u-boot/drivers/net/fm/fm.c

421 lines
11 KiB
C
Raw Normal View History

powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
/*
* Copyright 2009-2011 Freescale Semiconductor, Inc.
* Dave Liu <daveliu@freescale.com>
*
* SPDX-License-Identifier: GPL-2.0+
powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
*/
#include <common.h>
#include <malloc.h>
#include <asm/io.h>
#include <asm/errno.h>
#include "fm.h"
#include "../../qe/qe.h" /* For struct qe_firmware */
#ifdef CONFIG_SYS_QE_FMAN_FW_IN_NAND
powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
#include <nand.h>
#elif defined(CONFIG_SYS_QE_FW_IN_SPIFLASH)
#include <spi_flash.h>
#elif defined(CONFIG_SYS_QE_FMAN_FW_IN_MMC)
powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
#include <mmc.h>
#endif
struct fm_muram muram[CONFIG_SYS_NUM_FMAN];
u32 fm_muram_base(int fm_idx)
{
return muram[fm_idx].base;
}
u32 fm_muram_alloc(int fm_idx, u32 size, u32 align)
{
u32 ret;
u32 align_mask, off;
u32 save;
align_mask = align - 1;
save = muram[fm_idx].alloc;
off = save & align_mask;
if (off != 0)
muram[fm_idx].alloc += (align - off);
off = size & align_mask;
if (off != 0)
size += (align - off);
if ((muram[fm_idx].alloc + size) >= muram[fm_idx].top) {
muram[fm_idx].alloc = save;
printf("%s: run out of ram.\n", __func__);
}
ret = muram[fm_idx].alloc;
muram[fm_idx].alloc += size;
memset((void *)ret, 0, size);
return ret;
}
static void fm_init_muram(int fm_idx, void *reg)
{
u32 base = (u32)reg;
muram[fm_idx].base = base;
muram[fm_idx].size = CONFIG_SYS_FM_MURAM_SIZE;
muram[fm_idx].alloc = base + FM_MURAM_RES_SIZE;
muram[fm_idx].top = base + CONFIG_SYS_FM_MURAM_SIZE;
}
/*
* fm_upload_ucode - Fman microcode upload worker function
*
* This function does the actual uploading of an Fman microcode
* to an Fman.
*/
static void fm_upload_ucode(int fm_idx, struct fm_imem *imem,
u32 *ucode, unsigned int size)
{
unsigned int i;
unsigned int timeout = 1000000;
/* enable address auto increase */
out_be32(&imem->iadd, IRAM_IADD_AIE);
/* write microcode to IRAM */
for (i = 0; i < size / 4; i++)
out_be32(&imem->idata, ucode[i]);
/* verify if the writing is over */
out_be32(&imem->iadd, 0);
while ((in_be32(&imem->idata) != ucode[0]) && --timeout)
;
if (!timeout)
printf("Fman%u: microcode upload timeout\n", fm_idx + 1);
/* enable microcode from IRAM */
out_be32(&imem->iready, IRAM_READY);
}
/*
* Upload an Fman firmware
*
* This function is similar to qe_upload_firmware(), exception that it uploads
* a microcode to the Fman instead of the QE.
*
* Because the process for uploading a microcode to the Fman is similar for
* that of the QE, the QE firmware binary format is used for Fman microcode.
* It should be possible to unify these two functions, but for now we keep them
* separate.
