summaryrefslogtreecommitdiff
path: root/drivers
diff options
context:
space:
mode:
authorKhoronzhuk, Ivan <ivan.khoronzhuk@ti.com>2014-09-05 19:02:47 +0300
committerTom Rini <trini@ti.com>2014-10-23 11:27:04 -0400
commitef4547176d0fa4d43d060a58c0c902add7fe0aed (patch)
tree1098e15423022644bffc6e0beb9e6069584bf8bb /drivers
parented948e2959acbad5289de1c733b3aeddd93208fe (diff)
dma: keystone_nav: move keystone_nav driver to driver/dma/
The keystone_nav is used by drivers/net/keystone_net.c driver to send and receive packets, but currently it's placed at keystone arch sources. So it should be in the drivers directory also. It's separate driver that can be used for sending and receiving pktdma packets by others drivers also. This patch just move this driver to appropriate directory and doesn't add any functional changes. Acked-by: Murali Karicheri <m-karicheri2@ti.com> Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@ti.com>
Diffstat (limited to 'drivers')
-rw-r--r--drivers/dma/Makefile1
-rw-r--r--drivers/dma/keystone_nav.c354
-rw-r--r--drivers/net/keystone_net.c2
3 files changed, 356 insertions, 1 deletions
diff --git a/drivers/dma/Makefile b/drivers/dma/Makefile
index a79c3919dd..519dd0d347 100644
--- a/drivers/dma/Makefile
+++ b/drivers/dma/Makefile
@@ -8,3 +8,4 @@
obj-$(CONFIG_FSLDMAFEC) += MCD_tasksInit.o MCD_dmaApi.o MCD_tasks.o
obj-$(CONFIG_APBH_DMA) += apbh_dma.o
obj-$(CONFIG_FSL_DMA) += fsl_dma.o
+obj-$(CONFIG_TI_KSNAV) += keystone_nav.o
diff --git a/drivers/dma/keystone_nav.c b/drivers/dma/keystone_nav.c
new file mode 100644
index 0000000000..d960fbb266
--- /dev/null
+++ b/drivers/dma/keystone_nav.c
@@ -0,0 +1,354 @@
+/*
+ * Multicore Navigator driver for TI Keystone 2 devices.
+ *
+ * (C) Copyright 2012-2014
+ * Texas Instruments Incorporated, <www.ti.com>
+ *
+ * SPDX-License-Identifier: GPL-2.0+
+ */
+#include <common.h>
+#include <asm/io.h>
+#include <asm/ti-common/keystone_nav.h>
+
+struct qm_config qm_memmap = {
+ .stat_cfg = CONFIG_KSNAV_QM_QUEUE_STATUS_BASE,
+ .queue = (void *)CONFIG_KSNAV_QM_MANAGER_QUEUES_BASE,
+ .mngr_vbusm = CONFIG_KSNAV_QM_BASE_ADDRESS,
+ .i_lram = CONFIG_KSNAV_QM_LINK_RAM_BASE,
+ .proxy = (void *)CONFIG_KSNAV_QM_MANAGER_Q_PROXY_BASE,
+ .status_ram = CONFIG_KSNAV_QM_STATUS_RAM_BASE,
+ .mngr_cfg = (void *)CONFIG_KSNAV_QM_CONF_BASE,
+ .intd_cfg = CONFIG_KSNAV_QM_INTD_CONF_BASE,
+ .desc_mem = (void *)CONFIG_KSNAV_QM_DESC_SETUP_BASE,
+ .region_num = CONFIG_KSNAV_QM_REGION_NUM,
+ .pdsp_cmd = CONFIG_KSNAV_QM_PDSP1_CMD_BASE,
+ .pdsp_ctl = CONFIG_KSNAV_QM_PDSP1_CTRL_BASE,
+ .pdsp_iram = CONFIG_KSNAV_QM_PDSP1_IRAM_BASE,
+ .qpool_num = CONFIG_KSNAV_QM_QPOOL_NUM,
+};
+
+/*
