/****************************************************************************
* arch/arm/src/tiva/tm4c_ethernet.c
*
* Copyright (C) 2014-2015 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in
* the documentation and/or other materials provided with the
* distribution.
* 3. Neither the name NuttX nor the names of its contributors may be
* used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
* FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
* COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
* BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS
* OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
* AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
* ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
* POSSIBILITY OF SUCH DAMAGE.
*
****************************************************************************/
/****************************************************************************
* Included Files
****************************************************************************/
#include <nuttx/config.h>
#if defined(CONFIG_NET) && defined(CONFIG_TIVA_ETHERNET)
#include <stdint.h>
#include <stdbool.h>
#include <time.h>
#include <string.h>
#include <debug.h>
#include <queue.h>
#include <errno.h>
#include <arpa/inet.h>
#include <nuttx/arch.h>
#include <nuttx/irq.h>
#include <nuttx/wdog.h>
#ifdef CONFIG_NET_NOINTS
# include <nuttx/wqueue.h>
#endif
#include <nuttx/net/mii.h>
#include <nuttx/net/arp.h>
#include <nuttx/net/netdev.h>
#ifdef CONFIG_TIVA_PHY_INTERRUPTS
# include <nuttx/net/phy.h>
#endif
#ifdef CONFIG_NET_PKT
# include <nuttx/net/pkt.h>
#endif
#include "up_internal.h"
#include "chip.h"
#include "tiva_gpio.h"
#include "tiva_syscontrol.h"
#include "tiva_enablepwr.h"
#include "tiva_enableclks.h"
#include "tiva_periphrdy.h"
#include "tiva_ethernet.h"
#include "chip/tiva_pinmap.h"
#include <arch/board/board.h>
/* TIVA_NETHCONTROLLERS determines the number of physical interfaces
* that will be supported.
*/
#if TIVA_NETHCONTROLLERS > 0
/****************************************************************************
* Definitions
****************************************************************************/
/* Configuration ************************************************************/
#if TIVA_NETHCONTROLLERS > 1
# error Logic to support multiple Ethernet interfaces is incomplete
#endif
/* If processing is not done at the interrupt level, then high priority
* work queue support is required.
*/
#if defined(CONFIG_NET_NOINTS) && !defined(CONFIG_SCHED_HPWORK)
# error High priority work queue support is required
#endif
/* Are we using the internal PHY or an external PHY? */
#if defined(CONFIG_TIVA_PHY_INTERNAL)
/* Internal PHY */
# if defined(CONFIG_TIVA_PHY_MII) ||defined(CONFIG_TIVA_PHY_RMII)
# warning CONFIG_TIVA_PHY_MII or CONFIG_TIVA_PHY_RMII defined with internal PHY
# endif
# undef CONFIG_TIVA_PHY_MII
# undef CONFIG_TIVA_PHY_RMII
/* Properties of the internal PHY are hard-coded */
# undef CONFIG_TIVA_PHYADDR
# undef CONFIG_TIVA_PHYSR_ALTCONFIG
# undef CONFIG_TIVA_PHYSR_ALTMODE
# undef CONFIG_TIVA_PHYSR_10HD
# undef CONFIG_TIVA_PHYSR_100HD
# undef CONFIG_TIVA_PHYSR_10FD
# undef CONFIG_TIVA_PHYSR_100FD
# undef CONFIG_TIVA_PHYSR_SPEED
# undef CONFIG_TIVA_PHYSR_100MBPS
# undef CONFIG_TIVA_PHYSR_MODE
# undef CONFIG_TIVA_PHYSR_FULLDUPLEX
# define CONFIG_TIVA_PHYADDR 0
# define CONFIG_TIVA_PHYSR TIVA_EPHY_STS
# define CONFIG_TIVA_PHYSR_SPEED EPHY_STS_SPEED
# define CONFIG_TIVA_PHYSR_100MBPS 0
# define CONFIG_TIVA_PHYSR_MODE EPHY_STS_DUPLEX
# define CONFIG_TIVA_PHYSR_FULLDUPLEX EPHY_STS_DUPLEX
#else
/* External PHY. Properties must be provided in the configuration */
# if !defined(CONFIG_TIVA_PHY_MII) && !defined(CONFIG_TIVA_PHY_RMII)
# warning None of CONFIG_TIVA_PHY_INTERNAL, CONFIG_TIVA_PHY_MII, or CONFIG_TIVA_PHY_RMII defined
# endif
# if defined(CONFIG_TIVA_PHY_MII) && defined(CONFIG_TIVA_PHY_RMII)
# error Both CONFIG_TIVA_PHY_MII and CONFIG_TIVA_PHY_RMII defined
# endif
#endif
#ifndef CONFIG_TIVA_PHYADDR
# error CONFIG_TIVA_PHYADDR must be defined in the NuttX configuration
#endif
#ifdef CONFIG_TIVA_AUTONEG
# ifndef CONFIG_TIVA_PHYSR
# error CONFIG_TIVA_PHYSR must be defined in the NuttX configuration
# endif
# ifdef CONFIG_TIVA_PHYSR_ALTCONFIG
# ifndef CONFIG_TIVA_PHYSR_ALTMODE
# error CONFIG_TIVA_PHYSR_ALTMODE must be defined in the NuttX configuration
# endif
# ifndef CONFIG_TIVA_PHYSR_10HD
# error CONFIG_TIVA_PHYSR_10HD must be defined in the NuttX configuration
# endif
# ifndef CONFIG_TIVA_PHYSR_100HD
# error CONFIG_TIVA_PHYSR_100HD must be defined in the NuttX configuration
# endif
# ifndef CONFIG_TIVA_PHYSR_10FD
# error CONFIG_TIVA_PHYSR_10FD must be defined in the NuttX configuration
# endif
# ifndef CONFIG_TIVA_PHYSR_100FD
# error CONFIG_TIVA_PHYSR_100FD must be defined in the NuttX configuration
# endif
# else
# ifndef CONFIG_TIVA_PHYSR_SPEED
# error CONFIG_TIVA_PHYSR_SPEED must be defined in the NuttX configuration
# endif
# ifndef CONFIG_TIVA_PHYSR_100MBPS
# error CONFIG_TIVA_PHYSR_100MBPS must be defined in the NuttX configuration
# endif
# ifndef CONFIG_TIVA_PHYSR_MODE
# error CONFIG_TIVA_PHYSR_MODE must be defined in the NuttX configuration
# endif
# ifndef CONFIG_TIVA_PHYSR_FULLDUPLEX
# error CONFIG_TIVA_PHYSR_FULLDUPLEX must be defined in the NuttX configuration
# endif
# endif
#endif
#ifdef CONFIG_TIVA_EMAC_PTP
# warning CONFIG_TIVA_EMAC_PTP is not yet supported
#endif
/* This driver does not use enhanced descriptors. Enhanced descriptors must
* be used, however, if time stamping or and/or IPv4 checksum offload is
* supported.
*/
#undef CONFIG_TIVA_EMAC_ENHANCEDDESC
#undef CONFIG_TIVA_EMAC_HWCHECKSUM
/* Ethernet buffer sizes, number of buffers, and number of descriptors */
#ifndef CONFIG_NET_MULTIBUFFER
# error CONFIG_NET_MULTIBUFFER is required
#endif
#ifndef CONFIG_TIVA_EMAC_NRXDESC
# define CONFIG_TIVA_EMAC_NRXDESC 8
#endif
#ifndef CONFIG_TIVA_EMAC_NTXDESC
# define CONFIG_TIVA_EMAC_NTXDESC 4
#endif
/* Add 4 to the configured buffer size to account for the 2 byte checksum
* memory needed at the end of the maximum size packet. Buffer sizes must
* be an even multiple of 4, 8, or 16 bytes (depending on buswidth). We
* will use the 16-byte alignment in all cases.
*/
#define OPTIMAL_EMAC_BUFSIZE ((CONFIG_NET_ETH_MTU + 4 + 15) & ~15)
#if OPTIMAL_EMAC_BUFSIZE > EMAC_TDES1_TBS1_MASK
# error OPTIMAL_EMAC_BUFSIZE is too large
#endif
#if (OPTIMAL_EMAC_BUFSIZE & 15) != 0
# error OPTIMAL_EMAC_BUFSIZE must be aligned
#endif
#if OPTIMAL_EMAC_BUFSIZE != OPTIMAL_EMAC_BUFSIZE
# warning You using an incomplete/untested configuration
#endif
/* We need at least one more free buffer than transmit buffers */
#define TIVA_EMAC_NFREEBUFFERS (CONFIG_TIVA_EMAC_NTXDESC+1)
/* Extremely detailed register debug that you would normally never want
* enabled.
*/
#ifndef CONFIG_DEBUG
# undef CONFIG_TIVA_ETHERNET_REGDEBUG
#endif
/* Clocking *****************************************************************/
/* Set MIIADDR CR bits depending on SysClk frequency */
#if SYSCLK_FREQUENCY >= 20000000 && SYSCLK_FREQUENCY < 35000000
# define EMAC_MIIADDR_CR EMAC_MIIADDR_CR_20_35
#elif SYSCLK_FREQUENCY >= 35000000 && SYSCLK_FREQUENCY <= 64000000
# define EMAC_MIIADDR_CR EMAC_MIIADDR_CR_35_60
#elif SYSCLK_FREQUENCY >= 60000000 && SYSCLK_FREQUENCY <= 104000000
# define EMAC_MIIADDR_CR EMAC_MIIADDR_CR_60_100
#elif SYSCLK_FREQUENCY >= 100000000 && SYSCLK_FREQUENCY <= 150000000
# define EMAC_MIIADDR_CR EMAC_MIIADDR_CR_100_150
#elif SYSCLK_FREQUENCY >= 150000000 && SYSCLK_FREQUENCY <= 168000000
# define EMAC_MIIADDR_CR EMAC_MIIADDR_CR_150_168
#else
# error SYSCLK_FREQUENCY not supportable
#endif
/* Timing *******************************************************************/
/* TX poll delay = 1 seconds. CLK_TCK is the number of clock ticks per
* second
*/
#define TIVA_WDDELAY (1*CLK_TCK)
#define TIVA_POLLHSEC (1*2)
/* TX timeout = 1 minute */
#define TIVA_TXTIMEOUT (60*CLK_TCK)
/* PHY reset/configuration delays in milliseconds */
#define PHY_RESET_DELAY (65)
#define PHY_CONFIG_DELAY (1000)
/* PHY read/write delays in loop counts */
#define PHY_READ_TIMEOUT (0x0004ffff)
#define PHY_WRITE_TIMEOUT (0x0004ffff)
#define PHY_RETRY_TIMEOUT (0x0004ffff)
/* Register values **********************************************************/
/* Clear the MACCR bits that will be setup during MAC initialization (or that
* are cleared unconditionally). Per the reference manual, all reserved bits
* must be retained at their reset value.
*
* EMAC_CFG_RE Bit 2: Receiver enable
* EMAC_CFG_TE Bit 3: Transmitter enable
* EMAC_CFG_DC Bit 4: Deferral check
* EMAC_CFG_BL Bits 5-6: Back-off limit
* EMAC_CFG_ACS Bit 7: Automatic pad/CRC stripping
* EMAC_CFG_DR Bit 9: Retry disable
* EMAC_CFG_IPC Bit 10: IPv4 checksum offload
* EMAC_CFG_DUPM Bit 11: Duplex mode
* EMAC_CFG_LOOPBM Bit 12: Loopback mode
* EMAC_CFG_DRO Bit 13: Receive own disable
* EMAC_CFG_FES Bit 14: Fast Ethernet speed
* EMAC_CFG_PS Bit 15: Port Select
* EMAC_CFG_DISCRS Bit 16: Carrier sense disable
* EMAC_CFG_IFG Bits 17-19: Interframe gap
* EMAC_CFG_JFEN Bit 20: Jumbo Frame Enable
* EMAC_CFG_JD Bit 22: Jabber disable
* EMAC_CFG_WDDIS Bit 23: Watchdog disable
* EMAC_CFG_CST Bit 25: CRC stripping for Type frames
* EMAC_CFG_TWOKPEN Bit 27: IEEE 802
* EMAC_CFG_SADDR Bits 28-30: Source Address Insertion or Replacement Control
*/
#define MACCR_CLEAR_BITS \
(EMAC_CFG_RE | EMAC_CFG_TE | EMAC_CFG_DC | EMAC_CFG_BL_MASK | \
EMAC_CFG_ACS | EMAC_CFG_DR | EMAC_CFG_IPC | EMAC_CFG_DUPM | \
EMAC_CFG_LOOPBM | EMAC_CFG_DRO | EMAC_CFG_FES | EMAC_CFG_DISCRS | \
EMAC_CFG_IFG_MASK | EMAC_CFG_JD | EMAC_CFG_WDDIS | EMAC_CFG_CST)
/* The following bits are set or left zero unconditionally in all modes.
*
* EMAC_CFG_RE Receiver enable 0 (disabled)
* EMAC_CFG_TE Transmitter enable 0 (disabled)
* EMAC_CFG_DC Deferral check 0 (disabled)
* EMAC_CFG_BL Back-off limit 0 (10)
* EMAC_CFG_ACS Automatic pad/CRC stripping 0 (disabled)
* EMAC_CFG_DR Retry disable 1 (disabled)
* EMAC_CFG_IPC IPv4 checksum offload Depends on CONFIG_TIVA_EMAC_HWCHECKSUM
* EMAC_CFG_LOOPBM Loopback mode 0 (disabled)
* EMAC_CFG_DRO Receive own disable 0 (enabled)
* EMAC_CFG_PS Port Select (read-only)
* EMAC_CFG_DISCRS Carrier sense disable 0 (enabled)
* EMAC_CFG_IFG Interframe gap 0 (96 bits)
* EMAC_CFG_JFEN Jumbo Frame Enable 0 (jumbo frame creates error)
* EMAC_CFG_JD Jabber disable 0 (enabled)
* EMAC_CFG_WDDIS Watchdog disable 0 (enabled)
* EMAC_CFG_CST CRC stripping for Type frames 0 (disabled, F2/F4 only)
* EMAC_CFG_TWOKPEN IEEE 802 0 (>1518 == giant frame)
* EMAC_CFG_SADDR Source Address Insertion or
* Replacement Control
*
* The following are set conditionally based on mode and speed.
*
* EMAC_CFG_DUPM Duplex mode Depends on priv->fduplex
* EMAC_CFG_FES Fast Ethernet speed Depends on priv->mbps100
*/
#ifdef CONFIG_TIVA_EMAC_HWCHECKSUM
# define MACCR_SET_BITS \
(EMAC_CFG_BL_10 | EMAC_CFG_DR | EMAC_CFG_IPC | EMAC_CFG_IFG_96)
#else
# define MACCR_SET_BITS \
(EMAC_CFG_BL_10 | EMAC_CFG_DR | EMAC_CFG_IFG_96)
#endif
/* Clear the MACCR bits that will be setup during MAC initialization (or that
* are cleared unconditionally). Per the reference manual, all reserved bits
* must be retained at their reset value.
*
* EMAC_FRAMEFLTR_PR Bit 0: Promiscuous mode
* EMAC_FRAMEFLTR_HUC Bit 1: Hash unicast
* EMAC_FRAMEFLTR_HMC Bit 2: Hash multicast
* EMAC_FRAMEFLTR_DAIF Bit 3: Destination address inverse filtering
* EMAC_FRAMEFLTR_PM Bit 4: Pass all multicast
* EMAC_FRAMEFLTR_DBF Bit 5: Broadcast frames disable
* EMAC_FRAMEFLTR_PCF Bits 6-7: Pass control frames
* EMAC_FRAMEFLTR_SAIF Bit 8: Source address inverse filtering
* EMAC_FRAMEFLTR_SAF Bit 9: Source address filter
* EMAC_FRAMEFLTR_HPF Bit 10: Hash or perfect filter
* EMAC_FRAMEFLTR_VTFE Bit 16: VLAN Tag Filter Enable
* EMAC_FRAMEFLTR_RA Bit 31: Receive all
*/
#define FRAMEFLTR_CLEAR_BITS \
(EMAC_FRAMEFLTR_PR | EMAC_FRAMEFLTR_HUC | EMAC_FRAMEFLTR_HMC | EMAC_FRAMEFLTR_DAIF | \
EMAC_FRAMEFLTR_PM | EMAC_FRAMEFLTR_DBF | EMAC_FRAMEFLTR_PCF_MASK | EMAC_FRAMEFLTR_SAIF | \
EMAC_FRAMEFLTR_SAF | EMAC_FRAMEFLTR_HPF | EMAC_FRAMEFLTR_RA)
/* The following bits are set or left zero unconditionally in all modes.
*
* EMAC_FRAMEFLTR_PR Promiscuous mode 0 (disabled)
* EMAC_FRAMEFLTR_HUC Hash unicast 0 (perfect dest filtering)
* EMAC_FRAMEFLTR_HMC Hash multicast 0 (perfect dest filtering)
* EMAC_FRAMEFLTR_DAIF Destination address inverse filtering 0 (normal)
* EMAC_FRAMEFLTR_PM Pass all multicast 0 (Depends on HM bit)
* EMAC_FRAMEFLTR_DBF Broadcast frames disable 0 (enabled)
* EMAC_FRAMEFLTR_PCF Pass control frames 1 (block all but PAUSE)
* EMAC_FRAMEFLTR_SAIF Source address inverse filtering 0 (not used)
* EMAC_FRAMEFLTR_SAF Source address filter 0 (disabled)
* EMAC_FRAMEFLTR_HPF Hash or perfect filter 0 (Only matching frames passed)
* EMAC_FRAMEFLTR_VTFE VLAN Tag Filter Enable 0 (VLAN tag ignored)
* EMAC_FRAMEFLTR_RA Receive all 0 (disabled)
*/
#define FRAMEFLTR_SET_BITS (EMAC_FRAMEFLTR_PCF_PAUSE)
/* Clear the FLOWCTL bits that will be setup during MAC initialization (or that
* are cleared unconditionally). Per the reference manual, all reserved bits
* must be retained at their reset value.
