GNU bug report logs - #81012
30.2; Letters are sometimes unexpectedly not fully drawn

Please note: This is a static page, with minimal formatting, updated once a day.
Click here to see this page with the latest information and nicer formatting.

Package: emacs; Reported by: Markus Triska <triska@HIDDEN>; dated Sun, 10 May 2026 08:31:01 UTC; Maintainer for emacs is bug-gnu-emacs@HIDDEN.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 6 Jun 2026 11:28:59 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sat Jun 06 07:28:59 2026
Received: from localhost ([127.0.0.1]:37961 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wVpD0-0002XC-G2
	for submit <at> debbugs.gnu.org; Sat, 06 Jun 2026 07:28:59 -0400
Received: from eggs.gnu.org ([2001:470:142:3::10]:59646)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <eliz@HIDDEN>) id 1wVpCy-0002WS-52
 for 81012 <at> debbugs.gnu.org; Sat, 06 Jun 2026 07:28:56 -0400
Received: from fencepost.gnu.org ([2001:470:142:3::e])
 by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.90_1) (envelope-from <eliz@HIDDEN>)
 id 1wVpCs-0004YF-5X; Sat, 06 Jun 2026 07:28:50 -0400
DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=gnu.org;
 s=fencepost-gnu-org; h=References:Subject:In-Reply-To:To:From:Date:
 mime-version; bh=YgvOvGXXi5xUz8Pq1YkUSELPPT2/ZR9ANqEXxjevOvk=; b=rx7IJqRjZReE
 Rp7N+mOXag//UTqc9fuJbmLGWhd+eIAz9Ha/iYLnZUqaaeGU91+GmJc6gp9g3vF5+qBeyls4NKamj
 sf3o+kbBF3Pa8TNXSVCqUt0XPqVTJiopJXVLxaAy19/51o+ztgQK0RDVRkJyo7VXwXDQoZv8EEFUv
 /W3UAZwmX2DGcaKnNQC3EyvDQ5UE5Za2vVSl3YOSVC7Nt/ew4UfuxsPrph+bFJNFr8uoyqHPAGQwm
 2Q+0arvUxFCfVzD4oA1ZEanSQmwkx46+LrQ6wCYojxcFK4OEGY4GE5tiOuUngVVmfbUCHWTScn6yg
 +jMpoFFVxt0Oi/jnL7/R9g==;
Date: Sat, 06 Jun 2026 14:28:47 +0300
Message-Id: <86bjdnlxg0.fsf@HIDDEN>
From: Eli Zaretskii <eliz@HIDDEN>
To: Markus Triska <triska@HIDDEN>
In-Reply-To: <87bjdo3prl.fsf@HIDDEN> (message from Markus Triska on Sat, 
 06 Jun 2026 12:51:58 +0200)
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <87ik8nuqpk.fsf@HIDDEN> <87pl2vb0ib.fsf@HIDDEN>
 <877bp25gn9.fsf@HIDDEN> <86ik7wko4q.fsf@HIDDEN>
 <87bjdo3prl.fsf@HIDDEN>
X-Spam-Score: -2.3 (--)
X-Debbugs-Envelope-To: 81012
Cc: pipcet@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -3.3 (---)

> From: Markus Triska <triska@HIDDEN>
> Cc: pipcet@HIDDEN,  81012 <at> debbugs.gnu.org
> Date: Sat, 06 Jun 2026 12:51:58 +0200
> 
> Eli Zaretskii <eliz@HIDDEN> writes:
> 
> > understand how we should resolve this.  Is an entry in PROBLEMS
> > enough?  If so, would someone please propose the text for such an
> > entry?  If not, do we have a patch to fix this?
> 
> Updated input from me is as follows:
> 
>   1. For the past 3 weeks, I ran Emacs exclusively with Pip's patch applied.
>   2. With the patch applied, I have never again encountered the issue,
>      on neither of two different machines both running Debian Trixie.
>   3. With the patch applied, I noticed no regressions: Everything I use
>      Emacs for still works also with the patch applied.
> 
> >From a correctness perspective, it would thus seem appropriate to install
> Pip's patch, since it fixes a correctness issue, and the only drawback
> seems to be a potential slowdown in certain usage modes (which I did not
> encounter in what I use Emacs for).
> 
> If the patch were installed, the potential slowdown could be documented
> in PROBLEMS, and it could be mitigated with a customizeable flag that
> yields more speed at the cost of sometimes drawing characters only partially.
> 
> Would this be acceptable?

What kind of slowdown is that?  You describe no slowdown above, so I'm
unsure what does this allude to.




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 6 Jun 2026 10:52:02 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sat Jun 06 06:52:01 2026
Received: from localhost ([127.0.0.1]:37382 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wVodF-0008Mm-Be
	for submit <at> debbugs.gnu.org; Sat, 06 Jun 2026 06:52:01 -0400
Received: from [78.47.144.35] (port=51122 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wVodD-0008MZ-OJ
 for 81012 <at> debbugs.gnu.org; Sat, 06 Jun 2026 06:52:00 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id 3BFFB9C7AE; Sat,  6 Jun 2026 12:51:58 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Eli Zaretskii <eliz@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <87ik8nuqpk.fsf@HIDDEN> <87pl2vb0ib.fsf@HIDDEN>
 <877bp25gn9.fsf@HIDDEN> <86ik7wko4q.fsf@HIDDEN>
Date: Sat, 06 Jun 2026 12:51:58 +0200
Message-ID: <87bjdo3prl.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview: Eli Zaretskii <eliz@HIDDEN> writes: > understand how we
 should
 resolve this. Is an entry in PROBLEMS > enough? If so, would someone please
 propose the text for such an > entry? If not, do we have a patch to fix this?
 Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: pipcet@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Eli Zaretskii <eliz@HIDDEN> writes:

> understand how we should resolve this.  Is an entry in PROBLEMS
> enough?  If so, would someone please propose the text for such an
> entry?  If not, do we have a patch to fix this?

Updated input from me is as follows:

  1. For the past 3 weeks, I ran Emacs exclusively with Pip's patch applied.
  2. With the patch applied, I have never again encountered the issue,
     on neither of two different machines both running Debian Trixie.
  3. With the patch applied, I noticed no regressions: Everything I use
     Emacs for still works also with the patch applied.

From a correctness perspective, it would thus seem appropriate to install
Pip's patch, since it fixes a correctness issue, and the only drawback
seems to be a potential slowdown in certain usage modes (which I did not
encounter in what I use Emacs for).

If the patch were installed, the potential slowdown could be documented
in PROBLEMS, and it could be mitigated with a customizeable flag that
yields more speed at the cost of sometimes drawing characters only partially.

Would this be acceptable?

Thank you and all the best,
Markus




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 6 Jun 2026 09:35:29 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sat Jun 06 05:35:29 2026
Received: from localhost ([127.0.0.1]:36163 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wVnRB-00005g-4C
	for submit <at> debbugs.gnu.org; Sat, 06 Jun 2026 05:35:29 -0400
Received: from eggs.gnu.org ([2001:470:142:3::10]:37132)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <eliz@HIDDEN>) id 1wVnR8-00004t-Cd
 for 81012 <at> debbugs.gnu.org; Sat, 06 Jun 2026 05:35:27 -0400
Received: from fencepost.gnu.org ([2001:470:142:3::e])
 by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.90_1) (envelope-from <eliz@HIDDEN>)
 id 1wVnR2-0006oJ-9H; Sat, 06 Jun 2026 05:35:20 -0400
DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=gnu.org;
 s=fencepost-gnu-org; h=References:Subject:In-Reply-To:To:From:Date:
 mime-version; bh=hVHaRlSIo/r9D6OmQ7V7QUNz7QTKqw1PHZllG2ImiAA=; b=Q2CT7BtXukKe
 Nge/TzE1BykvySNkNPRUsV6uGzKet7UooA9srairbw4IzHSi1wJ+bpd0qzZYvn7XmZfi9t0A/+zzc
 RhVZnT6Uo6iGgkC4z6MfTvct6Y62bLZgy25HMnvGTVs5OqZ253YMF+i/EC7bftlxcNNeGVkMnQ6PU
 P1e7/v46cDCX4ZLtyx2E9eUbVLsFMO+9p/lWcfbK3oNXkNKpjDbAAPUR0CFVN9Emw6BZ+0xHmMEf6
 Fdmrfv4+LGpPKqF/R7ZTf8i+SND3xLPil3O3LkRv2FrpladtPuaJQbgGEqvgp1i2ofjwB4vKZnQrX
 YiiFFY0dvxC1d2UWr7uZeg==;
Date: Sat, 06 Jun 2026 12:35:17 +0300
Message-Id: <86ik7wko4q.fsf@HIDDEN>
From: Eli Zaretskii <eliz@HIDDEN>
To: pipcet@HIDDEN, Markus Triska <triska@HIDDEN>
In-Reply-To: <877bp25gn9.fsf@HIDDEN> (message from Markus Triska on Sun, 
 17 May 2026 15:07:06 +0200)
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <87ik8nuqpk.fsf@HIDDEN> <87pl2vb0ib.fsf@HIDDEN>
 <877bp25gn9.fsf@HIDDEN>
X-Spam-Score: -2.3 (--)
X-Debbugs-Envelope-To: 81012
Cc: 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -3.3 (---)

Ping!  The discussion seems to have died, and I'm not sure I
understand how we should resolve this.  Is an entry in PROBLEMS
enough?  If so, would someone please propose the text for such an
entry?  If not, do we have a patch to fix this?

> From: Markus Triska <triska@HIDDEN>
> Cc: Eli Zaretskii <eliz@HIDDEN>,  81012 <at> debbugs.gnu.org
> Date: Sun, 17 May 2026 15:07:06 +0200
> 
> Pip Cet <pipcet@HIDDEN> writes:
> 
> > If you could do that and report back if you experience any issues,
> > that'd be great!
> 
> Yes, I will do that! I will now switch to using Emacs with your patch
> applied. So far, everything seems to work nicely.
> 
> Regarding the version *without* your patch installed, one other thing I
> noticed is that the issue does *not* occur if two windows are displayed
> in the frame. It apparently does not make a difference which buffer is
> displayed in the other window, all that seems to matter is that there is
> a second window. For example, the issue I posted does not arise if I do
> C-x 2 (split-window-below) before evaluating the form.
> 
> I think this is also the reason why I can reproduce the issue only when
> launching Emacs with "-Q": "-Q" implies --no-splash, and therefore there
> is only one window in the frame, so the issue arises.
> 
> Maybe this finding can also help to mitigate the issue?
> 
> >
> > So I think it depends on your use case, but if this works and avoids the
> > issue we should look for a way to do the same thing in ftcrfont.c.
> 
> Awesome!
> 
> Thank you so much,
> Markus
> 




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 17 May 2026 13:07:10 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sun May 17 09:07:10 2026
Received: from localhost ([127.0.0.1]:42292 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wObD3-0005oS-VS
	for submit <at> debbugs.gnu.org; Sun, 17 May 2026 09:07:10 -0400
Received: from [78.47.144.35] (port=51654 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wObD1-0005o0-GV
 for 81012 <at> debbugs.gnu.org; Sun, 17 May 2026 09:07:08 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id 3BD309C7A2; Sun, 17 May 2026 15:07:06 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <87ik8nuqpk.fsf@HIDDEN> <87pl2vb0ib.fsf@HIDDEN>
Date: Sun, 17 May 2026 15:07:06 +0200
In-Reply-To: <87pl2vb0ib.fsf@HIDDEN> (Pip Cet's message of "Sat, 16
 May 2026 19:46:07 +0000")
Message-ID: <877bp25gn9.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview:  Pip Cet <pipcet@HIDDEN> writes: > If you could do
 that and report back if you experience any issues, > that'd be great! Yes,
 I will do that! I will now switch to using Emacs with your patch applied.
 So far, everything seems to work nicely. 
 Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Pip Cet <pipcet@HIDDEN> writes:

> If you could do that and report back if you experience any issues,
> that'd be great!

Yes, I will do that! I will now switch to using Emacs with your patch
applied. So far, everything seems to work nicely.

Regarding the version *without* your patch installed, one other thing I
noticed is that the issue does *not* occur if two windows are displayed
in the frame. It apparently does not make a difference which buffer is
displayed in the other window, all that seems to matter is that there is
a second window. For example, the issue I posted does not arise if I do
C-x 2 (split-window-below) before evaluating the form.

I think this is also the reason why I can reproduce the issue only when
launching Emacs with "-Q": "-Q" implies --no-splash, and therefore there
is only one window in the frame, so the issue arises.

Maybe this finding can also help to mitigate the issue?

>
> So I think it depends on your use case, but if this works and avoids the
> issue we should look for a way to do the same thing in ftcrfont.c.

Awesome!

Thank you so much,
Markus




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 16 May 2026 19:46:23 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sat May 16 15:46:23 2026
Received: from localhost ([127.0.0.1]:34934 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wOKxq-0002Td-Sg
	for submit <at> debbugs.gnu.org; Sat, 16 May 2026 15:46:23 -0400
Received: from mail-43102.protonmail.ch ([185.70.43.102]:64065)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <pipcet@HIDDEN>)
 id 1wOKxo-0002TJ-4J
 for 81012 <at> debbugs.gnu.org; Sat, 16 May 2026 15:46:21 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=protonmail.com;
 s=protonmail3; t=1778960773; x=1779219973;
 bh=Xow5dgJr7CtBhUNzwCd7yfRehvM4QMVmmxGrPk+H8dU=;
 h=Date:To:From:Cc:Subject:Message-ID:In-Reply-To:References:
 Feedback-ID:From:To:Cc:Date:Subject:Reply-To:Feedback-ID:
 Message-ID:BIMI-Selector;
 b=Bne2aJbmb7/7kQzPWoVvB5sld74wCXbIBV9pTkj6roO2DrcTszx+Brw3aEWsqnfWU
 CNwxirinKdA2s1YqUwoNl0+mRpxZ00iiz84hSF8mPQDmsH0hPkaNFRezxlang1q8LS
 5+o8oSel8FOkUlJYZAzimV/jV9XTpxvkilV7m+HWHSkUwRgEYHAUg+nqh7CD7G8bng
 16t+1SiIOyGVcTC0XkeJPikuVYUB1i+RtDak8V1Z4WaM6DAwjoGjhfg0XBDXR4Iows
 6h7obciva0Tiy66QKUJbAnAJssCTPtwBja+ImQDXGGbzMGWuBWSNTXZlbTVCE4cFja
 BmfhWKgOIlmuA==
Date: Sat, 16 May 2026 19:46:07 +0000
To: Markus Triska <triska@HIDDEN>
From: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
Message-ID: <87pl2vb0ib.fsf@HIDDEN>
In-Reply-To: <87ik8nuqpk.fsf@HIDDEN>
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <87ik8nuqpk.fsf@HIDDEN>
Feedback-ID: 112775352:user:proton
X-Pm-Message-ID: 2f58fa7955e2c99c2382f7ca1798afe89e5d5c0e
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: -0.7 (/)
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.7 (-)

"Markus Triska" <triska@HIDDEN> writes:

> Pip Cet <pipcet@HIDDEN> writes:
>
>> This is definitely not a fix, but could you try applying this patch and
>> seeing whether the problems still occur (it's specific to the font
>> driver you're using, and a different patch is required for ftcrfont.c) ?
>>
>> diff --git a/src/xftfont.c b/src/xftfont.c
>> index f15dbae1e7a..6b21ec72755 100644
>
> With this patch applied, I no longer see the issue!

Excellent news!

> Are there any problematic issues one should expect when running Emacs
> with this change applied? I'm considering enabling it on my setup.

If you could do that and report back if you experience any issues,
that'd be great!

Apart from (noticeably?) worse performance, I wouldn't expect any major
issues. I think it should be usable in what is the common case today: a
local X server running on a different CPU core from the client
application.  I tried using it with "ssh -Y" over tor, and that was
quite painful.

So I think it depends on your use case, but if this works and avoids the
issue we should look for a way to do the same thing in ftcrfont.c.

> Thank you and all the best,

Thank you!

