Fix GCC 4.7 warnings:

Fix GCC 4.7 warnings:

* src/debug.cc (Debug::SetBreakPoint): Compare value not pointer.

Committed: http://code.google.com/p/v8/source/detail?r=10296

[Tobias Burnus, 2011-12-15 20:43:40]

GCC 4.7 (experimental) found with "make ia32.debug x64.debug" some bugs: One
where a pointer instead of the value was compared with ">= 0" and three where
under certain conditions non-void functions didn't return a value.

[Tobias Burnus, 2011-12-20 15:38:09]

Identical patch - but this time the server didn't crash while uploading.

Thus, side-by-side diff is available, which should help with reviewing (esp. for
"src/ic.cc").

[Vyacheslav Egorov (Chromium), 2011-12-20 15:45:14]

http://codereview.chromium.org/8971002/diff/3001/src/ia32/disasm-ia32.cc
File src/ia32/disasm-ia32.cc (right):

http://codereview.chromium.org/8971002/diff/3001/src/ia32/disasm-ia32.cc#newc...
src/ia32/disasm-ia32.cc:465: return 1;
what kind of warning do you get here? this return can never be reached (cases
end with return, and default captures everything else). seems like a bug in the
compiler.

http://codereview.chromium.org/8971002/diff/3001/src/ic.cc
File src/ic.cc (right):

http://codereview.chromium.org/8971002/diff/3001/src/ic.cc#newcode625
src/ic.cc:625: return Handle<Code>::null();
ditto. what does compiler say? this can never be reached both branches of the if
statement end with returns.

http://codereview.chromium.org/8971002/diff/3001/src/x64/disasm-x64.cc
File src/x64/disasm-x64.cc (right):

http://codereview.chromium.org/8971002/diff/3001/src/x64/disasm-x64.cc#newcod...
src/x64/disasm-x64.cc:551: return 1;
and the same question again (default is present and all cases terminate with
returns).

[Tobias Burnus, 2011-12-21 10:38:54]

Updated patch: I only left in a patch for the clearly wrong src/debug.cc.

For the others, I concur that the extra return is superfluous as the warning is
bogus. Thanks for the careful patch reading. (Actually, I also stumbled over it
yesterday - albeit only after requesting the review.)

I have now filled a GCC PR for the bogus warning
  warning: control reaches end of non-void function [-Wreturn-type]
Cf. http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51647

[Vyacheslav Egorov (Chromium), 2011-12-21 12:26:30]

LGTM, I will land.

Thanks for dealing with GCC issues!