Bug 56865 – Webkit warning: converting to 'unsigned int' from 'double'

• Component: New Bugs
• Status: RESOLVED FIXED

[George Guo]

When I tried to compile webkit on RedHat Linux with this following command: 
    Tools/Scripts/build-webkit --qt –debug

I got the following error 

../../Source/JavaScriptCore/profiler -I../../../Source/JavaScriptCore/runtime -I../../../Source/JavaScriptCore/wtf -I../../../Source/JavaScriptCore/wtf/gobject -I/nokia/bs_nmp/groups/s40_builds/naguo/webkit2/trunk/Source/JavaScriptCore/wtf/symbian -I../../../Source/JavaScriptCore/wtf/unicode -I../../../Source/JavaScriptCore/yarr -I../../../Source/JavaScriptCore/API -I../../../Source/JavaScriptCore/ForwardingHeaders -Igenerated -I../../../Source -I../include/QtWebKit -I../include -I. -I../../../Source/JavaScriptCore -I. -o obj/debug/TimeoutChecker.o ../../../Source/JavaScriptCore/runtime/TimeoutChecker.cpp
cc1plus: warnings being treated as errors
../../../Source/JavaScriptCore/runtime/TimeoutChecker.cpp: In function 'unsigned int JSC::getCPUTime()':
../../../Source/JavaScriptCore/runtime/TimeoutChecker.cpp:104: warning: converting to 'unsigned int' from 'double'
make[1]: *** [obj/debug/TimeoutChecker.o] Error 1

[Laszlo Gombos]

This is not just for Qt, removing the Qt prefix.

[George Guo]

Created attachment 86553 [details]
A explicit cast in C++ style for review

[Darin Adler]

Comment on attachment 86553 [details]
A explicit cast in C++ style for review 

This casts to unsigned before multiplying by 1000, so now the function will only return multiples of 1000!

[George Guo]

Created attachment 86611 [details]
This patch resolves the issue pointed out by Adler during the review

[Darin Adler]

Comment on attachment 86611 [details]
This patch resolves the issue pointed out by Adler during the review

View in context: https://bugs.webkit.org/attachment.cgi?id=86611&action=review

> Source/JavaScriptCore/runtime/TimeoutChecker.cpp:104
> +    return static_cast<unsigned> ((currentTime() - firstTime) * 1000);

There’s an extra space here after the "<". This is not the normal way we format.

I’m also concerned. We have many places where we assign from double to an integer. Is this really the only failure in WebKit of this type?

[George Guo]

ok, preparing a new patch to get rid of the extra space. Yes, with the build it did pass the build after this change.

[WebKit Commit Bot]

Comment on attachment 86611 [details]
This patch resolves the issue pointed out by Adler during the review

Clearing flags on attachment: 86611

Committed r81803: <http://trac.webkit.org/changeset/81803>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.