On 03/21/18 07:20 AM, Boris Pismenny wrote:
>
>
> On 3/20/2018 7:54 PM, Dave Watson wrote:
> > + ctx->control = header[0];
> > +
> > + data_len = ((header[4] & 0xFF) | (header[3] << 8));
> > +
> > + cipher_overhead = tls_ctx->rx.tag_size + tls_ctx->rx.iv_size;
> > +
> > + if (data_len > TLS_MAX_PAYLOAD_SIZE + cipher_overhead) {
> > + ret = -EMSGSIZE;
> > + goto read_failure;
> > + }
> > + if (data_len < cipher_overhead) {
> > + ret = -EMSGSIZE;
>
> I think this should be considered EBADMSG, because this error is cipher
> dependent. At least, that's what happens within OpenSSL. Also, EMSGSIZE is
> usually used only for too long messages.
Ah, indeed. Thanks, will send v2.
