On 5/7/13 4:18 AM, Daniel Gruno wrote:
On 05/07/2013 05:37 AM, James Peach wrote:
On May 6, 2013, at 3:30 PM, Igor Galić <i.ga...@brainsware.org> wrote:


----- Original Message -----
Updated Branches:
  refs/heads/master c4815b077 -> da413c514


Try to fix Solaris wclear macro build breakage


Project: http://git-wip-us.apache.org/repos/asf/trafficserver/repo
Commit:
http://git-wip-us.apache.org/repos/asf/trafficserver/commit/da413c51
Tree:
http://git-wip-us.apache.org/repos/asf/trafficserver/tree/da413c51
Diff:
http://git-wip-us.apache.org/repos/asf/trafficserver/diff/da413c51

Branch: refs/heads/master
Commit: da413c5140e33d52866f64757c374257116548ab
Parents: c4815b0
Author: James Peach <jpe...@apache.org>
Authored: Mon May 6 14:56:23 2013 -0700
Committer: James Peach <jpe...@apache.org>
Committed: Mon May 6 14:56:23 2013 -0700

----------------------------------------------------------------------
tools/tstop/tstop.cc |    5 +++++
1 files changed, 5 insertions(+), 0 deletions(-)
----------------------------------------------------------------------


http://git-wip-us.apache.org/repos/asf/trafficserver/blob/da413c51/tools/tstop/tstop.cc
----------------------------------------------------------------------
diff --git a/tools/tstop/tstop.cc b/tools/tstop/tstop.cc
index 7828c61..0626764 100644
--- a/tools/tstop/tstop.cc
+++ b/tools/tstop/tstop.cc
@@ -46,6 +46,11 @@
#  error "SysV or X/Open-compatible Curses header file required"
#endif

+// Solaris curses #defines wclear, which breaks
std::string::wclear().
+#ifndef wclear
+#undef wclear
+#endif
+
#include "stats.h"

using namespace std;


should(n't) this be in ink_defs.h?
Nah it's too ugly for that ... and of course I spelled "ifdef" wrong :(

J

Isn't the real problem that curses.h defines clear as wclear??
std::string has no such thing as wclear as far as I can tell, but it
does have a clear() function, which curses.h steals:

curses.h:#define clear()                   wclear(stdscr)


I could have sworn I tried the #undef clear , but it broke things in curses later. So you might have to "redo" the #define after calling the std::string clear? But, I might also have been smoking crack.

-- Leif

Reply via email to