Release v1.95

[djzort]

Release v1.91

[Copilot]

Pull request overview

This release (v1.91) includes improvements to test reliability, security enhancements, and CI infrastructure updates for the DBD-Oracle Perl database driver.

Changes:

• Enhanced test portability by adding signal fallback mechanisms for systems without USR1/USR2 support
• Improved security by replacing unsafe sprintf calls with snprintf in C code
• Updated CI configuration to use Oracle Free instead of Oracle XE

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
t/92-segv-fork.t	Added documentation about SEGV issue, new thread count checking function, and reduced test job count for faster execution
t/92-segv-fork.pl	Reduced random sleep time range to speed up tests
t/91-segv-fork.t	Added signal portability with USR1/USR2 fallback to HUP/INT for OS compatibility
oci8.c	Security improvement: replaced sprintf with snprintf and increased buffer sizes
.github/workflows/action.yml	Updated CI to use Oracle Free instead of Oracle XE

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Typo in comment: "handedled" should be "handled".

Suggested change

	## that allows a SIGCHLD (any interrupt) to be handedled using a worker
	## that allows a SIGCHLD (any interrupt) to be handled using a worker

[Copilot]

Typo in comment: "eminent" should be "imminent". "Eminent" means distinguished or prominent, while "imminent" means about to happen, which is the intended meaning here.

Suggested change

	# Not important but an indication SEGV is eminent
	# Not important but an indication SEGV is imminent

[Copilot]

Typo in comment: "wont't" should be "won't" (incorrect apostrophe placement).

Suggested change

	# (wont't PASS if Perl is not built with threads support)
	# (won't PASS if Perl is not built with threads support)

[Copilot]

The comment states "1-3 seconds" but the code calculates int(rand(100000)) + 2000000, which produces values between 2,000,000 and 2,099,999 microseconds. This corresponds to approximately 2.0-2.1 seconds, not 1-3 seconds. The comment should be corrected to accurately reflect the actual range.

Suggested change

	my $usleep = int(rand(100000)) + 2000000; # 1-3 seconds (to speed up test!)
	my $usleep = int(rand(100000)) + 2000000; # about 2.0-2.1 seconds (to speed up test!)

[djzort]

I have tagged v1.91_3 and will shortly arrange to have it pushed to the cpan

[djzort]

1.91_3 is now on the cpan https://metacpan.org/release/ZARQUON/DBD-Oracle-1.91_3

[alexm]

t/92-segv-fork.t hangs on Debian unstable sometimes, and when it doesn't it fails:

