The Trunk: Kernel-mt.1211.mcz

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

The Trunk: Kernel-mt.1211.mcz

commits-2
Marcel Taeumel uploaded a new version of Kernel to project The Trunk:
http://source.squeak.org/trunk/Kernel-mt.1211.mcz

==================== Summary ====================

Name: Kernel-mt.1211
Author: mt
Time: 22 January 2019, 1:51:49.258809 pm
UUID: 2bc71a1e-02c1-824a-ae49-4db30171e2f9
Ancestors: Kernel-eem.1210

1) Restart timer-interrupt watcher process during system-wide clean-up.
2) Extract idle-process into own method because that's what we usually do: EventSensor (interrupt watcher), EventSensor (event tickler), WeakArray (finalization), etc.

=============== Diff against Kernel-eem.1210 ===============

Item was added:
+ ----- Method: Delay class>>cleanUp (in category 'class initialization') -----
+ cleanUp
+
+ self initialize.!

Item was added:
+ ----- Method: ProcessorScheduler class>>installIdleProcess (in category 'background process') -----
+ installIdleProcess
+ "Install a background process of the lowest possible priority that is always runnable."
+ "Details: The virtual machine requires that there is aways some runnable process that
+ can be scheduled; this background process ensures that this is the case."
+
+ BackgroundProcess ifNotNil: [BackgroundProcess terminate].
+ BackgroundProcess := [self idleProcess] newProcess.
+ BackgroundProcess priority: SystemRockBottomPriority.
+ BackgroundProcess resume.!

Item was changed:
  ----- Method: ProcessorScheduler class>>startUp: (in category 'background process') -----
  startUp: resuming
- "Install a background process of the lowest possible priority that is always runnable."
- "Details: The virtual machine requires that there is aways some runnable process that
- can be scheduled; this background process ensures that this is the case."
 
+ Smalltalk processPreemptionYields ifTrue: [
+ Smalltalk processPreemptionYields: false].
+
+ resuming ifTrue: [
+ Smalltalk installLowSpaceWatcher.
+ self installIdleProcess].!
- Smalltalk processPreemptionYields ifTrue:
- [Smalltalk processPreemptionYields: false].
- resuming ifTrue:
- [Smalltalk installLowSpaceWatcher.
- BackgroundProcess ifNotNil:
- [BackgroundProcess terminate].
- (BackgroundProcess := [self idleProcess] newProcess)
- priority: SystemRockBottomPriority;
- resume]!


Reply | Threaded
Open this post in threaded view
|

Re: The Trunk: Kernel-mt.1211.mcz

marcel.taeumel
Here is an overview of the usual suspects:

<http://forum.world.st/file/t370078/squeak-processes.png>

Best,
Marcel



--
Sent from: http://forum.world.st/Squeak-Dev-f45488.html

Reply | Threaded
Open this post in threaded view
|

Re: The Trunk: Kernel-mt.1211.mcz

David T. Lewis
On Tue, Jan 22, 2019 at 07:10:27AM -0600, marcel.taeumel wrote:
> Here is an overview of the usual suspects:
>
> <http://forum.world.st/file/t370078/squeak-processes.png>
>
> Best,
> Marcel
>

That looks very good, thank you for cleaning it up.

Should the changes go into the squeak52 stream also?

Dave


Reply | Threaded
Open this post in threaded view
|

Re: The Trunk: Kernel-mt.1211.mcz

marcel.taeumel
Yes, we should backport the new version of ReleaseBuilder to squeak52 and then update the CI script on GitHub to make use of "ReleaseBuilder prepareProcesses" and "ReleaseBuilder saveAndQuit", which would then produce a repaired release image.

Kernel-mt.1211 is not needed for it. Only ReleaseBuilder-mt.194

Best,
Marcel

Am 22.01.2019 16:22:29 schrieb David T. Lewis <[hidden email]>:

On Tue, Jan 22, 2019 at 07:10:27AM -0600, marcel.taeumel wrote:
> Here is an overview of the usual suspects:
>
>
>
> Best,
> Marcel
>

That looks very good, thank you for cleaning it up.

Should the changes go into the squeak52 stream also?

Dave




Reply | Threaded
Open this post in threaded view
|

Re: The Trunk: Kernel-mt.1211.mcz

David T. Lewis
Backporting ReleaseBuilder into the squeak52 update stream requires
adopting the four changes in ReleaseBuilder-mt.194, while skipping the
other changes in 191, 192, and 193.

In order to prevent file name conflicts between trunk and squeak52,
I propose saving the new version in squeak52 with the name
ReleaseBuilder.merge-XXX.191, and making a new update map to point
to ReleaseBuilder.merge.

If there are no objections I will make those changes soon.

