cmake: Bump required flux-core version

[trws]

problem: recent PRs have added things that require more recent flux versions and not added any check for the version being found

solution: add the required version into the pkg_check_modules call in cmake/FindFluxCore.cmake

Related to #1390

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
⚠️ Please upload report for BASE (master@5c04074). Learn more about missing BASE report.
⚠️ Report is 1 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff            @@
##             master   #1391   +/-   ##
========================================
  Coverage          ?   77.2%           
========================================
  Files             ?     114           
  Lines             ?   16320           
  Branches          ?       0           
========================================
  Hits              ?   12609           
  Misses            ?    3711           
  Partials          ?       0           

🚀 New features to boost your workflow:

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

[grondo]

Hm, I didn't think the intent of any recent PR was to bump the required version of flux-core (yet). Do we know what exactly caused the incompatibility? It might be nice to fix that and allow flux-sched to still be built alongside the deployed version of core.

[trws]

It may actually pass, I misread how the flags were being added in 1390, if so this version should change, but we should be tracking a working target version this way so that users get an appropriate message when trying to use one that's too old.

[garlick]

#1390 just sets a broker attribute in some tests. That broker attribute doesn't exist in previous releases but setting an unknown attribute isn't treated as an error, for better or worse.

[grondo]

I am getting errors running make check on a TOSS 4 system with flux-core v0.76.0. However, I went back to the latest flux-sched tag and still seem to get the same errors, so my feeling is that this is a Fluxion testsuite error and not any dependence on a given version of flux-core. I haven't run down the source of the errors yet though (actually I wonder if anyone else has issues, or perhaps it is user error on my part)

[trws]

I saw some issues in another PR that look related to the recently merged feasibility module seemingly interacting with the post-load error behavior. Not sure why it didn’t show up in the other PR though, maybe only shows up after the Core Image update? Get Outlook for iOS<https://aka.ms/o0ukef>

…

________________________________ From: Mark Grondona ***@***.***> Sent: Tuesday, August 19, 2025 1:33:45 PM To: flux-framework/flux-sched ***@***.***> Cc: Scogland, Tom ***@***.***>; Author ***@***.***> Subject: Re: [flux-framework/flux-sched] cmake: Bump required flux-core version (PR #1391) [https://avatars.githubusercontent.com/u/741970?s=20&v=4]grondo left a comment (flux-framework/flux-sched#1391)<https://urldefense.us/v3/__https://github.com/flux-framework/flux-sched/pull/1391*issuecomment-3202138770__;Iw!!G2kpM7uM-TzIFchu!xvOJw7YJmPjDywz7B3WClmLiovoXyzSDNMXrS1gQPF26wabELTxtrOw7Nb3WJ07Z4biNX_D3Kszbd-QFPlGZoATr6ZA$> I am getting errors running make check on a TOSS 4 system with flux-core v0.76.0. However, I went back to the latest flux-sched tag and still seem to get the same errors, so my feeling is that this is a Fluxion testsuite error and not any dependence on a given version of flux-core. I haven't run down the source of the errors yet though (actually I wonder if anyone else has issues, or perhaps it is user error on my part) — Reply to this email directly, view it on GitHub<https://urldefense.us/v3/__https://github.com/flux-framework/flux-sched/pull/1391*issuecomment-3202138770__;Iw!!G2kpM7uM-TzIFchu!xvOJw7YJmPjDywz7B3WClmLiovoXyzSDNMXrS1gQPF26wabELTxtrOw7Nb3WJ07Z4biNX_D3Kszbd-QFPlGZoATr6ZA$>, or unsubscribe<https://urldefense.us/v3/__https://github.com/notifications/unsubscribe-auth/AAFBFNJ4AW77EQL7NX4CZIT3OOC2TAVCNFSM6AAAAACEI7ZFXWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTEMBSGEZTQNZXGA__;!!G2kpM7uM-TzIFchu!xvOJw7YJmPjDywz7B3WClmLiovoXyzSDNMXrS1gQPF26wabELTxtrOw7Nb3WJ07Z4biNX_D3Kszbd-QFPlGZ4hGSlZo$>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[trws]

Were the errors like these? https://github.com/flux-framework/flux-sched/actions/runs/17078065102/job/48424613295#step:6:2084

I’m not 100% sure why, but a couple of different PRs got those errors for a while earlier today, then they seemed to be resolved. Not sure why…

[grondo]

It doesn't appear to be similar, though I may need to do more testing. Here's an example failure from t1007-recovery-full.t:

expecting success: 
    load_resource match-format=rv1 policy=high &&
    load_qmanager

