dri-devel Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Alan Mackenzie" <acm@muc.de>
Cc: linux-serial@vger.kernel.org, dri-devel@lists.freedesktop.org
Subject: Re: [Patch 5/9]: vt: Amend three Kconfig files
Date: Thu, 27 Aug 2026 19:12:57 +0000	[thread overview]
Message-ID: <20260827191257.AE0A51F000E9@smtp.kernel.org> (raw)
In-Reply-To: <apCG-zHyuMVYHJ6j@MAC.fritz.box>

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [High] Enabling CONFIG_FB_GLYPH_21BIT changes GLYPH_SZ to 4 globally, breaking hardware console drivers (like vgacon) and rotated fbcon drivers that still hardcode 16-bit screen buffer accesses.
- [Medium] FB_GLYPH_21BIT is a generic VT/fbdev feature but is inappropriately placed in the DRM clients menu, unnecessarily restricting its use.
--

commit 7b9fb605add2fa8c6f62e8ba79cee762cbcbd24c
Author: Alan Mackenzie <acm@muc.de>

vt: Amend three Kconfig files

Until it becomes clear how to integrate CONFIG_A11Y_BRAILLE_CONSOLE and
CONFIG_SPEAKUP, they will remain disabled when CONFIG_FB_GLYPH_21BIT is
#defined.

[ ... ]

> diff --git a/drivers/gpu/drm/clients/Kconfig b/drivers/gpu/drm/clients/Kconfig
> index 6096c623d9d5b..0ce8c1506eaad 100644
> --- a/drivers/gpu/drm/clients/Kconfig
> +++ b/drivers/gpu/drm/clients/Kconfig
> @@ -42,6 +42,15 @@ config DRM_FBDEV_EMULATION
>  
>  	  If in doubt, say "Y".
>  
> +config FB_GLYPH_21BIT
> +	bool "Enable 21-bit glyphs in the frame buffer device"
> +	depends on (DRM_FBDEV_EMULATION = y)

[Severity: Medium]
Could placing this generic VT/fbdev feature in the DRM clients menu
inappropriately restrict its use?

Native framebuffer drivers (e.g., vesafb, efifb, simplefb) might be prevented
from utilizing 21-bit glyph support, as this strictly requires
DRM_FBDEV_EMULATION. The feature modifies the generic Virtual Terminal and
fbcon subsystems, which are independent of DRM.

[Severity: High]
Can enabling CONFIG_FB_GLYPH_21BIT corrupt the screen buffer for hardware
console drivers and rotated fbcon drivers?

When this is enabled, GLYPH_SZ becomes 4 globally and VT uses 32-bit screen
buffer entries. However, the consw interface still passes pointers typed as
const unsigned short *, and some drivers still hardcode 16-bit accesses.

For example, rotated fbcon implementations iterate the buffer using scr_readw()
which advances the u16 pointer by 2 bytes instead of 4, reading half-characters:

drivers/video/fbdev/core/fbcon_cw.c:cw_putcs_aligned() {
    ...
    src = par->rotated.buf + (scr_readw(s++) & charmask) * cellsize;
    ...
}

Similarly, vgacon maps physical VGA VRAM assuming exactly 16 bits per character,
meaning 32-bit writes from scr_writeg in vt.c will overwrite adjacent cells
in hardware memory:

drivers/video/console/vgacon.c:vgacon_invert_region() {
    ...
    u16 a = scr_readw(p);
    ...
    scr_writew(a, p++);
    ...
}

> +	help
> +	  This option enables 21-bit glyphs in the fbdev framebuffer,
> +	  surmounting the traditional restriction to 256/512 glyphs dating back
> +	  to the VGA hardware of the 1980s and 1990s.  To use this effectively
> +	  you will need a suitable font and possibly a suitable keyboard layout.
> +

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/apCEDM2sWv_M354-@MAC.fritz.box?part=5

  reply	other threads:[~2026-08-27 19:13 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-08-27 18:38 vt: Enlarge the framebuffer glyph size from 16 to 32 bits Alan Mackenzie
2026-08-27 18:42 ` [Patch 1/9]: Make consolemap.c handle Unicode planes outside BMP Alan Mackenzie
2026-08-27 19:14   ` sashiko-bot
2026-08-28  4:57   ` Jiri Slaby
2026-08-29 11:51     ` Alan Mackenzie
2026-08-27 18:45 ` [Patch 2/9]: Glyph size: Use GLYPH_SZ/HW rather than hardcoded 2, 1 Alan Mackenzie
2026-08-27 19:22   ` sashiko-bot
2026-08-27 18:47 ` [Patch 3/9]: Replace scr_readw/writew by scr_readg/writeg, etc Alan Mackenzie
2026-08-27 19:18   ` sashiko-bot
2026-08-29 12:11   ` Greg Kroah-Hartman
2026-08-29 13:33     ` Alan Mackenzie
2026-08-27 18:48 ` [Patch 4/9]: Amend internal manipulation of glyph structure Alan Mackenzie
2026-08-27 19:19   ` sashiko-bot
2026-08-27 18:50 ` [Patch 5/9]: vt: Amend three Kconfig files Alan Mackenzie
2026-08-27 19:12   ` sashiko-bot [this message]
2026-08-27 18:52 ` [Patch 6/9]: vt: Use u32 and typedef u1632 to handle whole glyphs Alan Mackenzie
2026-08-27 19:22   ` sashiko-bot
2026-08-27 18:54 ` [Patch 7/9]: vt: Handle up to 2^21 glyphs, rather than 256/512 Alan Mackenzie
2026-08-27 19:14   ` sashiko-bot
2026-08-27 18:56 ` [Patch 8/9]: vt: Enhancements to the VT ioctl interface Alan Mackenzie
2026-08-27 19:25   ` sashiko-bot
2026-08-27 18:58 ` [Patch 9/9]: vt: Misc changes, e.g. to #include directives Alan Mackenzie
2026-08-27 19:25   ` sashiko-bot
2026-08-29 12:06   ` Greg Kroah-Hartman
2026-08-28  6:12 ` vt: Enlarge the framebuffer glyph size from 16 to 32 bits Thomas Zimmermann
2026-08-28 14:36   ` Alan Mackenzie
2026-08-29 12:09     ` Greg Kroah-Hartman
2026-08-29 14:39       ` Alan Mackenzie
2026-08-31  6:20         ` Thomas Zimmermann
2026-08-31 14:45           ` Alan Mackenzie
2026-08-31 17:21             ` Helge Deller
2026-09-01 11:14               ` Alan Mackenzie

Reply instructions:

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

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

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

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

  git send-email \
    --in-reply-to=20260827191257.AE0A51F000E9@smtp.kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=acm@muc.de \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=linux-serial@vger.kernel.org \
    --cc=sashiko-reviews@lists.linux.dev \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox