On Sun, Oct 04, 2026 at 11:27:12PM +0800, Xiaoyi Suo wrote:

> Same as sha1_hmac(): the key schedule and keybuf are locals cleared just
> before the function returns, so those stores are dead too. keybuf holds
> the key, or its digest when the key is longer than a block.
> 
> The memset() that initialises keybuf is not one of the clears, so leave
> it alone.
> 
> Cc: Tom Rini <[email protected]>
> Signed-off-by: Xiaoyi Suo <[email protected]>

Reviewed-by: Tom Rini <[email protected]>

-- 
Tom

Attachment: signature.asc
Description: PGP signature

Reply via email to