2002-10-04 20:31:23 +00:00
|
|
|
/* $OpenBSD: criov.c,v 1.9 2002/01/29 15:48:29 jason Exp $ */
|
|
|
|
|
2005-01-07 02:29:27 +00:00
|
|
|
/*-
|
2002-10-04 20:31:23 +00:00
|
|
|
* Copyright (c) 1999 Theo de Raadt
|
|
|
|
*
|
|
|
|
* Redistribution and use in source and binary forms, with or without
|
|
|
|
* modification, are permitted provided that the following conditions
|
|
|
|
* are met:
|
|
|
|
*
|
|
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer.
|
|
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
|
|
* documentation and/or other materials provided with the distribution.
|
|
|
|
* 3. The name of the author may not be used to endorse or promote products
|
|
|
|
* derived from this software without specific prior written permission.
|
|
|
|
*
|
|
|
|
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
|
|
|
|
* IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
|
|
|
|
* OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
|
|
|
|
* IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
|
|
|
|
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
|
|
|
|
* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
|
|
|
|
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
|
|
|
|
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
|
|
|
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
|
|
|
|
* THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
|
|
|
*/
|
|
|
|
|
2003-06-11 05:57:50 +00:00
|
|
|
#include <sys/cdefs.h>
|
|
|
|
__FBSDID("$FreeBSD$");
|
|
|
|
|
2002-10-04 20:31:23 +00:00
|
|
|
#include <sys/param.h>
|
|
|
|
#include <sys/systm.h>
|
|
|
|
#include <sys/proc.h>
|
|
|
|
#include <sys/errno.h>
|
|
|
|
#include <sys/malloc.h>
|
|
|
|
#include <sys/kernel.h>
|
2006-06-04 22:15:13 +00:00
|
|
|
#include <sys/mbuf.h>
|
2002-10-04 20:31:23 +00:00
|
|
|
#include <sys/uio.h>
|
2018-12-13 04:40:53 +00:00
|
|
|
#include <sys/limits.h>
|
|
|
|
#include <sys/lock.h>
|
2002-10-04 20:31:23 +00:00
|
|
|
|
|
|
|
#include <opencrypto/cryptodev.h>
|
|
|
|
|
2006-05-17 17:56:00 +00:00
|
|
|
/*
|
|
|
|
* This macro is only for avoiding code duplication, as we need to skip
|
|
|
|
* given number of bytes in the same way in three functions below.
|
|
|
|
*/
|
|
|
|
#define CUIO_SKIP() do { \
|
|
|
|
KASSERT(off >= 0, ("%s: off %d < 0", __func__, off)); \
|
|
|
|
KASSERT(len >= 0, ("%s: len %d < 0", __func__, len)); \
|
|
|
|
while (off > 0) { \
|
|
|
|
KASSERT(iol >= 0, ("%s: empty in skip", __func__)); \
|
|
|
|
if (off < iov->iov_len) \
|
|
|
|
break; \
|
|
|
|
off -= iov->iov_len; \
|
|
|
|
iol--; \
|
|
|
|
iov++; \
|
|
|
|
} \
|
|
|
|
} while (0)
|
|
|
|
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
static void
|
2002-10-04 20:31:23 +00:00
|
|
|
cuio_copydata(struct uio* uio, int off, int len, caddr_t cp)
|
|
|
|
{
|
|
|
|
struct iovec *iov = uio->uio_iov;
|
|
|
|
int iol = uio->uio_iovcnt;
|
|
|
|
unsigned count;
|
|
|
|
|
2006-05-17 17:56:00 +00:00
|
|
|
CUIO_SKIP();
|
2002-10-04 20:31:23 +00:00
|
|
|
while (len > 0) {
|
2006-05-17 17:56:00 +00:00
|
|
|
KASSERT(iol >= 0, ("%s: empty", __func__));
|
2002-10-04 20:31:23 +00:00
|
|
|
count = min(iov->iov_len - off, len);
|
|
|
|
bcopy(((caddr_t)iov->iov_base) + off, cp, count);
|
|
|
|
len -= count;
|
|
|
|
cp += count;
|
|
|
|
off = 0;
|
|
|
|
iol--;
|
|
|
|
iov++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
static void
|
2017-05-09 18:28:42 +00:00
|
|
|
cuio_copyback(struct uio* uio, int off, int len, c_caddr_t cp)
|
2002-10-04 20:31:23 +00:00
|
|
|
{
|
|
|
|
struct iovec *iov = uio->uio_iov;
|
|
|
|
int iol = uio->uio_iovcnt;
|
|
|
|
unsigned count;
|
|
|
|
|
2006-05-17 17:56:00 +00:00
|
|
|
CUIO_SKIP();
|
2002-10-04 20:31:23 +00:00
|
|
|
while (len > 0) {
|
2006-05-17 17:56:00 +00:00
|
|
|
KASSERT(iol >= 0, ("%s: empty", __func__));
|
2002-10-04 20:31:23 +00:00
|
|
|
count = min(iov->iov_len - off, len);
|
|
|
|
bcopy(cp, ((caddr_t)iov->iov_base) + off, count);
|
|
|
|
len -= count;
|
|
|
|
cp += count;
|
|
|
|
off = 0;
|
|
|
|
iol--;
|
|
|
|
iov++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2014-12-12 19:56:36 +00:00
|
|
|
* Return the index and offset of location in iovec list.
