* [PATCH v3] lib/crypto: chacha20poly1305: zeroize state in __chacha20poly1305_decrypt
@ 2026-10-05 15:03 Mohamad Raizudeen
2026-10-05 15:20 ` Eric Biggers
0 siblings, 1 reply; 2+ messages in thread
From: Mohamad Raizudeen @ 2026-10-05 15:03 UTC (permalink / raw)
To: ebiggers, ardb, Jason
Cc: skhan, me, jkoolstra, linux-crypto, linux-kernel, Mohamad Raizudeen
The `__chacha20poly1305_decrypt` function does not zeroize the chacha
state, unlike its encrypt counterpart. The regular
`chacha20poly1305_decrypt` function handles this by manually calling
chacha_zeroize_state(). However, `xchacha20poly1305_decrypt` returns
the result directly without clearing the state, leaving the derived
chacha20 subkey on the stack.
Because this function is EXPORT_SYMBOL()ed, this is an ABI robustness
improvement. To ensure consistent and safe cleanup for any future
callers, move the chacha_zeroize_state() call into
`__chacha20poly1305_decrypt()` itself, so the helper cleans up after its
own state just like `__chacha20poly1305_encrypt()` does.
Additionally, moved the src_len check to the callers so the state is not
initialized on invalid inputs, matching the logic in
chacha20poly1305_crypt_sg_inplace().
Suggested-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Mohamad Raizudeen <raizudeen.kerneldev@gmail.com>
---
Changes in v3:
- Remove Fixes and Cc: stable tags, as this is an ABI robustness
improvement rather than a fix.
- Move the src_len check to the callers so the state is not initialized
on invalid inputs.
- Fix whitespace and ordering of the chacha_zeroize_state() to match
chacha20poly1305_crypt_sg_inplace().
Link to v2: https://lore.kernel.org/all/asMfBGbckJAhCa45@kernel/T/
lib/crypto/chacha20poly1305.c | 11 +++++++----
1 file changed, 7 insertions(+), 4 deletions(-)
diff --git a/lib/crypto/chacha20poly1305.c b/lib/crypto/chacha20poly1305.c
index ea42a28f4ff7..c3995ac8acb7 100644
--- a/lib/crypto/chacha20poly1305.c
+++ b/lib/crypto/chacha20poly1305.c
@@ -137,9 +137,6 @@ __chacha20poly1305_decrypt(u8 *dst, const u8 *src, const size_t src_len,
__le64 lens[2];
} b;
- if (unlikely(src_len < POLY1305_DIGEST_SIZE))
- return false;
-
chacha20_crypt(chacha_state, b.block0, pad0, sizeof(b.block0));
poly1305_init(&poly1305_state, b.block0);
@@ -162,6 +159,7 @@ __chacha20poly1305_decrypt(u8 *dst, const u8 *src, const size_t src_len,
if (likely(!ret))
chacha20_crypt(chacha_state, dst, src, dst_len);
+ chacha_zeroize_state(chacha_state);
memzero_explicit(&b, sizeof(b));
return !ret;
@@ -177,6 +175,9 @@ bool chacha20poly1305_decrypt(u8 *dst, const u8 *src, const size_t src_len,
__le64 iv[2];
bool ret;
+ if (unlikely(src_len < POLY1305_DIGEST_SIZE))
+ return false;
+
chacha_load_key(k, key);
iv[0] = 0;
@@ -186,7 +187,6 @@ bool chacha20poly1305_decrypt(u8 *dst, const u8 *src, const size_t src_len,
ret = __chacha20poly1305_decrypt(dst, src, src_len, ad, ad_len,
&chacha_state);
- chacha_zeroize_state(&chacha_state);
memzero_explicit(iv, sizeof(iv));
memzero_explicit(k, sizeof(k));
return ret;
@@ -200,6 +200,9 @@ bool xchacha20poly1305_decrypt(u8 *dst, const u8 *src, const size_t src_len,
{
struct chacha_state chacha_state;
+ if (unlikely(src_len < POLY1305_DIGEST_SIZE))
+ return false;
+
xchacha_init(&chacha_state, key, nonce);
return __chacha20poly1305_decrypt(dst, src, src_len, ad, ad_len,
&chacha_state);
--
2.53.0
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [PATCH v3] lib/crypto: chacha20poly1305: zeroize state in __chacha20poly1305_decrypt
2026-10-05 15:03 [PATCH v3] lib/crypto: chacha20poly1305: zeroize state in __chacha20poly1305_decrypt Mohamad Raizudeen
@ 2026-10-05 15:20 ` Eric Biggers
0 siblings, 0 replies; 2+ messages in thread
From: Eric Biggers @ 2026-10-05 15:20 UTC (permalink / raw)
To: Mohamad Raizudeen
Cc: ardb, Jason, skhan, me, jkoolstra, linux-crypto, linux-kernel
On Mon, Oct 05, 2026 at 08:33:24PM +0530, Mohamad Raizudeen wrote:
> The `__chacha20poly1305_decrypt` function does not zeroize the chacha
> state, unlike its encrypt counterpart. The regular
> `chacha20poly1305_decrypt` function handles this by manually calling
> chacha_zeroize_state(). However, `xchacha20poly1305_decrypt` returns
> the result directly without clearing the state, leaving the derived
> chacha20 subkey on the stack.
>
> Because this function is EXPORT_SYMBOL()ed, this is an ABI robustness
> improvement. To ensure consistent and safe cleanup for any future
> callers, move the chacha_zeroize_state() call into
> `__chacha20poly1305_decrypt()` itself, so the helper cleans up after its
> own state just like `__chacha20poly1305_encrypt()` does.
>
> Additionally, moved the src_len check to the callers so the state is not
> initialized on invalid inputs, matching the logic in
> chacha20poly1305_crypt_sg_inplace().
>
> Suggested-by: Ard Biesheuvel <ardb@kernel.org>
> Signed-off-by: Mohamad Raizudeen <raizudeen.kerneldev@gmail.com>
> ---
> Changes in v3:
> - Remove Fixes and Cc: stable tags, as this is an ABI robustness
> improvement rather than a fix.
> - Move the src_len check to the callers so the state is not initialized
> on invalid inputs.
> - Fix whitespace and ordering of the chacha_zeroize_state() to match
> chacha20poly1305_crypt_sg_inplace().
Applied to https://git.kernel.org/pub/scm/linux/kernel/git/ebiggers/linux.git/log/?h=libcrypto-next
- Eric
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2026-10-05 15:20 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-10-05 15:03 [PATCH v3] lib/crypto: chacha20poly1305: zeroize state in __chacha20poly1305_decrypt Mohamad Raizudeen
2026-10-05 15:20 ` Eric Biggers
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®