Bug 59482 – Unused but set variable warning in MacroAssemberX86_64

• Component: JavaScriptCore
• Status: RESOLVED FIXED

[Xan Lopez]

../../Source/JavaScriptCore/assembler/MacroAssemblerX86_64.h: In member function ‘JSC::AbstractMacroAssembler<JSC::X86Assembler>::Call JSC::MacroAssemblerX86_64::call()’:
../../Source/JavaScriptCore/assembler/MacroAssemblerX86_64.h:118:22: warning: variable ‘label’ set but not used [-Wunused-but-set-variable]
../../Source/JavaScriptCore/assembler/MacroAssemblerX86_64.h: In member function ‘JSC::AbstractMacroAssembler<JSC::X86Assembler>::Call JSC::MacroAssemblerX86_64::tailRecursiveCall()’:
../../Source/JavaScriptCore/assembler/MacroAssemblerX86_64.h:126:22: warning: variable ‘label’ set but not used [-Wunused-but-set-variable]
../../Source/JavaScriptCore/assembler/MacroAssemblerX86_64.h: In member function ‘JSC::AbstractMacroAssembler<JSC::X86Assembler>::Call JSC::MacroAssemblerX86_64::makeTailRecursiveCall(JSC::AbstractMacroAssembler<JSC::X86Assembler>::Jump)’:
../../Source/JavaScriptCore/assembler/MacroAssemblerX86_64.h:135:22: warning: variable ‘label’ set but not used [-Wunused-but-set-variable]
  CXX    Source/JavaScriptCore/runtime/libJavaScriptCore_la-JSLock.lo
  CXX    Source/JavaScriptCore/runtime/libJavaScriptCore_la-JSNotAnObject.lo
../../Source/JavaScriptCore/runtime/JSLock.cpp: In static member function ‘static void JSC::JSLock::lock(JSC::JSLockBehavior)’:
../../Source/JavaScriptCore/runtime/JSLock.cpp:91:13: warning: variable ‘result’ set but not used [-Wunused-but-set-variable]
../../Source/JavaScriptCore/runtime/JSLock.cpp: In static member function ‘static void JSC::JSLock::unlock(JSC::JSLockBehavior)’:
../../Source/JavaScriptCore/runtime/JSLock.cpp:111:13: warning: variable ‘result’ set but not used [-Wunused-but-set-variable]


Trivial patch coming.

[Xan Lopez]

Created attachment 91143 [details]
Patch

[Daniel Bates]

Comment on attachment 91143 [details]
Patch

I wish there was a better way we could do this. This looks sane to me.

[Xan Lopez]

Comment on attachment 91143 [details]
Patch

Landed as r84977.

[Xan Lopez]

Closing.

[WebKit Review Bot]

http://trac.webkit.org/changeset/84977 might have broken SnowLeopard Intel Release (Tests)
The following tests are not passing:
accessibility/anchor-linked-anonymous-block-crash.html
accessibility/aria-activedescendant-crash.html
accessibility/aria-checkbox-checked.html
accessibility/aria-checkbox-text.html
accessibility/aria-combobox.html
accessibility/aria-controls-with-tabs.html
accessibility/aria-describedby-on-input.html
accessibility/aria-disabled.html
accessibility/aria-help.html
accessibility/aria-hidden-update.html
accessibility/aria-hidden-with-elements.html
accessibility/aria-hidden.html
accessibility/aria-label.html
accessibility/aria-labelledby-on-input.html
accessibility/aria-labelledby-overrides-label.html
accessibility/aria-labelledby-stay-within.html
accessibility/aria-link-supports-press.html
accessibility/aria-list-and-listitem.html
accessibility/aria-option-role.html
accessibility/aria-presentational-role.html

[Daniel Bates]

This patch was rolled out in <http://trac.webkit.org/changeset/85000> (bug #59568).

[Xan Lopez]

Created attachment 91234 [details]
Patch

[Xan Lopez]

OK, that was obviously broken. Unfortunately we even need an even slightly uglier patch here... I wonder if there's a way of flagging a variable as unused, we could do that for release builds here. Maybe just (void)var;, like ASSERT_UNUSED, but without using the macro because we need it in NDEBUG.

[Eric Seidel (no email)]

Comment on attachment 91234 [details]
Patch

We have an UNUSED() macro for this, no?

[Alexis Menard (darktears)]

Created attachment 92095 [details]
Patch

[Eric Seidel (no email)]

Comment on attachment 92095 [details]
Patch

OK.

[WebKit Commit Bot]

Comment on attachment 92095 [details]
Patch

Clearing flags on attachment: 92095

Committed r85718: <http://trac.webkit.org/changeset/85718>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[WebKit Commit Bot]

The commit-queue encountered the following flaky tests while processing attachment 92095 [details]:

http/tests/xmlhttprequest/cross-origin-no-authorization.html bug 33357 (author: [e-mail])
The commit-queue is continuing to process your patch.