|
2002-10-04 20:31:23 +00:00
|
|
|
*/
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
static int
|
2002-10-04 20:31:23 +00:00
|
|
|
cuio_getptr(struct uio *uio, int loc, int *off)
|
|
|
|
{
|
2014-12-12 19:56:36 +00:00
|
|
|
int ind, len;
|
2002-10-04 20:31:23 +00:00
|
|
|
|
2014-12-12 19:56:36 +00:00
|
|
|
ind = 0;
|
|
|
|
while (loc >= 0 && ind < uio->uio_iovcnt) {
|
|
|
|
len = uio->uio_iov[ind].iov_len;
|
|
|
|
if (len > loc) {
|
2002-10-04 20:31:23 +00:00
|
|
|
*off = loc;
|
2014-12-12 19:56:36 +00:00
|
|
|
return (ind);
|
2002-10-04 20:31:23 +00:00
|
|
|
}
|
2014-12-12 19:56:36 +00:00
|
|
|
loc -= len;
|
|
|
|
ind++;
|
|
|
|
}
|
2002-10-04 20:31:23 +00:00
|
|
|
|
2014-12-12 19:56:36 +00:00
|
|
|
if (ind > 0 && loc == 0) {
|
|
|
|
ind--;
|
|
|
|
*off = uio->uio_iov[ind].iov_len;
|
|
|
|
return (ind);
|
|
|
|
}
|
2002-10-04 20:31:23 +00:00
|
|
|
|
2014-12-12 19:56:36 +00:00
|
|
|
return (-1);
|
2002-10-04 20:31:23 +00:00
|
|
|
}
|
2006-05-17 17:56:00 +00:00
|
|
|
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
void
|
|
|
|
crypto_cursor_init(struct crypto_buffer_cursor *cc,
|
|
|
|
const struct crypto_buffer *cb)
|
|
|
|
{
|
|
|
|
memset(cc, 0, sizeof(*cc));
|
|
|
|
cc->cc_type = cb->cb_type;
|
|
|
|
switch (cc->cc_type) {
|
|
|
|
case CRYPTO_BUF_CONTIG:
|
|
|
|
cc->cc_buf = cb->cb_buf;
|
|
|
|
cc->cc_buf_len = cb->cb_buf_len;
|
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_MBUF:
|
|
|
|
cc->cc_mbuf = cb->cb_mbuf;
|
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_UIO:
|
|
|
|
cc->cc_iov = cb->cb_uio->uio_iov;
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
#ifdef INVARIANTS
|
|
|
|
panic("%s: invalid buffer type %d", __func__, cb->cb_type);
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
crypto_cursor_advance(struct crypto_buffer_cursor *cc, size_t amount)
|
|
|
|
{
|
|
|
|
size_t remain;
|
|
|
|
|
|
|
|
switch (cc->cc_type) {
|
|
|
|
case CRYPTO_BUF_CONTIG:
|
|
|
|
MPASS(cc->cc_buf_len >= amount);
|
|
|
|
cc->cc_buf += amount;
|
|
|
|
cc->cc_buf_len -= amount;
|
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_MBUF:
|
|
|
|
for (;;) {
|
|
|
|
remain = cc->cc_mbuf->m_len - cc->cc_offset;
|
|
|
|
if (amount < remain) {
|
|
|
|
cc->cc_offset += amount;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
amount -= remain;
|
|
|
|
cc->cc_mbuf = cc->cc_mbuf->m_next;
|
|
|
|
cc->cc_offset = 0;
|
|
|
|
if (amount == 0)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_UIO:
|
|
|
|
for (;;) {
|
|
|
|
remain = cc->cc_iov->iov_len - cc->cc_offset;
|
|
|
|
if (amount < remain) {
|
|
|
|
cc->cc_offset += amount;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
amount -= remain;
|
|
|
|
cc->cc_iov++;
|
|
|
|
cc->cc_offset = 0;
|
|
|
|