*
* EMAC_FLOWCTL_FCBBPA Bit 0: Flow control busy/back pressure activate
* EMAC_FLOWCTL_TFE Bit 1: Transmit flow control enable
* EMAC_FLOWCTL_RFE Bit 2: Receive flow control enable
* EMAC_FLOWCTL_UP Bit 3: Unicast pause frame detect
* EMAC_FLOWCTL_PLT Bits 4-5: Pause low threshold
* EMAC_FLOWCTL_DZQP Bit 7: Zero-quanta pause disable
* EMAC_FLOWCTL_PT Bits 16-31: Pause time
*/
#define FLOWCTL_CLEAR_MASK \
(EMAC_FLOWCTL_FCBBPA | EMAC_FLOWCTL_TFE | EMAC_FLOWCTL_RFE | EMAC_FLOWCTL_UP | \
EMAC_FLOWCTL_PLT_MASK | EMAC_FLOWCTL_DZQP | EMAC_FLOWCTL_PT_MASK)
/* The following bits are set or left zero unconditionally in all modes.
*
* EMAC_FLOWCTL_FCBBPA Flow control busy/back pressure activate 0 (no pause control frame)
* EMAC_FLOWCTL_TFE Transmit flow control enable 0 (disabled)
* EMAC_FLOWCTL_RFE Receive flow control enable 0 (disabled)
* EMAC_FLOWCTL_UP Unicast pause frame detect 0 (disabled)
* EMAC_FLOWCTL_PLT Pause low threshold 0 (pause time - 4)
* EMAC_FLOWCTL_DZQP Zero-quanta pause disable 1 (disabled)
* EMAC_FLOWCTL_PT Pause time 0
*/
#define FLOWCTL_SET_MASK (EMAC_FLOWCTL_PLT_M4 | EMAC_FLOWCTL_DZQP)
/* Clear the DMAOPMODE bits that will be setup during MAC initialization (or that
* are cleared unconditionally). Per the reference manual, all reserved bits
* must be retained at their reset value.
*
* EMAC_DMAOPMODE_SR Bit 1: Start/stop receive
* EMAC_DMAOPMODE_OSF Bit 2: Operate on second frame
* EMAC_DMAOPMODE_RTC Bits 3-4: Receive threshold control
* EMAC_DMAOPMODE_DGF Bit 5: Drop giant frames enable
* EMAC_DMAOPMODE_FUF Bit 6: Forward undersized good frames
* EMAC_DMAOPMODE_FEF Bit 7: Forward error frames
* EMAC_DMAOPMODE_ST Bit 13: Start/stop transmission
* EMAC_DMAOPMODE_TTC Bits 14-16: Transmit threshold control
* EMAC_DMAOPMODE_FTF Bit 20: Flush transmit FIFO
* EMAC_DMAOPMODE_TSF Bit 21: Transmit store and forward
* EMAC_DMAOPMODE_DFF Bit 24: Disable flushing of received frames
* EMAC_DMAOPMODE_RSF Bit 25: Receive store and forward
* EMAC_DMAOPMODE_DT Bit 26: Dropping of TCP/IP checksum error frames disable
*/
#define DMAOPMODE_CLEAR_MASK \
(EMAC_DMAOPMODE_SR | EMAC_DMAOPMODE_OSF | EMAC_DMAOPMODE_RTC_MASK | EMAC_DMAOPMODE_DGF | \
EMAC_DMAOPMODE_FUF | EMAC_DMAOPMODE_FEF | EMAC_DMAOPMODE_ST | EMAC_DMAOPMODE_TTC_MASK | \
EMAC_DMAOPMODE_FTF | EMAC_DMAOPMODE_TSF | EMAC_DMAOPMODE_DFF | EMAC_DMAOPMODE_RSF | \
EMAC_DMAOPMODE_DT)
/* The following bits are set or left zero unconditionally in all modes.
*
* EMAC_DMAOPMODE_SR Start/stop receive 0 (not running)
* EMAC_DMAOPMODE_OSF Operate on second frame 1 (enabled)
* EMAC_DMAOPMODE_RTC Receive threshold control 0 (64 bytes)
* EMAC_DMAOPMODE_FUF Forward undersized good frames 0 (disabled)
* EMAC_DMAOPMODE_FEF Forward error frames 0 (disabled)
* EMAC_DMAOPMODE_ST Start/stop transmission 0 (not running)
* EMAC_DMAOPMODE_TTC Transmit threshold control 0 (64 bytes)
* EMAC_DMAOPMODE_FTF Flush transmit FIFO 0 (no flush)
* EMAC_DMAOPMODE_TSF Transmit store and forward Depends on CONFIG_TIVA_EMAC_HWCHECKSUM
* EMAC_DMAOPMODE_DFF Disable flushing of received frames 0 (enabled)
* EMAC_DMAOPMODE_RSF Receive store and forward Depends on CONFIG_TIVA_EMAC_HWCHECKSUM
* EMAC_DMAOPMODE_DT Dropping of TCP/IP checksum error Depends on CONFIG_TIVA_EMAC_HWCHECKSUM
* frames disable
*
* When the checksum offload feature is enabled, we need to enable the Store
* and Forward mode: the store and forward guarantee that a whole frame is
* stored in the FIFO, so the MAC can insert/verify the checksum, if the
* checksum is OK the DMA can handle the frame otherwise the frame is dropped
*/
#if CONFIG_TIVA_EMAC_HWCHECKSUM
# define DMAOPMODE_SET_MASK \
(EMAC_DMAOPMODE_OSF | EMAC_DMAOPMODE_RTC_64 | EMAC_DMAOPMODE_TTC_64 | \
EMAC_DMAOPMODE_TSF | EMAC_DMAOPMODE_RSF)
#else
# define DMAOPMODE_SET_MASK \
(EMAC_DMAOPMODE_OSF | EMAC_DMAOPMODE_RTC_64 | EMAC_DMAOPMODE_TTC_64 | \
EMAC_DMAOPMODE_DT)
#endif
/* Clear the DMABUSMOD bits that will be setup during MAC initialization (or that
* are cleared unconditionally). Per the reference manual, all reserved bits
* must be retained at their reset value.
*
* EMAC_DMABUSMOD_SWR Bit 0: Software reset
* EMAC_DMABUSMOD_DA Bit 1: DMA Arbitration
* EMAC_DMABUSMOD_DSL Bits 2-6: Descriptor skip length
* EMAC_DMABUSMOD_ATDS Bit 7: Enhanced descriptor format enable
* EMAC_DMABUSMOD_PBL Bits 8-13: Programmable burst length
* EMAC_DMABUSMOD_PR Bits 14-15: RX TX priority ratio
* EMAC_DMABUSMOD_FB Bit 16: Fixed burst
* EMAC_DMABUSMOD_RPBL Bits 17-22: RX DMA programmable bust length
* EMAC_DMABUSMOD_USP Bit 23: Use separate PBL
* EMAC_DMABUSMOD_8XPBL Bit 24: 8x programmable burst length mode
* EMAC_DMABUSMOD_AAL Bit 25: Address-aligned beats
* EMAC_DMABUSMOD_MB Bit 26: Mixed burst (F2/F4 only)
* EMAC_DMABUSMOD_TXPR Bit 27: Transmit Priority
* EMAC_DMABUSMOD_RIB Bit 31: Rebuild Burst
*/
#define DMABUSMOD_CLEAR_MASK \
(EMAC_DMABUSMOD_SWR | EMAC_DMABUSMOD_DA | EMAC_DMABUSMOD_DSL_MASK | \
EMAC_DMABUSMOD_ATDS | EMAC_DMABUSMOD_PBL_MASK | EMAC_DMABUSMOD_PR_MASK | \
EMAC_DMABUSMOD_FB | EMAC_DMABUSMOD_RPBL_MASK | EMAC_DMABUSMOD_USP | \
EMAC_DMABUSMOD_8XPBL | EMAC_DMABUSMOD_AAL | EMAC_DMABUSMOD_MB |\
EMAC_DMABUSMOD_TXPR | EMAC_DMABUSMOD_RIB)
/* The following bits are set or left zero unconditionally in all modes.
*
* EMAC_DMABUSMOD_SWR Software reset 0 (no reset)
* EMAC_DMABUSMOD_DA DMA Arbitration 1 (fixed priority)
* EMAC_DMABUSMOD_DSL Descriptor skip length 0
* EMAC_DMABUSMOD_ATDS Enhanced descriptor format enable Depends on CONFIG_TIVA_EMAC_ENHANCEDDESC
* EMAC_DMABUSMOD_PBL Programmable burst length Depends on EMAC_DMA_RXBURST
* EMAC_DMABUSMOD_PR RX TX priority ratio 0 1:1
* EMAC_DMABUSMOD_FB Fixed burst 0 (disabled)
* EMAC_DMABUSMOD_RPBL RX DMA programmable burst length Depends on EMAC_DMA_TXBURST
* EMAC_DMABUSMOD_USP Use separate PBL Depends on EMAC_DMA_RX/TXBURST
* EMAC_DMABUSMOD_8XPBL 8x programmable burst length mode Depends on EMAC_DMA_RX/TXBURST
* EMAC_DMABUSMOD_AAL Address-aligned beats 0 (disabled)
* EMAC_DMABUSMOD_MB Mixed burst 1 (enabled)
* EMAC_DMABUSMOD_TXPR Transmit Priority 0 (RX DMA has priority over TX)
* EMAC_DMABUSMOD_RIB Rebuild Burst 0
*/
#define EMAC_DMA_RXBURST 4
#define EMAC_DMA_TXBURST 4
#if EMAC_DMA_RXBURST > 32 || EMAC_DMA_TXBURST > 32
# define __EMAC_DMABUSMOD_8XPBL 0
# define __EMAC_DMA_RXBURST EMAC_DMA_RXBURST
# define __EMAC_DMA_TXBURST EMAC_DMA_TXBURST
#else
/* Divide both burst lengths by 8 and set the 8X burst length multiplier */
# define __EMAC_DMABUSMOD_8XPBL EMAC_DMABUSMOD_8XPBL
# define __EMAC_DMA_RXBURST (EMAC_DMA_RXBURST >> 3)
# define __EMAC_DMA_TXBURST (EMAC_DMA_TXBURST >> 3)
#endif
#define __EMAC_DMABUSMOD_PBL EMAC_DMABUSMOD_PBL(__EMAC_DMA_RXBURST)
/* Are the receive and transmit burst lengths the same? */
#if __EMAC_DMA_RXBURST == __EMAC_DMA_TXBURST
/* Yes.. Set up to use a single burst length */
# define __EMAC_DMABUSMOD_USP 0
# define __EMAC_DMABUSMOD_RPBL 0
#else
/* No.. Use separate burst lengths for each */
# define __EMAC_DMABUSMOD_USP EMAC_DMABUSMOD_USP
# define __EMAC_DMABUSMOD_RPBL EMAC_DMABUSMOD_RPBL(__EMAC_DMA_TXBURST)
#endif
#ifdef CONFIG_TIVA_EMAC_ENHANCEDDESC
# define __EMAC_DMABUSMOD_ATDS EMAC_DMABUSMOD_ATDS
#else
# define __EMAC_DMABUSMOD_ATDS 0
#endif
#define DMABUSMOD_SET_MASK \
(EMAC_DMABUSMOD_DA | EMAC_DMABUSMOD_DSL(0) | __EMAC_DMABUSMOD_ATDS | \
__EMAC_DMABUSMOD_PBL | __EMAC_DMABUSMOD_RPBL | __EMAC_DMABUSMOD_USP | \
__EMAC_DMABUSMOD_8XPBL | EMAC_DMABUSMOD_MB)
/* Interrupt bit sets *******************************************************/
/* All interrupts in the normal and abnormal interrupt summary. Early transmit
* interrupt (ETI) is excluded from the abnormal set because it causes too
* many interrupts and is not interesting.
*/
#define EMAC_DMAINT_NORMAL \
(EMAC_DMAINT_TI | EMAC_DMAINT_TBUI |EMAC_DMAINT_RI | EMAC_DMAINT_ERI)
#define EMAC_DMAINT_ABNORMAL \
(EMAC_DMAINT_TPSI | EMAC_DMAINT_TJTI | EMAC_DMAINT_OVFI | EMAC_EMAINT_UNFI | \
EMAC_DMAINT_RBUI | EMAC_DMAINT_RPSI | EMAC_DMAINT_RWTI | /* EMAC_DMAINT_ETI | */ \
EMAC_DMAINT_FBEI)
/* Normal receive, transmit, error interrupt enable bit sets */
#define EMAC_DMAINT_RECV_ENABLE (EMAC_DMAINT_NIS | EMAC_DMAINT_RI)
#define EMAC_DMAINT_XMIT_ENABLE (EMAC_DMAINT_NIS | EMAC_DMAINT_TI)
#define EMAC_DMAINT_XMIT_DISABLE (EMAC_DMAINT_TI)
#ifdef CONFIG_DEBUG_NET
# define EMAC_DMAINT_ERROR_ENABLE (EMAC_DMAINT_AIS | EMAC_DMAINT_ABNORMAL)
#else
# define EMAC_DMAINT_ERROR_ENABLE (0)
#endif
/* Helpers ******************************************************************/
/* This is a helper pointer for accessing the contents of the Ethernet
* header
*/
#define BUF ((struct eth_hdr_s *)priv->dev.d_buf)
/****************************************************************************
* Private Types
****************************************************************************/
/* The tiva_ethmac_s encapsulates all state information for a single hardware
* interface
*/
struct tiva_ethmac_s
{
uint8_t ifup : 1; /* true:ifup false:ifdown */
uint8_t mbps100 : 1; /* 100MBps operation (vs 10 MBps) */
uint8_t fduplex : 1; /* Full (vs. half) duplex */
WDOG_ID txpoll; /* TX poll timer */
WDOG_ID txtimeout; /* TX timeout timer */
#ifdef CONFIG_NET_NOINTS
struct work_s work; /* For deferring work to the work queue */
#endif
#ifdef CONFIG_TIVA_PHY_INTERRUPTS
xcpt_t handler; /* Attached PHY interrupt handler */
#endif
/* This holds the information visible to uIP/NuttX */
struct net_driver_s dev; /* Interface understood by network subsystem */
/* Used to track transmit and receive descriptors */
struct emac_txdesc_s *txhead; /* Next available TX descriptor */
struct emac_rxdesc_s *rxhead; /* Next available RX descriptor */
struct emac_txdesc_s *txtail; /* First "in_flight" TX descriptor */
struct emac_rxdesc_s *rxcurr; /* First RX descriptor of the segment */
uint16_t segments; /* RX segment count */
uint16_t inflight; /* Number of TX transfers "in_flight" */
sq_queue_t freeb; /* The free buffer list */
/* Descriptor allocations */
struct emac_rxdesc_s rxtable[CONFIG_TIVA_EMAC_NRXDESC];
struct emac_txdesc_s txtable[CONFIG_TIVA_EMAC_NTXDESC];
/* Buffer allocations */
uint8_t rxbuffer[CONFIG_TIVA_EMAC_NRXDESC*OPTIMAL_EMAC_BUFSIZE];
uint8_t alloc[TIVA_EMAC_NFREEBUFFERS*OPTIMAL_EMAC_BUFSIZE];
};
/****************************************************************************
* Private Data
****************************************************************************/
static struct tiva_ethmac_s g_tiva_ethmac[TIVA_NETHCONTROLLERS];
/****************************************************************************
* Private Function Prototypes
****************************************************************************/
/* Register operations ******************************************************/
#if defined(CONFIG_TIVA_ETHERNET_REGDEBUG) && defined(CONFIG_DEBUG)
static uint32_t tiva_getreg(uint32_t addr);
static void tiva_putreg(uint32_t val, uint32_t addr);
static void tiva_checksetup(void);
#else
# define tiva_getreg(addr) getreg32(addr)
# define tiva_putreg(val,addr) putreg32(val,addr)
# define tiva_checksetup()
#endif
/* Free buffer management */
static void tiva_initbuffer(FAR struct tiva_ethmac_s *priv);
static inline uint8_t *tiva_allocbuffer(FAR struct tiva_ethmac_s *priv);
static inline void tiva_freebuffer(FAR struct tiva_ethmac_s *priv, uint8_t *buffer);
static inline bool tiva_isfreebuffer(FAR struct tiva_ethmac_s *priv);
/* Common TX logic */
static int tiva_transmit(FAR struct tiva_ethmac_s *priv);
static int tiva_txpoll(struct net_driver_s *dev);
static void tiva_dopoll(FAR struct tiva_ethmac_s *priv);
/* Interrupt handling */
static void tiva_enableint(FAR struct tiva_ethmac_s *priv, uint32_t ierbit);
static void tiva_disableint(FAR struct tiva_ethmac_s *priv, uint32_t ierbit);
static void tiva_freesegment(FAR struct tiva_ethmac_s *priv,
FAR struct emac_rxdesc_s *rxfirst, int segments);
static int tiva_recvframe(FAR struct tiva_ethmac_s *priv);
static void tiva_receive(FAR struct tiva_ethmac_s *priv);
static void tiva_freeframe(FAR struct tiva_ethmac_s *priv);
static void tiva_txdone(FAR struct tiva_ethmac_s *priv);
static inline void tiva_interrupt_process(FAR struct tiva_ethmac_s *priv);
#ifdef CONFIG_NET_NOINTS
static void tiva_interrupt_work(FAR void *arg);
#endif
static int tiva_interrupt(int irq, FAR void *context);
/* Watchdog timer expirations */
static inline void tiva_txtimeout_process(FAR struct tiva_ethmac_s *priv);
#ifdef CONFIG_NET_NOINTS
static void tiva_txtimeout_work(FAR void *arg);
#endif
static void tiva_txtimeout_expiry(int argc, uint32_t arg, ...);
static inline void tiva_poll_process(FAR struct tiva_ethmac_s *priv);
#ifdef CONFIG_NET_NOINTS
static void tiva_poll_work(FAR void *arg);
#endif
static void tiva_poll_expiry(int argc, uint32_t arg, ...);
/* NuttX callback functions */
static int tiva_ifup(struct net_driver_s *dev);
static int tiva_ifdown(struct net_driver_s *dev);
static inline void tiva_txavail_process(FAR struct tiva_ethmac_s *priv);
#ifdef CONFIG_NET_NOINTS
static void tiva_txavail_work(FAR void *arg);
#endif
static int tiva_txavail(struct net_driver_s *dev);
#if defined(CONFIG_NET_IGMP) || defined(CONFIG_NET_ICMPv6)
static int tiva_addmac(struct net_driver_s *dev, FAR const uint8_t *mac);
#endif
#ifdef CONFIG_NET_IGMP
static int tiva_rmmac(struct net_driver_s *dev, FAR const uint8_t *mac);
#endif
#ifdef CONFIG_NETDEV_PHY_IOCTL
static int tiva_ioctl(struct net_driver_s *dev, int cmd, long arg);
#endif
/* Descriptor Initialization */
static void tiva_txdescinit(FAR struct tiva_ethmac_s *priv);
static void tiva_rxdescinit(FAR struct tiva_ethmac_s *priv);
/* PHY Initialization */
#if CONFIG_TIVA_PHY_INTERRUPTS
static void tiva_phy_intenable(bool enable);
#endif
static int tiva_phyread(uint16_t phydevaddr, uint16_t phyregaddr, uint16_t *value);
static int tiva_phywrite(uint16_t phydevaddr, uint16_t phyregaddr, uint16_t value);
static int tiva_phyinit(FAR struct tiva_ethmac_s *priv);
/* MAC/DMA Initialization */
static void tiva_phy_configure(FAR struct tiva_ethmac_s *priv);
static inline void tiva_phy_initialize(FAR struct tiva_ethmac_s *priv);
static void tiva_ethreset(FAR struct tiva_ethmac_s *priv);
static int tiva_macconfig(FAR struct tiva_ethmac_s *priv);
static void tiva_macaddress(FAR struct tiva_ethmac_s *priv);
#ifdef CONFIG_NET_ICMPv6
static void tiva_ipv6multicast(FAR struct tiva_ethmac_s *priv);
#endif
static int tiva_macenable(FAR struct tiva_ethmac_s *priv);
static int tive_emac_configure(FAR struct tiva_ethmac_s *priv);
/****************************************************************************
* Private Functions
****************************************************************************/
/****************************************************************************
* Name: tiva_getreg
*
* Description:
* This function may to used to intercept an monitor all register accesses.
