;;;; gc tests ;;;; This software is part of the SBCL system. See the README file for ;;;; more information. ;;;; ;;;; While most of SBCL is derived from the CMU CL system, the test ;;;; files (like this one) were written from scratch after the fork ;;;; from CMU CL. ;;; ;;;; This software is in the public domain and is provided with ;;;; absoluely no warranty. See the COPYING and CREDITS files for ;;;; more information. (in-package :cl-user) (defparameter *x* ()) (defun cons-madly () (loop repeat 10000 do (setq *x* (make-string 100000)))) ;; check that WITHOUT-INTERRUPTS doesn't block the gc trigger (sb-sys:without-interrupts (cons-madly)) ;; check that WITHOUT-INTERRUPTS doesn't block SIG_STOP_FOR_GC #+sb-thread (sb-sys:without-interrupts (let ((thread (sb-thread:make-thread (lambda () (sb-ext:gc))))) (loop while (sb-thread:thread-alive-p thread)))) (let ((gc-happend nil)) (push (lambda () (setq gc-happend t)) sb-ext:*after-gc-hooks*) ;; check that WITHOUT-GCING defers explicit gc (sb-sys:without-gcing (gc) (assert (not gc-happend))) (assert gc-happend) ;; check that WITHOUT-GCING defers SIG_STOP_FOR_GC #+sb-thread (let ((in-without-gcing nil)) (setq gc-happend nil) (sb-thread:make-thread (lambda () (loop while (not in-without-gcing)) (sb-ext:gc))) (sb-sys:without-gcing (setq in-without-gcing t) (sleep 3) (assert (not gc-happend))) ;; give the hook time to run (sleep 1) (assert gc-happend))) ;;; SB-EXT:GENERATION-* accessors returned bogus values for generation > 0 (with-test (:name :bug-529014 :skipped-on '(not :gencgc)) (loop for i from 0 to sb-vm:+pseudo-static-generation+ do (assert (= (sb-ext:generation-bytes-consed-between-gcs i) (truncate (sb-ext:bytes-consed-between-gcs) sb-vm:+highest-normal-generation+))) ;; FIXME: These parameters are a) tunable in the source and b) ;; duplicated multiple times there and now here. It would be good to ;; OAOO-ify them (probably to src/compiler/generic/params.lisp). (assert (= (sb-ext:generation-minimum-age-before-gc i) 0.75)) (assert (= (sb-ext:generation-number-of-gcs-before-promotion i) 1)))) (defun stress-gc () ;; Kludge or not? I don't know whether the smaller allocation size ;; for sb-safepoint is a legitimate correction to the test case, or ;; rather hides the actual bug this test is checking for... It's also ;; not clear to me whether the issue is actually safepoint-specific. ;; But the main problem safepoint-related bugs tend to introduce is a ;; delay in the GC triggering -- and if bug-936304 fails, it also ;; causes bug-981106 to fail, even though there is a full GC in ;; between, which makes it seem unlikely to me that the problem is ;; delay- (and hence safepoint-) related. --DFL (let* ((x (make-array (truncate #-sb-safepoint (* 0.2 (dynamic-space-size)) #+sb-safepoint (* 0.1 (dynamic-space-size)) sb-vm:n-word-bytes)))) (elt x 0))) (with-test (:name :bug-936304) (gc :full t) (time (assert (eq :ok (handler-case (progn (loop repeat 50 do (stress-gc)) :ok) (storage-condition () :oom)))))) (with-test (:name :bug-981106) (gc :full t) (time (assert (eq :ok (handler-case (dotimes (runs 100 :ok) (let* ((n (truncate (dynamic-space-size) 1200)) (len (length (with-output-to-string (string) (dotimes (i n) (write-sequence "hi there!" string)))))) (assert (eql len (* n (length "hi there!")))))) (storage-condition () :oom)))))) (with-test (:name :gc-logfile) (assert (not (gc-logfile))) (let ((p #p"gc.log")) (assert (not (probe-file p))) (assert (equal p (setf (gc-logfile) p))) (gc) (let ((p2 (gc-logfile))) (assert (equal (truename p2) (truename p)))) (assert (not (setf (gc-logfile) nil))) (assert (not (gc-logfile))) (delete-file p)))