+ * We are going to use only one type of descriptors - host packet
+ * descriptors. We staticaly allocate memory for them here
+ */
+struct qm_host_desc desc_pool[HDESC_NUM] __aligned(sizeof(struct qm_host_desc));
+
+static struct qm_config *qm_cfg;
+
+inline int num_of_desc_to_reg(int num_descr)
+{
+ int j, num;
+
+ for (j = 0, num = 32; j < 15; j++, num *= 2) {
+ if (num_descr <= num)
+ return j;
+ }
+
+ return 15;
+}
+
+int _qm_init(struct qm_config *cfg)
+{
+ u32 j;
+
+ qm_cfg = cfg;
+
+ qm_cfg->mngr_cfg->link_ram_base0 = qm_cfg->i_lram;
+ qm_cfg->mngr_cfg->link_ram_size0 = HDESC_NUM * 8;
+ qm_cfg->mngr_cfg->link_ram_base1 = 0;
+ qm_cfg->mngr_cfg->link_ram_size1 = 0;
+ qm_cfg->mngr_cfg->link_ram_base2 = 0;
+
+ qm_cfg->desc_mem[0].base_addr = (u32)desc_pool;
+ qm_cfg->desc_mem[0].start_idx = 0;
+ qm_cfg->desc_mem[0].desc_reg_size =
+ (((sizeof(struct qm_host_desc) >> 4) - 1) << 16) |
+ num_of_desc_to_reg(HDESC_NUM);
+
+ memset(desc_pool, 0, sizeof(desc_pool));
+ for (j = 0; j < HDESC_NUM; j++)
+ qm_push(&desc_pool[j], qm_cfg->qpool_num);
+
+ return QM_OK;
+}
+
+int qm_init(void)
+{
+ return _qm_init(&qm_memmap);
+}
+
+void qm_close(void)
+{
+ u32 j;
+
+ if (qm_cfg == NULL)
+ return;
+
+ queue_close(qm_cfg->qpool_num);
+
+ qm_cfg->mngr_cfg->link_ram_base0 = 0;
+ qm_cfg->mngr_cfg->link_ram_size0 = 0;
+ qm_cfg->mngr_cfg->link_ram_base1 = 0;
+ qm_cfg->mngr_cfg->link_ram_size1 = 0;
+ qm_cfg->mngr_cfg->link_ram_base2 = 0;
+
+ for (j = 0; j < qm_cfg->region_num; j++) {
+ qm_cfg->desc_mem[j].base_addr = 0;
+ qm_cfg->desc_mem[j].start_idx = 0;
+ qm_cfg->desc_mem[j].desc_reg_size = 0;
+ }
+
+ qm_cfg = NULL;
+}
+
+void qm_push(struct qm_host_desc *hd, u32 qnum)
+{
+ u32 regd;
+
+ if (!qm_cfg)
+ return;
+
+ cpu_to_bus((u32 *)hd, sizeof(struct qm_host_desc)/4);
+ regd = (u32)hd | ((sizeof(struct qm_host_desc) >> 4) - 1);
+ writel(regd, &qm_cfg->queue[qnum].ptr_size_thresh);
+}
+
+void qm_buff_push(struct qm_host_desc *hd, u32 qnum,
+ void *buff_ptr, u32 buff_len)
+{
+ hd->orig_buff_len = buff_len;
+ hd->buff_len = buff_len;
+ hd->orig_buff_ptr = (u32)buff_ptr;
+ hd->buff_ptr = (u32)buff_ptr;
+ qm_push(hd, qnum);
+}
+
+struct qm_host_desc *qm_pop(u32 qnum)
+{
+ u32 uhd;
+
+ if (!qm_cfg)
+ return NULL;
+
+ uhd = readl(&qm_cfg->queue[qnum].ptr_size_thresh) & ~0xf;
+ if (uhd)
+ cpu_to_bus((u32 *)uhd, sizeof(struct qm_host_desc)/4);
+
+ return (struct qm_host_desc *)uhd;
+}
+
+struct qm_host_desc *qm_pop_from_free_pool(void)
+{
+ if (!qm_cfg)
+ return NULL;
+
+ return qm_pop(qm_cfg->qpool_num);
+}
+
+void queue_close(u32 qnum)
+{
+ struct qm_host_desc *hd;
+
+ while ((hd = qm_pop(qnum)))
+ ;
+}
+
+/**
+ * DMA API
+ */
+struct pktdma_cfg netcp_pktdma = {