*/
static int fman_upload_firmware(int fm_idx,
struct fm_imem *fm_imem,
const struct qe_firmware *firmware)
{
unsigned int i;
u32 crc;
size_t calc_size = sizeof(struct qe_firmware);
size_t length;
const struct qe_header *hdr;
if (!firmware) {
printf("Fman%u: Invalid address for firmware\n", fm_idx + 1);
return -EINVAL;
}
hdr = &firmware->header;
length = be32_to_cpu(hdr->length);
/* Check the magic */
if ((hdr->magic[0] != 'Q') || (hdr->magic[1] != 'E') ||
(hdr->magic[2] != 'F')) {
printf("Fman%u: Data at %p is not a firmware\n", fm_idx + 1,
firmware);
return -EPERM;
}
/* Check the version */
if (hdr->version != 1) {
printf("Fman%u: Unsupported firmware version %u\n", fm_idx + 1,
hdr->version);
return -EPERM;
}
/* Validate some of the fields */
if ((firmware->count != 1)) {
printf("Fman%u: Invalid data in firmware header\n", fm_idx + 1);
return -EINVAL;
}
/* Validate the length and check if there's a CRC */
calc_size += (firmware->count - 1) * sizeof(struct qe_microcode);
for (i = 0; i < firmware->count; i++)
/*
* For situations where the second RISC uses the same microcode
* as the first, the 'code_offset' and 'count' fields will be
* zero, so it's okay to add those.
*/
calc_size += sizeof(u32) *
be32_to_cpu(firmware->microcode[i].count);
/* Validate the length */
if (length != calc_size + sizeof(u32)) {
printf("Fman%u: Invalid length in firmware header\n",
fm_idx + 1);
return -EPERM;
}
/*
* Validate the CRC. We would normally call crc32_no_comp(), but that
* function isn't available unless you turn on JFFS support.
*/
crc = be32_to_cpu(*(u32 *)((void *)firmware + calc_size));
if (crc != (crc32(-1, (const void *)firmware, calc_size) ^ -1)) {
printf("Fman%u: Firmware CRC is invalid\n", fm_idx + 1);
return -EIO;
}
/* Loop through each microcode. */
for (i = 0; i < firmware->count; i++) {
const struct qe_microcode *ucode = &firmware->microcode[i];
/* Upload a microcode if it's present */
if (ucode->code_offset) {
u32 ucode_size;
u32 *code;
printf("Fman%u: Uploading microcode version %u.%u.%u\n",
fm_idx + 1, ucode->major, ucode->minor,
ucode->revision);
code = (void *)firmware + ucode->code_offset;
ucode_size = sizeof(u32) * ucode->count;
fm_upload_ucode(fm_idx, fm_imem, code, ucode_size);
}
}
return 0;
}
static u32 fm_assign_risc(int port_id)
{
u32 risc_sel, val;
risc_sel = (port_id & 0x1) ? FMFPPRC_RISC2 : FMFPPRC_RISC1;
val = (port_id << FMFPPRC_PORTID_SHIFT) & FMFPPRC_PORTID_MASK;
val |= ((risc_sel << FMFPPRC_ORA_SHIFT) | risc_sel);
return val;
}
static void fm_init_fpm(struct fm_fpm *fpm)
{
int i, port_id;
u32 val;
setbits_be32(&fpm->fmfpee, FMFPEE_EHM | FMFPEE_UEC |
FMFPEE_CER | FMFPEE_DER);
/* IM mode, each even port ID to RISC#1, each odd port ID to RISC#2 */
/* offline/parser port */
for (i = 0; i < MAX_NUM_OH_PORT; i++) {
port_id = OH_PORT_ID_BASE + i;
val = fm_assign_risc(port_id);
out_be32(&fpm->fpmprc, val);
}
/* Rx 1G port */
for (i = 0; i < MAX_NUM_RX_PORT_1G; i++) {
port_id = RX_PORT_1G_BASE + i;
val = fm_assign_risc(port_id);
out_be32(&fpm->fpmprc, val);
}
/* Tx 1G port */
for (i = 0; i < MAX_NUM_TX_PORT_1G; i++) {
