roll clang 163674:165787

roll clang 163674:165787

Also includes r165962 and r165963 on mac to fix a build issue.

BUG=156005, 151927
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=162917

[Nico, 2012-10-15 21:48:27]

http://codereview.chromium.org/11143020/diff/4/tools/clang/plugins/FindBadCon...
File tools/clang/plugins/FindBadConstructs.cpp (right):

http://codereview.chromium.org/11143020/diff/4/tools/clang/plugins/FindBadCon...
tools/clang/plugins/FindBadConstructs.cpp:420: llvm::errs() << "Unknown clang
plugin argument: " << args[i] << "\n";
Note: I made this change before building the packages, so it's not in this
week's plugin. But I forgot to remove the check-cc flag, and then found this
error message confusing, so I'm changing it here.

[Nico, 2012-10-17 22:22:00]

(cancelled due to asan issues)

On Mon, Oct 15, 2012 at 2:48 PM,  <thakis@chromium.org> wrote:
> Reviewers: hans,
>
>
>
http://codereview.chromium.org/11143020/diff/4/tools/clang/plugins/FindBadCon...
> File tools/clang/plugins/FindBadConstructs.cpp (right):
>
>
http://codereview.chromium.org/11143020/diff/4/tools/clang/plugins/FindBadCon...
> tools/clang/plugins/FindBadConstructs.cpp:420: llvm::errs() << "Unknown
> clang plugin argument: " << args[i] << "\n";
> Note: I made this change before building the packages, so it's not in
> this week's plugin. But I forgot to remove the check-cc flag, and then
> found this error message confusing, so I'm changing it here.
>
> Description:
> roll clang 163674:165787
>
> Also includes r165962 and r165963 on mac to fix a build issue.
>
> BUG=none
>
> Please review this at http://codereview.chromium.org/11143020/
>
> SVN Base: svn://svn.chromium.org/chrome/trunk/src
>
> Affected files:
>   M build/common.gypi
>   M tools/clang/plugins/FindBadConstructs.cpp
>   M tools/clang/scripts/plugin_flags.sh
>   M tools/clang/scripts/update.sh
>
>
> Index: build/common.gypi
> diff --git a/build/common.gypi b/build/common.gypi
> index
>
bfa63857ac08ea49f7a08a3234351aef542e5e61..fb4344921542dd5bac1e8decdf16c427d4b0cc5e
> 100644
> --- a/build/common.gypi
> +++ b/build/common.gypi
> @@ -2509,6 +2509,8 @@
>
>                # TODO(thakis): Remove this.
>                '-Wno-implicit-conversion-floating-point-to-bool',
> +              # TODO(thakis): Remove this once http://crbug.com/151927 is
> fixed.
> +              '-Wno-tautological-constant-out-of-range-compare',
>              ],
>              'cflags!': [
>                # Clang doesn't seem to know know this flag.
> @@ -3065,6 +3067,8 @@
>
>                  # TODO(thakis): Remove this.
>                  '-Wno-implicit-conversion-floating-point-to-bool',
> +                # TODO(thakis): Remove this once http://crbug.com/151927 is
> fixed.
> +                '-Wno-tautological-constant-out-of-range-compare',
>                ],
>              }],
>              ['clang==1 and clang_use_chrome_plugins==1', {
> Index: tools/clang/plugins/FindBadConstructs.cpp
> diff --git a/tools/clang/plugins/FindBadConstructs.cpp
> b/tools/clang/plugins/FindBadConstructs.cpp
> index
>
b79a64dbd19959bc449719681212961ed39b382b..0d2c7cd179f372ba5368fe4ff2e149b5d55a86f4
> 100644
> --- a/tools/clang/plugins/FindBadConstructs.cpp
> +++ b/tools/clang/plugins/FindBadConstructs.cpp
