vxworks程序示例和源码解释


\| 网站首页 \| 新闻 \| SOPC \| FPGA \| DSP \| ARM \| 嵌入式操作系统 \| 下载 \| 所有产品 \| 留言 \| 论坛 \| 购买指南 \| 网络协议 \| 驱动设计 \|

您现在的位置： 21嵌入式控制研究室 >> 嵌入式操作系统 >> Vxworks >> 文章正文

用户登录

新用户注册

vxworks程序示例和源码解释

荐 ★★★

【字体：小大】

vxworks程序示例和源码解释

作者：佚名文章来源：21control 点击数：更新时间：2005-12-22

这里是一些程序示例和源码解释.一些相关程序代码会陆续补上. 系统初始化 系统初始化对不同的CPU,基本步骤是类似的. 系统初始化的主要步骤如下 启动 关闭中断 放boot type到堆栈 清空缓存 VxWorks 系统的 PowerPC BSP,系统开机后执行的第一个函数 romInit(),在ROM的起点,这里是使用的PowerPC汇编语言 /* 定义内部函数 internals */ .globl romInit /* start of system code */ .globl _romInit /* start of system code */ /* 定义外部函数 externals */ .extern romStart /* system initialization routine */ .text .align 2 /******************************************************************************* * * romInit ( int startType /@ only used by 2nd entry point @/ ) */ romInit: _romInit: bl cold /* 冷启动 */ bl warm /* 热启动 */ cold: li p5, BOOT_COLD bl start /* skip over next instruction */ warm: or p5, p0, p0 /* startType to p5 */ start: /* 此处是系统启动开始 */ /* 屏蔽MSR中CE,EE位,关闭所有的外部中断 /* * Disable external interrupts */ mfmsr p0 /* p0 = msr */ INT_MASK (p0, p1) /* mask EE and CE bit */ ori p1,p1,_PPC_MSR_ME /* enable machine checks */ mtmsr p1 /* msr = p1 */ isync /* 下面两步是按照硬件定义初始化一些SPR,DCR寄存器,置0或置1 /* SPR是特殊功能寄存器,DCR为设备控制寄存器,还有MSR机器状态寄存器,这些是PowerPC内核中很重要的寄存器 /* 初始化SPR,DCR寄存器置0 * Initalize registers that need to be set to zero. */ addi r4,r0,0x0000 mtspr SGR,r4 /* 解锁所有存储区域 SPR 中 SGR 位置0 */ mtspr ESR, r4 /* SPR中的错误状态位 ESR 清0 */ mtspr TCR, r4 /* 关闭所有的 timers */ mtspr PIT, r4 /* 清0 PIT timer */ mtdcr UICER, r4 /* 关闭中断控制器(UIC)中的所有中断 */ mtspr XER, r4 /* 清0 integer exception 寄存器 */ /* 初始化另一些SPR,DCR寄存器置1 * Initalize registers that need to be cleared with 0xFFFFFFFF. */ addis r4,r0,0xffff ori r4,r4,0xffff mtspr TSR, r4 /* timer */ mtspr DBSR, r4 /* 调试状态位置1 */ mtdcr UICSR, r4 /* 清除中断控制器(UIC)中的所有 pending 中断 */ mtdcr dmasr, r4 /* DMA状态寄存器置1 */ /* PowerPC405用两个缓存,一个是16K指令缓存(ICU),一个是6K数据缓存(DCU),下面是清空着两个缓存,并根据硬件设置缓存 */ /* 清空指令缓存 */ /*BESR type regs ZZZZZZZZZZZZ * Invalidate the entire instruction cache. This can be done * with a single iccci instruction in the processor core. */ iccci r0, r0 /*清空数据缓存 * Invalidate the entire data cache. * The 405 processor core in the 405GP has 128 congruence classes. * Each cache line in the 405 processor is 32 bytes. */ /* * Turn the instruction cache on for faster boot-up. * Also, the icache is needed to help initialize Bank 0 * of the EBC to speed up accesses to flash. * address space 0x00000000-0x07ffffff is cached * address space 0xf8000000-0xffffffff is cached */ lis p0, HIADJ(_PPC403_ICCR_DEFAULT_VAL) addi