Re: OpenSSL legacy provider is broken
- Reply: Lexi Winter : "Re: OpenSSL legacy provider is broken"
- In reply to: Enji Cooper (yaneurabeya): "Re: OpenSSL legacy provider is broken"
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Sun, 10 Aug 2025 06:06:06 UTC
In message <B43DA54A-0017-42CA-A1FE-15F28048FEF0@gmail.com>, "Enji Cooper (yane urabeya)" writes: > > --Apple-Mail=_5B6A4863-E7AA-415B-BCF7-22A067F7F7C6 > Content-Transfer-Encoding: quoted-printable > Content-Type: text/plain; > charset=utf-8 > > > > On Aug 9, 2025, at 7:08=E2=80=AFAM, Ian FREISLICH = > <ianfreislich@gmail.com> wrote: > >=20 > > Hi > >=20 > > Previously this worked > >=20 > > [brane] /usr/ports # openssl list -providers -provider legacy > > Providers: > > legacy > > name: OpenSSL Legacy Provider > > version: 3.0.16 > > status: active > >=20 > > Since the build last night, > >=20 > > [router] /usr/ports/net/freeradius3 # openssl list -providers = > -provider legacy > > list: unable to load provider legacy > > Hint: use -provider-path option or OPENSSL_MODULES environment = > variable. > > 10B045DBE7340000:error:12800067:DSO support routines:dlfcn_load:could = > not load the shared = > library:/usr/src/crypto/openssl/crypto/dso/dso_dlfcn.c:118:filename(/usr/l= > ib/ossl-modules/legacy.so): /usr/lib/ossl-modules/legacy.so: Undefined = > symbol "ossl_kdf_pvk_functions" > > 10B045DBE7340000:error:12800067:DSO support routines:DSO_load:could = > not load the shared = > library:/usr/src/crypto/openssl/crypto/dso/dso_lib.c:147: > > 10B045DBE7340000:error:07880025:common libcrypto = > routines:provider_init:reason(37):/usr/src/crypto/openssl/crypto/provider_= > core.c:1019:name=3Dlegacy > >=20 > > and freeradius doesn't start because of this: > >=20 > > [router] /usr/ports/net/freeradius3 # radiusd -fX > > FreeRADIUS Version 3.2.7 > > ... > > (TLS) Failed loading legacy provider > >=20 > > I haven't yet figured out what part of my EAP configuration needs the = > legacy provider. It may be that EAP just needs a working legacy provider = > because it looks like the EAP module unconditionally attempts to load = > the provider and fails. This looks like it's due to MIT KRB5 in 15. The HEIMDAL option needs to be replaced with a BASE option and BASE should test for MIT KRB5 or HEIMDAL by looking at which kdc (kdc for Heimdal or krb5kdc for MIT) is installed. -- 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=0