I've noticed that on my local CI test platform, one of my tests is failing and I'm at a loss as to why. It's intermittent, but it fails far more often than not.
Here's the TAP output up until the failure is produced:
pi@pi-test:~/repos/rpi-wiringpi $ prove -v t/75-serial.t
t/75-serial.t ..
ok 1 - An object of class 'RPi::Serial' isa 'RPi::Serial'
ok 2 - putc() and getc() ok
not ok 3 - puts() and gets() ok
# Failed test 'puts() and gets() ok'
# at t/75-serial.t line 32.
# got: 'hello, world!'
# expected: 'hello, world!'
Notice how the got and expected are identical. What am I missing here?
The test is writing a string to, and then reading the string back from a serial interface. Here's the test code if it helps:
use strict;
use warnings;
use lib 't/';
use RPiTest qw(check_pin_status);
use RPi::WiringPi;
use RPi::Const qw(:all);
use Test::More;
my $mod = 'RPi::WiringPi';
if (! $ENV{RPI_SERIAL}){
plan skip_all => "RPI_SERIAL environment variable not set\n";
}
if (! $ENV{PI_BOARD}){
$ENV{NO_BOARD} = 1;
plan skip_all => "Not on a Pi board\n";
}
my $pi = $mod->new;
my $s = $pi->serial("/dev/ttyS0", 115200);
isa_ok $s, 'RPi::Serial';
$s->putc(254);
is $s->getc, 254, "putc() and getc() ok";
$s->puts("hello, world!");
# FAILING TEST IS BELOW
is $s->gets(13), "hello, world!", "puts() and gets() ok";
$pi->cleanup;
check_pin_status();
done_testing();