Activity for [incr Tcl]

Michał Mrozik created ticket #293

Michał Mrozik — Wed, 12 Jun 2024 08:59:58 -0000

Destructor silently fails to delete class object

[incr Tcl] released /[Incr Tcl_Tk]-source/itcl 3.4.4/itcl3.4.4.tar.gz

[incr Tcl] — Tue, 22 Nov 2022 18:55:03 -0000

[incr Tcl] released /[incr Tcl_Tk]-4-source/itcl 4.2.3/itcl4.2.3.tar.gz

[incr Tcl] — Tue, 22 Nov 2022 18:44:02 -0000

[incr Tcl] released /[incr Tcl_Tk]-4-source/itcl 4.2.2/itcl4.2.2.tar.gz

[incr Tcl] — Fri, 05 Nov 2021 21:08:03 -0000

Szabó László Zoltán created ticket #292

Szabó László Zoltán — Wed, 26 May 2021 12:05:19 -0000

Thread containing an itcl::ensemble crashes by releasing

[incr Tcl] released /[incr Tcl_Tk]-4-source/itcl 4.2.0/itcl4.2.0.tar.gz

[incr Tcl] — Thu, 21 Nov 2019 22:03:02 -0000

Don Porter modified ticket #291

Don Porter — Fri, 31 May 2019 15:46:09 -0000

memory leak in ItclCreateArgList function

Don Porter posted a comment on ticket #291

Don Porter — Fri, 31 May 2019 15:45:14 -0000

Sure looks like this was fixed more than 2 years ago. Upgrade to fix. https://core.tcl-lang.org/itcl/info/0b0d0e0b8cd8105c

Don Porter posted a comment on ticket #291

Don Porter — Fri, 31 May 2019 15:31:56 -0000

That said, the routine ItclCreateArgList() in current development certainly includes ckfree() calls on both argv and defaultArgv. Can you specify the code path that misses those calls and leaks?

