbg%29/build...:

esting ipc.cpp... failed: expected and actual differed

Actual:

clang.exe: error: clang frontend command failed due to signal (use -v to see
invocation)

clang version 3.9.0 (trunk 263117)

Target: x86_64-pc-windows-msvc

Thread model: posix

InstalledDir:
C:/b/build/slave/ClangToTWin_dbg_/build/src/third_party/llvm-build/Release+Asserts/bin

clang.exe: note: diagnostic msg: PLEASE submit a bug report to
http://llvm.org/bugs/ and include the crash backtrace, preprocessed source, and
associated run script.

clang.exe: note: diagnostic msg: 

********************



PLEASE ATTACH THE FOLLOWING FILES TO THE BUG REPORT:

Preprocessed source(s) and associated run script(s) are located at:

clang.exe: note: diagnostic msg:
C:\Users\CHROME~1\AppData\Local\Temp\ipc-a10ee7.cpp

clang.exe: note: diagnostic msg:
C:\Users\CHROME~1\AppData\Local\Temp\ipc-a10ee7.sh

clang.exe: note: diagnostic msg: 



********************

Expected:

ipc.cpp:83:26: error: [chromium-ipc] IPC::WriteParam() is called on blacklisted
type 'size_t'.

    WriteParam(pickle, p.size); // ERROR

                         ^

ipc.cpp:107:24: error: [chromium-ipc] IPC::WriteParam() is called on blacklisted
type 'long'.

    WriteParam(pickle, static_cast<long>(container.value)); // ERROR





Let me revert the IPC checkers for now.

[Nico, 2016-03-10 15:11:23]

A revert of this CL (patchset #2 id:20001) has been created in
https://codereview.chromium.org/1778383002/ by thakis@chromium.org.

The reason for reverting is: IPC plugin tests are failing on the clang/tot win
bots..

[hans, 2016-03-10 16:31:37]

On 2016/03/10 15:11:23, Nico wrote:
> A revert of this CL (patchset #2 id:20001) has been created in
> https://codereview.chromium.org/1778383002/ by mailto:thakis@chromium.org.
> 
> The reason for reverting is: IPC plugin tests are failing on the clang/tot win
> bots..

Weird, it passed for me when I built the Windows package. Anyway, I suppose this
will allow us to figure out the proper fix.