Pip





Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 16 May 2026 18:58:19 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sat May 16 14:58:18 2026
Received: from localhost ([127.0.0.1]:34355 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wOKDK-00040d-Gc
	for submit <at> debbugs.gnu.org; Sat, 16 May 2026 14:58:18 -0400
Received: from [78.47.144.35] (port=57708 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wOKDI-000404-Fi
 for 81012 <at> debbugs.gnu.org; Sat, 16 May 2026 14:58:17 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id 6FCC09C762; Sat, 16 May 2026 20:58:15 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
Date: Sat, 16 May 2026 20:58:15 +0200
In-Reply-To: <87wlxb9kuz.fsf@HIDDEN> (Pip Cet's message of "Sun, 10
 May 2026 18:32:17 +0000")
Message-ID: <87ik8nuqpk.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview: Pip Cet <pipcet@HIDDEN> writes: > This is definitely
 not a fix, but could you try applying this patch and > seeing whether the
 problems still occur (it's specific to the font > driver you're using, and
 a different patch is required [...] 
 Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Pip Cet <pipcet@HIDDEN> writes:

> This is definitely not a fix, but could you try applying this patch and
> seeing whether the problems still occur (it's specific to the font
> driver you're using, and a different patch is required for ftcrfont.c) ?
>
> diff --git a/src/xftfont.c b/src/xftfont.c
> index f15dbae1e7a..6b21ec72755 100644

With this patch applied, I no longer see the issue!

Are there any problematic issues one should expect when running Emacs
with this change applied? I'm considering enabling it on my setup.

Thank you and all the best,
Markus





Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 16 May 2026 18:30:26 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sat May 16 14:30:26 2026
Received: from localhost ([127.0.0.1]:34212 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wOJmL-0000Bs-9S
	for submit <at> debbugs.gnu.org; Sat, 16 May 2026 14:30:25 -0400
Received: from mout01.posteo.de ([185.67.36.65]:37969)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <yantar92@HIDDEN>)
 id 1wOJmH-00007g-U9
 for 81012 <at> debbugs.gnu.org; Sat, 16 May 2026 14:30:22 -0400
Received: from submission (posteo.de [185.67.36.169]) 
 by mout01.posteo.de (Postfix) with ESMTPS id AB911240029
 for <81012 <at> debbugs.gnu.org>; Sat, 16 May 2026 20:30:15 +0200 (CEST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=posteo.net; s=2017;
 t=1778956215; bh=XHrSNBEJ/+KAT3PbsnDixkZyKtREWWu/k2JB/6oBtqE=;
 h=From:To:Cc:Subject:Date:Message-ID:MIME-Version:Content-Type:
 From;
 b=dXFNZTwM3t3++8Xhy6YkVVvZVxGhotaIcV7nXwLSeDlN4z55X/GYFloFwA06Sdmsa
 gtdVBCOrAeH/GAzOq+d6a/+3J026Ajjk7l8w/2N1xinA+1D2YhburmkgdTFlb4U/5Z
 JFy+6SNgZ+MVsOulaYb4jmeqawzHy/KT6k4R56tSG4q24omjOFufH33/Ag70tjrdTY
 cO7f9twl2ELrPxcmGeEqkDtJpu6aBsy6a9XBo+neR4uPh7BmPhrCcjURzeTKa5g3wf
 csmorN69P32XQQdQVLmyfB43GYKIMXJcefOQspEg+TvkWQz/G5Me/Kl7zZYhVHWEeh
 2J0OuMiDFdBIw==
Received: from customer (localhost [127.0.0.1])
 by submission (posteo.de) with ESMTPSA id 4gHt0R09Vwz6twf;
 Sat, 16 May 2026 20:30:14 +0200 (CEST)
From: Ihor Radchenko <yantar92@HIDDEN>
To: Markus Triska <triska@HIDDEN>
Subject: Re: bug#81012: 30.2; Letters are sometimes unexpectedly not fully
 drawn
In-Reply-To: <874ik7ciud.fsf@HIDDEN>
References: <20260510083506.293CBF08697@triska-ser8>
 <86jytb38az.fsf@HIDDEN> <87se7z1kp6.fsf@HIDDEN>
 <87wlxb9kuz.fsf@HIDDEN> <87o6inp0me.fsf@HIDDEN>
 <87o6in9h6q.fsf@HIDDEN> <87tssfnets.fsf@HIDDEN>
 <86wlxa119r.fsf@HIDDEN> <87y0hp4vfv.fsf@HIDDEN>
 <868q9o28fc.fsf@HIDDEN> <87lddo95e3.fsf@HIDDEN>
 <874ik7ciud.fsf@HIDDEN>
Date: Sat, 16 May 2026 18:30:15 +0000
Message-ID: <87h5o71a2v.fsf@localhost>
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: -2.3 (--)
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, pipcet@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -3.3 (---)

Markus Triska <triska@HIDDEN> writes:

>> To summarize: The suprise is that using "-Q" does not affect the hinting
>> Emacs uses, but it affects the frequency of the present issue. This may
>> indicate that a different factor other than hinting also plays a role.
>
> I have now driven this to the extreme: I have deleted both ~/.emacs and
> ~/.Xresources, and - with the sample snippet I posted - I *still* can
> elicit the issue only when I invoke Emacs with "-Q". When I invoke Emacs
> as "emacs", I do not observe the issue with the instructions I gave.

The difference between -Q and no -Q with absent .Xresources and
.emacs/.emacs.d is system-wide site file. Maybe your distro is doing something?

-- 
Ihor Radchenko // yantar92,
Org mode maintainer,
Learn more about Org mode at <https://orgmode.org/>.
Support Org development at <https://liberapay.com/org-mode>,
or support my work at <https://liberapay.com/yantar92>




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 16 May 2026 18:25:34 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sat May 16 14:25:34 2026
Received: from localhost ([127.0.0.1]:34200 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wOJhe-0007sw-3e
	for submit <at> debbugs.gnu.org; Sat, 16 May 2026 14:25:34 -0400
Received: from [78.47.144.35] (port=57314 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wOJhb-0007rB-LI
 for 81012 <at> debbugs.gnu.org; Sat, 16 May 2026 14:25:32 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id 581439C79E; Sat, 16 May 2026 20:25:30 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Eli Zaretskii <eliz@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <87o6inp0me.fsf@HIDDEN> <87o6in9h6q.fsf@HIDDEN>
 <87tssfnets.fsf@HIDDEN> <86wlxa119r.fsf@HIDDEN>
 <87y0hp4vfv.fsf@HIDDEN> <868q9o28fc.fsf@HIDDEN>
 <87lddo95e3.fsf@HIDDEN>
Date: Sat, 16 May 2026 20:25:30 +0200
In-Reply-To: <87lddo95e3.fsf@HIDDEN> (Markus Triska's message of "Tue,
 12 May 2026 14:30:44 +0200")
Message-ID: <874ik7ciud.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview: > To summarize: The suprise is that using "-Q" does not
 affect
 the hinting > Emacs uses, but it affects the frequency of the present issue.
 This may > indicate that a different factor other than hinti [...] 
 Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: pipcet@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)


> To summarize: The suprise is that using "-Q" does not affect the hinting
> Emacs uses, but it affects the frequency of the present issue. This may
> indicate that a different factor other than hinting also plays a role.

I have now driven this to the extreme: I have deleted both ~/.emacs and
~/.Xresources, and - with the sample snippet I posted - I *still* can
elicit the issue only when I invoke Emacs with "-Q". When I invoke Emacs
as "emacs", I do not observe the issue with the instructions I gave.

All the best,
Markus




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 16 May 2026 06:59:10 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sat May 16 02:59:10 2026
Received: from localhost ([127.0.0.1]:54790 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wO8zO-0005eP-EN
	for submit <at> debbugs.gnu.org; Sat, 16 May 2026 02:59:10 -0400
Received: from [78.47.144.35] (port=49922 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wO8zM-0005eE-LE
 for 81012 <at> debbugs.gnu.org; Sat, 16 May 2026 02:59:09 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id 55C939C79E; Sat, 16 May 2026 08:59:07 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Mike Kupfer <kupfer@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8>
 <87wlxb9kuz.fsf@HIDDEN> <864ikf2hzo.fsf@HIDDEN>
 <878q9odcqa.fsf@HIDDEN> <86tssczt0e.fsf@HIDDEN>
 <87cxyzf1tn.fsf@HIDDEN> <87mry2ge1l.fsf@HIDDEN>
 <87jyt4erlq.fsf@HIDDEN> <874ik8911m.fsf@HIDDEN>
 <87bjegehgw.fsf@HIDDEN> <7209.1778880004@HIDDEN>
Date: Sat, 16 May 2026 08:59:07 +0200
In-Reply-To: <7209.1778880004@HIDDEN> (Mike Kupfer's message of
 "Fri, 15 May 2026 14:20:04 -0700")
Message-ID: <87pl2v4z7o.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview:  Mike Kupfer <kupfer@HIDDEN> writes: > I'd also like to
 see some mention of changing the cursor shape. I have > not seen any glitches
 since switching to 'bar'. I can confirm this: With the sample program I posted,
 if I add, after (fundamental-mode), the form (setq cursor-type 'bar), then
 the issue no longer appears whereas it otherwise appears consistently. 
 Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: Pip Cet <pipcet@HIDDEN>, Eli Zaretskii <eliz@HIDDEN>,
 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Mike Kupfer <kupfer@HIDDEN> writes:

> I'd also like to see some mention of changing the cursor shape.  I have
> not seen any glitches since switching to 'bar'.

I can confirm this: With the sample program I posted, if I add, after
(fundamental-mode), the form (setq cursor-type 'bar), then the issue
no longer appears whereas it otherwise appears consistently.

Thank you and all the best,
Markus




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 15 May 2026 21:20:15 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Fri May 15 17:20:15 2026
Received: from localhost ([127.0.0.1]:49706 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wNzx9-000557-GU
	for submit <at> debbugs.gnu.org; Fri, 15 May 2026 17:20:15 -0400
Received: from shell1.rawbw.com ([198.144.192.42]:44302 ident=root)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <kupfer@HIDDEN>) id 1wNzx6-00054L-Ng
 for 81012 <at> debbugs.gnu.org; Fri, 15 May 2026 17:20:13 -0400
Received: from bach.localdomain (135-180-175-182.dsl.dynamic.sonic.net
 [135.180.175.182] (may be forged)) (authenticated bits=0)
 by shell1.rawbw.com (8.15.1/8.15.1) with ESMTPSA id 64FLK408051709
 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NO);
 Fri, 15 May 2026 14:20:09 -0700 (PDT)
 (envelope-from kupfer@HIDDEN)
DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=rawbw.com; s=apr2026;
 t=1778880010; bh=PSiTdwBViondBgTP5d5AqqMh32pKVMqwUfidXZotErY=;
 h=From:To:cc:Subject:In-reply-to:References:Date;
 b=G8D6+GOI3ClAqyHcF1APloA1iQKG2BR3aZ5NICTTKH8swYaDtMlf3VwPW2vAkSDmm
 5eqaGotaLyDAFcAgxfRXhCyOFnlMqAZ2ciIYVjiAOKjOnGP0yApZRnq7tTcWaBry9K
 7uw1xoi56LgLhfyYNEBZbkYItUwPzLQn02gxvWZ0=
X-Authentication-Warning: shell1.rawbw.com: Host
 135-180-175-182.dsl.dynamic.sonic.net [135.180.175.182] (may be forged)
 claimed to be bach.localdomain
From: Mike Kupfer <kupfer@HIDDEN>
To: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
In-reply-to: <87bjegehgw.fsf@HIDDEN>
References: <20260510083506.293CBF08697@triska-ser8>
 <87wlxb9kuz.fsf@HIDDEN> <864ikf2hzo.fsf@HIDDEN>
 <878q9odcqa.fsf@HIDDEN> <86tssczt0e.fsf@HIDDEN>
 <87cxyzf1tn.fsf@HIDDEN> <87mry2ge1l.fsf@HIDDEN>
 <87jyt4erlq.fsf@HIDDEN> <874ik8911m.fsf@HIDDEN>
 <87bjegehgw.fsf@HIDDEN>
Comments: In-reply-to Pip Cet via "Bug reports for GNU Emacs,
 the Swiss army knife of text editors" <bug-gnu-emacs@HIDDEN>
 message dated "Fri, 15 May 2026 16:59:13 -0000."
X-Mailer: MH-E 8.6+git; nmh 1.8-RC2; Emacs 31.0.50
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <7208.1778880004.1@HIDDEN>
Date: Fri, 15 May 2026 14:20:04 -0700
Message-ID: <7209.1778880004@HIDDEN>
X-Spam-Score: 0.0 (/)
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, Markus Triska <triska@HIDDEN>,
 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.0 (-)

Pip Cet via "Bug reports for GNU Emacs, wrote:

> I'm not sure what we should write in etc/PROBLEMS, though. We should
> mention 'amdgpu', and I think we should recommend disabling double
> buffering, but if disabling GLX doesn't help, should we even mention it?
> Or font hinting?

I'd like to see font hinting mentioned.  Although I did see some
rendering glitches with slight hinting, the problem seemed less severe
than with medium hinting.  That is, incidents seemed less frequent, and
the mis-rendered characters didn't seem to stand out quite so much.

I'd also like to see some mention of changing the cursor shape.  I have
not seen any glitches since switching to 'bar'.

thanks,
mike




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 15 May 2026 20:38:05 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Fri May 15 16:38:05 2026
Received: from localhost ([127.0.0.1]:49231 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wNzIJ-0000Lq-BU
	for submit <at> debbugs.gnu.org; Fri, 15 May 2026 16:38:04 -0400
Received: from [78.47.144.35] (port=44062 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wNzIG-0000L6-HQ
 for 81012 <at> debbugs.gnu.org; Fri, 15 May 2026 16:38:01 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id 3DA479C762; Fri, 15 May 2026 22:37:59 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8>
 <87wlxb9kuz.fsf@HIDDEN> <864ikf2hzo.fsf@HIDDEN>
 <878q9odcqa.fsf@HIDDEN> <86tssczt0e.fsf@HIDDEN>
 <87cxyzf1tn.fsf@HIDDEN> <87mry2ge1l.fsf@HIDDEN>
 <87jyt4erlq.fsf@HIDDEN> <874ik8911m.fsf@HIDDEN>
 <87bjegehgw.fsf@HIDDEN>
Date: Fri, 15 May 2026 22:37:59 +0200
In-Reply-To: <87bjegehgw.fsf@HIDDEN> (Pip Cet's message of "Fri, 15
 May 2026 16:59:13 +0000")
Message-ID: <87wlx4o1co.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview: Pip Cet <pipcet@HIDDEN> writes: > I'm not sure what
 we should write in etc/PROBLEMS, though. We should > mention 'amdgpu', and
 I think we should recommend disabling double > buffering, but if disabling
 GLX doesn't help, should we ev [...] 
 Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Pip Cet <pipcet@HIDDEN> writes:

> I'm not sure what we should write in etc/PROBLEMS, though. We should
> mention 'amdgpu', and I think we should recommend disabling double
> buffering, but if disabling GLX doesn't help, should we even mention it?
> Or font hinting?

Personally, I would regard as extremely valuable to include in such
notes, if that is possible, a patch for Emacs that one could apply to
reliably stop the issue from occurring.

For instance, the "trapezoid" method which you mentioned earlier, is
there a way to apply it unconditionally and thus prevent the problem
once and for all and for all background colours, even though it may have
other drawbacks (such as maybe being computationally more expensive?).

Personally, I would greatly appreciate a pointer to such a patch, or
maybe even compilation option. As far as I can tell, the other options
mentioned so far do not reliably prevent the problem for all of us.