p0, p0, LO(_PPC403_ICCR_DEFAULT_VAL) mtspr _PPC403_ICCR, p0 isync /* 初始化外部总线控制器(EBC),跳转指令BL,执行extBusCntlrInit * / bl extBusCntlrInit /* * Now that the EBC Bank 0 has been set up, turn the I-cache off if * i-cache was not specified in config.h. It is also invalidated * again. */ #ifndef USER_I_CACHE_ENABLE li p0, 0 /* clear p0 */ mtspr _PPC403_ICCR, p0 /* turn off i-cache */ isync iccci r0, r0 /* invalidate the I-cache again */ #endif /* 初始化和SDRAM相关的IIC(inter-integrated circut)寄存器IIC0, * Initialize IIC0 for use in automatic SDRAM configuration */ #ifdef LOCAL_MEM_AUTOSIZE bl iic0Init #endif /* 初始化SDRAM,BL跳转执行sdramInit * Configure the SDRAM controller only if this is a cold boot. * If the SDRAM controller is reinitialized on a warm boot, the * boot line will get wiped out because of the ECC SDRAM memory * initialization. */ li p0, BOOT_COLD and. p0, p0, p5 /* p5 is saved at the entry of romInit */ beq skip bl sdramInit skip: /* * Clear the CPU reservation bit */ li r0, 0 lwarx p0, r0, r0 stwcx. p0, r0, r0 #ifdef PPC405GP_REVA /* 设置中断向量表到0x0000 */ li p0, 0x2100/4 mtctr p0 lis p0, WALNUT_EVPR_VAL li p1, 0x0000 zeroOut: stw p1,0x0(p0) addi p0, p0, 4 bdnz zeroOut #endif /* 初始化堆栈 /* Initialize the stack pointer (r1) */ lis sp, HIADJ(STACK_ADRS) addi sp, sp, LO(STACK_ADRS) #if FALSE /* SDA not supported */ /* initialize r2 and r13 according to EABI standard */ lis r2, HIADJ(_SDA2_BASE_) addi r2, r2, LO(_SDA2_BASE_) lis r13, HIADJ(_SDA_BASE_) addi r13, r13, LO(_SDA_BASE_) #endif /* 得到C程序romStart()在ROM中的地址,保证romInit执行结束后,系统跳转执行romStart() /* calculate C entry point: routine - entry point + ROM base */ lis p1, HIADJ(romStart) /* p1 = romstart */ addi p1, p1, LO(romStart) lis p2, HIADJ(romInit) /* p2 = romInit */ addi p2, p2, LO(romInit) /* ROM_TEXT_ADRS为ROM的入口地址,在文件makefile定义,为0xfff80100 lis p3, HIADJ(ROM_TEXT_ADRS) /* p3 = ROM_TEXT_ADRS */ addi p3, p3, LO(ROM_TEXT_ADRS) subf p1, p2, p1 /* p1 = p1 - p2 */ add p1, p1, p3 /* p1 = p1 + p3 */ /* p1中是romStart()的地址,这里把这个地址放到连接寄存器LR中. mtlr p1 /* link register = C entry point */ or p0, p5, p5 /* p0 = startType */ addi sp, sp, -FRAMEBASESZ /* get frame stack */ /* 跳转到LR中romStart()的地址,执行romStart() blr /* branch to link register */ 硬件中断 中断的产生和VxWorks系统的中断操作: 一般中断的产生是由硬件定义的,如串口中断的定义: 1.接收中断:当接收中断使能,接收数据存储器 RxData 存在有效数据,则产生中断. 2.发送中断:当发送中断使能,发送数据存储器 TxData 为空,则产生中断. 硬件发送中断产生逻辑示意 硬件接收中断产生逻辑示意 　 所以要产生一个串口中断,主要有两步: 1.使能这两个串口中断,RX Enable,TX Enable,函数intEnable(). 2.用intConnect()登记中断号,和相应的中断例程ISR. 