Hi Akhil,
> -----Original Message-----
> From: Akhil Goyal <gakhil@marvell.com>
> Sent: Wednesday, October 12, 2022 1:13 PM
> To: Chautru, Nicolas <nicolas.chautru@intel.com>; dev@dpdk.org;
> maxime.coquelin@redhat.com
> Cc: trix@redhat.com; mdr@ashroe.eu; Richardson, Bruce
> <bruce.richardson@intel.com>; hemant.agrawal@nxp.com;
> david.marchand@redhat.com; stephen@networkplumber.org; Vargas,
> Hernan <hernan.vargas@intel.com>; thomas@monjalon.net
> Subject: RE: [EXT] [PATCH v12 06/16] baseband/acc: add info get function
> for ACC200
>
> > diff --git a/drivers/baseband/acc/acc200_pmd.h
> > b/drivers/baseband/acc/acc200_pmd.h
> > index acead1d17c..587b6f6d5b 100644
> > --- a/drivers/baseband/acc/acc200_pmd.h
> > +++ b/drivers/baseband/acc/acc200_pmd.h
> > @@ -8,6 +8,7 @@
> > #include "acc_common.h"
> > #include "acc200_pf_enum.h"
> > #include "acc200_vf_enum.h"
> > +#include "rte_acc200_cfg.h"
> >
> You should really fix your build procedure.
> There have been 4-5 versions of this patchset due to build failures.
> Even this v12 is not built properly.
> Including this file "rte_acc200_cfg.h" here is not required and is not available
> until this patch.
> This was highlighted in v11 but is still not fixed in v12.
This was my intent but somehow I clearly didn't do that amend for that commit as intended. I should have triple checked.
> I removed this line while merging this series as I don't want to work again on
> this series.
>
> Please take care of builds in future.
Thanks and sorry for the churn. We have an issue in our CICD and the patch by patch build but I should have checked manually all meson build.
> I have not merged the last patch for test-bbdev as it is not reviewed yet and
> was introduced in v10 only.
> You should not introduce such new patches so late in the cycle of review of a
> patchset.
>
> With above build issue fixed, series applied to dpdk-next-crypto except the
> last patch.
>
> @Maxime can you please review the last patch of the series also.
>
Thanks again.
@@ -8,6 +8,7 @@
#include "acc_common.h"
#include "acc200_pf_enum.h"
#include "acc200_vf_enum.h"
+#include "rte_acc200_cfg.h"
/* Helper macro for logging */
#define rte_bbdev_log(level, fmt, ...) \
@@ -29,6 +29,189 @@ RTE_LOG_REGISTER_DEFAULT(acc200_logtype, DEBUG);
RTE_LOG_REGISTER_DEFAULT(acc200_logtype, NOTICE);
#endif
+/* Calculate the offset of the enqueue register. */
+static inline uint32_t
+queue_offset(bool pf_device, uint8_t vf_id, uint8_t qgrp_id, uint16_t aq_id)
+{
+ if (pf_device)
+ return ((vf_id << 12) + (qgrp_id << 7) + (aq_id << 3) +
+ HWPfQmgrIngressAq);
+ else
+ return ((qgrp_id << 7) + (aq_id << 3) +
+ HWVfQmgrIngressAq);
+}
+
+enum {UL_4G = 0, UL_5G, DL_4G, DL_5G, FFT, NUM_ACC};
+
+/* Return the queue topology for a Queue Group Index. */
+static inline void
+qtopFromAcc(struct rte_acc_queue_topology **qtop, int acc_enum, struct rte_acc_conf *acc_conf)
+{
+ struct rte_acc_queue_topology *p_qtop;
+ p_qtop = NULL;
+
+ switch (acc_enum) {
+ case UL_4G:
+ p_qtop = &(acc_conf->q_ul_4g);
+ break;
+ case UL_5G:
+ p_qtop = &(acc_conf->q_ul_5g);
+ break;
+ case DL_4G:
+ p_qtop = &(acc_conf->q_dl_4g);
+ break;
+ case DL_5G:
+ p_qtop = &(acc_conf->q_dl_5g);
+ break;
+ case FFT:
+ p_qtop = &(acc_conf->q_fft);
+ break;
+ default:
+ /* NOTREACHED. */
+ rte_bbdev_log(ERR, "Unexpected error evaluating %s using %d", __func__, acc_enum);
+ break;
+ }
+ *qtop = p_qtop;
+}
+
+static void
+initQTop(struct rte_acc_conf *acc_conf)
+{
+ acc_conf->q_ul_4g.num_aqs_per_groups = 0;
+ acc_conf->q_ul_4g.num_qgroups = 0;
+ acc_conf->q_ul_4g.first_qgroup_index = -1;
+ acc_conf->q_ul_5g.num_aqs_per_groups = 0;
+ acc_conf->q_ul_5g.num_qgroups = 0;
+ acc_conf->q_ul_5g.first_qgroup_index = -1;
+ acc_conf->q_dl_4g.num_aqs_per_groups = 0;
+ acc_conf->q_dl_4g.num_qgroups = 0;
+ acc_conf->q_dl_4g.first_qgroup_index = -1;
+ acc_conf->q_dl_5g.num_aqs_per_groups = 0;
+ acc_conf->q_dl_5g.num_qgroups = 0;
+ acc_conf->q_dl_5g.first_qgroup_index = -1;
+ acc_conf->q_fft.num_aqs_per_groups = 0;
+ acc_conf->q_fft.num_qgroups = 0;
+ acc_conf->q_fft.first_qgroup_index = -1;
+}
+
+static inline void
+updateQtop(uint8_t acc, uint8_t qg, struct rte_acc_conf *acc_conf, struct acc_device *d) {
+ uint32_t reg;
+ struct rte_acc_queue_topology *q_top = NULL;
+ uint16_t aq;
+
+ qtopFromAcc(&q_top, acc, acc_conf);
+ if (unlikely(q_top == NULL))
+ return;
+ q_top->num_qgroups++;
+ if (q_top->first_qgroup_index == -1) {
+ q_top->first_qgroup_index = qg;
+ /* Can be optimized to assume all are enabled by default. */
+ reg = acc_reg_read(d, queue_offset(d->pf_device, 0, qg, ACC200_NUM_AQS - 1));
+ if (reg & ACC_QUEUE_ENABLE) {
+ q_top->num_aqs_per_groups = ACC200_NUM_AQS;
+ return;
+ }
+ q_top->num_aqs_per_groups = 0;
+ for (aq = 0; aq < ACC200_NUM_AQS; aq++) {
+ reg = acc_reg_read(d, queue_offset(d->pf_device, 0, qg, aq));
+ if (reg & ACC_QUEUE_ENABLE)
+ q_top->num_aqs_per_groups++;
+ }
+ }
+}
+
+/* Fetch configuration enabled for the PF/VF using MMIO Read (slow). */
+static inline void
+fetch_acc200_config(struct rte_bbdev *dev)
+{
+ struct acc_device *d = dev->data->dev_private;
+ struct rte_acc_conf *acc_conf = &d->acc_conf;
+ const struct acc200_registry_addr *reg_addr;
+ uint8_t acc, qg;
+ uint32_t reg_aq, reg_len0, reg_len1, reg0, reg1;
+ uint32_t reg_mode, idx;
+ struct rte_acc_queue_topology *q_top = NULL;
+ int qman_func_id[ACC200_NUM_ACCS] = {ACC_ACCMAP_0, ACC_ACCMAP_1,
+ ACC_ACCMAP_2, ACC_ACCMAP_3, ACC_ACCMAP_4};
+
+ /* No need to retrieve the configuration is already done. */
+ if (d->configured)
+ return;
+
+ /* Choose correct registry addresses for the device type. */
+ if (d->pf_device)
+ reg_addr = &pf_reg_addr;
+ else
+ reg_addr = &vf_reg_addr;
+
+ d->ddr_size = 0;
+
+ /* Single VF Bundle by VF. */
+ acc_conf->num_vf_bundles = 1;
+ initQTop(acc_conf);
+
+ reg0 = acc_reg_read(d, reg_addr->qman_group_func);
+ reg1 = acc_reg_read(d, reg_addr->qman_group_func + 4);
+ for (qg = 0; qg < ACC200_NUM_QGRPS; qg++) {
+ reg_aq = acc_reg_read(d, queue_offset(d->pf_device, 0, qg, 0));
+ if (reg_aq & ACC_QUEUE_ENABLE) {
+ if (qg < ACC_NUM_QGRPS_PER_WORD)
+ idx = (reg0 >> (qg * 4)) & 0x7;
+ else
+ idx = (reg1 >> ((qg -
+ ACC_NUM_QGRPS_PER_WORD) * 4)) & 0x7;
+ if (idx < ACC200_NUM_ACCS) {
+ acc = qman_func_id[idx];
+ updateQtop(acc, qg, acc_conf, d);
+ }
+ }
+ }
+
+ /* Check the depth of the AQs. */
+ reg_len0 = acc_reg_read(d, reg_addr->depth_log0_offset);
+ reg_len1 = acc_reg_read(d, reg_addr->depth_log1_offset);
+ for (acc = 0; acc < NUM_ACC; acc++) {
+ qtopFromAcc(&q_top, acc, acc_conf);
+ if (q_top->first_qgroup_index < ACC_NUM_QGRPS_PER_WORD)
+ q_top->aq_depth_log2 = (reg_len0 >> (q_top->first_qgroup_index * 4)) & 0xF;
