1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
|
;; Copyright 2006-2016, Matthew Welland.
;;
;; This program is made available under the GNU GPL version 2.0 or
;; greater. See the accompanying file COPYING for details.
;;
;; This program is distributed WITHOUT ANY WARRANTY; without even the
;; implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
;; PURPOSE.
;; strftime('%m/%d/%Y %H:%M:%S','now','localtime')
(use sqlite3 srfi-1 posix regex regex-case srfi-69 dot-locking (srfi 18)
posix-extras directory-utils pathname-expand typed-records format)
(import (prefix sqlite3 sqlite3:))
(declare (unit runs))
(declare (uses db))
(declare (uses common))
(declare (uses items))
(declare (uses runconfig))
(declare (uses tests))
(declare (uses server))
(declare (uses mt))
(declare (uses archive))
(declare (uses keys))
;; (declare (uses filedb))
(include "common_records.scm")
(include "key_records.scm")
(include "db_records.scm")
(include "run_records.scm")
(include "test_records.scm")
|
<
|
>
|
>
>
>
>
|
|
>
|
>
>
|
<
<
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
|
;; Copyright 2006-2016, Matthew Welland.
;;
;; This file is part of Megatest.
;;
;; Megatest is free software: you can redistribute it and/or modify
;; it under the terms of the GNU General Public License as published by
;; the Free Software Foundation, either version 3 of the License, or
;; (at your option) any later version.
;;
;; Megatest is distributed in the hope that it will be useful,
;; but WITHOUT ANY WARRANTY; without even the implied warranty of
;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
;; GNU General Public License for more details.
;;
;; You should have received a copy of the GNU General Public License
;; along with Megatest. If not, see <http://www.gnu.org/licenses/>.
;; strftime('%m/%d/%Y %H:%M:%S','now','localtime')
(use (prefix sqlite3 sqlite3:) srfi-1 posix regex regex-case srfi-69 (srfi 18)
posix-extras directory-utils pathname-expand typed-records format)
(declare (unit runs))
(declare (uses db))
(declare (uses common))
(declare (uses items))
(declare (uses runconfig))
(declare (uses tests))
(declare (uses server))
(declare (uses mt))
(declare (uses archive))
;; (declare (uses filedb))
(include "common_records.scm")
(include "key_records.scm")
(include "db_records.scm")
(include "run_records.scm")
(include "test_records.scm")
|
︙ | | | ︙ | |
46
47
48
49
50
51
52
53
54
55
56
57
58
59
|
((can-run-more-tests-count 0) : fixnum))
(defstruct runs:testdat
hed tal reg reruns test-record
test-name item-path jobgroup
waitons testmode newtal itemmaps prereqs-not-met)
;; set up needed environment variables given a run-id and optionally a target, itempath etc.
;;
(define (runs:set-megatest-env-vars run-id #!key (inkeys #f)(inrunname #f)(inkeyvals #f)(intarget #f)(testname #f)(itempath #f))
;;(bb-check-path msg: "runs:set-megatest-env-vars entry")
(let* ((target (or intarget
(common:args-get-target)
(get-environment-variable "MT_TARGET")))
|
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
|
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
|
((can-run-more-tests-count 0) : fixnum))
(defstruct runs:testdat
hed tal reg reruns test-record
test-name item-path jobgroup
waitons testmode newtal itemmaps prereqs-not-met)
(define (runs:get-mt-env-alist run-id runname target testname itempath)
;;(bb-check-path msg: "runs:set-megatest-env-vars entry")
`(("MT_TEST_NAME" . ,testname)
("MT_ITEMPATH" . ,itempath)
("MT_TARGET" . ,target)
("MT_RUNNAME" . ,runname)
("MT_RUN_AREA_HOME" . ,*toppath*)
,@(let* ((link-tree (common:get-linktree))) ;; (configf:lookup *configdat* "setup" "linktree")))
(if link-tree
(list (cons "MT_LINKTREE" link-tree)
(cons "MT_TEST_RUN_DIR"
(conc link-tree "/" target "/" runname "/" testname
(if (and (string? itempath) (not (equal? itempath "")))
(conc "/" itempath)
"")))
)
'()))
,@(map
(lambda (key)
(cons (car key) (cadr key)))
(keys:target->keyval (rmt:get-keys) target))
,@(map (lambda (var)
(let ((val (configf:lookup *configdat* "env-override" var)))
(cons var val)))
(configf:section-vars *configdat* "env-override"))))
;; set up needed environment variables given a run-id and optionally a target, itempath etc.
;;
(define (runs:set-megatest-env-vars run-id #!key (inkeys #f)(inrunname #f)(inkeyvals #f)(intarget #f)(testname #f)(itempath #f))
;;(bb-check-path msg: "runs:set-megatest-env-vars entry")
(let* ((target (or intarget
(common:args-get-target)
(get-environment-variable "MT_TARGET")))
|
︙ | | | ︙ | |
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
|
exn
(let ((call-chain (get-call-chain))
(msg ((condition-property-accessor 'exn 'message) exn)))
(if (< count 5)
(begin ;; this call is colliding, do some crude stuff to fix it.
(debug:print 0 *default-log-port* "ERROR: *configdat* was inaccessible! This should never happen. Retry #" count)
(launch:setup force-reread: #t)
(fatal-loop (+ count 1)))
(begin
(debug:print 0 *default-log-port* "FATAL: *configdat* was inaccessible! This should never happen. Retried " count " times. Message: " msg)
(debug:print 0 *default-log-port* "Call chain:")
(with-output-to-port *default-log-port*
(lambda ()(pp call-chain)))
(exit 1))))
;;(bb-check-path msg: "runs:set-megatest-env-vars block 1.5")
(alist->env-vars (hash-table-ref/default *configdat* "env-override" '())))) ;;;; environment is tainted HERE in this let block.
;;(bb-check-path msg: "runs:set-megatest-env-vars block 2")
;; Lets use this as an opportunity to put MT_RUNNAME in the environment
(let ((runname (if inrunname inrunname (rmt:get-run-name-from-id run-id))))
(if runname
(setenv "MT_RUNNAME" runname)
(debug:print-error 0 *default-log-port* "no value for runname for id " run-id)))
|
|
>
|
>
>
>
>
>
>
>
>
>
|
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
|
exn
(let ((call-chain (get-call-chain))
(msg ((condition-property-accessor 'exn 'message) exn)))
(if (< count 5)
(begin ;; this call is colliding, do some crude stuff to fix it.
(debug:print 0 *default-log-port* "ERROR: *configdat* was inaccessible! This should never happen. Retry #" count)
(launch:setup force-reread: #t)
(fatal-loop (+ count 1)))
(begin
(debug:print 0 *default-log-port* "FATAL: *configdat* was inaccessible! This should never happen. Retried " count " times. Message: " msg)
(debug:print 0 *default-log-port* "Call chain:")
(with-output-to-port *default-log-port*
(lambda ()
(print "*configdat* is >>"*configdat*"<<")
(pp *configdat*)
(pp call-chain)))
(exit 1))))
;;(bb-check-path msg: "runs:set-megatest-env-vars block 1.5")
(when (or (not *configdat*) (not (hash-table? *configdat*)))
(debug:print 0 *default-log-port* "WARNING: *configdat* was inaccessible! This should never happen. Brute force reread.")
;;(BB> "ERROR: *configdat* was inaccessible! This should never happen. Brute force reread.")
(thread-sleep! 2) ;; assuming nfs lag.
(launch:setup force-reread: #t))
(alist->env-vars (hash-table-ref/default *configdat* "env-override" '())))) ;;;; environment is tainted HERE in this let block.
;;(bb-check-path msg: "runs:set-megatest-env-vars block 2")
;; Lets use this as an opportunity to put MT_RUNNAME in the environment
(let ((runname (if inrunname inrunname (rmt:get-run-name-from-id run-id))))
(if runname
(setenv "MT_RUNNAME" runname)
(debug:print-error 0 *default-log-port* "no value for runname for id " run-id)))
|
︙ | | | ︙ | |
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
|
(if (> (- currtime lasttime) waitval)
(begin
(hash-table-set! *runs:denoise* key currtime)
#t)
#f)))
(define (runs:can-run-more-tests runsdat run-id jobgroup max-concurrent-jobs)
;; Take advantage of a good place to exit if running the one-pass methodology
(if (and (> (runs:dat-can-run-more-tests-count runsdat) 20)
(args:get-arg "-one-pass"))
(exit 0))
(thread-sleep! (cond
((> (runs:dat-can-run-more-tests-count runsdat) 20)
(if (runs:lownoise "waiting on tasks" 60)(debug:print-info 2 *default-log-port* "waiting for tasks to complete, sleeping briefly ..."))
2);; obviously haven't had any work to do for a while
(else 0)))
(let* ((num-running (rmt:get-count-tests-running run-id))
(num-running-in-jobgroup (rmt:get-count-tests-running-in-jobgroup run-id jobgroup))
(job-group-limit (let ((jobg-count (config-lookup *configdat* "jobgroups" jobgroup)))
(if (string? jobg-count)
(string->number jobg-count)
jobg-count))))
(if (> (+ num-running num-running-in-jobgroup) 0)
(runs:inc-can-run-more-tests-count runsdat)) ;; (set! *runs:can-run-more-tests-count* (+ *runs:can-run-more-tests-count* 1)))
(if (not (eq? *last-num-running-tests* num-running))
(begin
(debug:print 2 *default-log-port* "max-concurrent-jobs: " max-concurrent-jobs ", num-running: " num-running)
(set! *last-num-running-tests* num-running)))
(if (not (eq? 0 *globalexitstatus*))
(list #f num-running num-running-in-jobgroup max-concurrent-jobs job-group-limit)
(let ((can-not-run-more (cond
;; if max-concurrent-jobs is set and the number running is greater
;; than it then cannot run more jobs
((and max-concurrent-jobs (>= num-running max-concurrent-jobs))
(if (runs:lownoise "mcj msg" 60)
(debug:print 0 *default-log-port* "WARNING: Max running jobs exceeded, current number running: " num-running
", max_concurrent_jobs: " max-concurrent-jobs))
#t)
;; if job-group-limit is set and number of jobs in the group is greater
;; than the limit then cannot run more jobs of this kind
((and job-group-limit
(>= num-running-in-jobgroup job-group-limit))
(if (runs:lownoise (conc "maxjobgroup " jobgroup) 60)
(debug:print 1 *default-log-port* "WARNING: number of jobs " num-running-in-jobgroup
" in jobgroup \"" jobgroup "\" exceeds limit of " job-group-limit))
#t)
(else #f))))
(list (not can-not-run-more) num-running num-running-in-jobgroup max-concurrent-jobs job-group-limit)))))
;; test-names: Comma separated patterns same as test-patts but used in selection
;; of tests to run. The item portions are not respected.
;; FIXME: error out if /patt specified
;;
(define (runs:run-tests target runname test-patts user flags #!key (run-count 1)) ;; test-names
(let* ((keys (keys:config-get-fields *configdat*))
(keyvals (keys:target->keyval keys target))
(run-id (rmt:register-run keyvals runname "new" "n/a" user (args:get-arg "-contour"))) ;; test-name)))
;; (deferred '()) ;; delay running these since they have a waiton clause
(runconfigf (conc *toppath* "/runconfigs.config"))
(dbfile (conc *toppath* "/megatest.db"))
(readonly-mode (not (file-write-access? dbfile)))
(test-records (make-hash-table))
;; need to process runconfigs before generating these lists
(all-tests-registry #f) ;; (tests:get-all)) ;; (tests:get-valid-tests (make-hash-table) test-search-path)) ;; all valid tests to check waiton names
(all-test-names #f) ;; (hash-table-keys all-tests-registry))
(test-names #f) ;; Generated by a call to (tests:filter-test-names all-test-names test-patts))
(required-tests #f) ;; Put fully qualified test/testpath names in this list to be done
(task-key (conc (hash-table->alist flags) " " (get-host-name) " " (current-process-id)))
;; (tdbdat (tasks:open-db))
(config-reruns (let ((x (configf:lookup *configdat* "setup" "reruns")))
(if x (string->number x) #f)))
(allowed-tests #f))
;; check if readonly
(when readonly-mode
(debug:print-error 0 *default-log-port* "megatest.db is readonly. Cannot proceed.")
(exit 1))
;; per user request. If less than 100Meg space on dbdir partition, bail out with error
;; this will reduce issues in database corruption
(common:check-db-dir-and-exit-if-insufficient)
;; override the number of reruns from the configs
(if (and config-reruns
(> run-count config-reruns))
(set! run-count config-reruns))
;; (if (tasks:need-server run-id)(tasks:start-and-wait-for-server tdbdat run-id 10))
(let ((sighand (lambda (signum)
;; (signal-mask! signum) ;; to mask or not? seems to cause issues in exiting
(set! *time-to-exit* #t)
(print "Received signal " signum ", cleaning up before exit. Please wait...")
(let ((th1 (make-thread (lambda ()
|
>
|
>
>
|
>
|
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
|
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
|
|
|
|
|
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
|
(if (> (- currtime lasttime) waitval)
(begin
(hash-table-set! *runs:denoise* key currtime)
#t)
#f)))
(define (runs:can-run-more-tests runsdat run-id jobgroup max-concurrent-jobs)
;; Take advantage of a good place to exit if running the one-pass methodology
(if (and (> (runs:dat-can-run-more-tests-count runsdat) 20)
(args:get-arg "-one-pass"))
(exit 0))
(thread-sleep! (cond ;; BB: check with Matt. Should this sleep move to cond clauses below where we determine we have too many jobs running rather than each time the and condition above is true (which seems like always)?
((> (runs:dat-can-run-more-tests-count runsdat) 20)
(if (runs:lownoise "waiting on tasks" 60)(debug:print-info 2 *default-log-port* "waiting for tasks to complete, sleeping briefly ..."))
(configf:lookup-number *configdat* "setup" "inter-test-delay" default: 0.1) ;; was 2
);; obviously haven't had any work to do for a while
(else 0)))
(let* ((num-running (rmt:get-count-tests-running run-id))
(num-running-in-jobgroup (rmt:get-count-tests-running-in-jobgroup run-id jobgroup))
(job-group-limit (let ((jobg-count (config-lookup *configdat* "jobgroups" jobgroup)))
(if (string? jobg-count)
(string->number jobg-count)
jobg-count))))
(if (> (+ num-running num-running-in-jobgroup) 0)
(runs:inc-can-run-more-tests-count runsdat)) ;; (set! *runs:can-run-more-tests-count* (+ *runs:can-run-more-tests-count* 1)))
(if (not (eq? *last-num-running-tests* num-running))
(begin
(debug:print 2 *default-log-port* "max-concurrent-jobs: " max-concurrent-jobs ", num-running: " num-running)
(set! *last-num-running-tests* num-running)))
(if (not (eq? 0 *globalexitstatus*))
(list #f num-running num-running-in-jobgroup max-concurrent-jobs job-group-limit)
(let* ((can-not-run-more (cond
;; if max-concurrent-jobs is set and the number running is greater
;; than it then cannot run more jobs
((and max-concurrent-jobs (>= num-running max-concurrent-jobs))
(if (runs:lownoise "mcj msg" 60)
(debug:print 0 *default-log-port* "WARNING: Max running jobs exceeded, current number running: " num-running
", max_concurrent_jobs: " max-concurrent-jobs))
#t)
;; if job-group-limit is set and number of jobs in the group is greater
;; than the limit then cannot run more jobs of this kind
((and job-group-limit
(>= num-running-in-jobgroup job-group-limit))
(if (runs:lownoise (conc "maxjobgroup " jobgroup) 60)
(debug:print 1 *default-log-port* "WARNING: number of jobs " num-running-in-jobgroup
" in jobgroup \"" jobgroup "\" exceeds limit of " job-group-limit))
#t)
(else #f))))
(list (not can-not-run-more) num-running num-running-in-jobgroup max-concurrent-jobs job-group-limit)))))
(define (runs:run-pre-hook run-id)
(let* ((run-pre-hook (configf:lookup *configdat* "runs" "pre-hook"))
(existing-tests (if run-pre-hook
(rmt:get-tests-for-run run-id "%" '() '() ;; run-id testpatt states statuses
#f #f ;; offset limit
#f ;; not-in
#f ;; sort-by
#f ;; sort-order
#f ;; get full data (not 'shortlist)
0 ;; (runs:gendat-inc-results-last-update *runs:general-data*) ;; last update time
'dashboard)
'()))
(log-dir (conc *toppath* "/logs"))
(log-file (conc "pre-hook-" (string-translate (getenv "MT_TARGET") "/" "-") "-" (getenv "MT_RUNNAME") ".log"))
(full-log-fname (conc log-dir "/" log-file)))
(if run-pre-hook
(if (null? existing-tests)
(let* ((use-log-dir (if (not (directory-exists? log-dir))
(handle-exceptions
exn
(begin
(debug:print 0 *default-log-port* "WARNING: Failed to create " log-dir)
#f)
(create-directory log-dir #t)
#t)
#t))
(start-time (current-seconds))
(actual-logf (if use-log-dir full-log-fname log-file)))
(handle-exceptions
exn
(begin
(print-call-chain *default-log-port*)
(debug:print 0 *default-log-port* "Message: " ((condition-property-accessor 'exn 'message) exn))
(debug:print 0 *default-log-port* "ERROR: failed to run pre-hook " run-pre-hook ", check the log " log-file))
(debug:print-info 0 *default-log-port* "running run-pre-hook: \"" run-pre-hook "\", log is " actual-logf)
(system (conc run-pre-hook " >> " actual-logf " 2>&1"))
(debug:print-info 0 *default-log-port* "pre-hook \"" run-pre-hook "\" took " (- (current-seconds) start-time) " seconds to run.")))
(debug:print 0 *default-log-port* "Skipping pre-hook call \"" run-pre-hook "\" as there are existing tests for this run.")))))
(define (runs:run-post-hook run-id)
(let* ((run-post-hook (configf:lookup *configdat* "runs" "post-hook"))
(existing-tests (if run-post-hook
(rmt:get-tests-for-run run-id "%" '() '() ;; run-id testpatt states statuses
#f #f ;; offset limit
#f ;; not-in
#f ;; sort-by
#f ;; sort-order
#f ;; get full data (not 'shortlist)
0 ;; (runs:gendat-inc-results-last-update *runs:general-data*) ;; last update time
'dashboard)
'()))
(log-dir (conc *toppath* "/logs"))
(log-file (conc "post-hook-" (string-translate (getenv "MT_TARGET") "/" "-") "-" (getenv "MT_RUNNAME") ".log"))
(full-log-fname (conc log-dir "/" log-file)))
(if run-post-hook
;; (if (null? existing-tests)
;; (debug:print 0 *default-log-port* "Skipping post-hook call \"" run-post-hook "\" as there are existing tests for this run.")))))
(let* ((use-log-dir (if (not (directory-exists? log-dir))
(handle-exceptions
exn
(begin
(debug:print 0 *default-log-port* "WARNING: Failed to create " log-dir)
#f)
(create-directory log-dir #t)
#t)
#t))
(start-time (current-seconds))
(actual-logf (if use-log-dir full-log-fname log-file)))
(handle-exceptions
exn
(begin
(print-call-chain *default-log-port*)
(debug:print 0 *default-log-port* "Message: " ((condition-property-accessor 'exn 'message) exn))
(debug:print 0 *default-log-port* "ERROR: failed to run post-hook " run-post-hook ", check the log " log-file))
(debug:print-info 0 *default-log-port* "running run-post-hook: \"" run-post-hook "\", log is " actual-logf)
(system (conc run-post-hook " >> " actual-logf " 2>&1"))
(debug:print-info 0 *default-log-port* "post-hook \"" run-post-hook "\" took " (- (current-seconds) start-time) " seconds to run."))))))
;; return #t when all items in waitors-upon list are represented in test-patt, #f otherwise.
(define (runs:testpatts-mention-waitors-upon? test-patt waitors-upon)
(null? (tests:filter-test-names-not-matched waitors-upon test-patt)))
;;======================================================================
;; runs:run-tests is called from megatest.scm and itself
;;======================================================================
;;
;; test-names: Comma separated patterns same as test-patts but used in selection
;; of tests to run. The item portions are not respected.
