On Tue May 21, 2024 at 8:36 AM EEST, Bharat Bhushan wrote:
> > - data_len -= 2;
> > + (*data++) = _tag(UNIV, PRIM, INT);
>
> Just for my clarification:
> First index of "data" is updated here with tag and data pointer
> incremented.
> Next comment for continuation
>
> > + data_len--;
> >
> > - data[0] = _tag(UNIV, PRIM, INT);
> > - if (integer == 0) {
> > - *d++ = 0;
> > - goto out;
> > + if (!memchr_inv(integer, 0, integer_len)) {
> > + data[1] = 1;
> > + data[2] = 0;
> > + return &data[2];
>
> Here we are effectively setting second and third index of original
> "data" pointer as "data" pointer was incremented earlier.
> So second index of original "data" pointer is not touched. Also
> returning 3rd index pointer of original data pointer
>
> Is that intentional?
No! I read the diff few times, and I think you have a point.
Indices should be 0 (length) and 1 (value). I.e. it forms an encoded
version of zero. The last index what it should be, i.e. return address
of the next byte after the encoded integer.
Thanks for pointing this out.
> Thanks
> -Bharat
BR, Jarkko