+ .global = (void *)CONFIG_KSNAV_NETCP_PDMA_CTRL_BASE,
+ .tx_ch = (void *)CONFIG_KSNAV_NETCP_PDMA_TX_BASE,
+ .tx_ch_num = CONFIG_KSNAV_NETCP_PDMA_TX_CH_NUM,
+ .rx_ch = (void *)CONFIG_KSNAV_NETCP_PDMA_RX_BASE,
+ .rx_ch_num = CONFIG_KSNAV_NETCP_PDMA_RX_CH_NUM,
+ .tx_sched = (u32 *)CONFIG_KSNAV_NETCP_PDMA_SCHED_BASE,
+ .rx_flows = (void *)CONFIG_KSNAV_NETCP_PDMA_RX_FLOW_BASE,
+ .rx_flow_num = CONFIG_KSNAV_NETCP_PDMA_RX_FLOW_NUM,
+ .rx_free_q = CONFIG_KSNAV_NETCP_PDMA_RX_FREE_QUEUE,
+ .rx_rcv_q = CONFIG_KSNAV_NETCP_PDMA_RX_RCV_QUEUE,
+ .tx_snd_q = CONFIG_KSNAV_NETCP_PDMA_TX_SND_QUEUE,
+};
+
+struct pktdma_cfg *netcp;
+
+static int netcp_rx_disable(void)
+{
+ u32 j, v, k;
+
+ for (j = 0; j < netcp->rx_ch_num; j++) {
+ v = readl(&netcp->rx_ch[j].cfg_a);
+ if (!(v & CPDMA_CHAN_A_ENABLE))
+ continue;
+
+ writel(v | CPDMA_CHAN_A_TDOWN, &netcp->rx_ch[j].cfg_a);
+ for (k = 0; k < TDOWN_TIMEOUT_COUNT; k++) {
+ udelay(100);
+ v = readl(&netcp->rx_ch[j].cfg_a);
+ if (!(v & CPDMA_CHAN_A_ENABLE))
+ continue;
+ }
+ /* TODO: teardown error on if TDOWN_TIMEOUT_COUNT is reached */
+ }
+
+ /* Clear all of the flow registers */
+ for (j = 0; j < netcp->rx_flow_num; j++) {
+ writel(0, &netcp->rx_flows[j].control);
+ writel(0, &netcp->rx_flows[j].tags);
+ writel(0, &netcp->rx_flows[j].tag_sel);
+ writel(0, &netcp->rx_flows[j].fdq_sel[0]);
+ writel(0, &netcp->rx_flows[j].fdq_sel[1]);
+ writel(0, &netcp->rx_flows[j].thresh[0]);
+ writel(0, &netcp->rx_flows[j].thresh[1]);
+ writel(0, &netcp->rx_flows[j].thresh[2]);
+ }
+
+ return QM_OK;
+}
+
+static int netcp_tx_disable(void)
+{
+ u32 j, v, k;
+
+ for (j = 0; j < netcp->tx_ch_num; j++) {
+ v = readl(&netcp->tx_ch[j].cfg_a);
+ if (!(v & CPDMA_CHAN_A_ENABLE))
+ continue;
+
+ writel(v | CPDMA_CHAN_A_TDOWN, &netcp->tx_ch[j].cfg_a);
+ for (k = 0; k < TDOWN_TIMEOUT_COUNT; k++) {
+ udelay(100);
+ v = readl(&netcp->tx_ch[j].cfg_a);
+ if (!(v & CPDMA_CHAN_A_ENABLE))
+ continue;
+ }
+ /* TODO: teardown error on if TDOWN_TIMEOUT_COUNT is reached */
+ }
+
+ return QM_OK;
+}
+
+static int _netcp_init(struct pktdma_cfg *netcp_cfg,
+ struct rx_buff_desc *rx_buffers)
+{
+ u32 j, v;
+ struct qm_host_desc *hd;
+ u8 *rx_ptr;
+
+ if (netcp_cfg == NULL || rx_buffers == NULL ||
+ rx_buffers->buff_ptr == NULL || qm_cfg == NULL)
+ return QM_ERR;
+
+ netcp = netcp_cfg;
+ netcp->rx_flow = rx_buffers->rx_flow;
+
+ /* init rx queue */
+ rx_ptr = rx_buffers->buff_ptr;
+
+ for (j = 0; j < rx_buffers->num_buffs; j++) {
+ hd = qm_pop(qm_cfg->qpool_num);
+ if (hd == NULL)
+ return QM_ERR;
+
+ qm_buff_push(hd, netcp->rx_free_q,
+ rx_ptr, rx_buffers->buff_len);
+
+ rx_ptr += rx_buffers->buff_len;
+ }
+
+ netcp_rx_disable();
+