;; FIXME: error out if /patt specified
;;
(define (runs:run-tests target runname test-patts user flags #!key (run-count 1)) ;; test-names
(let* ((keys (keys:config-get-fields *configdat*))
(keyvals (keys:target->keyval keys target))
(run-id (rmt:register-run keyvals runname "new" "n/a" user (args:get-arg "-contour"))) ;; test-name)))
;; (deferred '()) ;; delay running these since they have a waiton clause
(runconfigf (conc *toppath* "/runconfigs.config"))
(dbfile (conc *toppath* "/megatest.db"))
(readonly-mode (not (file-write-access? dbfile)))
(test-records (make-hash-table))
;; need to process runconfigs before generating these lists
(all-tests-registry #f) ;; (tests:get-all)) ;; (tests:get-valid-tests (make-hash-table) test-search-path)) ;; all valid tests to check waiton names
(all-test-names #f) ;; (hash-table-keys all-tests-registry))
(test-names #f) ;; Generated by a call to (tests:filter-test-names all-test-names test-patts))
(required-tests #f) ;; Put fully qualified test/testpath names in this list to be done
(waitors-upon (make-hash-table)) ;; given a test, return list of tests waiting upon this test.
(task-key (conc (hash-table->alist flags) " " (get-host-name) " " (current-process-id)))
;; (tdbdat (tasks:open-db))
(config-reruns (let ((x (configf:lookup *configdat* "setup" "reruns")))
(if x (string->number x) #f)))
(allowed-tests #f))
;; check if readonly
(when readonly-mode
(debug:print-error 0 *default-log-port* "megatest.db is readonly. Cannot proceed.")
(exit 1))
;; per user request. If less than 100Meg space on dbdir partition, bail out with error
;; this will reduce issues in database corruption
(common:check-db-dir-and-exit-if-insufficient)
;; override the number of reruns from the configs
;; this needs to be done at the place where is first runs:run-tests called
;(if (and config-reruns
; (> run-count config-reruns))
;(set! run-count config-reruns))
;; (if (tasks:need-server run-id)(tasks:start-and-wait-for-server tdbdat run-id 10))
(let ((sighand (lambda (signum)
;; (signal-mask! signum) ;; to mask or not? seems to cause issues in exiting
(set! *time-to-exit* #t)
(print "Received signal " signum ", cleaning up before exit. Please wait...")
(let ((th1 (make-thread (lambda ()
|
︙ | | | ︙ | |
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
|
(exit 4)))))
(thread-start! th2)
(thread-start! th1)
(thread-join! th2)))))
(set-signal-handler! signal/int sighand)
(set-signal-handler! signal/term sighand))
;; force the starting of a server
(debug:print 0 *default-log-port* "waiting on server...")
(server:start-and-wait *toppath*)
(runs:set-megatest-env-vars run-id inkeys: keys inrunname: runname) ;; these may be needed by the launching process
(set! runconf (if (file-exists? runconfigf)
(setup-env-defaults runconfigf run-id *already-seen-runconfig-info* keyvals target)
(begin
(debug:print 0 *default-log-port* "WARNING: You do not have a run config file: " runconfigf)
#f)))
(if (not test-patts) ;; first time in - adjust testpatt
(set! test-patts (common:args-get-testpatt runconf)))
(if (args:get-arg "-tagexpr")
(begin
(set! allowed-tests (string-join (runs:get-tests-matching-tags (args:get-arg "-tagexpr")) ","))
(debug:print-info 0 *default-log-port* "filtering initial test list with tagexpr: " (args:get-arg "-tagexpr") " => " allowed-tests)
));; tests will be ANDed with this list
;; register this run in monitor.db
(rmt:tasks-add "run-tests" user target runname test-patts task-key) ;; params)
(rmt:tasks-set-state-given-param-key task-key "running")
;; Now generate all the tests lists
(set! all-tests-registry (tests:get-all)) ;; hash of testname => path-to-test
(set! all-test-names (hash-table-keys all-tests-registry))
;; filter first for allowed-tests (from -tagexpr) then for test-patts.
(set! test-names (tests:filter-test-names
(if allowed-tests
(tests:filter-test-names all-test-names allowed-tests)
all-test-names)
test-patts))
;; I think seeding required-tests with all test-names makes sense but lack analysis to back that up.
;; NEW STRATEGY HERE:
;; 1. fill required tests with test-patts
;; 2. scan testconfigs and if waitons, itemwait, itempatt calc prior test test-patt
;; 3. repeat until all deps propagated
;; any tests with direct mention in test-patts can be added to required
;;
(set! required-tests (lset-intersection equal? (string-split test-patts ",") all-test-names))
;; (set! required-tests (lset-intersection equal? test-names all-test-names))
;; look up all tests matching the comma separated list of globs in
;; test-patts (using % as wildcard)
;; (set! test-names (delete-duplicates (tests:get-valid-tests *toppath* test-patts)))
(debug:print-info 0 *default-log-port* "tests search path: " (string-intersperse (tests:get-tests-search-path *configdat*) " "))
|
|
|
|
|
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
<
|
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
|
(exit 4)))))
(thread-start! th2)
(thread-start! th1)
(thread-join! th2)))))
(set-signal-handler! signal/int sighand)
(set-signal-handler! signal/term sighand))
;; force the starting of a server -- removed BB 17ww28 - no longer needed.
;;(debug:print 0 *default-log-port* "waiting on server...")
;;(server:start-and-wait *toppath*)
(runs:set-megatest-env-vars run-id inkeys: keys inrunname: runname) ;; these may be needed by the launching process
(set! runconf (if (common:file-exists? runconfigf)
(setup-env-defaults runconfigf run-id *already-seen-runconfig-info* keyvals target)
(begin
(debug:print 0 *default-log-port* "WARNING: You do not have a run config file: " runconfigf)
#f)))
(if (not test-patts) ;; first time in - adjust testpatt
(set! test-patts (common:args-get-testpatt runconf)))
;; if test-patts is #f at this point there is something wrong and we need to bail out
(if (not test-patts)
(begin
(debug:print 0 *default-log-port* "WARNING: there is no test pattern for this run. Exiting now.")
(exit 0)))
(if (args:get-arg "-tagexpr")
(begin
(set! allowed-tests (string-join (runs:get-tests-matching-tags (args:get-arg "-tagexpr")) ","))
(debug:print-info 0 *default-log-port* "filtering initial test list with tagexpr: " (args:get-arg "-tagexpr") " => " allowed-tests)
));; tests will be ANDed with this list
;; register this run in monitor.db
(rmt:tasks-add "run-tests" user target runname test-patts task-key) ;; params)
(rmt:tasks-set-state-given-param-key task-key "running")
(common:telemetry-log "run-tests"
payload:
`( (target . ,target)
(run-name . ,runname)
(test-patts . ,test-patts) ) )
;; Now generate all the tests lists
(set! all-tests-registry (tests:get-all)) ;; hash of testname => path-to-test
(set! all-test-names (hash-table-keys all-tests-registry))
;; filter first for allowed-tests (from -tagexpr) then for test-patts.
(set! test-names (tests:filter-test-names
(if allowed-tests
(tests:filter-test-names all-test-names allowed-tests)
all-test-names)
test-patts))
;; I think seeding required-tests with all test-names makes sense but lack analysis to back that up.
;; NEW STRATEGY HERE:
;; 1. fill required tests with test-patts
;; 2. scan testconfigs and if waitons, itemwait, itempatt calc prior test test-patt
;; 3. repeat until all deps propagated
;; any tests with direct mention in test-patts can be added to required
;;(set! required-tests (lset-intersection equal? (string-split test-patts ",") all-test-names))
(set! required-tests (tests:filter-test-names all-test-names test-patts))
;;
;; (set! required-tests (lset-intersection equal? test-names all-test-names))
;; look up all tests matching the comma separated list of globs in
;; test-patts (using % as wildcard)
;; (set! test-names (delete-duplicates (tests:get-valid-tests *toppath* test-patts)))
(debug:print-info 0 *default-log-port* "tests search path: " (string-intersperse (tests:get-tests-search-path *configdat*) " "))
|
︙ | | | ︙ | |
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
|
(rmt:set-tests-state-status run-id test-names state status "NOT_STARTED" status)))
;; list of state/status pairs separated by spaces
(string-split (or (configf:lookup *configdat* "setup" "allow-auto-rerun") "")))))
;; Ensure all tests are registered in the test_meta table
(runs:update-all-test_meta #f)
;; now add non-directly referenced dependencies (i.e. waiton)
;;======================================================================
;; refactoring this block into tests:get-full-data
;;
;; What happended, this code is now duplicated in tests!?
;;
;;======================================================================
(if (not (null? test-names))
(let loop ((hed (car test-names)) ;; NOTE: This is the main loop that iterates over the test-names
(tal (cdr test-names))) ;; 'return-procs tells the config reader to prep running system but return a proc
(change-directory *toppath*) ;; PLEASE OPTIMIZE ME!!! I think this should be a no-op but there are several places where change-directories could be happening.
(setenv "MT_TEST_NAME" hed) ;;
(let*-values (((waitons waitors config)(tests:get-waitons hed all-tests-registry)))
(debug:print-info 8 *default-log-port* "waitons: " waitons)
;; check for hed in waitons => this would be circular, remove it and issue an
;; error
(if (or (member hed waitons)
(member hed waitors))
(begin
(debug:print-error 0 *default-log-port* "test " hed " has listed itself as a waiton or waitor, please correct this!")
(set! waitons (filter (lambda (x)(not (equal? x hed))) waitons))
(set! waitors (filter (lambda (x)(not (equal? x hed))) waitors))))
;; (items (items:get-items-from-config config)))
(if (not (hash-table-ref/default test-records hed #f))
(hash-table-set! test-records
hed (vector hed ;; 0
config ;; 1
waitons ;; 2
(config-lookup config "requirements" "priority") ;; priority 3
(tests:get-items config) ;; expand the [items] and or [itemstable] into explict items
#f ;; itemsdat 5
#f ;; spare - used for item-path
waitors ;;
)))
(for-each
(lambda (waiton)
(if (and waiton (not (member waiton test-names)))
(let* ((waiton-record (hash-table-ref/default test-records waiton #f))
(waiton-tconfig (if waiton-record (vector-ref waiton-record 1) #f))
(waiton-itemized (and waiton-tconfig
(or (hash-table-ref/default waiton-tconfig "items" #f)
(hash-table-ref/default waiton-tconfig "itemstable" #f))))
(itemmaps (tests:get-itemmaps config)) ;; (configf:lookup config "requirements" "itemmap"))
(new-test-patts (tests:extend-test-patts test-patts hed waiton itemmaps)))
(debug:print-info 0 *default-log-port* "Test " waiton " has " (if waiton-record "a" "no") " waiton-record and" (if waiton-itemized " " " no ") "items")
;; need to account for test-patt here, if I am test "a", selected with a test-patt of "hed/b%"
;; and we are waiting on "waiton" we need to add "waiton/,waiton/b%" to test-patt
;; is this satisfied by merely appending "/" to the waiton name added to the list?
;;
;; This approach causes all of the items in an upstream test to be run
;; if we have this waiton already processed once we can analzye it for extending
;; tests to be run, since we can't properly process waitons unless they have been
;; initially added we add them again to be processed on second round AND add the hed
;; back in to also be processed on second round
;;
(if waiton-tconfig
(begin
(set! test-names (cons waiton test-names)) ;; need to process this one, only add once the waiton tconfig read
(if waiton-itemized
(begin
(debug:print-info 0 *default-log-port* "New test patts: " new-test-patts ", prev test patts: " test-patts)
(set! required-tests (cons (conc waiton "/") required-tests))
(set! test-patts new-test-patts))
(begin
(debug:print-info 0 *default-log-port* "Adding non-itemized test " waiton " to required-tests")
(set! required-tests (cons waiton required-tests))
(set! test-patts new-test-patts))))
(begin
(debug:print-info 0 *default-log-port* "No testconfig info yet for " waiton ", setting up to re-process it")
(set! tal (append (cons waiton tal)(list hed))))) ;; (cons (conc waiton "/") required-tests))
;; NOPE: didn't work. required needs to be plain test names. Try tacking on to test-patts
;; - doesn't work
;; (set! test-patts (conc test-patts "," waiton "/"))
;; (set! test-names (cons waiton test-names))))) ;; was an append, now a cons
)))
(delete-duplicates (append waitons waitors)))
(let ((remtests (delete-duplicates (append waitons tal))))
(if (not (null? remtests))
(begin
;; (debug:print-info 0 *default-log-port* "Preprocessing continues for " (string-intersperse remtests ", "))
(loop (car remtests)(cdr remtests))))))))
(if (not (null? required-tests))
(debug:print-info 1 *default-log-port* "Adding \"" (string-intersperse required-tests " ") "\" to the run queue"))
;; NOTE: these are all parent tests, items are not expanded yet.
(debug:print-info 4 *default-log-port* "test-records=" (hash-table->alist test-records))
(let ((reglen (configf:lookup *configdat* "setup" "runqueue")))
(if (> (length (hash-table-keys test-records)) 0)
(let* ((keep-going #t)
(run-queue-retries 5)
(th1 (make-thread (lambda ()
(runs:run-tests-queue run-id runname test-records keyvals flags test-patts required-tests (any->number reglen) all-tests-registry))
;; (handle-exceptions
;; exn
;; (begin
;; (print-call-chain (current-error-port))
;; (debug:print-error 0 *default-log-port* "failure in runs:run-tests-queue thread, error: " ((condition-property-accessor 'exn 'message) exn))
;; (if (> run-queue-retries 0)
;; (begin
;; (set! run-queue-retries (- run-queue-retries 1))
;; (runs:run-tests-queue run-id runname test-records keyvals flags test-patts required-tests (any->number reglen) all-tests-registry))))
;; (runs:run-tests-queue run-id runname test-records keyvals flags test-patts required-tests (any->number reglen) all-tests-registry)))
"runs:run-tests-queue"))
(th2 (make-thread (lambda ()
;; (rmt:find-and-mark-incomplete-all-runs))))) CAN'T INTERRUPT IT ...
(let ((run-ids (rmt:get-all-run-ids)))
(for-each (lambda (run-id)
(if keep-going
(handle-exceptions
exn
(debug:print 0 *default-log-port* "error in calling find-and-mark-incomplete for run-id " run-id)
(rmt:find-and-mark-incomplete run-id #f)))) ;; ovr-deadtime)))
run-ids)))
"runs: mark-incompletes")))
(thread-start! th1)
(thread-start! th2)
(thread-join! th1)
(set! keep-going #f)
(thread-join! th2)
;; if run-count > 0 call, set -preclean and -rerun STUCK/DEAD
(if (> run-count 0) ;; handle reruns
(begin
(if (not (hash-table-ref/default flags "-preclean" #f))
(hash-table-set! flags "-preclean" #t))
(if (not (hash-table-ref/default flags "-rerun" #f))
(hash-table-set! flags "-rerun" "STUCK/DEAD,n/a,ZERO_ITEMS"))
;; recursive call to self
(runs:run-tests target runname test-patts user flags run-count: (- run-count 1)))))
(debug:print-info 0 *default-log-port* "No tests to run")))
(debug:print-info 4 *default-log-port* "All done by here")
(rmt:tasks-set-state-given-param-key task-key "done")
;; (sqlite3:finalize! tasks-db)
))
;; loop logic. These are used in runs:run-tests-queue to make it a bit more readable.
;;
;; If reg not full and have items in tal then loop with (car tal)(cdr tal) reg reruns
;; If reg is full (i.e. length >= n
;; loop with (car reg) tal (cdr reg) reruns
;; If tal is empty
;; but have items in reg; loop with (car reg)(cdr reg) '() reruns
;; If reg is empty => all done
(define (runs:queue-next-hed tal reg n regfull)
(if regfull
(car reg)
(if (null? tal) ;; tal is used up, pop from reg
(car reg)
(car tal))))
(define (runs:queue-next-tal tal reg n regfull)
(if regfull
tal
(if (null? tal) ;; must transfer from reg
(cdr reg)
(cdr tal))))
(define (runs:queue-next-reg tal reg n regfull)
(if regfull
(cdr reg)
(if (null? tal) ;; if tal is null and reg not full then '() as reg contents moved to tal
'()
reg)))
(define runs:nothing-left-in-queue-count 0)
(define (runs:expand-items hed tal reg reruns regfull newtal jobgroup max-concurrent-jobs run-id waitons item-path testmode test-record can-run-more items runname tconfig reglen test-registry test-records itemmaps)
(let* ((loop-list (list hed tal reg reruns))
(prereqs-not-met (let ((res (rmt:get-prereqs-not-met run-id waitons hed item-path mode: testmode itemmaps: itemmaps)))
(if (list? res)
res
(begin
(debug:print 0 *default-log-port*
"ERROR: rmt:get-prereqs-not-met returned non-list!\n"
" res=" res " run-id=" run-id " waitons=" waitons " hed=" hed " item-path=" item-path " testmode=" testmode " itemmaps=" itemmaps)
'()))))
;; (prereqs-not-met (mt:lazy-get-prereqs-not-met run-id waitons item-path mode: testmode itemmap: itemmap))
(fails (runs:calc-fails prereqs-not-met))
(prereq-fails (runs:calc-prereq-fail prereqs-not-met))
(non-completed (runs:calc-not-completed prereqs-not-met))
(runnables (runs:calc-runnable prereqs-not-met)))
(debug:print-info 4 *default-log-port* "START OF INNER COND #2 "
"\n can-run-more: " can-run-more
"\n testname: " hed
"\n prereqs-not-met: " (runs:pretty-string prereqs-not-met)
"\n non-completed: " (runs:pretty-string non-completed)
"\n prereq-fails: " (runs:pretty-string prereq-fails)
"\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 can-run-more: " can-run-more)
(cond
;; all prereqs met, fire off the test
;; or, if it is a 'toplevel test and all prereqs not met are COMPLETED then launch
((and (not (member 'toplevel testmode))
(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
(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)))
(list (runs:queue-next-hed tal reg reglen regfull)
(runs:queue-next-tal tal reg reglen regfull)
(runs:queue-next-reg tal reg reglen regfull)
reruns)
(begin
(debug:print-info 0 *default-log-port* "Nothing left in the queue!")
;; If get here twice then we know we've tried to expand all items
;; since there must be a logic issue with the handling of loops in the
;; items expand phase we will brute force an exit here.
(if (> runs:nothing-left-in-queue-count 2)
(begin
(debug:print 0 *default-log-port* "WARNING: this condition is triggered when there were no items to expand and nothing to run. Please check your run for completeness")
(exit 0))
(set! runs:nothing-left-in-queue-count (+ runs:nothing-left-in-queue-count 1)))
#f)))
;;
((or (null? prereqs-not-met)
(and (member 'toplevel testmode)
(null? non-completed)))
(debug:print-info 4 *default-log-port* "runs:expand-items: (or (null? prereqs-not-met) (and (member 'toplevel testmode)(null? non-completed)))")
(let ((test-name (tests:testqueue-get-testname test-record)))
(setenv "MT_TEST_NAME" test-name) ;;
(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)))
(if (list? items-list)
(begin
(if (null? items-list)
(let ((test-id (rmt:get-test-id run-id test-name ""))
(num-items (rmt:test-toplevel-num-items run-id test-name)))
(if (and test-id
(not (> num-items 0)))
(mt:test-set-state-status-by-id run-id test-id "NOT_STARTED" "ZERO_ITEMS" "Failed to run due to failed prerequisites"))))
(tests:testqueue-set-items! test-record items-list)
(list hed tal reg reruns))
(begin
(debug:print-error 0 *default-log-port* "The proc from reading the items table did not yield a list - please report this")
(exit 1))))))
((and (null? fails)
(null? prereq-fails)
(not (null? non-completed)))
(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-not-met)))
;; a prereq that is not found in allinqueue will be put in the notinqueue list
;;
|
>
>
>
>
>
>
|
>
|
>
>
>
>
>
>
>
>
>
|
|
|
|
|
|
>
|
|
>
>
>
>
>
>
>
>
>
>
|
|
<
<
|
<
<
|
>
|
|
>
|
|
|
>
>
>
|
|
|
<
<
>
|
<
|
|
|
|
|
<
<
<
|
|
|
|
|
>
>
>
>
>
>
>
|
|
|
|
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
|
>
>
>
>
>
>
>
>
>
>
|
>
<
|
<
<
>
>
>
>
>
|
>
>
>
>
|
>
|
>
|
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
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
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
|
(rmt:set-tests-state-status run-id test-names state status "NOT_STARTED" status)))
;; list of state/status pairs separated by spaces
(string-split (or (configf:lookup *configdat* "setup" "allow-auto-rerun") "")))))
;; Ensure all tests are registered in the test_meta table
(runs:update-all-test_meta #f)
;; run the run prehook if there are no tests yet run for this run:
;;
(runs:run-pre-hook run-id)
;; mark all test launced flag as false in the meta table
(rmt:set-var (conc "lunch-complete-" run-id) "no")
(rmt:set-run-state-status run-id "new" "n/a")
;; now add non-directly referenced dependencies (i.e. waiton)
;;======================================================================
;; refactoring this block into tests:get-full-data
;;
;; What happended, this code is now duplicated in tests!?