$ env ORACLE_USERID=kermit/foobar ORACLE_DSN='dbi:Oracle://localhost:1521/FREEPDB1' prove --blib t/92-segv-fork.t -v
t/92-segv-fork.t .. 
# + --------------------------------------------- +
#  ORACLE - READY
# + --------------------------------------------- +
ok 1 - Expected thread count=1
ok 2 - ALTER SESSION SET NLS_DATE_FORMAT
ok 3 - ALTER SESSION SET NLS_TIMESTAMP_FORMAT
ok 4 - ALTER SESSION SET NLS_TIMESTAMP_TZ_FORMAT
# [
#   [
#     '2026-02-22T22:11:06Z'
#   ]
# ]
# + --------------------------------------------- +
#  QUEUE - BASICS
# + --------------------------------------------- +
ok 5 - Queue depth
ok 6 - Queue size
ok 7 - Queue running
ok 8 - Queue is idle
ok 9 - Queue is not busy
ok 10 - Queue has slots
ok 11 - Queue is not full
ok 12 - Enqueue 1
ok 13 - Queue size
ok 14 - Enqueue 2
ok 15 - Queue size
ok 16 - Queue running
ok 17 - Dequeue 1
ok 18 - Queue size
ok 19 - Dequeue 2
ok 20 - Queue size
ok 21 - Queue is idle
ok 22 - Queue is not busy
ok 23 - Queue has slots
# + --------------------------------------------- +
#  FORK - SEGV
# + --------------------------------------------- +
ok 24 - Queue depth
ok 25 - Queue size
ok 26 - Queue running
ok 27 - Queue is idle
ok 28 - Queue is not busy
ok 29 - Queue has slots
ok 30 - Queue is not full
ok 31 - Enqueue JOB-001
ok 32 - Enqueue JOB-002
ok 33 - Enqueue JOB-003
ok 34 - Enqueue JOB-004
ok 35 - Enqueue JOB-005
ok 36 - Enqueue JOB-006
ok 37 - Enqueue JOB-007
ok 38 - Enqueue JOB-008
ok 39 - Enqueue JOB-009
ok 40 - Enqueue JOB-010
ok 41 - Enqueue JOB-011
ok 42 - Enqueue JOB-012
ok 43 - Enqueue JOB-013
ok 44 - Enqueue JOB-014
ok 45 - Enqueue JOB-015
ok 46 - Enqueue JOB-016
ok 47 - Enqueue JOB-017
ok 48 - Enqueue JOB-018
ok 49 - Enqueue JOB-019
ok 50 - Enqueue JOB-020
ok 51 - Enqueue JOB-021
ok 52 - Enqueue JOB-022
ok 53 - Enqueue JOB-023
ok 54 - Enqueue JOB-024
ok 55 - Enqueue JOB-025
ok 56 - Enqueue JOB-026
ok 57 - Enqueue JOB-027
ok 58 - Enqueue JOB-028
ok 59 - Enqueue JOB-029
ok 60 - Enqueue JOB-030
ok 61 - Enqueue JOB-031
ok 62 - Enqueue JOB-032
ok 63 - Enqueue JOB-033
ok 64 - Enqueue JOB-034
ok 65 - Enqueue JOB-035
ok 66 - Enqueue JOB-036
ok 67 - Enqueue JOB-037
ok 68 - Enqueue JOB-038
ok 69 - Enqueue JOB-039
ok 70 - Enqueue JOB-040
ok 71 - Queue size
ok 72 - Queue running - zero
ok 73 - Start one child ->> 1
ok 74 - Queue size verified
ok 75 - Start one child ->> 2
ok 76 - Queue size verified
ok 77 - queue->run - start -DEPTH children
ok 78 - Queue running - 8 children started
ok 79 - Queue is full
# Child 27529 finished with code 0
{
  'CODE' => 0,
  '-PID' => 27529,
  '-ERR' => [],
  '-OUT' => [],
  '-JOB' => 'JOB-001'
}
# Child 27532 finished with code 0
{
  '-OUT' => [],
  '-ERR' => [],
  '-JOB' => 'JOB-004',
  '-PID' => 27532,
  'CODE' => 0
}
# Child 27535 finished with code 0
{
  '-PID' => 27535,
  'CODE' => 0,
  '-ERR' => [],
  '-OUT' => [],
  '-JOB' => 'JOB-007'
}
# Child 27534 finished with code 0
{
  '-OUT' => [],
  '-ERR' => [],
  '-JOB' => 'JOB-006',
  '-PID' => 27534,
  'CODE' => 0
}
# Child 27530 finished with code 0
{
  '-PID' => 27530,
  'CODE' => 0,
  '-JOB' => 'JOB-002',