Don Porter posted a comment on ticket #291

Don Porter — Fri, 31 May 2019 15:30:34 -0000

~~~ All work on Itcl 4.0.* is abandoned. Upgrade to Itcl 4.1.2 or development trunk.

Gevorg Tevosyan posted a comment on ticket #291

Gevorg Tevosyan — Fri, 31 May 2019 13:43:53 -0000

My Itcl version is 4.0.4

Gevorg Tevosyan created ticket #291

Gevorg Tevosyan — Fri, 31 May 2019 13:34:27 -0000

memory leak in ItclCreateArgList function

Joakim Tjernlund created ticket #54

Joakim Tjernlund — Tue, 28 May 2019 07:36:34 -0000

Please release 4.1.3

Tim Tomkinson created ticket #290

Tim Tomkinson — Thu, 28 Mar 2019 12:50:15 -0000

Class Procedures Can Be Called Like Methods

Tim Tomkinson created ticket #289

Tim Tomkinson — Thu, 28 Mar 2019 12:24:55 -0000

Composition Access Error

Tim Tomkinson created ticket #288

Tim Tomkinson — Thu, 28 Mar 2019 11:22:17 -0000

itcl::local does not support class names that start with a digit

Trey Jackson posted a comment on ticket #271

Trey Jackson — Tue, 09 Oct 2018 23:37:38 -0000

Related performance issue, using latest Itcl (4.1.1) versus 3.2, a 2.6X slowdown in simple script Itcl 3.2: 4.7 seconds Itcl 4.1.1: 12.2 seconds Script is: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17#!/bin/sh #-*-Tcl-*- the next line restarts using wish \ exec tclsh "$0" "$@" puts "Itcl version: [package require Itcl]" itcl::class myclass { public method member_function {} {} public method call_member_function {ntimes} { for {set i 0} {$i < $ntimes} {incr i} { member_function } } } myclass myobj puts...

Trey Jackson created ticket #287

Trey Jackson — Tue, 09 Oct 2018 21:25:06 -0000

itcl::scope and itcl::configbody interaction: can't scope variable: missing object context

Tim Tomkinson created ticket #286

Tim Tomkinson — Sat, 11 Aug 2018 00:49:03 -0000

Recursive method memory leak

Don Porter posted a comment on ticket #117

Don Porter — Wed, 08 Aug 2018 22:45:31 -0000

Once I set up a testbed with all those outdated releases on an OSX High Sierra system, my testing shows no failure running the test script. Sorry, but there continues to be nothing I can do when I cannot reproduce the failure.

Pan Chai posted a comment on ticket #117

Pan Chai — Mon, 06 Aug 2018 21:09:14 -0000

This is still happening. I have more information about this one, Xcode: 9.4+ Mac OSX: Sierra and High Sierra. The test scripts is failing now all the time, once I upgraded Xcode to 9.4.2. Again, I enclosed this script for you.. package require Iwidgets ::iwidgets::radiobox .pref -labeltext "Window Management" .pref add NONE -text "Leave existing windows open when opening new one" .pref add HIDE -text "Hide existing windows when opening new one" .pref add CLOSE -text "Close existing windows when opening...

Trey Jackson posted a comment on ticket #271

Trey Jackson — Tue, 17 Jul 2018 16:17:13 -0000

Can I request a new release of Itcl with this patch on it? 4.1.2? We're trying to finish our porting to Tcl 8.6, and this would help us improve our performance tests to come close to 8.4 performance.

Tim Tomkinson created ticket #285

Tim Tomkinson — Sun, 24 Jun 2018 20:09:45 -0000

Duplicate method/proc error message

Tim Tomkinson created ticket #284

Tim Tomkinson — Sun, 24 Jun 2018 19:46:19 -0000

Remove "my" from constructor error message

Trey Jackson posted a comment on ticket #271

Trey Jackson — Fri, 08 Jun 2018 18:17:48 -0000

That was huge. Thanks. The time spent in Tcl_Preserve/Tcl_Release is much reduced (from a total of >20 seconds to about 6 seconds). There are no obvious Itcl-related performance issues for this particular usage of my tool. On to the next issue for me!

Poor Yorick posted a comment on ticket #271

Poor Yorick — Fri, 08 Jun 2018 14:33:17 -0000

In commit 7388d7698118df3 ItclMemberCode uses reference counting instead of Itcl_PreserveData.

Trey Jackson posted a comment on ticket #271

Trey Jackson — Mon, 04 Jun 2018 19:46:53 -0000

Gladly.

Don Porter posted a comment on ticket #271

Don Porter — Mon, 04 Jun 2018 19:21:33 -0000

When an ItclMemberCode improvement is done, we'll want you to try it and determine how much it helps you.

Trey Jackson posted a comment on ticket #271

Trey Jackson — Mon, 04 Jun 2018 19:17:54 -0000

So... did you want something specific from me? I could update the Itcl code to see which calls are actually being taken - along with a counter. Let me know, I can modify things.

Don Porter posted a comment on ticket #271

Don Porter — Mon, 04 Jun 2018 12:58:16 -0000

Looks like the next low-hanging fruit in preserve elimination is ItclMemberCode. Give it a proper refcount as we've done with several other structs now.

Don Porter posted a comment on ticket #271

Don Porter — Mon, 04 Jun 2018 12:54:13 -0000

Sorry, no. At base that implementation also calls Tcl_Preserve.

Don Porter posted a comment on ticket #271

Don Porter — Mon, 04 Jun 2018 12:53:15 -0000

There's a build option in Itcl, ITCL_PRESERVE_DEBUG that makes it use its own implementation instead of a call to Tcl_Preserve. It's possible that will have a different performance impact.

Don Porter posted a comment on ticket #271

Don Porter — Mon, 04 Jun 2018 12:48:41 -0000

~~~~ Ok. It's only in the implementation of Itcl_PreserveData(), but that still gets significant use.