;;
;;======================================================================
(if (not (null? test-names)) ;; BEGIN test-names loop
(let loop ((hed (car test-names)) ;; NOTE: This is the main loop that iterates over the test-names
(tal (cdr test-names))) ;; 'return-procs tells the config reader to prep running system but return a proc
(debug:print-info 4 *default-log-port* "\n\ntestpatt elaboration loop => hed="hed " tal="tal" test-patts="test-patts" test-names="test-names)
(change-directory *toppath*) ;; PLEASE OPTIMIZE ME!!! I think this should be a no-op but there are several places where change-directories could be happening.
(setenv "MT_TEST_NAME" hed) ;;
(let*-values (((waitons waitors config) (tests:get-waitons hed all-tests-registry))
;; NOTE: Have the config - can extract [waitons] section
((hed-mode)
(let ((m (config-lookup config "requirements" "mode")))
(if m (map string->symbol (string-split m)) '(normal))))
((hed-itemized-waiton) ;; are items in hed waiting on items of waiton?
(not (null? (lset-intersection eq? hed-mode '(itemmatch itemwait)))))
)
(debug:print-info 8 *default-log-port* "waitons: " waitons)
;; check for hed in waitons => this would be circular, remove it and issue an
;; error
(if (or (member hed waitons)
(member hed waitors))
(begin
(debug:print-error 0 *default-log-port* "test " hed " has listed itself as a waiton or waitor, please correct this!")
(set! waitons (filter (lambda (x)(not (equal? x hed))) waitons))
(set! waitors (filter (lambda (x)(not (equal? x hed))) waitors))))
;; (items (items:get-items-from-config config)))
(if (not (hash-table-ref/default test-records hed #f)) ;; waiton-tconfig below will be #f until that test is visted here at least once
(hash-table-set! test-records ;; BB: we are doing a manual make-tests:testqueue
hed (vector hed ;; 0 ;; testname
config ;; 1
waitons ;; 2
(config-lookup config "requirements" "priority") ;; priority 3
(tests:get-items config) ;; 4 ;; expand the [items] and or [itemstable] into explict items
#f ;; itemsdat 5
#f ;; spare - used for item-path
waitors ;;
)))
;; update waitors-upon here
(for-each
(lambda (waiton)
(let* ((current-waitors-upon (hash-table-ref/default waitors-upon waiton '())))
(debug:print-info 8 *default-log-port* " current-waiters-upon["waiton"] is "current-waitors-upon )
(when (not (member hed current-waitors-upon))
(debug:print-info 8 *default-log-port* " current-waiters-upon["waiton"] << "hed )
(hash-table-set! waitors-upon waiton (cons hed current-waitors-upon)))))
(if (list? waitons) waitons '()))
(debug:print-info 8 *default-log-port* " process waitons&waitors of "hed": "(delete-duplicates (append waitons waitors)))
(for-each
(lambda (waiton)
(if (and waiton (not (member waiton test-names)))
(let* ((waitors-in-testpatt (runs:testpatts-mention-waitors-upon? test-patts (hash-table-ref/default waitors-upon waiton '())))
(waiton-record (hash-table-ref/default test-records waiton #f))
(waiton-tconfig (if waiton-record (vector-ref waiton-record 1) #f))
(waiton-itemized (and waiton-tconfig
(or (hash-table-ref/default waiton-tconfig "items" #f)
(hash-table-ref/default waiton-tconfig "itemstable" #f))))
(itemmaps (tests:get-itemmaps config)) ;; (configf:lookup config "requirements" "itemmap"))
(new-test-patts (tests:extend-test-patts test-patts hed waiton itemmaps hed-itemized-waiton)))
(debug:print-info 0 *default-log-port* "Test " waiton " has " (if waiton-record "a" "no") " waiton-record and" (if waiton-itemized " " " no ") "items")
;; need to account for test-patt here, if I am test "a", selected with a test-patt of "hed/b%"
;; and we are waiting on "waiton" we need to add "waiton/,waiton/b%" to test-patt
;; is this satisfied by merely appending "/" to the waiton name added to the list?
;;
;; This approach causes all of the items in an upstream test to be run
;; if we have this waiton already processed once we can analzye it for extending
;; tests to be run, since we can't properly process waitons unless they have been
;; initially added we add them again to be processed on second round AND add the hed
;; back in to also be processed on second round
(if waiton-tconfig ;; BB: waiter should be in test-patts as well as the waiton have a tconfig.
(if waiton-itemized
(if waitors-in-testpatt
(begin
(debug:print-info 0 *default-log-port* "New test patts: " new-test-patts ", prev test patts: " test-patts)
(set! test-names (cons waiton test-names)) ;; need to process this one, only add once the waiton tconfig read
(set! required-tests (cons (conc waiton "/") required-tests))
(set! test-patts new-test-patts))
(begin
(debug:print-info 0 *default-log-port* "Waitor(s) not yet on testpatt for " waiton ", setting up to re-process it")
(set! tal (append (cons waiton tal)(list hed)))))
(begin
(debug:print-info 0 *default-log-port* "Adding non-itemized test " waiton " to required-tests")
(set! required-tests (cons waiton required-tests))
(set! test-patts new-test-patts)))
(begin
(debug:print-info 0 *default-log-port* "No testconfig info yet for " waiton ", setting up to re-process it")
(set! tal (append (cons waiton tal)(list hed))))) ;; (cons (conc waiton "/") required-tests))
;; NOPE: didn't work. required needs to be plain test names. Try tacking on to test-patts
;; - doesn't work
;; (set! test-patts (conc test-patts "," waiton "/"))
;; (set! test-names (cons waiton test-names))))) ;; was an append, now a cons
)))
(delete-duplicates (append waitons waitors)))
(let ((remtests (delete-duplicates (append waitons tal))))
(debug:print-info 8 *default-log-port* " remtests are "remtests)
(if (not (null? remtests))
(begin
;; (debug:print-info 0 *default-log-port* "Preprocessing continues for " (string-intersperse remtests ", "))
(loop (car remtests)(cdr remtests)))))))) ;; END test-names loop
(if (not (null? required-tests))
(debug:print-info 1 *default-log-port* "Adding \"" (string-intersperse required-tests " ") "\" to the run queue"))
;; NOTE: these are all parent tests, items are not expanded yet.
(debug:print-info 4 *default-log-port* "test-records=" (hash-table->alist test-records))
(let ((reglen (configf:lookup *configdat* "setup" "runqueue")))
(if (> (length (hash-table-keys test-records)) 0)
(let* ((keep-going #t)
(run-queue-retries 5)
(th1 (make-thread (lambda ()
(handle-exceptions
exn
(begin
(print-call-chain)
(print " message: " ((condition-property-accessor 'exn 'message) exn)))
(runs:run-tests-queue run-id runname test-records keyvals flags test-patts required-tests
(any->number reglen) all-tests-registry)))
"runs:run-tests-queue"))
(th2 (make-thread (lambda () ;; BBQ: why are we visiting ALL runs here?
;; (rmt:find-and-mark-incomplete-all-runs))))) CAN'T INTERRUPT IT ...
(let ((run-ids (rmt:get-all-run-ids)))
(for-each (lambda (run-id)
(if keep-going
(handle-exceptions
exn
(debug:print 0 *default-log-port* "error in calling find-and-mark-incomplete for run-id " run-id)
(rmt:find-and-mark-incomplete run-id #f)))) ;; ovr-deadtime))) ;; could be root of https://hsdes.intel.com/appstore/article/#/220546828/main -- Title: Megatest jobs show DEAD even though they are still running (1.64/27)
run-ids)))
"runs: mark-incompletes")))
(thread-start! th1)
(thread-start! th2)
(thread-join! th1)
(set! keep-going #f)
(thread-join! th2)
;; if run-count > 0 call, set -preclean and -rerun STUCK/DEAD
(if (> run-count 0) ;; handle reruns
(begin
(if (not (hash-table-ref/default flags "-preclean" #f))
(hash-table-set! flags "-preclean" #t))
(if (not (hash-table-ref/default flags "-rerun" #f))
(hash-table-set! flags "-rerun" "STUCK/DEAD,n/a,ZERO_ITEMS"))
;; recursive call to self
(runs:run-tests target runname test-patts user flags run-count: (- run-count 1)))))
(debug:print-info 0 *default-log-port* "No tests to run")))
(debug:print-info 4 *default-log-port* "All done by here")
;; TODO: try putting post hook call here
;(if (eq? run-count 0)
; (begin
; (debug:print-info 0 *default-log-port* "Calling Post Hook")
; (debug:print-info 2 *default-log-port* " run-count " run-count)
; (runs:run-post-hook run-id))
; (debug:print-info 2 *default-log-port* "Not calling post hook runcount = " run-count ))
(rmt:tasks-set-state-given-param-key task-key "done")
;; (sqlite3:finalize! tasks-db)
))
;; loop logic. These are used in runs:run-tests-queue to make it a bit more readable.
;;
;; If reg not full and have items in tal then loop with (car tal)(cdr tal) reg reruns
;; If reg is full (i.e. length >= n
;; loop with (car reg) tal (cdr reg) reruns
;; If tal is empty
;; but have items in reg; loop with (car reg)(cdr reg) '() reruns
;; If reg is empty => all done
(define (runs:queue-next-hed tal reg n regfull)
(if regfull
(if (null? reg) #f (car reg))
(if (null? tal) ;; tal is used up, pop from reg
(if (null? reg) #f (car reg))
(car tal))))
(define (runs:queue-next-tal tal reg n regfull)
(if regfull
tal
(if (null? tal) ;; must transfer from reg
(if (null? reg) '() (cdr reg))
(cdr tal))))
(define (runs:queue-next-reg tal reg n regfull)
(if regfull
(if (null? reg) '() (cdr reg)) ;; EXPLORE: reorder (cdr reg) such that looping is more efficient
(if (null? tal) ;; if tal is null and reg not full then '() as reg contents moved to tal
'()
reg)))
;; this is the list of parameters to the named loop "loop" near the top of runs:run-tests-queue, look around line 1216
;;
(define (runs:loop-values tal reg reglen regfull reruns)
(list (runs:queue-next-hed tal reg reglen regfull) ;; hed
(runs:queue-next-tal tal reg reglen regfull) ;; tal
(runs:queue-next-reg tal reg reglen regfull) ;; reg
reruns)) ;; reruns
;; objective - iterate thru tests
;; => want to prioritize tests we haven't seen before
;; => sometimes need to squeeze things in (added to reg)
;; => review of a previously seen test is higher priority of never visited test
;; reg - list of previously visited tests
;; tal - list of never visited tests
;; prefer next hed to be from reg than tal.
(define runs:nothing-left-in-queue-count 0)
;;======================================================================
;; runs:expand-items is called by runs:run-tests-queue
;;======================================================================
;;
;; return value of runs:expand-items is passed back to runs-tests-queue and is fed to named loop with this signature:
;; (let loop ((hed (car sorted-test-names))
;; (tal (cdr sorted-test-names))
;; (reg '()) ;; registered, put these at the head of tal
;; (reruns '()))
(define (runs:expand-items hed tal reg reruns regfull newtal jobgroup max-concurrent-jobs run-id waitons item-path testmode test-record can-run-more items runname tconfig reglen test-registry test-records itemmaps)
(let* ((loop-list (list hed tal reg reruns))
(prereqs-not-met (let ((res (rmt:get-prereqs-not-met run-id waitons hed item-path mode: testmode itemmaps: itemmaps)))
(if (list? res)
res
(begin
(debug:print 0 *default-log-port*
"ERROR: rmt:get-prereqs-not-met returned non-list!\n"
" res=" res " run-id=" run-id " waitons=" waitons " hed=" hed " item-path=" item-path " testmode=" testmode " itemmaps=" itemmaps)
'()))))
(have-itemized (not (null? (lset-intersection eq? testmode '(itemmatch itemwait)))))
;; (prereqs-not-met (mt:lazy-get-prereqs-not-met run-id waitons item-path mode: testmode itemmap: itemmap))
(fails (runs:calc-fails prereqs-not-met))
(prereq-fails (runs:calc-prereq-fail prereqs-not-met))
(non-completed (runs:calc-not-completed prereqs-not-met))
(runnables (runs:calc-runnable prereqs-not-met))
(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))
)
(debug:print-info 4 *default-log-port* "START OF INNER COND #2 "
"\n can-run-more: " can-run-more
"\n testname: " hed
"\n prereqs-not-met: " (runs:pretty-string prereqs-not-met)
"\n non-completed: " (runs:pretty-string non-completed)
"\n prereq-fails: " (runs:pretty-string prereq-fails)
"\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 can-run-more: " can-run-more)
(cond
;; all prereqs met, fire off the test
;; or, if it is a 'toplevel test and all prereqs not met are COMPLETED then launch
((and (not (member 'toplevel testmode))
(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
(debug:print-info 4 *default-log-port* "cond branch - " "ei-1")
(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)
(begin
(debug:print-info 0 *default-log-port* "Nothing left in the queue!")
;; If get here twice then we know we've tried to expand all items
;; since there must be a logic issue with the handling of loops in the
;; items expand phase we will brute force an exit here.
(if (> runs:nothing-left-in-queue-count 2)
(begin
(debug:print 0 *default-log-port* "WARNING: this condition is triggered when there were no items to expand and nothing to run. Please check your run for completeness")
(exit 0))
(set! runs:nothing-left-in-queue-count (+ runs:nothing-left-in-queue-count 1)))
#f)))
;; desired result of below cond branch:
;; we want to expand items in our test of interest (hed) in the following cases:
;; case 1 - mode is itemmatch or itemwait:
;; - all prereq tests have been expanded
;; - at least one prereq's items have completed
;; case 2 - mode is toplevel
;; - prereqs are completed.
;; - or no prereqs can complete
;; case 3 - mode not specified
;; - prereqs are completed and passed (we could consider removing "and passed" -- it would change behavior from current)
((or (null? prereqs-not-met)
(and (member 'toplevel testmode)
(null? non-completed)))
(debug:print-info 4 *default-log-port* "cond branch - " "ei-2")
(debug:print-info 4 *default-log-port* "runs:expand-items: (or (null? prereqs-not-met) (and (member 'toplevel testmode)(null? non-completed)))")
(let ((test-name (tests:testqueue-get-testname test-record)))
(setenv "MT_TEST_NAME" test-name) ;;
(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)))
(if (list? items-list)
(begin
(if (null? items-list)
(let ((test-id (rmt:get-test-id run-id test-name ""))
(num-items (rmt:test-toplevel-num-items run-id test-name)))
(if (and test-id
(not (> num-items 0)))
(mt:test-set-state-status-by-id run-id test-id "NOT_STARTED" "ZERO_ITEMS" "Failed to run due to failed prerequisites"))))
(tests:testqueue-set-items! test-record items-list)
(list hed tal reg reruns))
(begin
(debug:print-error 0 *default-log-port* "The proc from reading the items table did not yield a list - please report this")
(exit 1))))))
((and (null? fails)
(null? prereq-fails)
(not (null? non-completed)))
(debug:print-info 4 *default-log-port* "cond branch - " "ei-3")
(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-not-met)))
;; a prereq that is not found in allinqueue will be put in the notinqueue list
;;
|
︙ | | | ︙ | |
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
|
(if (and give-up
(not (and (null? tal)(null? reg))))
(let ((trimmed-tal (mt:discard-blocked-tests run-id hed tal test-records))
(trimmed-reg (mt:discard-blocked-tests run-id hed reg test-records)))
(debug:print 1 *default-log-port* "WARNING: test " hed " has discarded prerequisites, removing it from the 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" "PREQ_DISCARDED" "Failed to run due to discarded prerequisites")))
(if (and (null? trimmed-tal)
(null? trimmed-reg))
#f
(list (runs:queue-next-hed trimmed-tal trimmed-reg reglen regfull)
(runs:queue-next-tal trimmed-tal trimmed-reg reglen regfull)
(runs:queue-next-reg trimmed-tal trimmed-reg reglen regfull)
reruns)))
(list (car newtal)(append (cdr newtal) reg) '() reruns))))
((and (null? fails)
(null? prereq-fails)
(null? non-completed))
(if (runs:can-keep-running? hed 20)
(begin
(runs:inc-cant-run-tests hed)
(debug:print-info 1 *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)
;; 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.")))
(list (runs:queue-next-hed tal reg reglen regfull)
(runs:queue-next-tal tal reg reglen regfull)
(runs:queue-next-reg tal reg reglen regfull)
reruns))))
((and
(or (not (null? fails))
(not (null? prereq-fails)))
(member 'normal testmode))
(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))
(mt:test-set-state-status-by-id run-id test-id "NOT_STARTED" "PREQ_DISCARDED" "Failed to run due to prior failed prerequisites")
(mt:test-set-state-status-by-id run-id test-id "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)
(list (runs:queue-next-hed tal reg reglen regfull)
(runs:queue-next-tal tal reg reglen regfull)
(runs:queue-next-reg tal reg reglen regfull)
(cons hed reruns)))
#f)) ;; #f flags do not loop
((and (not (null? fails))(member 'toplevel testmode))
(if (or (not (null? reg))(not (null? tal)))
(list (car newtal)(append (cdr newtal) reg) '() reruns)
#f))
((null? runnables) #f) ;; if we get here and non-completed is null then it is all over.
(else
(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 (runs:queue-next-hed tal reg reglen regfull)
;; (runs:queue-next-tal tal reg reglen regfull)
;; (runs:queue-next-reg tal reg reglen regfull)
;; reruns)
(list (car newtal)(cdr newtal) reg reruns)))))
(define (runs:mixed-list-testname-and-testrec->list-of-strings inlst)
(if (null? inlst)
'()
(map (lambda (t)
(cond
|
|
<
|
<
|
|
>
|
|
<
|
<
|
|
>
|
>
>
|
<
<
<
|
>
|
>
>
>
|
>
<
<
<
<
|
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
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
|
(if (and give-up
(not (and (null? tal)(null? reg))))
(let ((trimmed-tal (mt:discard-blocked-tests run-id hed tal test-records))
(trimmed-reg (mt:discard-blocked-tests run-id hed reg test-records)))
(debug:print 1 *default-log-port* "WARNING: test " hed " has discarded prerequisites, removing it from the 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 "COMPLETED" "PREQ_DISCARDED" "Failed to run due to discarded prerequisites")))
(if (and (null? trimmed-tal)
(null? trimmed-reg))
#f
(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))
(debug:print-info 4 *default-log-port* "cond branch - " "ei-4")
(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))
(debug:print-info 4 *default-log-port* "cond branch - " "ei-5")
(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))
(mt:test-set-state-status-by-id-unless-completed run-id test-id "COMPLETED" "PREQ_DISCARDED" "Failed to run due to prior failed prerequisites")
(mt:test-set-state-status-by-id-unless-completed run-id test-id "COMPLETED" "PREQ_FAIL" "Failed to run due to failed prerequisites"))))
;; (debug:print 4 *default-log-port*"BB> set PREQ_FAIL on "hed)
;; (mt:test-set-state-status-by-id run-id test-id "NOT_STARTED" "PREQ_FAIL" "Failed to run due to failed prerequisites")))) ;; BB: this works, btu equivalent for itemwait mode does not work.
(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))
(debug:print-info 4 *default-log-port* "cond branch - " "ei-6")
(if (or (not (null? reg))(not (null? tal)))
(list (car newtal)(append (cdr newtal) reg) '() reruns)
#f))
((null? runnables)
(debug:print-info 4 *default-log-port* "cond branch - " "ei-7")
#f) ;; if we get here and non-completed is null then it is all over.