if (amount == 0)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
#ifdef INVARIANTS
|
|
|
|
panic("%s: invalid buffer type %d", __func__, cc->cc_type);
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void *
|
|
|
|
crypto_cursor_segbase(struct crypto_buffer_cursor *cc)
|
|
|
|
{
|
|
|
|
switch (cc->cc_type) {
|
|
|
|
case CRYPTO_BUF_CONTIG:
|
|
|
|
return (cc->cc_buf);
|
|
|
|
case CRYPTO_BUF_MBUF:
|
|
|
|
if (cc->cc_mbuf == NULL)
|
|
|
|
return (NULL);
|
|
|
|
KASSERT((cc->cc_mbuf->m_flags & M_EXTPG) == 0,
|
|
|
|
("%s: not supported for unmapped mbufs", __func__));
|
|
|
|
return (mtod(cc->cc_mbuf, char *) + cc->cc_offset);
|
|
|
|
case CRYPTO_BUF_UIO:
|
|
|
|
return ((char *)cc->cc_iov->iov_base + cc->cc_offset);
|
|
|
|
default:
|
|
|
|
#ifdef INVARIANTS
|
|
|
|
panic("%s: invalid buffer type %d", __func__, cc->cc_type);
|
|
|
|
#endif
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
size_t
|
|
|
|
crypto_cursor_seglen(struct crypto_buffer_cursor *cc)
|
|
|
|
{
|
|
|
|
switch (cc->cc_type) {
|
|
|
|
case CRYPTO_BUF_CONTIG:
|
|
|
|
return (cc->cc_buf_len);
|
|
|
|
case CRYPTO_BUF_MBUF:
|
|
|
|
if (cc->cc_mbuf == NULL)
|
|
|
|
return (0);
|
|
|
|
return (cc->cc_mbuf->m_len - cc->cc_offset);
|
|
|
|
case CRYPTO_BUF_UIO:
|
|
|
|
return (cc->cc_iov->iov_len - cc->cc_offset);
|
|
|
|
default:
|
|
|
|
#ifdef INVARIANTS
|
|
|
|
panic("%s: invalid buffer type %d", __func__, cc->cc_type);
|
|
|
|
#endif
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
crypto_cursor_copyback(struct crypto_buffer_cursor *cc, int size,
|
|
|
|
const void *vsrc)
|
|
|
|
{
|
|
|
|
size_t remain, todo;
|
|
|
|
const char *src;
|
|
|
|
char *dst;
|
|
|
|
|
|
|
|
src = vsrc;
|
|
|
|
switch (cc->cc_type) {
|
|
|
|
case CRYPTO_BUF_CONTIG:
|
|
|
|
MPASS(cc->cc_buf_len >= size);
|
|
|
|
memcpy(cc->cc_buf, src, size);
|
|
|
|
cc->cc_buf += size;
|
|
|
|
cc->cc_buf_len -= size;
|
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_MBUF:
|
|
|
|
for (;;) {
|
|
|
|
KASSERT((cc->cc_mbuf->m_flags & M_EXTPG) == 0,
|
|
|
|
("%s: not supported for unmapped mbufs", __func__));
|
|
|
|
dst = mtod(cc->cc_mbuf, char *) + cc->cc_offset;
|
|
|
|
remain = cc->cc_mbuf->m_len - cc->cc_offset;
|
|
|
|
todo = MIN(remain, size);
|
|
|
|
memcpy(dst, src, todo);
|
2020-05-29 05:41:21 +00:00
|
|
|
src += todo;
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
if (todo < remain) {
|
|
|
|
cc->cc_offset += todo;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
size -= todo;
|
|
|
|
cc->cc_mbuf = cc->cc_mbuf->m_next;
|
|
|
|
cc->cc_offset = 0;
|
|
|
|
if (size == 0)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_UIO:
|
|
|
|
for (;;) {
|
|
|
|
dst = (char *)cc->cc_iov->iov_base + cc->cc_offset;
|
|
|
|
remain = cc->cc_iov->iov_len - cc->cc_offset;
|
|
|
|