Don Porter posted a comment on ticket #271

Don Porter — Mon, 04 Jun 2018 12:46:33 -0000

~~~~ I see only 3 calls to Tcl_Preserve() left in the Itcl trunk of development. You're really seeing one or more of them account for significant runtime? Please tell more.

Trey Jackson posted a comment on ticket #271

Trey Jackson — Fri, 01 Jun 2018 23:10:24 -0000

Side question - in my profiling, the vast majority of the time (between 3.2 and 4.1.1) appears to be in Tcl_Preserve/Tcl_Release, and those appear to be relatively unchanged between Tcl 8.4 and 8.6. Any idea why this might be? Just used more now? Perhaps teh Tcl_MutexLock/Unlock is slower? (I've got a single Tcl interpreter, there might be other threads in the application, but they're not Tcl related....)

Trey Jackson posted a comment on ticket #271

Trey Jackson — Fri, 01 Jun 2018 17:47:02 -0000

Object creation local to a proc is 3-4X slower in 4.1.1 vs. 3.2 (don' t have 3.4 numbers) Itcl 3.2 12 microseconds per iteration for 10000 creation of objects in a proc call Itcl 4.1.1 45.0342 microseconds per iteration for 10000 creation of objects in a proc call Note: it's only about 36 microseconds if the SimpleClass is an empty class, ala itcl::class SimpleClass {} But it's instructive that adding a variable and method to a class increases the runtime by ~30% in 4.1.1, whereas in 3.2, the runtime...

Trey Jackson posted a comment on ticket #271

Trey Jackson — Fri, 01 Jun 2018 17:37:50 -0000

Understood. Just pointing out areas of potential slowdown. I'll put in a couple more.

Don Porter posted a comment on ticket #271

Don Porter — Fri, 01 Jun 2018 17:35:30 -0000

Confirmed I see factor of 4 slower operation in Itcl 4.1.1 compared with 3.4.4. No idea yet whether there's anything reasonable that can be done about it.

Don Porter posted a comment on ticket #271

Don Porter — Fri, 01 Jun 2018 17:33:44 -0000

I don't think it's a bug, no. I think it's intrustion into an implementation detail. If there's documentation otherwise, let me know.

Trey Jackson posted a comment on ticket #271

Trey Jackson — Fri, 01 Jun 2018 17:30:27 -0000

So, where do I file a bug against 3.4.4? It works in 4.1.1, and 3.2, so it seems like a bug, no?

Don Porter posted a comment on ticket #271

Don Porter — Fri, 01 Jun 2018 17:30:04 -0000

ok, I see. "proc: -> "itcl::body" makes it work ok.

Don Porter posted a comment on ticket #271

Don Porter — Fri, 01 Jun 2018 17:28:29 -0000

Indeed these test scripts do not run in Itcl 3.4.4. That's what I'm willing to test.

Trey Jackson modified a comment on ticket #271

Trey Jackson — Fri, 01 Jun 2018 17:12:53 -0000

Fair enough. Here's the singleton example - no code change, works for me with Tcl 8.67 and Itcl 4.1.1 (actually latest fossil version), and in 8.4 3.2 Note: GetSingleton is defined using [proc], which has always worked for us (Itcl 3.2 or 4.1) - but it does have problems with Itcl 3.4. It's fixed by changing that to [itcl::body]. Perhaps that's the issue you saw (Itcl 3.4?). Anyway, output first, then script Tcl 8.6.7 Itcl 4.1.1 12.9345 microseconds per iteration per each 10000 class calls that use...

Trey Jackson posted a comment on ticket #271

Trey Jackson — Fri, 01 Jun 2018 17:08:55 -0000

Don Porter posted a comment on ticket #283

Don Porter — Fri, 01 Jun 2018 14:27:37 -0000

This is probably related: http://core.tcl.tk/tcl/tktview?name=3610400fff

Don Porter posted a comment on ticket #283

Don Porter — Fri, 01 Jun 2018 14:22:51 -0000

~~~~ This change is due to a re-implementation in terms of [tailcall]. Here's a Tcl-only demonstration: % proc cb args {puts TRACE:$args} % proc demo {} { tailcall string cat x } % trace add execution demo leave cb % demo TRACE:demo 0 {} leave x If execution traces and [tailcall] are not supposed to interact this way, it's a matter to get resolved in Tcl itself.

Don Porter posted a comment on ticket #271

Don Porter — Fri, 01 Jun 2018 14:01:44 -0000

Maybe we can try this with one focused, working example at a time?

