* [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver
@ 2018-02-24 13:14 Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 1/7] crypto/virtio: add virtio related fundamental functions Jay Zhou
` (7 more replies)
0 siblings, 8 replies; 11+ messages in thread
From: Jay Zhou @ 2018-02-24 13:14 UTC (permalink / raw)
To: dev
Cc: pablo.de.lara.guarch, roy.fan.zhang, thomas, arei.gonglei,
xin.zeng, weidong.huang, wangxinxin.wang, longpeng2,
jianjay.zhou
This patch series introduce virtio crypto poll mode driver.
Since it is limited by the backend of the virtio-crypto,
this patch series only supports a limited subset of crypto services.
Only the following algorithms are tested:
Cipher algorithms:
- RTE_CRYPTO_CIPHER_AES_CBC (128-bit, 192-bit and 256-bit keys)
- RTE_CRYPTO_CIPHER_AES_CTR (128-bit, 192-bit and 256-bit keys)
Firstly build QEMU with libgcrypt cryptography support.
QEMU can then be started using the following parameters:
qemu-system-x86_64 \
[...] \
-object cryptodev-backend-builtin,id=cryptodev0 \
-device virtio-crypto-pci,id=crypto0,cryptodev=cryptodev0 \
[...]
Bind the uio_generic driver for the virtio-crypto device.
For example, 0000:00:04.0 is the domain, bus, device and function
number of the virtio-crypto device:
modprobe uio_pci_generic
echo -n 0000:00:04.0 > /sys/bus/pci/drivers/virtio-pci/unbind
echo "1af4 1054" > /sys/bus/pci/drivers/uio_pci_generic/new_id
The front-end virtio crypto PMD driver can be installed:
cd to the top-level DPDK directory
sed -i 's,\(CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO\)=n,\1=y,' config/common_base
make config T=x86_64-native-linuxapp-gcc
make install T=x86_64-native-linuxapp-gcc
The test cases can be compiled as below:
cd to the top-level DPDK directory
export RTE_TARGET=x86_64-native-linuxapp-gcc
export RTE_SDK=`pwd`
cd to test/test
make
./test (MUST reserve enough huge pages memory)
type the command "cryptodev_virtio_autotest" to test
The result should be like this:
RTE>>cryptodev_virtio_autotest
+ ------------------------------------------------------- +
+ Test Suite : Crypto VIRTIO Unit Test Suite
+ ------------------------------------------------------- +
0) TestCase AES-128-CBC Encryption PASS
1) TestCase AES-128-CBC Decryption PASS
2) TestCase AES-192-CBC Encryption PASS
3) TestCase AES-192-CBC Decryption PASS
4) TestCase AES-256-CBC Encryption PASS
5) TestCase AES-256-CBC Decryption PASS
6) TestCase AES-256-CBC OOP Encryption PASS
7) TestCase AES-256-CBC OOP Decryption PASS
8) TestCase AES-128-CTR Encryption PASS
9) TestCase AES-128-CTR Decryption PASS
10) TestCase AES-192-CTR Encryption PASS
11) TestCase AES-192-CTR Decryption PASS
12) TestCase AES-256-CTR Encryption PASS
13) TestCase AES-256-CTR Decryption PASS
+ TestCase [ 0] : test_AES_cipheronly_virtio_all succeeded
+ ------------------------------------------------------- +
+ Test Suite Summary
+ Tests Total : 1
+ Tests Skipped : 0
+ Tests Executed : 1
+ Tests Unsupported: 0
+ Tests Passed : 1
+ Tests Failed : 0
+ ------------------------------------------------------- +
Test OK
Please help to review, thanks!
Changes in v2:
- using pre-allocated mempool instead of rte_malloc to improve performance [Fan]
- split the patch into a patchset [Fan]
- using linux/virtio_crypto.h instead of creating a copy of the file [Fan]
- update doc/guides/cryptodevs for describing virtio crypto PMD [Fan]
- update copyright
- delete virtio legacy mode code since virtio-crypto conforms to virtio-1.0
- refine the function and variable names
- fix errors and warnings reported by checkpatch
Jay Zhou (7):
crypto/virtio: add virtio related fundamental functions
crpyto/virtio: add crypto related session structure
cryptodev/virtio: core code of crypto devices
crypto/virtio: add makefile
cryptodev: add document for virtio crypto PMD
cryptodev: add function tests for virtio crypto PMD
MAINTAINERS: add myself as virtio crypto PMD maintainer
MAINTAINERS | 6 +
config/common_base | 20 +
doc/guides/cryptodevs/features/virtio.ini | 20 +
doc/guides/cryptodevs/virtio.rst | 93 ++
drivers/crypto/Makefile | 1 +
drivers/crypto/virtio/Makefile | 31 +
.../virtio/rte_pmd_virtio_crypto_version.map | 3 +
drivers/crypto/virtio/virtio_crypto_algs.h | 27 +
drivers/crypto/virtio/virtio_cryptodev.c | 1544 ++++++++++++++++++++
drivers/crypto/virtio/virtio_cryptodev.h | 66 +
drivers/crypto/virtio/virtio_logs.h | 47 +
drivers/crypto/virtio/virtio_pci.c | 460 ++++++
drivers/crypto/virtio/virtio_pci.h | 252 ++++
drivers/crypto/virtio/virtio_ring.h | 137 ++
drivers/crypto/virtio/virtio_rxtx.c | 533 +++++++
drivers/crypto/virtio/virtqueue.c | 43 +
drivers/crypto/virtio/virtqueue.h | 176 +++
mk/rte.app.mk | 1 +
test/test/test_cryptodev.c | 49 +
test/test/test_cryptodev.h | 1 +
test/test/test_cryptodev_aes_test_vectors.h | 45 +-
test/test/test_cryptodev_blockcipher.c | 9 +-
test/test/test_cryptodev_blockcipher.h | 1 +
23 files changed, 3549 insertions(+), 16 deletions(-)
create mode 100644 doc/guides/cryptodevs/features/virtio.ini
create mode 100644 doc/guides/cryptodevs/virtio.rst
create mode 100644 drivers/crypto/virtio/Makefile
create mode 100644 drivers/crypto/virtio/rte_pmd_virtio_crypto_version.map
create mode 100644 drivers/crypto/virtio/virtio_crypto_algs.h
create mode 100644 drivers/crypto/virtio/virtio_cryptodev.c
create mode 100644 drivers/crypto/virtio/virtio_cryptodev.h
create mode 100644 drivers/crypto/virtio/virtio_logs.h
create mode 100644 drivers/crypto/virtio/virtio_pci.c
create mode 100644 drivers/crypto/virtio/virtio_pci.h
create mode 100644 drivers/crypto/virtio/virtio_ring.h
create mode 100644 drivers/crypto/virtio/virtio_rxtx.c
create mode 100644 drivers/crypto/virtio/virtqueue.c
create mode 100644 drivers/crypto/virtio/virtqueue.h
--
1.8.3.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* [dpdk-dev] [PATCH v2 1/7] crypto/virtio: add virtio related fundamental functions
2018-02-24 13:14 [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver Jay Zhou
@ 2018-02-24 13:14 ` Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 2/7] crpyto/virtio: add crypto related session structure Jay Zhou
` (6 subsequent siblings)
7 siblings, 0 replies; 11+ messages in thread
From: Jay Zhou @ 2018-02-24 13:14 UTC (permalink / raw)
To: dev
Cc: pablo.de.lara.guarch, roy.fan.zhang, thomas, arei.gonglei,
xin.zeng, weidong.huang, wangxinxin.wang, longpeng2,
jianjay.zhou
Since there are not have the common virtio library, we have to put
these files here. They are basically the same with virtio net related files
with some minor changes.
Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com>
---
config/common_base | 20 ++
drivers/crypto/virtio/virtio_logs.h | 47 ++++
drivers/crypto/virtio/virtio_pci.c | 460 ++++++++++++++++++++++++++++++++++++
drivers/crypto/virtio/virtio_pci.h | 252 ++++++++++++++++++++
drivers/crypto/virtio/virtio_ring.h | 137 +++++++++++
drivers/crypto/virtio/virtqueue.c | 43 ++++
drivers/crypto/virtio/virtqueue.h | 176 ++++++++++++++
7 files changed, 1135 insertions(+)
create mode 100644 drivers/crypto/virtio/virtio_logs.h
create mode 100644 drivers/crypto/virtio/virtio_pci.c
create mode 100644 drivers/crypto/virtio/virtio_pci.h
create mode 100644 drivers/crypto/virtio/virtio_ring.h
create mode 100644 drivers/crypto/virtio/virtqueue.c
create mode 100644 drivers/crypto/virtio/virtqueue.h
diff --git a/config/common_base b/config/common_base
index ad03cf4..19d0cdd 100644
--- a/config/common_base
+++ b/config/common_base
@@ -482,6 +482,26 @@ CONFIG_RTE_LIBRTE_PMD_QAT_DEBUG_DRIVER=n
CONFIG_RTE_QAT_PMD_MAX_NB_SESSIONS=2048
#
+# Compile PMD for virtio crypto devices
+#
+CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO=n
+CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO_DEBUG_INIT=n
+CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO_DEBUG_SESSION=n
+CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO_DEBUG_TX=n
+CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO_DEBUG_RX=n
+CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO_DEBUG_DRIVER=n
+CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO_DEBUG_DUMP=n
+#
+# Number of maximum virtio crypto devices
+#
+CONFIG_RTE_MAX_VIRTIO_CRYPTO=32
+#
+# Number of sessions to create in the session memory pool
+# on a single virtio crypto device.
+#
+CONFIG_RTE_VIRTIO_CRYPTO_PMD_MAX_NB_SESSIONS=1024
+
+#
# Compile PMD for AESNI backed device
#
CONFIG_RTE_LIBRTE_PMD_AESNI_MB=n
diff --git a/drivers/crypto/virtio/virtio_logs.h b/drivers/crypto/virtio/virtio_logs.h
new file mode 100644
index 0000000..20582a4
--- /dev/null
+++ b/drivers/crypto/virtio/virtio_logs.h
@@ -0,0 +1,47 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 HUAWEI TECHNOLOGIES CO., LTD.
+ */
+
+#ifndef _VIRTIO_LOGS_H_
+#define _VIRTIO_LOGS_H_
+
+#include <rte_log.h>
+
+#ifdef RTE_LIBRTE_PMD_VIRTIO_CRYPTO_DEBUG_INIT
+#define PMD_INIT_LOG(level, fmt, args...) \
+ RTE_LOG(level, PMD, "%s(): " fmt "\n", __func__, ## args)
+#define PMD_INIT_FUNC_TRACE() PMD_INIT_LOG(DEBUG, " >>")
+#else
+#define PMD_INIT_LOG(level, fmt, args...) do { } while (0)
+#define PMD_INIT_FUNC_TRACE() do { } while (0)
+#endif
+
+#ifdef RTE_LIBRTE_PMD_VIRTIO_CRYPTO_DEBUG_SESSION
+#define PMD_SESSION_LOG(level, fmt, args...) \
+ RTE_LOG(level, PMD, "%s() session: " fmt "\n", __func__, ## args)
+#else
+#define PMD_SESSION_LOG(level, fmt, args...) do { } while (0)
+#endif
+
+#ifdef RTE_LIBRTE_PMD_VIRTIO_CRYPTO_DEBUG_RX
+#define PMD_RX_LOG(level, fmt, args...) \
+ RTE_LOG(level, PMD, "%s() rx: " fmt "\n", __func__, ## args)
+#else
+#define PMD_RX_LOG(level, fmt, args...) do { } while (0)
+#endif
+
+#ifdef RTE_LIBRTE_PMD_VIRTIO_CRYPTO_DEBUG_TX
+#define PMD_TX_LOG(level, fmt, args...) \
+ RTE_LOG(level, PMD, "%s() tx: " fmt "\n", __func__, ## args)
+#else
+#define PMD_TX_LOG(level, fmt, args...) do { } while (0)
+#endif
+
+#ifdef RTE_LIBRTE_PMD_VIRTIO_CRYPTO_DEBUG_DRIVER
+#define PMD_DRV_LOG(level, fmt, args...) \
+ RTE_LOG(level, PMD, "%s(): driver " fmt "\n", __func__, ## args)
+#else
+#define PMD_DRV_LOG(level, fmt, args...) do { } while (0)
+#endif
+
+#endif /* _VIRTIO_LOGS_H_ */
diff --git a/drivers/crypto/virtio/virtio_pci.c b/drivers/crypto/virtio/virtio_pci.c
new file mode 100644
index 0000000..7aa5cdd
--- /dev/null
+++ b/drivers/crypto/virtio/virtio_pci.c
@@ -0,0 +1,460 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 HUAWEI TECHNOLOGIES CO., LTD.
+ */
+
+#include <stdint.h>
+
+#ifdef RTE_EXEC_ENV_LINUXAPP
+ #include <dirent.h>
+ #include <fcntl.h>
+#endif
+
+#include <rte_io.h>
+#include <rte_bus.h>
+
+#include "virtio_pci.h"
+#include "virtqueue.h"
+
+/*
+ * Following macros are derived from linux/pci_regs.h, however,
+ * we can't simply include that header here, as there is no such
+ * file for non-Linux platform.
+ */
+#define PCI_CAPABILITY_LIST 0x34
+#define PCI_CAP_ID_VNDR 0x09
+#define PCI_CAP_ID_MSIX 0x11
+
+/*
+ * The remaining space is defined by each driver as the per-driver
+ * configuration space.
+ */
+#define VIRTIO_PCI_CONFIG(hw) \
+ (((hw)->use_msix == VIRTIO_MSIX_ENABLED) ? 24 : 20)
+
+static inline int
+check_vq_phys_addr_ok(struct virtqueue *vq)
+{
+ /* Virtio PCI device VIRTIO_PCI_QUEUE_PF register is 32bit,
+ * and only accepts 32 bit page frame number.
+ * Check if the allocated physical memory exceeds 16TB.
+ */
+ if ((vq->vq_ring_mem + vq->vq_ring_size - 1) >>
+ (VIRTIO_PCI_QUEUE_ADDR_SHIFT + 32)) {
+ PMD_INIT_LOG(ERR, "vring address shouldn't be above 16TB!");
+ return 0;
+ }
+
+ return 1;
+}
+
+static inline void
+io_write64_twopart(uint64_t val, uint32_t *lo, uint32_t *hi)
+{
+ rte_write32(val & ((1ULL << 32) - 1), lo);
+ rte_write32(val >> 32, hi);
+}
+
+static void
+modern_read_dev_config(struct virtio_crypto_hw *hw, size_t offset,
+ void *dst, int length)
+{
+ int i;
+ uint8_t *p;
+ uint8_t old_gen, new_gen;
+
+ do {
+ old_gen = rte_read8(&hw->common_cfg->config_generation);
+
+ p = dst;
+ for (i = 0; i < length; i++)
+ *p++ = rte_read8((uint8_t *)hw->dev_cfg + offset + i);
+
+ new_gen = rte_read8(&hw->common_cfg->config_generation);
+ } while (old_gen != new_gen);
+}
+
+static void
+modern_write_dev_config(struct virtio_crypto_hw *hw, size_t offset,
+ const void *src, int length)
+{
+ int i;
+ const uint8_t *p = src;
+
+ for (i = 0; i < length; i++)
+ rte_write8((*p++), (((uint8_t *)hw->dev_cfg) + offset + i));
+}
+
+static uint64_t
+modern_get_features(struct virtio_crypto_hw *hw)
+{
+ uint32_t features_lo, features_hi;
+
+ rte_write32(0, &hw->common_cfg->device_feature_select);
+ features_lo = rte_read32(&hw->common_cfg->device_feature);
+
+ rte_write32(1, &hw->common_cfg->device_feature_select);
+ features_hi = rte_read32(&hw->common_cfg->device_feature);
+
+ return ((uint64_t)features_hi << 32) | features_lo;
+}
+
+static void
+modern_set_features(struct virtio_crypto_hw *hw, uint64_t features)
+{
+ rte_write32(0, &hw->common_cfg->guest_feature_select);
+ rte_write32(features & ((1ULL << 32) - 1),
+ &hw->common_cfg->guest_feature);
+
+ rte_write32(1, &hw->common_cfg->guest_feature_select);
+ rte_write32(features >> 32,
+ &hw->common_cfg->guest_feature);
+}
+
+static uint8_t
+modern_get_status(struct virtio_crypto_hw *hw)
+{
+ return rte_read8(&hw->common_cfg->device_status);
+}
+
+static void
+modern_set_status(struct virtio_crypto_hw *hw, uint8_t status)
+{
+ rte_write8(status, &hw->common_cfg->device_status);
+}
+
+static void
+modern_reset(struct virtio_crypto_hw *hw)
+{
+ modern_set_status(hw, VIRTIO_CONFIG_STATUS_RESET);
+ modern_get_status(hw);
+}
+
+static uint8_t
+modern_get_isr(struct virtio_crypto_hw *hw)
+{
+ return rte_read8(hw->isr);
+}
+
+static uint16_t
+modern_set_config_irq(struct virtio_crypto_hw *hw, uint16_t vec)
+{
+ rte_write16(vec, &hw->common_cfg->msix_config);
+ return rte_read16(&hw->common_cfg->msix_config);
+}
+
+static uint16_t
+modern_set_queue_irq(struct virtio_crypto_hw *hw, struct virtqueue *vq,
+ uint16_t vec)
+{
+ rte_write16(vq->vq_queue_index, &hw->common_cfg->queue_select);
+ rte_write16(vec, &hw->common_cfg->queue_msix_vector);
+ return rte_read16(&hw->common_cfg->queue_msix_vector);
+}
+
+static uint16_t
+modern_get_queue_num(struct virtio_crypto_hw *hw, uint16_t queue_id)
+{
+ rte_write16(queue_id, &hw->common_cfg->queue_select);
+ return rte_read16(&hw->common_cfg->queue_size);
+}
+
+static int
+modern_setup_queue(struct virtio_crypto_hw *hw, struct virtqueue *vq)
+{
+ uint64_t desc_addr, avail_addr, used_addr;
+ uint16_t notify_off;
+
+ if (!check_vq_phys_addr_ok(vq))
+ return -1;
+
+ desc_addr = vq->vq_ring_mem;
+ avail_addr = desc_addr + vq->vq_nentries * sizeof(struct vring_desc);
+ used_addr = RTE_ALIGN_CEIL(avail_addr + offsetof(struct vring_avail,
+ ring[vq->vq_nentries]),
+ VIRTIO_PCI_VRING_ALIGN);
+
+ rte_write16(vq->vq_queue_index, &hw->common_cfg->queue_select);
+
+ io_write64_twopart(desc_addr, &hw->common_cfg->queue_desc_lo,
+ &hw->common_cfg->queue_desc_hi);
+ io_write64_twopart(avail_addr, &hw->common_cfg->queue_avail_lo,
+ &hw->common_cfg->queue_avail_hi);
+ io_write64_twopart(used_addr, &hw->common_cfg->queue_used_lo,
+ &hw->common_cfg->queue_used_hi);
+
+ notify_off = rte_read16(&hw->common_cfg->queue_notify_off);
+ vq->notify_addr = (void *)((uint8_t *)hw->notify_base +
+ notify_off * hw->notify_off_multiplier);
+
+ rte_write16(1, &hw->common_cfg->queue_enable);
+
+ PMD_INIT_LOG(DEBUG, "queue %u addresses:", vq->vq_queue_index);
+ PMD_INIT_LOG(DEBUG, "\t desc_addr: %" PRIx64, desc_addr);
+ PMD_INIT_LOG(DEBUG, "\t aval_addr: %" PRIx64, avail_addr);
+ PMD_INIT_LOG(DEBUG, "\t used_addr: %" PRIx64, used_addr);
+ PMD_INIT_LOG(DEBUG, "\t notify addr: %p (notify offset: %u)",
+ vq->notify_addr, notify_off);
+
+ return 0;
+}
+
+static void
+modern_del_queue(struct virtio_crypto_hw *hw, struct virtqueue *vq)
+{
+ rte_write16(vq->vq_queue_index, &hw->common_cfg->queue_select);
+
+ io_write64_twopart(0, &hw->common_cfg->queue_desc_lo,
+ &hw->common_cfg->queue_desc_hi);
+ io_write64_twopart(0, &hw->common_cfg->queue_avail_lo,
+ &hw->common_cfg->queue_avail_hi);
+ io_write64_twopart(0, &hw->common_cfg->queue_used_lo,
+ &hw->common_cfg->queue_used_hi);
+
+ rte_write16(0, &hw->common_cfg->queue_enable);
+}
+
+static void
+modern_notify_queue(struct virtio_crypto_hw *hw __rte_unused,
+ struct virtqueue *vq)
+{
+ rte_write16(vq->vq_queue_index, vq->notify_addr);
+}
+
+const struct virtio_pci_ops virtio_crypto_modern_ops = {
+ .read_dev_cfg = modern_read_dev_config,
+ .write_dev_cfg = modern_write_dev_config,
+ .reset = modern_reset,
+ .get_status = modern_get_status,
+ .set_status = modern_set_status,
+ .get_features = modern_get_features,
+ .set_features = modern_set_features,
+ .get_isr = modern_get_isr,
+ .set_config_irq = modern_set_config_irq,
+ .set_queue_irq = modern_set_queue_irq,
+ .get_queue_num = modern_get_queue_num,
+ .setup_queue = modern_setup_queue,
+ .del_queue = modern_del_queue,
+ .notify_queue = modern_notify_queue,
+};
+
+void
+vtpci_read_cryptodev_config(struct virtio_crypto_hw *hw, size_t offset,
+ void *dst, int length)
+{
+ VTPCI_OPS(hw)->read_dev_cfg(hw, offset, dst, length);
+}
+
+void
+vtpci_write_cryptodev_config(struct virtio_crypto_hw *hw, size_t offset,
+ const void *src, int length)
+{
+ VTPCI_OPS(hw)->write_dev_cfg(hw, offset, src, length);
+}
+
+uint64_t
+vtpci_cryptodev_negotiate_features(struct virtio_crypto_hw *hw,
+ uint64_t host_features)
+{
+ uint64_t features;
+
+ /*
+ * Limit negotiated features to what the driver, virtqueue, and
+ * host all support.
+ */
+ features = host_features & hw->guest_features;
+ VTPCI_OPS(hw)->set_features(hw, features);
+
+ return features;
+}
+
+void
+vtpci_cryptodev_reset(struct virtio_crypto_hw *hw)
+{
+ VTPCI_OPS(hw)->set_status(hw, VIRTIO_CONFIG_STATUS_RESET);
+ /* flush status write */
+ VTPCI_OPS(hw)->get_status(hw);
+}
+
+void
+vtpci_cryptodev_reinit_complete(struct virtio_crypto_hw *hw)
+{
+ vtpci_cryptodev_set_status(hw, VIRTIO_CONFIG_STATUS_DRIVER_OK);
+}
+
+void
+vtpci_cryptodev_set_status(struct virtio_crypto_hw *hw, uint8_t status)
+{
+ if (status != VIRTIO_CONFIG_STATUS_RESET)
+ status |= VTPCI_OPS(hw)->get_status(hw);
+
+ VTPCI_OPS(hw)->set_status(hw, status);
+}
+
+uint8_t
+vtpci_cryptodev_get_status(struct virtio_crypto_hw *hw)
+{
+ return VTPCI_OPS(hw)->get_status(hw);
+}
+
+uint8_t
+vtpci_cryptodev_isr(struct virtio_crypto_hw *hw)
+{
+ return VTPCI_OPS(hw)->get_isr(hw);
+}
+
+static void *
+get_cfg_addr(struct rte_pci_device *dev, struct virtio_pci_cap *cap)
+{
+ uint8_t bar = cap->bar;
+ uint32_t length = cap->length;
+ uint32_t offset = cap->offset;
+ uint8_t *base;
+
+ if (bar >= PCI_MAX_RESOURCE) {
+ PMD_INIT_LOG(ERR, "invalid bar: %u", bar);
+ return NULL;
+ }
+
+ if (offset + length < offset) {
+ PMD_INIT_LOG(ERR, "offset(%u) + length(%u) overflows",
+ offset, length);
+ return NULL;
+ }
+
+ if (offset + length > dev->mem_resource[bar].len) {
+ PMD_INIT_LOG(ERR,
+ "invalid cap: overflows bar space: %u > %" PRIu64,
+ offset + length, dev->mem_resource[bar].len);
+ return NULL;
+ }
+
+ base = dev->mem_resource[bar].addr;
+ if (base == NULL) {
+ PMD_INIT_LOG(ERR, "bar %u base addr is NULL", bar);
+ return NULL;
+ }
+
+ return base + offset;
+}
+
+#define PCI_MSIX_ENABLE 0x8000
+
+static int
+virtio_read_caps(struct rte_pci_device *dev, struct virtio_crypto_hw *hw)
+{
+ uint8_t pos;
+ struct virtio_pci_cap cap;
+ int ret;
+
+ if (rte_pci_map_device(dev)) {
+ PMD_INIT_LOG(DEBUG, "failed to map pci device!");
+ return -1;
+ }
+
+ ret = rte_pci_read_config(dev, &pos, 1, PCI_CAPABILITY_LIST);
+ if (ret < 0) {
+ PMD_INIT_LOG(DEBUG, "failed to read pci capability list");
+ return -1;
+ }
+
+ while (pos) {
+ ret = rte_pci_read_config(dev, &cap, sizeof(cap), pos);
+ if (ret < 0) {
+ PMD_INIT_LOG(ERR,
+ "failed to read pci cap at pos: %x", pos);
+ break;
+ }
+
+ if (cap.cap_vndr == PCI_CAP_ID_MSIX) {
+ /* Transitional devices would also have this capability,
+ * that's why we also check if msix is enabled.
+ * 1st byte is cap ID; 2nd byte is the position of next
+ * cap; next two bytes are the flags.
+ */
+ uint16_t flags = ((uint16_t *)&cap)[1];
+
+ if (flags & PCI_MSIX_ENABLE)
+ hw->use_msix = VIRTIO_MSIX_ENABLED;
+ else
+ hw->use_msix = VIRTIO_MSIX_DISABLED;
+ }
+
+ if (cap.cap_vndr != PCI_CAP_ID_VNDR) {
+ PMD_INIT_LOG(DEBUG,
+ "[%2x] skipping non VNDR cap id: %02x",
+ pos, cap.cap_vndr);
+ goto next;
+ }
+
+ PMD_INIT_LOG(DEBUG,
+ "[%2x] cfg type: %u, bar: %u, offset: %04x, len: %u",
+ pos, cap.cfg_type, cap.bar, cap.offset, cap.length);
+
+ switch (cap.cfg_type) {
+ case VIRTIO_PCI_CAP_COMMON_CFG:
+ hw->common_cfg = get_cfg_addr(dev, &cap);
+ break;
+ case VIRTIO_PCI_CAP_NOTIFY_CFG:
+ rte_pci_read_config(dev, &hw->notify_off_multiplier,
+ 4, pos + sizeof(cap));
+ hw->notify_base = get_cfg_addr(dev, &cap);
+ break;
+ case VIRTIO_PCI_CAP_DEVICE_CFG:
+ hw->dev_cfg = get_cfg_addr(dev, &cap);
+ break;
+ case VIRTIO_PCI_CAP_ISR_CFG:
+ hw->isr = get_cfg_addr(dev, &cap);
+ break;
+ }
+
+next:
+ pos = cap.cap_next;
+ }
+
+ if (hw->common_cfg == NULL || hw->notify_base == NULL ||
+ hw->dev_cfg == NULL || hw->isr == NULL) {
+ PMD_INIT_LOG(INFO, "no modern virtio pci device found.");
+ return -1;
+ }
+
+ PMD_INIT_LOG(INFO, "found modern virtio pci device.");
+
+ PMD_INIT_LOG(DEBUG, "common cfg mapped at: %p", hw->common_cfg);
+ PMD_INIT_LOG(DEBUG, "device cfg mapped at: %p", hw->dev_cfg);
+ PMD_INIT_LOG(DEBUG, "isr cfg mapped at: %p", hw->isr);
+ PMD_INIT_LOG(DEBUG, "notify base: %p, notify off multiplier: %u",
+ hw->notify_base, hw->notify_off_multiplier);
+
+ return 0;
+}
+
+/*
+ * Return -1:
+ * if there is error mapping with VFIO/UIO.
+ * if port map error when driver type is KDRV_NONE.
+ * if whitelisted but driver type is KDRV_UNKNOWN.
+ * Return 1 if kernel driver is managing the device.
+ * Return 0 on success.
+ */
+int
+vtpci_cryptodev_init(struct rte_pci_device *dev, struct virtio_crypto_hw *hw)
+{
+ /*
+ * Try if we can succeed reading virtio pci caps, which exists
+ * only on modern pci device. If failed, we fallback to legacy
+ * virtio handling.
+ */
+ if (virtio_read_caps(dev, hw) == 0) {
+ PMD_INIT_LOG(INFO, "modern virtio pci detected.");
+ virtio_hw_internal[hw->dev_id].vtpci_ops =
+ &virtio_crypto_modern_ops;
+ hw->modern = 1;
+ return 0;
+ }
+
+ /*
+ * virtio crypto conforms to virtio 1.0 and doesn't support
+ * legacy mode
+ */
+ return -1;
+}
diff --git a/drivers/crypto/virtio/virtio_pci.h b/drivers/crypto/virtio/virtio_pci.h
new file mode 100644
index 0000000..a469ea3
--- /dev/null
+++ b/drivers/crypto/virtio/virtio_pci.h
@@ -0,0 +1,252 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 HUAWEI TECHNOLOGIES CO., LTD.
+ */
+
+#ifndef _VIRTIO_PCI_H_
+#define _VIRTIO_PCI_H_
+
+#include <linux/virtio_crypto.h>
+
+#include <stdint.h>
+
+#include <rte_pci.h>
+#include <rte_bus_pci.h>
+#include <rte_cryptodev.h>
+
+struct virtqueue;
+
+/* VirtIO PCI vendor/device ID. */
+#define VIRTIO_CRYPTO_PCI_VENDORID 0x1AF4
+#define VIRTIO_CRYPTO_PCI_DEVICEID 0x1054
+
+/* VirtIO ABI version, this must match exactly. */
+#define VIRTIO_PCI_ABI_VERSION 0
+
+/*
+ * VirtIO Header, located in BAR 0.
+ */
+#define VIRTIO_PCI_HOST_FEATURES 0 /* host's supported features (32bit, RO)*/
+#define VIRTIO_PCI_GUEST_FEATURES 4 /* guest's supported features (32, RW) */
+#define VIRTIO_PCI_QUEUE_PFN 8 /* physical address of VQ (32, RW) */
+#define VIRTIO_PCI_QUEUE_NUM 12 /* number of ring entries (16, RO) */
+#define VIRTIO_PCI_QUEUE_SEL 14 /* current VQ selection (16, RW) */
+#define VIRTIO_PCI_QUEUE_NOTIFY 16 /* notify host regarding VQ (16, RW) */
+#define VIRTIO_PCI_STATUS 18 /* device status register (8, RW) */
+#define VIRTIO_PCI_ISR 19 /* interrupt status register, reading
+ * also clears the register (8, RO)
+ */
+/* Only if MSIX is enabled: */
+
+/* configuration change vector (16, RW) */
+#define VIRTIO_MSI_CONFIG_VECTOR 20
+/* vector for selected VQ notifications */
+#define VIRTIO_MSI_QUEUE_VECTOR 22
+
+/* The bit of the ISR which indicates a device has an interrupt. */
+#define VIRTIO_PCI_ISR_INTR 0x1
+/* The bit of the ISR which indicates a device configuration change. */
+#define VIRTIO_PCI_ISR_CONFIG 0x2
+/* Vector value used to disable MSI for queue. */
+#define VIRTIO_MSI_NO_VECTOR 0xFFFF
+
+/* Status byte for guest to report progress. */
+#define VIRTIO_CONFIG_STATUS_RESET 0x00
+#define VIRTIO_CONFIG_STATUS_ACK 0x01
+#define VIRTIO_CONFIG_STATUS_DRIVER 0x02
+#define VIRTIO_CONFIG_STATUS_DRIVER_OK 0x04
+#define VIRTIO_CONFIG_STATUS_FEATURES_OK 0x08
+#define VIRTIO_CONFIG_STATUS_FAILED 0x80
+
+/*
+ * Each virtqueue indirect descriptor list must be physically contiguous.
+ * To allow us to malloc(9) each list individually, limit the number
+ * supported to what will fit in one page. With 4KB pages, this is a limit
+ * of 256 descriptors. If there is ever a need for more, we can switch to
+ * contigmalloc(9) for the larger allocations, similar to what
+ * bus_dmamem_alloc(9) does.
+ *
+ * Note the sizeof(struct vring_desc) is 16 bytes.
+ */
+#define VIRTIO_MAX_INDIRECT ((int) (PAGE_SIZE / 16))
+
+/* Do we get callbacks when the ring is completely used, even if we've
+ * suppressed them?
+ */
+#define VIRTIO_F_NOTIFY_ON_EMPTY 24
+
+/* Can the device handle any descriptor layout? */
+#define VIRTIO_F_ANY_LAYOUT 27
+
+/* We support indirect buffer descriptors */
+#define VIRTIO_RING_F_INDIRECT_DESC 28
+
+#define VIRTIO_F_VERSION_1 32
+#define VIRTIO_F_IOMMU_PLATFORM 33
+
+/* The Guest publishes the used index for which it expects an interrupt
+ * at the end of the avail ring. Host should ignore the avail->flags field.
+ */
+/* The Host publishes the avail index for which it expects a kick
+ * at the end of the used ring. Guest should ignore the used->flags field.
+ */
+#define VIRTIO_RING_F_EVENT_IDX 29
+
+/* Common configuration */
+#define VIRTIO_PCI_CAP_COMMON_CFG 1
+/* Notifications */
+#define VIRTIO_PCI_CAP_NOTIFY_CFG 2
+/* ISR Status */
+#define VIRTIO_PCI_CAP_ISR_CFG 3
+/* Device specific configuration */
+#define VIRTIO_PCI_CAP_DEVICE_CFG 4
+/* PCI configuration access */
+#define VIRTIO_PCI_CAP_PCI_CFG 5
+
+/* This is the PCI capability header: */
+struct virtio_pci_cap {
+ uint8_t cap_vndr; /* Generic PCI field: PCI_CAP_ID_VNDR */
+ uint8_t cap_next; /* Generic PCI field: next ptr. */
+ uint8_t cap_len; /* Generic PCI field: capability length */
+ uint8_t cfg_type; /* Identifies the structure. */
+ uint8_t bar; /* Where to find it. */
+ uint8_t padding[3]; /* Pad to full dword. */
+ uint32_t offset; /* Offset within bar. */
+ uint32_t length; /* Length of the structure, in bytes. */
+};
+
+struct virtio_pci_notify_cap {
+ struct virtio_pci_cap cap;
+ uint32_t notify_off_multiplier; /* Multiplier for queue_notify_off. */
+};
+
+/* Fields in VIRTIO_PCI_CAP_COMMON_CFG: */
+struct virtio_pci_common_cfg {
+ /* About the whole device. */
+ uint32_t device_feature_select; /* read-write */
+ uint32_t device_feature; /* read-only */
+ uint32_t guest_feature_select; /* read-write */
+ uint32_t guest_feature; /* read-write */
+ uint16_t msix_config; /* read-write */
+ uint16_t num_queues; /* read-only */
+ uint8_t device_status; /* read-write */
+ uint8_t config_generation; /* read-only */
+
+ /* About a specific virtqueue. */
+ uint16_t queue_select; /* read-write */
+ uint16_t queue_size; /* read-write, power of 2. */
+ uint16_t queue_msix_vector; /* read-write */
+ uint16_t queue_enable; /* read-write */
+ uint16_t queue_notify_off; /* read-only */
+ uint32_t queue_desc_lo; /* read-write */
+ uint32_t queue_desc_hi; /* read-write */
+ uint32_t queue_avail_lo; /* read-write */
+ uint32_t queue_avail_hi; /* read-write */
+ uint32_t queue_used_lo; /* read-write */
+ uint32_t queue_used_hi; /* read-write */
+};
+
+struct virtio_crypto_hw;
+
+struct virtio_pci_ops {
+ void (*read_dev_cfg)(struct virtio_crypto_hw *hw, size_t offset,
+ void *dst, int len);
+ void (*write_dev_cfg)(struct virtio_crypto_hw *hw, size_t offset,
+ const void *src, int len);
+ void (*reset)(struct virtio_crypto_hw *hw);
+
+ uint8_t (*get_status)(struct virtio_crypto_hw *hw);
+ void (*set_status)(struct virtio_crypto_hw *hw, uint8_t status);
+
+ uint64_t (*get_features)(struct virtio_crypto_hw *hw);
+ void (*set_features)(struct virtio_crypto_hw *hw, uint64_t features);
+
+ uint8_t (*get_isr)(struct virtio_crypto_hw *hw);
+
+ uint16_t (*set_config_irq)(struct virtio_crypto_hw *hw, uint16_t vec);
+
+ uint16_t (*set_queue_irq)(struct virtio_crypto_hw *hw,
+ struct virtqueue *vq, uint16_t vec);
+
+ uint16_t (*get_queue_num)(struct virtio_crypto_hw *hw,
+ uint16_t queue_id);
+ int (*setup_queue)(struct virtio_crypto_hw *hw, struct virtqueue *vq);
+ void (*del_queue)(struct virtio_crypto_hw *hw, struct virtqueue *vq);
+ void (*notify_queue)(struct virtio_crypto_hw *hw, struct virtqueue *vq);
+};
+
+struct virtio_crypto_hw {
+ /* control queue */
+ struct virtqueue *cvq;
+ uint16_t dev_id;
+ uint16_t max_dataqueues;
+ uint64_t req_guest_features;
+ uint64_t guest_features;
+ uint8_t use_msix;
+ uint8_t modern;
+ uint32_t notify_off_multiplier;
+ uint8_t *isr;
+ uint16_t *notify_base;
+ struct virtio_pci_common_cfg *common_cfg;
+ struct virtio_crypto_config *dev_cfg;
+};
+
+/*
+ * While virtio_crypto_hw is stored in shared memory, this structure stores
+ * some infos that may vary in the multiple process model locally.
+ * For example, the vtpci_ops pointer.
+ */
+struct virtio_hw_internal {
+ const struct virtio_pci_ops *vtpci_ops;
+ struct rte_pci_ioport io;
+};
+
+#define VTPCI_OPS(hw) (virtio_hw_internal[(hw)->dev_id].vtpci_ops)
+#define VTPCI_IO(hw) (&virtio_hw_internal[(hw)->dev_id].io)
+
+extern struct virtio_hw_internal virtio_hw_internal[RTE_MAX_VIRTIO_CRYPTO];
+
+/*
+ * How many bits to shift physical queue address written to QUEUE_PFN.
+ * 12 is historical, and due to x86 page size.
+ */
+#define VIRTIO_PCI_QUEUE_ADDR_SHIFT 12
+
+/* The alignment to use between consumer and producer parts of vring. */
+#define VIRTIO_PCI_VRING_ALIGN 4096
+
+enum virtio_msix_status {
+ VIRTIO_MSIX_NONE = 0,
+ VIRTIO_MSIX_DISABLED = 1,
+ VIRTIO_MSIX_ENABLED = 2
+};
+
+static inline int
+vtpci_with_feature(struct virtio_crypto_hw *hw, uint64_t bit)
+{
+ return (hw->guest_features & (1ULL << bit)) != 0;
+}
+
+/*
+ * Function declaration from virtio_pci.c
+ */
+int vtpci_cryptodev_init(struct rte_pci_device *dev,
+ struct virtio_crypto_hw *hw);
+void vtpci_cryptodev_reset(struct virtio_crypto_hw *hw);
+
+void vtpci_cryptodev_reinit_complete(struct virtio_crypto_hw *hw);
+
+uint8_t vtpci_cryptodev_get_status(struct virtio_crypto_hw *hw);
+void vtpci_cryptodev_set_status(struct virtio_crypto_hw *hw, uint8_t status);
+
+uint64_t vtpci_cryptodev_negotiate_features(struct virtio_crypto_hw *hw,
+ uint64_t host_features);
+
+void vtpci_write_cryptodev_config(struct virtio_crypto_hw *hw, size_t offset,
+ const void *src, int length);
+
+void vtpci_read_cryptodev_config(struct virtio_crypto_hw *hw, size_t offset,
+ void *dst, int length);
+
+uint8_t vtpci_cryptodev_isr(struct virtio_crypto_hw *hw);
+
+#endif /* _VIRTIO_PCI_H_ */
diff --git a/drivers/crypto/virtio/virtio_ring.h b/drivers/crypto/virtio/virtio_ring.h
new file mode 100644
index 0000000..ee30674
--- /dev/null
+++ b/drivers/crypto/virtio/virtio_ring.h
@@ -0,0 +1,137 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 HUAWEI TECHNOLOGIES CO., LTD.
+ */
+
+#ifndef _VIRTIO_RING_H_
+#define _VIRTIO_RING_H_
+
+#include <stdint.h>
+
+#include <rte_common.h>
+
+/* This marks a buffer as continuing via the next field. */
+#define VRING_DESC_F_NEXT 1
+/* This marks a buffer as write-only (otherwise read-only). */
+#define VRING_DESC_F_WRITE 2
+/* This means the buffer contains a list of buffer descriptors. */
+#define VRING_DESC_F_INDIRECT 4
+
+/* The Host uses this in used->flags to advise the Guest: don't kick me
+ * when you add a buffer. It's unreliable, so it's simply an
+ * optimization. Guest will still kick if it's out of buffers.
+ */
+#define VRING_USED_F_NO_NOTIFY 1
+/* The Guest uses this in avail->flags to advise the Host: don't
+ * interrupt me when you consume a buffer. It's unreliable, so it's
+ * simply an optimization.
+ */
+#define VRING_AVAIL_F_NO_INTERRUPT 1
+
+/* VirtIO ring descriptors: 16 bytes.
+ * These can chain together via "next".
+ */
+struct vring_desc {
+ uint64_t addr; /* Address (guest-physical). */
+ uint32_t len; /* Length. */
+ uint16_t flags; /* The flags as indicated above. */
+ uint16_t next; /* We chain unused descriptors via this. */
+};
+
+struct vring_avail {
+ uint16_t flags;
+ uint16_t idx;
+ uint16_t ring[0];
+};
+
+/* id is a 16bit index. uint32_t is used here for ids for padding reasons. */
+struct vring_used_elem {
+ /* Index of start of used descriptor chain. */
+ uint32_t id;
+ /* Total length of the descriptor chain which was written to. */
+ uint32_t len;
+};
+
+struct vring_used {
+ uint16_t flags;
+ volatile uint16_t idx;
+ struct vring_used_elem ring[0];
+};
+
+struct vring {
+ unsigned int num;
+ struct vring_desc *desc;
+ struct vring_avail *avail;
+ struct vring_used *used;
+};
+
+/* The standard layout for the ring is a continuous chunk of memory which
+ * looks like this. We assume num is a power of 2.
+ *
+ * struct vring {
+ * // The actual descriptors (16 bytes each)
+ * struct vring_desc desc[num];
+ *
+ * // A ring of available descriptor heads with free-running index.
+ * __u16 avail_flags;
+ * __u16 avail_idx;
+ * __u16 available[num];
+ * __u16 used_event_idx;
+ *
+ * // Padding to the next align boundary.
+ * char pad[];
+ *
+ * // A ring of used descriptor heads with free-running index.
+ * __u16 used_flags;
+ * __u16 used_idx;
+ * struct vring_used_elem used[num];
+ * __u16 avail_event_idx;
+ * };
+ *
+ * NOTE: for VirtIO PCI, align is 4096.
+ */
+
+/*
+ * We publish the used event index at the end of the available ring, and vice
+ * versa. They are at the end for backwards compatibility.
+ */
+#define vring_used_event(vr) ((vr)->avail->ring[(vr)->num])
+#define vring_avail_event(vr) (*(uint16_t *)&(vr)->used->ring[(vr)->num])
+
+static inline size_t
+vring_size(unsigned int num, unsigned long align)
+{
+ size_t size;
+
+ size = num * sizeof(struct vring_desc);
+ size += sizeof(struct vring_avail) + (num * sizeof(uint16_t));
+ size = RTE_ALIGN_CEIL(size, align);
+ size += sizeof(struct vring_used) +
+ (num * sizeof(struct vring_used_elem));
+ return size;
+}
+
+static inline void
+vring_init(struct vring *vr, unsigned int num, uint8_t *p,
+ unsigned long align)
+{
+ vr->num = num;
+ vr->desc = (struct vring_desc *) p;
+ vr->avail = (struct vring_avail *) (p +
+ num * sizeof(struct vring_desc));
+ vr->used = (void *)
+ RTE_ALIGN_CEIL((uintptr_t)(&vr->avail->ring[num]), align);
+}
+
+/*
+ * The following is used with VIRTIO_RING_F_EVENT_IDX.
+ * Assuming a given event_idx value from the other size, if we have
+ * just incremented index from old to new_idx, should we trigger an
+ * event?
+ */
+static inline int
+vring_need_event(uint16_t event_idx, uint16_t new_idx, uint16_t old)
+{
+ return (uint16_t)(new_idx - event_idx - 1) < (uint16_t)(new_idx - old);
+}
+
+#endif /* _VIRTIO_RING_H_ */
diff --git a/drivers/crypto/virtio/virtqueue.c b/drivers/crypto/virtio/virtqueue.c
new file mode 100644
index 0000000..fd8be58
--- /dev/null
+++ b/drivers/crypto/virtio/virtqueue.c
@@ -0,0 +1,43 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 HUAWEI TECHNOLOGIES CO., LTD.
+ */
+
+#include <stdint.h>
+
+#include <rte_mbuf.h>
+#include <rte_crypto.h>
+#include <rte_malloc.h>
+
+#include "virtqueue.h"
+
+void
+virtqueue_disable_intr(struct virtqueue *vq)
+{
+ /*
+ * Set VRING_AVAIL_F_NO_INTERRUPT to hint host
+ * not to interrupt when it consumes packets
+ * Note: this is only considered a hint to the host
+ */
+ vq->vq_ring.avail->flags |= VRING_AVAIL_F_NO_INTERRUPT;
+}
+
+void
+virtqueue_detatch_unused(struct virtqueue *vq)
+{
+ struct rte_crypto_op *cop = NULL;
+
+ int idx;
+
+ if (vq != NULL)
+ for (idx = 0; idx < vq->vq_nentries; idx++) {
+ cop = vq->vq_descx[idx].crypto_op;
+ if (cop) {
+ if (cop->sym->m_src)
+ rte_pktmbuf_free(cop->sym->m_src);
+ if (cop->sym->m_dst)
+ rte_pktmbuf_free(cop->sym->m_dst);
+ rte_crypto_op_free(cop);
+ vq->vq_descx[idx].crypto_op = NULL;
+ }
+ }
+}
diff --git a/drivers/crypto/virtio/virtqueue.h b/drivers/crypto/virtio/virtqueue.h
new file mode 100644
index 0000000..1bd0e89
--- /dev/null
+++ b/drivers/crypto/virtio/virtqueue.h
@@ -0,0 +1,176 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 HUAWEI TECHNOLOGIES CO., LTD.
+ */
+
+#ifndef _VIRTQUEUE_H_
+#define _VIRTQUEUE_H_
+
+#include <linux/virtio_crypto.h>
+
+#include <stdint.h>
+
+#include <rte_atomic.h>
+#include <rte_memory.h>
+#include <rte_memzone.h>
+#include <rte_mempool.h>
+
+#include "virtio_pci.h"
+#include "virtio_ring.h"
+#include "virtio_logs.h"
+
+struct rte_mbuf;
+
+/*
+ * Per virtio_config.h in Linux.
+ * For virtio_pci on SMP, we don't need to order with respect to MMIO
+ * accesses through relaxed memory I/O windows, so smp_mb() et al are
+ * sufficient.
+ *
+ */
+#define virtio_mb() rte_smp_mb()
+#define virtio_rmb() rte_smp_rmb()
+#define virtio_wmb() rte_smp_wmb()
+
+#define VIRTQUEUE_MAX_NAME_SZ 32
+
+enum { VTCRYPTO_DATAQ = 0, VTCRYPTO_CTRLQ = 1 };
+
+/**
+ * The maximum virtqueue size is 2^15. Use that value as the end of
+ * descriptor chain terminator since it will never be a valid index
+ * in the descriptor table. This is used to verify we are correctly
+ * handling vq_free_cnt.
+ */
+#define VQ_RING_DESC_CHAIN_END 32768
+
+struct vq_desc_extra {
+ void *crypto_op;
+ void *cookie;
+ uint16_t ndescs;
+};
+
+struct virtqueue {
+ /**< virtio_crypto_hw structure pointer. */
+ struct virtio_crypto_hw *hw;
+ /**< mem zone to populate RX ring. */
+ const struct rte_memzone *mz;
+ /**< memzone to populate hdr and request. */
+ struct rte_mempool *mpool;
+ uint8_t dev_id; /**< Device identifier. */
+ uint16_t vq_queue_index; /**< PCI queue index */
+
+ void *vq_ring_virt_mem; /**< linear address of vring*/
+ unsigned int vq_ring_size;
+ phys_addr_t vq_ring_mem; /**< physical address of vring */
+
+ struct vring vq_ring; /**< vring keeping desc, used and avail */
+ uint16_t vq_free_cnt; /**< num of desc available */
+ uint16_t vq_nentries; /**< vring desc numbers */
+
+ /**
+ * Head of the free chain in the descriptor table. If
+ * there are no free descriptors, this will be set to
+ * VQ_RING_DESC_CHAIN_END.
+ */
+ uint16_t vq_desc_head_idx;
+ uint16_t vq_desc_tail_idx;
+ /**
+ * Last consumed descriptor in the used table,
+ * trails vq_ring.used->idx.
+ */
+ uint16_t vq_used_cons_idx;
+ uint16_t vq_avail_idx;
+
+ /* Statistics */
+ uint64_t packets_sent_total;
+ uint64_t packets_sent_failed;
+ uint64_t packets_received_total;
+ uint64_t packets_received_failed;
+
+ uint16_t *notify_addr;
+
+ struct vq_desc_extra vq_descx[0];
+};
+
+/**
+ * Tell the backend not to interrupt us.
+ */
+void virtqueue_disable_intr(struct virtqueue *vq);
+
+/**
+ * Get all mbufs to be freed.
+ */
+void virtqueue_detatch_unused(struct virtqueue *vq);
+
+static inline int
+virtqueue_full(const struct virtqueue *vq)
+{
+ return vq->vq_free_cnt == 0;
+}
+
+#define VIRTQUEUE_NUSED(vq) \
+ ((uint16_t)((vq)->vq_ring.used->idx - (vq)->vq_used_cons_idx))
+
+static inline void
+vq_update_avail_idx(struct virtqueue *vq)
+{
+ virtio_wmb();
+ vq->vq_ring.avail->idx = vq->vq_avail_idx;
+}
+
+static inline void
+vq_update_avail_ring(struct virtqueue *vq, uint16_t desc_idx)
+{
+ uint16_t avail_idx;
+ /*
+ * Place the head of the descriptor chain into the next slot and make
+ * it usable to the host. The chain is made available now rather than
+ * deferring to virtqueue_notify() in the hopes that if the host is
+ * currently running on another CPU, we can keep it processing the new
+ * descriptor.
+ */
+ avail_idx = (uint16_t)(vq->vq_avail_idx & (vq->vq_nentries - 1));
+ if (unlikely(vq->vq_ring.avail->ring[avail_idx] != desc_idx))
+ vq->vq_ring.avail->ring[avail_idx] = desc_idx;
+ vq->vq_avail_idx++;
+}
+
+static inline int
+virtqueue_kick_prepare(struct virtqueue *vq)
+{
+ return !(vq->vq_ring.used->flags & VRING_USED_F_NO_NOTIFY);
+}
+
+static inline void
+virtqueue_notify(struct virtqueue *vq)
+{
+ /*
+ * Ensure updated avail->idx is visible to host.
+ * For virtio on IA, the notificaiton is through io port operation
+ * which is a serialization instruction itself.
+ */
+ VTPCI_OPS(vq->hw)->notify_queue(vq->hw, vq);
+}
+
+/**
+ * Dump virtqueue internal structures, for debug purpose only.
+ */
+#ifdef RTE_LIBRTE_PMD_VIRTIO_CRYPTO_DEBUG_DUMP
+#define VIRTQUEUE_DUMP(vq) do { \
+ uint16_t used_idx, nused; \
+ used_idx = (vq)->vq_ring.used->idx; \
+ nused = (uint16_t)(used_idx - (vq)->vq_used_cons_idx); \
+ PMD_INIT_LOG(DEBUG, \
+ "VQ: - size=%d; free=%d; used=%d; desc_head_idx=%d;" \
+ " avail.idx=%d; used_cons_idx=%d; used.idx=%d;" \
+ " avail.flags=0x%x; used.flags=0x%x", \
+ (vq)->vq_nentries, (vq)->vq_free_cnt, nused, \
+ (vq)->vq_desc_head_idx, (vq)->vq_ring.avail->idx, \
+ (vq)->vq_used_cons_idx, (vq)->vq_ring.used->idx, \
+ (vq)->vq_ring.avail->flags, (vq)->vq_ring.used->flags); \
+} while (0)
+#else
+#define VIRTQUEUE_DUMP(vq) do { } while (0)
+#endif
+
+#endif /* _VIRTQUEUE_H_ */
--
1.8.3.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* [dpdk-dev] [PATCH v2 2/7] crpyto/virtio: add crypto related session structure
2018-02-24 13:14 [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 1/7] crypto/virtio: add virtio related fundamental functions Jay Zhou
@ 2018-02-24 13:14 ` Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 3/7] cryptodev/virtio: core code of crypto devices Jay Zhou
` (5 subsequent siblings)
7 siblings, 0 replies; 11+ messages in thread
From: Jay Zhou @ 2018-02-24 13:14 UTC (permalink / raw)
To: dev
Cc: pablo.de.lara.guarch, roy.fan.zhang, thomas, arei.gonglei,
xin.zeng, weidong.huang, wangxinxin.wang, longpeng2,
jianjay.zhou
This structure will be used in the following patches, especially
at creating and destroying crypto sessions.
Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com>
---
drivers/crypto/virtio/virtio_crypto_algs.h | 27 +++++++++++++++++++++++++++
1 file changed, 27 insertions(+)
create mode 100644 drivers/crypto/virtio/virtio_crypto_algs.h
diff --git a/drivers/crypto/virtio/virtio_crypto_algs.h b/drivers/crypto/virtio/virtio_crypto_algs.h
new file mode 100644
index 0000000..5f1e9df
--- /dev/null
+++ b/drivers/crypto/virtio/virtio_crypto_algs.h
@@ -0,0 +1,27 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 HUAWEI TECHNOLOGIES CO., LTD.
+ */
+
+#ifndef _VIRTIO_CRYPTO_ALGS_H_
+#define _VIRTIO_CRYPTO_ALGS_H_
+
+#include <rte_memory.h>
+#include <linux/virtio_crypto.h>
+
+struct virtio_crypto_session {
+ uint64_t session_id;
+
+ struct {
+ uint16_t offset;
+ uint16_t length;
+ } iv;
+
+ struct {
+ uint32_t length;
+ phys_addr_t phys_addr;
+ } aad;
+
+ struct virtio_crypto_op_ctrl_req ctrl;
+};
+
+#endif /* _VIRTIO_CRYPTO_ALGS_H_ */
--
1.8.3.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* [dpdk-dev] [PATCH v2 3/7] cryptodev/virtio: core code of crypto devices
2018-02-24 13:14 [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 1/7] crypto/virtio: add virtio related fundamental functions Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 2/7] crpyto/virtio: add crypto related session structure Jay Zhou
@ 2018-02-24 13:14 ` Jay Zhou
2018-03-21 4:08 ` Zhang, Roy Fan
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 4/7] crypto/virtio: add makefile Jay Zhou
` (4 subsequent siblings)
7 siblings, 1 reply; 11+ messages in thread
From: Jay Zhou @ 2018-02-24 13:14 UTC (permalink / raw)
To: dev
Cc: pablo.de.lara.guarch, roy.fan.zhang, thomas, arei.gonglei,
xin.zeng, weidong.huang, wangxinxin.wang, longpeng2,
jianjay.zhou
The idea comes from QAT and virtio-net devices.
Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com>
---
drivers/crypto/virtio/virtio_cryptodev.c | 1544 ++++++++++++++++++++++++++++++
drivers/crypto/virtio/virtio_cryptodev.h | 66 ++
drivers/crypto/virtio/virtio_rxtx.c | 533 +++++++++++
3 files changed, 2143 insertions(+)
create mode 100644 drivers/crypto/virtio/virtio_cryptodev.c
create mode 100644 drivers/crypto/virtio/virtio_cryptodev.h
create mode 100644 drivers/crypto/virtio/virtio_rxtx.c
diff --git a/drivers/crypto/virtio/virtio_cryptodev.c b/drivers/crypto/virtio/virtio_cryptodev.c
new file mode 100644
index 0000000..8d84526
--- /dev/null
+++ b/drivers/crypto/virtio/virtio_cryptodev.c
@@ -0,0 +1,1544 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 HUAWEI TECHNOLOGIES CO., LTD.
+ */
+
+#include <stdint.h>
+#include <string.h>
+#include <stdio.h>
+#include <stdbool.h>
+#include <errno.h>
+#include <unistd.h>
+#ifdef RTE_EXEC_ENV_LINUXAPP
+#include <dirent.h>
+#include <fcntl.h>
+#endif
+
+#include <rte_cryptodev.h>
+#include <rte_cryptodev_pmd.h>
+#include <rte_memcpy.h>
+#include <rte_string_fns.h>
+#include <rte_memzone.h>
+#include <rte_malloc.h>
+#include <rte_atomic.h>
+#include <rte_branch_prediction.h>
+#include <rte_pci.h>
+#include <rte_common.h>
+#include <rte_errno.h>
+
+#include <rte_memory.h>
+#include <rte_eal.h>
+#include <rte_dev.h>
+#include <rte_log.h>
+
+#include "virtio_cryptodev.h"
+#include "virtqueue.h"
+#include "virtio_crypto_algs.h"
+
+static int virtio_crypto_dev_configure(struct rte_cryptodev *dev,
+ struct rte_cryptodev_config *config);
+static int virtio_crypto_dev_start(struct rte_cryptodev *dev);
+static void virtio_crypto_dev_stop(struct rte_cryptodev *dev);
+static int virtio_crypto_dev_close(struct rte_cryptodev *dev);
+static void virtio_crypto_dev_info_get(struct rte_cryptodev *dev,
+ struct rte_cryptodev_info *dev_info);
+static void virtio_crypto_dev_stats_get(struct rte_cryptodev *dev,
+ struct rte_cryptodev_stats *stats);
+static void virtio_crypto_dev_stats_reset(struct rte_cryptodev *dev);
+static int virtio_crypto_qp_setup(struct rte_cryptodev *dev,
+ uint16_t queue_pair_id,
+ const struct rte_cryptodev_qp_conf *qp_conf,
+ int socket_id,
+ struct rte_mempool *session_pool);
+static int virtio_crypto_qp_release(struct rte_cryptodev *dev,
+ uint16_t queue_pair_id);
+static void virtio_crypto_dev_free_mbufs(struct rte_cryptodev *dev);
+static unsigned int virtio_crypto_sym_get_session_private_size(
+ struct rte_cryptodev *dev);
+static void virtio_crypto_sym_clear_session(struct rte_cryptodev *dev,
+ struct rte_cryptodev_sym_session *sess);
+static int virtio_crypto_sym_configure_session(struct rte_cryptodev *dev,
+ struct rte_crypto_sym_xform *xform,
+ struct rte_cryptodev_sym_session *session,
+ struct rte_mempool *mp);
+
+/*
+ * The set of PCI devices this driver supports
+ */
+static const struct rte_pci_id pci_id_virtio_crypto_map[] = {
+ { RTE_PCI_DEVICE(VIRTIO_CRYPTO_PCI_VENDORID,
+ VIRTIO_CRYPTO_PCI_DEVICEID) },
+ { .vendor_id = 0, /* sentinel */ },
+};
+
+uint8_t cryptodev_virtio_driver_id;
+
+#define NUM_ENTRY_SYM_CREATE_SESSION 4
+
+static int
+virtio_crypto_send_command(struct virtqueue *vq,
+ struct virtio_crypto_op_ctrl_req *ctrl, uint8_t *cipher_key,
+ uint8_t *auth_key, struct virtio_crypto_session *session)
+{
+ uint8_t idx = 0;
+ uint8_t needed = 1;
+ uint32_t head = 0;
+ uint32_t len_cipher_key = 0;
+ uint32_t len_auth_key = 0;
+ uint32_t len_ctrl_req = sizeof(struct virtio_crypto_op_ctrl_req);
+ uint32_t len_session_input = sizeof(struct virtio_crypto_session_input);
+ uint32_t len_total = 0;
+ uint32_t input_offset = 0;
+ void *virt_addr_started = NULL;
+ phys_addr_t phys_addr_started;
+ struct vring_desc *desc;
+ uint32_t desc_offset;
+ struct virtio_crypto_session_input *input;
+ int ret;
+
+ PMD_INIT_FUNC_TRACE();
+
+ if (session == NULL) {
+ PMD_SESSION_LOG(ERR, "session is NULL.");
+ return -EINVAL;
+ }
+ /* cipher only is supported, it is available if auth_key is NULL */
+ if (!cipher_key) {
+ PMD_SESSION_LOG(ERR, "cipher key is NULL.");
+ return -EINVAL;
+ }
+
+ head = vq->vq_desc_head_idx;
+ PMD_INIT_LOG(DEBUG, "vq->vq_desc_head_idx = %d, vq = %p", head, vq);
+
+ if (vq->vq_free_cnt < needed) {
+ PMD_SESSION_LOG(ERR, "Not enough entry");
+ return -ENOSPC;
+ }
+
+ /* calculate the length of cipher key */
+ if (cipher_key) {
+ switch (ctrl->u.sym_create_session.op_type) {
+ case VIRTIO_CRYPTO_SYM_OP_CIPHER:
+ len_cipher_key
+ = ctrl->u.sym_create_session.u.cipher
+ .para.keylen;
+ break;
+ case VIRTIO_CRYPTO_SYM_OP_ALGORITHM_CHAINING:
+ len_cipher_key
+ = ctrl->u.sym_create_session.u.chain
+ .para.cipher_param.keylen;
+ break;
+ default:
+ PMD_SESSION_LOG(ERR, "invalid op type");
+ return -EINVAL;
+ }
+ }
+
+ /* calculate the length of auth key */
+ if (auth_key) {
+ len_auth_key =
+ ctrl->u.sym_create_session.u.chain.para.u.mac_param
+ .auth_key_len;
+ }
+
+ /*
+ * malloc memory to store indirect vring_desc entries, including
+ * ctrl request, cipher key, auth key, session input and desc vring
+ */
+ desc_offset = len_ctrl_req + len_cipher_key + len_auth_key
+ + len_session_input;
+ virt_addr_started = rte_malloc(NULL,
+ desc_offset + NUM_ENTRY_SYM_CREATE_SESSION
+ * sizeof(struct vring_desc), RTE_CACHE_LINE_SIZE);
+ if (virt_addr_started == NULL) {
+ PMD_SESSION_LOG(ERR, "not enough heap memory");
+ return -ENOSPC;
+ }
+ phys_addr_started = rte_malloc_virt2iova(virt_addr_started);
+
+ /* address to store indirect vring desc entries */
+ desc = (struct vring_desc *)
+ ((uint8_t *)virt_addr_started + desc_offset);
+
+ /* ctrl req part */
+ memcpy(virt_addr_started, ctrl, len_ctrl_req);
+ desc[idx].addr = phys_addr_started;
+ desc[idx].len = len_ctrl_req;
+ desc[idx].flags = VRING_DESC_F_NEXT;
+ desc[idx].next = idx + 1;
+ idx++;
+ len_total += len_ctrl_req;
+ input_offset += len_ctrl_req;
+
+ /* cipher key part */
+ if (len_cipher_key > 0) {
+ memcpy((uint8_t *)virt_addr_started + len_total,
+ cipher_key, len_cipher_key);
+
+ desc[idx].addr = phys_addr_started + len_total;
+ desc[idx].len = len_cipher_key;
+ desc[idx].flags = VRING_DESC_F_NEXT;
+ desc[idx].next = idx + 1;
+ idx++;
+ len_total += len_cipher_key;
+ input_offset += len_cipher_key;
+ }
+
+ /* auth key part */
+ if (len_auth_key > 0) {
+ memcpy((uint8_t *)virt_addr_started + len_total,
+ auth_key, len_auth_key);
+
+ desc[idx].addr = phys_addr_started + len_total;
+ desc[idx].len = len_auth_key;
+ desc[idx].flags = VRING_DESC_F_NEXT;
+ desc[idx].next = idx + 1;
+ idx++;
+ len_total += len_auth_key;
+ input_offset += len_auth_key;
+ }
+
+ /* input part */
+ input = (struct virtio_crypto_session_input *)
+ ((uint8_t *)virt_addr_started + input_offset);
+ input->status = VIRTIO_CRYPTO_ERR;
+ input->session_id = ~0ULL;
+ desc[idx].addr = phys_addr_started + len_total;
+ desc[idx].len = len_session_input;
+ desc[idx].flags = VRING_DESC_F_WRITE;
+ idx++;
+
+ /* use a single desc entry */
+ vq->vq_ring.desc[head].addr = phys_addr_started + desc_offset;
+ vq->vq_ring.desc[head].len = idx * sizeof(struct vring_desc);
+ vq->vq_ring.desc[head].flags = VRING_DESC_F_INDIRECT;
+ vq->vq_free_cnt--;
+
+ vq->vq_desc_head_idx = vq->vq_ring.desc[head].next;
+
+ vq_update_avail_ring(vq, head);
+ vq_update_avail_idx(vq);
+
+ PMD_INIT_LOG(DEBUG, "vq->vq_queue_index = %d", vq->vq_queue_index);
+
+ virtqueue_notify(vq);
+
+ rte_rmb();
+ while (vq->vq_used_cons_idx == vq->vq_ring.used->idx) {
+ rte_rmb();
+ usleep(100);
+ }
+
+ while (vq->vq_used_cons_idx != vq->vq_ring.used->idx) {
+ uint32_t idx, desc_idx, used_idx;
+ struct vring_used_elem *uep;
+
+ used_idx = (uint32_t)(vq->vq_used_cons_idx
+ & (vq->vq_nentries - 1));
+ uep = &vq->vq_ring.used->ring[used_idx];
+ idx = (uint32_t) uep->id;
+ desc_idx = idx;
+
+ while (vq->vq_ring.desc[desc_idx].flags & VRING_DESC_F_NEXT) {
+ desc_idx = vq->vq_ring.desc[desc_idx].next;
+ vq->vq_free_cnt++;
+ }
+
+ vq->vq_ring.desc[desc_idx].next = vq->vq_desc_head_idx;
+ vq->vq_desc_head_idx = idx;
+
+ vq->vq_used_cons_idx++;
+ vq->vq_free_cnt++;
+ }
+
+ PMD_INIT_LOG(DEBUG, "vq->vq_free_cnt=%d\nvq->vq_desc_head_idx=%d",
+ vq->vq_free_cnt, vq->vq_desc_head_idx);
+
+ /* get the result */
+ if (input->status != VIRTIO_CRYPTO_OK) {
+ PMD_SESSION_LOG(ERR, "Something wrong on backend! "
+ "status=%"PRIu32", session_id=%"PRIu64"",
+ input->status, input->session_id);
+ rte_free(virt_addr_started);
+ ret = -1;
+ } else {
+ session->session_id = input->session_id;
+
+ PMD_SESSION_LOG(INFO, "Create session successfully, "
+ "session_id=%"PRIu64"", input->session_id);
+ rte_free(virt_addr_started);
+ ret = 0;
+ }
+
+ return ret;
+}
+
+void virtio_crypto_queue_release(struct virtqueue *vq)
+{
+ struct virtio_crypto_hw *hw;
+
+ PMD_INIT_FUNC_TRACE();
+
+ if (vq) {
+ hw = vq->hw;
+ /* Select and deactivate the queue */
+ VTPCI_OPS(hw)->del_queue(hw, vq);
+
+ rte_memzone_free(vq->mz);
+ rte_mempool_free(vq->mpool);
+ rte_free(vq);
+ }
+}
+
+#define MPOOL_MAX_NAME_SZ 32
+
+int virtio_crypto_queue_setup(struct rte_cryptodev *dev,
+ int queue_type,
+ uint16_t vtpci_queue_idx,
+ uint16_t nb_desc,
+ int socket_id,
+ struct virtqueue **pvq)
+{
+ char vq_name[VIRTQUEUE_MAX_NAME_SZ];
+ char mpool_name[MPOOL_MAX_NAME_SZ];
+ const struct rte_memzone *mz;
+ unsigned int vq_size, size;
+ struct virtio_crypto_hw *hw = dev->data->dev_private;
+ struct virtqueue *vq = NULL;
+ uint32_t i, j;
+
+ PMD_INIT_FUNC_TRACE();
+
+ PMD_INIT_LOG(DEBUG, "setting up queue: %u", vtpci_queue_idx);
+
+ /*
+ * Read the virtqueue size from the Queue Size field
+ * Always power of 2 and if 0 virtqueue does not exist
+ */
+ vq_size = VTPCI_OPS(hw)->get_queue_num(hw, vtpci_queue_idx);
+ if (vq_size == 0) {
+ PMD_INIT_LOG(ERR, "virtqueue does not exist");
+ return -EINVAL;
+ }
+ PMD_INIT_LOG(DEBUG, "vq_size: %u", vq_size);
+
+ if (!rte_is_power_of_2(vq_size)) {
+ PMD_INIT_LOG(ERR, "virtqueue size is not powerof 2");
+ return -EINVAL;
+ }
+
+ if (queue_type == VTCRYPTO_DATAQ) {
+ snprintf(vq_name, sizeof(vq_name), "dev%d_dataqueue%d",
+ dev->data->dev_id, vtpci_queue_idx);
+ snprintf(mpool_name, sizeof(mpool_name), "dev%d_dataqueue%d_mpool",
+ dev->data->dev_id, vtpci_queue_idx);
+ } else if (queue_type == VTCRYPTO_CTRLQ) {
+ snprintf(vq_name, sizeof(vq_name), "dev%d_controlqueue",
+ dev->data->dev_id);
+ snprintf(mpool_name, sizeof(mpool_name), "dev%d_controlqueue_mpool",
+ dev->data->dev_id);
+ }
+ size = RTE_ALIGN_CEIL(sizeof(*vq) +
+ vq_size * sizeof(struct vq_desc_extra),
+ RTE_CACHE_LINE_SIZE);
+ vq = rte_zmalloc_socket(vq_name, size, RTE_CACHE_LINE_SIZE,
+ socket_id);
+ if (vq == NULL) {
+ PMD_INIT_LOG(ERR, "Can not allocate virtqueue");
+ return -ENOMEM;
+ }
+
+ if (queue_type == VTCRYPTO_DATAQ) {
+ /* pre-allocate a mempool and use it in the data plane to
+ * improve performance
+ */
+ vq->mpool = rte_mempool_lookup(mpool_name);
+ if (vq->mpool == NULL)
+ vq->mpool = rte_mempool_create(mpool_name,
+ vq_size,
+ sizeof(struct virtio_crypto_op_cookie),
+ RTE_CACHE_LINE_SIZE, 0,
+ NULL, NULL, NULL, NULL, socket_id,
+ 0);
+ if (!vq->mpool) {
+ PMD_DRV_LOG(ERR, "Virtio Crypto PMD Cannot create"
+ " mempool");
+ goto mpool_create_err;
+ }
+ for (i = 0; i < vq_size; i++) {
+ vq->vq_descx[i].cookie =
+ rte_zmalloc("crypto PMD op cookie pointer",
+ sizeof(struct virtio_crypto_op_cookie),
+ RTE_CACHE_LINE_SIZE);
+ if (vq->vq_descx[i].cookie == NULL) {
+ PMD_DRV_LOG(ERR, "Failed to alloc mem for cookie");
+ goto cookie_alloc_err;
+ }
+ }
+ }
+
+ vq->hw = hw;
+ vq->dev_id = dev->data->dev_id;
+ vq->vq_queue_index = vtpci_queue_idx;
+ vq->vq_nentries = vq_size;
+
+ /*
+ * Using part of the vring entries is permitted, but the maximum
+ * is vq_size
+ */
+ if (nb_desc == 0 || nb_desc > vq_size)
+ nb_desc = vq_size;
+ vq->vq_free_cnt = nb_desc;
+
+ /*
+ * Reserve a memzone for vring elements
+ */
+ size = vring_size(vq_size, VIRTIO_PCI_VRING_ALIGN);
+ vq->vq_ring_size = RTE_ALIGN_CEIL(size, VIRTIO_PCI_VRING_ALIGN);
+ PMD_INIT_LOG(DEBUG, "%s vring_size: %d, rounded_vring_size: %d",
+ (queue_type == VTCRYPTO_DATAQ) ? "dataq" : "ctrlq",
+ size, vq->vq_ring_size);
+
+ mz = rte_memzone_reserve_aligned(vq_name, vq->vq_ring_size,
+ socket_id, 0, VIRTIO_PCI_VRING_ALIGN);
+ if (mz == NULL) {
+ if (rte_errno == EEXIST)
+ mz = rte_memzone_lookup(vq_name);
+ if (mz == NULL) {
+ PMD_INIT_LOG(ERR, "not enough memory");
+ goto mz_reserve_err;
+ }
+ }
+
+ /*
+ * Virtio PCI device VIRTIO_PCI_QUEUE_PF register is 32bit,
+ * and only accepts 32 bit page frame number.
+ * Check if the allocated physical memory exceeds 16TB.
+ */
+ if ((mz->phys_addr + vq->vq_ring_size - 1)
+ >> (VIRTIO_PCI_QUEUE_ADDR_SHIFT + 32)) {
+ PMD_INIT_LOG(ERR, "vring address shouldn't be above 16TB!");
+ goto vring_addr_err;
+ }
+
+ memset(mz->addr, 0, sizeof(mz->len));
+ vq->mz = mz;
+ vq->vq_ring_mem = mz->phys_addr;
+ vq->vq_ring_virt_mem = mz->addr;
+ PMD_INIT_LOG(DEBUG, "vq->vq_ring_mem(physical): 0x%"PRIx64,
+ (uint64_t)mz->phys_addr);
+ PMD_INIT_LOG(DEBUG, "vq->vq_ring_virt_mem: 0x%"PRIx64,
+ (uint64_t)(uintptr_t)mz->addr);
+
+ *pvq = vq;
+
+ return 0;
+
+vring_addr_err:
+ rte_memzone_free(mz);
+mz_reserve_err:
+cookie_alloc_err:
+ rte_mempool_free(vq->mpool);
+ if (i != 0) {
+ for (j = 0; j < i; j++)
+ rte_free(vq->vq_descx[j].cookie);
+ }
+mpool_create_err:
+ rte_free(vq);
+ return -ENOMEM;
+}
+
+static int
+virtio_crypto_ctrlq_setup(struct rte_cryptodev *dev, uint16_t queue_idx)
+{
+ int ret;
+ struct virtqueue *vq;
+ struct virtio_crypto_hw *hw = dev->data->dev_private;
+
+ /* if virtio device has started, do not touch the virtqueues */
+ if (dev->data->dev_started)
+ return 0;
+
+ PMD_INIT_FUNC_TRACE();
+
+ ret = virtio_crypto_queue_setup(dev, VTCRYPTO_CTRLQ, queue_idx,
+ 0, SOCKET_ID_ANY, &vq);
+ if (ret < 0) {
+ PMD_INIT_LOG(ERR, "control vq initialization failed");
+ return ret;
+ }
+
+ hw->cvq = vq;
+
+ return 0;
+}
+
+static void
+virtio_crypto_free_queues(struct rte_cryptodev *dev)
+{
+ unsigned int i;
+ struct virtio_crypto_hw *hw = dev->data->dev_private;
+
+ PMD_INIT_FUNC_TRACE();
+
+ /* control queue release */
+ virtio_crypto_queue_release(hw->cvq);
+
+ /* data queue release */
+ for (i = 0; i < hw->max_dataqueues; i++)
+ virtio_crypto_queue_release(dev->data->queue_pairs[i]);
+}
+
+static int
+virtio_crypto_dev_close(struct rte_cryptodev *dev __rte_unused)
+{
+ return 0;
+}
+
+/*
+ * dev_ops for virtio, bare necessities for basic operation
+ */
+static struct rte_cryptodev_ops virtio_crypto_dev_ops = {
+ /* Device related operations */
+ .dev_configure = virtio_crypto_dev_configure,
+ .dev_start = virtio_crypto_dev_start,
+ .dev_stop = virtio_crypto_dev_stop,
+ .dev_close = virtio_crypto_dev_close,
+ .dev_infos_get = virtio_crypto_dev_info_get,
+
+ .stats_get = virtio_crypto_dev_stats_get,
+ .stats_reset = virtio_crypto_dev_stats_reset,
+
+ .queue_pair_setup = virtio_crypto_qp_setup,
+ .queue_pair_release = virtio_crypto_qp_release,
+ .queue_pair_start = NULL,
+ .queue_pair_stop = NULL,
+ .queue_pair_count = NULL,
+
+ /* Crypto related operations */
+ .session_get_size = virtio_crypto_sym_get_session_private_size,
+ .session_configure = virtio_crypto_sym_configure_session,
+ .session_clear = virtio_crypto_sym_clear_session,
+ .qp_attach_session = NULL,
+ .qp_detach_session = NULL
+};
+
+static void
+virtio_crypto_update_stats(struct rte_cryptodev *dev,
+ struct rte_cryptodev_stats *stats)
+{
+ unsigned int i;
+ struct virtio_crypto_hw *hw = dev->data->dev_private;
+
+ PMD_INIT_FUNC_TRACE();
+
+ if (stats == NULL) {
+ PMD_DRV_LOG(ERR, "invalid pointer");
+ return;
+ }
+
+ for (i = 0; i < hw->max_dataqueues; i++) {
+ const struct virtqueue *data_queue
+ = dev->data->queue_pairs[i];
+ if (data_queue == NULL)
+ continue;
+
+ stats->enqueued_count += data_queue->packets_sent_total;
+ stats->enqueue_err_count += data_queue->packets_sent_failed;
+
+ stats->dequeued_count += data_queue->packets_received_total;
+ stats->dequeue_err_count
+ += data_queue->packets_received_failed;
+ }
+}
+
+static void
+virtio_crypto_dev_stats_get(struct rte_cryptodev *dev,
+ struct rte_cryptodev_stats *stats)
+{
+ PMD_INIT_FUNC_TRACE();
+
+ virtio_crypto_update_stats(dev, stats);
+}
+
+static void
+virtio_crypto_dev_stats_reset(struct rte_cryptodev *dev)
+{
+ unsigned int i;
+ struct virtio_crypto_hw *hw = dev->data->dev_private;
+
+ PMD_INIT_FUNC_TRACE();
+
+ for (i = 0; i < hw->max_dataqueues; i++) {
+ struct virtqueue *data_queue = dev->data->queue_pairs[i];
+ if (data_queue == NULL)
+ continue;
+
+ data_queue->packets_sent_total = 0;
+ data_queue->packets_sent_failed = 0;
+
+ data_queue->packets_received_total = 0;
+ data_queue->packets_received_failed = 0;
+ }
+}
+
+static int
+virtio_crypto_qp_setup(struct rte_cryptodev *dev, uint16_t queue_pair_id,
+ const struct rte_cryptodev_qp_conf *qp_conf,
+ int socket_id,
+ struct rte_mempool *session_pool __rte_unused)
+{
+ int ret;
+ struct virtqueue *vq;
+
+ PMD_INIT_FUNC_TRACE();
+
+ /* if virtio dev is started, do not touch the virtqueues */
+ if (dev->data->dev_started)
+ return 0;
+
+ ret = virtio_crypto_queue_setup(dev, VTCRYPTO_DATAQ, queue_pair_id,
+ qp_conf->nb_descriptors, socket_id, &vq);
+ if (ret < 0) {
+ PMD_INIT_LOG(ERR,
+ "virtio crypto data queue initialization failed\n");
+ return ret;
+ }
+
+ dev->data->queue_pairs[queue_pair_id] = vq;
+
+ return 0;
+}
+
+static int
+virtio_crypto_qp_release(struct rte_cryptodev *dev, uint16_t queue_pair_id)
+{
+ struct virtqueue *vq
+ = (struct virtqueue *)dev->data->queue_pairs[queue_pair_id];
+
+ PMD_INIT_FUNC_TRACE();
+
+ if (vq == NULL) {
+ PMD_DRV_LOG(DEBUG, "vq already freed");
+ return 0;
+ }
+
+ virtio_crypto_queue_release(vq);
+ return 0;
+}
+
+static int
+virtio_negotiate_features(struct virtio_crypto_hw *hw, uint64_t req_features)
+{
+ uint64_t host_features;
+
+ PMD_INIT_FUNC_TRACE();
+
+ /* Prepare guest_features: feature that driver wants to support */
+ PMD_INIT_LOG(DEBUG, "guest_features before negotiate = %" PRIx64,
+ req_features);
+
+ /* Read device(host) feature bits */
+ host_features = VTPCI_OPS(hw)->get_features(hw);
+ PMD_INIT_LOG(DEBUG, "host_features before negotiate = %" PRIx64,
+ host_features);
+
+ /*
+ * Negotiate features: Subset of device feature bits are written back
+ * guest feature bits.
+ */
+ hw->guest_features = req_features;
+ hw->guest_features = vtpci_cryptodev_negotiate_features(hw,
+ host_features);
+ PMD_INIT_LOG(DEBUG, "features after negotiate = %" PRIx64,
+ hw->guest_features);
+
+ if (hw->modern) {
+ if (!vtpci_with_feature(hw, VIRTIO_F_VERSION_1)) {
+ PMD_INIT_LOG(ERR,
+ "VIRTIO_F_VERSION_1 features is not enabled.");
+ return -1;
+ }
+ vtpci_cryptodev_set_status(hw,
+ VIRTIO_CONFIG_STATUS_FEATURES_OK);
+ if (!(vtpci_cryptodev_get_status(hw) &
+ VIRTIO_CONFIG_STATUS_FEATURES_OK)) {
+ PMD_INIT_LOG(ERR, "failed to set FEATURES_OK status!");
+ return -1;
+ }
+ }
+
+ hw->req_guest_features = req_features;
+
+ return 0;
+}
+
+/* reset device and renegotiate features if needed */
+static int
+virtio_crypto_init_device(struct rte_cryptodev *cryptodev,
+ uint64_t req_features)
+{
+ struct virtio_crypto_hw *hw = cryptodev->data->dev_private;
+ struct virtio_crypto_config local_config;
+ struct virtio_crypto_config *config = &local_config;
+
+ PMD_INIT_FUNC_TRACE();
+
+ /* Reset the device although not necessary at startup */
+ vtpci_cryptodev_reset(hw);
+
+ /* Tell the host we've noticed this device. */
+ vtpci_cryptodev_set_status(hw, VIRTIO_CONFIG_STATUS_ACK);
+
+ /* Tell the host we've known how to drive the device. */
+ vtpci_cryptodev_set_status(hw, VIRTIO_CONFIG_STATUS_DRIVER);
+ if (virtio_negotiate_features(hw, req_features) < 0)
+ return -1;
+
+ /* Get status of the device */
+ vtpci_read_cryptodev_config(hw,
+ offsetof(struct virtio_crypto_config, status),
+ &config->status, sizeof(config->status));
+ if (config->status != VIRTIO_CRYPTO_S_HW_READY) {
+ PMD_DRV_LOG(ERR, "accelerator hardware is "
+ "not ready");
+ return -1;
+ }
+
+ /* Get number of data queues */
+ vtpci_read_cryptodev_config(hw,
+ offsetof(struct virtio_crypto_config, max_dataqueues),
+ &config->max_dataqueues,
+ sizeof(config->max_dataqueues));
+ hw->max_dataqueues = config->max_dataqueues;
+
+ PMD_INIT_LOG(DEBUG, "hw->max_dataqueues=%d",
+ hw->max_dataqueues);
+
+ return 0;
+}
+
+/*
+ * This function is based on probe() function
+ * It returns 0 on success.
