On 4/15/13 4:35 PM, Leif Hedstrom wrote:
Hi all,
I've prepared a release for v3.3.2, which has quite a few improvements and
bug fixes over 3.3.0. Please see the CHANGES for more details. It's
important we get as much testing as possible for these dev releases for
the next few months, in preparation for v3.4.0. There are several known
bugs already, which are (now) scheduled for v3.3.3, v3.3.4 and v3.3.5
accordingly. Help triaging, commenting and closing those bugs are also
much needed.
This vote passes with 9 +1 votes (8 binding) and 1 -1 vote. For the -1 vote,
we have fixed the issue with certain (older compilers), it's committed on
master in 5b44534f7bdc2338c23c5cf89ddf86d6de3ae829. I'm including the tiny
patch below as well. The v.3.3.2 release has been pushed to the dist server,
and I'll make an official announcement tomorrow.
Thanks to everyone for voting! v3.3.3 in a month, keep up the great work
towards v3.4.0 stable.
-- Leif
diff --git a/lib/records/RecUtils.cc b/lib/records/RecUtils.cc
index d04fce7..3680fee 100644
--- a/lib/records/RecUtils.cc
+++ b/lib/records/RecUtils.cc
@@ -76,7 +76,7 @@ RecDataSetMax(RecDataT type, RecData * data)
#endif
case RECD_INT:
case RECD_COUNTER:
- data->rec_int = 0x7fffffffffffffff;
+ data->rec_int = INT64_MAX; // Assumes rec_int is int64_t, which it
currently is
break;
#if defined(STAT_PROCESSOR)
case RECD_CONST:
@@ -98,7 +98,7 @@ RecDataSetMin(RecDataT type, RecData * data)
#endif
case RECD_INT:
case RECD_COUNTER:
- data->rec_int = 0x8000000000000000;
+ data->rec_int = INT64_MIN; // Assumes rec_int is int64_t, which it
currently is
break;
#if defined(STAT_PROCESSOR)
case RECD_CONST:
