From: Ruslan Semchenko <uncleruc2075@gmail.com>
To: linux-kbuild@vger.kernel.org
Cc: nathan@kernel.org, nsc@kernel.org, julianbraha@gmail.com,
masahiroy@kernel.org, linux-kernel@vger.kernel.org,
Ruslan Semchenko <uncleruc2075@gmail.com>
Subject: [PATCH] kconfig: preprocess: fix UAF and SEGV on self-referencing variables
Date: Mon, 2 Nov 2026 12:05:00 +0300 [thread overview]
Message-ID: <20261102090500.54575-1-uncleruc2075@gmail.com> (raw)
In variable_add(), when updating an existing simply expanded variable (:=),
v->value was freed before calling expand_string(value). If the value
referred to the variable itself (e.g. A := $(A)b), variable_expand()
accessed the already freed memory, causing a heap-use-after-free.
Furthermore, for a new variable definition, an uninitialized variable
struct was inserted into variable_list before expanding value. If value
referred to the variable itself (e.g. B := $(B)x), variable_lookup()
found the incomplete variable and dereferenced uninitialized v->value,
causing a segmentation fault.
Fix both issues by computing the new value before modifying the existing
variable or inserting a new one into variable_list.
Signed-off-by: Ruslan Semchenko <uncleruc2075@gmail.com>
---
scripts/kconfig/preprocess.c | 23 ++++++++++++++++-------
1 file changed, 16 insertions(+), 7 deletions(-)
diff --git a/scripts/kconfig/preprocess.c b/scripts/kconfig/preprocess.c
index 783abcaa5..d48791921 100644
--- a/scripts/kconfig/preprocess.c
+++ b/scripts/kconfig/preprocess.c
@@ -293,27 +293,35 @@ void variable_add(const char *name, const char *value,
if (flavor == VAR_APPEND) {
flavor = v->flavor;
append = true;
- } else {
- free(v->value);
}
} else {
/* For undefined variables, += assumes the recursive flavor */
if (flavor == VAR_APPEND)
flavor = VAR_RECURSIVE;
+ }
+
+ /*
+ * Compute the new value before touching the variable. Expanding it may
+ * refer to the variable itself, which must still see the old value.
+ */
+ if (flavor == VAR_SIMPLE)
+ new_value = expand_string(value);
+ else
+ new_value = xstrdup(value);
+ if (!v) {
+ /* A new variable is not visible until it is fully defined. */
v = xmalloc(sizeof(*v));
v->name = xstrdup(name);
+ v->value = new_value;
+ v->flavor = flavor;
v->exp_count = 0;
list_add_tail(&v->node, &variable_list);
+ return;
}
v->flavor = flavor;
- if (flavor == VAR_SIMPLE)
- new_value = expand_string(value);
- else
- new_value = xstrdup(value);
-
if (append) {
v->value = xrealloc(v->value,
strlen(v->value) + strlen(new_value) + 2);
@@ -321,6 +329,7 @@ void variable_add(const char *name, const char *value,
strcat(v->value, new_value);
free(new_value);
} else {
+ free(v->value);
v->value = new_value;
}
}
--
2.47.3
next reply other threads:[~2026-11-02 9:05 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-02 9:05 Ruslan Semchenko [this message]
2026-11-02 13:08 ` Julian Braha
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=20261102090500.54575-1-uncleruc2075@gmail.com \
--to=uncleruc2075@gmail.com \
--cc=julianbraha@gmail.com \
--cc=linux-kbuild@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=masahiroy@kernel.org \
--cc=nathan@kernel.org \
--cc=nsc@kernel.org \
/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
all inboxes | Powered by JetHome®