392074 - Code still refers to JSTRACE_ATOM

Status: RESOLVED FIXED

(Core :: JavaScript Engine, defect)

Description

[Igor Bukanov]

[This is a spin-off of bug 386265 comment 41]

JS_IS_VALID_TRACE_KIND still refers to removed JSATOM_TRACE with !JS_HAS_XML_SUPPORT.

Comment 1

[Igor Bukanov]

Attachment: v1

The patch fixes JS_IS_VALID_TRACE_KIND macro and adds a static assert to prevent this kind of bugs to happen again.

Comment 2

[Igor Bukanov]

I checked in the patch from comment 1 to the trunk:

http://bonsai.mozilla.org/cvsquery.cgi?module=PhoenixTinderbox&branch=HEAD&cvsroot=%2Fcvsroot&date=explicit&mindate=1187036463&maxdate=1187036702&who=igor%25mir2.org

Comment 3

[Johnny Stenback (:jst)]

Based on Tinderbox it looks like this could have caused the Win32 orangeness. Any thoughts?

Comment 4

[Igor Bukanov]

(In reply to comment #3)
> Based on Tinderbox it looks like this could have caused the Win32 orangeness.

Sorry for not waiting until all the boxes finish the tests.

My assumption was that adding an extern declaration of a C function that was never accessed from the code (this is what JS_STATIC_ASSERT expands into) may only affect the compilation. So I just waited until one of Linux/Max/Win boxes finished.

The next time I should to take such shortcuts and wait for all tests to finish for the reason related to the the last orange. If orange happens during my commit, I better be available to help to figure out the reason for it like explaining the nature of the check in.