PPC/s390: modify if/else block to fix debug build

PPC/s390: modify if/else block to fix debug build

changed the if/else block to fix the debug build
failure on PPC/s390.

R=adamk@chromium.org, mbrandy@us.ibm.com

BUG=
LOG=N

Committed: https://crrev.com/d6229137bd79f5055f75ae57c75600cf5c4f4258
Cr-Commit-Position: refs/heads/master@{#35954}

[JaideepBajwa, 2016-05-02 21:28:28]

PTAL

[adamk, 2016-05-02 21:38:55]

https://codereview.chromium.org/1941683003/diff/1/src/parsing/preparser.cc
File src/parsing/preparser.cc (left):

https://codereview.chromium.org/1941683003/diff/1/src/parsing/preparser.cc#ol...
src/parsing/preparser.cc:285: } else {
Your compiler really can't tell that an unconditional 'else' ending in a
'return' is always hit? Why doesn't this affect Parser::ParseScopedStatement,
which has a similar code layout?

[JaideepBajwa, 2016-05-02 21:59:40]

On 2016/05/02 21:38:55, adamk wrote:
> https://codereview.chromium.org/1941683003/diff/1/src/parsing/preparser.cc
> File src/parsing/preparser.cc (left):
> 
>
https://codereview.chromium.org/1941683003/diff/1/src/parsing/preparser.cc#ol...
> src/parsing/preparser.cc:285: } else {
> Your compiler really can't tell that an unconditional 'else' ending in a
> 'return' is always hit? Why doesn't this affect Parser::ParseScopedStatement,
> which has a similar code layout?

we use g++ only, and only get this error for debug build on ppc/s390. I think
its related to the macro expansion of CHECK_OK. I see the following error:
../src/parsing/preparser.cc: In member function
‘v8::internal::PreParser::Statement
v8::internal::PreParser::ParseScopedStatement(bool, bool*)’:
../src/parsing/preparser.cc:290:1: error: control reaches end of non-void
function [-Werror=return-type]
 }
 ^
I'm not sure why it only affects our platforms.

[adamk, 2016-05-02 22:01:15]

https://codereview.chromium.org/1941683003/diff/1/src/parsing/preparser.cc
File src/parsing/preparser.cc (left):

https://codereview.chromium.org/1941683003/diff/1/src/parsing/preparser.cc#ol...
src/parsing/preparser.cc:288: return ParseFunctionDeclaration(CHECK_OK);
If it's only about CHECK_OK, then this can safely be changed from "CHECK_OK" to
"ok".

[JaideepBajwa, 2016-05-02 22:08:40]

On 2016/05/02 22:01:15, adamk wrote:
> https://codereview.chromium.org/1941683003/diff/1/src/parsing/preparser.cc
> File src/parsing/preparser.cc (left):
> 
>
https://codereview.chromium.org/1941683003/diff/1/src/parsing/preparser.cc#ol...
> src/parsing/preparser.cc:288: return ParseFunctionDeclaration(CHECK_OK);
> If it's only about CHECK_OK, then this can safely be changed from "CHECK_OK"
to
> "ok".

OK thanks, yes that works as well. I'll update my fix.

[JaideepBajwa, 2016-05-02 22:26:11]

updated, PTAL

[adamk, 2016-05-02 22:27:57]

lgtm, thanks

[JaideepBajwa, 2016-05-02 22:28:39]

The CQ bit was checked by bjaideep@ca.ibm.com

[commit-bot: I haz the power, 2016-05-02 22:28:51]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1941683003/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1941683003/20001

[commit-bot: I haz the power, 2016-05-02 22:55:18]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-05-02 22:57:58]

Description was changed from

==========
PPC/s390: modify if/else block to fix debug build

changed the if/else block to fix the debug build
failure on PPC/s390.

R=adamk@chromium.org, mbrandy@us.ibm.com

BUG=
LOG=N
==========

to

==========
PPC/s390: modify if/else block to fix debug build

changed the if/else block to fix the debug build
failure on PPC/s390.

R=adamk@chromium.org, mbrandy@us.ibm.com

BUG=
LOG=N

Committed: https://crrev.com/d6229137bd79f5055f75ae57c75600cf5c4f4258
Cr-Commit-Position: refs/heads/master@{#35954}
==========

[commit-bot: I haz the power, 2016-05-02 22:57:59]

Patchset 2 (id:??) landed as
https://crrev.com/d6229137bd79f5055f75ae57c75600cf5c4f4258
Cr-Commit-Position: refs/heads/master@{#35954}