Dave


On Tue, Jan 22, 2019 at 04:32:57PM +0100, Marcel Taeumel wrote:

> Yes, we should backport the new version of ReleaseBuilder to squeak52 and then update the CI script on GitHub to make use of "ReleaseBuilder prepareProcesses" and "ReleaseBuilder saveAndQuit", which would then produce a repaired release image.
>
> Kernel-mt.1211 is not needed for it. Only ReleaseBuilder-mt.194
>
>
> Best,
> Marcel
> Am 22.01.2019 16:22:29 schrieb David T. Lewis <[hidden email]>:
> On Tue, Jan 22, 2019 at 07:10:27AM -0600, marcel.taeumel wrote:
> > Here is an overview of the usual suspects:
> >
> >
> >
> > Best,
> > Marcel
> >
>
> That looks very good, thank you for cleaning it up.
>
> Should the changes go into the squeak52 stream also?
>
> Dave
>
>

>


Reply | Threaded
Open this post in threaded view
|

Re: The Trunk: Kernel-mt.1211.mcz

David T. Lewis
Done.

The ReleaseBuilder update is ReleaseBuilder.merge-dtl.191 and the
update map is update-dtl.437.mcm.

Dave

On Tue, Jan 22, 2019 at 11:44:10AM -0500, David T. Lewis wrote:

> Backporting ReleaseBuilder into the squeak52 update stream requires
> adopting the four changes in ReleaseBuilder-mt.194, while skipping the
> other changes in 191, 192, and 193.
>
> In order to prevent file name conflicts between trunk and squeak52,
> I propose saving the new version in squeak52 with the name
> ReleaseBuilder.merge-XXX.191, and making a new update map to point
> to ReleaseBuilder.merge.
>
> If there are no objections I will make those changes soon.
>
> Dave
>
>
> On Tue, Jan 22, 2019 at 04:32:57PM +0100, Marcel Taeumel wrote:
> > Yes, we should backport the new version of ReleaseBuilder to squeak52 and then update the CI script on GitHub to make use of "ReleaseBuilder prepareProcesses" and "ReleaseBuilder saveAndQuit", which would then produce a repaired release image.
> >
> > Kernel-mt.1211 is not needed for it. Only ReleaseBuilder-mt.194
> >
> >
> > Best,
> > Marcel
> > Am 22.01.2019 16:22:29 schrieb David T. Lewis <[hidden email]>:
> > On Tue, Jan 22, 2019 at 07:10:27AM -0600, marcel.taeumel wrote:
> > > Here is an overview of the usual suspects:
> > >
> > >
> > >
> > > Best,
> > > Marcel
> > >
> >
> > That looks very good, thank you for cleaning it up.
> >
> > Should the changes go into the squeak52 stream also?
> >
> > Dave
> >
> >
>
> >
>
>

Reply | Threaded
Open this post in threaded view
|

Re: The Trunk: Kernel-mt.1211.mcz

marcel.taeumel
Thank you. :-)

Am 22.01.2019 19:06:56 schrieb David T. Lewis <[hidden email]>:

Done.

The ReleaseBuilder update is ReleaseBuilder.merge-dtl.191 and the
update map is update-dtl.437.mcm.

Dave

On Tue, Jan 22, 2019 at 11:44:10AM -0500, David T. Lewis wrote:
> Backporting ReleaseBuilder into the squeak52 update stream requires
> adopting the four changes in ReleaseBuilder-mt.194, while skipping the
> other changes in 191, 192, and 193.
>
> In order to prevent file name conflicts between trunk and squeak52,
> I propose saving the new version in squeak52 with the name
> ReleaseBuilder.merge-XXX.191, and making a new update map to point
> to ReleaseBuilder.merge.
>
> If there are no objections I will make those changes soon.
>
> Dave
>
>
> On Tue, Jan 22, 2019 at 04:32:57PM +0100, Marcel Taeumel wrote:
> > Yes, we should backport the new version of ReleaseBuilder to squeak52 and then update the CI script on GitHub to make use of "ReleaseBuilder prepareProcesses" and "ReleaseBuilder saveAndQuit", which would then produce a repaired release image.
> >
> > Kernel-mt.1211 is not needed for it. Only ReleaseBuilder-mt.194
> >
> >
> > Best,
> > Marcel
> > Am 22.01.2019 16:22:29 schrieb David T. Lewis :
> > On Tue, Jan 22, 2019 at 07:10:27AM -0600, marcel.taeumel wrote:
> > > Here is an overview of the usual suspects:
> > >
> > >
> > >
> > > Best,
> > > Marcel
> > >
> >
> > That looks very good, thank you for cleaning it up.
> >
> > Should the changes go into the squeak52 stream also?
> >
> > Dave
> >
> >
>
> >
>
>