Bug 32939

style-queue ran check-webkit-style on attachment 45497 [details] without any errors.

Created attachment 45498 [details] WinMain function for WinCE WinCE doesn't support a main function, only a WinMain. This is a simple wrapper to call the usual main, which only needs to be linked additionaly. Maybe there is a better place for this file instead of os-win32? This file can be reused by other executables too, insted of only patching jsc.cpp

Attachment 45498 [details] did not pass style-queue: Failed to run "WebKitTools/Scripts/check-webkit-style" exit_code: 1 JavaScriptCore/os-win32/WinMain.cpp:22: Found other header before a header this file implements. Should be: config.h, primary header, blank line, and then alphabetically sorted. [build/include_order] [4] Total errors found: 1

Created attachment 45502 [details] Buildfix for MainThreadWin.cpp HWND_MESSAGE isn't supported on WinCE. See http://social.msdn.microsoft.com/forums/en-US/vssmartdevicesnative/thread/dadf671b-faa3-45f7-9b67-42c3b5959373

style-queue ran check-webkit-style on attachment 45502 [details] without any errors.

Created attachment 45504 [details] Some buildfixes for WebCore/platform/wince/ This are only the first (simple) fixes. Many of the other files don't compile too :-(

style-queue ran check-webkit-style on attachment 45504 [details] without any errors.

Comment on attachment 45498 [details] WinMain function for WinCE ok

Comment on attachment 45504 [details] Some buildfixes for WebCore/platform/wince/ ok... I like the previous patch's way of handling this slightly better.

Created attachment 45508 [details] Buildfix for RenderThemeWince.cpp

style-queue ran check-webkit-style on attachment 45508 [details] without any errors.

Created attachment 45509 [details] Buildfix for TextBreakIteratorWince.cpp

style-queue ran check-webkit-style on attachment 45509 [details] without any errors.

Comment on attachment 45497 [details] Buildfix for jsc.cpp Rejecting patch 45497 from commit-queue. Failed to run "['WebKitTools/Scripts/build-webkit']" exit_code: 1 Last 500 characters of output: uild/Release/testapi.build/Script-14D857B50A469C100032146C.sh === BUILDING AGGREGATE TARGET All OF PROJECT JavaScriptCore WITH CONFIGURATION Release === Checking Dependencies... ** BUILD FAILED ** The following build commands failed: jsc: Distributed-CompileC /Users/eseidel/Projects/CommitQueueSVN/WebKitBuild/JavaScriptCore.build/Release/jsc.build/Objects-normal/i386/jsc.o /Users/eseidel/Projects/CommitQueueSVN/JavaScriptCore/jsc.cpp normal i386 c++ com.apple.compilers.gcc.4_2 (1 failure) Full output: http://webkit-commit-queue.appspot.com/results/149141

Comment on attachment 45498 [details] WinMain function for WinCE Clearing flags on attachment: 45498 Committed r52570: <http://trac.webkit.org/changeset/52570>

Comment on attachment 45508 [details] Buildfix for RenderThemeWince.cpp LGTM.

Comment on attachment 45509 [details] Buildfix for TextBreakIteratorWince.cpp LGTM.

Each of these fixes really should have had its own bug. Or at least the ChangeLogs should have all referenced bugs, even if they were all the same bug.

Comment on attachment 45502 [details] Buildfix for MainThreadWin.cpp Clearing flags on attachment: 45502 Committed r52571: <http://trac.webkit.org/changeset/52571>

(In reply to comment #14) > (From update of attachment 45497 [details]) > Rejecting patch 45497 from commit-queue. Do you have more infos about the error? (In reply to comment #18) > Each of these fixes really should have had its own bug. Or at least the > ChangeLogs should have all referenced bugs, even if they were all the same bug. I started only with getting jsc running, but then i found more and more errors... I also have many too for posting it here. Should i realy open a bug for every single file or should i pack them into one patch? Sidenote: #32169 has a pending review, but no comments on it.

Comment on attachment 45504 [details] Some buildfixes for WebCore/platform/wince/ Clearing flags on attachment: 45504 Committed r52572: <http://trac.webkit.org/changeset/52572>

Comment on attachment 45508 [details] Buildfix for RenderThemeWince.cpp Clearing flags on attachment: 45508 Committed r52573: <http://trac.webkit.org/changeset/52573>

Comment on attachment 45509 [details] Buildfix for TextBreakIteratorWince.cpp Clearing flags on attachment: 45509 Committed r52574: <http://trac.webkit.org/changeset/52574>

Not every patch needs a new bug, but it's often helpful. In this case, the only problem is that all the ChangeLogs should have had a bug number in them. Neither is a big deal. Thanks again for the patches!

Comment on attachment 45497 [details] Buildfix for jsc.cpp Marking this patch r- since it would break the Mac build (at least according to the commit-queue output).

(In reply to comment #25) > (From update of attachment 45497 [details]) > Marking this patch r- since it would break the Mac build (at least according to > the commit-queue output). Can you please provide me the errormessage of the compiler. I don't know where the problem might be and don't have a Mac to compile it myself. Does Mac require no statement after an exit????

The bot already provided the error message in comment 14: https://bugs.webkit.org/show_bug.cgi?id=32939#c14

A little searching through the full output linked to from comment 14 found this: cc1plus: warnings being treated as errors /Users/eseidel/Projects/CommitQueueSVN/JavaScriptCore/jsc.cpp: In function ‘JSC::JSValue functionQuit(JSC::ExecState*, JSC::JSObject*, JSC::JSValue, const JSC::ArgList&)’: /Users/eseidel/Projects/CommitQueueSVN/JavaScriptCore/jsc.cpp:297: warning: function declared ‘noreturn’ has a ‘return’ statement distcc[4729] ERROR: compile /Users/eseidel/Projects/CommitQueueSVN/JavaScriptCore/jsc.cpp on localhost failed

(In reply to comment #27) > The bot already provided the error message in comment 14: > https://bugs.webkit.org/show_bug.cgi?id=32939#c14 Ok, i didn't found it between the many setenv, when i looked at it :-( Do you see a way to improve the output? A HTML-output with anchor tags for every target and a dictionary? With patch from bug 33056 i got a better error message. So the NO_RETUN must not be used, but I think that it won't compile with GCC afterwards? Can i try this without a mac?

Created attachment 45704 [details] Buildfix for jsc.cpp

style-queue ran check-webkit-style on attachment 45704 [details] without any errors.

Comment on attachment 45704 [details] Buildfix for jsc.cpp OK.

Created attachment 45930 [details] Buildfix for jsc.cpp The PLATFORM() macro changed to OS() in the meantime

style-queue ran check-webkit-style on attachment 45930 [details] without any errors.

Comment on attachment 45930 [details] Buildfix for jsc.cpp OK. I'm not sure why you moved --help to the end, but looks OK.

(In reply to comment #35) > I'm not sure why you moved --help to the end, but looks OK. There is no sorting and so all cases witch a continue at the end are together.

Comment on attachment 45930 [details] Buildfix for jsc.cpp Clearing flags on attachment: 45930 Committed r52824: <http://trac.webkit.org/changeset/52824>

All reviewed patches have been landed. Closing bug.