Re: [Libguestfs] [libnbd PATCH v3 05/22] states: Prepare to receive 64-bit replies

> > @@ -83,13 +96,18 @@ REPLY.STRUCTURED_REPLY.CHECK: > > * not worth keeping the connection alive. > > */ > > if (length > MAX_REQUEST_SIZE + sizeof h->sbuf.reply.payload.offset_data) { > > - set_error (0, "invalid server reply length %" PRIu32, length); > > + set_error (0, "invalid server reply length %" PRIu64, length); > > SET_NEXT_STATE (%.DEAD); > > return 0; > > } > > + /* For convenience, we now normalize extended replies into compact, > > + * doable since we validated that payload length fits in 32 bits. > > + */ > > + h->sbuf.reply.hdr.structured.length = length; > > (8) I'm very confused by this. For an extended reply, this will > overwrite the "offset" field. At one point, I considered normalizing in the opposite direction (take structured replies and widen them into the extended header form); it required touching more lines of code, so I didn't pursue it at the time. But I can see how compressing things down (intentionally throwing away information we will no longer reference) can be confusing without good comments, so at a minimum, I will be adding comments, if not outright going with the widening rather than narrowing approach.