Bug 306334 – Build failure: expected unqualified-id before numeric constant

• Component: New Bugs
• Status: RESOLVED FIXED

[Mike Gorse]

I am trying to get WebKitGTK 2.51.90 to build on SLE and am getting these errors with a couple of our older codestreams:

In file included from /usr/include/X11/Xlib.h:44,
                 from /usr/include/EGL/eglplatform.h:134,
                 from /usr/include/epoxy/egl_generated.h:11,
                 from /usr/include/epoxy/egl.h:46,
                 from ../Source/WebKit/WebProcess/WebPage/CoordinatedGraphics/ThreadedCompositor.cpp:50,
                 from DerivedSources/WebKit/unified-sources/UnifiedSource-54928a2b-32.cpp:7:
WebCore/PrivateHeaders/WebCore/FontAttributes.h:42:52: error: expected unqualified-id before numeric constant
   42 |     Style::ListStyleType styleType { CSS::Keyword::None { } };
      |                                                    ^~~~
...

In file included from /usr/include/X11/Xlib.h:44,
                 from /usr/include/EGL/eglplatform.h:134,
                 from /usr/include/epoxy/egl_generated.h:11,
                 from /usr/include/epoxy/egl.h:46,
                 from ../Source/WebCore/platform/graphics/skia/PlatformDisplaySkia.cpp:42,
                 from WebCore/DerivedSources/unified-sources/UnifiedSource-3c72abbe-66.cpp:3:
WebCore/PrivateHeaders/WebCore/SharedMemory.h:57:27: error: expected identifier before numeric constant
   57 | enum class MemoryLedger { None, Default, Network, Media, Graphics, Neural };
      |                           ^~~~

I get more errors like it in both cases, although, for the latter, some of them come from ShareableBitmapSkia.cpp.
I likely have a nonstandard build (I'm compiling with -DENABLE_MEDIA_STREAM=OFF, for instance), but these errors don't seem directly related to my changes or patches. Both sources build if I add @no-unify to them.

[Mike Gorse]

Pull request: https://github.com/WebKit/WebKit/pull/57333

[Michael Catanzaro]

Xlib.h is just horrible. Unfortunately using @no-unify to avoid it getting included is probably indeed the best solution here. That's nicer than renaming things.

[Radar WebKit Bug Importer]

<rdar://problem/169534529>

[EWS]

Committed 311526@main (470279182571): <https://commits.webkit.org/311526@main>

Reviewed commits have been landed. Closing PR #57333 and removing active labels.