Thank you and all the best,
Markus




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 15 May 2026 16:59:33 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Fri May 15 12:59:32 2026
Received: from localhost ([127.0.0.1]:47596 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wNvsn-0001Xv-C0
	for submit <at> debbugs.gnu.org; Fri, 15 May 2026 12:59:32 -0400
Received: from mail-10629.protonmail.ch ([79.135.106.29]:39021)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <pipcet@HIDDEN>)
 id 1wNvsj-0001Wh-8L
 for 81012 <at> debbugs.gnu.org; Fri, 15 May 2026 12:59:26 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=protonmail.com;
 s=protonmail3; t=1778864358; x=1779123558;
 bh=hIHPG4vUe6x/NmvPYlLB2VQXSGLHhDRw5tR8I1svNx0=;
 h=Date:To:From:Cc:Subject:Message-ID:In-Reply-To:References:
 Feedback-ID:From:To:Cc:Date:Subject:Reply-To:Feedback-ID:
 Message-ID:BIMI-Selector;
 b=B9DWKJvYehiIbzUNEXWdbRMVXGdJHij0lOmxgsfgOjlIQYK0x3mZnZy+MaA46Pahe
 ztB6b2IM5UJO6GaNXkMFX1w3gGFFejw1w/8VG35WbwikAdN2NZz6X3CHQp6rpEdX7M
 AJesFfU9Q+IC8EQ2rW8/QqppJ6AbIdekCy8iOipt80Xem7Q4o90rqDdZboCBNykRbP
 QOtkTXoyvR6lIBmo0eVTRjV9gBXalXcFmFT77XMRA5ZNunmX3/Uqglx1OB+hXqOWQy
 KQUnHEQdhyOx4SRi1LwELSztMDRrE3MO/23KL3sY3MmdL2Gh6oM33leuRmbzIjk2rK
 wzbVpUHXydPHg==
Date: Fri, 15 May 2026 16:59:13 +0000
To: Markus Triska <triska@HIDDEN>
From: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
Message-ID: <87bjegehgw.fsf@HIDDEN>
In-Reply-To: <874ik8911m.fsf@HIDDEN>
References: <20260510083506.293CBF08697@triska-ser8>
 <87wlxb9kuz.fsf@HIDDEN> <864ikf2hzo.fsf@HIDDEN>
 <878q9odcqa.fsf@HIDDEN> <86tssczt0e.fsf@HIDDEN>
 <87cxyzf1tn.fsf@HIDDEN> <87mry2ge1l.fsf@HIDDEN>
 <87jyt4erlq.fsf@HIDDEN> <874ik8911m.fsf@HIDDEN>
Feedback-ID: 112775352:user:proton
X-Pm-Message-ID: ea8423dff1985c2c80473ec7a2e9b56e99b164a5
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: -0.7 (/)
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.7 (-)

"Markus Triska" <triska@HIDDEN> writes:

> Pip Cet <pipcet@HIDDEN> writes:
>
>> Can you try
>>
>> Section "Module"
>>     Disable "glx"
>> EndSection
>>
>> and/or check /var/log/Xorg.0.log (or similar) ?
>
> Yes of course. I paste here all lines from /var/log/Xorg.0.log that
> case-insensitively contain "glx":
>
> First, with the way I tried and posted before, these lines are:
>
>     [     8.079] (**) Extension "GLX" is disabled
>     ...
>     [     8.110] (II) LoadModule: "glx"
>     [     8.110] (II) Loading /usr/lib/xorg/modules/extensions/libglx.so
>     [     8.123] (II) Module glx: vendor=3D"X.Org Foundation"
>     [     8.123] =09compiled for 1.21.1.16, module version =3D 1.0.0
>     [     8.123] =09ABI class: X.Org Server Extension, version 10.0
>
> Second, with the way you show above, the lines are:
>
>     [  5626.689] (WW) "glx" will not be loaded unless you've specified it=
 to be loaded elsewhere.
>     [  5626.689] (II) "glx" will be loaded even though the default is to =
disable it.
>     ...
>     [  5626.771] (II) Initializing extension GLX
>
>
> With *both* ways to disable GLX, "glxinfo" says:
>
>     $ glxinfo
>     name of display: :0
>     Error: couldn't find RGB GLX visual or fbconfig
>
> With both* ways the issue still arises, using the sample form I posted.

Thanks for confirming! I've rerun the tests, and it still looks like I
am not seeing any damage with GLX disabled. Mysterious, but of course
that detail is extremely important to anyone looking for the actual
underlying driver/hardware bug.

I'm not sure what we should write in etc/PROBLEMS, though. We should
mention 'amdgpu', and I think we should recommend disabling double
buffering, but if disabling GLX doesn't help, should we even mention it?
Or font hinting?

Pip





Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 15 May 2026 14:53:48 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Fri May 15 10:53:47 2026
Received: from localhost ([127.0.0.1]:46262 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wNtv9-0005RZ-4y
	for submit <at> debbugs.gnu.org; Fri, 15 May 2026 10:53:47 -0400
Received: from [78.47.144.35] (port=37652 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wNtv4-0005Qr-DN
 for 81012 <at> debbugs.gnu.org; Fri, 15 May 2026 10:53:45 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id 2F4C89C7A5; Fri, 15 May 2026 16:53:41 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <864ikf2hzo.fsf@HIDDEN> <878q9odcqa.fsf@HIDDEN>
 <86tssczt0e.fsf@HIDDEN> <87cxyzf1tn.fsf@HIDDEN>
 <87mry2ge1l.fsf@HIDDEN> <87jyt4erlq.fsf@HIDDEN>
Date: Fri, 15 May 2026 16:53:41 +0200
In-Reply-To: <87jyt4erlq.fsf@HIDDEN> (Pip Cet's message of "Fri, 15
 May 2026 13:20:18 +0000")
Message-ID: <874ik8911m.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview:  Pip Cet <pipcet@HIDDEN> writes: > Can you try > >
 Section "Module" > Disable "glx" > EndSection > > and/or check
 /var/log/Xorg.0.log
 (or similar) ? Yes of course. I paste here all lines from /var/log/Xorg.0.log
 that case-insensitively contain "glx": 
 Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Pip Cet <pipcet@HIDDEN> writes:

> Can you try
>
> Section "Module"
>     Disable "glx"
> EndSection
>
> and/or check /var/log/Xorg.0.log (or similar) ?

Yes of course. I paste here all lines from /var/log/Xorg.0.log that
case-insensitively contain "glx":

First, with the way I tried and posted before, these lines are:

    [     8.079] (**) Extension "GLX" is disabled
    ...
    [     8.110] (II) LoadModule: "glx"
    [     8.110] (II) Loading /usr/lib/xorg/modules/extensions/libglx.so
    [     8.123] (II) Module glx: vendor="X.Org Foundation"
    [     8.123] 	compiled for 1.21.1.16, module version = 1.0.0
    [     8.123] 	ABI class: X.Org Server Extension, version 10.0

Second, with the way you show above, the lines are:

    [  5626.689] (WW) "glx" will not be loaded unless you've specified it to be loaded elsewhere.
    [  5626.689] (II) "glx" will be loaded even though the default is to disable it.
    ...
    [  5626.771] (II) Initializing extension GLX


With *both* ways to disable GLX, "glxinfo" says:

    $ glxinfo
    name of display: :0
    Error: couldn't find RGB GLX visual or fbconfig

With both* ways the issue still arises, using the sample form I posted.

Thank you and all the best,
Markus




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 15 May 2026 13:20:40 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Fri May 15 09:20:40 2026
Received: from localhost ([127.0.0.1]:43976 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wNsSy-0002Iv-3c
	for submit <at> debbugs.gnu.org; Fri, 15 May 2026 09:20:39 -0400
Received: from mail-10630.protonmail.ch ([79.135.106.30]:55889)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <pipcet@HIDDEN>)
 id 1wNsSs-0002C3-Qb
 for 81012 <at> debbugs.gnu.org; Fri, 15 May 2026 09:20:34 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=protonmail.com;
 s=protonmail3; t=1778851224; x=1779110424;
 bh=F1Dly/eQgPZeuph1cOu/dnscDNVacrF8NE4ykBaEwoQ=;
 h=Date:To:From:Cc:Subject:Message-ID:In-Reply-To:References:
 Feedback-ID:From:To:Cc:Date:Subject:Reply-To:Feedback-ID:
 Message-ID:BIMI-Selector;
 b=TAfuqxBfIKWoSGAOUfJQ/go2EOuCFH4Hu3qziwImVaHMu4te8wz3Zc1OQfgHN6FN0
 fELIXaz808UPgB9cCnk6CTuqDXYoe29O8NOOrbeUa438YUXZxcqkzu5IS5tA982tyS
 CZEaQI6DBBOqV7dEK2WogSdjZCkl77zsPAl+UZT1VMk9bi2DtbhpXdyZNNoXftAkew
 +caaalsXH40eVdXq42Nb0gBhrISM6iKiqtSL8B1fvg1SGnjcJFVr9hCTszAFwoqsIX
 5vhMsCtSeBnn7tfz2RNvMPgPd7ibUliux2ufJ8RPo0hETsZb2rUsPH8tx9pfdvBh8O
 LW8ISpiKz4v1Q==
Date: Fri, 15 May 2026 13:20:18 +0000
To: Markus Triska <triska@HIDDEN>
From: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
Message-ID: <87jyt4erlq.fsf@HIDDEN>
In-Reply-To: <87mry2ge1l.fsf@HIDDEN>
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <864ikf2hzo.fsf@HIDDEN> <878q9odcqa.fsf@HIDDEN>
 <86tssczt0e.fsf@HIDDEN> <87cxyzf1tn.fsf@HIDDEN>
 <87mry2ge1l.fsf@HIDDEN>
Feedback-ID: 112775352:user:proton
X-Pm-Message-ID: a6e8719c242bfba4c6b399388f19359cbdd9c7a3
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: -0.7 (/)
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.7 (-)

"Markus Triska" <triska@HIDDEN> writes:

> Pip Cet <pipcet@HIDDEN> writes:
>
>> 1. Disabling GLX (which must be done explicitly in the xorg.conf file;
>> GLX is implicitly enabled) avoids the issue. However, it produces many
>> warnings from other programs.
>
> I think I successfully disabled GLX by putting the following contents in
> the file /etc/X11/xorg.conf.d/disable-gpu.conf:
>
>     Section "Extensions"
>         Option "GLX" "Disable"
>     EndSection

Can you try

Section "Module"
    Disable "glx"
EndSection

and/or check /var/log/Xorg.0.log (or similar) ?

> Yet, *the issue still arises* also with GLX disabled! Also, I see no
> warnings from other programs I started.

Interesting.

>> There probably isn't going to be much interest in fixing a problem on
>> the "obsolete" X11 window system now everyone is expected (or forced) to
>
> If possible, please still consider filing your findings there. It would
> be great to have a reliable fix for what seems to be an extremely
> unfortunate problem in an application whose purpose is to display text.

I ultimately gave up on trying to spot the small differences on my
screen, and changed xftfont_draw to draw the background and glyph
several times, grab an XImage of each, calculate checksums, and compare
them. That's very slow, but at least it detects the error more reliably.

I also found another workaround that appears to work (on this system, at
least): if we fill the background using a (rectangular) trapezoid rather
than the rectangle API, no damage!

Here's the code to detect damage and to avoid it when the Xft backend is
in use. Note that the damage detection kind of assumes a black
background, because it only detects whether pixels are !=3D 0.

diff --git a/src/xftfont.c b/src/xftfont.c
index f15dbae1e7a..9522df250cc 100644
--- a/src/xftfont.c
+++ b/src/xftfont.c
@@ -514,6 +514,8 @@ xftfont_draw (struct glyph_string *s, int from, int to,=
 int x, int y,
   int len =3D to - from;
   int i;
=20
+  Lisp_Object shas =3D Qnil;
+ again:
   if (s->font =3D=3D face->font)
     xftface_info =3D (struct xftface_info *) face->extra;
   xftfont_get_colors (f, face, s->gc, xftface_info,
@@ -547,7 +549,6 @@ xftfont_draw (struct glyph_string *s, int from, int to,=
 int x, int y,
=20
 #if defined HAVE_XRENDER && (RENDER_MAJOR > 0 || (RENDER_MINOR >=3D 2))
       if (with_background
-=09  && FRAME_DISPLAY_INFO (s->f)->alpha_bits
 =09  && FRAME_CHECK_XR_VERSION (s->f, 0, 2))
 =09{
 =09  x_xr_ensure_picture (s->f);
@@ -564,13 +565,51 @@ xftfont_draw (struct glyph_string *s, int from, int t=
o, int x, int y,
 =09      else
 =09=09x_xr_reset_ext_clip (f);
 =09      x_xrender_color_from_gc_background (s->f, s->gc, &xc, s->hl !=3D =
DRAW_CURSOR);
+#if 1
+=09      Pixmap=09=09=09    pix;
+=09      XRenderPictureAttributes    pa;
+
+=09      pix =3D XCreatePixmap (FRAME_X_DISPLAY (s->f),
+=09=09=09=09   FRAME_X_WINDOW (s->f), 1, 1,
+=09=09=09=09   (unsigned) FRAME_X_PICTURE_FORMAT (s->f)->depth);
+
+=09      pa.repeat =3D True;
+=09      Picture pict =3D XRenderCreatePicture (FRAME_X_DISPLAY (s->f),
+=09=09=09=09=09=09   pix,
+=09=09=09=09=09=09   FRAME_X_PICTURE_FORMAT (s->f),
+=09=09=09=09=09=09   CPRepeat, &pa);
+=09      XFreePixmap (FRAME_X_DISPLAY (s->f), pix);
+=09      XRenderFillRectangle (FRAME_X_DISPLAY (s->f),
+=09=09=09=09    PictOpSrc,
+=09=09=09=09    pict,
+=09=09=09=09    &xc, 0, 0, 1, 1);
+=09      XTrapezoid trapezoid;
+=09      trapezoid.top =3D XDoubleToFixed (y - ascent);
+=09      trapezoid.bottom =3D XDoubleToFixed (y - ascent + height);
+=09      trapezoid.left.p1.x =3D XDoubleToFixed (x);
+=09      trapezoid.left.p1.y =3D XDoubleToFixed (y - ascent);
+=09      trapezoid.left.p2.x =3D XDoubleToFixed (x);
+=09      trapezoid.left.p2.y =3D XDoubleToFixed (y - ascent + height);
+=09      trapezoid.right.p1.x =3D XDoubleToFixed (x + s->width);
+=09      trapezoid.right.p1.y =3D XDoubleToFixed (y - ascent + height);
+=09      trapezoid.right.p2.x =3D XDoubleToFixed (x + s->width);
+=09      trapezoid.right.p2.y =3D XDoubleToFixed (y - ascent);
+
+=09      XRenderCompositeTrapezoids (FRAME_X_DISPLAY (s->f),
+=09=09=09=09=09  PictOpSrc,
+=09=09=09=09=09  pict,
+=09=09=09=09=09  FRAME_X_PICTURE (s->f),
+=09=09=09=09=09  FRAME_X_PICTURE_FORMAT (s->f),
+=09=09=09=09=09  0, 0,
+=09=09=09=09=09  &trapezoid, 1);
+=09      XRenderFreePicture (FRAME_X_DISPLAY (s->f), pict);
+#else
 =09      XRenderFillRectangle (FRAME_X_DISPLAY (s->f),
 =09=09=09=09    PictOpSrc, FRAME_X_PICTURE (s->f),
 =09=09=09=09    &xc, x, y - ascent, s->width, height);
+#endif
 =09      x_xr_reset_ext_clip (f);
 =09      x_mark_frame_dirty (s->f);
-
-=09      with_background =3D false;
 =09    }
 =09  else
 =09    XftDrawRect (xft_draw, &bg, x, y - ascent, s->width, height);
@@ -579,6 +618,7 @@ xftfont_draw (struct glyph_string *s, int from, int to,=
 int x, int y,
 #endif
 =09XftDrawRect (xft_draw, &bg, x, y - ascent, s->width, height);
     }
+  // XSync (FRAME_X_DISPLAY (f), False);
   code =3D alloca (sizeof (FT_UInt) * len);
   for (i =3D 0; i < len; i++)
     code[i] =3D s->char2b[from + i];
@@ -595,6 +635,62 @@ xftfont_draw (struct glyph_string *s, int from, int to=
, int x, int y,
      XftDraw structure.  */
   x_mark_frame_dirty (f);
=20
+  static Lisp_Object ht;
+  if (NILP (ht))
+    {
+      staticpro (&ht);
+      ht =3D CALLN (Fmake_hash_table, QCtest, Qequal);
+    }
+#if 1
+  if (with_background)
+    {
+      int height =3D FONT_HEIGHT (s->font), ascent =3D FONT_BASE (s->font)=
;
+      XImage *image =3D NULL;
+      x_ignore_errors_for_next_request (FRAME_DISPLAY_INFO (f), 0);
+      image =3D XGetImage (FRAME_X_DISPLAY (f), FRAME_X_RAW_DRAWABLE (f),
+=09=09=09 x, y - ascent, s->width, height, 0xffffff,
+=09=09=09 XYPixmap);
+      x_stop_ignoring_errors (FRAME_DISPLAY_INFO (f));
+
+      if (image)
+=09{
+=09  if (s->width < 60 && height < 60)
+=09    {
+=09      char pixels[height * s->width];
+=09      for (int y =3D 0; y < height; y++)
+=09=09{
+=09=09  for (int x =3D 0; x < s->width; x++)
+=09=09    {
+=09=09      pixels[y * s->width + x] =3D (image->f.get_pixel (image, x, y)=
 =3D=3D 0) ? '.' : 'X';
+=09=09    }
+=09=09}
+=09      Lisp_Object str =3D make_specified_string (pixels, height * s->wi=
dth, height * s->width, false);
+=09      Lisp_Object sha =3D Fmd5 (str, Qnil, Qnil, Qnil, Qnil);
+=09      shas =3D Fcons (sha, shas);
+=09      if (XFIXNUM (Flength (shas)) < 2)
+=09=09{
+=09=09  image->f.destroy_image (image);
+=09=09  goto again;
+=09=09}
+=09      if (NILP (Fequal (Fnth (make_fixnum (0), shas),
+=09=09=09=09Fnth (make_fixnum (1), shas))))
+=09=09{
+=09=09  fprintf (stderr, "Damage detected: checksums do not match!\n");
+=09=09  Fputhash (sha, str, ht);
+=09=09  for (int y =3D 0; y < height; y++)
+=09=09    {
+=09=09      for (int x =3D 0; x < s->width; x++)
+=09=09=09{
+=09=09=09  fprintf (stderr, "%c", (image->f.get_pixel (image, x, y) =3D=3D=
 0) ? '.' : 'X');
+=09=09=09}
+=09=09      fprintf (stderr, "\n");
+=09=09    }
+=09=09}
+=09    }
+=09  image->f.destroy_image (image);
+=09}
+    }
+#endif
   if (bg_allocated_p)
     XftColorFree (FRAME_X_DISPLAY (f),
 =09=09  FRAME_X_VISUAL (f),





Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 14 May 2026 16:18:51 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Thu May 14 12:18:51 2026
Received: from localhost ([127.0.0.1]:38237 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wNYlu-0006de-Oj
	for submit <at> debbugs.gnu.org; Thu, 14 May 2026 12:18:51 -0400
Received: from [78.47.144.35] (port=51560 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wNYls-0006cx-Af
 for 81012 <at> debbugs.gnu.org; Thu, 14 May 2026 12:18:49 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id C273E9C795; Thu, 14 May 2026 18:18:46 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <864ikf2hzo.fsf@HIDDEN> <878q9odcqa.fsf@HIDDEN>
 <86tssczt0e.fsf@HIDDEN> <87cxyzf1tn.fsf@HIDDEN>
Date: Thu, 14 May 2026 18:18:46 +0200
In-Reply-To: <87cxyzf1tn.fsf@HIDDEN> (Pip Cet's message of "Wed, 13
 May 2026 21:14:59 +0000")
Message-ID: <87mry2ge1l.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview:  Pip Cet <pipcet@HIDDEN> writes: > 1. Disabling GLX
 (which must be done explicitly in the xorg.conf file; > GLX is implicitly
 enabled) avoids the issue. However, it produces many > warnings from other
 programs. 
 Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Pip Cet <pipcet@HIDDEN> writes:

> 1. Disabling GLX (which must be done explicitly in the xorg.conf file;
> GLX is implicitly enabled) avoids the issue. However, it produces many
> warnings from other programs.