+ */
+static int
+crypto_virtio_create(const char *name, struct rte_pci_device *pci_dev,
+ struct rte_cryptodev_pmd_init_params *init_params)
+{
+ struct rte_cryptodev *cryptodev;
+ struct virtio_crypto_hw *hw;
+
+ PMD_INIT_FUNC_TRACE();
+
+ cryptodev = rte_cryptodev_pmd_create(name, &pci_dev->device,
+ init_params);
+ if (cryptodev == NULL)
+ return -ENODEV;
+
+ cryptodev->driver_id = cryptodev_virtio_driver_id;
+ cryptodev->dev_ops = &virtio_crypto_dev_ops;
+
+ cryptodev->enqueue_burst = virtio_crypto_pkt_tx_burst;
+ cryptodev->dequeue_burst = virtio_crypto_pkt_rx_burst;
+
+ cryptodev->feature_flags = RTE_CRYPTODEV_FF_SYMMETRIC_CRYPTO |
+ RTE_CRYPTODEV_FF_SYM_OPERATION_CHAINING;
+
+ hw = cryptodev->data->dev_private;
+ hw->dev_id = cryptodev->data->dev_id;
+
+ PMD_INIT_LOG(DEBUG, "dev %d vendorID=0x%x deviceID=0x%x",
+ cryptodev->data->dev_id, pci_dev->id.vendor_id,
+ pci_dev->id.device_id);
+
+ /* pci device init */
+ if (vtpci_cryptodev_init(pci_dev, hw))
+ return -1;
+
+ if (virtio_crypto_init_device(cryptodev,
+ VIRTIO_CRYPTO_PMD_GUEST_FEATURES) < 0)
+ return -1;
+
+ return 0;
+}
+
+static int
+virtio_crypto_dev_uninit(struct rte_cryptodev *cryptodev)
+{
+ struct virtio_crypto_hw *hw = cryptodev->data->dev_private;
+
+ PMD_INIT_FUNC_TRACE();
+
+ if (rte_eal_process_type() == RTE_PROC_SECONDARY)
+ return -EPERM;
+
+ if (cryptodev->data->dev_started) {
+ virtio_crypto_dev_stop(cryptodev);
+ virtio_crypto_dev_close(cryptodev);
+ }
+
+ cryptodev->dev_ops = NULL;
+ cryptodev->enqueue_burst = NULL;
+ cryptodev->dequeue_burst = NULL;
+
+ /* release control queue */
+ virtio_crypto_queue_release(hw->cvq);
+
+ rte_free(cryptodev->data);
+ cryptodev->data = NULL;
+
+ PMD_DRV_LOG(INFO, "dev_uninit completed");
+
+ return 0;
+}
+
+static int
+virtio_crypto_dev_configure(struct rte_cryptodev *cryptodev,
+ struct rte_cryptodev_config *config __rte_unused)
+{
+ struct virtio_crypto_hw *hw = cryptodev->data->dev_private;
+
+ PMD_INIT_FUNC_TRACE();
+
+ if (virtio_crypto_init_device(cryptodev,
+ VIRTIO_CRYPTO_PMD_GUEST_FEATURES) < 0)
+ return -1;
+
+ /* setup control queue
+ * [0, 1, ... ,(config->max_dataqueues - 1)] are data queues
+ * config->max_dataqueues is the control queue
+ */
+ if (virtio_crypto_ctrlq_setup(cryptodev, hw->max_dataqueues) < 0) {
+ PMD_INIT_LOG(ERR, "control queue setup error");
+ return -1;
+ }
+ virtio_crypto_ctrlq_start(cryptodev);
+
+ return 0;
+}
+
+static void
+virtio_crypto_dev_stop(struct rte_cryptodev *dev)
+{
+ struct virtio_crypto_hw *hw = dev->data->dev_private;
+
+ PMD_INIT_FUNC_TRACE();
+ PMD_DRV_LOG(DEBUG, "virtio_dev_stop");
+
+ vtpci_cryptodev_reset(hw);
+
+ virtio_crypto_dev_free_mbufs(dev);
+ virtio_crypto_free_queues(dev);
+
+ dev->data->dev_started = 0;
+}
+
+static int
+virtio_crypto_dev_start(struct rte_cryptodev *dev)
+{
+ struct virtio_crypto_hw *hw = dev->data->dev_private;
+
+ if (dev->data->dev_started)
+ return 0;
+
+ /* Do final configuration before queue engine starts */
+ virtio_crypto_dataq_start(dev);
+ vtpci_cryptodev_reinit_complete(hw);
+
+ dev->data->dev_started = 1;
+
+ return 0;
+}
+
+static void virtio_crypto_dev_free_mbufs(struct rte_cryptodev *dev)
+{
+ uint32_t i;
+ struct virtio_crypto_hw *hw = dev->data->dev_private;
+
+ for (i = 0; i < hw->max_dataqueues; i++) {
+ PMD_INIT_LOG(DEBUG, "Before freeing dataq[%d] used "
+ "and unused buf", i);
+ VIRTQUEUE_DUMP((struct virtqueue *)
+ dev->data->queue_pairs[i]);
+
+ PMD_INIT_LOG(DEBUG, "queue_pairs[%d]=%p",
+ i, dev->data->queue_pairs[i]);
+
+ virtqueue_detatch_unused(dev->data->queue_pairs[i]);
+
+ PMD_INIT_LOG(DEBUG, "After freeing dataq[%d] used and "
+ "unused buf", i);
+ VIRTQUEUE_DUMP(
+ (struct virtqueue *)dev->data->queue_pairs[i]);
+ }
+}
+
+static unsigned int virtio_crypto_sym_get_session_private_size(
+ struct rte_cryptodev *dev __rte_unused)
+{
+ PMD_INIT_FUNC_TRACE();
+
+ return RTE_ALIGN_CEIL(sizeof(struct virtio_crypto_session), 16);
+}
+
+static int virtio_crypto_check_sym_session_paras(
+ struct rte_cryptodev *dev)
+{
+ struct virtio_crypto_hw *hw;
+
+ PMD_INIT_FUNC_TRACE();
+
+ if (unlikely(dev == NULL)) {
+ PMD_SESSION_LOG(ERR, "dev is NULL");
+ return -1;
+ }
+ if (unlikely(dev->data == NULL)) {
+ PMD_SESSION_LOG(ERR, "dev->data is NULL");
+ return -1;
+ }
+ hw = dev->data->dev_private;
+ if (unlikely(hw == NULL)) {
+ PMD_SESSION_LOG(ERR, "hw is NULL");
+ return -1;
+ }
+ if (unlikely(hw->cvq == NULL)) {
+ PMD_SESSION_LOG(ERR, "vq is NULL");
+ return -1;
+ }
+
+ return 0;
+}
+
+static int virtio_crypto_check_sym_clear_session_paras(
+ struct rte_cryptodev *dev,
+ struct rte_cryptodev_sym_session *sess)
+{
+ PMD_INIT_FUNC_TRACE();
+
+ if (sess == NULL) {
+ PMD_SESSION_LOG(ERR, "sym_session is NULL");
+ return -1;
+ }
+
+ return virtio_crypto_check_sym_session_paras(dev);
+}
+
+#define NUM_ENTRY_SYM_CLEAR_SESSION 2
+
+static void virtio_crypto_sym_clear_session(
+ struct rte_cryptodev *dev,
+ struct rte_cryptodev_sym_session *sess)
+{
+ struct virtio_crypto_hw *hw;
+ struct virtqueue *vq;
+ struct virtio_crypto_session *session;
+ struct virtio_crypto_op_ctrl_req *ctrl;
+ struct vring_desc *desc;
+ uint8_t *status;
+ uint8_t needed = 1;
+ uint32_t head;
+ uint8_t *malloc_virt_addr;
+ uint64_t malloc_phys_addr;
+ uint8_t len_inhdr = sizeof(struct virtio_crypto_inhdr);
+ uint32_t len_op_ctrl_req = sizeof(struct virtio_crypto_op_ctrl_req);
+ uint32_t desc_offset = len_op_ctrl_req + len_inhdr;
+
+ PMD_INIT_FUNC_TRACE();
+
+ if (virtio_crypto_check_sym_clear_session_paras(dev, sess) < 0)
+ return;
+
+ hw = dev->data->dev_private;
+ vq = hw->cvq;
+ session = (struct virtio_crypto_session *)get_session_private_data(
+ sess, cryptodev_virtio_driver_id);
+ if (session == NULL) {
+ PMD_SESSION_LOG(ERR, "Invalid session parameter");
+ return;
+ }
+
+ PMD_SESSION_LOG(INFO, "vq->vq_desc_head_idx = %d, "
+ "vq = %p", vq->vq_desc_head_idx, vq);
+
+ if (vq->vq_free_cnt < needed) {
+ PMD_SESSION_LOG(ERR,
+ "vq->vq_free_cnt = %d is less than %d, "
+ "not enough", vq->vq_free_cnt, needed);
+ return;
+ }
+
+ /*
+ * malloc memory to store information of ctrl request op,
+ * returned status and desc vring
+ */
+ malloc_virt_addr = rte_malloc(NULL, len_op_ctrl_req + len_inhdr
+ + NUM_ENTRY_SYM_CLEAR_SESSION
+ * sizeof(struct vring_desc), RTE_CACHE_LINE_SIZE);
+ if (malloc_virt_addr == NULL) {
+ PMD_SESSION_LOG(ERR, "not enough heap room");
+ return;
+ }
+ malloc_phys_addr = rte_malloc_virt2iova(malloc_virt_addr);
+
+ /* assign ctrl request op part */
+ ctrl = (struct virtio_crypto_op_ctrl_req *)malloc_virt_addr;
+ ctrl->header.opcode = VIRTIO_CRYPTO_CIPHER_DESTROY_SESSION;
+ /* default data virtqueue is 0 */
+ ctrl->header.queue_id = 0;
+ ctrl->u.destroy_session.session_id = session->session_id;
+
+ /* status part */
+ status = &(((struct virtio_crypto_inhdr *)
+ ((uint8_t *)malloc_virt_addr + len_op_ctrl_req))->status);
+ *status = VIRTIO_CRYPTO_ERR;
+
+ /* indirect desc vring part */
+ desc = (struct vring_desc *)((uint8_t *)malloc_virt_addr
+ + desc_offset);
+
+ /* ctrl request part */
+ desc[0].addr = malloc_phys_addr;
+ desc[0].len = len_op_ctrl_req;
+ desc[0].flags = VRING_DESC_F_NEXT;
+ desc[0].next = 1;
+
+ /* status part */
+ desc[1].addr = malloc_phys_addr + len_op_ctrl_req;
+ desc[1].len = len_inhdr;
+ desc[1].flags = VRING_DESC_F_WRITE;
+
+ /* use only a single desc entry */
+ head = vq->vq_desc_head_idx;
+ vq->vq_ring.desc[head].flags = VRING_DESC_F_INDIRECT;
+ vq->vq_ring.desc[head].addr = malloc_phys_addr + desc_offset;
+ vq->vq_ring.desc[head].len
+ = NUM_ENTRY_SYM_CLEAR_SESSION
+ * sizeof(struct vring_desc);
+
+ vq->vq_free_cnt -= needed;
+
+ vq->vq_desc_head_idx = vq->vq_ring.desc[head].next;
+
+ vq_update_avail_ring(vq, head);
+ vq_update_avail_idx(vq);
+
+ PMD_INIT_LOG(DEBUG, "vq->vq_queue_index = %d", vq->vq_queue_index);
+
+ virtqueue_notify(vq);
+
+ rte_rmb();
+ while (vq->vq_used_cons_idx == vq->vq_ring.used->idx) {
+ rte_rmb();
+ usleep(100);
+ }
+
+ while (vq->vq_used_cons_idx != vq->vq_ring.used->idx) {
+ uint32_t idx, desc_idx, used_idx;
+ struct vring_used_elem *uep;
+
+ used_idx = (uint32_t)(vq->vq_used_cons_idx
+ & (vq->vq_nentries - 1));
+ uep = &vq->vq_ring.used->ring[used_idx];
+ idx = (uint32_t) uep->id;
+ desc_idx = idx;
+ while (vq->vq_ring.desc[desc_idx].flags
+ & VRING_DESC_F_NEXT) {
+ desc_idx = vq->vq_ring.desc[desc_idx].next;
+ vq->vq_free_cnt++;
+ }
+
+ vq->vq_ring.desc[desc_idx].next = vq->vq_desc_head_idx;
+ vq->vq_desc_head_idx = idx;
+ vq->vq_used_cons_idx++;
+ vq->vq_free_cnt++;
+ }
+
+ if (*status != VIRTIO_CRYPTO_OK) {
+ PMD_SESSION_LOG(ERR, "Close session failed "
+ "status=%"PRIu32", session_id=%"PRIu64"",
+ *status, session->session_id);
+ rte_free(malloc_virt_addr);
+ return;
+ }
+
+ PMD_INIT_LOG(DEBUG, "vq->vq_free_cnt=%d\nvq->vq_desc_head_idx=%d",
+ vq->vq_free_cnt, vq->vq_desc_head_idx);
+
+ PMD_SESSION_LOG(INFO, "Close session %"PRIu64" successfully ",
+ session->session_id);
+
+ memset(sess, 0, sizeof(struct virtio_crypto_session));
+ rte_free(malloc_virt_addr);
+}
+
+static struct rte_crypto_cipher_xform *
+virtio_crypto_get_cipher_xform(struct rte_crypto_sym_xform *xform)
+{
+ do {
+ if (xform->type == RTE_CRYPTO_SYM_XFORM_CIPHER)
+ return &xform->cipher;
+
+ xform = xform->next;
+ } while (xform);
+
+ return NULL;
+}
+
+static struct rte_crypto_auth_xform *
+virtio_crypto_get_auth_xform(struct rte_crypto_sym_xform *xform)
+{
+ do {
+ if (xform->type == RTE_CRYPTO_SYM_XFORM_AUTH)
+ return &xform->auth;
+
+ xform = xform->next;
+ } while (xform);
+
+ return NULL;
+}
+
+/** Get xform chain order */
+static int
+virtio_crypto_get_chain_order(struct rte_crypto_sym_xform *xform)
+{
+ if (xform == NULL)
+ return -1;
+
+ /* Cipher Only */
+ if (xform->type == RTE_CRYPTO_SYM_XFORM_CIPHER &&
+ xform->next == NULL)
+ return VIRTIO_CRYPTO_CMD_CIPHER;
+
+ /* Authentication Only */
+ if (xform->type == RTE_CRYPTO_SYM_XFORM_AUTH &&
+ xform->next == NULL)
+ return VIRTIO_CRYPTO_CMD_AUTH;
+
+ /* Authenticate then Cipher */
+ if (xform->type == RTE_CRYPTO_SYM_XFORM_AUTH &&
+ xform->next->type == RTE_CRYPTO_SYM_XFORM_CIPHER)
+ return VIRTIO_CRYPTO_CMD_HASH_CIPHER;
+
+ /* Cipher then Authenticate */
+ if (xform->type == RTE_CRYPTO_SYM_XFORM_CIPHER &&
+ xform->next->type == RTE_CRYPTO_SYM_XFORM_AUTH)
+ return VIRTIO_CRYPTO_CMD_CIPHER_HASH;
+
+ return -1;
+}
+
+static int virtio_crypto_sym_pad_cipher_param(
+ struct virtio_crypto_cipher_session_para *para,
+ struct rte_crypto_cipher_xform *cipher_xform)
+{
+ switch (cipher_xform->algo) {
+ case RTE_CRYPTO_CIPHER_NULL:
+ para->algo = VIRTIO_CRYPTO_NO_CIPHER;
+ break;
+ case RTE_CRYPTO_CIPHER_3DES_CBC:
+ para->algo = VIRTIO_CRYPTO_CIPHER_3DES_CBC;
+ break;
+ case RTE_CRYPTO_CIPHER_3DES_CTR:
+ para->algo = VIRTIO_CRYPTO_CIPHER_3DES_CTR;
+ break;
+ case RTE_CRYPTO_CIPHER_3DES_ECB:
+ para->algo = VIRTIO_CRYPTO_CIPHER_3DES_ECB;
+ break;
+ case RTE_CRYPTO_CIPHER_AES_CBC:
+ para->algo = VIRTIO_CRYPTO_CIPHER_AES_CBC;
+ break;
+ case RTE_CRYPTO_CIPHER_AES_CTR:
+ para->algo = VIRTIO_CRYPTO_CIPHER_AES_CTR;
+ break;
+ case RTE_CRYPTO_CIPHER_AES_ECB:
+ para->algo = VIRTIO_CRYPTO_CIPHER_AES_ECB;
+ break;
+ case RTE_CRYPTO_CIPHER_AES_F8:
+ para->algo = VIRTIO_CRYPTO_CIPHER_AES_F8;
+ break;
+ case RTE_CRYPTO_CIPHER_AES_XTS:
+ para->algo = VIRTIO_CRYPTO_CIPHER_AES_XTS;
+ break;
+ case RTE_CRYPTO_CIPHER_ARC4:
+ para->algo = VIRTIO_CRYPTO_CIPHER_ARC4;
+ break;
+ case RTE_CRYPTO_CIPHER_KASUMI_F8:
+ para->algo = VIRTIO_CRYPTO_CIPHER_KASUMI_F8;
+ break;
+ case RTE_CRYPTO_CIPHER_SNOW3G_UEA2:
+ para->algo = VIRTIO_CRYPTO_CIPHER_SNOW3G_UEA2;
+ break;
+ case RTE_CRYPTO_CIPHER_ZUC_EEA3:
+ para->algo = VIRTIO_CRYPTO_CIPHER_ZUC_EEA3;
+ break;
+ case RTE_CRYPTO_CIPHER_DES_CBC:
+ para->algo = VIRTIO_CRYPTO_CIPHER_DES_CBC;
+ break;
+ default:
+ PMD_SESSION_LOG(ERR, "Crypto: Unsupported Cipher alg %u",
+ cipher_xform->algo);
+ return -1;
+ }
+
+ para->keylen = cipher_xform->key.length;
+ switch (cipher_xform->op) {
+ case RTE_CRYPTO_CIPHER_OP_ENCRYPT:
+ para->op = VIRTIO_CRYPTO_OP_ENCRYPT;
+ break;
+ case RTE_CRYPTO_CIPHER_OP_DECRYPT:
+ para->op = VIRTIO_CRYPTO_OP_DECRYPT;
+ break;
+ default:
+ PMD_SESSION_LOG(ERR, "Unsupported cipher operation parameter");
+ return -1;
+ }
+
+ return 0;
+}
+
+static int virtio_crypto_sym_pad_auth_param(
+ struct virtio_crypto_op_ctrl_req *ctrl,
+ struct rte_crypto_auth_xform *auth_xform)
+{
+ uint32_t *algo;
+ struct virtio_crypto_alg_chain_session_para *para =
+ &(ctrl->u.sym_create_session.u.chain.para);
+
+ switch (ctrl->u.sym_create_session.u.chain.para.hash_mode) {
+ case VIRTIO_CRYPTO_SYM_HASH_MODE_PLAIN:
+ algo = &(para->u.hash_param.algo);
+ break;
+ case VIRTIO_CRYPTO_SYM_HASH_MODE_AUTH:
+ algo = &(para->u.mac_param.algo);
+ break;
+ default:
+ PMD_SESSION_LOG(ERR, "Unsupported hash mode %u specified",
+ ctrl->u.sym_create_session.u.chain.para.hash_mode);
+ return -1;
+ }
+
+ switch (auth_xform->algo) {
+ case RTE_CRYPTO_AUTH_NULL:
+ *algo = VIRTIO_CRYPTO_NO_MAC;
+ break;
+ case RTE_CRYPTO_AUTH_AES_CBC_MAC:
+ *algo = VIRTIO_CRYPTO_MAC_CBCMAC_AES;
+ break;
+ case RTE_CRYPTO_AUTH_AES_CMAC:
+ *algo = VIRTIO_CRYPTO_MAC_CMAC_AES;
+ break;
+ case RTE_CRYPTO_AUTH_AES_GMAC:
+ *algo = VIRTIO_CRYPTO_MAC_GMAC_AES;
+ break;
+ case RTE_CRYPTO_AUTH_AES_XCBC_MAC:
+ *algo = VIRTIO_CRYPTO_MAC_XCBC_AES;
+ break;
+ case RTE_CRYPTO_AUTH_KASUMI_F9:
+ *algo = VIRTIO_CRYPTO_MAC_KASUMI_F9;
+ break;
+ case RTE_CRYPTO_AUTH_MD5:
+ *algo = VIRTIO_CRYPTO_HASH_MD5;
+ break;
+ case RTE_CRYPTO_AUTH_MD5_HMAC:
+ *algo = VIRTIO_CRYPTO_MAC_HMAC_MD5;
+ break;
+ case RTE_CRYPTO_AUTH_SHA1:
+ *algo = VIRTIO_CRYPTO_HASH_SHA1;
+ break;
+ case RTE_CRYPTO_AUTH_SHA1_HMAC:
+ *algo = VIRTIO_CRYPTO_MAC_HMAC_SHA1;
+ break;
+ case RTE_CRYPTO_AUTH_SHA224:
+ *algo = VIRTIO_CRYPTO_HASH_SHA_224;
+ break;
+ case RTE_CRYPTO_AUTH_SHA224_HMAC:
+ *algo = VIRTIO_CRYPTO_MAC_HMAC_SHA_224;
+ break;
+ case RTE_CRYPTO_AUTH_SHA256:
+ *algo = VIRTIO_CRYPTO_HASH_SHA_256;
+ break;
+ case RTE_CRYPTO_AUTH_SHA256_HMAC:
+ *algo = VIRTIO_CRYPTO_MAC_HMAC_SHA_256;
+ break;
+ case RTE_CRYPTO_AUTH_SHA384:
+ *algo = VIRTIO_CRYPTO_HASH_SHA_384;
+ break;
+ case RTE_CRYPTO_AUTH_SHA384_HMAC:
+ *algo = VIRTIO_CRYPTO_MAC_HMAC_SHA_384;
+ break;
+ case RTE_CRYPTO_AUTH_SHA512:
+ *algo = VIRTIO_CRYPTO_HASH_SHA_512;
+ break;
+ case RTE_CRYPTO_AUTH_SHA512_HMAC:
+ *algo = VIRTIO_CRYPTO_MAC_HMAC_SHA_512;
+ break;
+ case RTE_CRYPTO_AUTH_SNOW3G_UIA2:
+ *algo = VIRTIO_CRYPTO_MAC_SNOW3G_UIA2;
+ break;
+ default:
+ PMD_SESSION_LOG(ERR,
+ "Crypto: Undefined Hash algo %u specified",
+ auth_xform->algo);
+ return -1;
+ }
+
+ return 0;
+}
+
+static int virtio_crypto_sym_pad_op_ctrl_req(
+ struct virtio_crypto_op_ctrl_req *ctrl,
+ struct rte_crypto_sym_xform *xform, bool is_chainned,
+ uint8_t **cipher_key_data, uint8_t **auth_key_data,
+ struct virtio_crypto_session *session)
+{
+ int ret;
+ struct rte_crypto_auth_xform *auth_xform = NULL;
+ struct rte_crypto_cipher_xform *cipher_xform = NULL;
+
+ /* Get cipher xform from crypto xform chain */
+ cipher_xform = virtio_crypto_get_cipher_xform(xform);
+ if (cipher_xform) {
+ if (is_chainned)
+ ret = virtio_crypto_sym_pad_cipher_param(
+ &ctrl->u.sym_create_session.u.chain.para
+ .cipher_param, cipher_xform);
+ else
+ ret = virtio_crypto_sym_pad_cipher_param(
+ &ctrl->u.sym_create_session.u.cipher.para,
+ cipher_xform);
+
+ if (ret < 0) {
+ PMD_SESSION_LOG(ERR,
+ "pad cipher parameter failed");
+ return -1;
+ }
+
+ *cipher_key_data = cipher_xform->key.data;
+
+ session->iv.offset = cipher_xform->iv.offset;
+ session->iv.length = cipher_xform->iv.length;
+ }
+
+ /* Get auth xform from crypto xform chain */
+ auth_xform = virtio_crypto_get_auth_xform(xform);
+ if (auth_xform) {
+ /* FIXME: support VIRTIO_CRYPTO_SYM_HASH_MODE_NESTED */
+ struct virtio_crypto_alg_chain_session_para *para =
+ &(ctrl->u.sym_create_session.u.chain.para);
+ if (auth_xform->key.length) {
+ para->hash_mode = VIRTIO_CRYPTO_SYM_HASH_MODE_AUTH;
+ para->u.mac_param.auth_key_len =
+ (uint32_t)auth_xform->key.length;
+ para->u.mac_param.hash_result_len =
+ auth_xform->digest_length;
+
+ *auth_key_data = auth_xform->key.data;
+ } else {
+ para->hash_mode = VIRTIO_CRYPTO_SYM_HASH_MODE_PLAIN;
+ para->u.hash_param.hash_result_len =
+ auth_xform->digest_length;
+ }
+
+ ret = virtio_crypto_sym_pad_auth_param(ctrl, auth_xform);
+ if (ret < 0) {
+ PMD_SESSION_LOG(ERR, "pad auth parameter failed");
+ return -1;
+ }
+ }
+
+ return 0;
+}
+
+static int virtio_crypto_check_sym_configure_session_paras(
+ struct rte_cryptodev *dev,
+ struct rte_crypto_sym_xform *xform,
+ struct rte_cryptodev_sym_session *sym_sess,
+ struct rte_mempool *mempool)
+{
+ if (unlikely(xform == NULL) || unlikely(sym_sess == NULL) ||
+ unlikely(mempool == NULL)) {
+ PMD_SESSION_LOG(ERR, "NULL pointer");
+ return -1;
+ }
+
+ if (virtio_crypto_check_sym_session_paras(dev) < 0)
+ return -1;
+
+ return 0;
+}
+
+static int virtio_crypto_sym_configure_session(
+ struct rte_cryptodev *dev,
+ struct rte_crypto_sym_xform *xform,
+ struct rte_cryptodev_sym_session *sess,
+ struct rte_mempool *mempool)
+{
+ int ret;
+ struct virtio_crypto_session crypto_sess;
+ void *session_private = &crypto_sess;
+ struct virtio_crypto_session *session;
+ struct virtio_crypto_op_ctrl_req *ctrl_req;
+ enum virtio_crypto_cmd_id cmd_id;
+ uint8_t *cipher_key_data = NULL;
+ uint8_t *auth_key_data = NULL;
+ struct virtio_crypto_hw *hw;
+ struct virtqueue *control_vq;
+
+ PMD_INIT_FUNC_TRACE();
+
+ ret = virtio_crypto_check_sym_configure_session_paras(dev, xform,
+ sess, mempool);
+ if (ret < 0) {
+ PMD_SESSION_LOG(ERR, "Invalid parameters");
+ return ret;
+ }
+
+ if (rte_mempool_get(mempool, &session_private)) {
+ PMD_SESSION_LOG(ERR,
+ "Couldn't get object from session mempool");
+ return -ENOMEM;
+ }
+
+ session = (struct virtio_crypto_session *)session_private;
+ memset(session, 0, sizeof(struct virtio_crypto_session));
+ ctrl_req = &session->ctrl;
+ ctrl_req->header.opcode = VIRTIO_CRYPTO_CIPHER_CREATE_SESSION;
+ /* FIXME: support multiqueue */
+ ctrl_req->header.queue_id = 0;
+
+ hw = dev->data->dev_private;
+ control_vq = hw->cvq;
+
+ cmd_id = virtio_crypto_get_chain_order(xform);
+ if (cmd_id == VIRTIO_CRYPTO_CMD_CIPHER_HASH)
+ ctrl_req->u.sym_create_session.u.chain.para.alg_chain_order
+ = VIRTIO_CRYPTO_SYM_ALG_CHAIN_ORDER_CIPHER_THEN_HASH;
+ if (cmd_id == VIRTIO_CRYPTO_CMD_HASH_CIPHER)
+ ctrl_req->u.sym_create_session.u.chain.para.alg_chain_order
+ = VIRTIO_CRYPTO_SYM_ALG_CHAIN_ORDER_HASH_THEN_CIPHER;
+
+ switch (cmd_id) {
+ case VIRTIO_CRYPTO_CMD_CIPHER_HASH:
+ case VIRTIO_CRYPTO_CMD_HASH_CIPHER:
+ ctrl_req->u.sym_create_session.op_type
+ = VIRTIO_CRYPTO_SYM_OP_ALGORITHM_CHAINING;
+
+ ret = virtio_crypto_sym_pad_op_ctrl_req(ctrl_req,
+ xform, true, &cipher_key_data, &auth_key_data, session);
+ if (ret < 0) {
+ PMD_SESSION_LOG(ERR,
+ "padding sym op ctrl req failed");
+ goto error_out;
+ }
+ ret = virtio_crypto_send_command(control_vq, ctrl_req,
+ cipher_key_data, auth_key_data, session);
+ if (ret < 0) {
+ PMD_SESSION_LOG(ERR,
+ "create session failed: %d", ret);
+ goto error_out;
+ }
+ break;
+ case VIRTIO_CRYPTO_CMD_CIPHER:
+ ctrl_req->u.sym_create_session.op_type
+ = VIRTIO_CRYPTO_SYM_OP_CIPHER;
+ ret = virtio_crypto_sym_pad_op_ctrl_req(ctrl_req, xform,
+ false, &cipher_key_data, &auth_key_data, session);
+ if (ret < 0) {
+ PMD_SESSION_LOG(ERR,
+ "padding sym op ctrl req failed");
+ goto error_out;
+ }
+ ret = virtio_crypto_send_command(control_vq, ctrl_req,
+ cipher_key_data, NULL, session);
+ if (ret < 0) {
+ PMD_SESSION_LOG(ERR,
+ "create session failed: %d", ret);
+ goto error_out;
+ }
+ break;
+ default:
+ PMD_SESSION_LOG(ERR,
+ "Unsupported operation chain order parameter");
+ goto error_out;
+ }
+
+ set_session_private_data(sess, dev->driver_id,
+ session_private);
+
+ return 0;
+
+error_out:
+ return -1;
+}
+
+static void
+virtio_crypto_dev_info_get(struct rte_cryptodev *dev,
+ struct rte_cryptodev_info *info)
+{
+ struct virtio_crypto_hw *hw = dev->data->dev_private;
+
+ PMD_INIT_FUNC_TRACE();
+
+ if (info != NULL) {
+ info->driver_id = cryptodev_virtio_driver_id;
+ info->pci_dev = RTE_DEV_TO_PCI(dev->device);
+ info->feature_flags = dev->feature_flags;
+ info->max_nb_queue_pairs = hw->max_dataqueues;
+ info->sym.max_nb_sessions =
+ RTE_VIRTIO_CRYPTO_PMD_MAX_NB_SESSIONS;
+ }
+}
+
+static int crypto_virtio_pci_probe(
+ struct rte_pci_driver *pci_drv __rte_unused,
+ struct rte_pci_device *pci_dev)
+{
+ struct rte_cryptodev_pmd_init_params init_params = {
+ .name = "",
+ .socket_id = rte_socket_id(),
+ .private_data_size = sizeof(struct virtio_crypto_hw),
+ .max_nb_sessions = RTE_VIRTIO_CRYPTO_PMD_MAX_NB_SESSIONS
+ };
+ char name[RTE_CRYPTODEV_NAME_MAX_LEN];
+
+ PMD_DRV_LOG(DEBUG, "Found Crypto device at %02x:%02x.%x",
+ pci_dev->addr.bus,
+ pci_dev->addr.devid,
+ pci_dev->addr.function);
+
+ rte_pci_device_name(&pci_dev->addr, name, sizeof(name));
+
+ return crypto_virtio_create(name, pci_dev, &init_params);
+}
+
+static int crypto_virtio_pci_remove(struct rte_pci_device *pci_dev)
+{
+ struct rte_cryptodev *cryptodev;
+ char cryptodev_name[RTE_CRYPTODEV_NAME_MAX_LEN];
+
+ if (pci_dev == NULL)
+ return -EINVAL;
+
+ rte_pci_device_name(&pci_dev->addr, cryptodev_name,
+ sizeof(cryptodev_name));
+
+ cryptodev = rte_cryptodev_pmd_get_named_dev(cryptodev_name);
+ if (cryptodev == NULL)
+ return -ENODEV;
+
+ return virtio_crypto_dev_uninit(cryptodev);
+}
+
+static struct rte_pci_driver rte_virtio_crypto_driver = {
+ .id_table = pci_id_virtio_crypto_map,
+ .drv_flags = 0,
+ .probe = crypto_virtio_pci_probe,
+ .remove = crypto_virtio_pci_remove
+};
+
+static struct cryptodev_driver virtio_crypto_drv;
+
+RTE_PMD_REGISTER_PCI(CRYPTODEV_NAME_VIRTIO_PMD, rte_virtio_crypto_driver);
+RTE_PMD_REGISTER_CRYPTO_DRIVER(virtio_crypto_drv, rte_virtio_crypto_driver,
+ cryptodev_virtio_driver_id);
diff --git a/drivers/crypto/virtio/virtio_cryptodev.h b/drivers/crypto/virtio/virtio_cryptodev.h
new file mode 100644
index 0000000..875400c
--- /dev/null
+++ b/drivers/crypto/virtio/virtio_cryptodev.h
@@ -0,0 +1,66 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 HUAWEI TECHNOLOGIES CO., LTD.
