Date: Thu, 31 Jul 2025 08:53:16 -0700 From: Rick Macklem <rick.macklem@gmail.com> To: Cy Schubert <Cy.Schubert@cschubert.com> Cc: Benjamin Kaduk <bjkfbsd@gmail.com>, Konstantin Belousov <kostikbel@gmail.com>, Jessica Clarke <jrtc27@freebsd.org>, Cy Schubert <cy@freebsd.org>, "src-committers@freebsd.org" <src-committers@freebsd.org>, "dev-commits-src-all@freebsd.org" <dev-commits-src-all@freebsd.org>, "dev-commits-src-main@freebsd.org" <dev-commits-src-main@freebsd.org> Subject: Re: git: c7da9fb90b0b - main - KRB5: Enable MIT KRB5 by default Message-ID: <CAM5tNy4N301K4SK-Ow%2BT_ojg_xKfy2hy2TtNmLHDQVTse_VUyA@mail.gmail.com> In-Reply-To: <20250731152803.19F9ACD@slippy.cwsent.com> References: <202507211410.56LEAD6J066633@gitrepo.freebsd.org> <47C3CC37-6F32-4376-900A-B5387B9817D5@freebsd.org> <20250721144645.3BA391BE@slippy.cwsent.com> <aH98iNXobigu39On@kib.kiev.ua> <20250722155941.AC7EB121@slippy.cwsent.com> <CAM5tNy63Ri73x3ByJUPFh7a0eCVjWPGW1hQwrkG0wz6pJ6-W3Q@mail.gmail.com> <aIcyq6JuYngAm4Ko@kib.kiev.ua> <CAM5tNy6pAUS1HD4W1=rxLXvfctAs1Ms_fxwUWz1X3tFNuVTVZg@mail.gmail.com> <CAJ5_RoBjpX_BoKgZTdOof5b83cdxjZyvn4iPM4m5voPHEr2SeQ@mail.gmail.com> <CAJ5_RoBXD4rkudaqo4BA%2BrggYcy%2BLUSnPrK2_FO4Cp_xG0fS=Q@mail.gmail.com> <CAM5tNy5DjktV54JaNcXSwc=9W9kyKTbYANyWp3wmgWu1aMvmLg@mail.gmail.com> <CAJ5_RoDdhmpsLWtuCiaWJoR1yMvAq_4r-E%2BpUk03TiQofTZELQ@mail.gmail.com> <CAM5tNy6Va=edwT5TTa8NEtCHrpbAnN1NMDsBKNNbqNgPoCjVng@mail.gmail.com> <CAM5tNy78DEgmx4z=RJbMsBkRuO6sX%2BhK9OfKbg4ACcRMnrOQgA@mail.gmail.com> <CAM5tNy4XupPGXHMS0p0TK0Wf_zAg5bsOzx4C1K1e-_2b=3eyUw@mail.gmail.com> <20250731152803.19F9ACD@slippy.cwsent.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Jul 31, 2025 at 8:28=E2=80=AFAM Cy Schubert <Cy.Schubert@cschubert.=
com> wrote:
>
> In message <CAM5tNy4XupPGXHMS0p0TK0Wf_zAg5bsOzx4C1K1e-_2b=3D3eyUw@mail.gm=
ail.c
> om>
> , Rick Macklem writes:
> > On Thu, Jul 31, 2025 at 6:51=3DE2=3D80=3DAFAM Rick Macklem <rick.mackle=
m@gmail.co=3D
> > m> wrote:
> > >
> > > On Thu, Jul 31, 2025 at 6:07=3DE2=3D80=3DAFAM Rick Macklem <rick.mack=
lem@gmail.=3D
> > com> wrote:
> > > >
> > > > On Wed, Jul 30, 2025 at 9:24=3DE2=3D80=3DAFPM Benjamin Kaduk <bjkfb=
sd@gmail.c=3D
> > om> wrote:
> > > > >
> > > > > On Wed, Jul 30, 2025 at 10:36=3DE2=3D80=3DAFAM Rick Macklem <rick=
.macklem@g=3D
> > mail.com> wrote:
> > > > >>
> > > > >> On Mon, Jul 28, 2025 at 3:32=3DE2=3D80=3DAFPM Benjamin Kaduk <bj=
kfbsd@gmai=3D
> > l.com> wrote:
> > > > >> >
> > > > >> > On Mon, Jul 28, 2025 at 3:04=3DE2=3D80=3DAFPM Benjamin Kaduk <=
bjkfbsd@gm=3D
> > ail.com> wrote:
> > > > >> >>
> > > > >> >>
> > > > >> >> Note that MIT krb5 provides the gss_krb5_export_lucid_sec_con=
text=3D
> > () API that does a lot of the work of getting useful bits out of an est=
abli=3D
> > shed GSS security context.
