Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 15 Aug 2022 18:59:48 +0000
From:      bugzilla-noreply@freebsd.org
To:        bugs@FreeBSD.org
Subject:   [Bug 265852] qat: kernel crash using krb5p with nfs
Message-ID:  <bug-265852-227-aKdAZ6e4J0@https.bugs.freebsd.org/bugzilla/>
In-Reply-To: <bug-265852-227@https.bugs.freebsd.org/bugzilla/>
References:  <bug-265852-227@https.bugs.freebsd.org/bugzilla/>

next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D265852

Mark Johnston <markj@FreeBSD.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |markj@FreeBSD.org,
                   |                            |mav@FreeBSD.org
             Status|New                         |Open

--- Comment #1 from Mark Johnston <markj@FreeBSD.org> ---
I believe this is a bug in kgssapi code.  aes_checksum() and aes_encrypt_1()
can observe that CRYPTO_F_DONE is set before the callback worker thread has=
 had
a chance to invoke the callback.  Then the code frees the request, i.e., th=
ere
is a use-after-free.

kgssapi is ok with synchronous implementations of crypto routines, but not =
with
async implementations like QAT.  So the workaround for now is simply to not=
 use
QAT.  I can provide a patch for testing, but I can't provide a truenas buil=
d.

--=20
You are receiving this mail because:
You are the assignee for the bug.=



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-265852-227-aKdAZ6e4J0>