port_id = TX_PORT_1G_BASE + i;
val = fm_assign_risc(port_id);
out_be32(&fpm->fpmprc, val);
}
/* Rx 10G port */
port_id = RX_PORT_10G_BASE;
val = fm_assign_risc(port_id);
out_be32(&fpm->fpmprc, val);
/* Tx 10G port */
port_id = TX_PORT_10G_BASE;
val = fm_assign_risc(port_id);
out_be32(&fpm->fpmprc, val);
/* disable the dispatch limit in IM case */
out_be32(&fpm->fpmflc, FMFP_FLC_DISP_LIM_NONE);
/* clear events */
out_be32(&fpm->fmfpee, FMFPEE_CLEAR_EVENT);
/* clear risc events */
for (i = 0; i < 4; i++)
out_be32(&fpm->fpmcev[i], 0xffffffff);
/* clear error */
out_be32(&fpm->fpmrcr, FMFP_RCR_MDEC | FMFP_RCR_IDEC);
}
static int fm_init_bmi(int fm_idx, struct fm_bmi_common *bmi)
{
int blk, i, port_id;
u32 val, offset, base;
/* alloc free buffer pool in MURAM */
base = fm_muram_alloc(fm_idx, FM_FREE_POOL_SIZE, FM_FREE_POOL_ALIGN);
if (!base) {
printf("%s: no muram for free buffer pool\n", __func__);
return -ENOMEM;
}
offset = base - fm_muram_base(fm_idx);
/* Need 128KB total free buffer pool size */
val = offset / 256;
blk = FM_FREE_POOL_SIZE / 256;
/* in IM, we must not begin from offset 0 in MURAM */
val |= ((blk - 1) << FMBM_CFG1_FBPS_SHIFT);
out_be32(&bmi->fmbm_cfg1, val);
/* disable all BMI interrupt */
out_be32(&bmi->fmbm_ier, FMBM_IER_DISABLE_ALL);
/* clear all events */
out_be32(&bmi->fmbm_ievr, FMBM_IEVR_CLEAR_ALL);
/*
* set port parameters - FMBM_PP_x
* max tasks 10G Rx/Tx=12, 1G Rx/Tx 4, others is 1
* max dma 10G Rx/Tx=3, others is 1
* set port FIFO size - FMBM_PFS_x
* 4KB for all Rx and Tx ports
*/
/* offline/parser port */
for (i = 0; i < MAX_NUM_OH_PORT; i++) {
port_id = OH_PORT_ID_BASE + i - 1;
/* max tasks=1, max dma=1, no extra */
out_be32(&bmi->fmbm_pp[port_id], 0);
/* port FIFO size - 256 bytes, no extra */
out_be32(&bmi->fmbm_pfs[port_id], 0);
}
/* Rx 1G port */
for (i = 0; i < MAX_NUM_RX_PORT_1G; i++) {
port_id = RX_PORT_1G_BASE + i - 1;
/* max tasks=4, max dma=1, no extra */
out_be32(&bmi->fmbm_pp[port_id], FMBM_PP_MXT(4));
/* FIFO size - 4KB, no extra */
out_be32(&bmi->fmbm_pfs[port_id], FMBM_PFS_IFSZ(0xf));
}
/* Tx 1G port FIFO size - 4KB, no extra */
for (i = 0; i < MAX_NUM_TX_PORT_1G; i++) {
port_id = TX_PORT_1G_BASE + i - 1;
/* max tasks=4, max dma=1, no extra */
out_be32(&bmi->fmbm_pp[port_id], FMBM_PP_MXT(4));
/* FIFO size - 4KB, no extra */
out_be32(&bmi->fmbm_pfs[port_id], FMBM_PFS_IFSZ(0xf));
}
/* Rx 10G port */
port_id = RX_PORT_10G_BASE - 1;
/* max tasks=12, max dma=3, no extra */
out_be32(&bmi->fmbm_pp[port_id], FMBM_PP_MXT(12) | FMBM_PP_MXD(3));
/* FIFO size - 4KB, no extra */
out_be32(&bmi->fmbm_pfs[port_id], FMBM_PFS_IFSZ(0xf));
/* Tx 10G port */
port_id = TX_PORT_10G_BASE - 1;
/* max tasks=12, max dma=3, no extra */
out_be32(&bmi->fmbm_pp[port_id], FMBM_PP_MXT(12) | FMBM_PP_MXD(3));
/* FIFO size - 4KB, no extra */
out_be32(&bmi->fmbm_pfs[port_id], FMBM_PFS_IFSZ(0xf));
/* initialize internal buffers data base (linked list) */
out_be32(&bmi->fmbm_init, FMBM_INIT_START);
return 0;
}
static void fm_init_qmi(struct fm_qmi_common *qmi)
{
/* disable enqueue and dequeue of QMI */