I think I successfully disabled GLX by putting the following contents in
the file /etc/X11/xorg.conf.d/disable-gpu.conf:

    Section "Extensions"
        Option "GLX" "Disable"
    EndSection

I think GLX is disabled because glxinfo now only says:

    $ glxinfo
    name of display: :0
    Error: couldn't find RGB GLX visual or fbconfig

whereas it previously showed many other things, starting with:

    name of display: :0
    display: :0  screen: 0
    direct rendering: Yes
    server glx vendor string: SGI
    server glx version string: 1.4
    ....

Yet, *the issue still arises* also with GLX disabled! Also, I see no
warnings from other programs I started.


> There probably isn't going to be much interest in fixing a problem on
> the "obsolete" X11 window system now everyone is expected (or forced) to

If possible, please still consider filing your findings there. It would
be great to have a reliable fix for what seems to be an extremely
unfortunate problem in an application whose purpose is to display text.

Thank you a lot!
Markus




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 14 May 2026 04:51:53 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Thu May 14 00:51:52 2026
Received: from localhost ([127.0.0.1]:33772 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wNO33-0007G3-Fr
	for submit <at> debbugs.gnu.org; Thu, 14 May 2026 00:51:52 -0400
Received: from eggs.gnu.org ([2001:470:142:3::10]:50854)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <eliz@HIDDEN>) id 1wNO2y-0007Ef-Eh
 for 81012 <at> debbugs.gnu.org; Thu, 14 May 2026 00:51:48 -0400
Received: from fencepost.gnu.org ([2001:470:142:3::e])
 by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.90_1) (envelope-from <eliz@HIDDEN>)
 id 1wNO2s-00068B-0W; Thu, 14 May 2026 00:51:38 -0400
DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=gnu.org;
 s=fencepost-gnu-org; h=References:Subject:In-Reply-To:To:From:Date:
 mime-version; bh=FU3ewVsTsrVJULJaLbnXFovtrwQU+zsxZGxT8ztxU4o=; b=hhcr04IkqSLV
 VcIk0vhHY4DTPCUhskCPHTmEMVyMlSO01LwVq5+JvPg73LwKXaSPWt991npSQyomUmpUhPB7H/gTc
 eDi4SzQY6b1rSkPZ9L44dGQkF83n29ngRxY0z3t35X131ehkOt4ev9bXs5sQzojltIi+ME5hGrQna
 M8XwIW8hGmsgeDVqKTO5/Lir2pCwumQkuT/53/Fz4Sx1AHq8HwZ2b86mDYgcGQ/i4s0uht2Uj4Nq0
 dGeAM9Tc6fTNkMIVgMk1IIc7W+rSmxx395EEC0o3gmV+6H73sMWSmCDWnY24plMR8alGStTmsgr9o
 dgSZ4cJWzbDsc8goMTLErw==;
Date: Thu, 14 May 2026 07:51:33 +0300
Message-Id: <86v7cq8uga.fsf@HIDDEN>
From: Eli Zaretskii <eliz@HIDDEN>
To: Pip Cet <pipcet@HIDDEN>
In-Reply-To: <87cxyzf1tn.fsf@HIDDEN> (message from Pip Cet on Wed, 13
 May 2026 21:14:59 +0000)
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <864ikf2hzo.fsf@HIDDEN> <878q9odcqa.fsf@HIDDEN>
 <86tssczt0e.fsf@HIDDEN> <87cxyzf1tn.fsf@HIDDEN>
X-Spam-Score: -2.3 (--)
X-Debbugs-Envelope-To: 81012
Cc: triska@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -3.3 (---)

> Date: Wed, 13 May 2026 21:14:59 +0000
> From: Pip Cet <pipcet@HIDDEN>
> Cc: triska@HIDDEN, 81012 <at> debbugs.gnu.org
> 
> "Eli Zaretskii" <eliz@HIDDEN> writes:
> 
> > This seems to contradict what Markus observed on his system:
> >
> >>                                                     When I move the
> >> cursor over the letters at the end of the video, each of them is drawn
> >> correctly the moment it is under the cursor.
> >
> > Isn't it strange that in your case it seems to be the other way
> > around?
> 
> I think Markus explained that we were seeing very similar symptoms

Yes, this bug report was really confusing in the description of the
symptoms.

> (though it gets stranger with non-box cursors: the bar cursor, in
> particular, damages glyphs it moves over, which the box cursor and the
> hollow cursor do not do. In fact, the hollow cursor doesn't exhibit the
> problem at all).

I think it depends on how Emacs draws each cursor type: normal block
cursor just redraws the character (with different colors), whereas the
other kinds draw the cursor using the GUI drawing primitives without
touching the character, except where the pixels where the cursor is
drawn intersect with the pixels of the character glyph.  In
particular, hollow cursor by definition draws outside of the character
cell.

> I have discovered a few more things that may help us find a good way to
> put this into etc/PROBLEMS, and to report the bug upstream (all of these
> were observed on my system only; I think it would help to verify them on
> other systems.):

Thanks.  These should be in PROBLEMS, but with enough details for
users to be able to take the workarounds.  For example:

> 1. Disabling GLX (which must be done explicitly in the xorg.conf file;
> GLX is implicitly enabled) avoids the issue. However, it produces many
> warnings from other programs.

This should tell how to disable GLX.  Likewise for other aspects.

> >> Depending on how common this is, we might need to add that to
> >> etc/PROBLEMS?
> >
> > Yes, definitely.  I think the hinting settings in .Xdefaults should
> > also be mentioned there, as one potential solution (although the
> > results seem to be somewhat ambiguous, IIUC).
> 
> Well, there already is an entry in etc/PROBLEMS that recommends
> disabling double buffering; I think it should be updated to indicate
> that the problem is still current, and that on modern systems,
> flickering due to a lack of double buffering is unlikely to be a
> problem.

Agreed.  Can you suggest a patch for PROBLEMS along these lines?




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 13 May 2026 21:15:16 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Wed May 13 17:15:15 2026
Received: from localhost ([127.0.0.1]:56624 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wNGvD-0004Bd-2B
	for submit <at> debbugs.gnu.org; Wed, 13 May 2026 17:15:15 -0400
Received: from mail-244116.protonmail.ch ([109.224.244.116]:31953)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <pipcet@HIDDEN>)
 id 1wNGv9-0004AJ-K9
 for 81012 <at> debbugs.gnu.org; Wed, 13 May 2026 17:15:13 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=protonmail.com;
 s=protonmail3; t=1778706904; x=1778966104;
 bh=5FW/Drk8kaI3ExeT5HM9QfyKaqpsVru3qwKV1AhKOhc=;
 h=Date:To:From:Cc:Subject:Message-ID:In-Reply-To:References:
 Feedback-ID:From:To:Cc:Date:Subject:Reply-To:Feedback-ID:
 Message-ID:BIMI-Selector;
 b=tXeR4O1V/h4AND1z6PsNMgqZH9qdRmqRGl9IiVcy5KhhO6QCCzVam/4slGHjAKFTG
 wl5/aeAjT6ck9myvciJGoaKqv2r7TsK/tVa306Alg3+qqvxdflqOrGf0s2aohZTROg
 osr5IcTQHrp4eShiczF2hmNh4IAXrMKXYqCcsB1t+JsirgF8iIELOVqfJY7F530mdC
 ZeOJ5fs07KjVbxHeU177W3q56zvdGERJjS+luqUFfhWlXstFMuq/ctV5BDfJSyobFH
 3BboDq2IIweFaJqGVT0v5KgSxKqKxs3Qc8xdWt+56MqgLJw7QHid5sbOJEQGNnLIxZ
 6zM9gZbXjzdWQ==
Date: Wed, 13 May 2026 21:14:59 +0000
To: Eli Zaretskii <eliz@HIDDEN>
From: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
Message-ID: <87cxyzf1tn.fsf@HIDDEN>
In-Reply-To: <86tssczt0e.fsf@HIDDEN>
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <864ikf2hzo.fsf@HIDDEN> <878q9odcqa.fsf@HIDDEN>
 <86tssczt0e.fsf@HIDDEN>
Feedback-ID: 112775352:user:proton
X-Pm-Message-ID: a272115c8aaeb4b1ffe66a90fab9596779920df9
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: -0.7 (/)
X-Debbugs-Envelope-To: 81012
Cc: triska@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.7 (-)

"Eli Zaretskii" <eliz@HIDDEN> writes:

>> Date: Tue, 12 May 2026 12:38:51 +0000
>> From: Pip Cet <pipcet@HIDDEN>
>> Cc: triska@HIDDEN, 81012 <at> debbugs.gnu.org
>>
>> "Eli Zaretskii" <eliz@HIDDEN> writes:
>>
>> > Did you try to disable double-buffering?
>>
>> That does indeed avoid the issue, thank you!
>
> Thanks for trying it.
>
>> IIUC, that means it could still be a driver issue (just one exclusive to
>> off-screen buffers created by XDBE).
>>
>> I've investigated some more, but I haven't found out anything new or
>> exciting. The issue seems to be specific to glyphs created "under" the
>> cursor, disabling the cursor makes it go away, and so does an XSync
>> between XftDrawRect/XRenderFillRectangle and XftDrawGlyphs.
>
> This seems to contradict what Markus observed on his system:
>
>>                                                     When I move the
>> cursor over the letters at the end of the video, each of them is drawn
>> correctly the moment it is under the cursor.
>
> Isn't it strange that in your case it seems to be the other way
> around?

I think Markus explained that we were seeing very similar symptoms
(though it gets stranger with non-box cursors: the bar cursor, in
particular, damages glyphs it moves over, which the box cursor and the
hollow cursor do not do. In fact, the hollow cursor doesn't exhibit the
problem at all).

I have discovered a few more things that may help us find a good way to
put this into etc/PROBLEMS, and to report the bug upstream (all of these
were observed on my system only; I think it would help to verify them on
other systems.):

1. Disabling GLX (which must be done explicitly in the xorg.conf file;
GLX is implicitly enabled) avoids the issue. However, it produces many
warnings from other programs.

2. The issue still appears with the current freedesktop.org git versions
of xserver and xf86-driver-amdgpu.

3. A window manager appears to be required for the problem to
appear. Maybe not running one disables double buffering somehow?

4. Modifying libXft so it uses a single-pixel pixmap and
XRenderCreatePicture rather than XRenderCreateSolidFill avoids the
issue.

5. inhibit-double-buffering doesn't fully avoid the issue. I just saw a
dotless "i" that was supposed to have a dot. Running some more tests
with different glyphs shows reduced, but still noticeable problems even
with inhibit-double-buffering.

There probably isn't going to be much interest in fixing a problem on
the "obsolete" X11 window system now everyone is expected (or forced) to
switch to the (vastly inferior, IMHO) Wayland system, though.

>> (set-frame-parameter nil 'inhibit-double-buffering nil) for me, I think.

>> Depending on how common this is, we might need to add that to
>> etc/PROBLEMS?
>
> Yes, definitely.  I think the hinting settings in .Xdefaults should
> also be mentioned there, as one potential solution (although the
> results seem to be somewhat ambiguous, IIUC).

Well, there already is an entry in etc/PROBLEMS that recommends
disabling double buffering; I think it should be updated to indicate
that the problem is still current, and that on modern systems,
flickering due to a lack of double buffering is unlikely to be a
problem.

Pip





Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 12 May 2026 13:58:22 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Tue May 12 09:58:22 2026
Received: from localhost ([127.0.0.1]:35612 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wMncr-0005KJ-Sa
	for submit <at> debbugs.gnu.org; Tue, 12 May 2026 09:58:22 -0400
Received: from [78.47.144.35] (port=48070 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wMncp-0005Jx-1Z
 for 81012 <at> debbugs.gnu.org; Tue, 12 May 2026 09:58:20 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id 533359C795; Tue, 12 May 2026 15:58:17 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <87o6inp0me.fsf@HIDDEN> <87o6in9h6q.fsf@HIDDEN>
Date: Tue, 12 May 2026 15:58:17 +0200
In-Reply-To: <87o6in9h6q.fsf@HIDDEN> (Pip Cet's message of "Sun, 10
 May 2026 19:51:35 +0000")
Message-ID: <87pl30afwm.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview:  Pip Cet <pipcet@HIDDEN> writes: >> Xft.hintstyle:
 hintnone > > I'm seeing less damage with that option, but still some of it.
 Yes,
 I can confirm this now: Disabling hinting does not solve the issue completely
 also for me, in contrast to what I initially thought. In the test I posted,
 it is clear that characters still lack pi [...] 
 Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Pip Cet <pipcet@HIDDEN> writes:

>>     Xft.hintstyle: hintnone
>
> I'm seeing less damage with that option, but still some of it.

Yes, I can confirm this now: Disabling hinting does not solve the issue
completely also for me, in contrast to what I initially thought.  In the
test I posted, it is clear that characters still lack pixels especially
when a line of "t"s is finished, and the line is redrawn when the line
break occurs.

Thank you a lot for looking into this so carefully!

All the best,
Markus




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 12 May 2026 13:27:16 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Tue May 12 09:27:15 2026
Received: from localhost ([127.0.0.1]:35337 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wMn8k-0001v1-4s
	for submit <at> debbugs.gnu.org; Tue, 12 May 2026 09:27:15 -0400
Received: from [78.47.144.35] (port=47826 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wMn8e-0001uK-V6
 for 81012 <at> debbugs.gnu.org; Tue, 12 May 2026 09:27:11 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id 49C299C79D; Tue, 12 May 2026 15:27:07 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Eli Zaretskii <eliz@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <864ikf2hzo.fsf@HIDDEN> <878q9odcqa.fsf@HIDDEN>
 <86tssczt0e.fsf@HIDDEN>
Date: Tue, 12 May 2026 15:27:07 +0200
In-Reply-To: <86tssczt0e.fsf@HIDDEN> (Eli Zaretskii's message of "Tue, 12 May
 2026 15:56:01 +0300")
Message-ID: <87lddog3mc.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview: Eli Zaretskii <eliz@HIDDEN> writes: > This seems to
 contradict
 what Markus observed on his system: I think Pip and I are describing the
 same phenomenon: The glyph is wrongly created under the cursor. Later, when
 the cursor moves over it, it is drawn correctly, though as Pip noticed not
 necessarily [...] 
 Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: Pip Cet <pipcet@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Eli Zaretskii <eliz@HIDDEN> writes:

> This seems to contradict what Markus observed on his system:

I think Pip and I are describing the same phenomenon: The glyph is
wrongly created under the cursor. Later, when the cursor moves over it,
it is drawn correctly, though as Pip noticed not necessarily instantly!