(else
(debug:print-info 4 *default-log-port* "cond branch - " "ei-8")
(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)))))
(define (runs:mixed-list-testname-and-testrec->list-of-strings inlst)
(if (null? inlst)
'()
(map (lambda (t)
(cond
|
︙ | | | ︙ | |
728
729
730
731
732
733
734
735
736
737
738
739
740
741
|
(conc t))))
inlst)))
;; hed tal reg reruns reglen regfull test-record runname test-name item-path jobgroup max-concurrent-jobs run-id waitons item-path testmode test-patts required-tests test-registry registry-mutex flags keyvals run-info newtal all-tests-registry itemmaps)
(define (runs:process-expanded-tests runsdat testdat)
;; unroll the contents of runsdat and testdat (due to ongoing refactoring).
(let* ((hed (runs:testdat-hed testdat))
(tal (runs:testdat-tal testdat))
(reg (runs:testdat-reg testdat))
(reruns (runs:testdat-reruns testdat))
(test-name (runs:testdat-test-name testdat))
(item-path (runs:testdat-item-path testdat))
(jobgroup (runs:testdat-jobgroup testdat))
|
>
>
>
|
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
|
(conc t))))
inlst)))
;; hed tal reg reruns reglen regfull test-record runname test-name item-path jobgroup max-concurrent-jobs run-id waitons item-path testmode test-patts required-tests test-registry registry-mutex flags keyvals run-info newtal all-tests-registry itemmaps)
(define (runs:process-expanded-tests runsdat testdat)
;; unroll the contents of runsdat and testdat (due to ongoing refactoring).
(debug:print 2 *default-log-port* "runs:process-expanded-tests; testdat:" )
(debug:print 2 *default-log-port* (with-output-to-string
(lambda () (pp (runs:testdat->alist testdat) ))))
(let* ((hed (runs:testdat-hed testdat))
(tal (runs:testdat-tal testdat))
(reg (runs:testdat-reg testdat))
(reruns (runs:testdat-reruns testdat))
(test-name (runs:testdat-test-name testdat))
(item-path (runs:testdat-item-path testdat))
(jobgroup (runs:testdat-jobgroup testdat))
|
︙ | | | ︙ | |
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
|
(have-resources (car run-limits-info))
(num-running (list-ref run-limits-info 1))
(num-running-in-jobgroup(list-ref run-limits-info 2))
(max-concurrent-jobs (list-ref run-limits-info 3))
(job-group-limit (list-ref run-limits-info 4))
;; (prereqs-not-met (rmt:get-prereqs-not-met run-id waitons hed item-path mode: testmode itemmaps: itemmaps))
;; (prereqs-not-met (mt:lazy-get-prereqs-not-met run-id waitons item-path mode: testmode itemmap: itemmap))
(fails (if (list? prereqs-not-met)
(runs:calc-fails prereqs-not-met)
(begin
(debug:print-error 0 *default-log-port* "prereqs-not-met is not a list! " prereqs-not-met)
'())))
(non-completed (filter (lambda (x) ;; remove hed from not completed list, duh, of course it is not completed!
(not (equal? x hed)))
(runs:calc-not-completed prereqs-not-met)))
(loop-list (list hed tal reg reruns))
;; configure the load runner
(numcpus (common:get-num-cpus #f))
(maxload (string->number (or (configf:lookup *configdat* "jobtools" "maxload") "3")))
(waitdelay (string->number (or (configf:lookup *configdat* "jobtools" "waitdelay") "60"))))
(debug:print-info 4 *default-log-port* "have-resources: " have-resources " prereqs-not-met: ("
(string-intersperse
(map (lambda (t)
(if (vector? t)
(conc (db:test-get-state t) "/" (db:test-get-status t))
(conc " WARNING: t is not a vector=" t )))
prereqs-not-met)
", ") ") fails: " fails
"\nregistered? " (hash-table-ref/default test-registry (db:test-make-full-name test-name item-path) #f))
(if (and (not (null? prereqs-not-met))
(runs:lownoise (conc "waiting on tests " prereqs-not-met hed) 60))
(debug:print-info 2 *default-log-port* "waiting on tests; " (string-intersperse (runs:mixed-list-testname-and-testrec->list-of-strings prereqs-not-met) ", ")))
;; Don't know at this time if the test have been launched at some time in the past
;; i.e. is this a re-launch?
(debug:print-info 4 *default-log-port* "run-limits-info = " run-limits-info)
(cond
;; Check item path against item-patts,
;;
((not (tests:match test-patts (tests:testqueue-get-testname test-record) item-path required: required-tests)) ;; This test/itempath is not to be run
;; else the run is stuck, temporarily or permanently
;; but should check if it is due to lack of resources vs. prerequisites
(debug:print-info 1 *default-log-port* "Skipping " (tests:testqueue-get-testname test-record) " " item-path " as it doesn't match " test-patts)
(if (or (not (null? tal))(not (null? reg)))
(list (runs:queue-next-hed tal reg reglen regfull)
(runs:queue-next-tal tal reg reglen regfull)
(runs:queue-next-reg tal reg reglen regfull)
reruns)
#f))
;; Register tests
;;
((not (hash-table-ref/default test-registry (db:test-make-full-name test-name item-path) #f))
(debug:print-info 4 *default-log-port* "Pre-registering test " test-name "/" item-path " to create placeholder" )
;; always do firm registration now in v1.60 and greater ;; (eq? *transport-type* 'fs) ;; no point in parallel registration if use fs
|
|
|
>
|
|
<
|
<
<
|
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
|
(have-resources (car run-limits-info))
(num-running (list-ref run-limits-info 1))
(num-running-in-jobgroup(list-ref run-limits-info 2))
(max-concurrent-jobs (list-ref run-limits-info 3))
(job-group-limit (list-ref run-limits-info 4))
;; (prereqs-not-met (rmt:get-prereqs-not-met run-id waitons hed item-path mode: testmode itemmaps: itemmaps))
;; (prereqs-not-met (mt:lazy-get-prereqs-not-met run-id waitons item-path mode: testmode itemmap: itemmap))
(fails (if (list? prereqs-not-met) ;; TODO: rename fails to failed-prereqs
(runs:calc-fails prereqs-not-met)
(begin
(debug:print-error 0 *default-log-port* "prereqs-not-met is not a list! " prereqs-not-met)
'())))
(non-completed (filter (lambda (x) ;; remove hed from not completed list, duh, of course it is not completed!
(not (equal? x hed)))
(runs:calc-not-completed prereqs-not-met)))
(loop-list (list hed tal reg reruns))
;; configure the load runner
(numcpus (common:get-num-cpus #f))
(maxload (string->number (or (configf:lookup *configdat* "jobtools" "maxload") "3.0"))) ;; use a non-number string to disable
(maxhomehostload (string->number (or (configf:lookup *configdat* "jobtools" "maxhomehostload") "2.0"))) ;; use a non-number string to disable
(waitdelay (string->number (or (configf:lookup *configdat* "jobtools" "waitdelay") "60"))))
(debug:print-info 4 *default-log-port* "have-resources: " have-resources " prereqs-not-met: ("
(string-intersperse
(map (lambda (t)
(if (vector? t)
(conc (db:test-get-state t) "/" (db:test-get-status t))
(conc " WARNING: t is not a vector=" t )))
prereqs-not-met)
", ") ") fails: " fails
"\nregistered? " (hash-table-ref/default test-registry (db:test-make-full-name test-name item-path) #f))
(if (and (not (null? prereqs-not-met))
(runs:lownoise (conc "waiting on tests " prereqs-not-met hed) 60))
(debug:print-info 2 *default-log-port* "waiting on tests; " (string-intersperse (runs:mixed-list-testname-and-testrec->list-of-strings prereqs-not-met) ", ")))
;; Don't know at this time if the test have been launched at some time in the past
;; i.e. is this a re-launch?
(debug:print-info 4 *default-log-port* "run-limits-info = " run-limits-info)
(cond ; cond 894- 1067
;; Check item path against item-patts,
;;
((not (tests:match test-patts (tests:testqueue-get-testname test-record) item-path required: required-tests)) ;; This test/itempath is not to be run
;; else the run is stuck, temporarily or permanently
;; but should check if it is due to lack of resources vs. prerequisites
(debug:print-info 1 *default-log-port* "Skipping " (tests:testqueue-get-testname test-record) " " item-path " as it doesn't match " test-patts)
(if (or (not (null? tal))(not (null? reg)))
(runs:loop-values tal reg reglen regfull reruns)
#f))
;; Register tests
;;
((not (hash-table-ref/default test-registry (db:test-make-full-name test-name item-path) #f))
(debug:print-info 4 *default-log-port* "Pre-registering test " test-name "/" item-path " to create placeholder" )
;; always do firm registration now in v1.60 and greater ;; (eq? *transport-type* 'fs) ;; no point in parallel registration if use fs
|
︙ | | | ︙ | |
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
|
(begin
(rmt:register-test run-id test-name "")
(if (rmt:get-test-id run-id test-name "")
(hash-table-set! test-registry (db:test-make-full-name test-name "") 'done))))
(runs:shrink-can-run-more-tests-count runsdat) ;; DELAY TWEAKER (still needed?)
(if (and (null? tal)(null? reg))
(list hed tal (append reg (list hed)) reruns)
(list (runs:queue-next-hed tal reg reglen regfull)
(runs:queue-next-tal tal reg reglen regfull)
;; NB// Here we are building reg as we register tests
;; if regfull we must pop the front item off reg
(if regfull
(append (cdr reg) (list hed))
(append reg (list hed)))
reruns)))
|
|
|
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
|
(begin
(rmt:register-test run-id test-name "")
(if (rmt:get-test-id run-id test-name "")
(hash-table-set! test-registry (db:test-make-full-name test-name "") 'done))))
(runs:shrink-can-run-more-tests-count runsdat) ;; DELAY TWEAKER (still needed?)
(if (and (null? tal)(null? reg))
(list hed tal (append reg (list hed)) reruns)
(list (runs:queue-next-hed tal reg reglen regfull) ;; cannot replace with a call to runs:loop-values as the logic is different for reg
(runs:queue-next-tal tal reg reglen regfull)
;; NB// Here we are building reg as we register tests
;; if regfull we must pop the front item off reg
(if regfull
(append (cdr reg) (list hed))
(append reg (list hed)))
reruns)))
|
︙ | | | ︙ | |
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
|
(not (member 'exclusive testmode)))))
;; (hash-table-delete! *max-tries-hash* (db:test-make-full-name test-name item-path))
;; we are going to reset all the counters for test retries by setting a new hash table
;; this means they will increment only when nothing can be run
(set! *max-tries-hash* (make-hash-table))
;; well, first lets see if cpu load throttling is enabled. If so wait around until the
;; average cpu load is under the threshold before continuing
(if (configf:lookup *configdat* "jobtools" "maxload") ;; only gate if maxload is specified
(common:wait-for-cpuload maxload numcpus waitdelay))
(run:test run-id run-info keyvals runname test-record flags #f test-registry all-tests-registry)
(runs:incremental-print-results run-id)
(hash-table-set! test-registry (db:test-make-full-name test-name item-path) 'running)
(runs:shrink-can-run-more-tests-count runsdat) ;; DELAY TWEAKER (still needed?)
;; (thread-sleep! *global-delta*)
(if (or (not (null? tal))(not (null? reg)))
(list (runs:queue-next-hed tal reg reglen regfull)
(runs:queue-next-tal tal reg reglen regfull)
(runs:queue-next-reg tal reg reglen regfull)
reruns)
#f))
;; must be we have unmet prerequisites
;;
(else
(debug:print 4 *default-log-port* "FAILS: " fails)
;; If one or more of the prereqs-not-met are FAIL then we can issue
|
|
|
>
>
>
<
|
<
<
|
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
|
(not (member 'exclusive testmode)))))
;; (hash-table-delete! *max-tries-hash* (db:test-make-full-name test-name item-path))
;; we are going to reset all the counters for test retries by setting a new hash table
;; this means they will increment only when nothing can be run
(set! *max-tries-hash* (make-hash-table))
;; well, first lets see if cpu load throttling is enabled. If so wait around until the
;; average cpu load is under the threshold before continuing
(if maxload ;; only gate if maxload is specified
(common:wait-for-cpuload maxload numcpus waitdelay))
(if maxhomehostload
(common:wait-for-homehost-load maxhomehostload (conc "Waiting for homehost load to drop below normalized value of " maxhomehostload)))
(run:test run-id run-info keyvals runname test-record flags #f test-registry all-tests-registry)
(runs:incremental-print-results run-id)
(hash-table-set! test-registry (db:test-make-full-name test-name item-path) 'running)
(runs:shrink-can-run-more-tests-count runsdat) ;; DELAY TWEAKER (still needed?)
;; (thread-sleep! *global-delta*)
(if (or (not (null? tal))(not (null? reg)))
(runs:loop-values tal reg reglen regfull reruns)
#f))
;; must be we have unmet prerequisites
;;
(else
(debug:print 4 *default-log-port* "FAILS: " fails)
;; If one or more of the prereqs-not-met are FAIL then we can issue
|
︙ | | | ︙ | |
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
|
(begin
;; couldn't run, take a breather
(if (runs:lownoise "Waiting for more work to do..." 60)
(debug:print-info 0 *default-log-port* "Waiting for more work to do..."))
(thread-sleep! 1)
(list (car newtal)(cdr newtal) reg reruns))
;; the waiton is FAIL so no point in trying to run hed ever again
(if (or (not (null? reg))(not (null? tal)))
(if (vector? hed)
(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 (mt:test-set-state-status-by-id run-id test-id "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
(mt:test-set-state-status-by-testname run-id test-name item-path "NOT_STARTED" "BLOCKED" #f)
(hash-table-set! test-registry (db:test-make-full-name test-name item-path) 'removed)
(list (runs:queue-next-hed tal reg reglen regfull)
(runs:queue-next-tal tal reg reglen regfull)
(runs:queue-next-reg tal reg reglen regfull)
reruns ;; WAS: (cons hed reruns) ;; but that makes no sense?
))
(let ((nth-try (hash-table-ref/default test-registry hed 0)))
(cond
((member "RUNNING" (map db:test-get-state prereqs-not-met))
(if (runs:lownoise (conc "possible RUNNING prerequistes " hed) 60)
(debug:print 0 *default-log-port* "WARNING: test " hed " has possible RUNNING prerequisites, don't give up on it yet."))
(thread-sleep! 4)
(list (runs:queue-next-hed newtal reg reglen regfull)
(runs:queue-next-tal newtal reg reglen regfull)
(runs:queue-next-reg newtal reg reglen regfull)
reruns))
((or (not nth-try)
(and (number? nth-try)
(< nth-try 10)))
(hash-table-set! test-registry hed (if (number? nth-try)
(+ nth-try 1)
0))
(if (runs:lownoise (conc "not removing test " hed) 60)
(debug:print 1 *default-log-port* "WARNING: not removing test " hed " from queue although it may not be runnable due to FAILED prerequisites"))
;; may not have processed correctly. Could be a race condition in your test implementation? Dropping test " hed) ;; " as it has prerequistes that are FAIL. (NOTE: hed is not a vector)")
(runs:shrink-can-run-more-tests-count runsdat) ;; DELAY TWEAKER (still needed?)
;; (list hed tal reg reruns)
;; (list (car newtal)(cdr newtal) reg reruns)
;; (hash-table-set! test-registry hed 'removed)
(list (runs:queue-next-hed newtal reg reglen regfull)
(runs:queue-next-tal newtal reg reglen regfull)
(runs:queue-next-reg newtal reg reglen regfull)
reruns))
((symbol? nth-try)
(if (eq? nth-try 'removed) ;; removed is removed - drop it NOW
(if (null? tal)
#f ;; yes, really
(list (car tal)(cdr tal) reg reruns))
(begin
(if (runs:lownoise (conc "FAILED prerequisites or other issue" hed) 60)
(debug:print 0 *default-log-port* "WARNING: test " hed " has FAILED prerequisites or other issue. Internal state " nth-try " will be overridden and we'll retry."))
(mt:test-set-state-status-by-testname run-id test-name item-path "NOT_STARTED" "KEEP_TRYING" #f)
(hash-table-set! test-registry hed 0)
(list (runs:queue-next-hed newtal reg reglen regfull)
(runs:queue-next-tal newtal reg reglen regfull)
(runs:queue-next-reg newtal reg reglen regfull)
reruns))))
(else
(if (runs:lownoise (conc "FAILED prerequitests and we tried" hed) 60)
(debug:print 0 *default-log-port* "WARNING: test " hed " has FAILED prerequitests and we've tried at least 10 times to run it. Giving up now."))
;; (debug:print 0 *default-log-port* " prereqs: " prereqs-not-met)
(hash-table-set! test-registry hed 'removed)
(mt:test-set-state-status-by-testname run-id test-name item-path "NOT_STARTED" "TEN_STRIKES" #f)
;; I'm unclear on if this roll up is needed - it may be the root cause of the "all set to FAIL" bug.
(rmt:set-state-status-and-roll-up-items run-id test-name item-path #f "FAIL" #f) ;; treat as FAIL
(list (if (null? tal)(car newtal)(car tal))
tal
reg
reruns)))))
;; can't drop this - maybe running? Just keep trying
(let ((runable-tests (runs:runable-tests prereqs-not-met)))
(if (null? runable-tests)
#f ;; I think we are truly done here
(list (runs:queue-next-hed newtal reg reglen regfull)
(runs:queue-next-tal newtal reg reglen regfull)
(runs:queue-next-reg newtal reg reglen regfull)
reruns)))))))))
;; scan a list of tests looking to see if any are potentially runnable
;;
(define (runs:runable-tests tests)
(filter (lambda (t)
(if (not (vector? t))
t
|
>
>
>
>
>
>
|
|
|
>
>
>
>
|
>
<
|
<
<
<
|
>
|
<
|
<
<
|
|
<
<
<
<
|
<
<
|
|
|
>
|
<
|
|
|
|
>
>
|
|
|
<
|
<
>
>
|
|
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
|
(begin
;; couldn't run, take a breather
(if (runs:lownoise "Waiting for more work to do..." 60)
(debug:print-info 0 *default-log-port* "Waiting for more work to do..."))
(thread-sleep! 1)
(list (car newtal)(cdr newtal) reg reruns))
;; the waiton is FAIL so no point in trying to run hed ever again
(begin
(let ((my-test-id (rmt:get-test-id run-id test-name item-path)))
(mt:test-set-state-status-by-id-unless-completed run-id my-test-id "COMPLETED" "PREQ_FAIL" "Failed to run due to failed prerequisites2"))
(if (or (not (null? reg))(not (null? tal)))
(if (vector? hed)
(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 (mt:test-set-state-status-by-id-unless-completed run-id test-id "COMPLETED" "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)
(rmt:set-state-status-and-roll-up-items run-id test-name item-path "NOT_STARTED" "BLOCKED" #f) )
(hash-table-set! test-registry (db:test-make-full-name test-name item-path) 'removed)
(runs:loop-values tal reg reglen regfull reruns))
(let ((nth-try (hash-table-ref/default test-registry hed 0))) ;; hed not a vector...
(debug:print 2 *default-log-port* "nth-try("hed")="nth-try)
(cond
((member "RUNNING" (map db:test-get-state prereqs-not-met))
(if (runs:lownoise (conc "possible RUNNING prerequistes " hed) 60)
(debug:print 0 *default-log-port* "WARNING: test " hed " has possible RUNNING prerequisites, don't give up on it yet."))
(thread-sleep! 0.1)
(runs:loop-values tal reg reglen regfull reruns))
((or (not nth-try) ;; BB: condition on subsequent tries, condition below fires on first try
(and (number? nth-try)
(< nth-try 2)))
(hash-table-set! test-registry hed (if (number? nth-try)
(+ nth-try 1)
0))
(if (runs:lownoise (conc "not removing test " hed) 60)
(debug:print 1 *default-log-port* "WARNING: not removing test " hed " from queue although it may not be runnable due to FAILED prerequisites"))
;; may not have processed correctly. Could be a race condition in your test implementation? Dropping test " hed) ;; " as it has prerequistes that are FAIL. (NOTE: hed is not a vector)")
(runs:shrink-can-run-more-tests-count runsdat) ;; DELAY TWEAKER (still needed?)
(runs:loop-values newtal reg reglen regfull reruns))
((symbol? nth-try) ;; BB: 'done matches here in one case where prereq itemwait failed. This is first "try"
(if (eq? nth-try 'removed) ;; removed is removed - drop it NOW
(if (null? tal)
#f ;; yes, really
(list (car tal)(cdr tal) reg reruns))
(begin
(if (runs:lownoise (conc "FAILED prerequisites or other issue" hed) 60)
(debug:print 0 *default-log-port* "WARNING: test " hed " has FAILED prerequisites or other issue. Internal state >" nth-try "< will be overridden and we'll retry."))
;; was: (mt:test-set-state-status-by-testname run-id test-name item-path "NOT_STARTED" "KEEP_TRYING" #f)
(mt:test-set-state-status-by-testname-unless-completed run-id test-name item-path "COMPLETED" "PREQ_FAIL" #f)
(hash-table-set! test-registry hed 'removed) ;; was 0
(if (not (and (null? reg) (null? tal)))
(runs:loop-values tal reg reglen regfull reruns)
#f))))
(else
(if (runs:lownoise (conc "FAILED prerequitests and we tried" hed) 60)
(debug:print 0 *default-log-port* "WARNING: test " hed " has FAILED prerequitests and we've tried at least 10 times to run it. Giving up now."))
;; (debug:print 0 *default-log-port* " prereqs: " prereqs-not-met)
(hash-table-set! test-registry hed 'removed)
(mt:test-set-state-status-by-testname run-id test-name item-path "NOT_STARTED" "TEN_STRIKES" #f)
;; I'm unclear on if this roll up is needed - it may be the root cause of the "all set to FAIL" bug.
(rmt:set-state-status-and-roll-up-items run-id test-name item-path #f "FAIL" #f) ;; treat as FAIL
(list (if (null? tal)(car newtal)(car tal))
tal
reg
reruns)))))
;; ELSE: can't drop this - maybe running? Just keep trying
;;(if (not (or (not (null? reg))(not (null? tal)))) ;; old experiment
(let ((runable-tests (runs:runable-tests prereqs-not-met))) ;; SUSPICIOUS: Should look at more than just prereqs-not-met?
(if (null? runable-tests)
#f ;; I think we are truly done here
(runs:loop-values newtal reg reglen regfull reruns)))
;;) ;;from old experiment
) ;; end if (or (not (null? reg))(not (null? tal)))
))))))
;; scan a list of tests looking to see if any are potentially runnable
;;
(define (runs:runable-tests tests)
(filter (lambda (t)
(if (not (vector? t))
t
|
︙ | | | ︙ | |
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
|
inc-results: (make-hash-table)
inc-results-last-update: 0
inc-results-fmt: "~12a~12a~20a~12a~40a\n" ;; state status time duration test-name item-path
run-info: #f
runname: #f
target: #f
)
)
(define (runs:incremental-print-results run-id)
(let ((curr-sec (current-seconds)))
(if (> (- curr-sec (runs:gendat-inc-results-last-update *runs:general-data*)) 5) ;; at least five seconds since last update
(let* ((run-dat (or (runs:gendat-run-info *runs:general-data*)(rmt:get-run-info run-id)))
(runname (or (runs:gendat-runname *runs:general-data*)
(db:get-value-by-header (db:get-rows run-dat)
|
|
|
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
|
inc-results: (make-hash-table)
inc-results-last-update: 0
inc-results-fmt: "~12a~12a~20a~12a~40a\n" ;; state status time duration test-name item-path
run-info: #f
runname: #f
target: #f
)
)
(define (runs:incremental-print-results run-id)
(let ((curr-sec (current-seconds)))
(if (> (- curr-sec (runs:gendat-inc-results-last-update *runs:general-data*)) 5) ;; at least five seconds since last update
(let* ((run-dat (or (runs:gendat-run-info *runs:general-data*)(rmt:get-run-info run-id)))
(runname (or (runs:gendat-runname *runs:general-data*)
(db:get-value-by-header (db:get-rows run-dat)
|
︙ | | | ︙ | |
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
|
(runs:gendat-inc-results-last-update-set! *runs:general-data* (- curr-sec 10))))
;; every time though the loop increment the test/itempatt val.
;; when the min is > max-allowed and none running then force exit
;;
(define *max-tries-hash* (make-hash-table))
;; test-records is a hash table testname:item_path => vector < testname testconfig waitons priority items-info ... >
(define (runs:run-tests-queue run-id runname test-records keyvals flags test-patts required-tests reglen-in all-tests-registry)
;; At this point the list of parent tests is expanded
;; NB// Should expand items here and then insert into the run queue.
(debug:print 5 *default-log-port* "test-records: " test-records ", flags: " (hash-table->alist flags))
;; Do mark-and-find clean up of db before starting runing of quue
;;
;; (rmt:find-and-mark-incomplete)
(let* ((run-info (rmt:get-run-info run-id))
(tests-info (mt:get-tests-for-run run-id #f '() '())) ;; qryvals: "id,testname,item_path"))
(sorted-test-names (tests:sort-by-priority-and-waiton test-records))
(test-registry (make-hash-table))
(registry-mutex (make-mutex))
(num-retries 0)
(max-retries (config-lookup *configdat* "setup" "maxretries"))
(max-concurrent-jobs (let ((mcj (config-lookup *configdat* "setup" "max_concurrent_jobs")))
(if (and mcj (string->number mcj))
(string->number mcj)
1))) ;; length of the register queue ahead
(reglen (if (number? reglen-in) reglen-in 1))
(last-time-incomplete (- (current-seconds) 900)) ;; force at least one clean up cycle
(last-time-some-running (current-seconds))
;; (tdbdat (tasks:open-db))
(runsdat (make-runs:dat
;; hed: hed
;; tal: tal
;; reg: reg
;; reruns: reruns
reglen: reglen
regfull: #f ;; regfull
;; test-record: test-record
runname: runname
;; test-name: test-name
;; item-path: item-path
;; jobgroup: jobgroup
max-concurrent-jobs: max-concurrent-jobs
run-id: run-id
;; waitons: waitons
;; testmode: testmode
test-patts: test-patts
required-tests: required-tests
test-registry: test-registry
registry-mutex: registry-mutex
flags: flags
keyvals: keyvals
run-info: run-info
;; newtal: newtal
all-tests-registry: all-tests-registry
;; itemmaps: itemmaps
;; prereqs-not-met: (rmt:get-prereqs-not-met run-id waitons hed item-path mode: testmode itemmaps: itemmaps)
;; can-run-more-tests: (runs:can-run-more-tests run-id jobgroup max-concurrent-jobs) ;; look at the test jobgroup and tot jobs running
)))
;; Initialize the test-registery hash with tests that already have a record
;; convert state to symbol and use that as the hash value
(for-each (lambda (trec)
(let ((id (db:test-get-id trec))
(tn (db:test-get-testname trec))
(ip (db:test-get-item-path trec))
(st (db:test-get-state trec)))
(if (not (equal? st "DELETED"))
(hash-table-set! test-registry (db:test-make-full-name tn ip) (string->symbol st)))))
tests-info)
(set! max-retries (if (and max-retries (string->number max-retries))(string->number max-retries) 100))
(let loop ((hed (car sorted-test-names))
(tal (cdr sorted-test-names))
(reg '()) ;; registered, put these at the head of tal
(reruns '()))
(runs:incremental-print-results run-id)
(if (not (null? reruns))(debug:print-info 4 *default-log-port* "reruns=" reruns))
|
>
>
>
>
|
|
|
|
|
|
|
|
<
<
<
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
|
(runs:gendat-inc-results-last-update-set! *runs:general-data* (- curr-sec 10))))
;; every time though the loop increment the test/itempatt val.
;; when the min is > max-allowed and none running then force exit
;;
(define *max-tries-hash* (make-hash-table))
;;======================================================================
;; runs:run-tests-queue is called by runs:run-tests
;;======================================================================
;;
;; test-records is a hash table testname:item_path => vector < testname testconfig waitons priority items-info ... >
(define (runs:run-tests-queue run-id runname test-records keyvals flags test-patts required-tests reglen-in all-tests-registry)
;; At this point the list of parent tests is expanded
;; NB// Should expand items here and then insert into the run queue.
(debug:print 5 *default-log-port* "test-records: " test-records ", flags: " (hash-table->alist flags))
;; Do mark-and-find clean up of db before starting runing of quue
;;
;; (rmt:find-and-mark-incomplete)
(let* ((run-info (rmt:get-run-info run-id))
(tests-info (mt:get-tests-for-run run-id #f '() '())) ;; qryvals: "id,testname,item_path"))
(sorted-test-names (tests:sort-by-priority-and-waiton test-records))
(test-registry (make-hash-table))
(registry-mutex (make-mutex))
(num-retries 0)
(max-retries (config-lookup *configdat* "setup" "maxretries"))
(max-concurrent-jobs (configf:lookup-number *configdat* "setup" "max_concurrent_jobs" default: 50))
(reglen (if (number? reglen-in) reglen-in 1))
(last-time-incomplete (- (current-seconds) 900)) ;; force at least one clean up cycle
(last-time-some-running (current-seconds))
;; (tdbdat (tasks:open-db))
(runsdat (make-runs:dat
;; hed: hed
;; tal: tal
;; reg: reg
;; reruns: reruns
reglen: reglen
regfull: #f ;; regfull
;; test-record: test-record
runname: runname
;; test-name: test-name
;; item-path: item-path
;; jobgroup: jobgroup
max-concurrent-jobs: max-concurrent-jobs
run-id: run-id
;; waitons: waitons
;; testmode: testmode
test-patts: test-patts
required-tests: required-tests
test-registry: test-registry
registry-mutex: registry-mutex
flags: flags
keyvals: keyvals
run-info: run-info
;; newtal: newtal
all-tests-registry: all-tests-registry
;; itemmaps: itemmaps
;; prereqs-not-met: (rmt:get-prereqs-not-met run-id waitons hed item-path mode: testmode itemmaps: itemmaps)
;; can-run-more-tests: (runs:can-run-more-tests run-id jobgroup max-concurrent-jobs) ;; look at the test jobgroup and tot jobs running
)))
;; Initialize the test-registery hash with tests that already have a record
;; convert state to symbol and use that as the hash value
(for-each (lambda (trec)
(let ((id (db:test-get-id trec))
(tn (db:test-get-testname trec))
(ip (db:test-get-item-path trec))
(st (db:test-get-state trec)))
(if (not (equal? st "DELETED"))
(hash-table-set! test-registry (db:test-make-full-name tn ip) (string->symbol st)))))
tests-info)
(set! max-retries (if (and max-retries (string->number max-retries))(string->number max-retries) 100))
(let loop ((hed (car sorted-test-names))
(tal (cdr sorted-test-names))
(reg '()) ;; registered, put these at the head of tal
(reruns '()))
(runs:incremental-print-results run-id)
(if (not (null? reruns))(debug:print-info 4 *default-log-port* "reruns=" reruns))
|
︙ | | | ︙ | |
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
|
(let* ((test-record (hash-table-ref test-records hed))
(test-name (tests:testqueue-get-testname test-record))
(tconfig (tests:testqueue-get-testconfig test-record))
(jobgroup (config-lookup tconfig "test_meta" "jobgroup"))
(testmode (let ((m (config-lookup tconfig "requirements" "mode")))
(if m (map string->symbol (string-split m)) '(normal))))
(itemmaps (tests:get-itemmaps tconfig)) ;; (configf:lookup tconfig "requirements" "itemmap"))
(waitons (tests:testqueue-get-waitons test-record))
(priority (tests:testqueue-get-priority test-record))
(itemdat (tests:testqueue-get-itemdat test-record)) ;; itemdat can be a string, list or #f
(items (tests:testqueue-get-items test-record))
(item-path (item-list->path itemdat))
(tfullname (db:test-make-full-name test-name item-path))
(newtal (append tal (list hed)))
(regfull (>= (length reg) reglen))
(num-running (rmt:get-count-tests-running-for-run-id run-id))
(testdat (make-runs:testdat
hed: hed
tal: tal
reg: reg
reruns: reruns
test-record: test-record
test-name: test-name
item-path: item-path
jobgroup: jobgroup
waitons: waitons
testmode: testmode
newtal: newtal
itemmaps: itemmaps
;; prereqs-not-met: prereqs-not-met
)))
(runs:dat-regfull-set! runsdat regfull)
;; every 15 minutes verify the server is there for this run
(if (and (common:low-noise-print 240 "try start server" run-id)
(not (server:check-if-running *toppath*)))
(server:kind-run *toppath*))
(if (> num-running 0)
(set! last-time-some-running (current-seconds)))
(if (> (current-seconds)(+ last-time-some-running (or (configf:lookup *configdat* "setup" "give-up-waiting") 36000)))
(hash-table-set! *max-tries-hash* tfullname (+ (hash-table-ref/default *max-tries-hash* tfullname 0) 1)))
;; (debug:print 0 *default-log-port* "max-tries-hash: " (hash-table->alist *max-tries-hash*))
;; Ensure all top level tests get registered. This way they show up as "NOT_STARTED" on the dashboard
;; and it is clear they *should* have run but did not.
(if (not (hash-table-ref/default test-registry (db:test-make-full-name test-name "") #f))
(begin
(rmt:register-test run-id test-name "")
|
<
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
|
>
|
>
>
>
|
|
|
|
|
|
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
|
(let* ((test-record (hash-table-ref test-records hed))
(test-name (tests:testqueue-get-testname test-record))
(tconfig (tests:testqueue-get-testconfig test-record))
(jobgroup (config-lookup tconfig "test_meta" "jobgroup"))
(testmode (let ((m (config-lookup tconfig "requirements" "mode")))
(if m (map string->symbol (string-split m)) '(normal))))
(itemmaps (tests:get-itemmaps tconfig)) ;; (configf:lookup tconfig "requirements" "itemmap"))
(priority (tests:testqueue-get-priority test-record))
(itemdat (tests:testqueue-get-itemdat test-record)) ;; itemdat can be a string, list or #f
(items (tests:testqueue-get-items test-record))
(item-path (item-list->path itemdat))
(tfullname (db:test-make-full-name test-name item-path))
;; these are hard coded item-item waits test/item-path => test/item-path2 ...
(extra-waits (let* ((section (configf:get-section (tests:testqueue-get-testconfig test-record) "waitons"))
(myextra (alist-ref tfullname section equal?)))
(if myextra
(let ((extras (string-split (car myextra))))
(if (runs:lownoise (conc tfullname "extra-waitons" tfullname) 60)
(debug:print-info 0 *default-log-port* "HAVE EXTRA WAITONS for test " tfullname ": " myextra))
(for-each
(lambda (extra)
;; (debug:print 0 *default-log-port* "FYI: extra = " extra " reruns = " reruns)
(let ((basetestname (car (string-split extra "/"))))
#;(if (not (member extra tal))
(set! reruns (append tal (list extra))))
(if (not (member basetestname tal))
(set! reruns (append tal (list basetestname))))
))
extras)
extras)
'())))
(waitons (delete-duplicates (append (tests:testqueue-get-waitons test-record) extra-waits) equal?))
(newtal (append tal (list hed)))
(regfull (>= (length reg) reglen))
(num-running (rmt:get-count-tests-running-for-run-id run-id))
(testdat (make-runs:testdat
hed: hed
tal: tal
reg: reg
reruns: reruns
test-record: test-record
test-name: test-name
item-path: item-path
jobgroup: jobgroup
waitons: waitons
testmode: testmode
newtal: newtal
itemmaps: itemmaps
;; prereqs-not-met: prereqs-not-met
)))
(runs:dat-regfull-set! runsdat regfull)
;; -- removed BB 17ww28 - no longer needed.
;; every 15 minutes verify the server is there for this run
;; (if (and (common:low-noise-print 240 "try start server" run-id)
;; (not (or (and *runremote*
;; (remote-server-url *runremote*)
;; (server:ping (remote-server-url *runremote*)))
;; (server:check-if-running *toppath*))))
;; (server:kind-run *toppath*))
(if (> num-running 0)
(set! last-time-some-running (current-seconds)))
(if (> (current-seconds)(+ last-time-some-running (or (configf:lookup *configdat* "setup" "give-up-waiting") 36000)))
(hash-table-set! *max-tries-hash* tfullname (+ (hash-table-ref/default *max-tries-hash* tfullname 0) 1)))
;; (debug:print 0 *default-log-port* "max-tries-hash: " (hash-table->alist *max-tries-hash*))
;; Ensure all top level tests get registered. This way they show up as "NOT_STARTED" on the dashboard
;; and it is clear they *should* have run but did not.
(if (not (hash-table-ref/default test-registry (db:test-make-full-name test-name "") #f))
(begin
(rmt:register-test run-id test-name "")
|
︙ | | | ︙ | |
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
|
(if (runs:lownoise (conc "been marked do not run " tfullname) 60)
(debug:print-info 0 *default-log-port* "Skipping test " tfullname " as it has been marked do not run due to being completed or not runnable"))
(if (or (not (null? tal))(not (null? reg)))
(loop (runs:queue-next-hed tal reg reglen regfull)
(runs:queue-next-tal tal reg reglen regfull)
(runs:queue-next-reg tal reg reglen regfull)
reruns))))
;; (loop (car tal)(cdr tal) reg reruns))))
(runs:incremental-print-results run-id)
(debug:print 4 *default-log-port* "TOP OF LOOP => "
"test-name: " test-name
"\n test-record " test-record
"\n hed: " hed
"\n itemdat: " itemdat
"\n items: " items
"\n item-path: " item-path
"\n waitons: " waitons
"\n num-retries: " num-retries
"\n tal: " tal
"\n reruns: " reruns
"\n regfull: " regfull
"\n reglen: " reglen
"\n length reg: " (length reg)
"\n reg: " reg)
;; check for hed in waitons => this would be circular, remove it and issue an
;; error
(if (member test-name waitons)
(begin
(debug:print-error 0 *default-log-port* "test " test-name " has listed itself as a waiton, please correct this!")
(set! waiton (filter (lambda (x)(not (equal? x hed))) waitons))))
|
|
<
>
>
>
|
<
|
<
>
|
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
|
(if (runs:lownoise (conc "been marked do not run " tfullname) 60)
(debug:print-info 0 *default-log-port* "Skipping test " tfullname " as it has been marked do not run due to being completed or not runnable"))
(if (or (not (null? tal))(not (null? reg)))
(loop (runs:queue-next-hed tal reg reglen regfull)
(runs:queue-next-tal tal reg reglen regfull)
(runs:queue-next-reg tal reg reglen regfull)
reruns))))
;; (loop (car tal)(cdr tal) reg reruns))))
(runs:incremental-print-results run-id)
(debug:print 4 *default-log-port* "TOP OF LOOP => "
"test-name: " test-name
"\n hed: " hed
"\n tal: " tal
"\n reg: " reg
"\n test-record " test-record
"\n itemdat: " itemdat
"\n items: " items
"\n item-path: " item-path
"\n waitons: " waitons
"\n num-retries: " num-retries
"\n reruns: " reruns
"\n regfull: " regfull
"\n reglen: " reglen
"\n length reg: " (length reg)
)
;; check for hed in waitons => this would be circular, remove it and issue an
;; error
(if (member test-name waitons)
(begin
(debug:print-error 0 *default-log-port* "test " test-name " has listed itself as a waiton, please correct this!")
(set! waiton (filter (lambda (x)(not (equal? x hed))) waitons))))
|
︙ | | | ︙ | |
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
|
number?
(map (lambda (waiton)
(if (and (not (member waiton tal)) ;; this waiton is not in the list to be tried to run
(not (member waiton reruns)))
1
#f))
waitons))))) ;; could do this more elegantly with a marker....
(debug:print 0 *default-log-port* "WARNING: Marking test " tfullname " as not runnable. It is waiting on tests that cannot be run. Giving up now.")
(hash-table-set! test-registry tfullname 'removed))
;; items is #f then the test is ok to be handed off to launch (but not before)
;;
((not items)
(debug:print-info 4 *default-log-port* "OUTER COND: (not items)")
(if (and (not (tests:match test-patts (tests:testqueue-get-testname test-record) item-path required: required-tests))
(not (null? tal)))
(loop (car tal)(cdr tal) reg reruns))
(runs:testdat-prereqs-not-met-set! testdat (rmt:get-prereqs-not-met run-id waitons hed item-path mode: testmode itemmaps: itemmaps))
(runs:dat-can-run-more-tests-set! runsdat (runs:can-run-more-tests runsdat run-id jobgroup max-concurrent-jobs)) ;; look at the test jobgroup and tot jobs running
(let ((loop-list (runs:process-expanded-tests runsdat testdat)))
(if loop-list (apply loop loop-list))))
;; items processed into a list but not came in as a list been processed
;;
((and (list? items) ;; thus we know our items are already calculated
(not itemdat)) ;; and not yet expanded into the list of things to be done
(debug:print-info 4 *default-log-port* "OUTER COND: (and (list? items)(not itemdat))")
;; Must determine if the items list is valid. Discard the test if it is not.