clrbits_be32(&qmi->fmqm_gc, FMQM_GC_ENQ_EN | FMQM_GC_DEQ_EN);
/* disable all error interrupts */
out_be32(&qmi->fmqm_eien, FMQM_EIEN_DISABLE_ALL);
/* clear all error events */
out_be32(&qmi->fmqm_eie, FMQM_EIE_CLEAR_ALL);
/* disable all interrupts */
out_be32(&qmi->fmqm_ien, FMQM_IEN_DISABLE_ALL);
/* clear all interrupts */
out_be32(&qmi->fmqm_ie, FMQM_IE_CLEAR_ALL);
}
/* Init common part of FM, index is fm num# like fm as above */
int fm_init_common(int index, struct ccsr_fman *reg)
{
int rc;
#if defined(CONFIG_SYS_QE_FMAN_FW_IN_NOR)
void *addr = (void *)CONFIG_SYS_FMAN_FW_ADDR;
#elif defined(CONFIG_SYS_QE_FMAN_FW_IN_NAND)
size_t fw_length = CONFIG_SYS_QE_FMAN_FW_LENGTH;
void *addr = malloc(CONFIG_SYS_QE_FMAN_FW_LENGTH);
powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
rc = nand_read(&nand_info[0], (loff_t)CONFIG_SYS_FMAN_FW_ADDR,
powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
&fw_length, (u_char *)addr);
if (rc == -EUCLEAN) {
printf("NAND read of FMAN firmware at offset 0x%x failed %d\n",
CONFIG_SYS_FMAN_FW_ADDR, rc);
powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
}
#elif defined(CONFIG_SYS_QE_FW_IN_SPIFLASH)
struct spi_flash *ucode_flash;
void *addr = malloc(CONFIG_SYS_QE_FMAN_FW_LENGTH);
powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
int ret = 0;
ucode_flash = spi_flash_probe(CONFIG_ENV_SPI_BUS, CONFIG_ENV_SPI_CS,
CONFIG_ENV_SPI_MAX_HZ, CONFIG_ENV_SPI_MODE);
if (!ucode_flash)
printf("SF: probe for ucode failed\n");
else {
ret = spi_flash_read(ucode_flash, CONFIG_SYS_FMAN_FW_ADDR,
CONFIG_SYS_QE_FMAN_FW_LENGTH, addr);
powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
if (ret)
printf("SF: read for ucode failed\n");
spi_flash_free(ucode_flash);
}
#elif defined(CONFIG_SYS_QE_FMAN_FW_IN_MMC)
powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
int dev = CONFIG_SYS_MMC_ENV_DEV;
void *addr = malloc(CONFIG_SYS_QE_FMAN_FW_LENGTH);
u32 cnt = CONFIG_SYS_QE_FMAN_FW_LENGTH / 512;
u32 blk = CONFIG_SYS_FMAN_FW_ADDR / 512;
powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
struct mmc *mmc = find_mmc_device(CONFIG_SYS_MMC_ENV_DEV);
if (!mmc)
printf("\nMMC cannot find device for ucode\n");
else {
printf("\nMMC read: dev # %u, block # %u, count %u ...\n",
dev, blk, cnt);
mmc_init(mmc);
(void)mmc->block_dev.block_read(dev, blk, cnt, addr);
powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
/* flush cache after read */
flush_cache((ulong)addr, cnt * 512);
}
powerpc/corenet_ds: Slave module for boot from SRIO For the powerpc processors with SRIO interface, boot location can be configured from SRIO1 or SRIO2 by RCW. The processor booting from SRIO can do without flash for u-boot image. The image can be fetched from another processor's memory space by SRIO link connected between them. The processor boots from SRIO is slave, the processor boots from normal flash memory space and can help slave to boot from its memory space is master. They are different environments and requirements: master: 1. NOR flash for its own u-boot image, ucode and ENV space. 2. Slave's u-boot image in master NOR flash. 