+ */
+
+#ifndef _VIRTIO_CRYPTODEV_H_
+#define _VIRTIO_CRYPTODEV_H_
+
+#include <linux/virtio_crypto.h>
+
+#include "virtio_pci.h"
+#include "virtio_ring.h"
+
+#ifndef PAGE_SIZE
+#define PAGE_SIZE 4096
+#endif
+
+#define CRYPTODEV_NAME_VIRTIO_PMD crypto_virtio
+
+#define NUM_ENTRY_VIRTIO_CRYPTO_OP 7
+
+/* Features desired/implemented by this driver. */
+#define VIRTIO_CRYPTO_PMD_GUEST_FEATURES (1ULL << VIRTIO_F_VERSION_1)
+
+extern uint8_t cryptodev_virtio_driver_id;
+
+enum virtio_crypto_cmd_id {
+ VIRTIO_CRYPTO_CMD_CIPHER = 0,
+ VIRTIO_CRYPTO_CMD_AUTH = 1,
+ VIRTIO_CRYPTO_CMD_CIPHER_HASH = 2,
+ VIRTIO_CRYPTO_CMD_HASH_CIPHER = 3
+};
+
+struct virtio_crypto_op_cookie {
+ struct virtio_crypto_op_data_req data_req;
+ struct virtio_crypto_inhdr inhdr;
+ struct vring_desc desc[NUM_ENTRY_VIRTIO_CRYPTO_OP];
+};
+
+/*
+ * Control queue function prototype
+ */
+void virtio_crypto_ctrlq_start(struct rte_cryptodev *dev);
+
+/*
+ * Data queue function prototype
+ */
+void virtio_crypto_dataq_start(struct rte_cryptodev *dev);
+
+int virtio_crypto_queue_setup(struct rte_cryptodev *dev,
+ int queue_type,
+ uint16_t vtpci_queue_idx,
+ uint16_t nb_desc,
+ int socket_id,
+ struct virtqueue **pvq);
+
+void virtio_crypto_queue_release(struct virtqueue *vq);
+
+uint16_t virtio_crypto_pkt_tx_burst(void *tx_queue,
+ struct rte_crypto_op **tx_pkts,
+ uint16_t nb_pkts);
+
+uint16_t virtio_crypto_pkt_rx_burst(void *tx_queue,
+ struct rte_crypto_op **tx_pkts,
+ uint16_t nb_pkts);
+
+#endif /* _VIRTIO_CRYPTODEV_H_ */
diff --git a/drivers/crypto/virtio/virtio_rxtx.c b/drivers/crypto/virtio/virtio_rxtx.c
new file mode 100644
index 0000000..b7d8066
--- /dev/null
+++ b/drivers/crypto/virtio/virtio_rxtx.c
@@ -0,0 +1,533 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018 HUAWEI TECHNOLOGIES CO., LTD.
+ */
+
+#include <stdint.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <errno.h>
+
+#include <rte_cycles.h>
+#include <rte_memory.h>
+#include <rte_memcpy.h>
+#include <rte_memzone.h>
+#include <rte_branch_prediction.h>
+#include <rte_mempool.h>
+#include <rte_malloc.h>
+#include <rte_mbuf.h>
+#include <rte_cryptodev.h>
+#include <rte_prefetch.h>
+#include <rte_string_fns.h>
+#include <rte_errno.h>
+#include <rte_byteorder.h>
+#include <rte_cryptodev_pmd.h>
+
+#include "virtqueue.h"
+#include "virtio_cryptodev.h"
+#include "virtio_crypto_algs.h"
+
+#ifdef RTE_LIBRTE_PMD_VIRTIO_CRYPTO_DEBUG_DUMP
+#define VIRTIO_DUMP_PACKET(m, len) rte_pktmbuf_dump(stdout, m, len)
+#else
+#define VIRTIO_DUMP_PACKET(m, len) do { } while (0)
+#endif
+
+static void
+vq_ring_free_chain(struct virtqueue *vq, uint16_t desc_idx)
+{
+ struct vring_desc *dp, *dp_tail;
+ struct vq_desc_extra *dxp;
+ uint16_t desc_idx_last = desc_idx;
+
+ dp = &vq->vq_ring.desc[desc_idx];
+ dxp = &vq->vq_descx[desc_idx];
+ vq->vq_free_cnt = (uint16_t)(vq->vq_free_cnt + dxp->ndescs);
+ if ((dp->flags & VRING_DESC_F_INDIRECT) == 0) {
+ while (dp->flags & VRING_DESC_F_NEXT) {
+ desc_idx_last = dp->next;
+ dp = &vq->vq_ring.desc[dp->next];
+ }
+ }
+ dxp->ndescs = 0;
+
+ /*
+ * We must append the existing free chain, if any, to the end of
+ * newly freed chain. If the virtqueue was completely used, then
+ * head would be VQ_RING_DESC_CHAIN_END (ASSERTed above).
+ */
+ if (vq->vq_desc_tail_idx == VQ_RING_DESC_CHAIN_END) {
+ vq->vq_desc_head_idx = desc_idx;
+ } else {
+ dp_tail = &vq->vq_ring.desc[vq->vq_desc_tail_idx];
+ dp_tail->next = desc_idx;
+ }
+
+ vq->vq_desc_tail_idx = desc_idx_last;
+ dp->next = VQ_RING_DESC_CHAIN_END;
+}
+
+static uint16_t
+virtqueue_dequeue_burst_rx(struct virtqueue *vq,
+ struct rte_crypto_op **rx_pkts, uint16_t num)
+{
+ struct vring_used_elem *uep;
+ struct rte_crypto_op *cop;
+ uint16_t used_idx, desc_idx;
+ uint16_t i;
+ struct virtio_crypto_inhdr *inhdr;
+ struct virtio_crypto_op_cookie *op_cookie;
+
+ /* Caller does the check */
+ for (i = 0; i < num ; i++) {
+ used_idx = (uint16_t)(vq->vq_used_cons_idx
+ & (vq->vq_nentries - 1));
+ uep = &vq->vq_ring.used->ring[used_idx];
+ desc_idx = (uint16_t)uep->id;
+ cop = (struct rte_crypto_op *)
+ vq->vq_descx[desc_idx].crypto_op;
+ op_cookie = (struct virtio_crypto_op_cookie *)
+ vq->vq_descx[desc_idx].cookie;
+ inhdr = &(op_cookie->inhdr);
+ switch (inhdr->status) {
+ case VIRTIO_CRYPTO_OK:
+ cop->status = RTE_CRYPTO_OP_STATUS_SUCCESS;
+ break;
+ case VIRTIO_CRYPTO_ERR:
+ cop->status = RTE_CRYPTO_OP_STATUS_ERROR;
+ vq->packets_received_failed++;
+ break;
+ case VIRTIO_CRYPTO_BADMSG:
+ cop->status = RTE_CRYPTO_OP_STATUS_INVALID_ARGS;
+ vq->packets_received_failed++;
+ break;
+ case VIRTIO_CRYPTO_NOTSUPP:
+ cop->status = RTE_CRYPTO_OP_STATUS_INVALID_ARGS;
+ vq->packets_received_failed++;
+ break;
+ case VIRTIO_CRYPTO_INVSESS:
+ cop->status = RTE_CRYPTO_OP_STATUS_INVALID_SESSION;
+ vq->packets_received_failed++;
+ break;
+ default:
+ break;
+ }
+
+ vq->packets_received_total++;
+
+ rx_pkts[i] = cop;
+ rte_mempool_put(vq->mpool, op_cookie);
+
+ vq->vq_used_cons_idx++;
+ vq_ring_free_chain(vq, desc_idx);
+ vq->vq_descx[desc_idx].crypto_op = NULL;
+ }
+
+ return i;
+}
+
+static int
+virtqueue_crypto_sym_pkt_header_arrange(
+ struct rte_crypto_op *cop,
+ struct virtio_crypto_op_data_req *data,
+ struct virtio_crypto_session *session)
+{
+ struct rte_crypto_sym_op *sym_op = cop->sym;
+ struct virtio_crypto_op_data_req *req_data = data;
+ struct virtio_crypto_op_ctrl_req *ctrl = &session->ctrl;
+ struct virtio_crypto_sym_create_session_req *sym_sess_req =
+ &ctrl->u.sym_create_session;
+ struct virtio_crypto_alg_chain_session_para *chain_para =
+ &sym_sess_req->u.chain.para;
+ struct virtio_crypto_cipher_session_para *cipher_para;
+
+ req_data->header.session_id = session->session_id;
+
+ switch (sym_sess_req->op_type) {
+ case VIRTIO_CRYPTO_SYM_OP_CIPHER:
+ req_data->u.sym_req.op_type = VIRTIO_CRYPTO_SYM_OP_CIPHER;
+
+ cipher_para = &sym_sess_req->u.cipher.para;
+ if (cipher_para->op == VIRTIO_CRYPTO_OP_ENCRYPT)
+ req_data->header.opcode = VIRTIO_CRYPTO_CIPHER_ENCRYPT;
+ else
+ req_data->header.opcode = VIRTIO_CRYPTO_CIPHER_DECRYPT;
+
+ req_data->u.sym_req.u.cipher.para.iv_len
+ = session->iv.length;
+
+ req_data->u.sym_req.u.cipher.para.src_data_len =
+ (sym_op->cipher.data.length +
+ sym_op->cipher.data.offset);
+ req_data->u.sym_req.u.cipher.para.dst_data_len =
+ req_data->u.sym_req.u.cipher.para.src_data_len;
+ break;
+ case VIRTIO_CRYPTO_SYM_OP_ALGORITHM_CHAINING:
+ req_data->u.sym_req.op_type =
+ VIRTIO_CRYPTO_SYM_OP_ALGORITHM_CHAINING;
+
+ cipher_para = &chain_para->cipher_param;
+ if (cipher_para->op == VIRTIO_CRYPTO_OP_ENCRYPT)
+ req_data->header.opcode = VIRTIO_CRYPTO_CIPHER_ENCRYPT;
+ else
+ req_data->header.opcode = VIRTIO_CRYPTO_CIPHER_DECRYPT;
+
+ req_data->u.sym_req.u.chain.para.iv_len = session->iv.length;
+ req_data->u.sym_req.u.chain.para.aad_len = session->aad.length;
+
+ req_data->u.sym_req.u.chain.para.src_data_len =
+ (sym_op->cipher.data.length +
+ sym_op->cipher.data.offset);
+ req_data->u.sym_req.u.chain.para.dst_data_len =
+ req_data->u.sym_req.u.chain.para.src_data_len;
+ req_data->u.sym_req.u.chain.para.cipher_start_src_offset =
+ sym_op->cipher.data.offset;
+ req_data->u.sym_req.u.chain.para.len_to_cipher =
+ sym_op->cipher.data.length;
+ req_data->u.sym_req.u.chain.para.hash_start_src_offset =
+ sym_op->auth.data.offset;
+ req_data->u.sym_req.u.chain.para.len_to_hash =
+ sym_op->auth.data.length;
+ req_data->u.sym_req.u.chain.para.aad_len =
+ chain_para->aad_len;
+
+ if (chain_para->hash_mode == VIRTIO_CRYPTO_SYM_HASH_MODE_PLAIN)
+ req_data->u.sym_req.u.chain.para.hash_result_len =
+ chain_para->u.hash_param.hash_result_len;
+ if (chain_para->hash_mode ==
+ VIRTIO_CRYPTO_SYM_HASH_MODE_AUTH)
+ req_data->u.sym_req.u.chain.para.hash_result_len =
+ chain_para->u.mac_param.hash_result_len;
+ break;
+ default:
+ return -1;
+ }
+
+ return 0;
+}
+
+static int
+virtqueue_crypto_sym_enqueue_xmit(
+ struct virtqueue *txvq,
+ struct rte_crypto_op *cop)
+{
+ uint16_t idx = 0;
+ uint16_t num_entry;
+ uint16_t needed = 1;
+ uint16_t head_idx;
+ struct vq_desc_extra *dxp;
+ struct vring_desc *start_dp;
+ struct vring_desc *desc;
+ uint64_t indirect_op_data_req_phys_addr;
+ uint16_t req_data_len = sizeof(struct virtio_crypto_op_data_req);
+ uint32_t indirect_vring_addr_offset = req_data_len +
+ sizeof(struct virtio_crypto_inhdr);
+ struct rte_crypto_sym_op *sym_op = cop->sym;
+ struct virtio_crypto_session *session =
+ (struct virtio_crypto_session *)get_session_private_data(
+ cop->sym->session, cryptodev_virtio_driver_id);
+ struct virtio_crypto_op_data_req *op_data_req;
+ uint32_t hash_result_len = 0;
+ struct virtio_crypto_op_cookie *crypto_op_cookie;
+ struct virtio_crypto_alg_chain_session_para *para;
+
+ if (unlikely(sym_op->m_src->nb_segs != 1))
+ return -EMSGSIZE;
+ if (unlikely(txvq->vq_free_cnt == 0))
+ return -ENOSPC;
+ if (unlikely(txvq->vq_free_cnt < needed))
+ return -EMSGSIZE;
+ head_idx = txvq->vq_desc_head_idx;
+ if (unlikely(head_idx >= txvq->vq_nentries))
+ return -EFAULT;
+ if (unlikely(session == NULL))
+ return -EFAULT;
+
+ dxp = &txvq->vq_descx[head_idx];
+
+ if (rte_mempool_get(txvq->mpool, &dxp->cookie)) {
+ PMD_TX_LOG(ERR, "can not get cookie");
+ return -EFAULT;
+ }
+ crypto_op_cookie = dxp->cookie;
+ indirect_op_data_req_phys_addr =
+ rte_mempool_virt2iova(crypto_op_cookie);
+ op_data_req = (struct virtio_crypto_op_data_req *)crypto_op_cookie;
+
+ if (virtqueue_crypto_sym_pkt_header_arrange(cop, op_data_req, session))
+ return -EFAULT;
+
+ /* status is initialized to VIRTIO_CRYPTO_ERR */
+ ((struct virtio_crypto_inhdr *)
+ ((uint8_t *)op_data_req + req_data_len))->status =
+ VIRTIO_CRYPTO_ERR;
+
+ /* point to indirect vring entry */
+ desc = (struct vring_desc *)
+ ((uint8_t *)op_data_req + indirect_vring_addr_offset);
+ for (idx = 0; idx < (NUM_ENTRY_VIRTIO_CRYPTO_OP - 1); idx++)
+ desc[idx].next = idx + 1;
+ desc[NUM_ENTRY_VIRTIO_CRYPTO_OP - 1].next = VQ_RING_DESC_CHAIN_END;
+
+ idx = 0;
+
+ /* indirect vring: first part, virtio_crypto_op_data_req */
+ desc[idx].addr = indirect_op_data_req_phys_addr;
+ desc[idx].len = req_data_len;
+ desc[idx++].flags = VRING_DESC_F_NEXT;
+
+ /* indirect vring: iv of cipher */
+ if (session->iv.length) {
+ desc[idx].addr = cop->phys_addr + session->iv.offset;
+ desc[idx].len = session->iv.length;
+ desc[idx++].flags = VRING_DESC_F_NEXT;
+ }
+
+ /* indirect vring: additional auth data */
+ if (session->aad.length) {
+ desc[idx].addr = session->aad.phys_addr;
+ desc[idx].len = session->aad.length;
+ desc[idx++].flags = VRING_DESC_F_NEXT;
+ }
+
+ /* indirect vring: src data */
+ desc[idx].addr = rte_pktmbuf_mtophys_offset(sym_op->m_src, 0);
+ desc[idx].len = (sym_op->cipher.data.offset
+ + sym_op->cipher.data.length);
+ desc[idx++].flags = VRING_DESC_F_NEXT;
+
+ /* indirect vring: dst data */
+ if (sym_op->m_dst) {
+ desc[idx].addr = rte_pktmbuf_mtophys_offset(sym_op->m_dst, 0);
+ desc[idx].len = (sym_op->cipher.data.offset
+ + sym_op->cipher.data.length);
+ } else {
+ desc[idx].addr = rte_pktmbuf_mtophys_offset(sym_op->m_src, 0);
+ desc[idx].len = (sym_op->cipher.data.offset
+ + sym_op->cipher.data.length);
+ }
+ desc[idx++].flags = VRING_DESC_F_WRITE | VRING_DESC_F_NEXT;
+
+ /* indirect vring: digest result */
+ para = &(session->ctrl.u.sym_create_session.u.chain.para);
+ if (para->hash_mode == VIRTIO_CRYPTO_SYM_HASH_MODE_PLAIN)
+ hash_result_len = para->u.hash_param.hash_result_len;
+ if (para->hash_mode == VIRTIO_CRYPTO_SYM_HASH_MODE_AUTH)
+ hash_result_len = para->u.mac_param.hash_result_len;
+ if (hash_result_len > 0) {
+ desc[idx].addr = sym_op->auth.digest.phys_addr;
+ desc[idx].len = hash_result_len;
+ desc[idx++].flags = VRING_DESC_F_WRITE | VRING_DESC_F_NEXT;
+ }
+
+ /* indirect vring: last part, status returned */
+ desc[idx].addr = indirect_op_data_req_phys_addr + req_data_len;
+ desc[idx].len = sizeof(struct virtio_crypto_inhdr);
+ desc[idx++].flags = VRING_DESC_F_WRITE;
+
+ num_entry = idx;
+
+ /* save the infos to use when receiving packets */
+ dxp->crypto_op = (void *)cop;
+ dxp->ndescs = needed;
+
+ /* use a single buffer */
+ start_dp = txvq->vq_ring.desc;
+ start_dp[head_idx].addr = indirect_op_data_req_phys_addr +
+ indirect_vring_addr_offset;
+ start_dp[head_idx].len = num_entry * sizeof(struct vring_desc);
+ start_dp[head_idx].flags = VRING_DESC_F_INDIRECT;
+
+ idx = start_dp[head_idx].next;
+ txvq->vq_desc_head_idx = idx;
+ if (txvq->vq_desc_head_idx == VQ_RING_DESC_CHAIN_END)
+ txvq->vq_desc_tail_idx = idx;
+ txvq->vq_free_cnt = (uint16_t)(txvq->vq_free_cnt - needed);
+ vq_update_avail_ring(txvq, head_idx);
+
+ return 0;
+}
+
+static int
+virtqueue_crypto_enqueue_xmit(struct virtqueue *txvq,
+ struct rte_crypto_op *cop)
+{
+ int ret;
+
+ switch (cop->type) {
+ case RTE_CRYPTO_OP_TYPE_SYMMETRIC:
+ ret = virtqueue_crypto_sym_enqueue_xmit(txvq, cop);
+ break;
+ default:
+ PMD_TX_LOG(ERR, "invalid crypto op type %u", cop->type);
+ ret = -EFAULT;
+ break;
+ }
+
+ return ret;
+}
+
+static int
+virtio_crypto_vring_start(struct virtqueue *vq)
+{
+ struct virtio_crypto_hw *hw = vq->hw;
+ int i, size = vq->vq_nentries;
+ struct vring *vr = &vq->vq_ring;