* Clearly this is nothing you would want to do unless you are debugging
* this driver.
*
* Input Parameters:
* addr - The register address to read
*
* Returned Value:
* The value read from the register
*
****************************************************************************/
#if defined(CONFIG_TIVA_ETHERNET_REGDEBUG) && defined(CONFIG_DEBUG)
static uint32_t tiva_getreg(uint32_t addr)
{
static uint32_t prevaddr = 0;
static uint32_t preval = 0;
static uint32_t count = 0;
/* Read the value from the register */
uint32_t val = getreg32(addr);
/* Is this the same value that we read from the same register last time?
* Are we polling the register? If so, suppress some of the output.
*/
if (addr == prevaddr && val == preval)
{
if (count == 0xffffffff || ++count > 3)
{
if (count == 4)
{
lldbg("...\n");
}
return val;
}
}
/* No this is a new address or value */
else
{
/* Did we print "..." for the previous value? */
if (count > 3)
{
/* Yes.. then show how many times the value repeated */
lldbg("[repeats %d more times]\n", count-3);
}
/* Save the new address, value, and count */
prevaddr = addr;
preval = val;
count = 1;
}
/* Show the register value read */
lldbg("%08x->%08x\n", addr, val);
return val;
}
#endif
/****************************************************************************
* Name: tiva_putreg
*
* Description:
* This function may to used to intercept an monitor all register accesses.
* Clearly this is nothing you would want to do unless you are debugging
* this driver.
*
* Input Parameters:
* val - The value to write to the register
* addr - The register address to read
*
* Returned Value:
* None
*
****************************************************************************/
#if defined(CONFIG_TIVA_ETHERNET_REGDEBUG) && defined(CONFIG_DEBUG)
static void tiva_putreg(uint32_t val, uint32_t addr)
{
/* Show the register value being written */
lldbg("%08x<-%08x\n", addr, val);
/* Write the value */
putreg32(val, addr);
}
#endif
/****************************************************************************
* Name: tiva_checksetup
*
* Description:
* Show the state of critical configuration registers.
*
* Input Parameters:
* None
*
* Returned Value:
* None
*
****************************************************************************/
#if defined(CONFIG_TIVA_ETHERNET_REGDEBUG) && defined(CONFIG_DEBUG)
static void tiva_checksetup(void)
{
}
#endif
/****************************************************************************
* Function: tiva_initbuffer
*
* Description:
* Initialize the free buffer list.
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* None
*
* Assumptions:
* Called during early driver initialization before Ethernet interrupts
* are enabled.
*
****************************************************************************/
static void tiva_initbuffer(FAR struct tiva_ethmac_s *priv)
{
uint8_t *buffer;
int i;
/* Initialize the head of the free buffer list */
sq_init(&priv->freeb);
/* Add all of the pre-allocated buffers to the free buffer list */
for (i = 0, buffer = priv->alloc;
i < TIVA_EMAC_NFREEBUFFERS;
i++, buffer += OPTIMAL_EMAC_BUFSIZE)
{
sq_addlast((FAR sq_entry_t *)buffer, &priv->freeb);
}
}
/****************************************************************************
* Function: tiva_allocbuffer
*
* Description:
* Allocate one buffer from the free buffer list.
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* Pointer to the allocated buffer on success; NULL on failure
*
* Assumptions:
* May or may not be called from an interrupt handler. In either case,
* global interrupts are disabled, either explicitly or indirectly through
* interrupt handling logic.
*
****************************************************************************/
static inline uint8_t *tiva_allocbuffer(FAR struct tiva_ethmac_s *priv)
{
/* Allocate a buffer by returning the head of the free buffer list */
return (uint8_t *)sq_remfirst(&priv->freeb);
}
/****************************************************************************
* Function: tiva_freebuffer
*
* Description:
* Return a buffer to the free buffer list.
*
* Parameters:
* priv - Reference to the driver state structure
* buffer - A pointer to the buffer to be freed
*
* Returned Value:
* None
*
* Assumptions:
* May or may not be called from an interrupt handler. In either case,
* global interrupts are disabled, either explicitly or indirectly through
* interrupt handling logic.
*
****************************************************************************/
static inline void tiva_freebuffer(FAR struct tiva_ethmac_s *priv, uint8_t *buffer)
{
/* Free the buffer by adding it to to the end of the free buffer list */
sq_addlast((FAR sq_entry_t *)buffer, &priv->freeb);
}
/****************************************************************************
* Function: tiva_isfreebuffer
*
* Description:
* Return TRUE if the free buffer list is not empty.
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* True if there are one or more buffers in the free buffer list;
* false if the free buffer list is empty
*
* Assumptions:
* None.
*
****************************************************************************/
static inline bool tiva_isfreebuffer(FAR struct tiva_ethmac_s *priv)
{
/* Return TRUE if the free buffer list is not empty */
return !sq_empty(&priv->freeb);
}
/****************************************************************************
* Function: tiva_transmit
*
* Description:
* Start hardware transmission. Called either from the txdone interrupt
* handling or from watchdog based polling.
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* OK on success; a negated errno on failure
*
* Assumptions:
* May or may not be called from an interrupt handler. In either case,
* global interrupts are disabled, either explicitly or indirectly through
* interrupt handling logic.
*
****************************************************************************/
static int tiva_transmit(FAR struct tiva_ethmac_s *priv)
{
struct emac_txdesc_s *txdesc;
struct emac_txdesc_s *txfirst;
/* The internal (optimal) uIP buffer size may be configured to be larger
* than the Ethernet buffer size.
*/
#if OPTIMAL_EMAC_BUFSIZE > OPTIMAL_EMAC_BUFSIZE
uint8_t *buffer;
int bufcount;
int lastsize;
int i;
#endif
/* Verify that the hardware is ready to send another packet. If we get
* here, then we are committed to sending a packet; Higher level logic
* must have assured that there is no transmission in progress.
*/
txdesc = priv->txhead;
txfirst = txdesc;
nvdbg("d_len: %d d_buf: %p txhead: %p tdes0: %08x\n",
priv->dev.d_len, priv->dev.d_buf, txdesc, txdesc->tdes0);
DEBUGASSERT(txdesc && (txdesc->tdes0 & EMAC_TDES0_OWN) == 0);
/* Is the size to be sent greater than the size of the Ethernet buffer? */
DEBUGASSERT(priv->dev.d_len > 0 && priv->dev.d_buf != NULL);
#if OPTIMAL_EMAC_BUFSIZE > OPTIMAL_EMAC_BUFSIZE
if (priv->dev.d_len > OPTIMAL_EMAC_BUFSIZE)
{
/* Yes... how many buffers will be need to send the packet? */
bufcount = (priv->dev.d_len + (OPTIMAL_EMAC_BUFSIZE-1)) / OPTIMAL_EMAC_BUFSIZE;
lastsize = priv->dev.d_len - (bufcount - 1) * OPTIMAL_EMAC_BUFSIZE;
nvdbg("bufcount: %d lastsize: %d\n", bufcount, lastsize);
/* Set the first segment bit in the first TX descriptor */
txdesc->tdes0 |= EMAC_TDES0_FS;
/* Set up all but the last TX descriptor */
buffer = priv->dev.d_buf;
for (i = 0; i < bufcount; i++)
{
/* This could be a normal event but the design does not handle it */
DEBUGASSERT((txdesc->tdes0 & EMAC_TDES0_OWN) == 0);
/* Set the Buffer1 address pointer */
txdesc->tdes2 = (uint32_t)buffer;
/* Set the buffer size in all TX descriptors */
if (i == (bufcount-1))
{
/* This is the last segment. Set the last segment bit in the
* last TX descriptor and ask for an interrupt when this
* segment transfer completes.
*/
txdesc->tdes0 |= (EMAC_TDES0_LS | EMAC_TDES0_IC);
/* This segement is, most likely, of fractional buffersize */
txdesc->tdes1 = lastsize;
buffer += lastsize;
}
else
{
/* This is not the last segment. We don't want an interrupt
* when this segment transfer completes.
*/
txdesc->tdes0 &= ~EMAC_TDES0_IC;
/* The size of the transfer is the whole buffer */
txdesc->tdes1 = OPTIMAL_EMAC_BUFSIZE;
buffer += OPTIMAL_EMAC_BUFSIZE;
}
/* Give the descriptor to DMA */
txdesc->tdes0 |= EMAC_TDES0_OWN;
txdesc = (struct emac_txdesc_s *)txdesc->tdes3;
}
}
else
#endif
{
/* The single descriptor is both the first and last segment. And we do
* want an interrupt when the transfer completes.
*/
txdesc->tdes0 |= (EMAC_TDES0_FS | EMAC_TDES0_LS | EMAC_TDES0_IC);
/* Set frame size */
DEBUGASSERT(priv->dev.d_len <= CONFIG_NET_ETH_MTU);
txdesc->tdes1 = priv->dev.d_len;
/* Set the Buffer1 address pointer */
txdesc->tdes2 = (uint32_t)priv->dev.d_buf;
/* Set OWN bit of the TX descriptor tdes0. This gives the buffer to
* Ethernet DMA
*/
txdesc->tdes0 |= EMAC_TDES0_OWN;
/* Point to the next available TX descriptor */
txdesc = (struct emac_txdesc_s *)txdesc->tdes3;
}
/* Remember where we left off in the TX descriptor chain */
priv->txhead = txdesc;
/* Detach the buffer from priv->dev structure. That buffer is now
* "in-flight".
*/
priv->dev.d_buf = NULL;
priv->dev.d_len = 0;
/* If there is no other TX buffer, in flight, then remember the location
* of the TX descriptor. This is the location to check for TX done events.
*/
if (!priv->txtail)
{
DEBUGASSERT(priv->inflight == 0);
priv->txtail = txfirst;
}
/* Increment the number of TX transfer in-flight */
priv->inflight++;
nvdbg("txhead: %p txtail: %p inflight: %d\n",
priv->txhead, priv->txtail, priv->inflight);
/* If all TX descriptors are in-flight, then we have to disable receive interrupts
* too. This is because receive events can trigger more un-stoppable transmit
* events.
*/
if (priv->inflight >= CONFIG_TIVA_EMAC_NTXDESC)
{
tiva_disableint(priv, EMAC_DMAINT_RI);
}
/* Check if the TX Buffer unavailable flag is set */
if ((tiva_getreg(TIVA_EMAC_DMARIS) & EMAC_DMAINT_TBUI) != 0)
{
/* Clear TX Buffer unavailable flag */
tiva_putreg(EMAC_DMAINT_TBUI, TIVA_EMAC_DMARIS);
/* Resume DMA transmission */
tiva_putreg(0, TIVA_EMAC_TXPOLLD);
}
/* Enable TX interrupts */
tiva_enableint(priv, EMAC_DMAINT_TI);
/* Setup the TX timeout watchdog (perhaps restarting the timer) */
(void)wd_start(priv->txtimeout, TIVA_TXTIMEOUT, tiva_txtimeout_expiry, 1, (uint32_t)priv);
return OK;
}
/****************************************************************************
* Function: tiva_txpoll
*
* Description:
* The transmitter is available, check if uIP has any outgoing packets ready
* to send. This is a callback from devif_poll(). devif_poll() may be called:
*
* 1. When the preceding TX packet send is complete,
* 2. When the preceding TX packet send timesout and the interface is reset
* 3. During normal TX polling
*
* Parameters:
* dev - Reference to the NuttX driver state structure
*
* Returned Value:
* OK on success; a negated errno on failure
*
* Assumptions:
* May or may not be called from an interrupt handler. In either case,
* global interrupts are disabled, either explicitly or indirectly through
* interrupt handling logic.
*
****************************************************************************/
static int tiva_txpoll(struct net_driver_s *dev)
{
FAR struct tiva_ethmac_s *priv = (FAR struct tiva_ethmac_s *)dev->d_private;
DEBUGASSERT(priv->dev.d_buf != NULL);
/* If the polling resulted in data that should be sent out on the network,
* the field d_len is set to a value > 0.
*/
if (priv->dev.d_len > 0)
{
/* Look up the destination MAC address and add it to the Ethernet
* header.
*/
#ifdef CONFIG_NET_IPv4
#ifdef CONFIG_NET_IPv6
if (IFF_IS_IPv4(priv->dev.d_flags))
#endif
{
arp_out(&priv->dev);
}
#endif /* CONFIG_NET_IPv4 */
#ifdef CONFIG_NET_IPv6
#ifdef CONFIG_NET_IPv4
else
#endif
{
neighbor_out(&priv->dev);
}
#endif /* CONFIG_NET_IPv6 */
/* Send the packet */
tiva_transmit(priv);
DEBUGASSERT(dev->d_len == 0 && dev->d_buf == NULL);
/* Check if the next TX descriptor is owned by the Ethernet DMA or CPU. We
* cannot perform the TX poll if we are unable to accept another packet for
* transmission.
*
* In a race condition, EMAC_TDES0_OWN may be cleared BUT still not available
* because tiva_freeframe() has not yet run. If tiva_freeframe() has run,
* the buffer1 pointer (tdes2) will be nullified (and inflight should be <
* CONFIG_TIVA_EMAC_NTXDESC).
*/
if ((priv->txhead->tdes0 & EMAC_TDES0_OWN) != 0 ||
priv->txhead->tdes2 != 0)
{
/* We have to terminate the poll if we have no more descriptors
* available for another transfer.
*/
return -EBUSY;
}
/* We have the descriptor, we can continue the poll. Allocate a new
* buffer for the poll.
*/
dev->d_buf = tiva_allocbuffer(priv);
/* We can't continue the poll if we have no buffers */
if (dev->d_buf == NULL)
{
/* Terminate the poll. */
return -ENOMEM;
}
}
/* If zero is returned, the polling will continue until all connections have
* been examined.
*/
return 0;
}
/****************************************************************************
* Function: tiva_dopoll
*
* Description:
* The function is called when a frame is received using the DMA receive
* interrupt. It scans the RX descriptors to the received frame.
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* None
*
* Assumptions:
* Global interrupts are disabled by interrupt handling logic.