Don Porter posted a comment on ticket #271

Don Porter — Fri, 01 Jun 2018 14:01:11 -0000

Demo script doesn't work: invalid command name "GetSingleton" while executing "GetSingleton" (procedure "CWithProc::CallSomeMethod" line 2) invoked from within "CWithProc::CallSomeMethod" invoked from within "time {CWithProc::CallSomeMethod} $count" (procedure "test_calls_to_singleton_via_class_proc" line 2) invoked from within "test_calls_to_singleton_via_class_proc" (file "/home/dgp/sf271.tcl" line 274)

Trey Jackson posted a comment on ticket #271

Trey Jackson — Wed, 30 May 2018 21:27:59 -0000

I've been tracking down performance issues with Itcl, and while 4.1.1 (and top of trunk) definitely improve things, there are a couple of simple patterns which exhibit degradation from Itcl 3.2 local object creation/deletion in Tcl proc: 4.1.1: 48micros 3.2: 11micros, ratio: ~4:1 call to member method via eval and itcl::code: 4.1.1 4.8micros, 3.2: 1micros, ratio: ~5:1 a simple singleton pattern: 4.1.1 13.7micros, 3.2: 2 micros. ratio: ~7:1 other patterns have about a 30-50% performance penalty Output...

Trey Jackson created ticket #283

Trey Jackson — Wed, 30 May 2018 18:06:13 -0000

Itcl object creation interaction with [trace add execution Class leave]

Don Porter modified ticket #267

Don Porter — Mon, 02 Apr 2018 14:04:56 -0000

"info vars" fails with no pattern

Don Porter posted a comment on ticket #267

Don Porter — Mon, 02 Apr 2018 14:04:32 -0000

Fix committed to trunk. Will be in Itcl 4.1.2

Don Porter posted a comment on ticket #267

Don Porter — Mon, 02 Apr 2018 14:01:51 -0000

Don't know why it wasn't clear before, but this is a simple fix.

Don Porter posted a comment on ticket #267

Don Porter — Mon, 02 Apr 2018 13:43:43 -0000

$ tclsh % dict get [namespace ensemble configure info -map] vars ::tcl::info::vars % package require Itcl 4.1.1 % dict get [namespace ensemble configure info -map] vars ::itcl::builtin::Info::vars

Don Porter posted a comment on ticket #267

Don Porter — Mon, 02 Apr 2018 13:39:06 -0000

The "dgp-method-type" branch has become the Itcl 4.1.* branch which is now the trunk, so this needs a solution.

[incr Tcl] released /[Incr Widgets]/4.1.1/iwidgets-4.1.1.tar.gz

[incr Tcl] — Thu, 22 Mar 2018 16:35:03 -0000

Don Porter posted a comment on ticket #281

Don Porter — Thu, 15 Feb 2018 22:01:13 -0000

Fixed for Itcl 4.1.2

Don Porter modified ticket #281

Don Porter — Thu, 15 Feb 2018 22:01:13 -0000

Protected common variables not visible to children

Tim Tomkinson created ticket #282

Tim Tomkinson — Thu, 15 Feb 2018 20:16:54 -0000

Missing object name should generate an error

Don Porter modified ticket #119

Don Porter — Wed, 14 Feb 2018 20:06:33 -0000

Throwing an error from within an itcl::ensemble causes a segmentation fault (core dump)

Don Porter posted a comment on ticket #119

Don Porter — Wed, 14 Feb 2018 20:06:33 -0000

Fixed for Itcl 4.1.2

Don Porter posted a comment on ticket #119

Don Porter — Tue, 13 Feb 2018 19:23:34 -0000

By all appearances, in the entire Itcl 4 development history, this just hasn't been done. I suspect it's not hard to do, but I'm not ready to predict a completion date. You can keep using Itcl 3.4.4 with Tcl 8.6 in the meantime.

Trey Jackson posted a comment on ticket #119

Trey Jackson — Tue, 13 Feb 2018 18:46:22 -0000

Is there a way to work around this issue temporarily?

Don Porter posted a comment on ticket #119

Don Porter — Tue, 13 Feb 2018 17:06:00 -0000

Itcl_InvokeEnsembleMethod is failing to set a ProcErrorProc.