(if (and (list? items)
(> (length items) 0)
(and (list? (car items))
(> (length (car items)) 0))
(debug:debug-mode 1))
(debug:print 2 *default-log-port* (map (lambda (row)
(conc (string-intersperse
(map (lambda (varval)
(string-intersperse varval "="))
row)
" ")
"\n"))
items)))
(for-each
(lambda (my-itemdat)
(let* ((new-test-record (let ((newrec (make-tests:testqueue)))
(vector-copy! test-record newrec)
newrec))
(my-item-path (item-list->path my-itemdat)))
(if (tests:match test-patts hed my-item-path required: required-tests) ;; (patt-list-match my-item-path item-patts) ;; yes, we want to process this item, NOTE: Should not need this check here!
(let ((newtestname (db:test-make-full-name hed my-item-path))) ;; test names are unique on testname/item-path
(tests:testqueue-set-items! new-test-record #f)
(tests:testqueue-set-itemdat! new-test-record my-itemdat)
(tests:testqueue-set-item_path! new-test-record my-item-path)
(hash-table-set! test-records newtestname new-test-record)
(set! tal (append tal (list newtestname))))))) ;; since these are itemized create new test names testname/itempath
items)
;; (debug:print-info 0 *default-log-port* "Test " (tests:testqueue-get-testname test-record) " is itemized but has no items")
;; At this point we have possibly added items to tal but all must be handed off to
;; INNER COND logic. I think loop without rotating the queue
;; (loop hed tal reg reruns))
;; (let ((newtal (append tal (list hed)))) ;; We should discard hed as it has been expanded into it's items? Yes, but only if this *is* an itemized test
;; (loop (car newtal)(cdr newtal) reg reruns)
(if (null? tal)
#f
(loop (car tal)(cdr tal) reg reruns)))
;; if items is a proc then need to run items:get-items-from-config, get the list and loop
;; - but only do that if resources exist to kick off the job
;; EXPAND ITEMS
((or (procedure? items)(eq? items 'have-procedure))
(let ((can-run-more (runs:can-run-more-tests runsdat run-id jobgroup max-concurrent-jobs)))
(if (and (list? can-run-more)
(car can-run-more))
(let ((loop-list (runs:expand-items hed tal reg reruns regfull newtal jobgroup max-concurrent-jobs run-id waitons item-path testmode test-record can-run-more items runname tconfig reglen test-registry test-records itemmaps)))
(if loop-list
(apply loop loop-list)))
;; if can't run more just loop with next possible test
(loop (car newtal)(cdr newtal) reg reruns))))
;; this case should not happen, added to help catch any bugs
((and (list? items) itemdat)
(debug:print-error 0 *default-log-port* "Should not have a list of items in a test and the itemspath set - please report this")
(exit 1))
((not (null? reruns))
(let* ((newlst (tests:filter-non-runnable run-id tal test-records)) ;; i.e. not FAIL, WAIVED, INCOMPLETE, PASS, KILLED,
(junked (lset-difference equal? tal newlst)))
(debug:print-info 4 *default-log-port* "full drop through, if reruns is less than 100 we will force retry them, reruns=" reruns ", tal=" tal)
(if (< num-retries max-retries)
(set! newlst (append reruns newlst)))
(set! num-retries (+ num-retries 1))
;; (thread-sleep! (+ 1 *global-delta*))
(if (not (null? newlst))
;; since reruns have been tacked on to newlst create new reruns from junked
(loop (car newlst)(cdr newlst) reg (delete-duplicates junked)))))
((not (null? tal))
(debug:print-info 4 *default-log-port* "I'm pretty sure I shouldn't get here."))
((not (null? reg)) ;; could we get here with leftovers?
(debug:print-info 0 *default-log-port* "Have leftovers!")
(loop (car reg)(cdr reg) '() reruns))
(else
(debug:print-info 4 *default-log-port* "Exiting loop with...\n hed=" hed "\n tal=" tal "\n reruns=" reruns))
)))
;; now *if* -run-wait we wait for all tests to be done
;; Now wait for any RUNNING tests to complete (if in run-wait mode)
(thread-sleep! 5) ;; I think there is a race condition here. Let states/statuses settle
(let wait-loop ((num-running (rmt:get-count-tests-running-for-run-id run-id))
(prev-num-running 0))
;; (debug:print-info 13 *default-log-port* "num-running=" num-running ", prev-num-running=" prev-num-running)
(if (and (or (args:get-arg "-run-wait")
(equal? (configf:lookup *configdat* "setup" "run-wait") "yes"))
(> num-running 0))
(begin
|
>
>
|
>
|
|
|
|
|
|
|
|
>
>
|
>
>
>
>
>
>
>
>
>
>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
>
|
|
>
>
>
|
>
>
>
>
>
>
>
>
|
|
|
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
|
number?
(map (lambda (waiton)
(if (and (not (member waiton tal)) ;; this waiton is not in the list to be tried to run
(not (member waiton reruns)))
1
#f))
waitons))))) ;; could do this more elegantly with a marker....
(debug:print-info 4 *default-log-port* "cond branch - " "rtq-1")
(debug:print 0 *default-log-port* "WARNING: Marking test " tfullname " as not runnable. It is waiting on tests that cannot be run. Giving up now.")
(hash-table-set! test-registry tfullname 'removed))
;; items is #f then the test is ok to be handed off to launch (but not before)
;;
((not items)
(debug:print-info 4 *default-log-port* "cond branch - " "rtq-2")
(debug:print-info 4 *default-log-port* "OUTER COND: (not items)")
(if (and (not (tests:match test-patts (tests:testqueue-get-testname test-record) item-path required: required-tests))
(not (null? tal)))
(loop (car tal)(cdr tal) reg reruns))
(runs:testdat-prereqs-not-met-set! testdat (rmt:get-prereqs-not-met run-id waitons hed item-path mode: testmode itemmaps: itemmaps))
(runs:dat-can-run-more-tests-set! runsdat (runs:can-run-more-tests runsdat run-id jobgroup max-concurrent-jobs)) ;; look at the test jobgroup and tot jobs running
(let ((loop-list (runs:process-expanded-tests runsdat testdat)))
(if loop-list (apply loop loop-list))))
;; items processed into a list but not came in as a list been processed
;;
((and (list? items) ;; thus we know our items are already calculated
(not itemdat)) ;; and not yet expanded into the list of things to be done
(debug:print-info 4 *default-log-port* "cond branch - " "rtq-3")
(debug:print-info 4 *default-log-port* "OUTER COND: (and (list? items)(not itemdat))")
;; Must determine if the items list is valid. Discard the test if it is not.
(if (and (list? items)
(> (length items) 0)
(and (list? (car items))
(> (length (car items)) 0))
(debug:debug-mode 1))
(debug:print 2 *default-log-port* (map (lambda (row)
(conc (string-intersperse
(map (lambda (varval)
(string-intersperse varval "="))
row)
" ")
"\n"))
items)))
(let* ((items-in-testpatt
(filter
(lambda (my-itemdat)
(tests:match test-patts hed (item-list->path my-itemdat) ))
;; was: (tests:match test-patts hed (item-list->path my-itemdat) required: required-tests))
items) ))
(if (null? items-in-testpatt)
(let ((test-id (rmt:get-test-id run-id test-name "")))
(debug:print-info 0 *default-log-port* "Test " (tests:testqueue-get-testname test-record) " is itemized but has no items matching test pattern -- marking status ZERO_ITEMS")
(if test-id
(mt:test-set-state-status-by-id run-id test-id "NOT_STARTED" "ZERO_ITEMS" "This test has no items which match test pattern.")))
(for-each (lambda (my-itemdat)
(let* ((new-test-record (let ((newrec (make-tests:testqueue)))
(vector-copy! test-record newrec)
newrec))
(my-item-path (item-list->path my-itemdat))
(newtestname (db:test-make-full-name hed my-item-path))) ;; test names are unique on testname/item-path
(tests:testqueue-set-items! new-test-record #f)
(tests:testqueue-set-itemdat! new-test-record my-itemdat)
(tests:testqueue-set-item_path! new-test-record my-item-path)
(hash-table-set! test-records newtestname new-test-record)
(set! tal (append tal (list newtestname))))) ;; since these are itemized create new test names testname/itempath
items-in-testpatt)))
;; At this point we have possibly added items to tal but all must be handed off to
;; INNER COND logic. I think loop without rotating the queue
;; (loop hed tal reg reruns))
;; (let ((newtal (append tal (list hed)))) ;; We should discard hed as it has been expanded into it's items? Yes, but only if this *is* an itemized test
;; (loop (car newtal)(cdr newtal) reg reruns)
(if (null? tal)
#f
(loop (car tal)(cdr tal) reg reruns)))
;; if items is a proc then need to run items:get-items-from-config, get the list and loop
;; - but only do that if resources exist to kick off the job
;; EXPAND ITEMS
((or (procedure? items)(eq? items 'have-procedure))
(debug:print-info 4 *default-log-port* "cond branch - " "rtq-4")
(let ((can-run-more (runs:can-run-more-tests runsdat run-id jobgroup max-concurrent-jobs)))
(if (and (list? can-run-more)
(car can-run-more))
(let ((loop-list (runs:expand-items hed tal reg reruns regfull newtal jobgroup max-concurrent-jobs run-id waitons item-path testmode test-record can-run-more items runname tconfig reglen test-registry test-records itemmaps))) ;; itemized test expanded here
(if loop-list
(apply loop loop-list)
(debug:print-info 4 *default-log-port* " -- Can't expand hed="hed)
)
)
;; if can't run more just loop with next possible test
(loop (car newtal)(cdr newtal) reg reruns))))
;; this case should not happen, added to help catch any bugs
((and (list? items) itemdat)
(debug:print-info 4 *default-log-port* "cond branch - " "rtq-5")
(debug:print-error 0 *default-log-port* "Should not have a list of items in a test and the itemspath set - please report this")
(exit 1))
((not (null? reruns))
(debug:print-info 4 *default-log-port* "cond branch - " "rtq-6")
(let* ((newlst (tests:filter-non-runnable run-id tal test-records)) ;; i.e. not FAIL, WAIVED, INCOMPLETE, PASS, KILLED,
(junked (lset-difference equal? tal newlst)))
(debug:print-info 4 *default-log-port* "full drop through, if reruns is less than 100 we will force retry them, reruns=" reruns ", tal=" tal)
(if (< num-retries max-retries)
(set! newlst (append reruns newlst)))
(set! num-retries (+ num-retries 1))
;; (thread-sleep! (+ 1 *global-delta*))
(if (not (null? newlst))
;; since reruns have been tacked on to newlst create new reruns from junked
(loop (car newlst)(cdr newlst) reg (delete-duplicates junked)))))
((not (null? tal))
(debug:print-info 4 *default-log-port* "cond branch - " "rtq-7")
(debug:print-info 4 *default-log-port* "I'm pretty sure I shouldn't get here."))
((not (null? reg)) ;; could we get here with leftovers?
(debug:print-info 4 *default-log-port* "cond branch - " "rtq-8")
(debug:print-info 0 *default-log-port* "Have leftovers!")
(loop (car reg)(cdr reg) '() reruns))
(else
(debug:print-info 4 *default-log-port* "cond branch - " "rtq-9")
(debug:print-info 4 *default-log-port* "Exiting loop with...\n hed=" hed "\n tal=" tal "\n reruns=" reruns))
))) ;; end loop on sorted test names
;; this is the point where everything is launced and now you can mark the run in metadata table as all launced
(rmt:set-var (conc "lunch-complete-" run-id) "yes")
;; now *if* -run-wait we wait for all tests to be done
;; Now wait for any RUNNING tests to complete (if in run-wait mode)
(thread-sleep! 10) ;; I think there is a race condition here. Let states/statuses settle
(let wait-loop ((num-running (rmt:get-count-tests-running-for-run-id run-id))
(prev-num-running 0))
;; (debug:print-info 13 *default-log-port* "num-running=" num-running ", prev-num-running=" prev-num-running)
(if (and (or (args:get-arg "-run-wait")
(equal? (configf:lookup *configdat* "setup" "run-wait") "yes"))
(> num-running 0))
(begin
|
︙ | | | ︙ | |
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
|
(if (not (eq? num-running prev-num-running))
(debug:print-info 0 *default-log-port* "run-wait specified, waiting on " num-running " tests in RUNNING, REMOTEHOSTSTART or LAUNCHED state at " (time->string (seconds->local-time (current-seconds)))))
(thread-sleep! 5)
;; (wait-loop (rmt:get-count-tests-running-for-run-id run-id) num-running))))
(wait-loop (rmt:get-count-tests-running-for-run-id run-id) num-running))))
;; LET* ((test-record
;; we get here on "drop through". All done!
(debug:print-info 1 *default-log-port* "All tests launched")))
(define (runs:calc-fails prereqs-not-met)
(filter (lambda (test)
(and (vector? test) ;; not (string? test))
(member (db:test-get-state test) '("INCOMPLETE" "COMPLETED"))
(not (member (db:test-get-status test)
'("PASS" "WARN" "CHECK" "WAIVED" "SKIP")))))
prereqs-not-met))
(define (runs:calc-prereq-fail prereqs-not-met)
(filter (lambda (test)
(and (vector? test) ;; not (string? test))
(equal? (db:test-get-state test) "NOT_STARTED")
(not (member (db:test-get-status test)
'("n/a" "KEEP_TRYING")))))
prereqs-not-met))
|
>
>
>
|
|
|
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
|
(if (not (eq? num-running prev-num-running))
(debug:print-info 0 *default-log-port* "run-wait specified, waiting on " num-running " tests in RUNNING, REMOTEHOSTSTART or LAUNCHED state at " (time->string (seconds->local-time (current-seconds)))))
(thread-sleep! 5)
;; (wait-loop (rmt:get-count-tests-running-for-run-id run-id) num-running))))
(wait-loop (rmt:get-count-tests-running-for-run-id run-id) num-running))))
;; LET* ((test-record
;; we get here on "drop through". All done!
;; this is moved to runs:run-testes since this function is getting called twice to ensure everthing is completed.
;; (debug:print-info 0 *default-log-port* "Calling Post Hook")
;; (runs:run-post-hook run-id)
(debug:print-info 1 *default-log-port* "All tests launched")))
(define (runs:calc-fails prereqs-not-met)
(filter (lambda (test)
(and (vector? test) ;; not (string? test))
(member (db:test-get-state test) '("INCOMPLETE" "COMPLETED")) ;; TODO: pull from *common:stuff...*
(not (member (db:test-get-status test)
'("PASS" "WARN" "CHECK" "WAIVED" "SKIP")))))
prereqs-not-met))
(define (runs:calc-prereq-fail prereqs-not-met) ;; REMOVEME since NOT_STARTED/PREQ_FAIL is now COMPLETED/PREQ_FAIL
(filter (lambda (test)
(and (vector? test) ;; not (string? test))
(equal? (db:test-get-state test) "NOT_STARTED")
(not (member (db:test-get-status test)
'("n/a" "KEEP_TRYING")))))
prereqs-not-met))
|
︙ | | | ︙ | |
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
|
;;
;; There is now a single call to runs:update-all-test_meta and this
;; per-test call is not needed. Given the delicacy of the move to
;; v1.55 this code is being left in place for the time being.
;;
(if (not (hash-table-ref/default *test-meta-updated* test-name #f))
(begin
(hash-table-set! *test-meta-updated* test-name #t)
(runs:update-test_meta test-name test-conf)))
;; itemdat => ((ripeness "overripe") (temperature "cool") (season "summer"))
(let* ((new-test-path (string-intersperse (cons test-path (map cadr itemdat)) "/"))
(test-id (rmt:get-test-id run-id test-name item-path))
(testdat (if test-id (rmt:get-test-info-by-id run-id test-id) #f)))
(if (not testdat)
(let loop ()
|
|
|
|
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
|
;;
;; There is now a single call to runs:update-all-test_meta and this
;; per-test call is not needed. Given the delicacy of the move to
;; v1.55 this code is being left in place for the time being.
;;
(if (not (hash-table-ref/default *test-meta-updated* test-name #f))
(begin
(hash-table-set! *test-meta-updated* test-name #t)
(runs:update-test_meta test-name test-conf)))
;; itemdat => ((ripeness "overripe") (temperature "cool") (season "summer"))
(let* ((new-test-path (string-intersperse (cons test-path (map cadr itemdat)) "/"))
(test-id (rmt:get-test-id run-id test-name item-path))
(testdat (if test-id (rmt:get-test-info-by-id run-id test-id) #f)))
(if (not testdat)
(let loop ()
|
︙ | | | ︙ | |
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
|
(begin
(debug:print-info 0 *default-log-port* "WARNING: server is overloaded, trying again in one second")
(thread-sleep! 1)
(loop)))))
(if (not testdat) ;; should NOT happen
(debug:print-error 0 *default-log-port* "failed to get test record for test-id " test-id))
(set! test-id (db:test-get-id testdat))
(if (file-exists? test-path)
(change-directory test-path)
(begin
(debug:print-error 0 *default-log-port* "test run path not created before attempting to run the test. Perhaps you are running -remove-runs at the same time?")
(change-directory *toppath*)))
(case (if force ;; (args:get-arg "-force")
'NOT_STARTED
(if testdat
|
|
|
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
|
(begin
(debug:print-info 0 *default-log-port* "WARNING: server is overloaded, trying again in one second")
(thread-sleep! 1)
(loop)))))
(if (not testdat) ;; should NOT happen
(debug:print-error 0 *default-log-port* "failed to get test record for test-id " test-id))
(set! test-id (db:test-get-id testdat))
(if (common:file-exists? test-path)
(change-directory test-path)
(begin
(debug:print-error 0 *default-log-port* "test run path not created before attempting to run the test. Perhaps you are running -remove-runs at the same time?")
(change-directory *toppath*)))
(case (if force ;; (args:get-arg "-force")
'NOT_STARTED
(if testdat
|
︙ | | | ︙ | |
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
|
(if (not (string-search (regexp "testdat.db") f))
(runs:recursive-delete-with-error-msg fullname)))
(+ 1 x))
0 real-dir)
;; then the entire directory
(runs:recursive-delete-with-error-msg real-dir))
;; Remove runs
;; fields are passing in through
;; action:
;; 'remove-runs
;; 'set-state-status
;;
;; NB// should pass in keys?
;;
(define (runs:operate-on action target runnamepatt testpatt #!key (state #f)(status #f)(new-state-status #f)(mode 'remove-all)(options '()))
(common:clear-caches) ;; clear all caches
(let* ((db #f)
;; (tdbdat (tasks:open-db))
(keys (rmt:get-keys))
(rundat (mt:get-runs-by-patt keys runnamepatt target))
(header (vector-ref rundat 0))
(runs (vector-ref rundat 1))
(states (if state (string-split state ",") '()))
(statuses (if status (string-split status ",") '()))
(state-status (if (string? new-state-status) (string-split new-state-status ",") '(#f #f)))
(rp-mutex (make-mutex))
(bup-mutex (make-mutex)))
(let* ((write-access-actions '(remove-runs set-state-status archive run-wait))
(dbfile (conc *toppath* "/megatest.db"))
(readonly-mode (not (file-write-access? dbfile))))
(when (and readonly-mode
(member action write-access-actions))
(debug:print-error 0 *default-log-port* "megatest.db is readonly. Cannot proceed with action ["action"] in which write-access isrequired .")
