[squeak-dev] testOutOfMemorySignalExtreme too disruptive

Chris Muller asqueaker at gmail.com
Wed Jun 1 00:30:39 UTC 2022


If -memory only works on unix VM's, then I need to update that test I put
in the Inbox to check for that, too.  I didn't know that!

 - Chris


On Tue, May 31, 2022 at 5:25 PM David T. Lewis <lewis at mail.msen.com> wrote:

> This is a really good approach although I think it only works on
> unix VMs. Even so it would be better than the current state, and
> it seems better to have the test available on unix VMs than not to
> have it at all.
>
> Dave
>
> On Tue, May 31, 2022 at 04:32:52PM -0500, Chris Muller wrote:
> > The test could be changed to run ONLY if the vm was started with a
> limiting
> > -memory argument.  See the attached method, Smalltalk>>#imageMemoryLimit.
> > Let THAT be the way to enable the test, that the user started the image
> in
> > a way that doesn't allow it, even accidentally, to swallow up the whole
> > system.
> >
> > Best,
> >   Chris
> >
> >
> > On Wed, May 18, 2022 at 11:07 PM Lauren Pullen <drurowin at gmail.com>
> wrote:
> >
> > > Hi list,
> > >
> > > I can't remember if I already asked this, please ignore if it's a
> double.
> > >
> > > AllocationTest's testOutOfMemorySignalExtreme is too disruptive to the
> > > rest of the OS to run by default.  It will absolutely fill my hard
> drive
> > > with pagefile if the OS doesn't trap and ask me to start terminating
> > > processes first.  If the pagefile fills the hard drive many programs
> are
> > > simply incapable of not corrupting their persisted state... every time
> > > it happens I need to go through and delete preferences and cache
> > > folders.  The test has never completed, and I lack a 1 TB hard drive or
> > > unused ones to RAID together.
> > >
> > > I feel it's appropriate to disable that test and require users
> > > explicitly allow it to run.  Personally, I delete the method.  Can this
> > > be disabled?
> > >
> > >
>
>
> >
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20220531/21bc5dcc/attachment.html>


More information about the Squeak-dev mailing list