Differences From Artifact [3e262d9c09]:
- File runs.scm — part of check-in [89fedf98b8] at 2017-09-29 17:56:31 on branch v1.64-keep-running-fix — got further. noticed race condition when not stepping one at a time by setting launcher. noticed xor does not handle preq-fail. (user: bjbarcla, size: 124859) [annotate] [blame] [check-ins using]
To Artifact [59d5521d01]:
- File
runs.scm
— part of check-in
[6baf47004a]
at
2017-10-02 17:57:41
on branch v1.64-keep-running-fix
— upgraded mt: to rmt: in runs.scm for status setting, fleshed out statuses that qualify for killed & in progress in db:prereqs-not-met
.. dead end. learnings applied to v1.64-itemflow2 branch. (user: bjbarcla, size: 125476) [annotate] [blame] [check-ins using]
︙ | ︙ | |||
703 704 705 706 707 708 709 | (non-completed (runs:calc-not-completed prereqs)) (runnable-prereqs (runs:calc-runnable prereqs)) (unexpanded-prereqs (filter (lambda (testname) (let* ((test-rec (hash-table-ref test-records testname)) (items (tests:testqueue-get-items test-rec))) | | | > | | | | | 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 | (non-completed (runs:calc-not-completed prereqs)) (runnable-prereqs (runs:calc-runnable prereqs)) (unexpanded-prereqs (filter (lambda (testname) (let* ((test-rec (hash-table-ref test-records testname)) (items (tests:testqueue-get-items test-rec))) ;;(BB> "HEY " testname "=>"items) (or (procedure? items)(eq? items 'have-procedure)))) waitons)) (completed-prereq-items (let ((foo (begin ;;(BB> "hello prereqs: "prereqs) #t)) (res (filter (lambda (test) ;;(BB> "foo - "test) (and (vector? test) (equal? "COMPLETED" (db:test-get-state test)) (equal? "COMPLETED" (db:test-get-state test)) (not (equal? "" (db:test-get-item-path test))))) prereqs))) res)) ) (debug:print-info 1 *default-log-port* "START OF INNER COND #2 " "\n can-run-more: " can-run-more "\n testname: " hed "\n prereqs: " (runs:pretty-string prereqs) "\n completed-prereq-items: " (runs:pretty-string completed-prereq-items) "\n non-completed: " (runs:pretty-string non-completed) "\n prereq-fails: " (runs:pretty-string prereq-fails) "\n runnable-prereqs: " (runs:pretty-string runnable-prereqs) "\n fails: " (runs:pretty-string fails) "\n testmode: " testmode "\n (member 'toplevel testmode): " (member 'toplevel testmode) "\n (null? non-completed): " (null? non-completed) "\n reruns: " reruns "\n items: " items "\n unexpanded-prereqs: " unexpanded-prereqs ;;all-prereqs-expanded "\n completed-prereq-items: " completed-prereq-items "\n have-itemized: " have-itemized "\n can-run-more: " can-run-more) ;;(BB> "before runs:expand-items cond") (let ((res (cond ;; all prereqs met, fire off the test ;; or, if it is a 'toplevel test and all prereqs not met are COMPLETED then launch ;; runs:expand-items case: test of interest not toplevel and IS blackballed -> ??? ((and (not (member 'toplevel testmode)) ;; test has been blackballed elsewhere (member (hash-table-ref/default test-registry (db:test-make-full-name hed item-path) 'n/a) '(DONOTRUN removed CANNOTRUN))) ;; *common:cant-run-states-sym*) ;; '(COMPLETED KILLED WAIVED UNKNOWN INCOMPLETE)) ;; try to catch repeat processing of COMPLETED tests here ;;(BB> "cb1") (debug:print-info 1 *default-log-port* "Test " hed " set to \"" (hash-table-ref test-registry (db:test-make-full-name hed item-path)) "\". Removing it from the queue") (if (or (not (null? tal)) (not (null? reg))) (runs:loop-values tal reg reglen regfull reruns) ;; blackballed test - throw it away (begin (debug:print-info 0 *default-log-port* "Nothing left in the queue!") |
︙ | ︙ | |||