> @@ -29,7 +29,7 @@ using namespace clang;
>  namespace {
>
>  bool TypeHasNonTrivialDtor(const Type* type) {
> -  if (const CXXRecordDecl* cxx_r = type->getCXXRecordDeclForPointerType())
> +  if (const CXXRecordDecl* cxx_r = type->getPointeeCXXRecordDecl())
>      return cxx_r->hasTrivialDestructor();
>
>    return false;
> @@ -417,7 +417,7 @@ class FindBadConstructsAction : public PluginASTAction {
>          check_virtuals_in_implementations_ = false;
>        } else {
>          parsed = false;
> -        llvm::errs() << "Unknown argument: " << args[i] << "\n";
> +        llvm::errs() << "Unknown clang plugin argument: " << args[i] <<
> "\n";
>        }
>      }
>
> Index: tools/clang/scripts/plugin_flags.sh
> diff --git a/tools/clang/scripts/plugin_flags.sh
> b/tools/clang/scripts/plugin_flags.sh
> index
>
217c5c3bd616b4bb1504737e2d6d848b1811e40b..92eaad9626b059f37750177f8faa67d3195ba642
> 100755
> --- a/tools/clang/scripts/plugin_flags.sh
> +++ b/tools/clang/scripts/plugin_flags.sh
> @@ -19,6 +19,4 @@ fi
>  echo -Xclang -load -Xclang $CLANG_LIB_PATH/libFindBadConstructs.$LIBSUFFIX
> \
>    -Xclang -add-plugin -Xclang find-bad-constructs \
>    -Xclang -plugin-arg-find-bad-constructs \
> -  -Xclang skip-virtuals-in-implementations \
> -  -Xclang -plugin-arg-find-bad-constructs \
> -  -Xclang check-cc-directory
> +  -Xclang skip-virtuals-in-implementations
> Index: tools/clang/scripts/update.sh
> diff --git a/tools/clang/scripts/update.sh b/tools/clang/scripts/update.sh
> index
>
e9448236c8d28eff957722c2750628dbf383a145..3e568b9eac4d3e1f2e7e36808195a5a91d26d440
> 100755
> --- a/tools/clang/scripts/update.sh
> +++ b/tools/clang/scripts/update.sh
> @@ -8,7 +8,7 @@
>  # Do NOT CHANGE this if you don't know what you're doing -- see
>  # https://code.google.com/p/chromium/wiki/UpdatingClang
>  # Reverting problematic clang rolls is safe, though.
> -CLANG_REVISION=163674
> +CLANG_REVISION=165787
>
>  THIS_DIR="$(dirname "${0}")"
>  LLVM_DIR="${THIS_DIR}/../../../third_party/llvm"
>
>

[Nico, 2012-10-18 23:44:46]

+mark, hans is on vacation.

The linux_asan issue doesn't repro locally and happens only spuriously on the
trybots (the run the patch set one earlier is fine), I'm not sure what's up with
that. I'll try landing and see what the main waterfall bot says.

Note that the mac binary has a small downstream diff to get it to compile. This
is all worked out upstream (http://llvm.org/PR14102 and the issue linked from
there), and it only affects the arm compiler-rt bits which we don't use
anywhere, so I think that's fine. (The diff is in the build log in the zip.)

[Nico, 2012-10-18 23:52:06]

…nevermind, looks like the cc stuff is still too big to build. Ignore for now.

[Mark Mentovai, 2012-10-18 23:52:47]

LGTM

[Nico, 2012-10-19 00:05:15]

On Thu, Oct 18, 2012 at 4:52 PM,  <thakis@chromium.org> wrote:
> …nevermind, looks like the cc stuff is still too big to build. Ignore for
> now.

Looks like the bot picked a revision before r162796, so the failure
isn't surprising. I'm resending a tryjob for mac_asan at r162796.

>
> http://codereview.chromium.org/11143020/

[commit-bot: I haz the power, 2012-10-19 00:27:14]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/thakis@chromium.org/11143020/23003