Re: generic crypto - remarks

[Bruno Haible]

Simon Josefsson wrote:
> > you don't need it,
> > because the rules for struct layout in C guarantee that a structure field
> > is aligned to a multiple of the alignment of the previous field.
>
> Are you saying that even if we don't change the type of buffer to
> char, we don't need that alignment?  I would agree, but I'm not
> certain.

If the previous struct element is a 'uint32_t' then a struct element of
type char or char[] has the same alignment. You can rely on it, but it
deserves a comment because it would break if someone inserts, say, a
'short' field in between.

> > 17) gc.m4
> >
> > I don't understand what happens when someone specifies
> > --disable-random-device. Then NAME_OF_RANDOM_DEVICE will be set to "no",
> > the AC_CHECK_FILE will check for a file named "no" and bail out.
>
> Hm.  How can this be solved?  Is AC_ARG_ENABLE the proper method to
> set non-boolean parameters at all?  If it is

Yes it is. Just test the $enableval whether it is "no".

> installed the patch below to gnulib-tool.

Thanks. I updated that to use the recommended autoconf macro argument
quoting style.

Bruno