Tim Tomkinson created ticket #281

Tim Tomkinson — Tue, 13 Feb 2018 15:47:03 -0000

Protected common variables not visible to children

Trey Jackson created ticket #119

Trey Jackson — Mon, 12 Feb 2018 23:33:17 -0000

Throwing an error from within an itcl::ensemble causes a segmentation fault (core dump)

[incr Tcl] released /[incr Tcl_Tk]-4-source/itcl 4.1.1/itcl4.1.1.tar.gz

[incr Tcl] — Thu, 21 Dec 2017 20:04:02 -0000

Don Porter modified ticket #118

Don Porter — Mon, 18 Dec 2017 17:52:42 -0000

application error when using combobox dropdown

Don Porter posted a comment on ticket #118

Don Porter — Mon, 18 Dec 2017 17:52:22 -0000

Dup of bug 115. Already fixed.

Trey Jackson created ticket #118

Trey Jackson — Fri, 15 Dec 2017 21:48:13 -0000

application error when using combobox dropdown

Don Porter modified ticket #280

Don Porter — Tue, 28 Nov 2017 17:53:10 -0000

Obsolete references to "@itcl" syntax

Don Porter modified ticket #280

Don Porter — Tue, 28 Nov 2017 17:15:28 -0000

Obsolete references to "@itcl" syntax

Don Porter modified ticket #279

Don Porter — Tue, 28 Nov 2017 17:13:51 -0000

Extra quote in "scope" error message

Don Porter posted a comment on ticket #279

Don Porter — Tue, 28 Nov 2017 17:13:30 -0000

fixed for Itcl 3.4.4 and 4.1.1

Don Porter posted a comment on ticket #279

Don Porter — Tue, 28 Nov 2017 16:54:12 -0000

It appears this is an error inherited from Itcl 3.

Tim Tomkinson created ticket #280

Tim Tomkinson — Wed, 22 Nov 2017 19:32:33 -0000

Obsolete references to "@itcl" syntax

Tim Tomkinson created ticket #279

Tim Tomkinson — Wed, 22 Nov 2017 19:26:52 -0000

Extra quote in "scope" error message

Tim Tomkinson posted a comment on ticket #271

Tim Tomkinson — Tue, 31 Oct 2017 15:54:07 -0000

I don't think this can get much faster given its current design. There are a lot of new calls to the Tcl core that the old code didn't make (e.g. Tcl_NewObjectInstance, dictionary functions, etc.), which is where most of the time is spent. Too bad, but I guess that's the price of progress.

Don Porter posted a comment on ticket #271

Don Porter — Mon, 30 Oct 2017 16:35:06 -0000

tclconfig is a separate project core.tcl.tk/tclconfig/ See these instructions http://core.tcl.tk/itcl/wiki?name=Tracking+Itcl+Sources

Tim Tomkinson posted a comment on ticket #271

Tim Tomkinson — Mon, 30 Oct 2017 16:16:12 -0000

That's a lot better! The creation time is now constant, but it's still significantly slower than 3.4.3. I'll look around some more to see if there's anything obvious. BTW, the tclconfig directory seems to be missing from the trunk.

Don Porter posted a comment on ticket #271

Don Porter — Mon, 30 Oct 2017 11:45:43 -0000

ok, now synced.

Don Porter posted a comment on ticket #271

Don Porter — Sun, 29 Oct 2017 21:14:27 -0000

Improvement committed to trunk. See what you think.

Don Porter posted a comment on ticket #271

Don Porter — Sun, 29 Oct 2017 19:43:24 -0000

Guess we had the same plans for a Sunday afternoon. Will tackle this soon.

Don Porter modified ticket #271

Don Porter — Sun, 29 Oct 2017 19:42:00 -0000

Object creation/deletion speed and memory usage

Don Porter posted a comment on ticket #271

Don Porter — Sun, 29 Oct 2017 19:42:00 -0000

Well this may not be too bad. The line that creates the trouble is: https://core.tcl.tk/itcl/artifact/cf873c608adaa772?ln=272 Itcl_PreserveData() is crude and inefficient. It needs replacing with an appropriate refcounting scheme directly on the ioPtr argument.

Tim Tomkinson posted a comment on ticket #271

Tim Tomkinson — Sun, 29 Oct 2017 17:48:21 -0000

It looks like the problem is in the Itcl_PreserveData() and Itcl_ReleaseData() functions in generic/itclUtil.c, which in turn call Tcl_Preserve() and Tcl_Release(). These functions use a linear search, which is why the time is increasing linearly. The Tcl documentation even states, "The mechanism is intended only for short-term use (i.e. while procedures are pending on the stack); it will not work efficiently as a mechanism for long-term reference counts." Itcl 3.4 uses a hash table implementation,...