Aug 19 21:44:57.102922 UTC 2025 sched-fluxion-resource.info[0]: version 0.45.0
Aug 19 21:44:57.103778 UTC 2025 sched-fluxion-resource.info[0]: populate_resource_db: loaded resources from core's resource.acquire
Aug 19 21:44:57.135515 UTC 2025 sched-fluxion-qmanager.info[0]: version 0.45.0
ok 3 - recovery: loading flux-sched modules works (rv1)

expecting success: 
    jobid1=$(flux job submit basic.json) &&
    jobid2=$(flux job submit basic.json) &&
    jobid3=$(flux job submit basic.json) &&
    jobid4=$(flux job submit basic.json) &&
    jobid5=$(flux job submit basic.json) &&
    jobid6=$(flux job submit basic.json) &&
    jobid7=$(flux job submit basic.json) &&
    flux job wait-event -t 10 ${jobid4} start &&
    flux job wait-event -t 10 ${jobid7} submit

1755639897.500147 start
1755639897.639151 submit userid=6885 urgency=16 flags=0 version=1
ok 4 - recovery: submit to occupy resources fully (rv1)

expecting success: 
    remove_qmanager &&
    remove_resource &&
    flux cancel ${jobid1} &&
    flux job wait-event -t 10 ${jobid1} release

flux-job: wait-event timeout on event 'release'
not ok 5 - recovery: cancel one running job without fluxion
#	
#	    remove_qmanager &&
#	    remove_resource &&
#	    flux cancel ${jobid1} &&
#	    flux job wait-event -t 10 ${jobid1} release
#	

expecting success: 
    load_resource match-format=rv1 policy=high &&
    load_qmanager_sync &&
    test_must_fail flux ion-resource info ${jobid1} &&
    flux ion-resource info ${jobid2} | grep "ALLOCATED" &&
    flux ion-resource info ${jobid3} | grep "ALLOCATED" &&
    flux ion-resource info ${jobid4} | grep "ALLOCATED" &&
    flux job wait-event -t 10 ${jobid5} start &&
    test_expect_code 3 flux ion-resource info ${jobid6}

Aug 19 21:45:07.987805 UTC 2025 sched-fluxion-resource.info[0]: version 0.45.0
Aug 19 21:45:07.988870 UTC 2025 sched-fluxion-resource.info[0]: populate_resource_db: loaded resources from core's resource.acquire
Aug 19 21:45:08.018220 UTC 2025 sched-fluxion-qmanager.info[0]: version 0.45.0
{
 "queues": {
  "default": {
   "policy": "fcfs",
   "queue_depth": 32,
   "max_queue_depth": 1000000,
   "queue_parameters": {},
   "policy_parameters": {},
   "action_counts": {
    "pending": 3,
    "running": 4,
    "reserved": 0,
    "rejected": 0,
    "complete": 0,
    "cancelled": 0,
    "reprioritized": 0
   },
   "pending_queues": {
    "pending": [
     "fppqqGB",
     "frFs93V",
     "fsZUWR5"
    ],
    "pending_provisional": [],
    "blocked": []
   },
   "scheduled_queues": {
    "running": [
     "fkTKx5D",
     "fmxoDhZ",
     "fiyLfjD",
     "foLrYvB"
    ],
    "rejected": [],
    "canceled": []
   }
  }
 }
}
JOBID                STATUS               AT                   OVERHEAD (Secs)     
27548188672          ALLOCATED            2025-08-19T21:44:57  0.000239722         
test_must_fail: command succeeded: flux ion-resource info fiyLfjD
not ok 6 - recovery: works when both modules restart (rv1)
#	
#	    load_resource match-format=rv1 policy=high &&
#	    load_qmanager_sync &&
#	    test_must_fail flux ion-resource info ${jobid1} &&
#	    flux ion-resource info ${jobid2} | grep "ALLOCATED" &&
#	    flux ion-resource info ${jobid3} | grep "ALLOCATED" &&
#	    flux ion-resource info ${jobid4} | grep "ALLOCATED" &&
#	    flux job wait-event -t 10 ${jobid5} start &&
#	    test_expect_code 3 flux ion-resource info ${jobid6}
#	

(with more cascading failures as the test progresses)

[trws]

Huh... where were you testing? I just built and ran master on dane and everything passed first try. There must be something going on but I'm having trouble tracing it down.

[grondo]

Hm, this was tuolumne. I was also wondering if it is something in my environment. Sorry I haven't had a chance to run this down yet.