Szelethus requested changes to this revision.
Szelethus added a comment.
This revision now requires changes to proceed.
Uhh, it's such a chore to work on these things -- You can really feel that the
preprocessor must've been the first thing implemented in clang. Unfortunately,
you really have to counterweight it's shortcomings with excessive amount of
comments. I can see an infinite loop and a lot of string comparisons, but
someone without having this particular test case as a context, it's very hard
to follow what's happening.
Could you please add some code examples in the comments too?
================
Comment at: clang/lib/StaticAnalyzer/Core/BugReporterVisitors.cpp:1997-1999
LocStart = BRC.getSourceManager().getImmediateMacroCallerLoc(LocStart);
+ if (!LocStart.isMacroID())
+ break;
----------------
> Gets the location of the immediate macro caller, one level up the stack
> toward the initial macro typed into the source.
Hmm, is there a guarantee that at the end of the stack is not a macro location?
================
Comment at: clang/test/Analysis/diagnostics/macros.cpp:53
+void testNestedNullSplitMacro(int i, int *p) {
+ nested_null_split(i); // expected-note {{Assuming 'i' is equal to
UINT32_MAX}}
+ // expected-note@-1 {{Taking false branch}}
----------------
Ah, that looks pretty :D
Repository:
rC Clang
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D59121/new/
https://reviews.llvm.org/D59121
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