> > > > >> >>
> > > > >> >
> > > > >> >
> > > > >> >
> > > > >> >
> > > > >> > And a bit more context on what is going on here and why kgssap=
i ha=3D
> > s to care:
> > > > >> > The GSS-API (RFC 2743) is all about a way to "establish a secu=
rity=3D
> > context" (i.e., do crypto negotiation, authentication, sometimes autho=
riza=3D
> > tion, etc.) between two entities, the initiator and the acceprot, and t=
hen =3D
> > exchanging protected messages between the two (which can be either encr=
ypte=3D
> > d or just integrity protection tags for otherweise cleartext data); lat=
er e=3D
> > xtensions included the ability to produce identical PRF output on both =
part=3D
> > ies, etc.. The details are "mechanism-specific", and for this purpose =
we'r=3D
> > e exclusively talking about the krb5 mechanism. The steps to establish=
the=3D
> > security context are complicated and sometimes fiddly, and in the gene=
ral =3D
> > case can require a large number of round-trips between the initiator an=
d ac=3D
> > ceptor before the security context is established. The individual mess=
age-=3D
> > protection parts are comparatively simple and amendable to implementati=
on i=3D
> > n the kernel for processing efficiency.
> > > > >> > RFC 2743 also defines functions for GSS_Export_sec_context() a=
nd G=3D
> > SS_Import_sec_context(), that are designed essentially to pass informat=
ion =3D
> > about an established security context from one process to another on th=
e sa=3D
> > me machine (which are presumably using the same implementation and vers=
ion =3D
> > of the implementation), so the contents of the exported blob are opaque=
and=3D
> > implementation-specific. We are abusing that mechanism to export info=
rmat=3D
> > ion about the security context that gssd has established and feed that =
info=3D
> > rmation into the kernel implementation of the per-message processing ro=
utin=3D
> > es. At present, this necessarily entails knowing the details of the im=
plem=3D
> > entation-specific opaque blob that is the "export sec context token", w=
hich=3D
> > is what the sys/kgssapi/krb5/krb5_mech.c code is doing. But if we can=
get=3D
> > the information we want without breaking the abstraction barrier, such=
as =3D
> > via the gss_krb5_export_lucid_sec_context() API, we are in a more robus=
t po=3D
> > sture overall and somewhat future-proofed against future evolution by M=
IT k=3D
> > rb5.
> > > > >> > (I note that recent Heimdal versions seem to also expose a gss=
_krb=3D
> > 5_export_lucid_sec_context() API, so part of the problem is just that t=
he H=3D
> > eimdal in base is so old.)
> > > > >>
> > > > >> Well, here's some "not so good" news...
> > > > >> I've been trying to use gss_inquire_sec_context_by_oid(..) with =
the =3D
> > oid
> > > > >> for the GSS_KRB5_EXPORT_LUCID_SEC_CONTEXT_OID with version 1.
> > > > >> It kept failing.
> > > > >> The problem seems to be that "gctx->proto =3D3D=3D3D 4" in make_=
external=3D
> > _lucid_ctx_v1()
> > > > >> function. This function only knows about the 0 and 1 setting for=
gct=3D
> > x->proto.
> > > > >>
> > > > >> Any ideas, rick
> > > > >>
> > > > >
> > > > >
> > > > > I'm not seeing anything to suggest that a "gctx->proto" value of =
4 is=3D
> > ever expected; it looks like it's supposed to just be 0 (for the legac=
y RF=3D
> > C 1964 format) or 1 (for the "CFX" format of RFC 4121, with wider seque=
nce =3D
> > numbers for message-protection formats, etc.). So maybe it's worth pos=
ting=3D
> > your current WIP somewhere to take a closer look at what's going on.