*
****************************************************************************/
static void tiva_dopoll(FAR struct tiva_ethmac_s *priv)
{
FAR struct net_driver_s *dev = &priv->dev;
/* Check if the next TX descriptor is owned by the Ethernet DMA or
* CPU. We cannot perform the TX poll if we are unable to accept
* another packet for transmission.
*
* In a race condition, EMAC_TDES0_OWN may be cleared BUT still not available
* because tiva_freeframe() has not yet run. If tiva_freeframe() has run,
* the buffer1 pointer (tdes2) will be nullified (and inflight should be <
* CONFIG_TIVA_EMAC_NTXDESC).
*/
if ((priv->txhead->tdes0 & EMAC_TDES0_OWN) == 0 &&
priv->txhead->tdes2 == 0)
{
/* If we have the descriptor, then poll uIP for new XMIT data.
* Allocate a buffer for the poll.
*/
DEBUGASSERT(dev->d_len == 0 && dev->d_buf == NULL);
dev->d_buf = tiva_allocbuffer(priv);
/* We can't poll if we have no buffers */
if (dev->d_buf)
{
(void)devif_poll(dev, tiva_txpoll);
/* We will, most likely end up with a buffer to be freed. But it
* might not be the same one that we allocated above.
*/
if (dev->d_buf)
{
DEBUGASSERT(dev->d_len == 0);
tiva_freebuffer(priv, dev->d_buf);
dev->d_buf = NULL;
}
}
}
}
/****************************************************************************
* Function: tiva_enableint
*
* Description:
* Enable a "normal" interrupt
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* None
*
* Assumptions:
* Global interrupts are disabled by interrupt handling logic.
*
****************************************************************************/
static void tiva_enableint(FAR struct tiva_ethmac_s *priv, uint32_t ierbit)
{
uint32_t regval;
/* Enable the specified "normal" interrupt */
regval = tiva_getreg(TIVA_EMAC_DMAIM);
regval |= (EMAC_DMAINT_NIS | ierbit);
tiva_putreg(regval, TIVA_EMAC_DMAIM);
}
/****************************************************************************
* Function: tiva_disableint
*
* Description:
* Disable a normal interrupt.
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* None
*
* Assumptions:
* Global interrupts are disabled by interrupt handling logic.
*
****************************************************************************/
static void tiva_disableint(FAR struct tiva_ethmac_s *priv, uint32_t ierbit)
{
uint32_t regval;
/* Disable the "normal" interrupt */
regval = tiva_getreg(TIVA_EMAC_DMAIM);
regval &= ~ierbit;
/* Are all "normal" interrupts now disabled? */
if ((regval & EMAC_DMAINT_NORMAL) == 0)
{
/* Yes.. disable normal interrupts */
regval &= ~EMAC_DMAINT_NIS;
}
tiva_putreg(regval, TIVA_EMAC_DMAIM);
}
/****************************************************************************
* Function: tiva_freesegment
*
* Description:
* The function is called when a frame is received using the DMA receive
* interrupt. It scans the RX descriptors to the received frame.
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* None
*
* Assumptions:
* Global interrupts are disabled by interrupt handling logic.
*
****************************************************************************/
static void tiva_freesegment(FAR struct tiva_ethmac_s *priv,
FAR struct emac_rxdesc_s *rxfirst, int segments)
{
struct emac_rxdesc_s *rxdesc;
int i;
nvdbg("rxfirst: %p segments: %d\n", rxfirst, segments);
/* Set OWN bit in RX descriptors. This gives the buffers back to DMA */
rxdesc = rxfirst;
for (i = 0; i < segments; i++)
{
rxdesc->rdes0 = EMAC_RDES0_OWN;
rxdesc = (struct emac_rxdesc_s *)rxdesc->rdes3;
}
/* Reset the segment managment logic */
priv->rxcurr = NULL;
priv->segments = 0;
/* Check if the RX Buffer unavailable flag is set */
if ((tiva_getreg(TIVA_EMAC_DMARIS) & EMAC_DMAINT_RBUI) != 0)
{
/* Clear RBUS Ethernet DMA flag */
tiva_putreg(EMAC_DMAINT_RBUI, TIVA_EMAC_DMARIS);
/* Resume DMA reception */
tiva_putreg(0, TIVA_EMAC_RXPOLLD);
}
}
/****************************************************************************
* Function: tiva_recvframe
*
* Description:
* The function is called when a frame is received using the DMA receive
* interrupt. It scans the RX descriptors of the received frame.
*
* NOTE: This function will silently discard any packets containing errors.
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* OK if a packet was successfully returned; -EAGAIN if there are no
* further packets available
*
* Assumptions:
* Global interrupts are disabled by interrupt handling logic.
*
****************************************************************************/
static int tiva_recvframe(FAR struct tiva_ethmac_s *priv)
{
struct emac_rxdesc_s *rxdesc;
struct emac_rxdesc_s *rxcurr;
uint8_t *buffer;
int i;
nvdbg("rxhead: %p rxcurr: %p segments: %d\n",
priv->rxhead, priv->rxcurr, priv->segments);
/* Check if there are free buffers. We cannot receive new frames in this
* design unless there is at least one free buffer.
*/
if (!tiva_isfreebuffer(priv))
{
nlldbg("No free buffers\n");
return -ENOMEM;
}
/* Scan descriptors owned by the CPU. Scan until:
*
* 1) We find a descriptor still owned by the DMA,
* 2) We have examined all of the RX descriptors, or
* 3) All of the TX descriptors are in flight.
*
* This last case is obscure. It is due to that fact that each packet
* that we receive can generate an unstoppable transmisson. So we have
* to stop receiving when we can not longer transmit. In this case, the
* transmit logic should also have disabled further RX interrupts.
*/
rxdesc = priv->rxhead;
for (i = 0;
(rxdesc->rdes0 & EMAC_RDES0_OWN) == 0 &&
i < CONFIG_TIVA_EMAC_NRXDESC &&
priv->inflight < CONFIG_TIVA_EMAC_NTXDESC;
i++)
{
/* Check if this is the first segment in the frame */
if ((rxdesc->rdes0 & EMAC_RDES0_FS) != 0 &&
(rxdesc->rdes0 & EMAC_RDES0_LS) == 0)
{
priv->rxcurr = rxdesc;
priv->segments = 1;
}
/* Check if this is an intermediate segment in the frame */
else if (((rxdesc->rdes0 & EMAC_RDES0_LS) == 0)&&
((rxdesc->rdes0 & EMAC_RDES0_FS) == 0))
{
priv->segments++;
}
/* Otherwise, it is the last segment in the frame */
else
{
priv->segments++;
/* Check if the there is only one segment in the frame */
if (priv->segments == 1)
{
rxcurr = rxdesc;
}
else
{
rxcurr = priv->rxcurr;
}
nvdbg("rxhead: %p rxcurr: %p segments: %d\n",
priv->rxhead, priv->rxcurr, priv->segments);
/* Check if any errors are reported in the frame */
if ((rxdesc->rdes0 & EMAC_RDES0_ES) == 0)
{
struct net_driver_s *dev = &priv->dev;
/* Get the Frame Length of the received packet: substruct 4
* bytes of the CRC
*/
dev->d_len = ((rxdesc->rdes0 & EMAC_RDES0_FL_MASK) >> EMAC_RDES0_FL_SHIFT) - 4;
/* Get a buffer from the free list. We don't even check if
* this is successful because we already assure the free
* list is not empty above.
*/
buffer = tiva_allocbuffer(priv);
/* Take the buffer from the RX descriptor of the first free
* segment, put it into the uIP device structure, then replace
* the buffer in the RX descriptor with the newly allocated
* buffer.
*/
DEBUGASSERT(dev->d_buf == NULL);
dev->d_buf = (uint8_t*)rxcurr->rdes2;
rxcurr->rdes2 = (uint32_t)buffer;
/* Return success, remebering where we should re-start scanning
* and resetting the segment scanning logic
*/
priv->rxhead = (struct emac_rxdesc_s*)rxdesc->rdes3;
tiva_freesegment(priv, rxcurr, priv->segments);
nvdbg("rxhead: %p d_buf: %p d_len: %d\n",
priv->rxhead, dev->d_buf, dev->d_len);
return OK;
}
else
{
/* Drop the frame that contains the errors, reset the segment
* scanning logic, and continue scanning with the next frame.
*/
nlldbg("DROPPED: RX descriptor errors: %08x\n", rxdesc->rdes0);
tiva_freesegment(priv, rxcurr, priv->segments);
}
}
/* Try the next descriptor */
rxdesc = (struct emac_rxdesc_s*)rxdesc->rdes3;
}
/* We get here after all of the descriptors have been scanned or when rxdesc points
* to the first descriptor owned by the DMA. Remember where we left off.
*/
priv->rxhead = rxdesc;
nvdbg("rxhead: %p rxcurr: %p segments: %d\n",
priv->rxhead, priv->rxcurr, priv->segments);
return -EAGAIN;
}
/****************************************************************************
* Function: tiva_receive
*
* Description:
* An interrupt was received indicating the availability of a new RX packet
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* None
*
* Assumptions:
* Global interrupts are disabled by interrupt handling logic.
*
****************************************************************************/
static void tiva_receive(FAR struct tiva_ethmac_s *priv)
{
struct net_driver_s *dev = &priv->dev;
/* Loop while while tiva_recvframe() successfully retrieves valid
* Ethernet frames.
*/
while (tiva_recvframe(priv) == OK)
{
#ifdef CONFIG_NET_PKT
/* When packet sockets are enabled, feed the frame into the packet tap */
pkt_input(&priv->dev);
#endif
/* Check if the packet is a valid size for the uIP buffer configuration
* (this should not happen)
*/
if (dev->d_len > CONFIG_NET_ETH_MTU)
{
nlldbg("DROPPED: Too big: %d\n", dev->d_len);
}
else
/* We only accept IP packets of the configured type and ARP packets */
#ifdef CONFIG_NET_IPv4
if (BUF->type == HTONS(ETHTYPE_IP))
{
nllvdbg("IPv4 frame\n");
/* Handle ARP on input then give the IPv4 packet to the network
* layer
*/
arp_ipin(&priv->dev);
ipv4_input(&priv->dev);
/* If the above function invocation resulted in data that should be
* sent out on the network, the field d_len will set to a value > 0.
*/
if (priv->dev.d_len > 0)
{
/* Update the Ethernet header with the correct MAC address */
#ifdef CONFIG_NET_IPv6
if (IFF_IS_IPv4(priv->dev.d_flags))
#endif
{
arp_out(&priv->dev);
}
#ifdef CONFIG_NET_IPv6
else
{
neighbor_out(&priv->dev);
}
#endif
/* And send the packet */
tiva_transmit(priv);
}
}
else
#endif
#ifdef CONFIG_NET_IPv6
if (BUF->type == HTONS(ETHTYPE_IP6))
{
nllvdbg("IPv6 frame\n");
/* Give the IPv6 packet to the network layer */
ipv6_input(&priv->dev);
/* If the above function invocation resulted in data that should be
* sent out on the network, the field d_len will set to a value > 0.
*/
if (priv->dev.d_len > 0)
{
/* Update the Ethernet header with the correct MAC address */
#ifdef CONFIG_NET_IPv4
if (IFF_IS_IPv4(priv->dev.d_flags))
{
arp_out(&priv->dev);
}
else
#endif
#ifdef CONFIG_NET_IPv6
{
neighbor_out(&priv->dev);
}
#endif
/* And send the packet */
tiva_transmit(priv);
}
}
else
#endif
#ifdef CONFIG_NET_ARP
if (BUF->type == htons(ETHTYPE_ARP))
{
nvdbg("ARP frame\n");
/* Handle ARP packet */
arp_arpin(&priv->dev);
/* If the above function invocation resulted in data that should be
* sent out on the network, the field d_len will set to a value > 0.
*/
if (priv->dev.d_len > 0)
{
tiva_transmit(priv);
}
}
else
#endif
{
nlldbg("DROPPED: Unknown type: %04x\n", BUF->type);
}
/* We are finished with the RX buffer. NOTE: If the buffer is
* re-used for transmission, the dev->d_buf field will have been
* nullified.
*/
if (dev->d_buf)
{
/* Free the receive packet buffer */
tiva_freebuffer(priv, dev->d_buf);
dev->d_buf = NULL;
dev->d_len = 0;
}
}
}
/****************************************************************************
* Function: tiva_freeframe
*
* Description:
* Scans the TX descriptors and frees the buffers of completed TX transfers.
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* None.
*
* Assumptions:
* Global interrupts are disabled by interrupt handling logic.
*
****************************************************************************/
static void tiva_freeframe(FAR struct tiva_ethmac_s *priv)
{
FAR struct emac_txdesc_s *txdesc;
int i;
nvdbg("txhead: %p txtail: %p inflight: %d\n",
priv->txhead, priv->txtail, priv->inflight);
/* Scan for "in-flight" descriptors owned by the CPU */
txdesc = priv->txtail;
if (txdesc)
{
DEBUGASSERT(priv->inflight > 0);
for (i = 0; (txdesc->tdes0 & EMAC_TDES0_OWN) == 0; i++)
{
/* There should be a buffer assigned to all in-flight
* TX descriptors.
*/
nvdbg("txtail: %p tdes0: %08x tdes2: %08x tdes3: %08x\n",
txdesc, txdesc->tdes0, txdesc->tdes2, txdesc->tdes3);
DEBUGASSERT(txdesc->tdes2 != 0);
/* Check if this is the first segment of a TX frame. */
if ((txdesc->tdes0 & EMAC_TDES0_FS) != 0)
{
/* Yes.. Free the buffer */
tiva_freebuffer(priv, (uint8_t*)txdesc->tdes2);
}
/* In any event, make sure that TDES2 is nullified. */
txdesc->tdes2 = 0;
/* Check if this is the last segment of a TX frame */
if ((txdesc->tdes0 & EMAC_TDES0_LS) != 0)
{
/* Yes.. Decrement the number of frames "in-flight". */
priv->inflight--;
/* If all of the TX descriptors were in-flight, then RX interrupts
* may have been disabled... we can re-enable them now.
*/
tiva_enableint(priv, EMAC_DMAINT_RI);
/* If there are no more frames in-flight, then bail. */
if (priv->inflight <= 0)
{
priv->txtail = NULL;
priv->inflight = 0;
return;
}
}
/* Try the next descriptor in the TX chain */
txdesc = (struct emac_txdesc_s*)txdesc->tdes3;
}
/* We get here if (1) there are still frames "in-flight". Remember
* where we left off.
*/
priv->txtail = txdesc;
nvdbg("txhead: %p txtail: %p inflight: %d\n",
priv->txhead, priv->txtail, priv->inflight);
}
}
/****************************************************************************
* Function: tiva_txdone
*
* Description:
* An interrupt was received indicating that the last TX packet(s) is done
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* None
*
* Assumptions:
* Global interrupts are disabled by the watchdog logic.
*
****************************************************************************/
static void tiva_txdone(FAR struct tiva_ethmac_s *priv)
{
FAR struct net_driver_s *dev = &priv->dev;
DEBUGASSERT(priv->txtail != NULL);
/* Scan the TX descriptor change, returning buffers to free list */
tiva_freeframe(priv);
dev->d_buf = NULL;
dev->d_len = 0;
/* If no further xmits are pending, then cancel the TX timeout */
if (priv->inflight <= 0)
{
wd_cancel(priv->txtimeout);
/* And disable further TX interrupts. */
tiva_disableint(priv, EMAC_DMAINT_TI);
}
/* Then poll uIP for new XMIT data */
tiva_dopoll(priv);
}
/****************************************************************************
* Function: tiva_interrupt_process
*
* Description:
* Interrupt processing. This may be performed either within the interrupt
* handler or on the worker thread, depending upon the configuration
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* None
*
* Assumptions:
* Ethernet interrupts are disabled
*
****************************************************************************/
static inline void tiva_interrupt_process(FAR struct tiva_ethmac_s *priv)
{
uint32_t dmaris;
/* Get the DMA interrupt status bits (no MAC interrupts are expected) */
dmaris = tiva_getreg(TIVA_EMAC_DMARIS);
/* Mask only enabled interrupts. This depends on the fact that the interrupt
* related bits (0-16) correspond in these two registers.
*/
dmaris &= tiva_getreg(TIVA_EMAC_DMAIM);
/* Check if there are pending "normal" interrupts */
if ((dmaris & EMAC_DMAINT_NIS) != 0)
{
/* Yes.. Check if we received an incoming packet, if so, call
* tiva_receive()
*/
if ((dmaris & EMAC_DMAINT_RI) != 0)
{
/* Clear the pending receive interrupt */
tiva_putreg(EMAC_DMAINT_RI, TIVA_EMAC_DMARIS);
/* Handle the received package */
tiva_receive(priv);
}
/* Check if a packet transmission just completed. If so, call
* tiva_txdone(). This may disable further TX interrupts if there
* are no pending transmissions.