781 782 783 784 785 786 787 | ;; runs:expand-items case: toplevel or else no dangling prerequeistes -- expand items now. ((or (and have-itemized (null? unexpanded-prereqs) (not (null? completed-prereq-items))) (null? prereqs) ;; nothing is in our way to proceed (need to expand this to an item level check.) (and (member 'toplevel testmode) ;; for toplevel test - proceed (nothing in our way) (null? non-completed))) | | | 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 | ;; runs:expand-items case: toplevel or else no dangling prerequeistes -- expand items now. ((or (and have-itemized (null? unexpanded-prereqs) (not (null? completed-prereq-items))) (null? prereqs) ;; nothing is in our way to proceed (need to expand this to an item level check.) (and (member 'toplevel testmode) ;; for toplevel test - proceed (nothing in our way) (null? non-completed))) ;;(BB> "cb2") (debug:print-info 4 *default-log-port* "runs:expand-items: (or (null? prereqs) (and (member 'toplevel testmode)(null? non-completed)))") (let ((test-name (tests:testqueue-get-testname test-record))) (setenv "MT_TEST_NAME" test-name) ;; hack to give context to get-items-from-config TODO: call-with-environment-variables (setenv "MT_RUNNAME" runname) (runs:set-megatest-env-vars run-id inrunname: runname) ;; these may be needed by the launching process (let ((items-list (items:get-items-from-config tconfig))) ;; BB: RIGHT HERE is where item expansion occurs.. target vars are not expanded to env vars at this point (item expansion using [items]\nwhatever [system echo $TARGETVAR] doesnt work right whereas [system echo #{targetvar}] does.. Tal and Randy have tix on this. |
︙ | ︙ | |||
807 808 809 810 811 812 813 | (debug:print-error 0 *default-log-port* "The proc from reading the items table did not yield a list - please report this") (exit 1)))))) ;; runs:expand-items case: no fails, no prereq-fails, some non-completed ((and (null? fails) (null? prereq-fails) (not (null? non-completed))) | | | 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 | (debug:print-error 0 *default-log-port* "The proc from reading the items table did not yield a list - please report this") (exit 1)))))) ;; runs:expand-items case: no fails, no prereq-fails, some non-completed ((and (null? fails) (null? prereq-fails) (not (null? non-completed))) ;;(BB> "cb3") (let* ((allinqueue (map (lambda (x)(if (string? x) x (db:test-get-testname x))) (append newtal reruns))) ;; prereqstrs is a list of test names as strings that are prereqs for hed (prereqstrs (delete-duplicates (map (lambda (x)(if (string? x) x (db:test-get-testname x))) prereqs))) ;; a prereq that is not found in allinqueue will be put in the notinqueue list ;; |
︙ | ︙ | |||
848 849 850 851 852 853 854 | (runs:loop-values trimmed-tal trimmed-reg reglen regfull reruns) )) (list (car newtal)(append (cdr newtal) reg) '() reruns)))) ((and (null? fails) ;; have not-started tests, but unable to run them. everything looks completed with no prospect of unsticking something that is stuck. we should mark hed as moribund and exit or continue if there are more tests to consider (null? prereq-fails) (null? non-completed)) | | | | > | > > | > > | | | | | 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 | (runs:loop-values trimmed-tal trimmed-reg reglen regfull reruns) )) (list (car newtal)(append (cdr newtal) reg) '() reruns)))) ((and (null? fails) ;; have not-started tests, but unable to run them. everything looks completed with no prospect of unsticking something that is stuck. we should mark hed as moribund and exit or continue if there are more tests to consider (null? prereq-fails) (null? non-completed)) ;;(BB> "cb4") (if (runs:can-keep-running? hed 20) (begin (runs:inc-cant-run-tests hed) (debug:print-info 0 *default-log-port* "no fails in prerequisites for " hed " but also none running, keeping " hed " for now. Try count: " (hash-table-ref/default *seen-cant-run-tests* hed 0)) ;; ;; getting here likely means the system is way overloaded, kill a full minute before continuing (thread-sleep! 