todo = MIN(remain, size);
|
|
|
|
memcpy(dst, src, todo);
|
2020-05-29 05:41:21 +00:00
|
|
|
src += todo;
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
if (todo < remain) {
|
|
|
|
cc->cc_offset += todo;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
size -= todo;
|
|
|
|
cc->cc_iov++;
|
|
|
|
cc->cc_offset = 0;
|
|
|
|
if (size == 0)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
#ifdef INVARIANTS
|
|
|
|
panic("%s: invalid buffer type %d", __func__, cc->cc_type);
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
crypto_cursor_copydata(struct crypto_buffer_cursor *cc, int size, void *vdst)
|
|
|
|
{
|
|
|
|
size_t remain, todo;
|
|
|
|
const char *src;
|
|
|
|
char *dst;
|
|
|
|
|
|
|
|
dst = vdst;
|
|
|
|
switch (cc->cc_type) {
|
|
|
|
case CRYPTO_BUF_CONTIG:
|
|
|
|
MPASS(cc->cc_buf_len >= size);
|
|
|
|
memcpy(dst, cc->cc_buf, size);
|
|
|
|
cc->cc_buf += size;
|
|
|
|
cc->cc_buf_len -= size;
|
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_MBUF:
|
|
|
|
for (;;) {
|
|
|
|
KASSERT((cc->cc_mbuf->m_flags & M_EXTPG) == 0,
|
|
|
|
("%s: not supported for unmapped mbufs", __func__));
|
|
|
|
src = mtod(cc->cc_mbuf, const char *) + cc->cc_offset;
|
|
|
|
remain = cc->cc_mbuf->m_len - cc->cc_offset;
|
|
|
|
todo = MIN(remain, size);
|
|
|
|
memcpy(dst, src, todo);
|
|
|
|
dst += todo;
|
|
|
|
if (todo < remain) {
|
|
|
|
cc->cc_offset += todo;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
size -= todo;
|
|
|
|
cc->cc_mbuf = cc->cc_mbuf->m_next;
|
|
|
|
cc->cc_offset = 0;
|
|
|
|
if (size == 0)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_UIO:
|
|
|
|
for (;;) {
|
|
|
|
src = (const char *)cc->cc_iov->iov_base +
|
|
|
|
cc->cc_offset;
|
|
|
|
remain = cc->cc_iov->iov_len - cc->cc_offset;
|
|
|
|
todo = MIN(remain, size);
|
|
|
|
memcpy(dst, src, todo);
|
|
|
|
dst += todo;
|
|
|
|
if (todo < remain) {
|
|
|
|
cc->cc_offset += todo;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
size -= todo;
|
|
|
|
cc->cc_iov++;
|
|
|
|
cc->cc_offset = 0;
|
|
|
|
if (size == 0)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
#ifdef INVARIANTS
|
|
|
|
panic("%s: invalid buffer type %d", __func__, cc->cc_type);
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* To avoid advancing 'cursor', make a local copy that gets advanced
|
|
|
|
* instead.
|
|
|
|
*/
|
|
|
|
void
|
|
|
|
crypto_cursor_copydata_noadv(struct crypto_buffer_cursor *cc, int size,
|
|
|
|
void *vdst)
|
|
|
|
{
|
|
|
|
struct crypto_buffer_cursor copy;
|
|
|
|
|
|
|
|
copy = *cc;
|
|
|
|
crypto_cursor_copydata(©, size, vdst);
|
|
|
|
}
|
|
|
|
|
2006-05-17 17:56:00 +00:00
|
|
|
/*
|
|
|
|
* Apply function f to the data in an iovec list starting "off" bytes from
|
|
|
|
* the beginning, continuing for "len" bytes.
