From: sashiko-bot@kernel.org
To: "Thomas Zimmermann" <tzimmermann@suse.de>
Cc: dri-devel@lists.freedesktop.org
Subject: Re: [PATCH v2 3/6] lib/fonts, fbcon: Use font_data_t for rotated fonts
Date: Fri, 20 Nov 2026 12:01:07 +0000 [thread overview]
Message-ID: <sashiko-outbox-225104@kernel.org> (raw)
In-Reply-To: <20261120115215.52092-4-tzimmermann@suse.de>
Thank you for your contribution! Sashiko AI review found 2 potential
issues to consider:
- [High] fbcon: invalid kfree() of font_data_t pointer in
fbcon_do_set_font()
- [Low] lib/fonts: contradictory kernel-doc comment for
font_data_rotate()
--
commit b5f0d07cd7c39fad0c72c3148ca9444f977625ca
Author: Thomas Zimmermann <tzimmermann@suse.de>
lib/fonts, fbcon: Use font_data_t for rotated fonts
This commit converts font_data_rotate() and struct fbcon_rotated.buf to
use font_data_t instead of unsigned char *, releasing rotated font data in
fbcon with font_data_put().
> diff --git a/drivers/video/fbdev/core/fbcon.h b/drivers/video/fbdev/core/fbcon.h
> index 289a3d18f917f..28111c706e348 100644
> --- a/drivers/video/fbdev/core/fbcon.h
> +++ b/drivers/video/fbdev/core/fbcon.h
> @@ -75,8 +75,7 @@ struct fbcon_bitops {
> #ifdef CONFIG_FRAMEBUFFER_CONSOLE_ROTATION
> struct fbcon_rotated {
> font_data_t *fontdata; /* source font */
> - unsigned char *buf; /* rotated glyphs */
> - size_t bufsize;
> + font_data_t *buf; /* rotated font */
> int buf_rotate; /* rotation of buf */
> unsigned int buf_width; /* geometry of source font */
> unsigned int buf_height;
[Severity: High]
Does fbcon_do_set_font() still pass struct fbcon_rotated.buf directly to
kfree() when changing the font on a visible rotated console?
drivers/video/fbdev/core/fbcon.c:fbcon_do_set_font() {
...
#ifdef CONFIG_FRAMEBUFFER_CONSOLE_ROTATION
if (rotated_saved)
kfree(saved_rotated.buf);
#endif
...
err_out:
#ifdef CONFIG_FRAMEBUFFER_CONSOLE_ROTATION
if (rotated_saved) {
kfree(par->rotated.buf);
par->rotated = saved_rotated;
}
#endif
...
}
Because __font_data_alloc() returns &font_data->data[0] (offset 16 bytes
past the start of struct font_data), calling kfree() on saved_rotated.buf
or par->rotated.buf frees an interior pointer and corrupts the slab. Both
paths need to check for a non-NULL buffer and release it with
font_data_put() instead.
[ ... ]
> diff --git a/lib/fonts/font_rotate.c b/lib/fonts/font_rotate.c
> index 275406008823b..ea54bd618ffb0 100644
> --- a/lib/fonts/font_rotate.c
> +++ b/lib/fonts/font_rotate.c
> @@ -181,8 +181,7 @@ EXPORT_SYMBOL_GPL(font_glyph_rotate_270);
> * @height: The number of scanlines in the glyph
> * @charcount: The number of glyphs in the font
> * @steps: Number of rotation steps of 90°
> - * @buf: Preallocated output buffer; can be NULL
> - * @bufsize: The size of @buf in bytes; can be NULL
> + * @buf: Preallocated font data; can be NULL
> *
> * The parameters @width and @height refer to the visible number of pixels
> * and scanlines in a single glyph. The number of glyphs is given in @charcount.
> @@ -194,16 +193,14 @@ EXPORT_SYMBOL_GPL(font_glyph_rotate_270);
> * as returned by font_data_buf(). Callers are responsible to free the returned
> * buffer with kfree(). Font rotation typically happens when displays get
> * re-oriented. To avoid unnecessary re-allocation of the memory buffer, the
> - * caller can pass in an earlier result buffer in @buf for reuse. The old and
> - * new buffer sizes are given and retrieved by the caller in @bufsize. The
> - * allocation semantics are compatible with krealloc().
> + * caller can pass in an earlier result in @buf for reuse. The returned font
> + * data should be released with font_data_put().
[Severity: Low]
This isn't a bug, but the earlier sentences in this kernel-doc comment
still say that the returned buffer is not compatible with font_data_t and
must be freed with kfree(), contradicting the new sentence at the end.
--
Sashiko AI review · https://sashiko.dev/#/patchset/20261120115215.52092-1-tzimmermann@suse.de?part=3
next prev parent reply other threads:[~2026-11-20 12:01 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-20 11:49 [PATCH v2 0/6] fbcon: Clean up glyph handling on redraws Thomas Zimmermann
2026-11-20 11:49 ` [PATCH v2 1/6] fbcon: Use unsigned char for glyph data Thomas Zimmermann
2026-11-21 17:37 ` Helge Deller
2026-11-20 11:49 ` [PATCH v2 2/6] lib/fonts: Provide internal helpers for allocating font_data_t Thomas Zimmermann
2026-11-20 11:49 ` [PATCH v2 3/6] lib/fonts, fbcon: Use font_data_t for rotated fonts Thomas Zimmermann
2026-11-20 12:01 ` sashiko-bot [this message]
2026-11-20 11:49 ` [PATCH v2 4/6] fbcon: Refactor size calculation for cursor buffers Thomas Zimmermann
2026-11-20 11:49 ` [PATCH v2 5/6] fbcon: Look up glyphs using the font interface Thomas Zimmermann
2026-11-20 12:22 ` sashiko-bot
2026-11-20 11:49 ` [PATCH v2 6/6] fbcon: Handle characters with no glyph defined in a consistent way Thomas Zimmermann
2026-11-20 12:28 ` sashiko-bot
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=sashiko-outbox-225104@kernel.org \
--to=sashiko-bot@kernel.org \
--cc=dri-devel@lists.freedesktop.org \
--cc=sashiko-reviews@lists.linux.dev \
--cc=tzimmermann@suse.de \
/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