| Subject: | Multiple failed tests (segfaults) in 3.0007 on OpenBSD, pass after applying patch |
See bug #21962 for my environment (summary: OpenBSD 3.9, perl 5.8.6,
mysql 5.0.22).
With 3.0007, make test'ing has the following errors:
# make test
PERL_DL_NONLAZY=1 /usr/bin/perl "-MExtUtils::Command::MM" "-e"
"test_harness(0, 'blib/lib', 'blib/arch')" t/*.t
t/00base.............ok
t/10dsnlist..........ok
t/20createdrop.......ok
t/30insertfetch......ok
t/35limit............dubious
Test returned status 0 (wstat 139, 0x8b)
test program seems to have generated a core
t/35prepare..........ok
t/40bindparam........ok
t/40bindparam2.......ok
t/40blobs............ok
t/40listfields.......ok
t/40nulls............ok
t/40numrows..........ok
t/41bindparam........ok
t/41blobs_prepare....dubious
Test returned status 0 (wstat 139, 0x8b)
test program seems to have generated a core
t/42bindparam........ok
t/50chopblanks.......ok
t/50commit...........dubious
Test returned status 0 (wstat 139, 0x8b)
test program seems to have generated a core
t/60leaks............skipped
all skipped: $ENV{SLOW_TESTS} is not set or Proc::ProcessTable
not installed
t/dbdadmin...........ok
t/insertid...........dubious
Test returned status 0 (wstat 139, 0x8b)
test program seems to have generated a core
t/param_values.......ok
t/prepare_noerror....skipped
all skipped: test - will only run with MySQL 5.1 and above.
t/texecute...........dubious
Test returned status 0 (wstat 139, 0x8b)
test program seems to have generated a core
Failed Test Stat Wstat Total Fail Failed List of Failed
-------------------------------------------------------------------------------
t/35limit.t 0 139 ?? ?? % ??
t/41blobs_prepare.t 0 139 ?? ?? % ??
t/50commit.t 0 139 ?? ?? % ??
t/insertid.t 0 139 ?? ?? % ??
t/texecute.t 0 139 ?? ?? % ??
2 tests skipped.
Failed 5/23 test scripts, 78.26% okay. 0/252 subtests failed, 100.00% okay.
*** Error code 2
Stop in /root/.cpan/build/DBD-mysql-3.0007 (line 870 of Makefile).
When I make the following changes:
# diff -u dbdimp.c.orig dbdimp.c.mod
--- dbdimp.c.orig Fri Sep 8 16:25:59 2006
+++ dbdimp.c.mod Sun Oct 8 17:08:03 2006
@@ -2277,11 +2277,12 @@
imp_sth->has_been_bound= 0;
/* Initialize ph variables with NULL values */
- for (bind= imp_sth->bind,
+ for (i = 0,
+ bind= imp_sth->bind,
fbind= imp_sth->fbind,
bind_end= bind+DBIc_NUM_PARAMS(imp_sth);
bind < bind_end ;
- bind++, fbind++ )
+ bind++, fbind++, i++ )
{
/*
if this statement has a result set, field types will be
@@ -2385,7 +2386,7 @@
&slen,
params,
num_params,
- imp_dbh->bind_type_guessing);
+ bind_type_guessing);
my_ulonglong rows= 0;
Make test passes:
# make test
PERL_DL_NONLAZY=1 /usr/bin/perl "-MExtUtils::Command::MM" "-e"
"test_harness(0, 'blib/lib', 'blib/arch')" t/*.t
t/00base.............ok
t/10dsnlist..........ok
t/20createdrop.......ok
t/30insertfetch......ok
t/35limit............ok
t/35prepare..........ok
t/40bindparam........ok
t/40bindparam2.......ok
t/40blobs............ok
t/40listfields.......ok
t/40nulls............ok
t/40numrows..........ok
t/41bindparam........ok
t/41blobs_prepare....ok
t/42bindparam........ok
t/50chopblanks.......ok
t/50commit...........ok
t/60leaks............skipped
all skipped: $ENV{SLOW_TESTS} is not set or Proc::ProcessTable
not installed
t/dbdadmin...........ok
t/insertid...........ok
t/param_values.......ok
t/prepare_noerror....skipped
all skipped: test - will only run with MySQL 5.1 and above.
t/texecute...........ok
All tests successful, 2 tests skipped.
Files=23, Tests=440, 23 wallclock secs ( 2.08 cusr + 0.28 csys = 2.36 CPU)
The change from "imp_dbh->bind_type_guessing" to "bind_type_guessing"
has already been made in 3.0007_1 and 3.0007_2. The change to the for
loop has not been made and I am not sure that it is entirely correct
(what is "i" supposed to be?; see my comments about this in a related
bug #21962) but it does stop the segfaults.