|
|
|
|
*/
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
static int
|
2006-05-17 17:56:00 +00:00
|
|
|
cuio_apply(struct uio *uio, int off, int len, int (*f)(void *, void *, u_int),
|
|
|
|
void *arg)
|
|
|
|
{
|
|
|
|
struct iovec *iov = uio->uio_iov;
|
|
|
|
int iol = uio->uio_iovcnt;
|
|
|
|
unsigned count;
|
|
|
|
int rval;
|
|
|
|
|
|
|
|
CUIO_SKIP();
|
|
|
|
while (len > 0) {
|
|
|
|
KASSERT(iol >= 0, ("%s: empty", __func__));
|
|
|
|
count = min(iov->iov_len - off, len);
|
|
|
|
rval = (*f)(arg, ((caddr_t)iov->iov_base) + off, count);
|
|
|
|
if (rval)
|
|
|
|
return (rval);
|
|
|
|
len -= count;
|
|
|
|
off = 0;
|
|
|
|
iol--;
|
|
|
|
iov++;
|
|
|
|
}
|
|
|
|
return (0);
|
|
|
|
}
|
2006-06-04 22:15:13 +00:00
|
|
|
|
|
|
|
void
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
crypto_copyback(struct cryptop *crp, int off, int size, const void *src)
|
2006-06-04 22:15:13 +00:00
|
|
|
{
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
struct crypto_buffer *cb;
|
2006-06-04 22:15:13 +00:00
|
|
|
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
if (crp->crp_obuf.cb_type != CRYPTO_BUF_NONE)
|
|
|
|
cb = &crp->crp_obuf;
|
|
|
|
else
|
|
|
|
cb = &crp->crp_buf;
|
|
|
|
switch (cb->cb_type) {
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
case CRYPTO_BUF_MBUF:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
m_copyback(cb->cb_mbuf, off, size, src);
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_UIO:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
cuio_copyback(cb->cb_uio, off, size, src);
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_CONTIG:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
MPASS(off + size <= cb->cb_buf_len);
|
|
|
|
bcopy(src, cb->cb_buf + off, size);
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
break;
|
|
|
|
default:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
#ifdef INVARIANTS
|
|
|
|
panic("invalid crp buf type %d", cb->cb_type);
|
|
|
|
#endif
|
|
|
|
break;
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
}
|
2006-06-04 22:15:13 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
void
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
crypto_copydata(struct cryptop *crp, int off, int size, void *dst)
|
2006-06-04 22:15:13 +00:00
|
|
|
{
|
|
|
|
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
switch (crp->crp_buf.cb_type) {
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
case CRYPTO_BUF_MBUF:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
m_copydata(crp->crp_buf.cb_mbuf, off, size, dst);
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_UIO:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
cuio_copydata(crp->crp_buf.cb_uio, off, size, dst);
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_CONTIG:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
MPASS(off + size <= crp->crp_buf.cb_buf_len);
|
|
|
|
bcopy(crp->crp_buf.cb_buf + off, dst, size);
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
break;
|
|
|
|
default:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
#ifdef INVARIANTS
|
|
|
|
panic("invalid crp buf type %d", crp->crp_buf.cb_type);
|
|
|
|
#endif
|
|
|
|
break;
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
}
|
2006-06-04 22:15:13 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
int
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
crypto_apply_buf(struct crypto_buffer *cb, int off, int len,
|
2006-06-04 22:15:13 +00:00
|
|
|
int (*f)(void *, void *, u_int), void *arg)
|
|
|
|
{
|
|
|
|
int error;
|
|
|
|
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
switch (cb->cb_type) {
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
case CRYPTO_BUF_MBUF:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
error = m_apply(cb->cb_mbuf, off, len, f, arg);
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_UIO:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
error = cuio_apply(cb->cb_uio, off, len, f, arg);
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
break;
|
|
|
|
case CRYPTO_BUF_CONTIG:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
MPASS(off + len <= cb->cb_buf_len);
|
|
|
|
error = (*f)(arg, cb->cb_buf + off, len);
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
break;
|
|
|
|
default:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
#ifdef INVARIANTS
|
|
|
|
panic("invalid crypto buf type %d", cb->cb_type);
