GNU bug report logs - #75672
31.0.50; scratch/igc memory usage/collection issues

Previous Next

Full log

Message #20 received at 75672 <at> debbugs.gnu.org (full text, mbox):

From: Pip Cet <pipcet <at> protonmail.com>
To: Stefan Kangas <stefankangas <at> gmail.com>
Cc: Alexis Purslane <alexispurslane <at> pm.me>, 75672 <at> debbugs.gnu.org,
 acorallo <at> gnu.org, gerd <at> gnu.org
Subject: Re: bug#75672: 31.0.50; scratch/igc memory usage/collection issues
Date: Sun, 19 Jan 2025 17:43:51 +0000

"Stefan Kangas" <stefankangas <at> gmail.com> writes:

> Alexis Purslane via "Bug reports for GNU Emacs, the Swiss army knife of
> text editors" <bug-gnu-emacs <at> gnu.org> writes:
>
>> I've been using scratch/igc for the past few days (see details below for
>> the exact version and situation) and have been having some /interesting/
>> experiences:
>>
>>     1. It seems slower than Emacs 29.4 with GCMH (gcmh-idle-timer set to
>>     'auto, so it GCs usually after less than a second of idle time, and
>>     gcmh-max-cons-threshold set to 200MB so that if there's a very
>>     intense operation going on that's allocating a lot of memory, it
>>     will actually GC before it gets out of hand and the idle GC pause
>>     would freeze Emacs, but most of the time it'll almost never GC
>>     unless I'm idle) when opening a lot of files (for instance, when
>>     running org-agenda). About twice as slow. Startup is faster, though.
>
> Thanks for the report.  I'm repyling only to the GCMH part here.
>
> For context, GCMH is a package by Andrea that tries to avoid GC by
> setting a high `gc-cons-threshold` when Emacs is in use, and then
> triggers GC when Emacs is idle.  Details here:
>
>     https://elpa.gnu.org/packages/gcmh.html
>
> Do we expect that running GCMH will be useful with IGC?

No.  We want to optimize IGC's usage of MPS.

> First, I'm not sure that `gc-cons-threshold` will even affect MPS.

It shouldn't.  The decision what to do with an incremental garbage
collector is so different that we cannot do anything useful with that
variable.

> Second, we already run 'igc_on_idle' periodically.

We'll have a background thread soon, to see how that works :-)

> Third, GCMH will run `garbage-collect`, which doesn't trigger a full GC
> with MPS, AFAIU by design.  GCMH would need to run `igc--collect`, but
> according to Gerd we should treat that function as a debugging tool.
>
> Maybe GCMH should simply do nothing if (featurep 'mps)?

SGTM.  Andrea?

Pip

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.