All the best,
Markus




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 12 May 2026 12:56:21 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Tue May 12 08:56:21 2026
Received: from localhost ([127.0.0.1]:35132 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wMmeq-0006GQ-Qh
	for submit <at> debbugs.gnu.org; Tue, 12 May 2026 08:56:21 -0400
Received: from eggs.gnu.org ([2001:470:142:3::10]:39790)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <eliz@HIDDEN>) id 1wMmem-0006FA-8E
 for 81012 <at> debbugs.gnu.org; Tue, 12 May 2026 08:56:19 -0400
Received: from fencepost.gnu.org ([2001:470:142:3::e])
 by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.90_1) (envelope-from <eliz@HIDDEN>)
 id 1wMmef-0001Px-Qh; Tue, 12 May 2026 08:56:10 -0400
DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=gnu.org;
 s=fencepost-gnu-org; h=References:Subject:In-Reply-To:To:From:Date:
 mime-version; bh=LjHPEuRyKZTkgsW8q9OcXaeCcfj+0onUpvLM5tl8qvA=; b=d9lFHpiOA2/K
 sTygLdC+K7I5AafBTLLTOyRlvPDDJHBxrzEgXCJpj0cpeXvuBTlKlIsnckdNxNqq+T/Mzpl/hQ1mN
 KQrFy2XOLwJGFPM/bH6h4B4A//CZhTmOvO5SFXrcieyBU02B62TiQIpFn2Z3arJaUJB39Tj5nvTSB
 CiLOmb2Iih2bYuqV25Wzb5KOi3UHcZHZxa6Fyuv8B6+TY4aBPQWFXbsk8lqExZeTv1ONvPdoehc3g
 vNogQmm6SeP7e7hZKUyeiBQRuOW5Fe2wLqXLjPExreecZjH4Ay4mMW0oYzFEyWSc4/AA9B0GVtLod
 KRS4wX/2Wj0YOxfIRZMLKA==;
Date: Tue, 12 May 2026 15:56:01 +0300
Message-Id: <86tssczt0e.fsf@HIDDEN>
From: Eli Zaretskii <eliz@HIDDEN>
To: Pip Cet <pipcet@HIDDEN>
In-Reply-To: <878q9odcqa.fsf@HIDDEN> (message from Pip Cet on Tue, 12
 May 2026 12:38:51 +0000)
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <864ikf2hzo.fsf@HIDDEN> <878q9odcqa.fsf@HIDDEN>
X-Spam-Score: -2.3 (--)
X-Debbugs-Envelope-To: 81012
Cc: triska@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -3.3 (---)

> Date: Tue, 12 May 2026 12:38:51 +0000
> From: Pip Cet <pipcet@HIDDEN>
> Cc: triska@HIDDEN, 81012 <at> debbugs.gnu.org
> 
> "Eli Zaretskii" <eliz@HIDDEN> writes:
> 
> > Did you try to disable double-buffering?
> 
> That does indeed avoid the issue, thank you!

Thanks for trying it.

> IIUC, that means it could still be a driver issue (just one exclusive to
> off-screen buffers created by XDBE).
> 
> I've investigated some more, but I haven't found out anything new or
> exciting. The issue seems to be specific to glyphs created "under" the
> cursor, disabling the cursor makes it go away, and so does an XSync
> between XftDrawRect/XRenderFillRectangle and XftDrawGlyphs.

This seems to contradict what Markus observed on his system:

>                                                     When I move the
> cursor over the letters at the end of the video, each of them is drawn
> correctly the moment it is under the cursor.

Isn't it strange that in your case it seems to be the other way
around?

> (set-frame-parameter nil 'inhibit-double-buffering nil) for me, I think.
> 
> Depending on how common this is, we might need to add that to
> etc/PROBLEMS?

Yes, definitely.  I think the hinting settings in .Xdefaults should
also be mentioned there, as one potential solution (although the
results seem to be somewhat ambiguous, IIUC).




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 12 May 2026 12:39:05 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Tue May 12 08:39:05 2026
Received: from localhost ([127.0.0.1]:35012 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wMmO8-0004AX-VG
	for submit <at> debbugs.gnu.org; Tue, 12 May 2026 08:39:05 -0400
Received: from mail-106118.protonmail.ch ([79.135.106.118]:35959)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <pipcet@HIDDEN>)
 id 1wMmO6-00049e-Be
 for 81012 <at> debbugs.gnu.org; Tue, 12 May 2026 08:39:03 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=protonmail.com;
 s=protonmail3; t=1778589535; x=1778848735;
 bh=XQe5gWg8bYFrCA/kR2sKedVVL2lOP68Zp6LRfMkQ5ek=;
 h=Date:To:From:Cc:Subject:Message-ID:In-Reply-To:References:
 Feedback-ID:From:To:Cc:Date:Subject:Reply-To:Feedback-ID:
 Message-ID:BIMI-Selector;
 b=fdrBmf2WpGQxcaVyURWbDZRdLi1nIBZ11q0Jsp39XZgZtirG+7MLBJZY0r9nK541v
 Ch8XtKtx2HHseea16q5qX6hK0z/1am885efgDAzdhoHFk09GGpHsYskXc7eKb49F29
 yY0siIuSSXa+Ysfr0zIn5cYt+RRCfFj+oDIqaqLZUkM/dzCuebzHr88CxrnbLqyqbF
 oKcpiLvpXfAWTmiqFzi+Kp+ImPDt1stf3G+Y2zYvtwk3u3oWN/Y2A0QbaaOih3lmag
 Ap+S7JIJRXozyIOs2cMxxtwtEVHhuq2iBAevhCaNz8eX0FReenofcI213POFYcNOeS
 E+excpHKyaG0w==
Date: Tue, 12 May 2026 12:38:51 +0000
To: Eli Zaretskii <eliz@HIDDEN>
From: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
Message-ID: <878q9odcqa.fsf@HIDDEN>
In-Reply-To: <864ikf2hzo.fsf@HIDDEN>
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <864ikf2hzo.fsf@HIDDEN>
Feedback-ID: 112775352:user:proton
X-Pm-Message-ID: 081ebaf606df60f228fec6d768da7df751598e49
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: -0.7 (/)
X-Debbugs-Envelope-To: 81012
Cc: triska@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.7 (-)

"Eli Zaretskii" <eliz@HIDDEN> writes:

>> Date: Sun, 10 May 2026 18:32:17 +0000
>> From: Pip Cet <pipcet@HIDDEN>
>> Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
>>
>> This is definitely not a fix, but could you try applying this patch and
>> seeing whether the problems still occur (it's specific to the font
>> driver you're using, and a different patch is required for ftcrfont.c) ?
>>
>> diff --git a/src/xftfont.c b/src/xftfont.c
>> index f15dbae1e7a..6b21ec72755 100644
>> --- a/src/xftfont.c
>> +++ b/src/xftfont.c
>> @@ -579,6 +579,8 @@ xftfont_draw (struct glyph_string *s, int from, int =
to, int x, int y,
>>  #endif
>>  =09XftDrawRect (xft_draw, &bg, x, y - ascent, s->width, height);
>>      }
>> +  struct x_display_info *dpyinfo =3D FRAME_DISPLAY_INFO (s->f);
>> +  XSync (dpyinfo->display, false);
>>    code =3D alloca (sizeof (FT_UInt) * len);
>>    for (i =3D 0; i < len; i++)
>>      code[i] =3D s->char2b[from + i];
>>
>> It seems to avoid the issue here, but my eyes have gone kind of blurry
>> at this point so maybe it's just that :-)
>>
>> For me, the issue appears with both ftcrfont and xft, but goes away when
>> I use Xnest.
>
> Did you try to disable double-buffering?

That does indeed avoid the issue, thank you!

IIUC, that means it could still be a driver issue (just one exclusive to
off-screen buffers created by XDBE).

I've investigated some more, but I haven't found out anything new or
exciting. The issue seems to be specific to glyphs created "under" the
cursor, disabling the cursor makes it go away, and so does an XSync
between XftDrawRect/XRenderFillRectangle and XftDrawGlyphs.

(set-frame-parameter nil 'inhibit-double-buffering nil) for me, I think.

Depending on how common this is, we might need to add that to
etc/PROBLEMS?

Thanks again
Pip





Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 12 May 2026 12:30:47 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Tue May 12 08:30:47 2026
Received: from localhost ([127.0.0.1]:34952 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wMmG7-0003K0-Dm
	for submit <at> debbugs.gnu.org; Tue, 12 May 2026 08:30:47 -0400
Received: from [78.47.144.35] (port=47012 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wMmG5-0003Jd-ML
 for 81012 <at> debbugs.gnu.org; Tue, 12 May 2026 08:30:46 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id 42BA29C7A5; Tue, 12 May 2026 14:30:44 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Eli Zaretskii <eliz@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <87o6inp0me.fsf@HIDDEN> <87o6in9h6q.fsf@HIDDEN>
 <87tssfnets.fsf@HIDDEN> <86wlxa119r.fsf@HIDDEN>
 <87y0hp4vfv.fsf@HIDDEN> <868q9o28fc.fsf@HIDDEN>
Date: Tue, 12 May 2026 14:30:44 +0200
In-Reply-To: <868q9o28fc.fsf@HIDDEN> (Eli Zaretskii's message of "Tue, 12 May
 2026 14:07:03 +0300")
Message-ID: <87lddo95e3.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview: Eli Zaretskii <eliz@HIDDEN> writes: > Given that
 ~/.Xdefaults
 is ignored in "emacs -Q", why is this a > surprise? From everything I can
 tell, at least some of the specific parameters I have in ~/.Xdefaults are
 taken into account also when I start Emacs with "-Q". In particular, the
 configured hinting style is "hi [...] 
 Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: pipcet@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Eli Zaretskii <eliz@HIDDEN> writes:

> Given that ~/.Xdefaults is ignored in "emacs -Q", why is this a
> surprise?

From everything I can tell, at least some of the specific parameters I
have in ~/.Xdefaults are taken into account also when I start Emacs with
"-Q". In particular, the configured hinting style is "hintslight",
whether I launch Emacs with "-Q" or without it.

When I do not specify any hinting style in ~/.Xresources, the used
hinting style (both with and without the "-Q" switch) is, as far as I
can tell from comparing screenshots, equivalent to when I configure the
style to be "hintfull". This is in itself a surprise, because I expect
the default hinting style in fact to be "hintslight" based on the XFT
configuration in Debian. I filed this as an FVWM3 issue:

    https://github.com/fvwmorg/fvwm3/issues/1313

To summarize: The suprise is that using "-Q" does not affect the hinting
Emacs uses, but it affects the frequency of the present issue. This may
indicate that a different factor other than hinting also plays a role.

All the best,
Markus




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 12 May 2026 11:07:55 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Tue May 12 07:07:55 2026
Received: from localhost ([127.0.0.1]:34508 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wMkxv-00005r-1a
	for submit <at> debbugs.gnu.org; Tue, 12 May 2026 07:07:55 -0400
Received: from eggs.gnu.org ([2001:470:142:3::10]:38536)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <eliz@HIDDEN>) id 1wMkxs-000051-L8
 for 81012 <at> debbugs.gnu.org; Tue, 12 May 2026 07:07:53 -0400
Received: from fencepost.gnu.org ([2001:470:142:3::e])
 by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.90_1) (envelope-from <eliz@HIDDEN>)
 id 1wMkxl-0007h3-Mu; Tue, 12 May 2026 07:07:45 -0400
DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=gnu.org;
 s=fencepost-gnu-org; h=References:Subject:In-Reply-To:To:From:Date:
 mime-version; bh=hO7IH9cNmqC09kPV2U0jZz7tLO98gJ8stXFZkZCDf4s=; b=bgCPpTy/LOpn
 LR3A1QIMj8BEiNaw+3yqfbYbldh9GlPRTdTZ/wMLRYFaqMmNIicxxp0qg/VfUm8bzzqHl+Pa+cf2L
 QtEgDTkyCleZTmc6617gw5WBIdtXHfTzNLObPd5pxKLda7TVgg1HTBX6jrsLfkdQ306LS8TNsZINy
 Wt6E6NZ9ExqY8/BN/M4zIVYwfKGtBQo670yXc6IDddZuxGvMBSxg2NhcBvoHCLiq3XZdr+hPD28ka
 Ey+P1OJdvHU7e4rhiF8Took5tyS+2est3F2C1xPYUmtRtdXcaaX0BCrYG0fYg3VW2IGqTiohePSru
 iyp1fwFJS9m0gCzN4Qm05A==;
Date: Tue, 12 May 2026 14:07:03 +0300
Message-Id: <868q9o28fc.fsf@HIDDEN>
From: Eli Zaretskii <eliz@HIDDEN>
To: Markus Triska <triska@HIDDEN>
In-Reply-To: <87y0hp4vfv.fsf@HIDDEN> (message from Markus Triska on Mon, 
 11 May 2026 21:07:00 +0200)
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <87o6inp0me.fsf@HIDDEN> <87o6in9h6q.fsf@HIDDEN>
 <87tssfnets.fsf@HIDDEN> <86wlxa119r.fsf@HIDDEN>
 <87y0hp4vfv.fsf@HIDDEN>
X-Spam-Score: -2.3 (--)
X-Debbugs-Envelope-To: 81012
Cc: pipcet@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -3.3 (---)

> From: Markus Triska <triska@HIDDEN>
> Cc: pipcet@HIDDEN,  81012 <at> debbugs.gnu.org
> Date: Mon, 11 May 2026 21:07:00 +0200
> 
> Eli Zaretskii <eliz@HIDDEN> writes:
> 
> > I think this explains what you saw, doesn't it?
> 
> The suprise is that all the parameters I set in ~/.Xresources have the
> same effect on how Emacs appears (notably the hinting style) in all
> cases, in other words: Whether Emacs is started with "-Q" or without it
> does not change the appearance of the frame or the text in it, yet I had
> seen the issue I reported only when I launched Emacs with "-Q", and in
> that case it consistently occurs within a few "t"s.

"emacs -Q" makes Emacs ignore ~/.Xdefaults, so the parameters therein
are not in effect.  Doesn't this resolve the surprise you describe
above?

> Phrased a bit differently, the surprise is that the only discernible
> effect of the "-Q" switch on Emacs is the frequency of issue #81012.

Given that ~/.Xdefaults is ignored in "emacs -Q", why is this a
surprise?




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 11 May 2026 19:07:05 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Mon May 11 15:07:05 2026
Received: from localhost ([127.0.0.1]:52451 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wMVy4-0006Af-SX
	for submit <at> debbugs.gnu.org; Mon, 11 May 2026 15:07:05 -0400
Received: from [78.47.144.35] (port=36670 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wMVy1-0006AG-Mp
 for 81012 <at> debbugs.gnu.org; Mon, 11 May 2026 15:07:03 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id 342CE9C79A; Mon, 11 May 2026 21:07:00 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Eli Zaretskii <eliz@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <87o6inp0me.fsf@HIDDEN> <87o6in9h6q.fsf@HIDDEN>
 <87tssfnets.fsf@HIDDEN> <86wlxa119r.fsf@HIDDEN>
Date: Mon, 11 May 2026 21:07:00 +0200
In-Reply-To: <86wlxa119r.fsf@HIDDEN> (Eli Zaretskii's message of "Mon, 11 May
 2026 17:14:40 +0300")
Message-ID: <87y0hp4vfv.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview: Eli Zaretskii <eliz@HIDDEN> writes: > I think this explains
 what you saw, doesn't it? The suprise is that all the parameters I set in
 ~/.Xresources have the same effect on how Emacs appears (notably the hinting
 style) in all cases, in other words: Whether Emacs is started with "-Q" or
 [...] Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: pipcet@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Eli Zaretskii <eliz@HIDDEN> writes:

> I think this explains what you saw, doesn't it?

The suprise is that all the parameters I set in ~/.Xresources have the
same effect on how Emacs appears (notably the hinting style) in all
cases, in other words: Whether Emacs is started with "-Q" or without it
does not change the appearance of the frame or the text in it, yet I had
seen the issue I reported only when I launched Emacs with "-Q", and in
that case it consistently occurs within a few "t"s.

For comparison, when I launch Emacs without "-Q", i.e., just as "emacs",
then it appears in exactly the same way (the XFT-specific settings seem
to take effect no matter how I launch Emacs), yet the issue I reported
tends to arise, if at all, then much more rarely: All "t"s are drawn
exactly as expected, with one single exception so far: I just now had
one test run where even though I started Emacs as "emacs", one of the
"t"s was not fully drawn. That was the only exception so far, and I did
not manage to repeat the issue despite trying, while in contrast I can
consistently elicit it when invoking Emacs with "-Q".