|
|
|
|
#endif
|
|
|
|
error = 0;
|
|
|
|
break;
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
}
|
2006-06-04 22:15:13 +00:00
|
|
|
return (error);
|
|
|
|
}
|
2014-12-12 19:56:36 +00:00
|
|
|
|
2015-07-07 18:45:32 +00:00
|
|
|
int
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
crypto_apply(struct cryptop *crp, int off, int len,
|
|
|
|
int (*f)(void *, void *, u_int), void *arg)
|
2014-12-12 19:56:36 +00:00
|
|
|
{
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
return (crypto_apply_buf(&crp->crp_buf, off, len, f, arg));
|
2014-12-12 19:56:36 +00:00
|
|
|
}
|
2018-12-13 04:40:53 +00:00
|
|
|
|
|
|
|
static inline void *
|
|
|
|
m_contiguous_subsegment(struct mbuf *m, size_t skip, size_t len)
|
|
|
|
{
|
|
|
|
int rel_off;
|
|
|
|
|
|
|
|
MPASS(skip <= INT_MAX);
|
|
|
|
|
|
|
|
m = m_getptr(m, (int)skip, &rel_off);
|
|
|
|
if (m == NULL)
|
|
|
|
return (NULL);
|
|
|
|
|
|
|
|
MPASS(rel_off >= 0);
|
|
|
|
skip = rel_off;
|
|
|
|
if (skip + len > m->m_len)
|
|
|
|
return (NULL);
|
|
|
|
|
|
|
|
return (mtod(m, char*) + skip);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void *
|
|
|
|
cuio_contiguous_segment(struct uio *uio, size_t skip, size_t len)
|
|
|
|
{
|
|
|
|
int rel_off, idx;
|
|
|
|
|
|
|
|
MPASS(skip <= INT_MAX);
|
|
|
|
idx = cuio_getptr(uio, (int)skip, &rel_off);
|
|
|
|
if (idx < 0)
|
|
|
|
return (NULL);
|
|
|
|
|
|
|
|
MPASS(rel_off >= 0);
|
|
|
|
skip = rel_off;
|
|
|
|
if (skip + len > uio->uio_iov[idx].iov_len)
|
|
|
|
return (NULL);
|
|
|
|
return ((char *)uio->uio_iov[idx].iov_base + skip);
|
|
|
|
}
|
|
|
|
|
|
|
|
void *
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
crypto_buffer_contiguous_subsegment(struct crypto_buffer *cb, size_t skip,
|
|
|
|
size_t len)
|
2018-12-13 04:40:53 +00:00
|
|
|
{
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
switch (cb->cb_type) {
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
case CRYPTO_BUF_MBUF:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
return (m_contiguous_subsegment(cb->cb_mbuf, skip, len));
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
case CRYPTO_BUF_UIO:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
return (cuio_contiguous_segment(cb->cb_uio, skip, len));
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
case CRYPTO_BUF_CONTIG:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
MPASS(skip + len <= cb->cb_buf_len);
|
|
|
|
return (cb->cb_buf + skip);
|
Refactor driver and consumer interfaces for OCF (in-kernel crypto).
- The linked list of cryptoini structures used in session
initialization is replaced with a new flat structure: struct
crypto_session_params. This session includes a new mode to define
how the other fields should be interpreted. Available modes
include:
- COMPRESS (for compression/decompression)
- CIPHER (for simply encryption/decryption)
- DIGEST (computing and verifying digests)
- AEAD (combined auth and encryption such as AES-GCM and AES-CCM)
- ETA (combined auth and encryption using encrypt-then-authenticate)
Additional modes could be added in the future (e.g. if we wanted to
support TLS MtE for AES-CBC in the kernel we could add a new mode
for that. TLS modes might also affect how AAD is interpreted, etc.)
The flat structure also includes the key lengths and algorithms as
before. However, code doesn't have to walk the linked list and
switch on the algorithm to determine which key is the auth key vs
encryption key. The 'csp_auth_*' fields are always used for auth
keys and settings and 'csp_cipher_*' for cipher. (Compression
algorithms are stored in csp_cipher_alg.)
- Drivers no longer register a list of supported algorithms. This
doesn't quite work when you factor in modes (e.g. a driver might
support both AES-CBC and SHA2-256-HMAC separately but not combined
for ETA). Instead, a new 'crypto_probesession' method has been
added to the kobj interface for symmteric crypto drivers. This
method returns a negative value on success (similar to how
device_probe works) and the crypto framework uses this value to pick
the "best" driver. There are three constants for hardware
(e.g. ccr), accelerated software (e.g. aesni), and plain software
(cryptosoft) that give preference in that order. One effect of this
is that if you request only hardware when creating a new session,
you will no longer get a session using accelerated software.
