Bumping the thread... Please take a look below... On Tue, Nov 23, 2021 at 05:24:31PM +0100, Daniel Kiper wrote: > CC-ing Vladimir... > > On Tue, Nov 09, 2021 at 02:34:25PM -0500, Robbie Harwood wrote: > > Paul Eggert <egg...@cs.ucla.edu> writes: > > > On 10/28/21 12:32, Robbie Harwood wrote: > > > > > >> I don't know why Patrick chose to > > >> not use that instead, but a local test seems to work. > > > > > > Is grub2 intended to be portable to compilers that don't support > > > <stdbool.h>? If that's the issue, I suggest that grub2 stop worrying > > > that. Surely every compiler of interest to grub2 supports <stdbool.h> > > > already. And if you really need to support older compilers, the Gnulib > > > stdbool module should suffice. > > > > > >> grub2 shims out config.h for some build targets (e.g., when not building > > >> utilities). > > > > > > Why does it need to do that? Is this because of cross-building, and > > > where <config.h> is for the utilities platform which is not the same as > > > the target platform? If so, that suggests that you should run two > > > 'configure' instances, one for the utilities and one for the target, and > > > compile the base64 module twice if it's used in both places. > > > > I'll defer to Daniel on why things are the way they are, but I don't > > disagree with you. > > Vladimir told me once we are doing that because otherwise we would be > leaking too many "OS specific things" into the GRUB core and modules > which run on top firmware/bare metal instead of the OS. I hope he will > be able to tell us more here... > > Vladimir?
Vladimir, could you give us examples of leaks? Daniel