Phrased a bit differently, the surprise is that the only discernible
effect of the "-Q" switch on Emacs is the frequency of issue #81012.

I tested everything so far in the FVWM3 window manager.




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 11 May 2026 14:14:59 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Mon May 11 10:14:58 2026
Received: from localhost ([127.0.0.1]:49974 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wMRPL-000740-B6
	for submit <at> debbugs.gnu.org; Mon, 11 May 2026 10:14:58 -0400
Received: from eggs.gnu.org ([2001:470:142:3::10]:47060)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <eliz@HIDDEN>) id 1wMRPH-00072d-EL
 for 81012 <at> debbugs.gnu.org; Mon, 11 May 2026 10:14:52 -0400
Received: from fencepost.gnu.org ([2001:470:142:3::e])
 by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.90_1) (envelope-from <eliz@HIDDEN>)
 id 1wMRPB-00013y-B7; Mon, 11 May 2026 10:14:45 -0400
DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=gnu.org;
 s=fencepost-gnu-org; h=MIME-version:References:Subject:In-Reply-To:To:From:
 Date; bh=bN6v34iktixjy/JH2u9K0sZ4GLqS6SXiR7xbXnFqczE=; b=HxbZ8983piiKDSec3FP1
 30OoRbMzAnkK+f5cslEloBOj7xhOnPljvF/epyv1mW0uw+bkfeYG62IwihpiMHyaAIDwdB/xhrtEr
 em68fy0VBEaC2k2Vq313voX5875z0gDZ04f+sNHKv/c0RTBue44LkgCkMoA59MN5cGK6D/tB/gigC
 dqydVXZ0/jwjFVT0e77klmmN96CR7QgblNhbyCj60U0H+Z6aoupIFxvc2Zz42ca8nAYAbTVfYbZHF
 vGqiXnmdubNQauDh7GRE/FaIJIza+wBFqkAIcbbiLWwxSv2fIDZeywI5YT8wF97+Nij7VE1RamlOt
 Ld0iaNndmScV1g==;
Date: Mon, 11 May 2026 17:14:40 +0300
Message-Id: <86wlxa119r.fsf@HIDDEN>
From: Eli Zaretskii <eliz@HIDDEN>
To: Markus Triska <triska@HIDDEN>
In-Reply-To: <87tssfnets.fsf@HIDDEN> (message from Markus Triska on Sun, 
 10 May 2026 23:18:39 +0200)
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <87o6inp0me.fsf@HIDDEN> <87o6in9h6q.fsf@HIDDEN>
 <87tssfnets.fsf@HIDDEN>
MIME-version: 1.0
Content-type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
X-Spam-Score: -2.3 (--)
X-Debbugs-Envelope-To: 81012
Cc: pipcet@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -3.3 (---)

> From: Markus Triska <triska@HIDDEN>
> Cc: Eli Zaretskii <eliz@HIDDEN>,  81012 <at> debbugs.gnu.org
> Date: Sun, 10 May 2026 23:18:39 +0200
> 
> In fact, there was also another aspect to this issue that surprised me:
> 
> The contents of my ~/.Xresources file when I reported this issue were:
> 
>     *FontBackend:	xft
>     Xft.dpi:	96
>     Xft.hinting:	true
>     Xft.hintstyle: hintslight
> 
> So, I had hintstyle set to hintslight. I deliberately had commented out
> *everything* in ~/.emacs.
> 
> Still, I observed the issue *only* when launching with "emacs -Q". I did
> not observe the issue when launching "emacs" without any arguments.
> 
> I found this surprising because from what I can tell, "hintslight" was
> enabled both when using "-Q" and not using it, and the options in
> ~/.Xresources seem not specific to Emacs.
> 
> Maybe it could be worth considering to make the settings better
> inspectable from within Elisp? Trying to detect from within Emacs
> whether it uses hinting, I found "hinting" not to occur in the manual.

From the Emacs user manual:

     Emacs does not process X resources at all if you set the variable
  ‘inhibit-x-resources’ to a non-‘nil’ value.  If you invoke Emacs with
  the ‘-Q’ (or ‘--quick’) command-line option, ‘inhibit-x-resources’ is
  automatically set to ‘t’ (*note Initial Options::).

I think this explains what you saw, doesn't it?




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 10 May 2026 22:38:40 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sun May 10 18:38:40 2026
Received: from localhost ([127.0.0.1]:41853 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wMCnH-0006KZ-NU
	for submit <at> debbugs.gnu.org; Sun, 10 May 2026 18:38:40 -0400
Received: from shell1.rawbw.com ([198.144.192.42]:55418 ident=root)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <kupfer@HIDDEN>) id 1wMCnD-0006Jb-Pd
 for 81012 <at> debbugs.gnu.org; Sun, 10 May 2026 18:38:37 -0400
Received: from bach.localdomain (135-180-175-182.dsl.dynamic.sonic.net
 [135.180.175.182] (may be forged)) (authenticated bits=0)
 by shell1.rawbw.com (8.15.1/8.15.1) with ESMTPSA id 64AMcRNR040609
 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NO);
 Sun, 10 May 2026 15:38:33 -0700 (PDT)
 (envelope-from kupfer@HIDDEN)
DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=rawbw.com; s=apr2026;
 t=1778452714; bh=Mp4jd5uwwpksPRfXSLCB4RbiFEvm1J6IAyPASt2YH+s=;
 h=From:To:cc:Subject:In-reply-to:References:Date;
 b=f58d5aBUfHH8HevnBhKcjqpeL8zG3GQ1W+M1eKs/LrlsZrk5YKPrQVRk4mPQkYX5G
 eph4UWOzdD/AZeYfR76aaMJ6gBR6t5rUCcIZUHuXl5o0d0ASeC4lQz9gcyCmGeSXxc
 9jncosB8KlvsbHB44XDDmDnaHmo6OaOQZjVcziZQ=
X-Authentication-Warning: shell1.rawbw.com: Host
 135-180-175-182.dsl.dynamic.sonic.net [135.180.175.182] (may be forged)
 claimed to be bach.localdomain
From: Mike Kupfer <kupfer@HIDDEN>
To: Markus Triska <triska@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
In-reply-to: <87se7z1kp6.fsf@HIDDEN>
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN>
Comments: In-reply-to Markus Triska <triska@HIDDEN>
 message dated "Sun, 10 May 2026 15:02:45 +0200."
X-Mailer: MH-E 8.6+git; nmh 1.8-RC2; Emacs 31.0.50
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="=-=-="
Date: Sun, 10 May 2026 15:38:27 -0700
Message-ID: <23428.1778452707@HIDDEN>
X-Spam-Score: 0.0 (/)
X-Debbugs-Envelope-To: 81012
Cc: Pip Cet <pipcet@HIDDEN>, Eli Zaretskii <eliz@HIDDEN>,
 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.0 (-)

--=-=-=
Content-Type: text/plain

Markus Triska wrote:

> Can anyone else reproduce the problem? I would greatly appreciate any
> feedback and help.

I'm probably seeing the same problem.  I don't see the problem using
your reproducer, but that could be due to running MATE at the moment
(more on that below).  

I have been seeing similar glitches with text rendering with i3.  For
example, in the attached screenshot, the second "o" in "soon" is missing
some "ink".  I've been collecting data in preparation for opening a bug.

Like you, I'm using AMD graphics.  I'm using Debian 12 (oldstable) and
Emacs master, but I'm pretty sure I also saw it with Emacs 30.  C-l
doesn't fix the problem, sometimes the problem goes away on its own, and
the problem always goes away if I highlight the region (e.g., C-SPC
C-a).

Some other comments:

- I see the problem most often when running i3.  I don't know that I've
  ever seen it with a more conventional environment (Cinnamon, MATE,
  Xfce).  I have a handful of screenshots that show the problem,and
  they were all taken when running i3.

- The problem seems to happen less often when hinting is set to
  "slight", but it does happen.  I don't think I've tested with hinting
  set to "none".

- I've seen the problem with the Lucid build, both Cairo+HarfBuzz and
  Cairo-only, and with the GTK3 build (Cairo+HarfBuzz).

- I only see the problem with Emacs.  But I suppose that could be
  because Emacs is the application I spend the most time on.  I don't
  have a reproducer.

mike

--=-=-=
Content-Type: image/png
Content-Disposition: attachment; filename=text-glitch-1-medium-hinting.png
Content-Transfer-Encoding: base64
Content-Description: sample text rendering problem
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==
--=-=-=--




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 10 May 2026 21:18:44 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sun May 10 17:18:44 2026
Received: from localhost ([127.0.0.1]:40925 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wMBXv-0003NG-Vt
	for submit <at> debbugs.gnu.org; Sun, 10 May 2026 17:18:44 -0400
Received: from [78.47.144.35] (port=49632 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wMBXt-0003Md-0O
 for 81012 <at> debbugs.gnu.org; Sun, 10 May 2026 17:18:42 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id C0BD09C793; Sun, 10 May 2026 23:18:39 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <87o6inp0me.fsf@HIDDEN> <87o6in9h6q.fsf@HIDDEN>
Date: Sun, 10 May 2026 23:18:39 +0200
In-Reply-To: <87o6in9h6q.fsf@HIDDEN> (Pip Cet's message of "Sun, 10
 May 2026 19:51:35 +0000")
Message-ID: <87tssfnets.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview: Pip Cet <pipcet@HIDDEN> writes: > Maybe it's two
 different
 bugs. In fact, there was also another aspect to this issue that surprised
 me: Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Pip Cet <pipcet@HIDDEN> writes:

> Maybe it's two different bugs.

In fact, there was also another aspect to this issue that surprised me:

The contents of my ~/.Xresources file when I reported this issue were:

    *FontBackend:	xft
    Xft.dpi:	96
    Xft.hinting:	true
    Xft.hintstyle: hintslight

So, I had hintstyle set to hintslight. I deliberately had commented out
*everything* in ~/.emacs.

Still, I observed the issue *only* when launching with "emacs -Q". I did
not observe the issue when launching "emacs" without any arguments.

I found this surprising because from what I can tell, "hintslight" was
enabled both when using "-Q" and not using it, and the options in
~/.Xresources seem not specific to Emacs.

Maybe it could be worth considering to make the settings better
inspectable from within Elisp? Trying to detect from within Emacs
whether it uses hinting, I found "hinting" not to occur in the manual.

Thank you and all the best,
Markus




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 10 May 2026 19:51:52 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sun May 10 15:51:52 2026
Received: from localhost ([127.0.0.1]:39925 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wMABr-00087L-8J
	for submit <at> debbugs.gnu.org; Sun, 10 May 2026 15:51:52 -0400
Received: from mail-24418.protonmail.ch ([109.224.244.18]:13959)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <pipcet@HIDDEN>)
 id 1wMABl-00085d-O6
 for 81012 <at> debbugs.gnu.org; Sun, 10 May 2026 15:51:49 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=protonmail.com;
 s=protonmail3; t=1778442698; x=1778701898;
 bh=azptfLo0Ri8aEH61rI44PebWn8t5/gTt/1ELyWI+3X0=;
 h=Date:To:From:Cc:Subject:Message-ID:In-Reply-To:References:
 Feedback-ID:From:To:Cc:Date:Subject:Reply-To:Feedback-ID:
 Message-ID:BIMI-Selector;
 b=tjjpoVAuhay/ff2z2fjrjwwwQ88McdzS10tQheWeSwX0P662DABBD6Pt+1vbEce+R
 PoG3pnnsOzGqIe0aP/uDDBjvI6q/nD6wa8Cgrx2a7TJ/V5FprOdEIV6xstCvrTdMX/
 Ds44eOyAtkTrE7kp659D7hKnCNbKJAkpB3vHLBM5F1GjpnuIjccAKlL6I8MxINEm6v
 jHaXiBlsSXOSR6fcNuW6iUmmE3SOergXvNZutvvoN8KxrxzKLdcSh1Icy+UWdYNFFz
 u40EEW3QuVO7AuVv8OyYsXUjzghKwKpP8X6+EYGXkOhkcsxjHDuJl/FFKTt06KE9WN
 2RxYvexUDrbOQ==
Date: Sun, 10 May 2026 19:51:35 +0000
To: Markus Triska <triska@HIDDEN>
From: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
Message-ID: <87o6in9h6q.fsf@HIDDEN>
In-Reply-To: <87o6inp0me.fsf@HIDDEN>
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
 <87o6inp0me.fsf@HIDDEN>
Feedback-ID: 112775352:user:proton
X-Pm-Message-ID: 2d559a806414a32f21001f5fb90e7fdd8b64ac5a
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: -0.7 (/)
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.7 (-)

"Markus Triska" <triska@HIDDEN> writes:

> Pip Cet <pipcet@HIDDEN> writes:
>
>> This is definitely not a fix, but could you try applying this patch and
>> seeing whether the problems still occur (it's specific to the font
>> driver you're using, and a different patch is required for ftcrfont.c) ?
>
> Thank you a lot, meanwhile I tested it (i.e., unpatched Emacs) with both
> xft and xfth, and both exhibit the issue, *but* also good news: I found
> the issue, at least with the example I posted, does *not* arise if I
> disable hinting by putting the following in ~/.Xresources:
>
>     Xft.hintstyle: hintnone

I'm seeing less damage with that option, but still some of it. I'm using
"Source Code Pro", and I have no idea whether that font includes a
hinting program.

Maybe it's two different bugs.

> Maybe reliably disabling hardware acceleration for font rendering could
> be a candidate for an interesting command line switch or configurable
> frame parameter in Emacs? We can trust the Emacs documentation.

We default to disabling font hinting on Android, IIRC, when we use Po
Lu's font driver (rather than freetype, which requires a few
changes). Maybe a fast way to do so on Xft/ftcr would be good, too.
>
> Since disabling hinting suffices to work around the problem for me,
> please close the issue any time you conclude it is not an Emacs issue.

I think Emacs is simply unusual in clearing very small rectangles right
before rendering glyphs into them. My theory is that that somehow gets
reordered and part of the rectangle is filled with the background color
after the glyph is written, resulting in the damage.

Without hinting, characters are less aligned to pixel boundaries, so the
damage may still be there but less visible.

So I suspect it is an Emacs issue, and we're mixing X drawing primitives
and font library calls in a way that doesn't quite work.

Pip





Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 10 May 2026 19:16:11 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sun May 10 15:16:10 2026
Received: from localhost ([127.0.0.1]:39718 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wM9dK-0003fW-Ef
	for submit <at> debbugs.gnu.org; Sun, 10 May 2026 15:16:10 -0400
Received: from eggs.gnu.org ([2001:470:142:3::10]:51372)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <eliz@HIDDEN>) id 1wM9dH-0003eY-UW
 for 81012 <at> debbugs.gnu.org; Sun, 10 May 2026 15:16:08 -0400
Received: from fencepost.gnu.org ([2001:470:142:3::e])
 by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.90_1) (envelope-from <eliz@HIDDEN>)
 id 1wM9d8-0002oB-HJ; Sun, 10 May 2026 15:16:01 -0400
DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=gnu.org;
 s=fencepost-gnu-org; h=References:Subject:In-Reply-To:To:From:Date:
 mime-version; bh=kpuW17e2UonyrIYfyqNg2bklGsIhBMiCR3XXS03iBlk=; b=sIDpyO4BuPdt
 YEkF6JmY33htSjqu1bH+M1bn/ftFYftmVBcbkaNcV1E6fwUYNj3rbdfu8SJhZy/2NGl0QiyqK9Wz/
 3wcsgNpyz/k0Del2gFDF6A995/FRgFYjhJq33ntvV63ZFTBhtMD7fFqQEChblsYeYF/dGF7xQK4Uh
 OIQQutG8dOaW39vYhTzk29NSKOdpiQZKU9EnE3mr52v9yupMNq949tmDlf79WAbeGwN/ArT2h/Hra
 A+GN8XLZeS8Lhl8+vCiS9iZoSocS1bdZ2RLoYFUurDRljzfUC/uvwwGsEbAxAwAoHcKqtd35Nx6+6
 7REOoTjGwP/x21zBpySrWA==;
Date: Sun, 10 May 2026 22:15:55 +0300
Message-Id: <864ikf2hzo.fsf@HIDDEN>
From: Eli Zaretskii <eliz@HIDDEN>
To: Pip Cet <pipcet@HIDDEN>
In-Reply-To: <87wlxb9kuz.fsf@HIDDEN> (message from Pip Cet on Sun, 10
 May 2026 18:32:17 +0000)
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
X-Spam-Score: -2.3 (--)
X-Debbugs-Envelope-To: 81012
Cc: triska@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -3.3 (---)

> Date: Sun, 10 May 2026 18:32:17 +0000
> From: Pip Cet <pipcet@HIDDEN>
> Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
> 
> This is definitely not a fix, but could you try applying this patch and
> seeing whether the problems still occur (it's specific to the font
> driver you're using, and a different patch is required for ftcrfont.c) ?
> 
> diff --git a/src/xftfont.c b/src/xftfont.c
> index f15dbae1e7a..6b21ec72755 100644
> --- a/src/xftfont.c
> +++ b/src/xftfont.c
> @@ -579,6 +579,8 @@ xftfont_draw (struct glyph_string *s, int from, int to, int x, int y,
>  #endif
>  	XftDrawRect (xft_draw, &bg, x, y - ascent, s->width, height);
>      }
> +  struct x_display_info *dpyinfo = FRAME_DISPLAY_INFO (s->f);
> +  XSync (dpyinfo->display, false);
>    code = alloca (sizeof (FT_UInt) * len);
>    for (i = 0; i < len; i++)
>      code[i] = s->char2b[from + i];
> 
> It seems to avoid the issue here, but my eyes have gone kind of blurry
> at this point so maybe it's just that :-)
> 
> For me, the issue appears with both ftcrfont and xft, but goes away when
> I use Xnest.