*/
if ((dmaris & EMAC_DMAINT_TI) != 0)
{
/* Clear the pending receive interrupt */
tiva_putreg(EMAC_DMAINT_TI, TIVA_EMAC_DMARIS);
/* Check if there are pending transmissions */
tiva_txdone(priv);
}
/* Clear the pending normal summary interrupt */
tiva_putreg(EMAC_DMAINT_NIS, TIVA_EMAC_DMARIS);
}
/* Handle error interrupt only if CONFIG_DEBUG_NET is eanbled */
#ifdef CONFIG_DEBUG_NET
/* Check if there are pending "abnormal" interrupts */
if ((dmaris & EMAC_DMAINT_AIS) != 0)
{
/* Just let the user know what happened */
nlldbg("Abnormal event(s): %08x\n", dmaris);
/* Clear all pending abnormal events */
tiva_putreg(EMAC_DMAINT_ABNORMAL, TIVA_EMAC_DMARIS);
/* Clear the pending abnormal summary interrupt */
tiva_putreg(EMAC_DMAINT_AIS, TIVA_EMAC_DMARIS);
}
#endif
}
/****************************************************************************
* Function: tiva_interrupt_work
*
* Description:
* Perform interrupt related work from the worker thread
*
* Parameters:
* arg - The argument passed when work_queue() was called.
*
* Returned Value:
* OK on success
*
* Assumptions:
* Ethernet interrupts are disabled
*
****************************************************************************/
#ifdef CONFIG_NET_NOINTS
static void tiva_interrupt_work(FAR void *arg)
{
FAR struct tiva_ethmac_s *priv = (FAR struct tiva_ethmac_s *)arg;
net_lock_t state;
DEBUGASSERT(priv);
/* Process pending Ethernet interrupts */
state = net_lock();
tiva_interrupt_process(priv);
net_unlock(state);
/* Re-enable Ethernet interrupts at the NVIC */
up_enable_irq(TIVA_IRQ_ETHCON);
}
#endif
/****************************************************************************
* Function: tiva_interrupt
*
* Description:
* Hardware interrupt handler
*
* Parameters:
* irq - Number of the IRQ that generated the interrupt
* context - Interrupt register state save info (architecture-specific)
*
* Returned Value:
* OK on success
*
* Assumptions:
*
****************************************************************************/
static int tiva_interrupt(int irq, FAR void *context)
{
FAR struct tiva_ethmac_s *priv = &g_tiva_ethmac[0];
#ifdef CONFIG_NET_NOINTS
uint32_t dmaris;
/* Get the raw interrupt status. */
dmaris = tiva_getreg(TIVA_EMAC_DMARIS);
if (dmaris != 0)
{
/* Disable further Ethernet interrupts. Because Ethernet interrupts
* are also disabled if the TX timeout event occurs, there can be no
* race condition here.
*/
up_disable_irq(TIVA_IRQ_ETHCON);
/* Check if a packet transmission just completed. */
if ((dmaris & EMAC_DMAINT_TI) != 0)
{
/* If a TX transfer just completed, then cancel the TX timeout so
* there will be no race condition between any subsequent timeout
* expiration and the deferred interrupt processing.
*/
wd_cancel(priv->txtimeout);
}
/* Cancel any pending poll work */
work_cancel(HPWORK, &priv->work);
/* Schedule to perform the interrupt processing on the worker thread. */
work_queue(HPWORK, &priv->work, tiva_interrupt_work, priv, 0);
}
#else
/* Process the interrupt now */
tiva_interrupt_process(priv);
#endif
#ifdef CONFIG_TIVA_PHY_INTERRUPTS
/* Check for pending PHY interrupts */
if ((tiva_getreg(TIVA_EPHY_MISC) & EMAC_PHYMISC_INT) != 0)
{
/* Clear the pending PHY interrupt */
tiva_putreg(EMAC_PHYMISC_INT, TIVA_EPHY_MISC);
/* Dispatch to the registered handler */
if (priv->handler)
{
(void)priv->handler(irq, context);
}
}
#endif
return OK;
}
/****************************************************************************
* Function: tiva_txtimeout_process
*
* Description:
* Process a TX timeout. Called from the either the watchdog timer
* expiration logic or from the worker thread, depending upon the
* configuration. The timeout means that the last TX never completed.
* Reset the hardware and start again.
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* None
*
* Assumptions:
* Global interrupts are disabled by the watchdog logic.
*
****************************************************************************/
static inline void tiva_txtimeout_process(FAR struct tiva_ethmac_s *priv)
{
/* Reset the hardware. Just take the interface down, then back up again. */
tiva_ifdown(&priv->dev);
tiva_ifup(&priv->dev);
/* Then poll uIP for new XMIT data */
tiva_dopoll(priv);
}
/****************************************************************************
* Function: tiva_txtimeout_work
*
* Description:
* Perform TX timeout related work from the worker thread
*
* Parameters:
* arg - The argument passed when work_queue() as called.
*
* Returned Value:
* OK on success
*
* Assumptions:
* Ethernet interrupts are disabled
*
****************************************************************************/
#ifdef CONFIG_NET_NOINTS
static void tiva_txtimeout_work(FAR void *arg)
{
FAR struct tiva_ethmac_s *priv = (FAR struct tiva_ethmac_s *)arg;
net_lock_t state;
/* Process pending Ethernet interrupts */
state = net_lock();
tiva_txtimeout_process(priv);
net_unlock(state);
}
#endif
/****************************************************************************
* Function: tiva_txtimeout_expiry
*
* Description:
* Our TX watchdog timed out. Called from the timer interrupt handler.
* The last TX never completed. Reset the hardware and start again.
*
* Parameters:
* argc - The number of available arguments
* arg - The first argument
*
* Returned Value:
* None
*
* Assumptions:
* Global interrupts are disabled by the watchdog logic.
*
****************************************************************************/
static void tiva_txtimeout_expiry(int argc, uint32_t arg, ...)
{
FAR struct tiva_ethmac_s *priv = (FAR struct tiva_ethmac_s *)arg;
nlldbg("Timeout!\n");
#ifdef CONFIG_NET_NOINTS
/* Disable further Ethernet interrupts. This will prevent some race
* conditions with interrupt work. There is still a potential race
* condition with interrupt work that is already queued and in progress.
*
* Interrupts will be re-enabled when tiva_ifup() is called.
*/
up_disable_irq(TIVA_IRQ_ETHCON);
/* Cancel any pending poll or interrupt work. This will have no effect
* on work that has already been started.
*/
work_cancel(HPWORK, &priv->work);
/* Schedule to perform the TX timeout processing on the worker thread. */
work_queue(HPWORK, &priv->work, tiva_txtimeout_work, priv, 0);
#else
/* Process the timeout now */
tiva_txtimeout_process(priv);
#endif
}
/****************************************************************************
* Function: tiva_poll_process
*
* Description:
* Perform the periodic poll. This may be called either from watchdog
* timer logic or from the worker thread, depending upon the configuration.
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* None
*
* Assumptions:
*
****************************************************************************/
static inline void tiva_poll_process(FAR struct tiva_ethmac_s *priv)
{
FAR struct net_driver_s *dev = &priv->dev;
/* Check if the next TX descriptor is owned by the Ethernet DMA or CPU. We
* cannot perform the timer poll if we are unable to accept another packet
* for transmission. Hmmm.. might be bug here. Does this mean if there is
* a transmit in progress, we will miss TCP time state updates?
*
* In a race condition, EMAC_TDES0_OWN may be cleared BUT still not available
* because tiva_freeframe() has not yet run. If tiva_freeframe() has run,
* the buffer1 pointer (tdes2) will be nullified (and inflight should be <
* CONFIG_TIVA_EMAC_NTXDESC).
*/
if ((priv->txhead->tdes0 & EMAC_TDES0_OWN) == 0 &&
priv->txhead->tdes2 == 0)
{
/* If we have the descriptor, then perform the timer poll. Allocate a
* buffer for the poll.
*/
DEBUGASSERT(dev->d_len == 0 && dev->d_buf == NULL);
dev->d_buf = tiva_allocbuffer(priv);
/* We can't poll if we have no buffers */
if (dev->d_buf)
{
/* Update TCP timing states and poll uIP for new XMIT data.
*/
(void)devif_timer(dev, tiva_txpoll, TIVA_POLLHSEC);
/* We will, most likely end up with a buffer to be freed. But it
* might not be the same one that we allocated above.
*/
if (dev->d_buf)
{
DEBUGASSERT(dev->d_len == 0);
tiva_freebuffer(priv, dev->d_buf);
dev->d_buf = NULL;
}
}
}
/* Setup the watchdog poll timer again */
(void)wd_start(priv->txpoll, TIVA_WDDELAY, tiva_poll_expiry, 1, (uint32_t)priv);
}
/****************************************************************************
* Function: tiva_poll_work
*
* Description:
* Perform periodic polling from the worker thread
*
* Parameters:
* arg - The argument passed when work_queue() as called.
*
* Returned Value:
* OK on success
*
* Assumptions:
* Ethernet interrupts are disabled
*
****************************************************************************/
#ifdef CONFIG_NET_NOINTS
static void tiva_poll_work(FAR void *arg)
{
FAR struct tiva_ethmac_s *priv = (FAR struct tiva_ethmac_s *)arg;
net_lock_t state;
/* Perform the poll */
state = net_lock();
tiva_poll_process(priv);
net_unlock(state);
}
#endif
/****************************************************************************
* Function: tiva_poll_expiry
*
* Description:
* Periodic timer handler. Called from the timer interrupt handler.
*
* Parameters:
* argc - The number of available arguments
* arg - The first argument
*
* Returned Value:
* None
*
* Assumptions:
* Global interrupts are disabled by the watchdog logic.
*
****************************************************************************/
static void tiva_poll_expiry(int argc, uint32_t arg, ...)
{
FAR struct tiva_ethmac_s *priv = (FAR struct tiva_ethmac_s *)arg;
#ifdef CONFIG_NET_NOINTS
/* Is our single work structure available? It may not be if there are
* pending interrupt actions.
*/
if (work_available(&priv->work))
{
/* Schedule to perform the interrupt processing on the worker thread. */
work_queue(HPWORK, &priv->work, tiva_poll_work, priv, 0);
}
else
{
/* No.. Just re-start the watchdog poll timer, missing one polling
* cycle.
*/
(void)wd_start(priv->txpoll, TIVA_WDDELAY, tiva_poll_expiry, 1, (uint32_t)priv);
}
#else
/* Process the interrupt now */
tiva_poll_process(priv);
#endif
}
/****************************************************************************
* Function: tiva_ifup
*
* Description:
* NuttX Callback: Bring up the Ethernet interface when an IP address is
* provided
*
* Parameters:
* dev - Reference to the NuttX driver state structure
*
* Returned Value:
* None
*
* Assumptions:
*
****************************************************************************/
static int tiva_ifup(struct net_driver_s *dev)
{
FAR struct tiva_ethmac_s *priv = (FAR struct tiva_ethmac_s *)dev->d_private;
int ret;
#ifdef CONFIG_NET_IPv4
ndbg("Bringing up: %d.%d.%d.%d\n",
dev->d_ipaddr & 0xff, (dev->d_ipaddr >> 8) & 0xff,
(dev->d_ipaddr >> 16) & 0xff, dev->d_ipaddr >> 24);
#endif
#ifdef CONFIG_NET_IPv6
ndbg("Bringing up: %04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x\n",
dev->d_ipv6addr[0], dev->d_ipv6addr[1], dev->d_ipv6addr[2],
dev->d_ipv6addr[3], dev->d_ipv6addr[4], dev->d_ipv6addr[5],
dev->d_ipv6addr[6], dev->d_ipv6addr[7]);
#endif
/* Configure the Ethernet interface for DMA operation. */
ret = tive_emac_configure(priv);
if (ret < 0)
{
return ret;
}
/* Set and activate a timer process */
(void)wd_start(priv->txpoll, TIVA_WDDELAY, tiva_poll_expiry, 1, (uint32_t)priv);
/* Enable the Ethernet interrupt */
priv->ifup = true;
up_enable_irq(TIVA_IRQ_ETHCON);
tiva_checksetup();
return OK;
}
/****************************************************************************
* Function: tiva_ifdown
*
* Description:
* NuttX Callback: Stop the interface.
*
* Parameters:
* dev - Reference to the NuttX driver state structure
*
* Returned Value:
* None
*
* Assumptions:
*
****************************************************************************/
static int tiva_ifdown(struct net_driver_s *dev)
{
FAR struct tiva_ethmac_s *priv = (FAR struct tiva_ethmac_s *)dev->d_private;
irqstate_t flags;
nvdbg("Taking the network down\n");
/* Disable the Ethernet interrupt */
flags = irqsave();
up_disable_irq(TIVA_IRQ_ETHCON);
/* Cancel the TX poll timer and TX timeout timers */
wd_cancel(priv->txpoll);
wd_cancel(priv->txtimeout);
/* Put the EMAC in its reset, non-operational state. This should be
* a known configuration that will guarantee the tiva_ifup() always
* successfully brings the interface back up.
*/
tiva_ethreset(priv);
/* Mark the device "down" */
priv->ifup = false;
irqrestore(flags);
return OK;
}
/****************************************************************************
* Function: tiva_txavail_process
*
* Description:
* Perform an out-of-cycle poll.
*
* Parameters:
* priv - Reference to the NuttX driver state structure
*
* Returned Value:
* None
*
* Assumptions:
* Called in normal user mode
*
****************************************************************************/
static inline void tiva_txavail_process(FAR struct tiva_ethmac_s *priv)
{
nvdbg("ifup: %d\n", priv->ifup);
/* Ignore the notification if the interface is not yet up */
if (priv->ifup)
{
/* Poll uIP for new XMIT data */
tiva_dopoll(priv);
}
}
/****************************************************************************
* Function: tiva_txavail_work
*
* Description:
* Perform an out-of-cycle poll on the worker thread.
*
* Parameters:
* arg - Reference to the NuttX driver state structure (cast to void*)
*
* Returned Value:
* None
*
* Assumptions:
* Called on the higher priority worker thread.
*
****************************************************************************/
#ifdef CONFIG_NET_NOINTS
static void tiva_txavail_work(FAR void *arg)
{
FAR struct tiva_ethmac_s *priv = (FAR struct tiva_ethmac_s *)arg;
net_lock_t state;
/* Perform the poll */
state = net_lock();
tiva_txavail_process(priv);
net_unlock(state);
}
#endif
/****************************************************************************
* Function: tiva_txavail
*
* Description:
* Driver callback invoked when new TX data is available. This is a
* stimulus perform an out-of-cycle poll and, thereby, reduce the TX
* latency.
*
* Parameters:
* dev - Reference to the NuttX driver state structure
*
* Returned Value:
* None
*
* Assumptions:
* Called in normal user mode
*
****************************************************************************/
static int tiva_txavail(struct net_driver_s *dev)
{
FAR struct tiva_ethmac_s *priv = (FAR struct tiva_ethmac_s *)dev->d_private;
#ifdef CONFIG_NET_NOINTS
/* Is our single work structure available? It may not be if there are
* pending interrupt actions and we will have to ignore the Tx
* availability action.
*/
if (work_available(&priv->work))
{
/* Schedule to serialize the poll on the worker thread. */
work_queue(HPWORK, &priv->work, tiva_txavail_work, priv, 0);
}
#else
irqstate_t flags;
/* Disable interrupts because this function may be called from interrupt
* level processing.
*/
flags = irqsave();
/* Perform the out-of-cycle poll now */
tiva_txavail_process(priv);
irqrestore(flags);
#endif
return OK;
}
/****************************************************************************
* Function: tiva_calcethcrc
*
* Description:
* Function to calculate the CRC used to check an ethernet frame
*
* Parameters:
* data - the data to be checked
* length - length of the data
*
* Returned Value:
* None
*
* Assumptions:
*
****************************************************************************/
#if defined(CONFIG_NET_IGMP) || defined(CONFIG_NET_ICMPv6)
static uint32_t tiva_calcethcrc(const uint8_t *data, size_t length)
{
uint32_t crc = 0xffffffff;
size_t i;
int j;
for (i = 0; i < length; i++)
{
for (j = 0; j < 8; j++)
{
if (((crc >> 31) ^ (data[i] >> j)) & 0x01)
{
/* x^26+x^23+x^22+x^16+x^12+x^11+x^10+x^8+x^7+x^5+x^4+x^2+x+1 */
crc = (crc << 1) ^ 0x04c11db7;
}
else
{
crc = crc << 1;
}
}
}
return ~crc;
}
#endif /* CONFIG_NET_IGMP || CONFIG_NET_ICMPv6 */
/****************************************************************************
* Function: tiva_addmac
*
* Description:
* NuttX Callback: Add the specified MAC address to the hardware multicast
* address filtering
*
* Parameters:
* dev - Reference to the NuttX driver state structure
* mac - The MAC address to be added
*
* Returned Value:
* None
*
* Assumptions:
*
****************************************************************************/
#if defined(CONFIG_NET_IGMP) || defined(CONFIG_NET_ICMPv6)
static int tiva_addmac(struct net_driver_s *dev, FAR const uint8_t *mac)
{
uint32_t crc;
uint32_t hashindex;
uint32_t temp;
uint32_t registeraddress;
nvdbg("MAC: %02x:%02x:%02x:%02x:%02x:%02x\n",
mac[0], mac[1], mac[2], mac[3], mac[4], mac[5]);
/* Add the MAC address to the hardware multicast hash table */
crc = tiva_calcethcrc(mac, 6);
hashindex = (crc >> 26) & 0x3F;
if (hashindex > 31)
{
registeraddress = TIVA_EMAC_HASHTBLH;
hashindex -= 32;
}
else
{
registeraddress = TIVA_EMAC_HASHTBLL;
}
temp = tiva_getreg(registeraddress);
temp |= 1 << hashindex;
tiva_putreg(temp, registeraddress);
temp = tiva_getreg(TIVA_EMAC_FRAMEFLTR);
temp |= (EMAC_FRAMEFLTR_HMC | EMAC_FRAMEFLTR_HPF);
tiva_putreg(temp, TIVA_EMAC_FRAMEFLTR);
return OK;
}
#endif
/****************************************************************************
* Function: tiva_rmmac
*
* Description:
* NuttX Callback: Remove the specified MAC address from the hardware multicast
* address filtering
*
* Parameters:
* dev - Reference to the NuttX driver state structure
* mac - The MAC address to be removed
*
* Returned Value:
* None
*
* Assumptions:
*
****************************************************************************/
#ifdef CONFIG_NET_IGMP
static int tiva_rmmac(struct net_driver_s *dev, FAR const uint8_t *mac)
{
uint32_t crc;
uint32_t hashindex;
uint32_t temp;
uint32_t registeraddress;
nvdbg("MAC: %02x:%02x:%02x:%02x:%02x:%02x\n",
mac[0], mac[1], mac[2], mac[3], mac[4], mac[5]);
/* Remove the MAC address to the hardware multicast hash table */
crc = tiva_calcethcrc(mac, 6);
hashindex = (crc >> 26) & 0x3F;
if (hashindex > 31)
{
registeraddress = TIVA_EMAC_HASHTBLH;
hashindex -= 32;
}
else
{
registeraddress = TIVA_EMAC_HASHTBLL;
}
temp = tiva_getreg(registeraddress);
temp &= ~(1 << hashindex);
tiva_putreg(temp, registeraddress);
/* If there is no address registered any more, delete multicast filtering */
if (tiva_getreg(TIVA_EMAC_HASHTBLH) == 0 &&
tiva_getreg(TIVA_EMAC_HASHTBLL) == 0)
{
temp = tiva_getreg(TIVA_EMAC_FRAMEFLTR);
temp &= ~(EMAC_FRAMEFLTR_HMC | EMAC_FRAMEFLTR_HPF);
tiva_putreg(temp, TIVA_EMAC_FRAMEFLTR);
}
return OK;
}
#endif
/****************************************************************************
* Function: tiva_txdescinit
*
* Description:
* Initializes the DMA TX descriptors in chain mode.