+ else
+ q_top->aq_depth_log2 = (reg_len1 >> ((q_top->first_qgroup_index -
+ ACC_NUM_QGRPS_PER_WORD) * 4)) & 0xF;
+ }
+
+ /* Read PF mode. */
+ if (d->pf_device) {
+ reg_mode = acc_reg_read(d, HWPfHiPfMode);
+ acc_conf->pf_mode_en = (reg_mode == ACC_PF_VAL) ? 1 : 0;
+ } else {
+ reg_mode = acc_reg_read(d, reg_addr->hi_mode);
+ acc_conf->pf_mode_en = reg_mode & 1;
+ }
+
+ rte_bbdev_log_debug(
+ "%s Config LLR SIGN IN/OUT %s %s QG %u %u %u %u %u AQ %u %u %u %u %u Len %u %u %u %u %u\n",
+ (d->pf_device) ? "PF" : "VF",
+ (acc_conf->input_pos_llr_1_bit) ? "POS" : "NEG",
+ (acc_conf->output_pos_llr_1_bit) ? "POS" : "NEG",
+ acc_conf->q_ul_4g.num_qgroups,
+ acc_conf->q_dl_4g.num_qgroups,
+ acc_conf->q_ul_5g.num_qgroups,
+ acc_conf->q_dl_5g.num_qgroups,
+ acc_conf->q_fft.num_qgroups,
+ acc_conf->q_ul_4g.num_aqs_per_groups,
+ acc_conf->q_dl_4g.num_aqs_per_groups,
+ acc_conf->q_ul_5g.num_aqs_per_groups,
+ acc_conf->q_dl_5g.num_aqs_per_groups,
+ acc_conf->q_fft.num_aqs_per_groups,
+ acc_conf->q_ul_4g.aq_depth_log2,
+ acc_conf->q_dl_4g.aq_depth_log2,
+ acc_conf->q_ul_5g.aq_depth_log2,
+ acc_conf->q_dl_5g.aq_depth_log2,
+ acc_conf->q_fft.aq_depth_log2);
+}
+
+/* Free memory used for software rings. */
static int
acc200_dev_close(struct rte_bbdev *dev)
{
@@ -38,9 +221,57 @@ acc200_dev_close(struct rte_bbdev *dev)
return 0;
}
+/* Get ACC200 device info. */
+static void
+acc200_dev_info_get(struct rte_bbdev *dev,
+ struct rte_bbdev_driver_info *dev_info)
+{
+ struct acc_device *d = dev->data->dev_private;
+ int i;
+ static const struct rte_bbdev_op_cap bbdev_capabilities[] = {
+ RTE_BBDEV_END_OF_CAPABILITIES_LIST()
+ };
+
+ static struct rte_bbdev_queue_conf default_queue_conf;
+ default_queue_conf.socket = dev->data->socket_id;
+ default_queue_conf.queue_size = ACC_MAX_QUEUE_DEPTH;
+
+ dev_info->driver_name = dev->device->driver->name;
+
+ /* Read and save the populated config from ACC200 registers. */
+ fetch_acc200_config(dev);
+
+ /* Exposed number of queues. */
+ dev_info->num_queues[RTE_BBDEV_OP_NONE] = 0;
+ dev_info->num_queues[RTE_BBDEV_OP_TURBO_DEC] = 0;
+ dev_info->num_queues[RTE_BBDEV_OP_TURBO_ENC] = 0;
+ dev_info->num_queues[RTE_BBDEV_OP_LDPC_DEC] = 0;
+ dev_info->num_queues[RTE_BBDEV_OP_LDPC_ENC] = 0;
+ dev_info->num_queues[RTE_BBDEV_OP_FFT] = 0;
+ dev_info->queue_priority[RTE_BBDEV_OP_TURBO_DEC] = 0;
+ dev_info->queue_priority[RTE_BBDEV_OP_TURBO_ENC] = 0;
+ dev_info->queue_priority[RTE_BBDEV_OP_LDPC_DEC] = 0;
+ dev_info->queue_priority[RTE_BBDEV_OP_LDPC_ENC] = 0;
+ dev_info->queue_priority[RTE_BBDEV_OP_FFT] = 0;
+ dev_info->max_num_queues = 0;
+ for (i = RTE_BBDEV_OP_NONE; i <= RTE_BBDEV_OP_FFT; i++)
+ dev_info->max_num_queues += dev_info->num_queues[i];
+ dev_info->queue_size_lim = ACC_MAX_QUEUE_DEPTH;
+ dev_info->hardware_accelerated = true;
+ dev_info->max_dl_queue_priority =
+ d->acc_conf.q_dl_4g.num_qgroups - 1;
+ dev_info->max_ul_queue_priority =
+ d->acc_conf.q_ul_4g.num_qgroups - 1;
+ dev_info->default_queue_conf = default_queue_conf;
+ dev_info->cpu_flag_reqs = NULL;
+ dev_info->min_alignment = 1;
+ dev_info->capabilities = bbdev_capabilities;
+ dev_info->harq_buffer_size = 0;
+}
static const struct rte_bbdev_ops acc200_bbdev_ops = {
.close = acc200_dev_close,
+ .info_get = acc200_dev_info_get,
};
/* ACC200 PCI PF address map. */