help / color / mirror / code / Atom feed
From: Bart Schaefer <schaefer@brasslantern.com>
To: "Johan Ström" <johan@stromnet.se>
Cc: Jun T <takimoto-j@kba.biglobe.ne.jp>,
	Zsh hackers list <zsh-workers@zsh.org>
Subject: Re: 5.9 core dumps under zleentry->zleread->zlecore
Date: Sat, 4 Jun 2022 13:06:39 -0700	[thread overview]
Message-ID: <CAH+w=7akr+BoyGJRTG30UWpuE9iHh_Tc4QmubMCJMFT-VRYXpg@mail.gmail.com> (raw)
In-Reply-To: <54c2c561-a3dc-5710-d6b6-6e235a6396b8@stromnet.se>

On Sat, Jun 4, 2022 at 12:39 AM Johan Ström <johan@stromnet.se> wrote:
> Hm, I had cases where terminals have been on the same virtual desktop,
> with no re-arranging going on at the moment, or since last being used
> (basically only resuming work in existing windows).
> Not sure if SIGWINCH would be sent by just switching virtual desktop (sway).

One reason I'm speculating about this is because of this from your
previous stack trace:

                 #9  0x000055ff7ab040f8 zalloc (zsh + 0x730f8)
                 #10 0x00007fb08a549417 calclist (complete.so + 0x1b417)
                 #11 0x00007fb08a54b040 ilistmatches (complete.so + 0x1d040)

There's only one zalloc() in calclist:

            ws = g->widths = (int *) zalloc(zterm_columns * sizeof(int));

If zterm_columns has somehow become messed up, this might run you out of memory.

Then in the other trace you have:

                 #10 0x000055d4aa005493 zhandler (zsh + 0xa8493)
                 #11 0x00007f2f935388e0 n/a (libc.so.6 + 0x3e8e0)
                 #12 0x00007f2f935fbef2 read (libc.so.6 + 0x101ef2)

That definitely means a signal was received, and SIGWINCH is
explicitly unblocked around that read() call, but it goes straight on

                 #9  0x000055d4a9fc3874 zleentry (zsh + 0x66874)

The only way to get to that from a signal handler is when a
user-defined trap is run:

     * If zle was running while the trap was executed, see if we
     * need to restore the display.
    if (zleactive && resetneeded)

However (I suspect) the compiler has optimized out a couple of
intervening function calls, which is why I hope you can get this to
reoccur with a debug build.

> In /etc/zsh, only zprofile:
> https://github.com/archlinux/svntogit-packages/blob/packages/zsh/trunk/zprofile
> Which references this:
> https://github.com/archlinux/svntogit-packages/blob/packages/filesystem/trunk/profile

Yes, which does
  emulate sh -c 'source /etc/profile'
though it seems unlikely that anything too weird is coming in that way.

      reply	other threads:[~2022-06-04 20:07 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-02  7:11 Johan Ström
2022-06-02  7:31 ` Bart Schaefer
2022-06-02  9:25   ` Johan Ström
2022-06-03  7:14 ` Jun T
2022-06-03  7:19   ` Johan Ström
2022-06-03  9:53     ` Jun T
2022-06-03 10:02       ` Johan Ström
2022-06-04  3:32         ` Bart Schaefer
2022-06-04  7:39           ` Johan Ström
2022-06-04 20:06             ` Bart Schaefer [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAH+w=7akr+BoyGJRTG30UWpuE9iHh_Tc4QmubMCJMFT-VRYXpg@mail.gmail.com' \
    --to=schaefer@brasslantern.com \
    --cc=johan@stromnet.se \
    --cc=takimoto-j@kba.biglobe.ne.jp \
    --cc=zsh-workers@zsh.org \
    --subject='Re: 5.9 core dumps under zleentry->zleread->zlecore' \


* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

Code repositories for project(s) associated with this inbox:


This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).