60) ;; TODO: gate by normalized server load > 1.0 (maxload config thing) ;; num-retries code was here ;; we use this opportunity to move contents of reg to tal (list (car newtal)(append (cdr newtal) reg) '() reruns)) ;; an issue with prereqs not yet met? (begin (debug:print-info 1 *default-log-port* "no fails in prerequisites for " hed " but nothing seen running in a while, dropping test " hed " from the run queue") (let ((test-id (rmt:get-test-id run-id hed ""))) (if test-id (mt:test-set-state-status-by-id run-id test-id "NOT_STARTED" "TIMED_OUT" "Nothing seen running in a while."))) (runs:loop-values tal reg reglen regfull reruns) ))) ((and (or (not (null? fails)) (not (null? prereq-fails))) (member 'normal testmode)) ;;(BB> "cb5") (debug:print-info 1 *default-log-port* "test " hed " (mode=" testmode ") has failed prerequisite(s); " (string-intersperse (map (lambda (t)(conc (db:test-get-testname t) ":" (db:test-get-state t)"/"(db:test-get-status t))) fails) ", ") ", removing it from to-do list") (let ((test-id (rmt:get-test-id run-id hed ""))) (if test-id (if (not (null? prereq-fails)) ;;was: (mt:test-set-state-status-by-id run-id test-id "NOT_STARTED" "PREQ_DISCARDED" "Failed to run due to prior failed prerequisites") (rmt:set-state-status-and-roll-up-items run-id test-id #f "NOT_STARTED" "PREQ_DISCARDED" "Failed to run due to prior failed prerequisites") (begin ;;(debug:print 4 *default-log-port*"BB> set PREQ_FAIL on "hed) ;;was: (mt:test-set-state-status-by-id run-id test-id "NOT_STARTED" "PREQ_FAIL" "Failed to run due to failed prerequisites") (rmt:set-state-status-and-roll-up-items run-id test-id #f "NOT_STARTED" "PREQ_FAIL" "Failed to run due to failed prerequisites") )))) (if (or (not (null? reg))(not (null? tal))) (begin (hash-table-set! test-registry hed 'CANNOTRUN) (runs:loop-values tal reg reglen regfull (cons hed reruns)) ) #f)) ;; #f flags do not loop ((and (not (null? fails))(member 'toplevel testmode)) ;;(BB> "cb6") (if (or (not (null? reg))(not (null? tal))) (list (car newtal)(append (cdr newtal) reg) '() reruns) #f)) ((null? runnable-prereqs) ;;(BB> "cb7") #f) ;; if we get here and non-completed is null then it is all over. (else ;;(BB> "cb8") (debug:print 0 *default-log-port* "WARNING: FAILS or incomplete tests maybe preventing completion of this run. Watch for issues with test " hed ", continuing for now") (list (car newtal)(cdr newtal) reg reruns))))) ;;(BB> "after runs:expand-items big cond") res))) (define (runs:mixed-list-testname-and-testrec->list-of-strings inlst) (if (null? inlst) '() (map (lambda (t) (cond |
︙ | ︙ | |||
1120 1121 1122 1123 1124 1125 1126 | ;; the waiton is FAIL so no point in trying to run hed ever again (if (or (not (null? reg))(not (null? tal))) (if (or (vector? hed) (not (null? fails))) ;; BB: why do we need a vector? in my case, fails is populated (prereq failed), reg is not nul, and we really want to drop this one (begin (debug:print 1 *default-log-port* "WARNING: Dropping test " test-name "/" item-path " from the launch list as it has prerequistes that are FAIL") (let ((test-id (rmt:get-test-id run-id hed ""))) | > | > > | | 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 | ;; the waiton is FAIL so no point in trying to run hed ever again (if (or (not (null? reg))(not (null? tal))) (if (or (vector? hed) (not (null? fails))) ;; BB: why do we need a vector? in my case, fails is populated (prereq failed), reg is not nul, and we really want to drop this one (begin (debug:print 1 *default-log-port* "WARNING: Dropping test " test-name "/" item-path " from the launch list as it has prerequistes that are FAIL") (let ((test-id (rmt:get-test-id run-id hed ""))) (if test-id ;; was: ;(mt:test-set-state-status-by-id run-id test-id "NOT_STARTED" "PREQ_FAIL" "Failed to run due to failed prerequisites"))) (rmt:set-state-status-and-roll-up-items run-id test-id #f "NOT_STARTED" "PREQ_FAIL" "Failed to run due to failed prerequisites"))) (runs:shrink-can-run-more-tests-count runsdat) ;; DELAY TWEAKER (still needed?) ;; (thread-sleep! *global-delta*) ;; This next is for the items (if (not (null? fails)) ;;(mt:test-set-state-status-by-testname run-id test-name item-path "NOT_STARTED" "PREQ_FAIL" #f) (rmt:set-state-status-and-roll-up-items run-id test-name item-path "NOT_STARTED" "PREQ_FAIL" #f) ;;(mt:test-set-state-status-by-testname run-id test-name item-path "NOT_STARTED" "BLOCKED" #f) |
︙ | ︙ |