LLVM Bugzilla is read-only and represents the historical archive of all LLVM issues filled before November 26, 2021. Use github to submit LLVM bugs

Bug 23482 - revision 236900 appears to break compilation on Solaris
Summary: revision 236900 appears to break compilation on Solaris
Status: RESOLVED FIXED
Alias: None
Product: libraries
Classification: Unclassified
Component: Support Libraries (show other bugs)
Version: trunk
Hardware: PC Solaris
: P normal
Assignee: Renato Golin
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2015-05-11 13:02 PDT by Norm Jacobs
Modified: 2015-05-12 05:35 PDT (History)
2 users (show)

See Also:
Fixed By Commit(s):

Attachments
patch to workaround naming conflict between sys/time.h and TargetParser.h (479 bytes, patch)
2015-05-11 13:02 PDT, Norm Jacobs 		Details
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Norm Jacobs 2015-05-11 13:02:56 PDT 
Created attachment 14310 [details]
patch to workaround naming conflict between sys/time.h and TargetParser.h

Changset 2369000 appears to introduce a new enumeration into TargetParser.h
+  // Arch extension modifiers for CPUs.
+  enum ArchExtKind {
+    INVALID_ARCHEXT = 0,
+    CRC,
+    CRYPTO,
+    FP,
+    HWDIV,
+    MP,
+    SEC,
+    VIRT,
+    LAST_ARCHEXT
+  };

sys/time.h on Solaris (and possibly other systems) defines "SEC" as "1" using a cpp macro.  The result is that this fails to compile.

A temporary workaround seems to be to #undef SEC in Solaris.h, but a better answer might be to prefix the enumeration values so there were no conflicts.
Comment 1 Reid Kleckner 2015-05-11 14:32:45 PDT 
Generally our solution is to undef the macro. The enum is already namespaced inside ARM::, so adding a prefix would be pretty redundant.
Comment 2 Renato Golin 2015-05-11 14:53:13 PDT 
I'm ok with adding the prefix, if that's preferred. I just kept the same names for consistency with the previous code.

Thinking about macros, conflicts with things like CRC and CRYPTO might be more common than we think...
Comment 3 Renato Golin 2015-05-12 05:35:03 PDT 
Adding prefixes to the enums in r237112 / r237113.