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* None
*
* Assumptions:
*
****************************************************************************/
static void tiva_txdescinit(FAR struct tiva_ethmac_s *priv)
{
struct emac_txdesc_s *txdesc;
int i;
/* priv->txhead will point to the first, available TX descriptor in the chain.
* Set the priv->txhead pointer to the first descriptor in the table.
*/
priv->txhead = priv->txtable;
/* priv->txtail will point to the first segment of the oldest pending
* "in-flight" TX transfer. NULL means that there are no active TX
* transfers.
*/
priv->txtail = NULL;
priv->inflight = 0;
/* Initialize each TX descriptor */
for (i = 0; i < CONFIG_TIVA_EMAC_NTXDESC; i++)
{
txdesc = &priv->txtable[i];
/* Set Second Address Chained bit */
txdesc->tdes0 = EMAC_TDES0_TCH;
#ifdef CHECKSUM_BY_HARDWARE
/* Enable the checksum insertion for the TX frames */
txdesc->tdes0 |= EMAC_TDES0_CIC_ALL;
#endif
/* Clear Buffer1 address pointer (buffers will be assigned as they
* are used)
*/
txdesc->tdes2 = 0;
/* Initialize the next descriptor with the Next Descriptor Polling Enable */
if (i < (CONFIG_TIVA_EMAC_NTXDESC-1))
{
/* Set next descriptor address register with next descriptor base
* address
*/
txdesc->tdes3 = (uint32_t)&priv->txtable[i+1];
}
else
{
/* For last descriptor, set next descriptor address register equal
* to the first descriptor base address
*/
txdesc->tdes3 = (uint32_t)priv->txtable;
}
}
/* Set Transmit Descriptor List Address Register */
tiva_putreg((uint32_t)priv->txtable, TIVA_EMAC_TXDLADDR);
}
/****************************************************************************
* Function: tiva_rxdescinit
*
* Description:
* Initializes the DMA RX descriptors in chain mode.
*
* Parameters:
* priv - Reference to the driver state structure
*
* Returned Value:
* None
*
* Assumptions:
*
****************************************************************************/
static void tiva_rxdescinit(FAR struct tiva_ethmac_s *priv)
{
struct emac_rxdesc_s *rxdesc;
int i;
/* priv->rxhead will point to the first, RX descriptor in the chain.
* This will be where we receive the first incomplete frame.
*/
priv->rxhead = priv->rxtable;
/* If we accumulate the frame in segments, priv->rxcurr points to the
* RX descriptor of the first segment in the current TX frame.
*/
priv->rxcurr = NULL;
priv->segments = 0;
/* Initialize each TX descriptor */
for (i = 0; i < CONFIG_TIVA_EMAC_NRXDESC; i++)
{
rxdesc = &priv->rxtable[i];
/* Set Own bit of the RX descriptor rdes0 */
rxdesc->rdes0 = EMAC_RDES0_OWN;
/* Set Buffer1 size and Second Address Chained bit and enabled DMA
* RX desc receive interrupt
*/
rxdesc->rdes1 = EMAC_RDES1_RCH | (uint32_t)OPTIMAL_EMAC_BUFSIZE;
/* Set Buffer1 address pointer */
rxdesc->rdes2 = (uint32_t)&priv->rxbuffer[i*OPTIMAL_EMAC_BUFSIZE];
/* Initialize the next descriptor with the Next Descriptor Polling Enable */
if (i < (CONFIG_TIVA_EMAC_NRXDESC-1))
{
/* Set next descriptor address register with next descriptor base
* address
*/
rxdesc->rdes3 = (uint32_t)&priv->rxtable[i+1];
}
else
{
/* For last descriptor, set next descriptor address register equal
* to the first descriptor base address
*/
rxdesc->rdes3 = (uint32_t)priv->rxtable;
}
}
/* Set Receive Descriptor List Address Register */
tiva_putreg((uint32_t)priv->rxtable, TIVA_EMAC_RXDLADDR);
}
/****************************************************************************
* Function: tiva_ioctl
*
* Description:
* Executes the SIOCxMIIxxx command and responds using the request struct
* that must be provided as its 2nd parameter.
*
* When called with SIOCGMIIPHY it will get the PHY address for the device
* and write it to the req->phy_id field of the request struct.
*
* When called with SIOCGMIIREG it will read a register of the PHY that is
* specified using the req->reg_no struct field and then write its output
* to the req->val_out field.
*
* When called with SIOCSMIIREG it will write to a register of the PHY that
* is specified using the req->reg_no struct field and use req->val_in as
* its input.
*
* Parameters:
* dev - Ethernet device structure
* cmd - SIOCxMIIxxx command code
* arg - Request structure also used to return values
*
* Returned Value: Negated errno on failure.
*
* Assumptions:
*
****************************************************************************/
#ifdef CONFIG_NETDEV_PHY_IOCTL
static int tiva_ioctl(struct net_driver_s *dev, int cmd, long arg)
{
int ret;
switch (cmd)
{
#ifdef CONFIG_TIVA_PHY_INTERRUPTS
case SIOCMIINOTIFY: /* Set up for PHY event notifications */
{
struct mii_iotcl_notify_s *req = (struct mii_iotcl_notify_s *)((uintptr_t)arg);
ret = phy_notify_subscribe(dev->d_ifname, req->pid, req->signo, req->arg);
if (ret == OK)
{
/* Enable PHY link up/down interrupts */
tiva_phy_intenable(true);
}
}
break;
#endif
case SIOCGMIIPHY: /* Get MII PHY address */
{
struct mii_ioctl_data_s *req = (struct mii_ioctl_data_s *)((uintptr_t)arg);
req->phy_id = CONFIG_TIVA_PHYADDR;
ret = OK;
}
break;
case SIOCGMIIREG: /* Get register from MII PHY */
{
struct mii_ioctl_data_s *req = (struct mii_ioctl_data_s *)((uintptr_t)arg);
ret = tiva_phyread(req->phy_id, req->reg_num, &req->val_out);
}
break;
case SIOCSMIIREG: /* Set register in MII PHY */
{
struct mii_ioctl_data_s *req = (struct mii_ioctl_data_s *)((uintptr_t)arg);
ret = tiva_phywrite(req->phy_id, req->reg_num, req->val_in);
}
break;
default:
ret = -ENOTTY;
break;
}
return ret;
}
#endif /* CONFIG_NETDEV_PHY_IOCTL */
/****************************************************************************
* Function: tiva_phy_intenable
*
* Description:
* Enable link up/down PHY interrupts. The interrupt protocol is like this:
*
* - Interrupt status is cleared when the interrupt is enabled.
* - Interrupt occurs. Interrupt is disabled (at the processor level) when
* is received.
* - Interrupt status is cleared when the interrupt is re-enabled.
*
* Parameters:
* priv - A reference to the private driver state structure
*
* Returned Value:
* OK on success; Negated errno (-ETIMEDOUT) on failure.
*
****************************************************************************/
#ifdef CONFIG_TIVA_PHY_INTERRUPTS
static void tiva_phy_intenable(bool enable)
{
#ifdef CONFIG_TIVA_PHY_INTERNAL
uint16_t phyval;
int ret;
/* Disable further PHY interrupts until we complete this setup */
tiva_putreg(0, TIVA_EPHY_IM);
/* Enable/disable event based PHY interrupts */
/* REVISIT: There is an issue here: The PHY interrupt handler is called
* from the interrupt level and it, in turn, will call this function to
* disabled further interrupts. Subsequent link status processing will
* also call tiva_phyread() to access PHY registers and will, eventually,
* call this function again to re-enable the PHY interrupt. The control
* between interrupt level access to the PHY and non-interrupt level
* access to the PHY is not well enforced but is probably okay just due
* to the sequencing of things.
*/
if (enable)
{
/* Configure interrupts on link status change events */
ret = tiva_phywrite(CONFIG_TIVA_PHYADDR, TIVA_EPHY_MISR1,
EPHY_MISR1_LINKSTATEN);
if (ret == OK)
{
/* Enable PHY event based interrupts */
ret = tiva_phyread(CONFIG_TIVA_PHYADDR, TIVA_EPHY_SCR, &phyval);
if (ret == OK)
{
phyval |= EPHY_SCR_INTEN;
ret = tiva_phywrite(CONFIG_TIVA_PHYADDR, TIVA_EPHY_SCR, phyval);
if (ret == OK)
{
/* Enable PHY interrupts */
tiva_putreg(EMAC_PHYIM_INT, TIVA_EPHY_IM);
}
}
}
}
else
{
/* Read the MISR1 register in order to clear any pending link status
* interrupts.
*/
ret = tiva_phyread(CONFIG_TIVA_PHYADDR, TIVA_EPHY_MISR1, &phyval);
if (ret == OK)
{
/* Disable PHY event based interrupts */
ret = tiva_phyread(CONFIG_TIVA_PHYADDR, TIVA_EPHY_SCR, &phyval);
if (ret == OK)
{
phyval |= EPHY_SCR_INTEN;
(void)tiva_phywrite(CONFIG_TIVA_PHYADDR, TIVA_EPHY_SCR, phyval);
}
}
}
#else
/* Interrupt configuration logic for external PHYs depends on the
* particular PHY part connected.
*/
#warning Missing logic
return -ENOSYS;
#endif
}
#endif
/****************************************************************************
* Function: tiva_phyread
*
* Description:
* Read a PHY register.
*
* Parameters:
* phydevaddr - The PHY device address
* phyregaddr - The PHY register address
* value - The location to return the 16-bit PHY register value.
*
* Returned Value:
* OK on success; Negated errno on failure.
*
* Assumptions:
*
****************************************************************************/
static int tiva_phyread(uint16_t phydevaddr, uint16_t phyregaddr, uint16_t *value)
{
volatile uint32_t timeout;
uint32_t regval;
/* Configure the MIIADDR register, preserving CSR Clock Range CR[2:0] bits */
regval = tiva_getreg(TIVA_EMAC_MIIADDR);
regval &= EMAC_MIIADDR_CR_MASK;
/* Set the PHY device address, PHY register address, and set the buy bit.
* the EMAC_MIIADDR_MIIW is clear, indicating a read operation.
*/
regval |= (((uint32_t)phydevaddr << EMAC_MIIADDR_PLA_SHIFT) & EMAC_MIIADDR_PLA_MASK);
regval |= (((uint32_t)phyregaddr << EMAC_MIIADDR_MII_SHIFT) & EMAC_MIIADDR_MII_MASK);
regval |= EMAC_MIIADDR_MIIB;
tiva_putreg(regval, TIVA_EMAC_MIIADDR);
/* Wait for the transfer to complete */
for (timeout = 0; timeout < PHY_READ_TIMEOUT; timeout++)
{
if ((tiva_getreg(TIVA_EMAC_MIIADDR) & EMAC_MIIADDR_MIIB) == 0)
{
*value = (uint16_t)tiva_getreg(TIVA_EMAC_MIIDATA);
return OK;
}
}
ndbg("MII transfer timed out: phydevaddr: %04x phyregaddr: %04x\n",
phydevaddr, phyregaddr);
return -ETIMEDOUT;
}
/****************************************************************************
* Function: tiva_phywrite
*
* Description:
* Write to a PHY register.
*
* Parameters:
* phydevaddr - The PHY device address
* phyregaddr - The PHY register address
* value - The 16-bit value to write to the PHY register value.
*
* Returned Value:
* OK on success; Negated errno on failure.
*
* Assumptions:
*
****************************************************************************/
static int tiva_phywrite(uint16_t phydevaddr, uint16_t phyregaddr, uint16_t value)
{
volatile uint32_t timeout;
uint32_t regval;
/* Configure the MIIADDR register, preserving CSR Clock Range CR[2:0] bits */
regval = tiva_getreg(TIVA_EMAC_MIIADDR);
regval &= EMAC_MIIADDR_CR_MASK;
/* Set the PHY device address, PHY register address, and set the busy bit.
* the EMAC_MIIADDR_MIIW is set, indicating a write operation.
*/
regval |= (((uint32_t)phydevaddr << EMAC_MIIADDR_PLA_SHIFT) & EMAC_MIIADDR_PLA_MASK);
regval |= (((uint32_t)phyregaddr << EMAC_MIIADDR_MII_SHIFT) & EMAC_MIIADDR_MII_MASK);
regval |= (EMAC_MIIADDR_MIIB | EMAC_MIIADDR_MIIW);
/* Write the value into the MACIIDR register before setting the new MIIADDR
* register value.
*/
tiva_putreg(value, TIVA_EMAC_MIIDATA);
tiva_putreg(regval, TIVA_EMAC_MIIADDR);
/* Wait for the transfer to complete */
for (timeout = 0; timeout < PHY_WRITE_TIMEOUT; timeout++)
{
if ((tiva_getreg(TIVA_EMAC_MIIADDR) & EMAC_MIIADDR_MIIB) == 0)
{
return OK;
}
}
ndbg("MII transfer timed out: phydevaddr: %04x phyregaddr: %04x value: %04x\n",
phydevaddr, phyregaddr, value);
return -ETIMEDOUT;
}
/****************************************************************************
* Function: tiva_phyinit
*
* Description:
* Configure the PHY and determine the link speed/duplex.
*
* Parameters:
* priv - A reference to the private driver state structure
*
* Returned Value:
* OK on success; Negated errno on failure.
*
* Assumptions:
*
****************************************************************************/
static int tiva_phyinit(FAR struct tiva_ethmac_s *priv)
{
#ifdef CONFIG_TIVA_AUTONEG
volatile uint32_t timeout;
#endif
uint32_t regval;
uint16_t phyval;
int ret;
/* Assume 10MBps and half duplex */
priv->mbps100 = 0;
priv->fduplex = 0;
/* Setup up PHY clocking by setting the CR field in the MIIADDR register */
regval = tiva_getreg(TIVA_EMAC_MIIADDR);
regval &= ~EMAC_MIIADDR_CR_MASK;
regval |= EMAC_MIIADDR_CR;
tiva_putreg(regval, TIVA_EMAC_MIIADDR);
/* Put the PHY in reset mode */
ret = tiva_phywrite(CONFIG_TIVA_PHYADDR, MII_MCR, MII_MCR_RESET);
if (ret < 0)
{
ndbg("Failed to reset the PHY: %d\n", ret);
return ret;
}
up_mdelay(PHY_RESET_DELAY);
/* Perform auto-negotiation if so configured */
#ifdef CONFIG_TIVA_AUTONEG
/* Wait for link status */
for (timeout = 0; timeout < PHY_RETRY_TIMEOUT; timeout++)
{
ret = tiva_phyread(CONFIG_TIVA_PHYADDR, MII_MSR, &phyval);
if (ret < 0)
{
ndbg("Failed to read the PHY MSR: %d\n", ret);
return ret;
}
else if ((phyval & MII_MSR_LINKSTATUS) != 0)
{
break;
}
}
if (timeout >= PHY_RETRY_TIMEOUT)
{
ndbg("Timed out waiting for link status: %04x\n", phyval);
return -ETIMEDOUT;
}
/* Enable auto-negotiation */
ret = tiva_phywrite(CONFIG_TIVA_PHYADDR, MII_MCR, MII_MCR_ANENABLE);
if (ret < 0)
{
ndbg("Failed to enable auto-negotiation: %d\n", ret);
return ret;
}
/* Wait until auto-negotiation completes */
for (timeout = 0; timeout < PHY_RETRY_TIMEOUT; timeout++)
{
ret = tiva_phyread(CONFIG_TIVA_PHYADDR, MII_MSR, &phyval);
if (ret < 0)
{
ndbg("Failed to read the PHY MSR: %d\n", ret);
return ret;
}
else if ((phyval & MII_MSR_ANEGCOMPLETE) != 0)
{
break;
}
}
if (timeout >= PHY_RETRY_TIMEOUT)
{
ndbg("Timed out waiting for auto-negotiation\n");
return -ETIMEDOUT;
}
/* Read the result of the auto-negotiation from the PHY-specific register */
ret = tiva_phyread(CONFIG_TIVA_PHYADDR, CONFIG_TIVA_PHYSR, &phyval);
if (ret < 0)
{
ndbg("Failed to read PHY status register\n");
return ret;
}
/* Remember the selected speed and duplex modes */
nvdbg("PHYSR[%d]: %04x\n", CONFIG_TIVA_PHYSR, phyval);
/* Different PHYs present speed and mode information in different ways. IF
* This CONFIG_TIVA_PHYSR_ALTCONFIG is selected, this indicates that the PHY
* represents speed and mode information are combined, for example, with
* separate bits for 10HD, 100HD, 10FD and 100FD.
