"Daniel P. Berrange" <berra...@redhat.com> writes: > On Fri, Nov 20, 2015 at 03:09:25PM -0700, Eric Blake wrote: >> On 11/20/2015 11:04 AM, Daniel P. Berrange wrote: >> > + >> > +static const char *base64_valid_chars = >> > + "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/="; >> > + >> > +static int >> > +qcrypto_secret_validate_base64(const uint8_t *input, >> > + size_t inputlen, >> > + Error **errp) >> >> Don't we already have base64 utility methods available? > > We normally use glib, g_base64_encode/decode. Unfortunately the > decode method doesn't provide any usefull error reporting facility. > It just silently skips any characters that are outside the valid > set. So the only way I could get any kind of sensible error report > was to do this validation myself against the set of permitted base64 > characters.
Yes. Same problem elsewhere, e.g. ringbuf-write. qapi-schema.json: # - base64: data must be base64 encoded text. Its binary # decoding gets written. # Bug: invalid base64 is currently not rejected. # Whitespace *is* invalid. This suggests that we shouldn't bury this in crypto/, but instead add it to util/. A replacement for g_base64_decode() could be easier to use than a checker function to use in addition to g_base64_decode(), [...]