Don Porter posted a comment on ticket #271

Don Porter — Fri, 27 Oct 2017 15:45:35 -0000

On 10/27/2017 11:19 AM, Tim Tomkinson wrote: This is still a problem with Itcl 4.1.0. Itcl 3.4's object creation time is constant at about 2 µs, while Itcl 4.1.0's time increases linearly (see attachments). Is this going to be too difficult to fix based on Itcl 4's internal design? Looks like object creation must still be doing something linear in the number of existing objects. Any idea what that is? -- | Don Porter Applied and Computational Mathematics Division | | donald.porter@nist.gov Information...

Tim Tomkinson posted a comment on ticket #271

Tim Tomkinson — Fri, 27 Oct 2017 15:19:47 -0000

This is still a problem with Itcl 4.1.0. Itcl 3.4's object creation time is constant at about 2 µs, while Itcl 4.1.0's time increases linearly (see attachments). Is this going to be too difficult to fix based on Itcl 4's internal design?

Tim Tomkinson posted a comment on ticket #278

Tim Tomkinson — Tue, 24 Oct 2017 20:58:11 -0000

Thanks for the quick response!

Don Porter modified ticket #278

Don Porter — Tue, 24 Oct 2017 16:23:58 -0000

"is object -class" fails in namespace

Don Porter posted a comment on ticket #278

Don Porter — Tue, 24 Oct 2017 16:23:29 -0000

Fixed for Itcl 4.1.1

Don Porter posted a comment on ticket #278

Don Porter — Tue, 24 Oct 2017 15:20:11 -0000

Confirmed. Appears this has never worked in the Itcl 4 development series.

Tim Tomkinson created ticket #278

Tim Tomkinson — Tue, 24 Oct 2017 15:09:10 -0000

"is object -class" fails in namespace

Don Porter modified ticket #114

Don Porter — Mon, 02 Oct 2017 14:19:18 -0000

iwidgets fails for TCL 6.4.5

Don Porter posted a comment on ticket #117

Don Porter — Mon, 02 Oct 2017 13:31:43 -0000

I am sorry to report that my builds of all those packages on an OSX Sierra system do not demonstrate a crash on the submitted script. If you can supply more information, I can try again, but if I cannot reproduce the trouble I cannot really begin to chase down what might be the problem.

Pan Chai posted a comment on ticket #117

Pan Chai — Thu, 28 Sep 2017 13:02:12 -0000

Don, As I indicated in the bug report, this issue only happens in Mac OS X (not in Linux nor Cgywin.) And the script does not crash every time, only randomly but quite often (mostly 7 out of 10.) From my own testing, I would suggest the possibility of issue between itk 4.1.0 and TK 8.6.6 (I had reported one earlier, #115.) I can try to upgrade TK to 8.6.7 or downgrade itk to 4.0.6 as you suggested to see if it helps. However, in a broader sense, since using Tcl/Tk is not limited to one specific project...

Don Porter posted a comment on ticket #117

Don Porter — Thu, 28 Sep 2017 11:36:48 -0000

Thank you for a good report. The demo script seems to work fine on a linux system. I will try on an OS X system later when I am on one. Perhaps next week. My underinformed guess at this point is you've run into some platform-specific trouble in Itk 4.1.0. or possibly Tk 8.6.6. If you have time to blindly try things, you might look into using Tcl/Tk 8.6.7 and/or Itcl/Itk 4.0.6. Will be able to give better answers after I reproduce the trouble. Should I expect a crash each time?

Pan Chai created ticket #117

Pan Chai — Thu, 28 Sep 2017 01:11:52 -0000

Mac OS X: random crashes of wish (tcl/tk 8.6.6) using Iwidgets (4.1.0) radiobox class

Don Porter modified ticket #116

Don Porter — Wed, 06 Sep 2017 13:42:27 -0000

iwidgets 4.1.0 configure error in line 6515

Thomas Perschak posted a comment on ticket #116

Thomas Perschak — Wed, 06 Sep 2017 11:21:10 -0000

This fixed the issue. thx!