Fix overflow in EVP_EncodeFinal

https://scan5.scan.coverity.com/#/project-view/60762/10222?selectedIssue=1677829

With recent changes, evp_encodeblock_int may return a negative value,
which EVP_EncodeFinal does not anticipate.  As the latter sets out[ret]
to "\0" where ret is the return value of evp_encodeblock_int, we may
underflow the array index and access invalid memory locations.

Only update the output buffer if the return value is greater or equal to
zero.

Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Nikola Pajkovsky <nikolap@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/29525)
This commit is contained in:
Neil Horman
2025-12-30 14:52:08 -05:00
parent fb99acc994
commit b6aed64e47

View File

@@ -457,10 +457,12 @@ void EVP_EncodeFinal(EVP_ENCODE_CTX *ctx, unsigned char *out, int *outl)
if (ctx->num != 0) {
ret = evp_encodeblock_int(ctx, out, ctx->enc_data, ctx->num,
&wrap_cnt);
if ((ctx->flags & EVP_ENCODE_CTX_NO_NEWLINES) == 0)
out[ret++] = '\n';
out[ret] = '\0';
ctx->num = 0;
if (ossl_assert(ret >= 0)) {
if ((ctx->flags & EVP_ENCODE_CTX_NO_NEWLINES) == 0)
out[ret++] = '\n';
out[ret] = '\0';
ctx->num = 0;
}
}
*outl = ret;
}