  '-ERR' => [],
  '-OUT' => []
}
# Child 27533 finished with code 0
{
  '-PID' => 27533,
  'CODE' => 0,
  '-JOB' => 'JOB-005',
  '-ERR' => [],
  '-OUT' => []
}
# Child 27531 finished with code 0
{
  '-OUT' => [],
  '-ERR' => [],
  '-JOB' => 'JOB-003',
  'CODE' => 0,
  '-PID' => 27531
}
# Child 27536 finished with code 0
{
  '-ERR' => [],
  '-JOB' => 'JOB-008',
  '-OUT' => [],
  '-PID' => 27536,
  'CODE' => 0
}
# Child 27561 finished with code 0
{
  '-ERR' => [],
  '-OUT' => [],
  '-JOB' => 'JOB-009',
  'CODE' => 0,
  '-PID' => 27561
}
# Child 27562 finished with code 0
{
  '-OUT' => [],
  '-ERR' => [],
  '-JOB' => 'JOB-010',
  'CODE' => 0,
  '-PID' => 27562
}
# Child 27566 finished with code 0
{
  '-ERR' => [],
  '-OUT' => [],
  '-JOB' => 'JOB-014',
  'CODE' => 0,
  '-PID' => 27566
}
# Child 27565 finished with code 0
{
  '-PID' => 27565,
  'CODE' => 0,
  '-OUT' => [],
  '-ERR' => [],
  '-JOB' => 'JOB-013'
}
# Child 27563 finished with code 0
{
  '-ERR' => [],
  '-OUT' => [],
  '-JOB' => 'JOB-011',
  '-PID' => 27563,
  'CODE' => 0
}
# Child 27567 finished with code 0
{
  'CODE' => 0,
  '-PID' => 27567,
  '-ERR' => [],
  '-OUT' => [],
  '-JOB' => 'JOB-015'
}
# Child 27564 finished with code 0
{
  '-ERR' => [],
  '-JOB' => 'JOB-012',
  '-OUT' => [],
  'CODE' => 0,
  '-PID' => 27564
}
# Child 27568 finished with code 0
{
  'CODE' => 0,
  '-PID' => 27568,
  '-JOB' => 'JOB-016',
  '-ERR' => [],
  '-OUT' => []
}
# Child 27593 finished with code 0
{
  '-JOB' => 'JOB-017',
  '-ERR' => [],
  '-OUT' => [],
  '-PID' => 27593,
  'CODE' => 0
}
# Child 27595 finished with code 0
{
  'CODE' => 0,
  '-PID' => 27595,
  '-ERR' => [],
  '-JOB' => 'JOB-019',
  '-OUT' => []
}
# Child 27594 finished with code 0
{
  'CODE' => 0,
  '-PID' => 27594,
  '-ERR' => [],
  '-JOB' => 'JOB-018',
  '-OUT' => []
}
# Child 27600 finished with code 0
{
  '-OUT' => [],
  '-ERR' => [],
  '-JOB' => 'JOB-022',
  'CODE' => 0,
  '-PID' => 27600
}
# Child 27599 finished with code 0
{
  '-JOB' => 'JOB-021',
  '-ERR' => [],
  '-OUT' => [],
  'CODE' => 0,
  '-PID' => 27599
}
ORA-24550: signal received: [si_signo=11] [si_errno=0] [si_code=1] [si_int=0] [si_ptr=(nil)] [si_addr=0x4ac]
kpedbg_dmp_stack()+415<-kpeDbgCrash()+220<-kpeDbgSignalHandler()+685<-skgesig_sigactionHandler()+295<-__sighandler()<-Perl_csighandler3()+56<-__sighandler()<-sem_trywait()+94?<-__libc_alloca_cutoff()+260?<-__nptl_death_event()+140?<-pthread_cond_timedwait()+333<-sltspctimewait()+147<-kpucpincrtime()+113<-pthread_condattr_setpshared()+1465?
All 79 subtests passed 

Test Summary Report
-------------------
t/92-segv-fork.t (Wstat: 11 (Signal: SEGV) Tests: 79 Failed: 0)
  Non-zero wait status: 11
  Parse errors: No plan found in TAP output
Files=1, Tests=79,  7 wallclock secs ( 0.05 usr  0.00 sys +  3.77 cusr  0.87 csys =  4.69 CPU)
Result: FAIL

[djzort]

I have tagged 1.91_5 and its being released to the cpan. Lots of changes in that.

I suspect it will be the last.

[djzort]

To give it some distance lets jump to 1.95

1.90->1.91 implies less distance between releases than there is.

[djzort]

I think i will ship this tonight.