t: Skip if pgrep is not installed

[perlpunk]

Otherwise the test might fail, and sometimes we don't even see an error message with the reason, like observed in OBS or probably here on cpantesters:
http://www.cpantesters.org/cpan/report/47446f5e-103b-11f0-98b0-b3c3213a625c http://www.cpantesters.org/cpan/report/57cba9ba-1017-11f0-98b0-b3c3213a625c

[okurz]

Oh cool. And now we have a CI failure which I have never seen before: https://github.com/openSUSE/Mojo-IOLoop-ReadWriteProcess/actions/runs/14251136067/job/39943920707?pr=78#step:6:88

[perlpunk]

We might want to use https://metacpan.org/pod/Proc::ProcessTable to replace pgrep

[Martchus]

If that supports process groups.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 97.89%. Comparing base (73c1df6) to head (a91dae8).
Report is 2 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master      #78      +/-   ##
==========================================
+ Coverage   97.78%   97.89%   +0.11%     
==========================================
  Files          27       27              
  Lines         903      903              
==========================================
+ Hits          883      884       +1     
+ Misses         20       19       -1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[perlpunk]

For the record:
https://github.com/openSUSE/Mojo-IOLoop-ReadWriteProcess/actions/runs/14251136067/job/39943920707

    #   Failed test 'There are 20 pids and the start word (21)'
    #   at t/13_shared.t line 203.
    #          got: '20'
    #     expected: '21'
    # Looks like you failed 1 test of 3.

#   Failed test 'concurrent memory read/write'
#   at t/13_shared.t line 208.
# {
    #   '7679' => 1,
    #   '7680' => 1,
    #   '7681' => 1,
    #   '7682' => 1,
    #   '7683' => 1,
    #   '7684' => 1,
    #   '7685' => 1,
    #   '7686' => 1,
    #   '7687' => 1,
    #   '7688' => 1,
    #   '7689' => 1,
    #   '7691' => 1,
    #   '7694' => 1,
    #   '7696' => 1,
    #   '7697' => 1,
    #   '7698' => 1,
    #   '7700' => 1,
    #   '7702' => 1,
    #   '7706' => 1,
    #   '7707' => 1
    # }
    # {
    #   '7715' => 1,
    #   '7716' => 1,
    #   '7717' => 1,
    #   '7718' => 1,
    #   '7719' => 1,
    #   '7720' => 1,
    #   '7721' => 1,
    #   '7722' => 1,
    #   '7723' => 1,
    #   '7724' => 1,
    #   '7725' => 1,
    #   '7727' => 1,
    #   '7730' => 1,
    #   '7731' => 1,
    #   '7732' => 1,
    #   '7733' => 1,
    #   '7734' => 1,
    #   '7735' => 1,
    #   '7738' => 1,
    #   '7739' => 1
    # }
# Looks like you failed 1 test of 6.
t/13_shared.t ............ 
Dubious, test returned 1 (wstat 256, 0x100)

[okurz]

We might want to use https://metacpan.org/pod/Proc::ProcessTable to replace pgrep

I guess that would still mean needing to read multiple lines from that table to find out all processes in a group. Not really better. PR is ok as is