Megatest

Check-in [d5bb9fca6e]
Login
Overview
Comment:remove-mutex From: 022801133154585551c60feee4baf6758b91f993 User: matt
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | v1.65-lazyqueue-items-rollup
Files: files | file ages | folders
SHA1: d5bb9fca6eba65aad22b57c7a5259b29ec5bb06a
User & Date: matt on 2021-02-25 22:05:47
Other Links: branch diff | manifest | tags
Context
2021-02-25
22:05
remove-mutex From: 022801133154585551c60feee4baf6758b91f993 User: matt Closed-Leaf check-in: d5bb9fca6e user: matt tags: v1.65-lazyqueue-items-rollup
22:05
add finalize of no-sync and re-enable the mutex From: 1775254e3f45db753fe942e0f142c3ec732dbcb6 User: matt check-in: d616d48e39 user: matt tags: v1.65-lazyqueue-items-rollup
Changes

Modified db.scm from [d64b3f8e76] to [e5323ac36a].

3996
3997
3998
3999
4000
4001
4002
4003
4004
4005
4006
4007
4008
4009
4010
;; 	  (("WARN" "FAIL")  '("COMPLETED" "FAIL"))
;; 	  (("WARN" "CHECK") '("COMPLETED" "CHECK"))
;; 	  (("WARN" "DEAD")
       
(define (db:set-state-status-and-roll-up-items dbstruct run-id test-name item-path state status comment)
  ;; establish info on incoming test followed by info on top level test
  ;; BBnote - for mode itemwait, linkage between upstream test & matching item status is propagated to run queue in db:prereqs-not-met
  (mutex-lock! *db-transaction-mutex*) ;; why do we need a mutex?
  (let* ((testdat      (if (number? test-name)
			   (db:get-test-info-by-id dbstruct run-id test-name) ;; test-name is actually a test-id
			   (db:get-test-info       dbstruct run-id test-name item-path)))
	 (test-id      (db:test-get-id testdat))
	 (test-name    (if (number? test-name)
			   (db:test-get-testname testdat)
			   test-name))







|







3996
3997
3998
3999
4000
4001
4002
4003
4004
4005
4006
4007
4008
4009
4010
;; 	  (("WARN" "FAIL")  '("COMPLETED" "FAIL"))
;; 	  (("WARN" "CHECK") '("COMPLETED" "CHECK"))
;; 	  (("WARN" "DEAD")
       
(define (db:set-state-status-and-roll-up-items dbstruct run-id test-name item-path state status comment)
  ;; establish info on incoming test followed by info on top level test
  ;; BBnote - for mode itemwait, linkage between upstream test & matching item status is propagated to run queue in db:prereqs-not-met
  ;; (mutex-lock! *db-transaction-mutex*) ;; why do we need a mutex?
  (let* ((testdat      (if (number? test-name)
			   (db:get-test-info-by-id dbstruct run-id test-name) ;; test-name is actually a test-id
			   (db:get-test-info       dbstruct run-id test-name item-path)))
	 (test-id      (db:test-get-id testdat))
	 (test-name    (if (number? test-name)
			   (db:test-get-testname testdat)
			   test-name))
4044
4045
4046
4047
4048
4049
4050
4051
4052
4053
4054
4055
4056
4057
4058
4059
4060
4061
4062
4063
4064
4065
4066
4067
		    (loop (+ count 1))
		    (sqlite3:with-transaction
		     no-sync-db
		     (lambda ()
		       (db:no-sync-set no-sync-db rollup-lock-key (current-seconds))
		       (db:no-sync-del! no-sync-db waiting-lock-key))))))
	  ;; now the rollup
	  (mutex-unlock! *db-transaction-mutex*) ;; why do we need a mutex?
	  (if rollup-flag ;; put this into a thread
	      (begin
		;; (thread-start! (make-thread
		;; 	      (lambda ()
		(db:roll-up-test-state-status dbstruct run-id test-name state status)
		(db:no-sync-del! no-sync-db rollup-lock-key))
	      ;; (conc "thread for run-id: " run-id " test-name: " test-name))))))))
	      ))
	(mutex-unlock! *db-transaction-mutex*) ;; why do we need a mutex?
	)))
	      
;; I'd like to remove the need for item-path - it is logically not needed here
;; for now we pass in state and status - NOTE: There is a possible race if a test
;; is rapidly re-run while an earlier run is waiting to rollup.
;;
(define (db:roll-up-test-state-status dbstruct run-id test-name state status)







|








|







4044
4045
4046
4047
4048
4049
4050
4051
4052
4053
4054
4055
4056
4057
4058
4059
4060
4061
4062
4063
4064
4065
4066
4067
		    (loop (+ count 1))
		    (sqlite3:with-transaction
		     no-sync-db
		     (lambda ()
		       (db:no-sync-set no-sync-db rollup-lock-key (current-seconds))
		       (db:no-sync-del! no-sync-db waiting-lock-key))))))
	  ;; now the rollup
	  ;; (mutex-unlock! *db-transaction-mutex*) ;; why do we need a mutex?
	  (if rollup-flag ;; put this into a thread
	      (begin
		;; (thread-start! (make-thread
		;; 	      (lambda ()
		(db:roll-up-test-state-status dbstruct run-id test-name state status)
		(db:no-sync-del! no-sync-db rollup-lock-key))
	      ;; (conc "thread for run-id: " run-id " test-name: " test-name))))))))
	      ))
	;; (mutex-unlock! *db-transaction-mutex*) ;; why do we need a mutex?
	)))
	      
;; I'd like to remove the need for item-path - it is logically not needed here
;; for now we pass in state and status - NOTE: There is a possible race if a test
;; is rapidly re-run while an earlier run is waiting to rollup.
;;
(define (db:roll-up-test-state-status dbstruct run-id test-name state status)