*/
#ifdef CONFIG_TIVA_PHYSR_ALTCONFIG
switch (phyval & CONFIG_TIVA_PHYSR_ALTMODE)
{
default:
case CONFIG_TIVA_PHYSR_10HD:
priv->fduplex = 0;
priv->mbps100 = 0;
break;
case CONFIG_TIVA_PHYSR_100HD:
priv->fduplex = 0;
priv->mbps100 = 1;
break;
case CONFIG_TIVA_PHYSR_10FD:
priv->fduplex = 1;
priv->mbps100 = 0;
break;
case CONFIG_TIVA_PHYSR_100FD:
priv->fduplex = 1;
priv->mbps100 = 1;
break;
}
/* Different PHYs present speed and mode information in different ways. Some
* will present separate information for speed and mode (this is the default).
* Those PHYs, for example, may provide a 10/100 Mbps indication and a separate
* full/half duplex indication.
*/
#else
if ((phyval & CONFIG_TIVA_PHYSR_MODE) == CONFIG_TIVA_PHYSR_FULLDUPLEX)
{
priv->fduplex = 1;
}
if ((phyval & CONFIG_TIVA_PHYSR_SPEED) == CONFIG_TIVA_PHYSR_100MBPS)
{
priv->mbps100 = 1;
}
#endif
#else /* Auto-negotion not selected */
phyval = 0;
#ifdef CONFIG_TIVA_ETHFD
phyval |= MII_MCR_FULLDPLX;
#endif
#ifdef CONFIG_TIVA_ETH100MBPS
phyval |= MII_MCR_SPEED100;
#endif
ret = tiva_phywrite(CONFIG_TIVA_PHYADDR, MII_MCR, phyval);
if (ret < 0)
{
ndbg("Failed to write the PHY MCR: %d\n", ret);
return ret;
}
up_mdelay(PHY_CONFIG_DELAY);
/* Remember the selected speed and duplex modes */
#ifdef CONFIG_TIVA_ETHFD
priv->fduplex = 1;
#endif
#ifdef CONFIG_TIVA_ETH100MBPS
priv->mbps100 = 1;
#endif
#endif
ndbg("Duplex: %s Speed: %d MBps\n",
priv->fduplex ? "FULL" : "HALF",
priv->mbps100 ? 100 : 10);
return OK;
}
/****************************************************************************
* Function: tiva_phy_configure
*
* Description:
* Configure to support the selected PHY. Called after each reset since
* many properties of the PHY configuration are lost at each reset.
*
* Parameters:
* priv - A reference to the private driver state structure
*
* Returned Value:
* None.
*
* Assumptions:
*
****************************************************************************/
static void tiva_phy_configure(FAR struct tiva_ethmac_s *priv)
{
uint32_t regval;
/* Set up the PHY configuration */
#if defined(CONFIG_TIVA_PHY_RMII)
regval = EMAC_PC_PINTFS_RMII | EMAC_PC_PHYEXT;
#elif defined(CONFIG_TIVA_PHY_MII)
regval = EMAC_PC_PINTFS_MII | EMAC_PC_PHYEXT;
#else /* defined(CONFIG_TIVA_PHY_INTERNAL) */
regval = EMAC_PC_MDIXEN | EMAC_PC_ANMODE_100FD | EMAC_PC_ANEN |
EMAC_PC_PINTFS_MII;
#endif
tiva_putreg(regval, TIVA_EMAC_PC);
#ifdef CONFIG_TIVA_PHY_INTERNAL
/* If we are using the internal PHY, reset it to ensure that new
* configuration is latched.
*/
regval = tiva_getreg(TIVA_SYSCON_SREPHY);
regval |= SYSCON_SREPHY_R0;
tiva_putreg(regval, TIVA_SYSCON_SREPHY);
regval &= ~SYSCON_SREPHY_R0;
tiva_putreg(regval, TIVA_SYSCON_SREPHY);
/* Wait for the reset to complete */
while (!tiva_ephy_periphrdy());
up_udelay(250);
#endif
/* Disable all MMC interrupts as these are enabled by default at reset */
tiva_putreg(0xffffffff, TIVA_EMAC_MMCRXIM);
tiva_putreg(0xffffffff, TIVA_EMAC_MMCTXIM);
/* If using an external RMII PHY, we must enable the external clock */
regval = tiva_getreg(TIVA_EMAC_CC);
#if defined(CONFIG_TIVA_PHY_RMII)
/* Enable the external clock source input to the RMII interface signal
* EN0RREF_CLK by setting both the CLKEN bit in the Ethernet Clock
* Configuration (EMACCC) register. The external clock source must be
* 50 MHz with a frequency tolerance of 50 PPM.
*/
regval = tiva_getreg(TIVA_EMAC_CC);
#else
/* Disable the external clock */
regval &= ~EMAC_CC_CLKEN;
#endif
tiva_putreg(regval, TIVA_EMAC_CC);
}
/****************************************************************************
* Function: tiva_phy_initialize
*
* Description:
* Perform one-time PHY initialization
*
* Parameters:
* priv - A reference to the private driver state structure
*
* Returned Value:
* None.
*
* Assumptions:
*
****************************************************************************/
static inline void tiva_phy_initialize(FAR struct tiva_ethmac_s *priv)
{
/* Enable the clock to the PHY module */
nllvdbg("Enable EPHY clocking\n");
tiva_ephy_enableclk();
/* What until the PREPHY register indicates that the PHY is ready before
* continuing.
*/
while (!tiva_ephy_periphrdy());
up_udelay(250);
/* Enable power to the Ethernet PHY */
nllvdbg("Enable EPHY power\n");
tiva_ephy_enablepwr();
/* What until the PREPHY register indicates that the PHY registers are ready
* to be accessed.
*/
while (!tiva_ephy_periphrdy());
up_udelay(250);
nllvdbg("RCGCEPHY: %08x PCEPHY: %08x PREPHY: %08x\n",
getreg32(TIVA_SYSCON_RCGCEPHY),
getreg32(TIVA_SYSCON_PCEPHY),
getreg32(TIVA_SYSCON_PREPHY));
nllvdbg("Configure PHY GPIOs\n");
#ifdef CONFIG_TIVA_PHY_INTERNAL
/* Integrated PHY:
*
* "The Ethernet Controller Module and Integrated PHY receive two clock inputs:
* - A gated system clock acts as the clock source to the Control and Status
* registers (CSR) of the Ethernet MAC. The SYSCLK frequency for Run, Sleep
* and Deep Sleep mode is programmed in the System Control module. ...
* - The PHY receives the main oscillator (MOSC) which must be 25 MHz ± 50 ppm
* for proper operation. The MOSC source can be a single-ended source or a
* crystal."
*
* These are currently set up in tiva_clockconfig() before this function runs.
*
* MII/RMII Clocking:
*
* External PHY support is not yet implemented.
*/
/* PHY interface pins:
*
* EN0TXOP - Fixed pin assignment
* EN0TXON - Fixed pin assignment
* EN0RXIP - Fixed pin assignment
* EN0RXIN - Fixed pin assignment
* RBIAS - Fixed pin assignment
* EN0LED0 - Configured GPIO output
* EN0LED1 - Configured GPIO output
* EN0LED2 - Configured GPIO output
*/
tiva_configgpio(GPIO_EN0_LED0);
tiva_configgpio(GPIO_EN0_LED1);
tiva_configgpio(GPIO_EN0_LED2);
#else /* if defined(CONFIG_TIVA_PHY_MII) || defined(CONFIG_TIVA_PHY_RMII) */
/* External PHY interrupt pin */
tiva_configgpio(GPIO_EN0_INTRN);
/* Configure GPIO pins to support MII or RMII */
/* MDC and MDIO are common to both modes */
tiva_configgpio(GPIO_EN0_MDC);
tiva_configgpio(GPIO_EN0_MDIO);
#if defined(CONFIG_TIVA_PHY_MII)
/* Set up the MII interface */
/* "Four clock inputs are driven into the Ethernet MAC when the MII
* configuration is enabled. The clocks are described as follows:
*
* - Gated system clock (SYSCLK): The SYSCLK signal acts as the clock
* source to the Control and Status registers (CSR) of the Ethernet
* MAC. The SYSCLK frequency for Run, Sleep and Deep Sleep mode is
* programmed in the System Control module. ...
* - MOSC: A gated version of the MOSC clock is provided as the Precision
* Time Protocol (PTP) reference clock (PTPREF_CLK). The MOSC clock
* source can be a single-ended source on the OSC0 pin or a crystal
* on the OSC0 and OSC1 pins. When advanced timestamping is used and
* the Precision Timer Protocol (PTP) module has been enabled by setting
* the PTPCEN bit in the EMACCC register, the MOSC drives PTPREF_CLK.
* PTPREF_CLK has a minimum frequency requirement of 5 MHz and a
* maximum frequency of 25 MHz. ...
* - EN0RXCK: This clock signal is driven by the external PHY oscillator
* and is either 2.5 or 25 MHz depending on whether the device is
* operating at 10 Mbps or 100 Mbps.
* - EN0TXCK This clock signal is driven by the external PHY oscillator
* and is either 2.5 or 25 MHz depending on whether the device is
* operating at 10 Mbps or 100 Mbps."
*/
/* MII interface pins (17):
*
* MII_TX_CLK, MII_TXD[3:0], MII_TX_EN, MII_RX_CLK, MII_RXD[3:0], MII_RX_ER,
* MII_RX_DV, MII_CRS, MII_COL, MDC, MDIO
*/
tiva_configgpio(GPIO_EN0_MII_COL);
tiva_configgpio(GPIO_EN0_MII_CRS);
tiva_configgpio(GPIO_EN0_MII_RXD0);
tiva_configgpio(GPIO_EN0_MII_RXD1);
tiva_configgpio(GPIO_EN0_MII_RXD2);
tiva_configgpio(GPIO_EN0_MII_RXD3);
tiva_configgpio(GPIO_EN0_MII_RX_CLK);
tiva_configgpio(GPIO_EN0_MII_RX_DV);
tiva_configgpio(GPIO_EN0_MII_RX_ER);
tiva_configgpio(GPIO_EN0_MII_TXD0);
tiva_configgpio(GPIO_EN0_MII_TXD1);
tiva_configgpio(GPIO_EN0_MII_TXD2);
tiva_configgpio(GPIO_EN0_MII_TXD3);
tiva_configgpio(GPIO_EN0_MII_TX_CLK);
tiva_configgpio(GPIO_EN0_MII_TX_EN);
#elif defined(CONFIG_TIVA_PHY_RMII)
/* Set up the RMII interface. */
/* "There are three clock sources that interface to the Ethernet MAC in
* an RMII configuration:
*
* - Gated system clock (SYSCLK): The SYSCLK signal acts as the clock
* source to the Control and Status registers (CSR) of the Ethernet MAC.
* The SYSCLK frequency for Run, Sleep and Deep Sleep mode is programmed
* in the System Control module. ...
* - MOSC: A gated version of the MOSC clock is provided as the Precision
* Time Protocol (PTP) reference clock (PTPREF_CLK). The MOSC clock
* source can be a single-ended source on the OSC0 pin or a crystal on
* the OSC0 and OSC1 pins. When advanced timestamping is used and
* the PTP module has been enabled by setting the PTPCEN bit in the
* EMACCC register, the MOSC drives PTPREF_CLK. PTPREF_CLK has a minimum
* frequency requirement of 5 MHz and a maximum frequency of 25 MHz. ...
* - EN0REF_CLK: When using RMII, a 50 MHz external reference clock must
* drive the EN0REF_CLK input signal and the external PHY. Depending on
* the configuration of the FES bit in the Ethernet MAC Configuration
* (EMACCFG) register, the reference clock input (EN0REF_CLK) is divided
* by 20 for 10 Mbps or 2 for 100 Mbps operation and used as the clock
* for receive and transmit data."
*/
/* RMII interface pins (7):
*
* RMII_TXD[1:0], RMII_TX_EN, RMII_RXD[1:0], RMII_CRS_DV, MDC, MDIO,
* RMII_REF_CLK
*/
tiva_configgpio(GPIO_EN0_RMII_CRS_DV);
tiva_configgpio(GPIO_EN0_RMII_REF_CLK);
tiva_configgpio(GPIO_EN0_RMII_RXD0);
tiva_configgpio(GPIO_EN0_RMII_RXD1);
tiva_configgpio(GPIO_EN0_RMII_TXD0);
tiva_configgpio(GPIO_EN0_RMII_TXD1);
/* tiva_configgpio(GPIO_EN0_RMII_TX_CLK); not needed? */
tiva_configgpio(GPIO_EN0_RMII_TX_EN);
#endif
/* Enable pulse-per-second (PPS) output signal */
tiva_configgpio(GPIO_EN0_PPS);
#endif
}
/****************************************************************************
* Function: tiva_ethreset
*
* Description:
* Reset the Ethernet block.
*
* Parameters:
* priv - A reference to the private driver state structure
*
* Returned Value:
* None.
*
* Assumptions:
*
****************************************************************************/
static void tiva_ethreset(FAR struct tiva_ethmac_s *priv)
{
uint32_t regval;
#if 0 /* REVISIT: This causes the DMABUSMOD reset to hang. */
/* Reset the Ethernet MAC */
regval = tiva_getreg(TIVA_SYSCON_SREMAC);
regval |= SYSCON_SREMAC_R0;
tiva_putreg(regval, TIVA_SYSCON_SREMAC);
regval &= ~SYSCON_SREMAC_R0;
tiva_putreg(regval, TIVA_SYSCON_SREMAC);
/* Wait for the reset to complete */
while (!tiva_emac_periphrdy());
up_udelay(250);
#endif
/* Perform a software reset by setting the SWR bit in the DMABUSMOD register.
* This Resets all MAC subsystem internal registers and logic. After this
* reset all the registers holds their reset values.
*/
regval = tiva_getreg(TIVA_EMAC_DMABUSMOD);
regval |= EMAC_DMABUSMOD_SWR;
tiva_putreg(regval, TIVA_EMAC_DMABUSMOD);
/* Wait for software reset to complete. The SWR bit is cleared automatically
* after the reset operation has completed in all of the core clock domains.
*/
while ((tiva_getreg(TIVA_EMAC_DMABUSMOD) & EMAC_DMABUSMOD_SWR) != 0);
up_udelay(250);
/* Reconfigure the PHY. Some PHY configurations will be lost as a
* consequence of the EMAC reset
*/
tiva_phy_configure(priv);
}
/****************************************************************************
* Function: tiva_macconfig
*
* Description:
* Configure the Ethernet MAC for DMA operation.
*
* Parameters:
* priv - A reference to the private driver state structure
*
* Returned Value:
* OK on success; Negated errno on failure.
*
* Assumptions:
*
****************************************************************************/
static int tiva_macconfig(FAR struct tiva_ethmac_s *priv)
{
uint32_t regval;
/* Set up the MACCR register */
regval = tiva_getreg(TIVA_EMAC_CFG);
regval &= ~MACCR_CLEAR_BITS;
regval |= MACCR_SET_BITS;
if (priv->fduplex)
{
/* Set the DM bit for full duplex support */
regval |= EMAC_CFG_DUPM;
}
if (priv->mbps100)
{
/* Set the FES bit for 100Mbps fast Ethernet support */
regval |= EMAC_CFG_FES;
}
tiva_putreg(regval, TIVA_EMAC_CFG);
/* Set up the FRAMEFLTR register */
regval = tiva_getreg(TIVA_EMAC_FRAMEFLTR);
regval &= ~FRAMEFLTR_CLEAR_BITS;
regval |= FRAMEFLTR_SET_BITS;
tiva_putreg(regval, TIVA_EMAC_FRAMEFLTR);
/* Set up the HASHTBLH and HASHTBLL registers */
tiva_putreg(0, TIVA_EMAC_HASHTBLH);
tiva_putreg(0, TIVA_EMAC_HASHTBLL);
/* Setup up the FLOWCTL register */
regval = tiva_getreg(TIVA_EMAC_FLOWCTL);
regval &= ~FLOWCTL_CLEAR_MASK;
regval |= FLOWCTL_SET_MASK;
tiva_putreg(regval, TIVA_EMAC_FLOWCTL);
/* Setup up the VLANTG register */
tiva_putreg(0, TIVA_EMAC_VLANTG);
/* DMA Configuration */
/* Set up the DMAOPMODE register */
regval = tiva_getreg(TIVA_EMAC_DMAOPMODE);
regval &= ~DMAOPMODE_CLEAR_MASK;
regval |= DMAOPMODE_SET_MASK;
tiva_putreg(regval, TIVA_EMAC_DMAOPMODE);
/* Set up the DMABUSMOD register */
regval = tiva_getreg(TIVA_EMAC_DMABUSMOD);
regval &= ~DMABUSMOD_CLEAR_MASK;
regval |= DMABUSMOD_SET_MASK;
tiva_putreg(regval, TIVA_EMAC_DMABUSMOD);
return OK;
}
/****************************************************************************
* Function: tiva_macaddress
*
* Description:
* Configure the selected MAC address.