Did you try to disable double-buffering?




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 10 May 2026 18:42:38 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sun May 10 14:42:38 2026
Received: from localhost ([127.0.0.1]:39495 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wM96r-0007aO-JM
	for submit <at> debbugs.gnu.org; Sun, 10 May 2026 14:42:38 -0400
Received: from [78.47.144.35] (port=48022 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wM96o-0007Zi-Sv
 for 81012 <at> debbugs.gnu.org; Sun, 10 May 2026 14:42:35 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id 862299C78B; Sun, 10 May 2026 20:42:33 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <87wlxb9kuz.fsf@HIDDEN>
Date: Sun, 10 May 2026 20:42:33 +0200
In-Reply-To: <87wlxb9kuz.fsf@HIDDEN> (Pip Cet's message of "Sun, 10
 May 2026 18:32:17 +0000")
Message-ID: <87o6inp0me.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview: Pip Cet <pipcet@HIDDEN> writes: > This is definitely
 not a fix, but could you try applying this patch and > seeing whether the
 problems still occur (it's specific to the font > driver you're using, and
 a different patch is required [...] 
 Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Pip Cet <pipcet@HIDDEN> writes:

> This is definitely not a fix, but could you try applying this patch and
> seeing whether the problems still occur (it's specific to the font
> driver you're using, and a different patch is required for ftcrfont.c) ?

Thank you a lot, meanwhile I tested it (i.e., unpatched Emacs) with both
xft and xfth, and both exhibit the issue, *but* also good news: I found
the issue, at least with the example I posted, does *not* arise if I
disable hinting by putting the following in ~/.Xresources:

    Xft.hintstyle: hintnone

For me, this suffices to work around the problem. Unfortunately I don't
know how to resolve any hardware or driver issues, I tried to disable
GPU acceleration but found so many conflicting kinds of advice that I
have no idea where to start or where to report any issues.

For instance, one thing that was suggested somewhere is to start Emacs
as follows:

    LIBGL_ALWAYS_SOFTWARE=1 emacs

whether this disables hardware acceleration, who can tell at this point
when most text on the Internet is now randomly generated?

Maybe reliably disabling hardware acceleration for font rendering could
be a candidate for an interesting command line switch or configurable
frame parameter in Emacs? We can trust the Emacs documentation.

Since disabling hinting suffices to work around the problem for me,
please close the issue any time you conclude it is not an Emacs issue.

Thank you all for looking into it and your responses and suggestions, I
greatly appreciate it!

All the best,
Markus




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 10 May 2026 18:32:37 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sun May 10 14:32:36 2026
Received: from localhost ([127.0.0.1]:39435 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wM8x8-0006G9-JE
	for submit <at> debbugs.gnu.org; Sun, 10 May 2026 14:32:36 -0400
Received: from mail-244122.protonmail.ch ([109.224.244.122]:50657)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <pipcet@HIDDEN>)
 id 1wM8x5-0006F7-L7
 for 81012 <at> debbugs.gnu.org; Sun, 10 May 2026 14:32:32 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=protonmail.com;
 s=protonmail3; t=1778437942; x=1778697142;
 bh=iNZDlkyxhjXBTG5+qP1P0eW96958ImcJW/cLyLUDaBQ=;
 h=Date:To:From:Cc:Subject:Message-ID:In-Reply-To:References:
 Feedback-ID:From:To:Cc:Date:Subject:Reply-To:Feedback-ID:
 Message-ID:BIMI-Selector;
 b=ybVsxFQFbYZkhQY83AUBt2HmqRNro+6NZQNWHbCjUcZarHJhAXpvfO0UW4A2g2+GE
 2pPYyhxqNJm/tnj6/bXvgaZDw4fNUVuK6HB9OrpeNg8fbHlzwsMQviURmrpRqMzMML
 oNPMP/dsLpmrDtomzXl6VhSeijPK9EW9oovJ+Ll/GY82W9d7g6Uqte9RgQBNHLRfGV
 GsVpkk5ZO/2z4kPEJ3gv8SYW8BkCrDWTIM9BHtyIFuoqrUb7Yq/ANe+TaLy85QaFB2
 1u8tmKupxkNvAY0tewNZZsyaynGkbAVSexIrJH8hHbJsMH9aJ/maWEDrDAixW0XeoC
 OlZEAHUJQhMwA==
Date: Sun, 10 May 2026 18:32:17 +0000
To: Markus Triska <triska@HIDDEN>
From: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
Message-ID: <87wlxb9kuz.fsf@HIDDEN>
In-Reply-To: <87se7z1kp6.fsf@HIDDEN>
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN>
Feedback-ID: 112775352:user:proton
X-Pm-Message-ID: 447b9a3c342df7e0c924814d459df7dde81cc43e
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: -0.7 (/)
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.7 (-)

Pip Cet <pipcet@HIDDEN> writes:

> "Markus Triska" <triska@HIDDEN> writes:
>
>> Eli Zaretskii <eliz@HIDDEN> writes:
>>
>>> Not reproducible here.  Are you sure this is not due to some
>>> "optimizations" done by your video driver software?
>>
>> I'm quite sure: I'm using amdgpu, to the best of my knowledge among the
>> best supported video drivers in free software environments, and the
>> issue arises only in very specific situations in Emacs and I have so far
>> never seen it in any other program.
>>
>> Also, I originally forgot to add: To evaluate the form, please paste it
>> into the *scratch* buffer, so that enough content is present for the
>> included (recenter -1) call to have the intended effect.
>>
>> In case it helps, here is a (software-based) screen recording of the iss=
ue:
>>
>>     https://www.metalevel.at/ei/type_after_scroll.mp4
>>
>> The issue is visible for example on the 2nd, 3rd and 5th "t" in the
>> line, and many others too, while for example the 1st and 4th "t" are
>> drawn exactly as expected. Note that the issue persists even after I
>> press a key and evaluation of the form is finished. When I move the
>> cursor over the letters at the end of the video, each of them is drawn
>> correctly the moment it is under the cursor.
>
> I don't think that's true: if you step through the single frames of the
> video, you'll see that sometimes the damaged character is inverted with
> the damage for a single frame, then fixes itself a frame later.
>
>> I hope that the recording software thus also "seeing" the problem rules
>> out hardware issues with the display itself.
>>
>> Can anyone else reproduce the problem? I would greatly appreciate any
>> feedback and help.
>
> I think I can. It might still be hardware/driver/memory corruption.
>
> I've seen similar problems, but I'd always blamed this on the amdgpu
> driver. Unfortunately, this is a hidpi screen and the errors seem to be
> single pixels at most, so it's kind of hard to detect this without
> laborious screenshotting.
>
> But it does happen: in fact, when I started typing this, the "I" in
> "I've seen" was noticeably incomplete. (This is my production session).
>
> It's very, very strange: when I move the cursor over a damaged
> character, it sometimes gets inverted with the corruption, but fixes
> itself after another second or so, still inverted. But sometimes it gets
> fixed immediately.
>
> (redisplay) and (clear-font-cache) don't fix the characters, and C-l
> moves the damage along with the characters at least some of the time.
>
> TBH, this looks like a driver/hardware issue to me. I've only ever seen
> this on this one machine, and I've seen corrupt pixels assume different
> colors from the normal text. Maybe something goes wrong on the way to
> video RAM and the screenshot software accesses the damaged pixels?
>
> Are the damaged pixels ever in the middle of a glyph, or do they happen
> only along the border of the glyph's bbox?
>
> Pip

This is definitely not a fix, but could you try applying this patch and
seeing whether the problems still occur (it's specific to the font
driver you're using, and a different patch is required for ftcrfont.c) ?

diff --git a/src/xftfont.c b/src/xftfont.c
index f15dbae1e7a..6b21ec72755 100644
--- a/src/xftfont.c
+++ b/src/xftfont.c
@@ -579,6 +579,8 @@ xftfont_draw (struct glyph_string *s, int from, int to,=
 int x, int y,
 #endif
 =09XftDrawRect (xft_draw, &bg, x, y - ascent, s->width, height);
     }
+  struct x_display_info *dpyinfo =3D FRAME_DISPLAY_INFO (s->f);
+  XSync (dpyinfo->display, false);
   code =3D alloca (sizeof (FT_UInt) * len);
   for (i =3D 0; i < len; i++)
     code[i] =3D s->char2b[from + i];

It seems to avoid the issue here, but my eyes have gone kind of blurry
at this point so maybe it's just that :-)

For me, the issue appears with both ftcrfont and xft, but goes away when
I use Xnest.

Pip





Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 10 May 2026 15:55:51 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sun May 10 11:55:50 2026
Received: from localhost ([127.0.0.1]:38356 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wM6VQ-0001XQ-UW
	for submit <at> debbugs.gnu.org; Sun, 10 May 2026 11:55:50 -0400
Received: from eggs.gnu.org ([2001:470:142:3::10]:58986)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <eliz@HIDDEN>) id 1wM6VO-0001W6-FM
 for 81012 <at> debbugs.gnu.org; Sun, 10 May 2026 11:55:47 -0400
Received: from fencepost.gnu.org ([2001:470:142:3::e])
 by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.90_1) (envelope-from <eliz@HIDDEN>)
 id 1wM6VF-0005Hw-3Y; Sun, 10 May 2026 11:55:38 -0400
DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=gnu.org;
 s=fencepost-gnu-org; h=References:Subject:In-Reply-To:To:From:Date:
 mime-version; bh=q/hqH/NC8oZYecMhrMvZEcpZux4QIR24wP4IeZnKXBY=; b=oxFRUbToxh49
 VFTPwg83pt+apWan6qYkUiaTJkElD0HRA/9D9xm/1l+YrvEg/3scfdCk+0AsT1Q8ddiTkqDbi12Xd
 IzSQhG/N8BK01gHbsRU2pmLqfeG/2IiwQhEa9/7dHjz5qpzs/iHHEiFNVtBtyKjh2bngJoBrS/tTf
 ULtPQYQ0ZMoyp+3x/0+VCGfvyn8rgQNEenHPcJGuzmg2UiGAKqH0AymC3+QExmgs6iaIm/ry+ztfe
 PtbkDSrRNr6fyR64swnbinDAGp1Ps4Q2IEhWXJFzWmW/udeFxo6eSZeac3xp9liE63KNDXSJcebHt
 qaJoJ4A8a7OkLqYBah1crg==;
Date: Sun, 10 May 2026 18:55:34 +0300
Message-Id: <86cxz32r9l.fsf@HIDDEN>
From: Eli Zaretskii <eliz@HIDDEN>
To: Pip Cet <pipcet@HIDDEN>
In-Reply-To: <878q9rb9jk.fsf@HIDDEN> (message from Pip Cet on Sun, 10
 May 2026 14:53:49 +0000)
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN> <878q9rb9jk.fsf@HIDDEN>
X-Spam-Score: -2.3 (--)
X-Debbugs-Envelope-To: 81012
Cc: triska@HIDDEN, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -3.3 (---)

> Date: Sun, 10 May 2026 14:53:49 +0000
> From: Pip Cet <pipcet@HIDDEN>
> Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
> 
> TBH, this looks like a driver/hardware issue to me. I've only ever seen
> this on this one machine, and I've seen corrupt pixels assume different
> colors from the normal text. Maybe something goes wrong on the way to
> video RAM and the screenshot software accesses the damaged pixels?

I suggest to look at the settings of the video driver software, and if
it has some "optimization" features, turn them off and see if that
fixes the problem.




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 10 May 2026 15:20:04 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sun May 10 11:20:04 2026
Received: from localhost ([127.0.0.1]:38038 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wM5wn-0004zg-GU
	for submit <at> debbugs.gnu.org; Sun, 10 May 2026 11:20:04 -0400
Received: from eggs.gnu.org ([2001:470:142:3::10]:57140)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <eliz@HIDDEN>) id 1wM5wk-0004yF-0v
 for 81012 <at> debbugs.gnu.org; Sun, 10 May 2026 11:20:00 -0400
Received: from fencepost.gnu.org ([2001:470:142:3::e])
 by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.90_1) (envelope-from <eliz@HIDDEN>)
 id 1wM5wd-0005IZ-Te; Sun, 10 May 2026 11:19:51 -0400
DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=gnu.org;
 s=fencepost-gnu-org; h=References:Subject:In-Reply-To:To:From:Date:
 mime-version; bh=3K+B7zbHxL022prnxXiHr06iALnv6XPZJKT3tGoYu/4=; b=CZTtarGuqD2T
 H4FxTeMp4hKcg2SSQIVIftmHQj7aXkzaK41h0xuhV60tB7QrfMlqtCOXPXiF+uJMyyKwSaphEAzqU
 ePE/vJiFFppZU1YSwV/iAw2saqWBp6NpWdhBz23c8ymyH/KW0mi3/Fo5YsPdxIG+7QH/wXiEsx4h4
 k58uioaihyoFQWgGS10cypFigBYQjn/4MQBACOg79r8/yLqrtAtiPM6NPrX06MrZ92+J80P4WUA3H
 B+VDnwCZdnW2gV8du0BenS6UfpZC28pgj11j6jjp7VHHXdTImhrpDhohPnHawoIJRIW9eSrYbIc7N
 A0dU6DDszgVLdk6riQNfLA==;
Date: Sun, 10 May 2026 18:19:49 +0300
Message-Id: <86fr3z2sx6.fsf@HIDDEN>
From: Eli Zaretskii <eliz@HIDDEN>
To: Markus Triska <triska@HIDDEN>
In-Reply-To: <87se7z1kp6.fsf@HIDDEN> (message from Markus Triska on Sun, 
 10 May 2026 15:02:45 +0200)
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN>
X-Spam-Score: -2.3 (--)
X-Debbugs-Envelope-To: 81012
Cc: 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -3.3 (---)

> From: Markus Triska <triska@HIDDEN>
> Cc: 81012 <at> debbugs.gnu.org
> Date: Sun, 10 May 2026 15:02:45 +0200
> 
> Eli Zaretskii <eliz@HIDDEN> writes:
> 
> > Not reproducible here.  Are you sure this is not due to some
> > "optimizations" done by your video driver software?
> 
> I'm quite sure: I'm using amdgpu, to the best of my knowledge among the
> best supported video drivers in free software environments, and the
> issue arises only in very specific situations in Emacs and I have so far
> never seen it in any other program.

That doesn't mean the problem is not there.  Emacs doesn't draw
partial glyphs, so problems like what you see are unlikely to be
caused by Emacs's own code.

> Also, I originally forgot to add: To evaluate the form, please paste it
> into the *scratch* buffer, so that enough content is present for the
> included (recenter -1) call to have the intended effect.

That's what I did to begin with.

> In case it helps, here is a (software-based) screen recording of the issue:
> 
>     https://www.metalevel.at/ei/type_after_scroll.mp4

I believe you that you see what you describe even without the
recording.  I just don't see it on my system.

> The issue is visible for example on the 2nd, 3rd and 5th "t" in the
> line, and many others too, while for example the 1st and 4th "t" are
> drawn exactly as expected. Note that the issue persists even after I
> press a key and evaluation of the form is finished. When I move the
> cursor over the letters at the end of the video, each of them is drawn
> correctly the moment it is under the cursor.

When you move cursor, Emacs redraws each character the cursor is on,
so this basically redraws all the line, one characters at a time.