(exit 1)))
(debug:print-info 4 *default-log-port* "runs:operate-on => Header: " header " action: " action " new-state-status: " new-state-status)
(if (> 2 (length state-status))
(begin
(debug:print-error 0 *default-log-port* "the parameter to -set-state-status is a comma delimited string. E.g. COMPLETED,FAIL")
(exit)))
(for-each
(lambda (run)
(let ((runkey (string-intersperse (map (lambda (k)
(db:get-value-by-header run header k)) keys) "/"))
(dirs-to-remove (make-hash-table))
(proc-get-tests (lambda (run-id)
(mt:get-tests-for-run run-id
testpatt states statuses
not-in: #f
sort-by: (case action
((remove-runs) 'rundir)
(else 'event_time))))))
(let* ((run-id (db:get-value-by-header run header "id"))
(run-state (db:get-value-by-header run header "state"))
(run-name (db:get-value-by-header run header "runname"))
(tests (if (not (equal? run-state "locked"))
(proc-get-tests run-id)
'()))
(lasttpath "/does/not/exist/I/hope")
(worker-thread #f))
(debug:print-info 4 *default-log-port* "runs:operate-on run=" run ", header=" header)
(if (not (null? tests))
(begin
(case action
((remove-runs)
;; (if (tasks:need-server run-id)(tasks:start-and-wait-for-server tdbdat run-id 10))
;; seek and kill in flight -runtests with % as testpatt here
;; (if (equal? testpatt "%")
(tasks:kill-runner target run-name testpatt)
;; (debug:print 0 *default-log-port* "not attempting to kill any run launcher processes as testpatt is " testpatt))
(debug:print 1 *default-log-port* "Removing tests for run: " runkey " " (db:get-value-by-header run header "runname")))
((set-state-status)
;; (if (tasks:need-server run-id)(tasks:start-and-wait-for-server tdbdat run-id 10))
(debug:print 1 *default-log-port* "Modifying state and staus for tests for run: " runkey " " (db:get-value-by-header run header "runname")))
((print-run)
(debug:print 1 *default-log-port* "Printing info for run " runkey ", run=" run ", tests=" tests ", header=" header)
action)
((run-wait)
(debug:print 1 *default-log-port* "Waiting for run " runkey ", run=" runnamepatt " to complete"))
((archive)
(debug:print 1 *default-log-port* "Archiving/restoring (" (args:get-arg "-archive") ") data for run: " runkey " " (db:get-value-by-header run header "runname"))
(set! worker-thread (make-thread (lambda ()
(case (string->symbol (args:get-arg "-archive"))
((save save-remove keep-html)(archive:run-bup (args:get-arg "-archive") run-id run-name tests rp-mutex bup-mutex))
((restore)(archive:bup-restore (args:get-arg "-archive") run-id run-name tests rp-mutex bup-mutex))
(else
(debug:print-error 0 *default-log-port* "unrecognised sub command to -archive. Run \"megatest\" to see help")
(exit))))
"archive-bup-thread"))
(thread-start! worker-thread))
(else
(debug:print-info 0 *default-log-port* "action not recognised " action)))
;; actions that operate on one test at a time can be handled below
;;
(let ((sorted-tests (filter
vector?
(sort tests (lambda (a b)(let ((dira ;; (rmt:sdb-qry 'getstr
(db:test-get-rundir a)) ;; ) ;; (filedb:get-path *fdb* (db:test-get-rundir a)))
(dirb ;; (rmt:sdb-qry 'getstr
(db:test-get-rundir b))) ;; ) ;; ((filedb:get-path *fdb* (db:test-get-rundir b))))
(if (and (string? dira)(string? dirb))
(> (string-length dira)(string-length dirb))
#f))))))
(toplevel-retries (make-hash-table)) ;; try three times to loop through and remove top level tests
(test-retry-time (make-hash-table))
(allow-run-time 10)) ;; seconds to allow for killing tests before just brutally killing 'em
(let loop ((test (car sorted-tests))
(tal (cdr sorted-tests)))
(let* ((test-id (db:test-get-id test))
(new-test-dat (rmt:get-test-info-by-id run-id test-id)))
(if (not new-test-dat)
(begin
(debug:print-error 0 *default-log-port* "We have a test-id of " test-id " but no record was found. NOTE: No locking of records is done between processes, do not simultaneously remove the same run from two processes!")
(if (not (null? tal))
(loop (car tal)(cdr tal))))
(let* ((item-path (db:test-get-item-path new-test-dat))
(test-name (db:test-get-testname new-test-dat))
(run-dir ;;(filedb:get-path *fdb*
;; (rmt:sdb-qry 'getid
(db:test-get-rundir new-test-dat)) ;; ) ;; run dir is from the link tree
(test-state (db:test-get-state new-test-dat))
(test-fulln (db:test-get-fullname new-test-dat))
(uname (db:test-get-uname new-test-dat))
(toplevel-with-children (and (db:test-get-is-toplevel test)
(> (rmt:test-toplevel-num-items run-id test-name) 0))))
(case action
((remove-runs)
;; if the test is a toplevel-with-children issue an error and do not remove
(if toplevel-with-children
(begin
(debug:print 0 *default-log-port* "WARNING: skipping removal of " test-fulln " with run-id " run-id " as it has sub tests")
(hash-table-set! toplevel-retries test-fulln (+ (hash-table-ref/default toplevel-retries test-fulln 0) 1))
(if (> (hash-table-ref toplevel-retries test-fulln) 3)
(if (not (null? tal))
(loop (car tal)(cdr tal))) ;; no else clause - drop it if no more in queue and > 3 tries
(let ((newtal (append tal (list test))))
(loop (car newtal)(cdr newtal))))) ;; loop with test still in queue
(begin
(debug:print-info 0 *default-log-port* "test: " test-name " itest-state: " test-state)
(if (member test-state (list "RUNNING" "LAUNCHED" "REMOTEHOSTSTART" "KILLREQ"))
(begin
(if (not (hash-table-ref/default test-retry-time test-fulln #f))
(begin
;; want to set to REMOVING BUT CANNOT do it here?
(hash-table-set! test-retry-time test-fulln (current-seconds))))
(if (> (- (current-seconds)(hash-table-ref test-retry-time test-fulln)) allow-run-time)
;; This test is not in a correct state for cleaning up. Let's try some graceful shutdown steps first
;; Set the test to "KILLREQ" and wait five seconds then try again. Repeat up to five times then give
;; up and blow it away.
(begin
(debug:print 0 *default-log-port* "WARNING: could not gracefully remove test " test-fulln ", tried to kill it to no avail. Forcing state to FAILEDKILL and continuing")
(mt:test-set-state-status-by-id run-id (db:test-get-id test) "FAILEDKILL" "n/a" #f)
(thread-sleep! 1))
(begin
(mt:test-set-state-status-by-id run-id (db:test-get-id test) "KILLREQ" "n/a" #f)
(thread-sleep! 1)))
;; NOTE: This is suboptimal as the testdata will be used later and the state/status may have changed ...
(if (null? tal)
(loop new-test-dat tal)
(loop (car tal)(append tal (list new-test-dat)))))
(begin
(runs:remove-test-directory new-test-dat mode) ;; 'remove-all)
(if (not (null? tal))
(loop (car tal)(cdr tal)))))))
(rmt:update-run-stats run-id (rmt:get-raw-run-stats run-id)))
((set-state-status)
(debug:print-info 2 *default-log-port* "new state " (car state-status) ", new status " (cadr state-status))
(mt:test-set-state-status-by-id run-id (db:test-get-id test) (car state-status)(cadr state-status) #f)
(if (not (null? tal))
(loop (car tal)(cdr tal))))
((run-wait)
(debug:print-info 2 *default-log-port* "still waiting, " (length tests) " tests still running")
(thread-sleep! 10)
(let ((new-tests (proc-get-tests run-id)))
(if (null? new-tests)
(debug:print-info 1 *default-log-port* "Run completed according to zero tests matching provided criteria.")
(loop (car new-tests)(cdr new-tests)))))
((archive)
(if (and run-dir (not toplevel-with-children))
(let ((ddir (conc run-dir "/")))
(case (string->symbol (args:get-arg "-archive"))
((save save-remove keep-html)
(if (file-exists? ddir)
(debug:print-info 0 *default-log-port* "Estimating disk space usage for " test-fulln ": " (common:get-disk-space-used ddir)))))))
(if (not (null? tal))
(loop (car tal)(cdr tal))))
)))
)
(if worker-thread (thread-join! worker-thread))))))
;; remove the run if zero tests remain
(if (eq? action 'remove-runs)
(let ((remtests (mt:get-tests-for-run (db:get-value-by-header run header "id") #f '("DELETED") '("n/a") not-in: #t)))
(if (null? remtests) ;; no more tests remaining
(let* ((dparts (string-split lasttpath "/"))
(runpath (conc "/" (string-intersperse
(take dparts (- (length dparts) 1))
"/"))))
(debug:print 1 *default-log-port* "Removing run: " runkey " " (db:get-value-by-header run header "runname") " and related record")
(rmt:delete-run run-id)
(rmt:delete-old-deleted-test-records)
;; (rmt:set-var "DELETED_TESTS" (current-seconds))
;; need to figure out the path to the run dir and remove it if empty
;; (if (null? (glob (conc runpath "/*")))
;; (begin
;; (debug:print 1 *default-log-port* "Removing run dir " runpath)
;; (system (conc "rmdir -p " runpath))))
)))))
))
runs)
;; (sqlite3:finalize! (db:delay-if-busy tdbdat))
)
#t)
(define (runs:remove-test-directory test mode) ;; remove-data-only)
(let* ((run-dir (db:test-get-rundir test)) ;; run dir is from the link tree
(real-dir (if (file-exists? run-dir)
;; (resolve-pathname run-dir)
(common:nice-path run-dir)
#f)))
(case mode
((remove-data-only)(mt:test-set-state-status-by-id (db:test-get-run_id test)(db:test-get-id test) "CLEANING" "LOCKED" #f))
((remove-all) (mt:test-set-state-status-by-id (db:test-get-run_id test)(db:test-get-id test) "REMOVING" "LOCKED" #f))
((archive-remove) (mt:test-set-state-status-by-id (db:test-get-run_id test)(db:test-get-id test) "ARCHIVE_REMOVING" #f #f)))
(debug:print-info 1 *default-log-port* "Attempting to remove " (if real-dir (conc " dir " real-dir " and ") "") " link " run-dir)
(if (and real-dir
(> (string-length real-dir) 5)
(file-exists? real-dir)) ;; bad heuristic but should prevent /tmp /home etc.
(begin ;; let* ((realpath (resolve-pathname run-dir)))
(debug:print-info 1 *default-log-port* "Recursively removing " real-dir)
(if (file-exists? real-dir)
(runs:safe-delete-test-dir real-dir)
(debug:print 0 *default-log-port* "WARNING: test dir " real-dir " appears to not exist or is not readable")))
(if real-dir
(debug:print 0 *default-log-port* "WARNING: directory " real-dir " does not exist")
(debug:print 0 *default-log-port* "WARNING: no real directory corrosponding to link " run-dir ", nothing done")))
(if (symbolic-link? run-dir)
(begin
|
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
|
|
>
|
<
|
|
|
|
|
|
>
>
>
>
<
|
>
>
|
>
|
>
|
|
|
|
|
>
>
>
|
>
>
>
>
>
|
<
|
|
|
|
|
|
|
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
|
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
|
|
>
>
|
|
>
>
|
>
>
|
|
|
|
|
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
|
|
|
|
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
|
(if (not (string-search (regexp "testdat.db") f))
(runs:recursive-delete-with-error-msg fullname)))
(+ 1 x))
0 real-dir)
;; then the entire directory
(runs:recursive-delete-with-error-msg real-dir))
;; cleanup often needs to remove all but the last N runs per target
;;
;; target-patts a1/b1/c1,a2/b2/c2 ...
;;
;; This will fail if called with empty target or a bad target (i.e. missing or extra fields)
;;
(define (runs:get-hash-by-target target-patts runpatt)
(let* ((targets (string-split target-patts ","))
(keys (rmt:get-keys))
(res-ht (make-hash-table))) ;; target -> ( runrecord1 runrecord2 ... )
(for-each
(lambda (target-patt)
(let ((runs (rmt:simple-get-runs runpatt #f #f target-patt)))
(for-each
(lambda (run)
(let ((target (simple-run-target run)))
(hash-table-set! res-ht target (cons run (hash-table-ref/default res-ht target '())))))
runs)))
targets)
res-ht))
;; delete runs older than X (weeks, days, months years etc.)
;; delete redundant runs within a target - N is the input
;; delete redundant runs within a target IFF older than given date/time AND keep at least N
;;
(define (runs:remove-all-but-last-n-runs-per-target target-patts runpatt num-to-keep #!key (actions '(print)))
(let* ((runs-ht (runs:get-hash-by-target target-patts runpatt))
(age (if (args:get-arg "-age")(common:hms-string->seconds (args:get-arg "-age")) #f))
(age-mark (if age (- (current-seconds) age) (+ (current-seconds) 86400)))
(precmd (or (args:get-arg "-precmd") "")))
(print "Actions: " actions)
(for-each
(lambda (target)
(let* ((runs (hash-table-ref runs-ht target))
(sorted (sort runs (lambda (a b)(< (simple-run-event_time a)(simple-run-event_time b)))))
(to-remove (let* ((len (length sorted))
(trim-amt (- len num-to-keep)))
(if (> trim-amt 0)
(take sorted trim-amt)
'()))))
(hash-table-set! runs-ht target to-remove)
(print target ":")
(for-each
(lambda (run)
(let ((remove (member run to-remove (lambda (a b)
(eq? (simple-run-id a)
(simple-run-id b))))))
(if (and age (> (simple-run-event_time run) age-mark))
(print "Skipping handling of " target "/" (simple-run-runname run) " as it is younger than " (args:get-arg "-age"))
(for-each
(lambda (action)
(case action
((print)
(print " " (simple-run-runname run)
" " (time->string (seconds->local-time (simple-run-event_time run)) "WW%V.%u %H:%M:%S")
" " (if remove "REMOVE" "")))
((remove-runs)
(if remove (system (conc precmd " megatest -remove-runs -target " target " -runname " (simple-run-runname run) " -testpatt %"
(if (member 'kill-runs actions) ;; if kill-runs is specified then set -kill-wait to 0
" -kill-wait 0"
"")))))
((archive)
(if remove (system (conc precmd " megatest -archive save-remove -target " target " -runname " (simple-run-runname run) " -testpatt %"))))
((kill-runs)
(if remove (system (conc precmd " megatest -kill-runs -target " target " -runname " (simple-run-runname run) " -testpatt %"))))
))
actions))))
sorted)))
;; (print "Sorted: " (map simple-run-event_time sorted))
;; (print "Remove: " (map simple-run-event_time to-remove))))
(hash-table-keys runs-ht))
runs-ht))
;; (define (runs:remove-all-but-last-n-runs-per-target target-patts runpatt num-to-keep)
;; (let ((data (runs:get-all-but-most-recent-n-per-target target-patts runpatt num-to-keep)))
;; (for-each
;; (lambda (target)
;; (let ((runs-to-remove (hash-table-ref data target )))
;; (for-each
;; (lambda (run)
;; (print "megatest -remove-runs -target " target " -runname " (simple-run-runname run) " -testpatt %"))
;; runs-to-remove)))
;; (hash-table-keys data))))
;; Remove runs
;; fields are passing in through
;; action:
;; 'remove-runs
;; 'set-state-status
;;
;; NB// should pass in keys?
;;
(define (runs:operate-on action target runnamepatt testpatt #!key (state #f)(status #f)(new-state-status #f)(mode #f)(options '()))
(common:clear-caches) ;; clear all caches
(let* ((db #f)
;; (tdbdat (tasks:open-db))
(keys (rmt:get-keys))
(rundat (mt:get-runs-by-patt keys runnamepatt target))
(header (vector-ref rundat 0))
(runs (vector-ref rundat 1))
(states (if state (string-split state ",") '()))
(statuses (if status (string-split status ",") '()))
(state-status (if (string? new-state-status) (string-split new-state-status ",") '(#f #f)))
(rp-mutex (make-mutex))
(bup-mutex (make-mutex))
(keep-records (args:get-arg "-keep-records"))) ;; used in conjunction with -remove-runs to keep the records, TODO: consolidate this with "mode".
(let* ((write-access-actions '(remove-runs set-state-status archive run-wait kill-runs))
(dbfile (conc *toppath* "/megatest.db"))
(readonly-mode (not (file-write-access? dbfile))))
(when (and readonly-mode
(member action write-access-actions))
(debug:print-error 0 *default-log-port* "megatest.db is readonly. Cannot proceed with action ["action"] in which write-access isrequired .")
(exit 1)))
(debug:print-info 4 *default-log-port* "runs:operate-on => Header: " header " action: " action " new-state-status: " new-state-status)
(if (> 2 (length state-status))
(begin
(debug:print-error 0 *default-log-port* "the parameter to -set-state-status is a comma delimited string. E.g. COMPLETED,FAIL")
(exit)))
(for-each
(lambda (run)
(let ((runkey (string-intersperse (map (lambda (k)
(db:get-value-by-header run header k)) keys) "/"))
(dirs-to-remove (make-hash-table))
(proc-get-tests (lambda (run-id)
(mt:get-tests-for-run run-id
testpatt states statuses
not-in: #f
sort-by: (case action
((remove-runs) 'rundir)
(else 'event_time))))))
(let* ((run-id (db:get-value-by-header run header "id"))
(run-state (db:get-value-by-header run header "state"))
(run-name (db:get-value-by-header run header "runname"))
(tests (if (not (equal? run-state "locked"))
(proc-get-tests run-id)
'()))
(lasttpath "/does/not/exist/I/hope")
(worker-thread #f))
(debug:print-info 4 *default-log-port* "runs:operate-on run=" run ", header=" header)
(if (not (null? tests))
(begin
(case action
((kill-runs)
(tasks:kill-runner target run-name "%")
(debug:print 1 *default-log-port* "Killing tests for run: " runkey " " (db:get-value-by-header run header "runname"))
)
((remove-runs)
;; (if (tasks:need-server run-id)(tasks:start-and-wait-for-server tdbdat run-id 10))
;; seek and kill in flight -runtests with % as testpatt here
;; (if (equal? testpatt "%")
(tasks:kill-runner target run-name testpatt)
;; (debug:print 0 *default-log-port* "not attempting to kill any run launcher processes as testpatt is " testpatt))
(debug:print 1 *default-log-port* "Removing tests for run: " runkey " " (db:get-value-by-header run header "runname")))
((set-state-status)
;; (if (tasks:need-server run-id)(tasks:start-and-wait-for-server tdbdat run-id 10))
(debug:print 1 *default-log-port* "Modifying state and staus for tests for run: " runkey " " (db:get-value-by-header run header "runname")))
((print-run)
(debug:print 1 *default-log-port* "Printing info for run " runkey ", run=" run ", tests=" tests ", header=" header)
action)
((run-wait)
(debug:print 1 *default-log-port* "Waiting for run " runkey ", run=" runnamepatt " to complete"))
((archive)
(debug:print 1 *default-log-port* "Archiving/restoring (" (args:get-arg "-archive") ") data for run: " runkey " " (db:get-value-by-header run header "runname"))
(set! worker-thread
(make-thread
(lambda ()
(case (string->symbol (args:get-arg "-archive"))
((save save-remove keep-html)
(archive:run-bup (args:get-arg "-archive") run-id run-name tests rp-mutex bup-mutex))
((restore)
(archive:bup-restore (args:get-arg "-archive") run-id run-name tests rp-mutex bup-mutex))
(else
(debug:print-error 0 *default-log-port* "unrecognised sub command to -archive. Run \"megatest\" to see help")
(exit))))
"archive-bup-thread"))
(thread-start! worker-thread))
(else
(debug:print-info 0 *default-log-port* "action not recognised " action)))
;; actions that operate on one test at a time can be handled below
;;
(let ((sorted-tests (filter
vector?
(sort tests (lambda (a b)(let ((dira ;; (rmt:sdb-qry 'getstr
(db:test-get-rundir a)) ;; ) ;; (filedb:get-path *fdb* (db:test-get-rundir a)))
(dirb ;; (rmt:sdb-qry 'getstr
(db:test-get-rundir b))) ;; ) ;; ((filedb:get-path *fdb* (db:test-get-rundir b))))
(if (and (string? dira)(string? dirb))
(> (string-length dira)(string-length dirb))
#f))))))
(toplevel-retries (make-hash-table)) ;; try three times to loop through and remove top level tests
(test-retry-time (make-hash-table))
(backgrounded-remove-status (make-hash-table))
(backgrounded-remove-last-visit (make-hash-table))
(backgrounded-remove-result (make-hash-table))
(allow-run-time (string->number (or (args:get-arg "-kill-wait") "10")))) ;; seconds to allow for killing tests before just brutally killing 'em
(let loop ((test (car sorted-tests))
(tal (cdr sorted-tests)))
(let* ((test-id (db:test-get-id test))
(new-test-dat (rmt:get-test-info-by-id run-id test-id)))
(if (not new-test-dat)
(begin
(debug:print-error 0 *default-log-port* "We have a test-id of " test-id " but no record was found. NOTE: No locking of records is done between processes, do not simultaneously remove the same run from two processes!")