*
* Parameters:
* priv - A reference to the private driver state structure
*
* Returned Value:
* OK on success; Negated errno on failure.
*
* Assumptions:
*
****************************************************************************/
static void tiva_macaddress(FAR struct tiva_ethmac_s *priv)
{
FAR struct net_driver_s *dev = &priv->dev;
uint32_t regval;
nvdbg("%s MAC: %02x:%02x:%02x:%02x:%02x:%02x\n",
dev->d_ifname,
dev->d_mac.ether_addr_octet[0], dev->d_mac.ether_addr_octet[1],
dev->d_mac.ether_addr_octet[2], dev->d_mac.ether_addr_octet[3],
dev->d_mac.ether_addr_octet[4], dev->d_mac.ether_addr_octet[5]);
/* Set the MAC address high register */
regval = ((uint32_t)dev->d_mac.ether_addr_octet[5] << 8) |
(uint32_t)dev->d_mac.ether_addr_octet[4];
tiva_putreg(regval, TIVA_EMAC_ADDR0H);
/* Set the MAC address low register */
regval = ((uint32_t)dev->d_mac.ether_addr_octet[3] << 24) |
((uint32_t)dev->d_mac.ether_addr_octet[2] << 16) |
((uint32_t)dev->d_mac.ether_addr_octet[1] << 8) |
(uint32_t)dev->d_mac.ether_addr_octet[0];
tiva_putreg(regval, TIVA_EMAC_ADDR0L);
}
/****************************************************************************
* Function: tiva_ipv6multicast
*
* Description:
* Configure the IPv6 multicast MAC address.
*
* Parameters:
* priv - A reference to the private driver state structure
*
* Returned Value:
* OK on success; Negated errno on failure.
*
* Assumptions:
*
****************************************************************************/
#ifdef CONFIG_NET_ICMPv6
static void tiva_ipv6multicast(FAR struct tiva_ethmac_s *priv)
{
struct net_driver_s *dev;
uint16_t tmp16;
uint8_t mac[6];
/* For ICMPv6, we need to add the IPv6 multicast address
*
* For IPv6 multicast addresses, the Ethernet MAC is derived by
* the four low-order octets OR'ed with the MAC 33:33:00:00:00:00,
* so for example the IPv6 address FF02:DEAD:BEEF::1:3 would map
* to the Ethernet MAC address 33:33:00:01:00:03.
*
* NOTES: This appears correct for the ICMPv6 Router Solicitation
* Message, but the ICMPv6 Neighbor Solicitation message seems to
* use 33:33:ff:01:00:03.
*/
mac[0] = 0x33;
mac[1] = 0x33;
dev = &priv->dev;
tmp16 = dev->d_ipv6addr[6];
mac[2] = 0xff;
mac[3] = tmp16 >> 8;
tmp16 = dev->d_ipv6addr[7];
mac[4] = tmp16 & 0xff;
mac[5] = tmp16 >> 8;
nvdbg("IPv6 Multicast: %02x:%02x:%02x:%02x:%02x:%02x\n",
mac[0], mac[1], mac[2], mac[3], mac[4], mac[5]);
(void)tiva_addmac(dev, mac);
#ifdef CONFIG_NET_ICMPv6_AUTOCONF
/* Add the IPv6 all link-local nodes Ethernet address. This is the
* address that we expect to receive ICMPv6 Router Advertisement
* packets.
*/
(void)tiva_addmac(dev, g_ipv6_ethallnodes.ether_addr_octet);
#endif /* CONFIG_NET_ICMPv6_AUTOCONF */
#ifdef CONFIG_NET_ICMPv6_ROUTER
/* Add the IPv6 all link-local routers Ethernet address. This is the
* address that we expect to receive ICMPv6 Router Solicitation
* packets.
*/
(void)tiva_addmac(dev, g_ipv6_ethallrouters.ether_addr_octet);
#endif /* CONFIG_NET_ICMPv6_ROUTER */
}
#endif /* CONFIG_NET_ICMPv6 */
/****************************************************************************
* Function: tiva_macenable
*
* Description:
* Enable normal MAC operation.
*
* Parameters:
* priv - A reference to the private driver state structure
*
* Returned Value:
* OK on success; Negated errno on failure.
*
* Assumptions:
*
****************************************************************************/
static int tiva_macenable(FAR struct tiva_ethmac_s *priv)
{
uint32_t regval;
/* Set the MAC address */
tiva_macaddress(priv);
#ifdef CONFIG_NET_ICMPv6
/* Set up the IPv6 multicast address */
tiva_ipv6multicast(priv);
#endif
/* Enable transmit state machine of the MAC for transmission on the MII */
regval = tiva_getreg(TIVA_EMAC_CFG);
regval |= EMAC_CFG_TE;
tiva_putreg(regval, TIVA_EMAC_CFG);
/* Flush Transmit FIFO */
regval = tiva_getreg(TIVA_EMAC_DMAOPMODE);
regval |= EMAC_DMAOPMODE_FTF;
tiva_putreg(regval, TIVA_EMAC_DMAOPMODE);
/* Enable receive state machine of the MAC for reception from the MII */
/* Enables or disables the MAC reception. */
regval = tiva_getreg(TIVA_EMAC_CFG);
regval |= EMAC_CFG_RE;
tiva_putreg(regval, TIVA_EMAC_CFG);
/* Start DMA transmission */
regval = tiva_getreg(TIVA_EMAC_DMAOPMODE);
regval |= EMAC_DMAOPMODE_ST;
tiva_putreg(regval, TIVA_EMAC_DMAOPMODE);
/* Start DMA reception */
regval = tiva_getreg(TIVA_EMAC_DMAOPMODE);
regval |= EMAC_DMAOPMODE_SR;
tiva_putreg(regval, TIVA_EMAC_DMAOPMODE);
/* Enable Ethernet DMA interrupts. */
tiva_putreg(EMAC_IM_ALLINTS, TIVA_EMAC_IM);
/* Ethernet DMA supports two classes of interrupts: Normal interrupt
* summary (NIS) and Abnormal interrupt summary (AIS) with a variety
* individual normal and abnormal interrupting events. Here only
* the normal receive event is enabled (unless DEBUG is enabled). Transmit
* events will only be enabled when a transmit interrupt is expected.
*/
tiva_putreg((EMAC_DMAINT_RECV_ENABLE | EMAC_DMAINT_ERROR_ENABLE),
TIVA_EMAC_DMAIM);
return OK;
}
/****************************************************************************
* Function: tive_emac_configure
*
* Description:
* Configure the Ethernet interface for DMA operation.
*
* Parameters:
* priv - A reference to the private driver state structure
*
* Returned Value:
* OK on success; Negated errno on failure.
*
* Assumptions:
*
****************************************************************************/
static int tive_emac_configure(FAR struct tiva_ethmac_s *priv)
{
int ret;
/* NOTE: The Ethernet clocks were initialized earlier in the start-up
* sequence.
*/
/* Reset the Ethernet block */
nvdbg("Reset the Ethernet block\n");
tiva_ethreset(priv);
/* Initialize the PHY */
nvdbg("Initialize the PHY\n");
ret = tiva_phyinit(priv);
if (ret < 0)
{
return ret;
}
/* Initialize the MAC and DMA */
nvdbg("Initialize the MAC and DMA\n");
ret = tiva_macconfig(priv);
if (ret < 0)
{
return ret;
}
/* Initialize the free buffer list */
tiva_initbuffer(priv);
/* Initialize TX Descriptors list: Chain Mode */
tiva_txdescinit(priv);
/* Initialize RX Descriptors list: Chain Mode */
tiva_rxdescinit(priv);
/* Enable normal MAC operation */
nvdbg("Enable normal operation\n");
return tiva_macenable(priv);
}
/****************************************************************************
* Public Functions
****************************************************************************/
/****************************************************************************
* Function: tiva_ethinitialize
*
* Description:
* Initialize the Ethernet driver for one interface. If the Tiva chip
* supports multiple Ethernet controllers, then board specific logic
* must implement up_netinitialize() and call this function to initialize
* the desired interfaces.
*
* Parameters:
* intf - In the case where there are multiple EMACs, this value
* identifies which EMAC is to be initialized.
*
* Returned Value:
* OK on success; Negated errno on failure.
*
* Assumptions:
*
****************************************************************************/
#if TIVA_NETHCONTROLLERS == 1
static inline
#endif
int tiva_ethinitialize(int intf)
{
struct tiva_ethmac_s *priv;
uint32_t regval;
nllvdbg("intf: %d\n", intf);
/* Get the interface structure associated with this interface number. */
DEBUGASSERT(intf < TIVA_NETHCONTROLLERS);
priv = &g_tiva_ethmac[intf];
/* Initialize the driver structure */
memset(priv, 0, sizeof(struct tiva_ethmac_s));
priv->dev.d_ifup = tiva_ifup; /* I/F up (new IP address) callback */
priv->dev.d_ifdown = tiva_ifdown; /* I/F down callback */
priv->dev.d_txavail = tiva_txavail; /* New TX data callback */
#ifdef CONFIG_NET_IGMP
priv->dev.d_addmac = tiva_addmac; /* Add multicast MAC address */
priv->dev.d_rmmac = tiva_rmmac; /* Remove multicast MAC address */
#endif
#ifdef CONFIG_NETDEV_PHY_IOCTL
priv->dev.d_ioctl = tiva_ioctl; /* Support PHY ioctl() calls */
#endif
priv->dev.d_private = (void*)g_tiva_ethmac; /* Used to recover private state from dev */
/* Create a watchdog for timing polling for and timing of transmissions */
priv->txpoll = wd_create(); /* Create periodic poll timer */
priv->txtimeout = wd_create(); /* Create TX timeout timer */
#ifdef CONFIG_TIVA_BOARDMAC
/* If the board can provide us with a MAC address, get the address
* from the board now. The MAC will not be applied until tiva_ifup()
* is called (and the MAC can be overwritten with a netdev ioctl call).
*/
tiva_ethernetmac(&priv->dev.d_mac);
#endif
/* Enable power and clocking to the Ethernet MAC
*
* - Enable Power: Applies power (only) to the EMAC peripheral. This is not
* an essential step since enabling clocking will also apply power. The
* only significance is that the EMAC state will be retained if the EMAC
* clocking is subsequently disabled.
* - Enable Clocking: Applies both power and clocking to the EMAC peripheral,
* bringing it a fully functional state.
*/
nllvdbg("Enable EMAC clocking\n");
tiva_emac_enablepwr(); /* Ethernet MAC Power Control */
tiva_emac_enableclk(); /* Ethernet MAC Run Mode Clock Gating Control */
/* What until the PREMAC register indicates that the EMAC registers are ready
* to be accessed.
*/
while (!tiva_emac_periphrdy());
up_udelay(250);
/* Show all EMAC clocks */
nllvdbg("RCGCEMAC: %08x PCEMAC: %08x PREMAC: %08x MOSCCTL: %08x\n",
getreg32(TIVA_SYSCON_RCGCEMAC),
getreg32(TIVA_SYSCON_PCEMAC),
getreg32(TIVA_SYSCON_PREMAC),
getreg32(TIVA_SYSCON_MOSCCTL));
/* Configure clocking and GPIOs to support the internal/eternal PHY */
tiva_phy_initialize(priv);
/* Attach the IRQ to the driver */
if (irq_attach(TIVA_IRQ_ETHCON, tiva_interrupt))
{
/* We could not attach the ISR to the interrupt */
return -EAGAIN;
}
/* Wait for EMAC to come out of reset. The SWR bit is cleared automatically
* after the reset operation has completed in all of the core clock domains.
*/
while ((tiva_getreg(TIVA_EMAC_DMABUSMOD) & EMAC_DMABUSMOD_SWR) != 0);
up_udelay(250);
#if 0 /* REVISIT: Part of work around to avoid DMABUSMOD SWR hangs */
/* Put the interface in the down state. */
tiva_ifdown(&priv->dev);
#else
/* Reset the Ethernet MAC */
regval = tiva_getreg(TIVA_SYSCON_SREMAC);
regval |= SYSCON_SREMAC_R0;
tiva_putreg(regval, TIVA_SYSCON_SREMAC);
regval &= ~SYSCON_SREMAC_R0;
tiva_putreg(regval, TIVA_SYSCON_SREMAC);
/* Wait for the reset to complete */
while (!tiva_emac_periphrdy());
up_udelay(250);
#endif
/* Register the device with the OS so that socket IOCTLs can be performed */
nllvdbg("Registering Ethernet device\n");
return netdev_register(&priv->dev, NET_LL_ETHERNET);
}
/****************************************************************************
* Function: up_netinitialize
*
* Description:
* This is the "standard" network initialization logic called from the
* low-level initialization logic in up_initialize.c. If TIVA_NETHCONTROLLERS
* greater than one, then board specific logic will have to supply a
* version of up_netinitialize() that calls tiva_ethinitialize() with
* the appropriate interface number.
*
* Parameters:
* None.
*
* Returned Value:
* None.
*
* Assumptions:
*
****************************************************************************/
#if TIVA_NETHCONTROLLERS == 1
void up_netinitialize(void)
{
(void)tiva_ethinitialize(0);
}
#endif
/****************************************************************************
* Name: arch_phy_irq
*
* Description:
* This function may be called to register an interrupt handler that will
* be called when a PHY interrupt occurs. This function both attaches
* the interrupt handler and enables the interrupt if 'handler' is non-
* NULL. If handler is NULL, then the interrupt is detached and disabled
* instead.
*
* The PHY interrupt is always disabled upon return. The caller must
* call back through the enable function point to control the state of
* the interrupt.
*
* This interrupt may or may not be available on a given platform depending
* on how the network hardware architecture is implemented. In a typical
* case, the PHY interrupt is provided to board-level logic as a GPIO
* interrupt (in which case this is a board-specific interface and really
* should be called board_phy_irq()); In other cases, the PHY interrupt
* may be cause by the chip's MAC logic (in which case arch_phy_irq()) is
* an appropriate name. Other other boards, there may be no PHY interrupts
* available at all. If client attachable PHY interrupts are available
* from the board or from the chip, then CONFIG_ARCH_PHY_INTERRUPT should
* be defined to indicate that fact.
*
* Typical usage:
* a. OS service logic (not application logic*) attaches to the PHY
* PHY interrupt and enables the PHY interrupt.
* b. When the PHY interrupt occurs: (1) the interrupt should be
* disabled and () work should be scheduled on the worker thread (or
* perhaps a dedicated application thread).
* c. That worker thread should use the SIOCGMIIPHY, SIOCGMIIREG,
* and SIOCSMIIREG ioctl calls** to communicate with the PHY,
* determine what network event took place (Link Up/Down?), and
* take the appropriate actions.
* d. It should then interact the the PHY to clear any pending
* interrupts, then re-enable the PHY interrupt.
*
* * This is an OS internal interface and should not be used from
* application space. Rather applications should use the SIOCMIISIG
* ioctl to receive a signal when a PHY event occurs.
* ** This interrupt is really of no use if the Ethernet MAC driver
* does not support these ioctl calls.
*
* Input Parameters:
* intf - Identifies the network interface. For example "eth0". Only
* useful on platforms that support multiple Ethernet interfaces
* and, hence, multiple PHYs and PHY interrupts.
* handler - The client interrupt handler to be invoked when the PHY
* asserts an interrupt. Must reside in OS space, but can
* signal tasks in user space. A value of NULL can be passed
* in order to detach and disable the PHY interrupt.
* enable - A function pointer that be unsed to enable or disable the
* PHY interrupt.
*
* Returned Value:
* The previous PHY interrupt handler address is returned. This allows you
* to temporarily replace an interrupt handler, then restore the original
* interrupt handler. NULL is returned if there is was not handler in
* place when the call was made.
*
****************************************************************************/
#ifdef CONFIG_TIVA_PHY_INTERRUPTS
xcpt_t arch_phy_irq(FAR const char *intf, xcpt_t handler, phy_enable_t *enable)
{
struct tiva_ethmac_s *priv;
irqstate_t flags;
xcpt_t oldhandler;
DEBUGASSERT(intf);
nvdbg("%s: handler=%p\n", intf, handler);
/* Get the interface structure associated with this interface. */
#if TIVA_NETHCONTROLLERS > 1
/* REVISIT: Additional logic needed if there are multiple EMACs */
warning Missing logic
#endif
priv = g_tiva_ethmac;
/* Disable interrupts until we are done. This guarantees that the
* following operations are atomic.
*/
flags = irqsave();
/* Get the old interrupt handler and save the new one */
oldhandler = priv->handler;
priv->handler = handler;
/* Return with the interrupt disabled in any case */
tiva_phy_intenable(false);
/* Return the enabling function pointer */
if (enable)
{
*enable = handler ? tiva_phy_intenable : NULL;;
}
/* Return the old handler (so that it can be restored) */
irqrestore(flags);
return oldhandler;
}
#endif /* CONFIG_TIVA_PHY_INTERRUPTS */
#endif /* TIVA_NETHCONTROLLERS > 0 */
#endif /* CONFIG_NET && CONFIG_TIVA_ETHERNET */