> > > >
> > > > Yea, the debugging I did was flawed (I probably got the wrong offse=
t
> > > > in the structure).
> > > > It is weird, though. If I do gss_inquire_sec_context_by_oid(&minor,=
ctx=3D
> > ,
> > > > OID_FOR_GSS_INQUIRE_SSPI_SESSION_KEY, &key), it
> > > > works and gives me the key and encryption type.
> > > >
> > > > If I do the same, but with the 12 byte OID for LUCID v1 (the 11 byt=
es f=3D
> > rom the
> > > > string + a 1 byte), it returns major =3D3D=3D3D GSS_S_COMPLETE, but=
no data=3D
> > and
> > > > a weird 39756046(decimal) or 0x25ea10e(hex) minor.
> > > > (Oh, and I tried gss_krb5_export_lucid_sec_context() and got the sa=
me
> > > > weird error.)
> > > --> Now (after doing a "make buildworld"), gss_krb5_export_lucid_sec_=
cont=3D
> > ext()
> > > returns GSS_S_BAD_MECH. Looking at the src, that error has to be=
fro=3D
> > m
> > > gss_inquire_sec_context_by_oid(). So, same function fails, but a=
dif=3D
> > ferent
> > > error return?
> > >
> > > It looks like "gssint_get_mechanism (ctx->mech_type)" is failing.
> > > I'm currently just passing GSS_C_NULL_OID into gss_init_sec_context()=
,
> > > but I've also tried the Kerberos 9 byte OID (both work, in the sense =
that
> > > gss_init_sec_context() seems to work, except that the actual_mech_typ=
e
> > > returned by it has a bogus pointer in the reply).
> > > --> It looks like the "mech_type" field of "ctx" is busted, for some =
reas=3D
> > on?
> > >
> > > I'm going to try building krb5 from ports and linking to that, to see=
if =3D
> > it
> > > does the same thing.
> > Finally some good news...
> > All I did was "pkg install krb5" and then linked the gssd to the librar=
ies =3D
> > in
> > /usr/local/lib and it worked!!
>
> gssapi/gssapi.h from krb5/lib/gssapi/generic is overwritten by our
> lib/libgssapi. As we have two the MIT gssapi.h is put in
> /usr/include/gssapi_krb5/gssapi.h.
>
> This patch should fix the problem. I haven't tested this yet.
>
> diff --git a/usr.sbin/gssd/Makefile b/usr.sbin/gssd/Makefile
> index 569e2c7e18f5..4c9d342c48c3 100644
> --- a/usr.sbin/gssd/Makefile
> +++ b/usr.sbin/gssd/Makefile
> @@ -14,7 +14,7 @@ LIBADD=3D gssapi
> .if ${MK_MITKRB5} !=3D "no"
> # MIT KRB5
> LIBADD+=3D krb5 k5crypto krb5profile krb5support
> -CFLAGS+=3D -DMK_MITKRB5=3Dyes
> +CFLAGS+=3D -DMK_MITKRB5=3Dyes -Iinclude/gssapi_krb5
> .else
> # Heimdal
> LIBADD+=3D krb5 roken
Just to be clear to everyone, this might allow it to be built after
being patched for MIT, but it does not fix it so that it works.
I will be debugging the patches that makes it works later to-day.
You state that Heimdal didn't have a gssapi.h, but it does and it
has always been included in gssd.c. (It was the other ones like
gssapi_krb5.h, which needs the MIT gssapi.h.)
rick
>
>
> >
> > Now I can test/debug the changes.
> >
> > Btw, the stuff in /usr/local/include/gssapi are correct and not messed =
up
> > like the stuff in /usr/include/gssapi. (The ones in /usr/local/include =
defi=3D
> > ne
> > GSS_DLLIMP for example.)
> >
> > I'm going to leave figuring out why the libraries in /usr/lib are messe=
d up
> > to someone else.
> >
> > rick
> >
>
>
> --
> Cheers,
> Cy Schubert <Cy.Schubert@cschubert.com>
> FreeBSD UNIX: <cy@FreeBSD.org> Web: https://FreeBSD.org
> NTP: <cy@nwtime.org> Web: https://nwtime.org
>
> e**(i*pi)+1=3D0
>
>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAM5tNy4N301K4SK-Ow%2BT_ojg_xKfy2hy2TtNmLHDQVTse_VUyA>