bug-gnu-emacs
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

bug#62794: 28.2; jit-lock-function enters busy loop when open """ near t


From: Nick Urbanik
Subject: bug#62794: 28.2; jit-lock-function enters busy loop when open """ near top of large Python program
Date: Thu, 13 Apr 2023 15:39:18 +1000

On 13/04/23 07:58 +0300, Eli Zaretskii wrote:
Date: Thu, 13 Apr 2023 09:55:42 +1000
From: Nick Urbanik <nicku@nicku.org>
Cc: 62794@debbugs.gnu.org

>Please post such a file, or a URL where it could be downloaded.  I
>couldn't easily find such a large file on my system.

It turns out that it is not the size, but some pathological
arrangement of the code.  I have provided a much smaller sample here
that I can reproduce the problem with by:

1. Move the cursor to below the from elesewhere import ... line.
2. Press Enter
3. Type:
def function():
     """

4. At this point, CPU usage goes to 100%, I cannot type anything else
into the buffer, but I can send a USR2 signal to end the 100% CPU
usage, but the buffer has only partial syntax colouring.

OK, thanks.  I can indeed reproduce this in Emacs 28.2, but not in
what soon will become Emacs 29.1.  So I think this problem was already
solved between these two version.

>I suspect this is already fixed in Emacs 29, but I cannot verify
>without a file that reproduces the problem in Emacs 28.

I will build an RPM package of emacs 29 and test that.

Thanks.  I will wait for your confirmation before closing the bug as
fixed.

I spent a couple of hours attempting to build an RPM package from git,
but each iteration takes a super long time, and I cannot afford to
spend a longer time from my work.  I may try again on the weekend.

For me this is a really big problem, as it has happened in other cases
of editing Python programs, and it is a big slowdown and interruption
to my workflow to restart emacs with lots of frames open.  Emacs has
always been so solid till this.

Fedora will not package emacs 29.1 before it is released.  I hope that
is soon.  F38 is due in 13 days, so this issue is likely to remain.

It would be fantastic to backport the fix.
--
Nick Urbanik             http://nicku.org           nicku@nicku.org
GPG: 7FFA CDC7 5A77 0558 DC7A 790A 16DF EC5B BB9D 2C24 ID: BB9D2C24





reply via email to

[Prev in Thread] Current Thread [Next in Thread]