+ /* configure rx channels */
+ v = CPDMA_REG_VAL_MAKE_RX_FLOW_A(1, 1, 0, 0, 0, 0, 0, netcp->rx_rcv_q);
+ writel(v, &netcp->rx_flows[netcp->rx_flow].control);
+ writel(0, &netcp->rx_flows[netcp->rx_flow].tags);
+ writel(0, &netcp->rx_flows[netcp->rx_flow].tag_sel);
+
+ v = CPDMA_REG_VAL_MAKE_RX_FLOW_D(0, netcp->rx_free_q, 0,
+ netcp->rx_free_q);
+
+ writel(v, &netcp->rx_flows[netcp->rx_flow].fdq_sel[0]);
+ writel(v, &netcp->rx_flows[netcp->rx_flow].fdq_sel[1]);
+ writel(0, &netcp->rx_flows[netcp->rx_flow].thresh[0]);
+ writel(0, &netcp->rx_flows[netcp->rx_flow].thresh[1]);
+ writel(0, &netcp->rx_flows[netcp->rx_flow].thresh[2]);
+
+ for (j = 0; j < netcp->rx_ch_num; j++)
+ writel(CPDMA_CHAN_A_ENABLE, &netcp->rx_ch[j].cfg_a);
+
+ /* configure tx channels */
+ /* Disable loopback in the tx direction */
+ writel(0, &netcp->global->emulation_control);
+
+ /* Set QM base address, only for K2x devices */
+ writel(CONFIG_KSNAV_QM_BASE_ADDRESS, &netcp->global->qm_base_addr[0]);
+
+ /* Enable all channels. The current state isn't important */
+ for (j = 0; j < netcp->tx_ch_num; j++) {
+ writel(0, &netcp->tx_ch[j].cfg_b);
+ writel(CPDMA_CHAN_A_ENABLE, &netcp->tx_ch[j].cfg_a);
+ }
+
+ return QM_OK;
+}
+
+int netcp_init(struct rx_buff_desc *rx_buffers)
+{
+ return _netcp_init(&netcp_pktdma, rx_buffers);
+}
+
+int netcp_close(void)
+{
+ if (!netcp)
+ return QM_ERR;
+
+ netcp_tx_disable();
+ netcp_rx_disable();
+
+ queue_close(netcp->rx_free_q);
+ queue_close(netcp->rx_rcv_q);
+ queue_close(netcp->tx_snd_q);
+
+ return QM_OK;
+}
+
+int netcp_send(u32 *pkt, int num_bytes, u32 swinfo2)
+{
+ struct qm_host_desc *hd;
+
+ hd = qm_pop(qm_cfg->qpool_num);
+ if (hd == NULL)
+ return QM_ERR;
+
+ hd->desc_info = num_bytes;
+ hd->swinfo[2] = swinfo2;
+ hd->packet_info = qm_cfg->qpool_num;
+
+ qm_buff_push(hd, netcp->tx_snd_q, pkt, num_bytes);
+
+ return QM_OK;
+}
+
+void *netcp_recv(u32 **pkt, int *num_bytes)
+{
+ struct qm_host_desc *hd;
+
+ hd = qm_pop(netcp->rx_rcv_q);
+ if (!hd)
+ return NULL;
+
+ *pkt = (u32 *)hd->buff_ptr;
+ *num_bytes = hd->desc_info & 0x3fffff;
+
+ return hd;
+}
+
+void netcp_release_rxhd(void *hd)
+{
+ struct qm_host_desc *_hd = (struct qm_host_desc *)hd;
+
+ _hd->buff_len = _hd->orig_buff_len;
+ _hd->buff_ptr = _hd->orig_buff_ptr;
+
+ qm_push(_hd, netcp->rx_free_q);
+}
diff --git a/drivers/net/keystone_net.c b/drivers/net/keystone_net.c
index d22b722bc5..1cfe6542d8 100644
--- a/drivers/net/keystone_net.c
+++ b/drivers/net/keystone_net.c
@@ -14,7 +14,7 @@
#include <malloc.h>
#include <asm/arch/emac_defs.h>
#include <asm/arch/psc_defs.h>
-#include <asm/arch/keystone_nav.h>
+#include <asm/ti-common/keystone_nav.h>
unsigned int emac_dbg;