AFAICS, there is no problem with e.msg.

diff --git a/src/simix/smx_deployment.c b/src/simix/smx_deployment.c
index dcd4497..f3536d7 100644 (file)
--- a/src/simix/smx_deployment.c
+++ b/src/simix/smx_deployment.c
@@ -126,8 +126,8 @@ void SIMIX_launch_application(const char *file)
          surf_parse_close();
          xbt_assert(!parse_status, "Parse error at %s:%d", file,surf_parse_lineno);
   } CATCH(e) {
          surf_parse_close();
          xbt_assert(!parse_status, "Parse error at %s:%d", file,surf_parse_lineno);
   } CATCH(e) {

-         xbt_die("Unrecoverable error at %s:%d: %s", file,surf_parse_lineno,
-                         __xbt_running_ctx_fetch()->exception.msg); //FIXME: that pimple is due to the fact that e.msg does not seem to be set on CATCH(e). The pimple should be removed when the bug is gone.
+         xbt_die("Unrecoverable error at %s:%d: %s",
+                  file, surf_parse_lineno, e.msg);

   }
 }
 
   }
 }