3. Normally boot from local NOR flash. 4. Configure SRIO switch system if needed. slave: 1. Just has EEPROM for RCW. No flash for u-boot image, ucode and ENV. 2. Boot location should be set to SRIO1 or SRIO2 by RCW. 3. RCW should configure the SerDes, SRIO interfaces correctly. 4. Slave must be powered on after master's boot. 5. Must define CONFIG_SYS_QE_FMAN_FW_IN_REMOTE because of no ucode locally. For the slave module, need to finish these processes: 1. Set the boot location to SRIO1 or SRIO2 by RCW. 2. Set a specific TLB entry for the boot process. 3. Set a LAW entry with the TargetID SRIO1 or SRIO2 for the boot. 4. Slave's u-boot image should be generated specifically by make xxxx_SRIOBOOT_SLAVE_config. This will set SYS_TEXT_BASE=0xFFF80000 and other configurations. Signed-off-by: Liu Gang <Gang.Liu@freescale.com> Signed-off-by: Shaohui Xie <Shaohui.Xie@freescale.com>
2012-03-08 00:33:18 +00:00
#elif defined(CONFIG_SYS_QE_FMAN_FW_IN_REMOTE)
void *addr = (void *)CONFIG_SYS_FMAN_FW_ADDR;
#else
void *addr = NULL;
powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
#endif
/* Upload the Fman microcode if it's present */
rc = fman_upload_firmware(index, &reg->fm_imem, addr);
if (rc)
return rc;
setenv_addr("fman_ucode", addr);
powerpc/85xx: Add support for FMan ethernet in Independent mode The Frame Manager (FMan) on QorIQ SoCs with DPAA (datapath acceleration architecture) is the ethernet contoller block. Normally it is utilized via Queue Manager (Qman) and Buffer Manager (Bman). However for boot usage the FMan supports a mode similar to QE or CPM ethernet collers called Independent mode. Additionally the FMan block supports multiple 1g and 10g interfaces as a single entity in the system rather than each controller being managed uniquely. This means we have to initialize all of Fman regardless of the number of interfaces we utilize. Different SoCs support different combinations of the number of FMan as well as the number of 1g & 10g interfaces support per Fman. We add support for the following SoCs: * P1023 - 1 Fman, 2x1g * P4080 - 2 Fman, each Fman has 4x1g and 1x10g * P204x/P3041/P5020 - 1 Fman, 5x1g, 1x10g Signed-off-by: Dave Liu <daveliu@freescale.com> Signed-off-by: Andy Fleming <afleming@freescale.com> Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Roy Zang <tie-fei.zang@freescale.com> Signed-off-by: Dai Haruki <dai.haruki@freescale.com> Signed-off-by: Kim Phillips <kim.phillips@freescale.com> Signed-off-by: Ioana Radulescu <ruxandra.radulescu@freescale.com> Signed-off-by: Lei Xu <B33228@freescale.com> Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com> Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Shaohui Xie <b21989@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
2011-04-13 13:37:44 +00:00
fm_init_muram(index, &reg->muram);
fm_init_qmi(&reg->fm_qmi_common);
fm_init_fpm(&reg->fm_fpm);
/* clear DMA status */
setbits_be32(&reg->fm_dma.fmdmsr, FMDMSR_CLEAR_ALL);
/* set DMA mode */
setbits_be32(&reg->fm_dma.fmdmmr, FMDMMR_SBER);
return fm_init_bmi(index, &reg->fm_bmi_common);
}