+ uint8_t *ring_mem = vq->vq_ring_virt_mem;
+
+ PMD_INIT_FUNC_TRACE();
+
+ vring_init(vr, size, ring_mem, VIRTIO_PCI_VRING_ALIGN);
+ vq->vq_desc_tail_idx = (uint16_t)(vq->vq_nentries - 1);
+ vq->vq_free_cnt = vq->vq_nentries;
+
+ /* Chain all the descriptors in the ring with an END */
+ for (i = 0; i < size - 1; i++)
+ vr->desc[i].next = (uint16_t)(i + 1);
+ vr->desc[i].next = VQ_RING_DESC_CHAIN_END;
+
+ /*
+ * Disable device(host) interrupting guest
+ */
+ virtqueue_disable_intr(vq);
+
+ /*
+ * Set guest physical address of the virtqueue
+ * in VIRTIO_PCI_QUEUE_PFN config register of device
+ * to share with the backend
+ */
+ if (VTPCI_OPS(hw)->setup_queue(hw, vq) < 0) {
+ PMD_INIT_LOG(ERR, "setup_queue failed");
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+void
+virtio_crypto_ctrlq_start(struct rte_cryptodev *dev)
+{
+ struct virtio_crypto_hw *hw = dev->data->dev_private;
+
+ if (hw->cvq) {
+ virtio_crypto_vring_start(hw->cvq);
+ VIRTQUEUE_DUMP((struct virtqueue *)hw->cvq);
+ }
+}
+
+void
+virtio_crypto_dataq_start(struct rte_cryptodev *dev)
+{
+ /*
+ * Start data vrings
+ * - Setup vring structure for data queues
+ */
+ uint16_t i;
+ struct virtio_crypto_hw *hw = dev->data->dev_private;
+
+ PMD_INIT_FUNC_TRACE();
+
+ /* Start data vring. */
+ for (i = 0; i < hw->max_dataqueues; i++) {
+ virtio_crypto_vring_start(dev->data->queue_pairs[i]);
+ VIRTQUEUE_DUMP((struct virtqueue *)dev->data->queue_pairs[i]);
+ }
+}
+
+/* vring size of data queue is 1024 */
+#define VIRTIO_MBUF_BURST_SZ 1024
+
+uint16_t
+virtio_crypto_pkt_rx_burst(void *tx_queue, struct rte_crypto_op **rx_pkts,
+ uint16_t nb_pkts)
+{
+ struct virtqueue *txvq = tx_queue;
+ uint16_t nb_used, num, nb_rx;
+
+ nb_used = VIRTQUEUE_NUSED(txvq);
+
+ virtio_rmb();
+
+ num = (uint16_t)(likely(nb_used <= nb_pkts) ? nb_used : nb_pkts);
+ num = (uint16_t)(likely(num <= VIRTIO_MBUF_BURST_SZ)
+ ? num : VIRTIO_MBUF_BURST_SZ);
+
+ if (num == 0)
+ return 0;
+
+ nb_rx = virtqueue_dequeue_burst_rx(txvq, rx_pkts, num);
+ PMD_RX_LOG(DEBUG, "used:%d dequeue:%d", nb_used, num);
+
+ return nb_rx;
+}
+
+uint16_t
+virtio_crypto_pkt_tx_burst(void *tx_queue, struct rte_crypto_op **tx_pkts,
+ uint16_t nb_pkts)
+{
+ struct virtqueue *txvq;
+ uint16_t nb_tx;
+ int error;
+
+ if (unlikely(nb_pkts < 1))
+ return nb_pkts;
+ if (unlikely(tx_queue == NULL)) {
+ PMD_TX_LOG(ERR, "tx_queue is NULL");
+ return 0;
+ }
+ txvq = tx_queue;
+
+ PMD_TX_LOG(DEBUG, "%d packets to xmit", nb_pkts);
+
+ for (nb_tx = 0; nb_tx < nb_pkts; nb_tx++) {
+ struct rte_mbuf *txm = tx_pkts[nb_tx]->sym->m_src;
+ /* nb_segs is always 1 at virtio crypto situation */
+ int need = txm->nb_segs - txvq->vq_free_cnt;
+
+ /*
+ * Positive value indicates it hasn't enough space in vring
+ * descriptors
+ */
+ if (unlikely(need > 0)) {
+ /*
+ * try it again because the receive process may be
+ * free some space
+ */
+ need = txm->nb_segs - txvq->vq_free_cnt;
+ if (unlikely(need > 0)) {
+ PMD_TX_LOG(ERR, "No free tx descriptors "
+ "to transmit");
+ break;
+ }
+ }
+
+ txvq->packets_sent_total++;
+
+ /* Enqueue Packet buffers */
+ error = virtqueue_crypto_enqueue_xmit(txvq, tx_pkts[nb_tx]);
+ if (unlikely(error)) {
+ if (error == ENOSPC)
+ PMD_TX_LOG(ERR,
+ "virtqueue_enqueue Free count = 0");
+ else if (error == EMSGSIZE)
+ PMD_TX_LOG(ERR,
+ "virtqueue_enqueue Free count < 1");
+ else
+ PMD_TX_LOG(ERR,
+ "virtqueue_enqueue error: %d", error);
+ txvq->packets_sent_failed++;
+ break;
+ }
+ }
+
+ if (likely(nb_tx)) {
+ vq_update_avail_idx(txvq);
+
+ if (unlikely(virtqueue_kick_prepare(txvq))) {
+ virtqueue_notify(txvq);
+ PMD_TX_LOG(DEBUG, "Notified backend after xmit");
+ }
+ }
+
+ return nb_tx;
+}
--
1.8.3.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* [dpdk-dev] [PATCH v2 4/7] crypto/virtio: add makefile
2018-02-24 13:14 [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver Jay Zhou
` (2 preceding siblings ...)
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 3/7] cryptodev/virtio: core code of crypto devices Jay Zhou
@ 2018-02-24 13:14 ` Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 5/7] cryptodev: add document for virtio crypto PMD Jay Zhou
` (3 subsequent siblings)
7 siblings, 0 replies; 11+ messages in thread
From: Jay Zhou @ 2018-02-24 13:14 UTC (permalink / raw)
To: dev
Cc: pablo.de.lara.guarch, roy.fan.zhang, thomas, arei.gonglei,
xin.zeng, weidong.huang, wangxinxin.wang, longpeng2,
jianjay.zhou
The virtio crypto PMD driver can be compiled now.
Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com>
---
drivers/crypto/Makefile | 1 +
drivers/crypto/virtio/Makefile | 31 ++++++++++++++++++++++
.../virtio/rte_pmd_virtio_crypto_version.map | 3 +++
mk/rte.app.mk | 1 +
4 files changed, 36 insertions(+)
create mode 100644 drivers/crypto/virtio/Makefile
create mode 100644 drivers/crypto/virtio/rte_pmd_virtio_crypto_version.map
diff --git a/drivers/crypto/Makefile b/drivers/crypto/Makefile
index 628bd14..81f91f5 100644
--- a/drivers/crypto/Makefile
+++ b/drivers/crypto/Makefile
@@ -16,5 +16,6 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_MRVL_CRYPTO) += mrvl
DIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL_CRYPTO) += null
DIRS-$(CONFIG_RTE_LIBRTE_PMD_DPAA2_SEC) += dpaa2_sec
DIRS-$(CONFIG_RTE_LIBRTE_PMD_DPAA_SEC) += dpaa_sec
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO) += virtio
include $(RTE_SDK)/mk/rte.subdir.mk
diff --git a/drivers/crypto/virtio/Makefile b/drivers/crypto/virtio/Makefile
new file mode 100644
index 0000000..c4727ea
--- /dev/null
+++ b/drivers/crypto/virtio/Makefile
@@ -0,0 +1,31 @@
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright(c) 2018 HUAWEI TECHNOLOGIES CO., LTD.
+
+include $(RTE_SDK)/mk/rte.vars.mk
+
+#
+# library name
+#
+LIB = librte_pmd_virtio_crypto.a
+
+CFLAGS += -O3
+CFLAGS += $(WERROR_FLAGS)
+
+EXPORT_MAP := rte_pmd_virtio_crypto_version.map
+
+LIBABIVER := 1
+
+#
+# all source are stored in SRCS-y
+#
+SRCS-$(CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO) += virtqueue.c
+SRCS-$(CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO) += virtio_pci.c
+SRCS-$(CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO) += virtio_rxtx.c
+SRCS-$(CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO) += virtio_cryptodev.c
+
+# this lib depends upon:
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO) += lib/librte_eal
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO) += lib/librte_mempool lib/librte_mbuf
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO) += lib/librte_cryptodev
+
+include $(RTE_SDK)/mk/rte.lib.mk
diff --git a/drivers/crypto/virtio/rte_pmd_virtio_crypto_version.map b/drivers/crypto/virtio/rte_pmd_virtio_crypto_version.map
new file mode 100644
index 0000000..de8e412
--- /dev/null
+++ b/drivers/crypto/virtio/rte_pmd_virtio_crypto_version.map
@@ -0,0 +1,3 @@
+DPDK_18.05 {
+ local: *;
+};
diff --git a/mk/rte.app.mk b/mk/rte.app.mk
index 3eb41d1..4aaf93a 100644
--- a/mk/rte.app.mk
+++ b/mk/rte.app.mk
@@ -194,6 +194,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_AESNI_GCM) += -L$(AESNI_MULTI_BUFFER_LIB_PATH)
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_OPENSSL) += -lrte_pmd_openssl -lcrypto
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_NULL_CRYPTO) += -lrte_pmd_null_crypto
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_QAT) += -lrte_pmd_qat -lcrypto
+_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO) += -lrte_pmd_virtio_crypto -lcrypto
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SNOW3G) += -lrte_pmd_snow3g
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_SNOW3G) += -L$(LIBSSO_SNOW3G_PATH)/build -lsso_snow3g
_LDLIBS-$(CONFIG_RTE_LIBRTE_PMD_KASUMI) += -lrte_pmd_kasumi
--
1.8.3.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* [dpdk-dev] [PATCH v2 5/7] cryptodev: add document for virtio crypto PMD
2018-02-24 13:14 [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver Jay Zhou
` (3 preceding siblings ...)
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 4/7] crypto/virtio: add makefile Jay Zhou
@ 2018-02-24 13:14 ` Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 6/7] cryptodev: add function tests " Jay Zhou
` (2 subsequent siblings)
7 siblings, 0 replies; 11+ messages in thread
From: Jay Zhou @ 2018-02-24 13:14 UTC (permalink / raw)
To: dev
Cc: pablo.de.lara.guarch, roy.fan.zhang, thomas, arei.gonglei,
xin.zeng, weidong.huang, wangxinxin.wang, longpeng2,
jianjay.zhou
Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com>
---
doc/guides/cryptodevs/features/virtio.ini | 20 +++++++
doc/guides/cryptodevs/virtio.rst | 93 +++++++++++++++++++++++++++++++
2 files changed, 113 insertions(+)
create mode 100644 doc/guides/cryptodevs/features/virtio.ini
create mode 100644 doc/guides/cryptodevs/virtio.rst
diff --git a/doc/guides/cryptodevs/features/virtio.ini b/doc/guides/cryptodevs/features/virtio.ini
new file mode 100644
index 0000000..43396ff
--- /dev/null
+++ b/doc/guides/cryptodevs/features/virtio.ini
@@ -0,0 +1,20 @@
+; Supported features of the 'virtio' crypto driver.
+;
+; Refer to default.ini for the full list of available PMD features.
+;
+[Features]
+Symmetric crypto = Y
+Sym operation chaining = N
+HW Accelerated = N
+Mbuf scatter gather = N
+
+;
+; Supported crypto algorithms of the 'virtio' crypto driver.
+;
+[Cipher]
+AES CBC (128) = Y
+AES CBC (192) = Y
+AES CBC (256) = Y
+AES CTR (128) = Y
+AES CTR (192) = Y
+AES CTR (256) = Y
diff --git a/doc/guides/cryptodevs/virtio.rst b/doc/guides/cryptodevs/virtio.rst
new file mode 100644
index 0000000..8c06eb5
--- /dev/null
+++ b/doc/guides/cryptodevs/virtio.rst
@@ -0,0 +1,93 @@
+.. SPDX-License-Identifier: BSD-3-Clause
+ Copyright(c) 2015-2016 Intel Corporation.
+
+Virtio Crypto Poll Mode Driver
+==============================
+
+The virtio crypto PMD provides poll mode driver support for the virtio crypto
+device.
+
+Features
+--------
+
+The virtio crypto PMD has support for:
+
+Cipher algorithms:
+
+* ``RTE_CRYPTO_CIPHER_AES_CBC``
+* ``RTE_CRYPTO_CIPHER_AES_CTR``
+
+Limitations
+-----------
+
+* Only supports the session-oriented API implementation (session-less APIs are
+ not supported).
+* Only supports modern mode since virtio crypto conforms to virtio-1.0.
+* Only has two types of queues: data queue and control queue. These two queues
+ only support indirect buffers to communication with the virtio backend.
+* Only supports cipher only algorithms since the virtio backend of qemu side
+ only supports cipher only now.
+* Does not support Link State interrupt.
+* Does not support runtime configuration.
+
+Virtio crypto PMD Rx/Tx Callbacks
+---------------------------------
+
+Rx callbacks:
+
+* ``virtio_crypto_pkt_rx_burst``
+
+Tx callbacks:
+
+* ``virtio_crypto_pkt_tx_burst``
+
+Installation
+------------
+
+Quick instructions are as follows:
+
+Firstly build QEMU with libgcrypt cryptography support.
+QEMU can then be started using the following parameters:
+
+.. code-block:: console
+
+ qemu-system-x86_64 \
+ [...] \
+ -object cryptodev-backend-builtin,id=cryptodev0 \
+ -device virtio-crypto-pci,id=crypto0,cryptodev=cryptodev0 \
+ [...]
+
+Secondly bind the uio_generic driver for the virtio-crypto device.
+For example, 0000:00:04.0 is the domain, bus, device and function
+number of the virtio-crypto device:
+
+.. code-block:: console
+
+ modprobe uio_pci_generic
+ echo -n 0000:00:04.0 > /sys/bus/pci/drivers/virtio-pci/unbind
+ echo "1af4 1054" > /sys/bus/pci/drivers/uio_pci_generic/new_id
+
+Finally the front-end virtio crypto PMD driver can be installed:
+
+.. code-block:: console
+
+ cd to the top-level DPDK directory
+ sed -i 's,\(CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO\)=n,\1=y,' config/common_base
+ make config T=x86_64-native-linuxapp-gcc
+ make install T=x86_64-native-linuxapp-gcc
+
+Tests
+-----
+
+The test cases can be tested as below:
+
+.. code-block:: console
+
+ cd to the top-level DPDK directory
+ export RTE_TARGET=x86_64-native-linuxapp-gcc
+ export RTE_SDK=`pwd`
+ cd to test/test
+ type the command "make" to compile
+ reserve enough huge pages
+ run the tests with "./test"
+ type the command "cryptodev_virtio_autotest" to test
--
1.8.3.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* [dpdk-dev] [PATCH v2 6/7] cryptodev: add function tests for virtio crypto PMD
2018-02-24 13:14 [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver Jay Zhou
` (4 preceding siblings ...)
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 5/7] cryptodev: add document for virtio crypto PMD Jay Zhou
@ 2018-02-24 13:14 ` Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 7/7] MAINTAINERS: add myself as virtio crypto PMD maintainer Jay Zhou
2018-03-21 1:44 ` [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver Zhoujian (jay)
7 siblings, 0 replies; 11+ messages in thread
From: Jay Zhou @ 2018-02-24 13:14 UTC (permalink / raw)
To: dev
Cc: pablo.de.lara.guarch, roy.fan.zhang, thomas, arei.gonglei,
xin.zeng, weidong.huang, wangxinxin.wang, longpeng2,
jianjay.zhou
Only RTE_CRYPTO_CIPHER_AES_CBC and RTE_CRYPTO_CIPHER_AES_CTR cipher
algorithms are tested now, it is limited by the cryptodev-backend-builtin
of qemu side.
Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com>
---
test/test/test_cryptodev.c | 49 +++++++++++++++++++++++++++++
test/test/test_cryptodev.h | 1 +
test/test/test_cryptodev_aes_test_vectors.h | 45 +++++++++++++++++---------
test/test/test_cryptodev_blockcipher.c | 9 +++++-
test/test/test_cryptodev_blockcipher.h | 1 +
5 files changed, 89 insertions(+), 16 deletions(-)
diff --git a/test/test/test_cryptodev.c b/test/test/test_cryptodev.c
index 1417482..357e1df 100644
--- a/test/test/test_cryptodev.c
+++ b/test/test/test_cryptodev.c
@@ -1765,6 +1765,26 @@ struct crypto_unittest_params {
}
static int
+test_AES_cipheronly_virtio_all(void)
+{
+ struct crypto_testsuite_params *ts_params = &testsuite_params;
+ int status;
+
+ status = test_blockcipher_all_tests(ts_params->mbuf_pool,
+ ts_params->op_mpool,
+ ts_params->session_mpool,
+ ts_params->valid_devs[0],
+ rte_cryptodev_driver_id_get(
+ RTE_STR(CRYPTODEV_NAME_VIRTIO_PMD)),
+ BLKCIPHER_AES_CIPHERONLY_TYPE);
+
+ TEST_ASSERT_EQUAL(status, 0, "Test failed");
+
+ return TEST_SUCCESS;
+}
+
+
+static int
test_AES_chain_dpaa_sec_all(void)
{
struct crypto_testsuite_params *ts_params = &testsuite_params;
@@ -8767,6 +8787,18 @@ struct test_crypto_vector {
}
};
+static struct unit_test_suite cryptodev_virtio_testsuite = {
+ .suite_name = "Crypto VIRTIO Unit Test Suite",
+ .setup = testsuite_setup,
+ .teardown = testsuite_teardown,
+ .unit_test_cases = {
+ TEST_CASE_ST(ut_setup, ut_teardown,
+ test_AES_cipheronly_virtio_all),
+
+ TEST_CASES_END() /**< NULL terminate unit test array */
+ }
+};
+
static struct unit_test_suite cryptodev_aesni_mb_testsuite = {
.suite_name = "Crypto Device AESNI MB Unit Test Suite",
.setup = testsuite_setup,
@@ -9664,6 +9696,22 @@ struct test_crypto_vector {
}
static int
+test_cryptodev_virtio(void /*argv __rte_unused, int argc __rte_unused*/)
+{
+ gbl_driver_id = rte_cryptodev_driver_id_get(
+ RTE_STR(CRYPTODEV_NAME_VIRTIO_PMD));
+
+ if (gbl_driver_id == -1) {
+ RTE_LOG(ERR, USER1, "VIRTIO PMD must be loaded. Check if "
+ "CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO is enabled "
+ "in config file to run this testsuite.\n");
+ return TEST_FAILED;
+ }
+
+ return unit_test_suite_runner(&cryptodev_virtio_testsuite);
+}
+
+static int
test_cryptodev_aesni_mb(void /*argv __rte_unused, int argc __rte_unused*/)
{
gbl_driver_id = rte_cryptodev_driver_id_get(
@@ -9879,3 +9927,4 @@ struct test_crypto_vector {
REGISTER_TEST_COMMAND(cryptodev_sw_mrvl_autotest, test_cryptodev_mrvl);
REGISTER_TEST_COMMAND(cryptodev_dpaa2_sec_autotest, test_cryptodev_dpaa2_sec);
REGISTER_TEST_COMMAND(cryptodev_dpaa_sec_autotest, test_cryptodev_dpaa_sec);
+REGISTER_TEST_COMMAND(cryptodev_virtio_autotest, test_cryptodev_virtio);
diff --git a/test/test/test_cryptodev.h b/test/test/test_cryptodev.h
index 8cdc087..c311277 100644
--- a/test/test/test_cryptodev.h
+++ b/test/test/test_cryptodev.h
@@ -61,6 +61,7 @@
#define CRYPTODEV_NAME_DPAA2_SEC_PMD crypto_dpaa2_sec
#define CRYPTODEV_NAME_SCHEDULER_PMD crypto_scheduler
#define CRYPTODEV_NAME_MRVL_PMD crypto_mrvl
+#define CRYPTODEV_NAME_VIRTIO_PMD crypto_virtio
/**
* Write (spread) data from buffer to mbuf data
diff --git a/test/test/test_cryptodev_aes_test_vectors.h b/test/test/test_cryptodev_aes_test_vectors.h
index 3577ef4..5e33b51 100644
--- a/test/test/test_cryptodev_aes_test_vectors.h
+++ b/test/test/test_cryptodev_aes_test_vectors.h
@@ -1249,7 +1249,8 @@
BLOCKCIPHER_TEST_TARGET_PMD_SCHEDULER |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_MRVL
+ BLOCKCIPHER_TEST_TARGET_PMD_MRVL |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-128-CBC HMAC-SHA1 Encryption Digest "
@@ -1526,7 +1527,8 @@
BLOCKCIPHER_TEST_TARGET_PMD_SCHEDULER |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_MRVL
+ BLOCKCIPHER_TEST_TARGET_PMD_MRVL |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-128-CBC Decryption",
@@ -1538,7 +1540,8 @@
BLOCKCIPHER_TEST_TARGET_PMD_SCHEDULER |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_MRVL
+ BLOCKCIPHER_TEST_TARGET_PMD_MRVL |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-192-CBC Encryption",
@@ -1549,7 +1552,8 @@
BLOCKCIPHER_TEST_TARGET_PMD_MB |
BLOCKCIPHER_TEST_TARGET_PMD_SCHEDULER |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC
+ BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-192-CBC Encryption Scater gather",
@@ -1570,7 +1574,8 @@
BLOCKCIPHER_TEST_TARGET_PMD_MB |
BLOCKCIPHER_TEST_TARGET_PMD_SCHEDULER |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC
+ BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-192-CBC Decryption Scatter Gather",
@@ -1590,7 +1595,8 @@
BLOCKCIPHER_TEST_TARGET_PMD_SCHEDULER |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_MRVL
+ BLOCKCIPHER_TEST_TARGET_PMD_MRVL |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-256-CBC Decryption",
@@ -1602,7 +1608,8 @@
BLOCKCIPHER_TEST_TARGET_PMD_SCHEDULER |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_MRVL
+ BLOCKCIPHER_TEST_TARGET_PMD_MRVL |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-256-CBC OOP Encryption",
@@ -1612,7 +1619,8 @@
.pmd_mask = BLOCKCIPHER_TEST_TARGET_PMD_OPENSSL |
BLOCKCIPHER_TEST_TARGET_PMD_QAT |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC
+ BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-256-CBC OOP Decryption",
@@ -1622,7 +1630,8 @@
.pmd_mask = BLOCKCIPHER_TEST_TARGET_PMD_OPENSSL |
BLOCKCIPHER_TEST_TARGET_PMD_QAT |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC
+ BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-128-CTR Encryption",
@@ -1634,7 +1643,8 @@
BLOCKCIPHER_TEST_TARGET_PMD_SCHEDULER |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_MRVL
+ BLOCKCIPHER_TEST_TARGET_PMD_MRVL |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-128-CTR Decryption",
@@ -1646,7 +1656,8 @@
BLOCKCIPHER_TEST_TARGET_PMD_SCHEDULER |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_MRVL
+ BLOCKCIPHER_TEST_TARGET_PMD_MRVL |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-192-CTR Encryption",
@@ -1657,7 +1668,8 @@
BLOCKCIPHER_TEST_TARGET_PMD_MB |
BLOCKCIPHER_TEST_TARGET_PMD_SCHEDULER |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC
+ BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-192-CTR Decryption",
@@ -1668,7 +1680,8 @@
BLOCKCIPHER_TEST_TARGET_PMD_MB |
BLOCKCIPHER_TEST_TARGET_PMD_SCHEDULER |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC
+ BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-256-CTR Encryption",
@@ -1680,7 +1693,8 @@
BLOCKCIPHER_TEST_TARGET_PMD_SCHEDULER |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_MRVL
+ BLOCKCIPHER_TEST_TARGET_PMD_MRVL |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-256-CTR Decryption",
@@ -1692,7 +1706,8 @@
BLOCKCIPHER_TEST_TARGET_PMD_SCHEDULER |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC |
BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC |
- BLOCKCIPHER_TEST_TARGET_PMD_MRVL
+ BLOCKCIPHER_TEST_TARGET_PMD_MRVL |
+ BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO
},
{
.test_descr = "AES-128-CTR Encryption (12-byte IV)",
diff --git a/test/test/test_cryptodev_blockcipher.c b/test/test/test_cryptodev_blockcipher.c
index ed06618..9eee350 100644
--- a/test/test/test_cryptodev_blockcipher.c
+++ b/test/test/test_cryptodev_blockcipher.c
@@ -68,6 +68,8 @@
RTE_STR(CRYPTODEV_NAME_DPAA_SEC_PMD));
int mrvl_pmd = rte_cryptodev_driver_id_get(
RTE_STR(CRYPTODEV_NAME_MRVL_PMD));
+ int virtio_pmd = rte_cryptodev_driver_id_get(
+ RTE_STR(CRYPTODEV_NAME_VIRTIO_PMD));
int nb_segs = 1;
@@ -94,7 +96,8 @@
driver_id == qat_pmd ||
driver_id == openssl_pmd ||
driver_id == armv8_pmd ||
- driver_id == mrvl_pmd) { /* Fall through */
+ driver_id == mrvl_pmd ||
+ driver_id == virtio_pmd) { /* Fall through */
digest_len = tdata->digest.len;
} else if (driver_id == aesni_mb_pmd ||
driver_id == scheduler_pmd) {
@@ -569,6 +572,8 @@
RTE_STR(CRYPTODEV_NAME_QAT_SYM_PMD));
int mrvl_pmd = rte_cryptodev_driver_id_get(
RTE_STR(CRYPTODEV_NAME_MRVL_PMD));
+ int virtio_pmd = rte_cryptodev_driver_id_get(
+ RTE_STR(CRYPTODEV_NAME_VIRTIO_PMD));
switch (test_type) {
case BLKCIPHER_AES_CHAIN_TYPE:
@@ -631,6 +636,8 @@
target_pmd_mask = BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC;
else if (driver_id == mrvl_pmd)
target_pmd_mask = BLOCKCIPHER_TEST_TARGET_PMD_MRVL;
+ else if (driver_id == virtio_pmd)
+ target_pmd_mask = BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO;
else
TEST_ASSERT(0, "Unrecognized cryptodev type");
diff --git a/test/test/test_cryptodev_blockcipher.h b/test/test/test_cryptodev_blockcipher.h
index edbdaab..49a35bd 100644
--- a/test/test/test_cryptodev_blockcipher.h
+++ b/test/test/test_cryptodev_blockcipher.h
@@ -27,6 +27,7 @@
#define BLOCKCIPHER_TEST_TARGET_PMD_DPAA2_SEC 0x0020 /* DPAA2_SEC flag */
#define BLOCKCIPHER_TEST_TARGET_PMD_DPAA_SEC 0x0040 /* DPAA_SEC flag */
#define BLOCKCIPHER_TEST_TARGET_PMD_MRVL 0x0080 /* Marvell flag */
+#define BLOCKCIPHER_TEST_TARGET_PMD_VIRTIO 0x0100 /* VIRTIO flag */
#define BLOCKCIPHER_TEST_OP_CIPHER (BLOCKCIPHER_TEST_OP_ENCRYPT | \
BLOCKCIPHER_TEST_OP_DECRYPT)
--
1.8.3.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* [dpdk-dev] [PATCH v2 7/7] MAINTAINERS: add myself as virtio crypto PMD maintainer
2018-02-24 13:14 [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver Jay Zhou
` (5 preceding siblings ...)
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 6/7] cryptodev: add function tests " Jay Zhou
@ 2018-02-24 13:14 ` Jay Zhou
2018-03-21 1:44 ` [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver Zhoujian (jay)
7 siblings, 0 replies; 11+ messages in thread
From: Jay Zhou @ 2018-02-24 13:14 UTC (permalink / raw)
To: dev
Cc: pablo.de.lara.guarch, roy.fan.zhang, thomas, arei.gonglei,
xin.zeng, weidong.huang, wangxinxin.wang, longpeng2,
jianjay.zhou
Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com>
---
MAINTAINERS | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 4f1f33b..3efc8df 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -718,6 +718,12 @@ F: drivers/crypto/snow3g/
F: doc/guides/cryptodevs/snow3g.rst
F: doc/guides/cryptodevs/features/snow3g.ini
+Virtio
+M: Jay Zhou <jianjay.zhou@huawei.com>
+F: drivers/crypto/virtio/
+F: doc/guides/cryptodevs/virtio.rst
+F: doc/guides/cryptodevs/features/virtio.ini
+
ZUC
M: Pablo de Lara <pablo.de.lara.guarch@intel.com>
F: drivers/crypto/zuc/
--
1.8.3.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver
2018-02-24 13:14 [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver Jay Zhou
` (6 preceding siblings ...)
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 7/7] MAINTAINERS: add myself as virtio crypto PMD maintainer Jay Zhou
@ 2018-03-21 1:44 ` Zhoujian (jay)
7 siblings, 0 replies; 11+ messages in thread
From: Zhoujian (jay) @ 2018-03-21 1:44 UTC (permalink / raw)
To: dev
Cc: pablo.de.lara.guarch, roy.fan.zhang, thomas, Gonglei (Arei),
xin.zeng, Huangweidong (C), wangxin (U),
longpeng
Gentle ping...
> -----Original Message-----
> From: Zhoujian (jay)
> Sent: Saturday, February 24, 2018 9:14 PM
> To: dev@dpdk.org
> Cc: pablo.de.lara.guarch@intel.com; roy.fan.zhang@intel.com;
> thomas@monjalon.net; Gonglei (Arei) <arei.gonglei@huawei.com>;
> xin.zeng@intel.com; Huangweidong (C) <weidong.huang@huawei.com>; wangxin (U)
> <wangxinxin.wang@huawei.com>; longpeng <longpeng2@huawei.com>; Zhoujian (jay)
> <jianjay.zhou@huawei.com>
> Subject: [PATCH v2 0/7] crypto: add virtio poll mode driver
>
> This patch series introduce virtio crypto poll mode driver.
>
> Since it is limited by the backend of the virtio-crypto, this patch series
> only supports a limited subset of crypto services.
> Only the following algorithms are tested:
>
> Cipher algorithms:
> - RTE_CRYPTO_CIPHER_AES_CBC (128-bit, 192-bit and 256-bit keys)
> - RTE_CRYPTO_CIPHER_AES_CTR (128-bit, 192-bit and 256-bit keys)
>
> Firstly build QEMU with libgcrypt cryptography support.
> QEMU can then be started using the following parameters:
> qemu-system-x86_64 \
> [...] \
> -object cryptodev-backend-builtin,id=cryptodev0 \
> -device virtio-crypto-pci,id=crypto0,cryptodev=cryptodev0 \
> [...]
>
> Bind the uio_generic driver for the virtio-crypto device.
> For example, 0000:00:04.0 is the domain, bus, device and function number of
> the virtio-crypto device:
> modprobe uio_pci_generic
> echo -n 0000:00:04.0 > /sys/bus/pci/drivers/virtio-pci/unbind
> echo "1af4 1054" > /sys/bus/pci/drivers/uio_pci_generic/new_id
>
> The front-end virtio crypto PMD driver can be installed:
> cd to the top-level DPDK directory
> sed -i 's,\(CONFIG_RTE_LIBRTE_PMD_VIRTIO_CRYPTO\)=n,\1=y,'
> config/common_base
> make config T=x86_64-native-linuxapp-gcc
> make install T=x86_64-native-linuxapp-gcc
>
> The test cases can be compiled as below:
> cd to the top-level DPDK directory
> export RTE_TARGET=x86_64-native-linuxapp-gcc
> export RTE_SDK=`pwd`
> cd to test/test
> make
> ./test (MUST reserve enough huge pages memory)
> type the command "cryptodev_virtio_autotest" to test
>
> The result should be like this:
> RTE>>cryptodev_virtio_autotest
> + ------------------------------------------------------- + + Test Suite :
> Crypto VIRTIO Unit Test Suite + --------------------------------------------
> ----------- +
> 0) TestCase AES-128-CBC Encryption PASS
> 1) TestCase AES-128-CBC Decryption PASS
> 2) TestCase AES-192-CBC Encryption PASS
> 3) TestCase AES-192-CBC Decryption PASS
> 4) TestCase AES-256-CBC Encryption PASS
> 5) TestCase AES-256-CBC Decryption PASS
> 6) TestCase AES-256-CBC OOP Encryption PASS
> 7) TestCase AES-256-CBC OOP Decryption PASS
> 8) TestCase AES-128-CTR Encryption PASS
> 9) TestCase AES-128-CTR Decryption PASS
> 10) TestCase AES-192-CTR Encryption PASS
> 11) TestCase AES-192-CTR Decryption PASS
> 12) TestCase AES-256-CTR Encryption PASS
> 13) TestCase AES-256-CTR Decryption PASS + TestCase [ 0] :
> test_AES_cipheronly_virtio_all succeeded + ---------------------------------
> ---------------------- + + Test Suite Summary
> + Tests Total : 1
> + Tests Skipped : 0
> + Tests Executed : 1
> + Tests Unsupported: 0
> + Tests Passed : 1
> + Tests Failed : 0
> + ------------------------------------------------------- + Test OK
>
> Please help to review, thanks!
>
> Changes in v2:
> - using pre-allocated mempool instead of rte_malloc to improve performance
> [Fan]
> - split the patch into a patchset [Fan]
> - using linux/virtio_crypto.h instead of creating a copy of the file [Fan]
> - update doc/guides/cryptodevs for describing virtio crypto PMD [Fan]
> - update copyright
> - delete virtio legacy mode code since virtio-crypto conforms to virtio-1.0
> - refine the function and variable names
> - fix errors and warnings reported by checkpatch
>
> Jay Zhou (7):
> crypto/virtio: add virtio related fundamental functions
> crpyto/virtio: add crypto related session structure
> cryptodev/virtio: core code of crypto devices
> crypto/virtio: add makefile
> cryptodev: add document for virtio crypto PMD
> cryptodev: add function tests for virtio crypto PMD
> MAINTAINERS: add myself as virtio crypto PMD maintainer
>
> MAINTAINERS | 6 +
> config/common_base | 20 +
> doc/guides/cryptodevs/features/virtio.ini | 20 +
> doc/guides/cryptodevs/virtio.rst | 93 ++
> drivers/crypto/Makefile | 1 +
> drivers/crypto/virtio/Makefile | 31 +
> .../virtio/rte_pmd_virtio_crypto_version.map | 3 +
> drivers/crypto/virtio/virtio_crypto_algs.h | 27 +
> drivers/crypto/virtio/virtio_cryptodev.c | 1544
> ++++++++++++++++++++
> drivers/crypto/virtio/virtio_cryptodev.h | 66 +
> drivers/crypto/virtio/virtio_logs.h | 47 +
> drivers/crypto/virtio/virtio_pci.c | 460 ++++++
> drivers/crypto/virtio/virtio_pci.h | 252 ++++
> drivers/crypto/virtio/virtio_ring.h | 137 ++
> drivers/crypto/virtio/virtio_rxtx.c | 533 +++++++
> drivers/crypto/virtio/virtqueue.c | 43 +
> drivers/crypto/virtio/virtqueue.h | 176 +++
> mk/rte.app.mk | 1 +
> test/test/test_cryptodev.c | 49 +
> test/test/test_cryptodev.h | 1 +
> test/test/test_cryptodev_aes_test_vectors.h | 45 +-
> test/test/test_cryptodev_blockcipher.c | 9 +-
> test/test/test_cryptodev_blockcipher.h | 1 +
> 23 files changed, 3549 insertions(+), 16 deletions(-) create mode 100644
> doc/guides/cryptodevs/features/virtio.ini
> create mode 100644 doc/guides/cryptodevs/virtio.rst create mode 100644
> drivers/crypto/virtio/Makefile create mode 100644
> drivers/crypto/virtio/rte_pmd_virtio_crypto_version.map
> create mode 100644 drivers/crypto/virtio/virtio_crypto_algs.h
> create mode 100644 drivers/crypto/virtio/virtio_cryptodev.c
> create mode 100644 drivers/crypto/virtio/virtio_cryptodev.h
> create mode 100644 drivers/crypto/virtio/virtio_logs.h
> create mode 100644 drivers/crypto/virtio/virtio_pci.c
> create mode 100644 drivers/crypto/virtio/virtio_pci.h
> create mode 100644 drivers/crypto/virtio/virtio_ring.h
> create mode 100644 drivers/crypto/virtio/virtio_rxtx.c
> create mode 100644 drivers/crypto/virtio/virtqueue.c create mode 100644
> drivers/crypto/virtio/virtqueue.h
>
> --
> 1.8.3.1
>
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [dpdk-dev] [PATCH v2 3/7] cryptodev/virtio: core code of crypto devices
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 3/7] cryptodev/virtio: core code of crypto devices Jay Zhou
@ 2018-03-21 4:08 ` Zhang, Roy Fan
2018-03-21 4:58 ` Zhoujian (jay)
0 siblings, 1 reply; 11+ messages in thread
From: Zhang, Roy Fan @ 2018-03-21 4:08 UTC (permalink / raw)
To: 'Jay Zhou', dev
Cc: De Lara Guarch, Pablo, thomas, arei.gonglei, Zeng, Xin,
weidong.huang, wangxinxin.wang, longpeng2, Zhang, Roy Fan
Hi Jay,
Excellent work! The patch, although need some minor rework, has improved the performance.
Some comments:
1. you need to set up capabilities for virtio_crypto PMD. As in Qemu vhost crypto proxy backend only AESCBC and SHA1 are supported (in cryptodev_vhost_user_init() definition), I believe in this version these two algorithms shall be enough. Actually for the same reason I suggest you to remove all AES_CTR test cases in the virtio_crypto PMD functional test, as they will fail when vhost_user crypto backend is used.
You may use driver/crypto/qat/qat_crypto_capabilities.h as example. The const capabilities array shall be returned to the application when virtio_crypto_dev_info_get() is called.
2. there is a bug in virtio_crypto_queue_setup(), you declared " uint32_t i, j;" while i may be used uninitialized later.
Regards,
Fan
> diff --git a/drivers/crypto/virtio/virtio_rxtx.c
> b/drivers/crypto/virtio/virtio_rxtx.c
> new file mode 100644
> index 0000000..b7d8066
> --- /dev/null
> +++ b/drivers/crypto/virtio/virtio_rxtx.c
> @@ -0,0 +1,533 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(c) 2018 HUAWEI TECHNOLOGIES CO., LTD.
> + */
> +
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [dpdk-dev] [PATCH v2 3/7] cryptodev/virtio: core code of crypto devices
2018-03-21 4:08 ` Zhang, Roy Fan
@ 2018-03-21 4:58 ` Zhoujian (jay)
0 siblings, 0 replies; 11+ messages in thread
From: Zhoujian (jay) @ 2018-03-21 4:58 UTC (permalink / raw)
To: Zhang, Roy Fan, dev
Cc: De Lara Guarch, Pablo, thomas, Gonglei (Arei),
Zeng, Xin, Huangweidong (C), wangxin (U),
longpeng
Hi Fan,
Really thank for your response and testing!
> -----Original Message-----
> From: Zhang, Roy Fan [mailto:roy.fan.zhang@intel.com]
> Sent: Wednesday, March 21, 2018 12:09 PM
> To: Zhoujian (jay) <jianjay.zhou@huawei.com>; dev@dpdk.org
> Cc: De Lara Guarch, Pablo <pablo.de.lara.guarch@intel.com>;
> thomas@monjalon.net; Gonglei (Arei) <arei.gonglei@huawei.com>; Zeng, Xin
> <xin.zeng@intel.com>; Huangweidong (C) <weidong.huang@huawei.com>; wangxin (U)
> <wangxinxin.wang@huawei.com>; longpeng <longpeng2@huawei.com>; Zhang, Roy Fan
> <roy.fan.zhang@intel.com>
> Subject: RE: [PATCH v2 3/7] cryptodev/virtio: core code of crypto devices
>
> Hi Jay,
>
> Excellent work! The patch, although need some minor rework, has improved the
> performance.
> Some comments:
>
> 1. you need to set up capabilities for virtio_crypto PMD. As in Qemu vhost
> crypto proxy backend only AESCBC and SHA1 are supported (in
> cryptodev_vhost_user_init() definition), I believe in this version these two
> algorithms shall be enough. Actually for the same reason I suggest you to
> remove all AES_CTR test cases in the virtio_crypto PMD functional test, as
> they will fail when vhost_user crypto backend is used.
>
> You may use driver/crypto/qat/qat_crypto_capabilities.h as example. The const
> capabilities array shall be returned to the application when
> virtio_crypto_dev_info_get() is called.
Okay. Thank for your suggestion.
> 2. there is a bug in virtio_crypto_queue_setup(), you declared " uint32_t i,
> j;" while i may be used uninitialized later.
Maybe the compiled result is different using different GCC, but will fix then.
Regards,
Jay
>
> Regards,
> Fan
>
> > diff --git a/drivers/crypto/virtio/virtio_rxtx.c
> > b/drivers/crypto/virtio/virtio_rxtx.c
> > new file mode 100644
> > index 0000000..b7d8066
> > --- /dev/null
> > +++ b/drivers/crypto/virtio/virtio_rxtx.c
> > @@ -0,0 +1,533 @@
> > +/* SPDX-License-Identifier: BSD-3-Clause
> > + * Copyright(c) 2018 HUAWEI TECHNOLOGIES CO., LTD.
> > + */
> > +
^ permalink raw reply [flat|nested] 11+ messages in thread
end of thread, other threads:[~2018-03-21 4:59 UTC | newest]
Thread overview: 11+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2018-02-24 13:14 [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 1/7] crypto/virtio: add virtio related fundamental functions Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 2/7] crpyto/virtio: add crypto related session structure Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 3/7] cryptodev/virtio: core code of crypto devices Jay Zhou
2018-03-21 4:08 ` Zhang, Roy Fan
2018-03-21 4:58 ` Zhoujian (jay)
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 4/7] crypto/virtio: add makefile Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 5/7] cryptodev: add document for virtio crypto PMD Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 6/7] cryptodev: add function tests " Jay Zhou
2018-02-24 13:14 ` [dpdk-dev] [PATCH v2 7/7] MAINTAINERS: add myself as virtio crypto PMD maintainer Jay Zhou
2018-03-21 1:44 ` [dpdk-dev] [PATCH v2 0/7] crypto: add virtio poll mode driver Zhoujian (jay)
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).