Another effect is that the default setting to disallow software
crypto via /dev/crypto now disables accelerated software.
Once a driver is chosen, 'crypto_newsession' is invoked as before.
- Crypto operations are now solely described by the flat 'cryptop'
structure. The linked list of descriptors has been removed.
A separate enum has been added to describe the type of data buffer
in use instead of using CRYPTO_F_* flags to make it easier to add
more types in the future if needed (e.g. wired userspace buffers for
zero-copy). It will also make it easier to re-introduce separate
input and output buffers (in-kernel TLS would benefit from this).
Try to make the flags related to IV handling less insane:
- CRYPTO_F_IV_SEPARATE means that the IV is stored in the 'crp_iv'
member of the operation structure. If this flag is not set, the
IV is stored in the data buffer at the 'crp_iv_start' offset.
- CRYPTO_F_IV_GENERATE means that a random IV should be generated
and stored into the data buffer. This cannot be used with
CRYPTO_F_IV_SEPARATE.
If a consumer wants to deal with explicit vs implicit IVs, etc. it
can always generate the IV however it needs and store partial IVs in
the buffer and the full IV/nonce in crp_iv and set
CRYPTO_F_IV_SEPARATE.
The layout of the buffer is now described via fields in cryptop.
crp_aad_start and crp_aad_length define the boundaries of any AAD.
Previously with GCM and CCM you defined an auth crd with this range,
but for ETA your auth crd had to span both the AAD and plaintext
(and they had to be adjacent).
crp_payload_start and crp_payload_length define the boundaries of
the plaintext/ciphertext. Modes that only do a single operation
(COMPRESS, CIPHER, DIGEST) should only use this region and leave the
AAD region empty.
If a digest is present (or should be generated), it's starting
location is marked by crp_digest_start.
Instead of using the CRD_F_ENCRYPT flag to determine the direction
of the operation, cryptop now includes an 'op' field defining the
operation to perform. For digests I've added a new VERIFY digest
mode which assumes a digest is present in the input and fails the
request with EBADMSG if it doesn't match the internally-computed
digest. GCM and CCM already assumed this, and the new AEAD mode
requires this for decryption. The new ETA mode now also requires
this for decryption, so IPsec and GELI no longer do their own
authentication verification. Simple DIGEST operations can also do
this, though there are no in-tree consumers.
To eventually support some refcounting to close races, the session
cookie is now passed to crypto_getop() and clients should no longer
set crp_sesssion directly.
- Assymteric crypto operation structures should be allocated via
crypto_getkreq() and freed via crypto_freekreq(). This permits the
crypto layer to track open asym requests and close races with a
driver trying to unregister while asym requests are in flight.
- crypto_copyback, crypto_copydata, crypto_apply, and
crypto_contiguous_subsegment now accept the 'crp' object as the
first parameter instead of individual members. This makes it easier
to deal with different buffer types in the future as well as
separate input and output buffers. It's also simpler for driver
writers to use.
- bus_dmamap_load_crp() loads a DMA mapping for a crypto buffer.
This understands the various types of buffers so that drivers that
use DMA do not have to be aware of different buffer types.
- Helper routines now exist to build an auth context for HMAC IPAD
and OPAD. This reduces some duplicated work among drivers.
- Key buffers are now treated as const throughout the framework and in
device drivers. However, session key buffers provided when a session
is created are expected to remain alive for the duration of the
session.
- GCM and CCM sessions now only specify a cipher algorithm and a cipher
key. The redundant auth information is not needed or used.
- For cryptosoft, split up the code a bit such that the 'process'
callback now invokes a function pointer in the session. This
function pointer is set based on the mode (in effect) though it
simplifies a few edge cases that would otherwise be in the switch in
'process'.
It does split up GCM vs CCM which I think is more readable even if there
is some duplication.
- I changed /dev/crypto to support GMAC requests using CRYPTO_AES_NIST_GMAC
as an auth algorithm and updated cryptocheck to work with it.
- Combined cipher and auth sessions via /dev/crypto now always use ETA
mode. The COP_F_CIPHER_FIRST flag is now a no-op that is ignored.
