[squeak-dev] Serious garbage/storage leak issue with MCInfoProxy

Marcel Taeumel marcel.taeumel at hpi.de
Tue Feb 20 13:00:42 UTC 2018


Hmm... I do not see value in cleaning up projects (to free memory) when the user just switches between them back and forth. Especially in the light of worlds-in-worlds or background projects, this seems like an unreasonable path to follow. We should not make too much assumptions about application code or user needs at this point. There could always be a "Oh, please render that background project into this file. Oh, why is it so slow at the moment..."

Cleaning up resources for save-and-quit, however, seems reasonable because it is at the environment's boundary and not some arbitrary in-image boundary.

Best,
Marcel
Am 20.02.2018 13:58:38 schrieb David T. Lewis <lewis at mail.msen.com>:
This might also justify adding one new method to Project:

Project>>cleanUpForExit

Default implementation would be do nothing, and for MorphicProject it
would clean up the drop morphs. This would be called from #finalExitActions:
for ongoing cleanup, and when saving the image you would do this:

Project current cleanupForExit

Dave


On Tue, Feb 20, 2018 at 07:50:57AM -0500, David T. Lewis wrote:
> To handle all of the morphic projects, it would be better to do the
> cleanup when exiting the projects:
>
> MorphicProject>>finalExitActions: enteringProject
> world allMorphsDo: [:ea | ea removeProperty: #dropShadow].
> ...
>
> That will leave only the current project to worry about when saving
> the image. Of course, you do not want to to cleam up the morphs if
> the current world is a ControlManager, so it might look something
> like this:
>
> | world |
> (world := Project current world) isMorph
> ifTrue: [world allMorphsDo: [:ea | ea removeProperty: #dropShadow]]
>
> Dave
>
> On Tue, Feb 20, 2018 at 11:26:19AM +0100, Marcel Taeumel wrote:
> > We have to consider all open projects:
> >
> > Project allMorphicProjects do: [:project |
> > ?? ??project world allMorphsDo: [:morph |
> > ?? ?? ?? morph removeProperty: #dropShadow]].
> >
> > Yet, there can be hidden morphs. If we want to ignore those, we can stick with only cleaning up top-level morphs, which are usually the system windows:
> >
> > Project allMorphicProjects do: [:project |
> > ?? ??project world submorphsDo: [:morph |
> > ?? ?? ?? morph removeProperty: #dropShadow]].
> >
> > Best,
> > Marcel
> > Am 20.02.2018 10:57:08 schrieb Bob Arning :
> > Probably good enough might be
> > World allMorphsDo: [:ea | ea removeProperty: #dropShadow]
> > takes 0 ms for me
> >
> >
> > On 2/20/18 2:46 AM, Marcel Taeumel wrote:
> >
> > Hi Eliot,
> >
> > sure. Removing the potential drop shadow of all kinds of morphs takes time:
> >
> > Morph allSubInstancesDo: [:ea |??ea removeProperty: #dropShadow].
> >
> > About 3 seconds here in a quite clean image.
> >
> > SystemWindow allSubInstancesDo: [:ea |??ea removeProperty: #dropShadow].
> >
> > Works at 100 milliseconds.
> >
> > What has the biggest effect in your image?
> >
> > Best,
> > Marcel
> > Am 20.02.2018 02:51:06 schrieb Eliot Miranda [mailto:eliot.miranda at gmail.com]:
> > Hi Marcel,
> >
> > ?? ?? I've finally worked out that the space overhead here comes from the dropShadow cache system in Morphs (in otherProperties in MorphExtensions).?? Would it be possible to arrange to flush all drop shadows in the current project when:
> > - exiting a project
> > - snapshotting
> > ?
> >
> > On Thu, Jan 18, 2018 at 3:24 PM, Eliot Miranda wrote:
> >
> > Hi All,
> >
> > ?? ?? I've been experiencing image save slowdowns recently and finally my work image reached 1.%Gb and I thought I better take a look:
> >
> > Sisyphus.Cog$ ls -lh SpurWork64.* save/SpurWork64-*
> > -rw-r--r--@ 1 eliot ??staff ?? ??28M Jan 18 12:47 SpurWork64.changes
> > -rw-r--r--@ 1 eliot ??staff ?? 1.6G Jan 18 12:48 SpurWork64.image
> > -rw-r--r--@ 1 eliot ??staff ?? ??28M Jan 18 12:03 save/SpurWork64-2018-01-18.changes
> > -rw-r--r--@ 1 eliot ??staff ?? 1.5G Jan 18 12:03 save/SpurWork64-2018-01-18.image
> >
> > I ran a space analysis and found that Bitmap and ByteArray were the top two, so I looked for large Bitmaps.?? I found three that fit this criterion:
> >
> >
> > ?? ?? Bitmap allInstances select: [:bm| bm size >= 1000000 and: [bm ~~ Display bits]]
> >
> > I inspected the three and did a chase pointers on one of them.?? As I did that suddenly
> > a) the inspector on the Array became empty (still an array but zero elements)
> > b) the progress bar for Downloading FlexibleVocabularies-who.NN appeared
> >
> > I interrupted this and did a very cursory stack examination. Some object had not understood isLiteral and from there what looked like an attempt to turn this stub into a real object caused FlexibleVocabularies-who.NN to start to download.
> >
> > I threw away the debugger, ran the GC and suddenly all my free space was back.?? So now on disc I have
> >
> > Sisyphus.Cog$ ls -lh SpurWork64.* save/SpurWork64-*
> > -rw-r--r--@ 1 eliot ??staff ?? ??28M Jan 18 15:17 SpurWork64.changes
> > -rw-r--r--@ 1 eliot ??staff ?? ??57M Jan 18 15:17 SpurWork64.image
> > -rw-r--r--@ 1 eliot ??staff ?? ??28M Jan 18 12:03 save/SpurWork64-2018-01-18.changes
> > -rw-r--r--@ 1 eliot ??staff ?? 1.5G Jan 18 12:03 save/SpurWork64-2018-01-18.image
> >
> > What is going on here??? There seems to be a very bad storage leak.?? Can we please discuss this??? This doesn't seem like healthy behaviour at all :-)
> >
> > _,,,^..^,,,_
> >
> > best,??Eliot
> >
> >
> >
> > --
> >
> > _,,,^..^,,,_
> >
> > best,??Eliot
> >
> >
>
> >
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20180220/658859e1/attachment.html>


More information about the Squeak-dev mailing list