程序示例 在VxWorks系统上登记,使能串口中断 intConnect((VOIDFUNCPTR *)5,ComISR,0); //登记中断服务程序ComISR()到外部中断号5, intEnable((VOIDFUNCPTR *)5); //使能外部中断5 //enable 使能 UART1 这里直接用32位地址表示了 *(unsigned long *)0x20000014 &= ~0x01000100; //INT_FORCE use pending *(unsigned long *)0x20000010 = 0x01000100; //INT_PENDING clear *(unsigned long *)0x20000018 = 0x01000100; //INT_MASK only enable UART1 RX //control register 使能 RX TX *(unsigned long *)0x20000068 = 0x00070007; //RX and TX ENABLE //divider register,baut rate 19200 设置波特率 *(unsigned long *)0x2000006c = 59; //完成, 这样当串口有中断发生时,硬件系统会自动跳转到中断号5的地址0x18,调用程序ComISP(). 注意:中断程序不能单步执行,或跟踪调试,中断服务程序中与函数库或系统有关的函数不可用如print()等,因为中断调用时,所有其它的任务都被挂起停止运行. VxWorks系统的网络驱动(END) VxWorks网络配置参见VxWorks网络驱动配置及分析 VxWorks系统网络驱动在BSP中完成,写驱动时应参考BSP develop kit,在VxWorks中叫做END( Enhanced Network Driver),编写程序使用由VxWorks定义的MUX接口 MUX是数据链路层和网络协议层之间的接口 主要调用过程和步骤如下: VxWorks系统执行的第一个任务target\config\all\usrConfig.c文件中 usrRoot()=======>> target\src\config\usrNetwork.c文件(该文件初始化TCP/IP)中 usrNetInit(BOOT_LINE_ADRS)(该函数作用是添加MUX END)========>> pcooki = pCookie = muxDevLoad(pDevTbl->unit,.....)其中pDevTbl在BSP网络配置文件configNet.h中定义.END_TBL_ENTRY endDevTbl[]={...},该表定义了网络设备的具体参数,在这里调用了网络驱动 END_TBL_ENTRY endDevTbl [] = { {0, IBM_EMAC_LOAD_FUNC, IBM_EMAC_LOAD_STR_0, TRUE, NULL, FALSE}, {0, END_TBL_END, NULL, 0, NULL, FALSE}, }; 其中IBM_EMAC_LOAD_FUNC就是 ibmEmacEndLoad() ========>>muxDevStart(pcooki)==========>>ibmEmacEndLoad() 　 ibmEmacEndLoad()初始化系统为网络驱动运行做准备 MUX调用ibmEmacStart() ibmEmacStart() 登记中断服务程序ibmEmacInit(),启动设备运行在中断模式下. LOCAL STATUS ibmEmacStart ( EMAC_DRV_CTRL * pDrvCtrl ) { int rc; SYS_INT_CONNECT (pDrvCtrl, ibmEmacInt, pDrvCtrl, &rc); SYS_OUT_LONG(pDrvCtrl, EMAC_ISR, 0xFFFFFFFF); SYS_INT_ENABLE (); /* Allow MAL EOB and Descriptor error interrupts */ malChannelIntMaskSet(MAL_TX_TYPE, pDrvCtrl->txChn0MalChannel, MAL_EOB_INT_EN | MAL_DE_INT_EN | MAL_SERR_INT_EN); malChannelIntMaskSet(MAL_RX_TYPE, pDrvCtrl->rxChn0MalChannel, MAL_EOB_INT_EN | MAL_DE_INT_EN | MAL_SERR_INT_EN); return (OK); } 中断服务程序ibmEmacInit() 处理EMAC控制器的中断,主要是 TX,RX状态错误 LOCAL void ibmEmacInt ( EMAC_DRV_CTRL * pDrvCtrl ) { UINT isrReg; /* Read the EMAC interrupt status register */ SYS_IN_LONG(pDrvCtrl, EMAC_ISR, isrReg); pDrvCtrl->errorEmac = isrReg; /* * Check to see if there was a TX error. If there was, the Dead bit * will be set. Clear the status bits for the TX error, and clear the dead * bit. Keep count of these errors in the main device structure. */ if (isrReg & EMAC_ISR_TX_INTS) { pDrvCtrl->intErrorTX++; SYS_OUT_LONG(pDrvCtrl, EMAC_ISR, EMAC_ISR_TX_INTS); } /* * Check to see if there was a RX error. Clear the status bits for the RX * error. Keep count of these errors in the main device structure. */ if (isrReg & EMAC_ISR_RX_INTS) { pDrvCtrl->intErrorRX++; SYS_OUT_LONG(pDrvCtrl, EMAC_ISR, EMAC_ISR_RX_INTS); } return; } (未完) 　 