(if (not (null? tal))
(loop (car tal)(cdr tal))))
(let* ((item-path (db:test-get-item-path new-test-dat))
(test-name (db:test-get-testname new-test-dat))
(run-dir ;;(filedb:get-path *fdb*
;; (rmt:sdb-qry 'getid
(db:test-get-rundir new-test-dat)) ;; ) ;; run dir is from the link tree
(has-subrun (and (subrun:subrun-test-initialized? run-dir)
(not (subrun:subrun-removed? run-dir))))
(test-state (db:test-get-state new-test-dat))
(test-status (db:test-get-status new-test-dat))
(test-fulln (db:test-get-fullname new-test-dat))
(uname (db:test-get-uname new-test-dat))
(toplevel-with-children (and (db:test-get-is-toplevel test)
(> (rmt:test-toplevel-num-items run-id test-name) 0))))
(case action
((remove-runs)
;; if the test is a toplevel-with-children issue an error and do not remove
(cond
(toplevel-with-children
(debug:print 0 *default-log-port* "WARNING: skipping removal of " test-fulln " with run-id " run-id " as it has sub tests")
(hash-table-set! toplevel-retries test-fulln (+ (hash-table-ref/default toplevel-retries test-fulln 0) 1))
(if (> (hash-table-ref toplevel-retries test-fulln) 3)
(if (not (null? tal))
(loop (car tal)(cdr tal))) ;; no else clause - drop it if no more in queue and > 3 tries
(let ((newtal (append tal (list test))))
(loop (car newtal)(cdr newtal))))) ;; loop with test still in queue
(has-subrun
;;
(let ((last-visit (hash-table-ref/default backgrounded-remove-last-visit test-fulln 0))
(now (current-seconds))
(rem-status (hash-table-ref/default backgrounded-remove-status test-fulln 'not-started)))
(case rem-status
((not-started)
(debug:print 0 *default-log-port* "WARNING: postponing removal of " test-fulln " with run-id " run-id " as it has a subrun")
(hash-table-set! backgrounded-remove-status test-fulln 'started)
(hash-table-set! backgrounded-remove-last-visit test-fulln (current-seconds))
(common:send-thunk-to-background-thread
(lambda ()
(let* ((subrun-remove-succeeded
(subrun:remove-subrun run-dir keep-records)))
(hash-table-set! backgrounded-remove-result test-fulln subrun-remove-succeeded)
(hash-table-set! backgrounded-remove-status test-fulln 'done)))
name: (conc "remove-subrun:"test-fulln))
;; send to back of line, loop
(let ((newtal (append tal (list test))))
(loop (car newtal)(cdr newtal)))
)
((started)
;; if last visit was within last second, sleep 1 second
(if (< (- now last-visit) 1.0)
(thread-sleep! 1.0))
(hash-table-set! backgrounded-remove-last-visit test-fulln (current-seconds))
;; send to back of line, loop
(let ((newtal (append tal (list test))))
(loop (car newtal)(cdr newtal)))
)
((done)
;; drop this one; if remaining, loop, else finish
(hash-table-set! backgrounded-remove-last-visit test-fulln (current-seconds))
(let ((subrun-remove-succeeded (hash-table-ref/default backgrounded-remove-result test-fulln 'exception)))
(cond
((eq? subrun-remove-succeeded 'exception)
(let* ((logfile (subrun:get-log-path run-dir "remove")))
(debug:print 0 *default-log-port* "ERROR: removing subrun of of " test-fulln " with run-id " run-id " ; see logfile @ "logfile))
(if (not (null? tal))
(loop (car tal)(cdr tal))))
(subrun-remove-succeeded
(debug:print 0 *default-log-port* "Now removing of " test-fulln " with run-id " run-id " since subrun was removed.")
;;(runs:remove-test-directory new-test-dat mode) ;; let normal case handle this. it will go thru loop again as non-subrun
(let ((newtal (append tal (list test))))
(loop (car newtal)(cdr newtal))))
(else
(let* ((logfile (subrun:get-log-path run-dir "remove")))
(debug:print 0 *default-log-port* "WARNING: removal of subrun failed. Please check "logfile" for details."))
;; send to back of line, loop (will not match has-subrun next time through)
(if (not (null? tal))
(loop (car tal)(cdr tal))))))
)
) ; end case rem-status
) ; end let
); end cond has-subrun
(else
;; BB - TODO - consider backgrounding to threads to delete tests (work below)
(debug:print-info 0 *default-log-port* "test: " test-name " itest-state: " test-state)
(if (member test-state (list "RUNNING" "LAUNCHED" "REMOTEHOSTSTART" "KILLREQ"))
(begin
(if (not (hash-table-ref/default test-retry-time test-fulln #f))
(begin
;; want to set to REMOVING BUT CANNOT do it here?
(hash-table-set! test-retry-time test-fulln (current-seconds))))
(if (> (- (current-seconds)(hash-table-ref test-retry-time test-fulln)) allow-run-time)
;; This test is not in a correct state for cleaning up. Let's try some graceful shutdown steps first
;; Set the test to "KILLREQ" and wait five seconds then try again. Repeat up to five times then give
;; up and blow it away.
(begin
(debug:print 0 *default-log-port* "WARNING: could not gracefully remove test " test-fulln ", tried to kill it to no avail. Forcing state to FAILEDKILL and continuing")
(mt:test-set-state-status-by-id run-id (db:test-get-id test) "FAILEDKILL" "n/a" #f)
(thread-sleep! 1))
(begin
(mt:test-set-state-status-by-id run-id (db:test-get-id test) "KILLREQ" "n/a" #f)
(thread-sleep! 1)))
;; NOTE: This is suboptimal as the testdata will be used later and the state/status may have changed ...
(if (null? tal)
(loop new-test-dat tal)
(loop (car tal)(append tal (list new-test-dat)))))
(begin
(runs:remove-test-directory new-test-dat mode) ;; 'remove-all)
(if (not (null? tal))
(loop (car tal)(cdr tal)))))))
(rmt:update-run-stats run-id (rmt:get-raw-run-stats run-id)))
((kill-runs)
;; RUNNING -> KILLREQ
;; LAUNCHED,RUNNING,REMOTEHOSTSTART -> NOT STARTED
(cond
((and has-subrun (member test-state (list "RUNNING" "LAUNCHED" "REMOTEHOSTSTART" "KILLREQ")))
(common:send-thunk-to-background-thread
(lambda ()
(let* ((subrun-remove-succeeded
(subrun:kill-subrun run-dir keep-records)))
#t)))
(if (not (null? tal))
(loop (car tal)(cdr tal)))
)
((member test-state (list "RUNNING" "LAUNCHED" "REMOTEHOSTSTART" "KILLREQ"))
(debug:print 1 *default-log-port* "INFO: issuing killreq to test "test-fulln)
(mt:test-set-state-status-by-id run-id (db:test-get-id test) "KILLREQ" "n/a" #f)
(if (not (null? tal))
(loop (car tal)(cdr tal))))
((and (member test-status '("PREQ_FAIL" "PREQ_DISCARDED" "BLOCKED" "ZERO_ITEMS" "KEEP_TRYING" "TEN_STRIKES" "TIMED_OUT")))
(rmt:set-state-status-and-roll-up-items run-id (db:test-get-id test) 'foo "NOT_STARTED" "n/a" (conc "kill-run moved from "test-state":"test-status" to NOT_STARTED:n/a"))
;;(mt:test-set-state-status-by-id run-id (db:test-get-id test) "NOT_STARTED" "n/a" (conc "kill-run moved from "test-state":"test-status" to NOT_STARTED:n/a"))
(if (not (null? tal))
(loop (car tal)(cdr tal)))
)
(else
(if (not (null? tal))
(loop (car tal)(cdr tal)))
)))
((set-state-status)
(let* ((new-state (car state-status))
(new-status (cadr state-status))
(test-id (db:test-get-id test))
(test-run-dir (db:test-get-rundir new-test-dat))
(has-subrun (and (subrun:subrun-test-initialized? test-run-dir)
(not (subrun:subrun-removed? test-run-dir)))))
(when has-subrun
(common:send-thunk-to-background-thread
(lambda ()
(subrun:set-state-status test-run-dir state status new-state-status)
)
)
)
(debug:print-info 2 *default-log-port* "new state " new-state ", new status " new-status )
(mt:test-set-state-status-by-id run-id test-id new-state new-status #f))
(if (not (null? tal))
(loop (car tal)(cdr tal))))
((run-wait)
;; BB TODO - manage has-subrun case
(debug:print-info 2 *default-log-port* "still waiting, " (length tests) " tests still running")
(thread-sleep! 10)
(let ((new-tests (proc-get-tests run-id)))
(if (null? new-tests)
(debug:print-info 1 *default-log-port* "Run completed according to zero tests matching provided criteria.")
(loop (car new-tests)(cdr new-tests)))))
((archive)
;; BB TODO - manage has-subrun case
(if (and run-dir (not toplevel-with-children))
(let ((ddir (conc run-dir "/")))
(case (string->symbol (args:get-arg "-archive"))
((save save-remove keep-html)
(if (common:file-exists? ddir)
(debug:print-info 0 *default-log-port* "Estimating disk space usage for " test-fulln ": " (common:get-disk-space-used ddir)))))))
(if (not (null? tal))
(loop (car tal)(cdr tal))))
)))
)
(if worker-thread (thread-join! worker-thread)))
(common:join-backgrounded-threads))))
;; remove the run if zero tests remain
(if (eq? action 'remove-runs)
(let* ((run-id (db:get-value-by-header run header "id")) ;; NB// masks run-id from above?
(remtests (mt:get-tests-for-run run-id #f '("DELETED") '("n/a") not-in: #t)))
(if (null? remtests) ;; no more tests remaining
(let* ((dparts (string-split lasttpath "/"))
(runpath (conc "/" (string-intersperse
(take dparts (- (length dparts) 1))
"/"))))
(debug:print 1 *default-log-port* "Removing run: " runkey " " (db:get-value-by-header run header "runname") " and related record")
(if (not keep-records)
(begin
(rmt:delete-run run-id)
(rmt:delete-old-deleted-test-records)))
;; (rmt:set-var "DELETED_TESTS" (current-seconds))
;; need to figure out the path to the run dir and remove it if empty
;; (if (null? (glob (conc runpath "/*")))
;; (begin
;; (debug:print 1 *default-log-port* "Removing run dir " runpath)
;; (system (conc "rmdir -p " runpath))))
)))))
))
runs)
;; (sqlite3:finalize! (db:delay-if-busy tdbdat))
)
#t)
(define (runs:remove-test-directory test mode) ;; remove-data-only)
(let* ((run-dir (db:test-get-rundir test)) ;; run dir is from the link tree
(real-dir (if (common:file-exists? run-dir)
;; (resolve-pathname run-dir)
(common:nice-path run-dir)
#f))
(clean-mode (or mode 'remove-all))
(test-id (db:test-get-id test))
;; (lock-key (conc "test-" test-id))
;; (got-lock (let loop ((lock (rmt:no-sync-get-lock lock-key))
;; (expire-time (+ (current-seconds) 30))) ;; give up on getting the lock and steal it after 15 seconds
;; (if (car lock)
;; #t
;; (if (> (current-seconds) expire-time)
;; (begin
;; (debug:print-info 0 *default-log-port* "Timed out waiting for a lock to clean test with id " test-id)
;; (rmt:no-sync-del! lock-key) ;; destroy the lock
;; (loop (rmt:no-sync-get-lock lock-key) expire-time)) ;;
;; (begin
;; (thread-sleep! 1)
;; (loop (rmt:no-sync-get-lock lock-key) expire-time)))))))
)
(case clean-mode
((remove-data-only)(mt:test-set-state-status-by-id (db:test-get-run_id test)(db:test-get-id test) "CLEANING" "LOCKED" #f))
((remove-all) (mt:test-set-state-status-by-id (db:test-get-run_id test)(db:test-get-id test) "REMOVING" "LOCKED" #f))
((archive-remove) (mt:test-set-state-status-by-id (db:test-get-run_id test)(db:test-get-id test) "ARCHIVE_REMOVING" #f #f)))
(debug:print-info 1 *default-log-port* "Attempting to remove " (if real-dir (conc " dir " real-dir " and ") "") " link " run-dir)
(if (and real-dir
(> (string-length real-dir) 5)
(common:file-exists? real-dir)) ;; bad heuristic but should prevent /tmp /home etc.
(begin ;; let* ((realpath (resolve-pathname run-dir)))
(debug:print-info 1 *default-log-port* "Recursively removing " real-dir)
(if (common:file-exists? real-dir)
(runs:safe-delete-test-dir real-dir)
(debug:print 0 *default-log-port* "WARNING: test dir " real-dir " appears to not exist or is not readable")))
(if real-dir
(debug:print 0 *default-log-port* "WARNING: directory " real-dir " does not exist")
(debug:print 0 *default-log-port* "WARNING: no real directory corrosponding to link " run-dir ", nothing done")))
(if (symbolic-link? run-dir)
(begin
|
︙ | | | ︙ | |
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
|
(delete-directory run-dir)))
(if (and run-dir
(not (member run-dir (list "n/a" "/tmp/badname"))))
(debug:print 0 *default-log-port* "WARNING: not removing " run-dir " as it either doesn't exist or is not a symlink")
(debug:print 0 *default-log-port* "NOTE: the run dir for this test is undefined. Test may have already been deleted."))
))
;; Only delete the records *after* removing the directory. If things fail we have a record
(case mode
((remove-data-only)(mt:test-set-state-status-by-id (db:test-get-run_id test)(db:test-get-id test) "NOT_STARTED" "n/a" #f))
((archive-remove) (mt:test-set-state-status-by-id (db:test-get-run_id test)(db:test-get-id test) "ARCHIVED" #f #f))
(else (rmt:delete-test-records (db:test-get-run_id test) (db:test-get-id test))))))
;;======================================================================
;; Routines for manipulating runs
;;======================================================================
;; Since many calls to a run require pretty much the same setup
;; this wrapper is used to reduce the replication of code
|
|
|
|
>
>
|
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
|
(delete-directory run-dir)))
(if (and run-dir
(not (member run-dir (list "n/a" "/tmp/badname"))))
(debug:print 0 *default-log-port* "WARNING: not removing " run-dir " as it either doesn't exist or is not a symlink")
(debug:print 0 *default-log-port* "NOTE: the run dir for this test is undefined. Test may have already been deleted."))
))
;; Only delete the records *after* removing the directory. If things fail we have a record
(case clean-mode
((remove-data-only)(mt:test-set-state-status-by-id (db:test-get-run_id test)(db:test-get-id test) (db:test-get-state test)(db:test-get-status test) #f))
((archive-remove) (mt:test-set-state-status-by-id (db:test-get-run_id test)(db:test-get-id test) "ARCHIVED" #f #f))
(else (rmt:delete-test-records (db:test-get-run_id test) (db:test-get-id test))))
;; (rmt:no-sync-del! lock-key)
))
;;======================================================================
;; Routines for manipulating runs
;;======================================================================
;; Since many calls to a run require pretty much the same setup
;; this wrapper is used to reduce the replication of code
|
︙ | | | ︙ | |
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
|
(exit 3))
(else
(let (;; (db #f)
(keys #f))
(if (launch:setup)
(begin
(full-runconfigs-read) ;; cache the run config
(launch:cache-config)) ;; do not cache here - need to be sure runconfigs is processed
(begin
(debug:print 0 *default-log-port* "Failed to setup, exiting")
(exit 1)))
(set! keys (keys:config-get-fields *configdat*))
;; have enough to process -target or -reqtarg here
|
>
|
|
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
|
(exit 3))
(else
(let (;; (db #f)
(keys #f))
(if (launch:setup)
(begin
(full-runconfigs-read) ;; cache the run config
;; (launch:cache-config) ;; there are two independent config cache locations, turning this one off for now. MRW.
) ;; do not cache here - need to be sure runconfigs is processed
(begin
(debug:print 0 *default-log-port* "Failed to setup, exiting")
(exit 1)))
(set! keys (keys:config-get-fields *configdat*))
;; have enough to process -target or -reqtarg here
|
︙ | | | ︙ | |
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
|
(define (runs:handle-locking target keys runname lock unlock user)
(let* ((db #f)
(rundat (mt:get-runs-by-patt keys runname target))
(header (vector-ref rundat 0))
(runs (vector-ref rundat 1)))
(for-each (lambda (run)
(let ((run-id (db:get-value-by-header run header "id")))
(if (or lock
(and unlock
(begin
(print "Do you really wish to unlock run " run-id "?\n y/n: ")
(equal? "y" (read-line)))))
(rmt:lock/unlock-run run-id lock unlock user)
(debug:print-info 0 *default-log-port* "Skipping lock/unlock on " run-id))))
runs)))
;;======================================================================
;; Rollup runs
;;======================================================================
;; Update the test_meta table for this test
|
|
>
>
>
>
|
|
>
>
|
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
|
(define (runs:handle-locking target keys runname lock unlock user)
(let* ((db #f)
(rundat (mt:get-runs-by-patt keys runname target))
(header (vector-ref rundat 0))
(runs (vector-ref rundat 1)))
(for-each (lambda (run)
(let ((run-id (db:get-value-by-header run header "id"))
(str (if lock
"lock"
"unlock")))
(if (or lock
(and unlock
(or (args:get-arg "-force")
(begin
(print "Do you really wish to unlock run " run-id "?\n y/n: ")
(equal? "y" (read-line))))))
(begin
(rmt:lock/unlock-run run-id lock unlock user)
(debug:print-info 0 *default-log-port* "Done " str " on run id " run-id))
(debug:print-info 0 *default-log-port* "Skipping lock/unlock on " run-id))))
runs)))
;;======================================================================
;; Rollup runs
;;======================================================================
;; Update the test_meta table for this test
|
︙ | | | ︙ | |
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
|
;; find tests with matching tags, tagpatt is a string "tagpatt1,tagpatt2%, ..."
;;
(define (runs:get-tests-matching-tags tagpatt)
(let* ((tagdata (rmt:get-tests-tags))
(res '())) ;; list of tests that match one or more tags
(for-each
(lambda (tag)
(if (patt-list-match tag tagpatt)
(set! res (append (hash-table-ref tagdata tag) res))))
(hash-table-keys tagdata))
res))
;; Update test_meta for all tests
(define (runs:update-all-test_meta db)
(let ((test-names (tests:get-all))) ;; (tests:get-valid-tests)))
(for-each
|
|
>
>
|
|
|
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
|
;; find tests with matching tags, tagpatt is a string "tagpatt1,tagpatt2%, ..."
;;
(define (runs:get-tests-matching-tags tagpatt)
(let* ((tagdata (rmt:get-tests-tags))
(res '())) ;; list of tests that match one or more tags
(for-each
(lambda (row)
(let* ((tag (car row))
(tests (cdr row)))
(if (patt-list-match tag tagpatt)
(set! res (append tests res)))))
tagdata)
res))
;; Update test_meta for all tests
(define (runs:update-all-test_meta db)
(let ((test-names (tests:get-all))) ;; (tests:get-valid-tests)))
(for-each
|
︙ | | | ︙ | |
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
|
;; clean cache files
(define (runs:clean-cache target runname toppath)
(if target
(if runname
(let* ((linktree (common:get-linktree)) ;; (if toppath (configf:lookup *configdat* "setup" "linktree")))
(runtop (conc linktree "/" target "/" runname))
(files (if (file-exists? runtop)
(append (glob (conc runtop "/.megatest*"))
(glob (conc runtop "/.runconfig*")))
'())))
(if (null? files)
(debug:print-info 0 *default-log-port* "No cached megatest or runconfigs files found. None removed.")
(begin
(debug:print-info 0 *default-log-port* "Removing cached files:\n " (string-intersperse files "\n "))
|
|
|
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
|
;; clean cache files
(define (runs:clean-cache target runname toppath)
(if target
(if runname
(let* ((linktree (common:get-linktree)) ;; (if toppath (configf:lookup *configdat* "setup" "linktree")))
(runtop (conc linktree "/" target "/" runname))
(files (if (common:file-exists? runtop)
(append (glob (conc runtop "/.megatest*"))
(glob (conc runtop "/.runconfig*")))
'())))
(if (null? files)
(debug:print-info 0 *default-log-port* "No cached megatest or runconfigs files found. None removed.")
(begin
(debug:print-info 0 *default-log-port* "Removing cached files:\n " (string-intersperse files "\n "))
|
︙ | | | ︙ | |