This was actually documented as being true in crypto(4) before, but
the code had not implemented this before I added the CIPHER_FIRST
flag.
- I have not yet updated /dev/crypto to be aware of explicit modes for
sessions. I will probably do that at some point in the future as well
as teach it about IV/nonce and tag lengths for AEAD so we can support
all of the NIST KAT tests for GCM and CCM.
- I've split up the exising crypto.9 manpage into several pages
of which many are written from scratch.
- I have converted all drivers and consumers in the tree and verified
that they compile, but I have not tested all of them. I have tested
the following drivers:
- cryptosoft
- aesni (AES only)
- blake2
- ccr
and the following consumers:
- cryptodev
- IPsec
- ktls_ocf
- GELI (lightly)
I have not tested the following:
- ccp
- aesni with sha
- hifn
- kgssapi_krb5
- ubsec
- padlock
- safe
- armv8_crypto (aarch64)
- glxsb (i386)
- sec (ppc)
- cesa (armv7)
- cryptocteon (mips64)
- nlmsec (mips64)
Discussed with: cem
Relnotes: yes
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D23677
2020-03-27 18:25:23 +00:00
|
|
|
default:
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
#ifdef INVARIANTS
|
|
|
|
panic("invalid crp buf type %d", cb->cb_type);
|
|
|
|
#endif
|
|
|
|
return (NULL);
|
2018-12-13 04:40:53 +00:00
|
|
|
}
|
|
|
|
}
|
Add support for optional separate output buffers to in-kernel crypto.
Some crypto consumers such as GELI and KTLS for file-backed sendfile
need to store their output in a separate buffer from the input.
Currently these consumers copy the contents of the input buffer into
the output buffer and queue an in-place crypto operation on the output
buffer. Using a separate output buffer avoids this copy.
- Create a new 'struct crypto_buffer' describing a crypto buffer
containing a type and type-specific fields. crp_ilen is gone,
instead buffers that use a flat kernel buffer have a cb_buf_len
field for their length. The length of other buffer types is
inferred from the backing store (e.g. uio_resid for a uio).
Requests now have two such structures: crp_buf for the input buffer,
and crp_obuf for the output buffer.
- Consumers now use helper functions (crypto_use_*,
e.g. crypto_use_mbuf()) to configure the input buffer. If an output
buffer is not configured, the request still modifies the input
buffer in-place. A consumer uses a second set of helper functions
(crypto_use_output_*) to configure an output buffer.
- Consumers must request support for separate output buffers when
creating a crypto session via the CSP_F_SEPARATE_OUTPUT flag and are
only permitted to queue a request with a separate output buffer on
sessions with this flag set. Existing drivers already reject
sessions with unknown flags, so this permits drivers to be modified
to support this extension without requiring all drivers to change.
- Several data-related functions now have matching versions that
operate on an explicit buffer (e.g. crypto_apply_buf,
crypto_contiguous_subsegment_buf, bus_dma_load_crp_buf).
- Most of the existing data-related functions operate on the input
buffer. However crypto_copyback always writes to the output buffer
if a request uses a separate output buffer.
- For the regions in input/output buffers, the following conventions
are followed:
- AAD and IV are always present in input only and their
fields are offsets into the input buffer.
- payload is always present in both buffers. If a request uses a
separate output buffer, it must set a new crp_payload_start_output
field to the offset of the payload in the output buffer.
- digest is in the input buffer for verify operations, and in the
output buffer for compute operations. crp_digest_start is relative
to the appropriate buffer.
- Add a crypto buffer cursor abstraction. This is a more general form
of some bits in the cryptosoft driver that tried to always use uio's.
However, compared to the original code, this avoids rewalking the uio
iovec array for requests with multiple vectors. It also avoids
allocate an iovec array for mbufs and populating it by instead walking
the mbuf chain directly.
- Update the cryptosoft(4) driver to support separate output buffers
making use of the cursor abstraction.
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D24545
2020-05-25 22:12:04 +00:00
|
|
|
|
|
|
|
void *
|
|
|
|
crypto_contiguous_subsegment(struct cryptop *crp, size_t skip, size_t len)
|
|
|
|
{
|
|
|
|
return (crypto_buffer_contiguous_subsegment(&crp->crp_buf, skip, len));
|
|
|
|
}
|