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 10 May 2026 14:54:09 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sun May 10 10:54:08 2026
Received: from localhost ([127.0.0.1]:37698 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wM5Xj-0001Uc-43
	for submit <at> debbugs.gnu.org; Sun, 10 May 2026 10:54:08 -0400
Received: from mail-43102.protonmail.ch ([185.70.43.102]:32821)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <pipcet@HIDDEN>)
 id 1wM5Xe-0001T1-GM
 for 81012 <at> debbugs.gnu.org; Sun, 10 May 2026 10:54:04 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=protonmail.com;
 s=protonmail3; t=1778424835; x=1778684035;
 bh=0Yq/7AwIj28zZzH49Ycr//3ZIfpsBSGZwbkSNv1lgAw=;
 h=Date:To:From:Cc:Subject:Message-ID:In-Reply-To:References:
 Feedback-ID:From:To:Cc:Date:Subject:Reply-To:Feedback-ID:
 Message-ID:BIMI-Selector;
 b=vrkx824+CFD6nb/5VeSWXOk4yssq3vLkv/o20T5S7mutpo/g2fvo2NAJDtgg3g462
 kAkYCtlsty2U08Z+o+rKDlpQ0zRqj55fJSnw1KgWYdylzg829yRwH24VZbMWyepSX1
 5YH7x9suq4REvl3DyYQYAIUwBC+MT3u4yz15ps4AlUVcu5KdNFq18Ht0cJgVYJnhAk
 m7VZAFNxFoPXGPcGpjPmDzIKdssjSju30HBBVsSg7wc1jd9Fidj8kkhPAp9hYKtsz0
 9w0IjV1WSDSg79bc6d3YM+MZ8IwmU1JOXOnqseGe2igYkXpLIfpEQPLxMB51mkKvCq
 em8jY5ism5zqg==
Date: Sun, 10 May 2026 14:53:49 +0000
To: Markus Triska <triska@HIDDEN>
From: Pip Cet <pipcet@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
Message-ID: <878q9rb9jk.fsf@HIDDEN>
In-Reply-To: <87se7z1kp6.fsf@HIDDEN>
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
 <87se7z1kp6.fsf@HIDDEN>
Feedback-ID: 112775352:user:proton
X-Pm-Message-ID: debef9fda730e7c310478e4f2ba96e0fcbfc444b
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: -0.7 (/)
X-Debbugs-Envelope-To: 81012
Cc: Eli Zaretskii <eliz@HIDDEN>, 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.7 (-)

"Markus Triska" <triska@HIDDEN> writes:

> Eli Zaretskii <eliz@HIDDEN> writes:
>
>> Not reproducible here.  Are you sure this is not due to some
>> "optimizations" done by your video driver software?
>
> I'm quite sure: I'm using amdgpu, to the best of my knowledge among the
> best supported video drivers in free software environments, and the
> issue arises only in very specific situations in Emacs and I have so far
> never seen it in any other program.
>
> Also, I originally forgot to add: To evaluate the form, please paste it
> into the *scratch* buffer, so that enough content is present for the
> included (recenter -1) call to have the intended effect.
>
> In case it helps, here is a (software-based) screen recording of the issu=
e:
>
>     https://www.metalevel.at/ei/type_after_scroll.mp4
>
> The issue is visible for example on the 2nd, 3rd and 5th "t" in the
> line, and many others too, while for example the 1st and 4th "t" are
> drawn exactly as expected. Note that the issue persists even after I
> press a key and evaluation of the form is finished. When I move the
> cursor over the letters at the end of the video, each of them is drawn
> correctly the moment it is under the cursor.

I don't think that's true: if you step through the single frames of the
video, you'll see that sometimes the damaged character is inverted with
the damage for a single frame, then fixes itself a frame later.

> I hope that the recording software thus also "seeing" the problem rules
> out hardware issues with the display itself.
>
> Can anyone else reproduce the problem? I would greatly appreciate any
> feedback and help.

I think I can. It might still be hardware/driver/memory corruption.

I've seen similar problems, but I'd always blamed this on the amdgpu
driver. Unfortunately, this is a hidpi screen and the errors seem to be
single pixels at most, so it's kind of hard to detect this without
laborious screenshotting.

But it does happen: in fact, when I started typing this, the "I" in
"I've seen" was noticeably incomplete. (This is my production session).

It's very, very strange: when I move the cursor over a damaged
character, it sometimes gets inverted with the corruption, but fixes
itself after another second or so, still inverted. But sometimes it gets
fixed immediately.

(redisplay) and (clear-font-cache) don't fix the characters, and C-l
moves the damage along with the characters at least some of the time.

TBH, this looks like a driver/hardware issue to me. I've only ever seen
this on this one machine, and I've seen corrupt pixels assume different
colors from the normal text. Maybe something goes wrong on the way to
video RAM and the screenshot software accesses the damaged pixels?

Are the damaged pixels ever in the middle of a glyph, or do they happen
only along the border of the glyph's bbox?

Pip





Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 10 May 2026 13:02:51 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sun May 10 09:02:51 2026
Received: from localhost ([127.0.0.1]:36777 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wM3o2-00032T-PP
	for submit <at> debbugs.gnu.org; Sun, 10 May 2026 09:02:51 -0400
Received: from [78.47.144.35] (port=44736 helo=metalevel.at)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <triska@HIDDEN>) id 1wM3nz-000321-9J
 for 81012 <at> debbugs.gnu.org; Sun, 10 May 2026 09:02:49 -0400
Received: by metalevel.at (Postfix, from userid 1000)
 id B68939C789; Sun, 10 May 2026 15:02:45 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: Eli Zaretskii <eliz@HIDDEN>
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8> <86jytb38az.fsf@HIDDEN>
Date: Sun, 10 May 2026 15:02:45 +0200
In-Reply-To: <86jytb38az.fsf@HIDDEN> (Eli Zaretskii's message of "Sun, 10 May
 2026 12:47:32 +0300")
Message-ID: <87se7z1kp6.fsf@HIDDEN>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: 1.3 (+)
X-Spam-Report: Spam detection software, running on the system "debbugs.gnu.org",
 has NOT identified this incoming email as spam.  The original
 message has been attached to this so you can view it or label
 similar future email.  If you have any questions, see
 the administrator of that system for details.
 Content preview: Eli Zaretskii <eliz@HIDDEN> writes: > Not reproducible here.
 Are you sure this is not due to some > "optimizations" done by your video
 driver software? I'm quite sure: I'm using amdgpu, to the best of my knowledge
 among the best supported video drivers in free software environments, and
 the issue arises only in very specific situations in Emacs and I [...] 
 Content analysis details:   (1.3 points, 10.0 required)
 pts rule name              description
 ---- ---------------------- --------------------------------------------------
 0.0 SPF_NONE               SPF: sender does not publish an SPF Record
 0.0 SPF_HELO_NONE          SPF: HELO does not publish an SPF Record
 1.3 RDNS_NONE Delivered to internal network by a host with no rDNS
X-Debbugs-Envelope-To: 81012
Cc: 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: 0.3 (/)

Eli Zaretskii <eliz@HIDDEN> writes:

> Not reproducible here.  Are you sure this is not due to some
> "optimizations" done by your video driver software?

I'm quite sure: I'm using amdgpu, to the best of my knowledge among the
best supported video drivers in free software environments, and the
issue arises only in very specific situations in Emacs and I have so far
never seen it in any other program.

Also, I originally forgot to add: To evaluate the form, please paste it
into the *scratch* buffer, so that enough content is present for the
included (recenter -1) call to have the intended effect.

In case it helps, here is a (software-based) screen recording of the issue:

    https://www.metalevel.at/ei/type_after_scroll.mp4

The issue is visible for example on the 2nd, 3rd and 5th "t" in the
line, and many others too, while for example the 1st and 4th "t" are
drawn exactly as expected. Note that the issue persists even after I
press a key and evaluation of the form is finished. When I move the
cursor over the letters at the end of the video, each of them is drawn
correctly the moment it is under the cursor.

I hope that the recording software thus also "seeing" the problem rules
out hardware issues with the display itself.

Can anyone else reproduce the problem? I would greatly appreciate any
feedback and help.

Thank you and all the best,
Markus




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at 81012 <at> debbugs.gnu.org:


Received: (at 81012) by debbugs.gnu.org; 10 May 2026 09:47:45 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sun May 10 05:47:44 2026
Received: from localhost ([127.0.0.1]:34881 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wM0lD-0006lL-SB
	for submit <at> debbugs.gnu.org; Sun, 10 May 2026 05:47:44 -0400
Received: from eggs.gnu.org ([2001:470:142:3::10]:36298)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <eliz@HIDDEN>) id 1wM0lB-0006k8-59
 for 81012 <at> debbugs.gnu.org; Sun, 10 May 2026 05:47:41 -0400
Received: from fencepost.gnu.org ([2001:470:142:3::e])
 by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.90_1) (envelope-from <eliz@HIDDEN>)
 id 1wM0l5-0004In-BX; Sun, 10 May 2026 05:47:35 -0400
DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=gnu.org;
 s=fencepost-gnu-org; h=References:Subject:In-Reply-To:To:From:Date:
 mime-version; bh=iP/WP3MuI6HpSP+63sIQgteTgRprSvYwNHGFBtDgY8c=; b=Wk1Mu0m5O5PH
 she7B2pJ8GEeqSafxS+NC/8yVjOGf5bn8VP55jHH7sNYxCTUJ/bCnIVB1zUjLMmYRD+s8Xxn77GXX
 PUdzNc5MD1mRFh5bxQVXWmYcmejQMwLOK8nUBva+vmCwVFGGMJwZ2o2JPDcfxH0yEhVeThvGoFO0Y
 qdvo/lm36vmjL25YZ6B8a6PYyAUDVLO2HgbP6dhQ9kE+Iei1PGMG7aPvCzCneiebSv4Q+E37kxfYn
 0S/ikPlgi0GviGojpzHPRteS6EeWQTMnrdEz0Pf0At85w2DBvV0jJ7Sh1VzAvx9bHWXDDW9OPwWrB
 HFEICn3NaunLz2ZIbhHzcg==;
Date: Sun, 10 May 2026 12:47:32 +0300
Message-Id: <86jytb38az.fsf@HIDDEN>
From: Eli Zaretskii <eliz@HIDDEN>
To: Markus Triska <triska@HIDDEN>
In-Reply-To: <20260510083506.293CBF08697@triska-ser8> (message from Markus
 Triska on Sun, 10 May 2026 10:35:06 +0200 (CEST))
Subject: Re: bug#81012: 30.2;
 Letters are sometimes unexpectedly not fully drawn
References: <20260510083506.293CBF08697@triska-ser8>
X-Spam-Score: -2.3 (--)
X-Debbugs-Envelope-To: 81012
Cc: 81012 <at> debbugs.gnu.org
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -3.3 (---)

> From: Markus Triska <triska@HIDDEN>
> Date: Sun, 10 May 2026 10:35:06 +0200 (CEST)
> 
> To reproduce this issue, please start Emacs with:
> 
>     $ emacs -Q
> 
> and then evaluate the form shown below.
> 
> The form prepares the buffer as needed, and then asks you to press a
> key. After you press any key, the form creates a "page break" by
> inserting 2 newlines, and then inserts a sequence of "t"s in the
> buffer: "ttttttt ....". Press any key to interrupt this.
> 
> Many of the "t"s are shown exactly as expected. But some of them are
> only partially drawn, missing several pixels in their horizontal bar
> or also in their stems or tails. By resizing the frame, I noticed that
> the parts that are not fully drawn as expected may also depend on the
> frame geometry.

Not reproducible here.  Are you sure this is not due to some
"optimizations" done by your video driver software?




Information forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.

Message received at submit <at> debbugs.gnu.org:


Received: (at submit) by debbugs.gnu.org; 10 May 2026 08:30:26 +0000
From debbugs-submit-bounces <at> debbugs.gnu.org Sun May 10 04:30:25 2026
Received: from localhost ([127.0.0.1]:34443 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1wLzYP-0003Md-5b
	for submit <at> debbugs.gnu.org; Sun, 10 May 2026 04:30:25 -0400
Received: from lists1p.gnu.org ([2001:470:142::17]:35330)
 by debbugs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.84_2) (envelope-from <mt@triska-ser8>) id 1wLzYL-0002lc-0g
 for submit <at> debbugs.gnu.org; Sun, 10 May 2026 04:30:24 -0400
Received: from eggs.gnu.org ([2001:470:142:3::10])
 by lists1p.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)
 (Exim 4.90_1) (envelope-from <mt@triska-ser8>) id 1wLzYB-0002FY-IJ
 for bug-gnu-emacs@HIDDEN; Sun, 10 May 2026 04:30:13 -0400
Received: from [78.47.144.35] (helo=metalevel.at)
 by eggs.gnu.org with esmtp (Exim 4.90_1)
 (envelope-from <mt@triska-ser8>) id 1wLzY5-0006YG-U8
 for bug-gnu-emacs@HIDDEN; Sun, 10 May 2026 04:30:10 -0400
Received: from triska-ser8 (localhost [127.0.0.1])
 by metalevel.at (Postfix) with ESMTP id 2382D9C776
 for <bug-gnu-emacs@HIDDEN>; Sun, 10 May 2026 10:30:04 +0200 (CEST)
Received: by triska-ser8 (Postfix, from userid 1000)
 id 293CBF08697; Sun, 10 May 2026 10:35:06 +0200 (CEST)
From: Markus Triska <triska@HIDDEN>
To: bug-gnu-emacs@HIDDEN
Subject: 30.2; Letters are sometimes unexpectedly not fully drawn
X-Debbugs-Cc: 
Message-Id: <20260510083506.293CBF08697@triska-ser8>
Date: Sun, 10 May 2026 10:35:06 +0200 (CEST)
X-Host-Lookup-Failed: Reverse DNS lookup failed for 78.47.144.35 (failed)
Received-SPF: none client-ip=78.47.144.35; envelope-from=mt@triska-ser8;
 helo=metalevel.at
X-Spam_score_int: -8
X-Spam_score: -0.9
X-Spam_bar: /
X-Spam_report: (-0.9 / 5.0 requ) BAYES_00=-1.9,
 HEADER_FROM_DIFFERENT_DOMAINS=0.249, RDNS_NONE=0.793, SPF_HELO_NONE=0.001,
 SPF_NONE=0.001 autolearn=no autolearn_force=no
X-Spam_action: no action
X-Spam-Score: 0.2 (/)
X-Debbugs-Envelope-To: submit
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -0.8 (/)

To reproduce this issue, please start Emacs with:

    $ emacs -Q

and then evaluate the form shown below.

The form prepares the buffer as needed, and then asks you to press a
key. After you press any key, the form creates a "page break" by
inserting 2 newlines, and then inserts a sequence of "t"s in the
buffer: "ttttttt ....". Press any key to interrupt this.

Many of the "t"s are shown exactly as expected. But some of them are
only partially drawn, missing several pixels in their horizontal bar
or also in their stems or tails. By resizing the frame, I noticed that
the parts that are not fully drawn as expected may also depend on the
frame geometry.

Unexpectedly to me, the issue seems also to depend on the content that
already appears in the buffer, hence this specific sample text.

The form follows below.

Thank you and all the best!
Markus

(let ((text "



We have:

    ?- collatz_next(3, N).
       N = 10.
       
On the other hand: 

    ?- collatz_next(10, N).
       N = 5
    ;  false.% redundant choicepoint 
    
We can eliminate the choicepoint with reification: "))
  (fundamental-mode)
  (buffer-face-set :family "DejaVu Sans Mono" :height 218)
  (insert "\n" text)
  (recenter -1)
  (read-key)
  (insert "\n\n    ")
  (let (read)
    (while (not read) 
      (insert "t")
      (setq read (read-event nil t 0.3)))))



In GNU Emacs 30.2 (build 9, x86_64-pc-linux-gnu, X toolkit, Xaw scroll
 bars) of 2026-01-09 built on beelink-ser8
Windowing system distributor 'The X.Org Foundation', version 11.0.12101016
System Description: Debian GNU/Linux 13 (trixie)

Configured using:
 'configure --without-cairo --with-x-toolkit=lucid
 --with-xpm=ifavailable --with-gif=ifavailable --with-tiff=ifavailable
 --with-gnutls=ifavailable'

Configured features:
FREETYPE GIF GMP GNUTLS JPEG LIBXML2 MODULES NATIVE_COMP NOTIFY
INOTIFY PDUMPER PNG SECCOMP SOUND THREADS TOOLKIT_SCROLL_BARS X11 XDBE
XFT XIM XPM LUCID ZLIB





Acknowledgement sent to Markus Triska <triska@HIDDEN>:
New bug report received and forwarded. Copy sent to bug-gnu-emacs@HIDDEN. Full text available.
Report forwarded to bug-gnu-emacs@HIDDEN:
bug#81012; Package emacs. Full text available.
Please note: This is a static page, with minimal formatting, updated once a day.
Click here to see this page with the latest information and nicer formatting.
Last modified: Sat, 6 Jun 2026 11:30:03 UTC

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