DrRacket GUI for plai/gc2/mutator crashes

[jameswhang]

DrRacket GUI that displays the heap for plai/gc2/mutator language seems to run out of memory when the heap has sufficiently large number of floats, but the mutator program itself doesn't stop executing and causes DrRacket to display a ton of out of memory window.

An example mutator that causes this error is shown below.

#lang plai/gc2/mutator

(allocator-setup "mark-and-sweep.rkt" 500) ; This is a standard mark-and-sweep gc collector as described in https://www.eecs.northwestern.edu/~robby/courses/321-2015-winter/lecture12.pdf

(define guess 0)

(define (sqrt x) ; dumb way
  (set! guess x)
  (sqrt-rec (/ x 2) x))

(define (sqrt-rec guess x)
  (if (close? guess x)
      guess
      (begin
        (set! guess (improve guess x))
        (sqrt-rec guess x))))

(define (average x y)
  (/ (+ x y) 2))

(define (square x)
  (* x x))

(define (abs x)
  (if (< x 0)
      (- 0 x)
      x))

(define (improve guess x)
  (average guess (/ x guess)))

(define (close? guess x)
  (< (abs (- (square guess) x)) 0.001))

(define (run num)
  (if (= 0 num)
      0
      (begin (sqrt 1000000)
             (run (- num 1)))))

(define a (run 100))

Closing the heap window doesn't cause DrRacket to crash, so the problem seems to be with the rendering of heap window.

[jeapostrophe]

It seems like this is the expected behavior... if you allocate a huge amount of memory, while imposing a memory limit, then you should expect an error that you violated the memory limit?

[rfindler]

Yes, but maybe not millions of times?

…

On Tue, Feb 21, 2017 at 2:01 PM Jay McCarthy ***@***.***> wrote: It seems like this is the expected behavior... if you allocate a huge amount of memory, while imposing a memory limit, then you should expect an error that you violated the memory limit? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#94 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAYWsHzJVZoFweOGHnXguoZdr2rKoWjcks5re0KkgaJpZM4MHm0d> .

[jameswhang]

The expected behavior that I expected was for the out of memory error message to show up once and stop the mutator program from running..

[jeapostrophe]

Sorry I missed that part of the message.

@rfindler Can you tell what's happening? Am I catching an error and letting it continue so DrRacket makes another one?

[rfindler]

Oh I don't think it is possible for this to be a plai bug. But I don't know what is happening. If you had energy to eliminate plai from the mix that would be great. But i can do it too (later).

…

On Tue, Feb 21, 2017 at 4:12 PM Jay McCarthy ***@***.***> wrote: Sorry I missed that part of the message. @rfindler <https://github.com/rfindler> Can you tell what's happening? Am I catching an error and letting it continue so DrRacket makes another one? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#94 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAYWsHkobQW-0KtSJ_gcSD5hWJyVYaX7ks5re2FkgaJpZM4MHm0d> .