Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Peter Korsgaard <peter@korsgaard.com>
To: buildroot@buildroot.org
Subject: [Buildroot] [PATCH] boot/shim: add patch to fix builds with -O3 / -Ofast
Date: Mon, 15 Jun 2026 22:25:31 +0200	[thread overview]
Message-ID: <20260615202534.1281579-1-peter@korsgaard.com> (raw)

Fixes:
https://autobuild.buildroot.net/results/c6ad52a048ba030bb1adc3ee1aeb69ef880cbdad/
https://autobuild.buildroot.net/results/136d1afc2ecacea6a4e12b915073e3eaf5c8ca01/
https://autobuild.buildroot.net/results/a3f293371f6bd0125dc7baa5afb64373e4383f4f/

Shim contains an embedded copy of OpenSSL.  Depending on the optimization
level, an error is triggered from make_kn():

In function 'make_kn',
    inlined from 'make_kn' at crypto/cmac/cmac.c:81:13,
    inlined from 'CMAC_Init' at crypto/cmac/cmac.c:205:9:
crypto/cmac/cmac.c:92:20: error: writing 1 byte into a region of size 0 [-Werror=stringop-overflow=]
   92 |         k1[bl - 1] ^= bl == 16 ? 0x87 : 0x1b;
      |         ~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~
crypto/cmac/cmac.c: In function 'CMAC_Init':
crypto/cmac/cmac.c:69:19: note: at offset 2147483647 into destination object 'k1' of size 32
   69 |     unsigned char k1[EVP_MAX_BLOCK_LENGTH];
      |                   ^~

Fix it by including a patch submitted upstream which backports an upstream
OpenSSL change to workaround this.

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
---
 ...make_kn-and-move-zero_iv-to-const-se.patch | 72 +++++++++++++++++++
 1 file changed, 72 insertions(+)
 create mode 100644 boot/shim/0002-cmac.c-optimize-make_kn-and-move-zero_iv-to-const-se.patch

diff --git a/boot/shim/0002-cmac.c-optimize-make_kn-and-move-zero_iv-to-const-se.patch b/boot/shim/0002-cmac.c-optimize-make_kn-and-move-zero_iv-to-const-se.patch
new file mode 100644
index 0000000000..2baa0de56c
--- /dev/null
+++ b/boot/shim/0002-cmac.c-optimize-make_kn-and-move-zero_iv-to-const-se.patch
@@ -0,0 +1,72 @@
+From 7aee873d6f6b0737dfe566fc8c5cd4eaada9bd5e Mon Sep 17 00:00:00 2001
+From: Fabrice Fontaine <fontaine.fabrice@gmail.com>
+Date: Mon, 8 Apr 2024 11:56:06 +0200
+Subject: [PATCH] cmac.c: optimize make_kn and move zero_iv to const segment.
+
+Backport
+https://github.com/openssl/openssl/commit/03cf7e784caa4c61febbf249be63cbae3e368ac9
+to fix the following k1 stringop-overflow:
+
+In function 'make_kn',
+    inlined from 'make_kn' at crypto/cmac/cmac.c:81:13,
+    inlined from 'CMAC_Init' at crypto/cmac/cmac.c:205:9:
+crypto/cmac/cmac.c:92:20: error: writing 1 byte into a region of size 0 [-Werror=stringop-overflow=]
+   92 |         k1[bl - 1] ^= bl == 16 ? 0x87 : 0x1b;
+      |         ~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~
+crypto/cmac/cmac.c: In function 'CMAC_Init':
+crypto/cmac/cmac.c:69:19: note: at offset [-2147483649, -1] into destination object 'k1' of size 32
+   69 |     unsigned char k1[EVP_MAX_BLOCK_LENGTH];
+      |                   ^~
+
+Fixes:
+ - http://autobuild.buildroot.org/results/97b6333cdc7bad24aba7af1b04890679e0058299
+
+Upstream: https://github.com/rhboot/shim/pull/652
+Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
+Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
+---
+ Cryptlib/OpenSSL/crypto/cmac/cmac.c | 17 ++++++++---------
+ 1 file changed, 8 insertions(+), 9 deletions(-)
+
+diff --git a/Cryptlib/OpenSSL/crypto/cmac/cmac.c b/Cryptlib/OpenSSL/crypto/cmac/cmac.c
+index 2954b6eb..27dbfaac 100644
+--- a/Cryptlib/OpenSSL/crypto/cmac/cmac.c
++++ b/Cryptlib/OpenSSL/crypto/cmac/cmac.c
+@@ -78,18 +78,17 @@ struct CMAC_CTX_st {
+ 
+ /* Make temporary keys K1 and K2 */
+ 
+-static void make_kn(unsigned char *k1, unsigned char *l, int bl)
++static void make_kn(unsigned char *k1, const unsigned char *l, int bl)
+ {
+     int i;
++    unsigned char c = l[0], carry = c>>7, cnext;
++
+     /* Shift block to left, including carry */
+-    for (i = 0; i < bl; i++) {
+-        k1[i] = l[i] << 1;
+-        if (i < bl - 1 && l[i + 1] & 0x80)
+-            k1[i] |= 1;
+-    }
++    for (i = 0; i < bl-1; i++, c = cnext)
++        k1[i] = (c << 1) | ((cnext=l[i+1]) >> 7);
++
+     /* If MSB set fixup with R */
+-    if (l[0] & 0x80)
+-        k1[bl - 1] ^= bl == 16 ? 0x87 : 0x1b;
++    k1[i] = (c << 1) ^ ((0-carry)&(bl==16?0x87:0x1b));
+ }
+ 
+ CMAC_CTX *CMAC_CTX_new(void)
+@@ -151,7 +150,7 @@ int CMAC_CTX_copy(CMAC_CTX *out, const CMAC_CTX *in)
+ int CMAC_Init(CMAC_CTX *ctx, const void *key, size_t keylen,
+               const EVP_CIPHER *cipher, ENGINE *impl)
+ {
+-    static unsigned char zero_iv[EVP_MAX_BLOCK_LENGTH];
++    static const unsigned char zero_iv[EVP_MAX_BLOCK_LENGTH] = {0};
+ #ifdef OPENSSL_FIPS
+     if (FIPS_mode()) {
+         /* If we have an ENGINE need to allow non FIPS */
+-- 
+2.47.3
+
-- 
2.47.3

_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

             reply	other threads:[~2026-06-15 20:25 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-06-15 20:25 Peter Korsgaard [this message]
2026-06-15 20:54 ` [Buildroot] [PATCH] boot/shim: add patch to fix builds with -O3 / -Ofast Julien Olivain via buildroot

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=20260615202534.1281579-1-peter@korsgaard.com \
    --to=peter@korsgaard.com \
    --cc=buildroot@buildroot.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