Cillus网卡CS8900A Linux驱动 下面是我为一个网友解释的CS8900A网卡驱动文件中的部分函数,操作系统为ucLinux,CPU是国内常用的Motorola龙珠系列MC68EZ328(16M),相比PowerPC和ARM来说,它的结构简单,不带MMU,较易理解 CS8900A是一个16位网卡,支持ISA总线,10-BastT. static inline void outw(unsigned short value,unsigned int addr) { unsigned short newvalue; unsigned char *_src=(unsigned char*)(&value),*_dest=(unsigned char*)(&newvalue); //这里为什么要交换一下? ////++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ //这里不是交换,而是赋值,将16位分成两个8位的数组,分别赋值,16位寄存器是只有后8位可读写, //前8位是寄存器ID,所以这里吧后8位放在前面"_dest[0]=_src[1];" //然后通过I/O口,赋给相应的存储器,即"*(volatile unsigned short*)addr=newvalue;" ///+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ _dest[0]=_src[1]; _dest[1]=_src[0]; *(volatile unsigned short*)addr=newvalue; } static int cs89x0_probe1(struct device *dev, int ioaddr) { struct net_local *lp; static unsigned version_printed = 0; int i; unsigned rev_type = 0; irq2dev_map[0] = dev; /* set up the chip select */ //下面这一段该怎么理解?看不懂阿 ////++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ //这里是对CPU进行操作,对cs89进行初始化,我没有见到你的硬件原理图,我只能按CPU //的结构定义来解释一下,你要想搞清楚,只有看硬件原理图,还要参考CPU的硬件手册 ////++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ //连通PORT F I/O功能管脚1,SEL1(-IRQ5/PF1), *(volatile unsigned char *)0xfffff42b |= 0x02; /* output /sleep */ //设置PORT F 管脚0为输出,输出高电平 *(volatile unsigned short *)0xfffff428 |= 0x0101; /* not sleeping */ //连通PORT F 中断功能管脚1,SEL1(-IRQ5/PF1) *(volatile unsigned char *)0xfffff42b &= ~0x02; /* input irq5 */ //PORT F 管脚1,SEL1(-IRQ5/PF1) 为输入方式,置低电平有效 *(volatile unsigned short *)0xfffff428 &= ~0x0202; /* irq5 fcn on */ //在寄存器CSGBB中定义片选及cs89的基地址为:0x10000000.++++非常重要+++++++ *(volatile unsigned short *)0xfffff102 = 0x8000; /* 0x04000000 */ //在寄存器CSB中,片选使能,cs89地址空间大小(1M),数据线宽16位,6个等待周期,FLASH..., //非保护存储空间(128K)...... *(volatile unsigned short *)0xfffff112 = 0x01e7; /* 128k, 2ws, FLASH, en */ //对PORT G操作........ *(volatile unsigned int *)0xfffff430 = 0x023c3d0a; ////++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ //这是I/O模式的主要的两个PORT口,一个是地址口,一个是数据口,这里对0x22操作, //中断IRQ0 ///+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ *(volatile unsigned short*)0x1000030a=0x2200; *(volatile unsigned short*)0x1000030c=0; //*(volatile unsigned short *)0xfffff302 |= 0x0080; ...................... ..................... void reset_chip(struct device *dev) { int reset_start_time; writereg(dev, PP_SelfCTL, readreg(dev, PP_SelfCTL) | POWER_ON_RESET); /* wait 30 ms */ current->state = TASK_INTERRUPTIBLE; current->timeout = jiffies + 3; schedule(); /* Wait until the chip is reset */ //这里jiffies说是timestamp,时间标记,这个变量和current //都在哪里定义? //jiffies主要起什么作用? ////+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ //这里是个延时等待,reset以后必须有等待,以便reset彻底完成,在reset过程中所有寄存器是关闭的 //不但对cs89,对其他芯片也是一样的 ///++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ reset_start_time = jiffies; while( (readreg(dev, PP_SelfST) & INIT_DONE) == 0 && jiffies - reset_start_time< 2); } static int net_open(struct device *dev) { struct net_local *lp = (struct net_local *)dev->priv; int result = 0; int i; write_irq(dev, lp->chip_type, 0); irq2dev_map[/* FIXME */ 0] = dev; writereg(dev, PP_BusCTL, 0); /* ints off! */ //这里和上面那个一样,怎么看懂? ////++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ //这里也是CPU对CS89操作,我也根据CPU硬件手册给你大概说一下 ///+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ //中断信号,正电平触发(POL5) *(volatile unsigned short *)0xfffff302 |= 0x0080; /* +ve pol irq */ //调整PORT G的输入输出状态,及设置管脚相应的电平信号 *(volatile unsigned int *)0xfffff430 = 0x023c3d0a; /* low -> high */ *(volatile unsigned int *)0xfffff430 = 0x023e3d0a; *(volatile unsigned int *)0xfffff430 = 0x023c3d0a; /* high -> low */ ////++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ //这是I/O模式的主要的两个PORT口,一个是地址口,一个是数据口,这里对0x22操作, //中断IRQ0 ///+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ *(volatile unsigned short *)0x1000030a = 0x2200; /* index window, REG index : 0022h */ *(volatile unsigned short *)0x1000030c = 0x0000; /* data window, REG data : 0000h,irq0 */ ......................... ......................... static void net_rx(struct device *dev) { struct net_local *lp = (struct net_local *)dev->priv; int ioaddr = dev->base_addr; struct sk_buff *skb; int status, length; //下面两行怎么右边是一样的? //是作者写错了吗? ////++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++| //不是,这是一个包中前后不同的两个数据,前一个数据是该数据包的状态,后一个是数据包的长度 ////++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ status = inw(ioaddr + RX_FRAME_PORT); length = inw(ioaddr + RX_FRAME_PORT); if ((status & RX_OK) == 0) { lp->stats.rx_errors++; if (status & RX_RUNT) lp->stats.rx_length_errors++; if (status & RX_EXTRA_DATA) lp->stats.rx_length_errors++; if (status & RX_CRC_ERROR) if (!(status & (RX_EXTRA_DATA|RX_RUNT))) /* per str 172 */ lp->stats.rx_crc_errors++; if (status & RX_DRIBBLE) lp->stats.rx_frame_errors++; return; } ...................... static int set_mac_address(struct device *dev, void *addr) { int i; if (dev->start) return -EBUSY; if(get_arena_addr()) { memcpy(dev->dev_addr,0x1100000,6); } else { //这里的00hhcnl和前面 //出现的00hhcn是什么意思? ////+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ //注意:在这里他更改了网卡的MAC值,随便给了一个值"0x00,0x00,'h',h','c','n'", //不过这个48位值必须是世界唯一的,或者用网卡自带的也可以.我不能确定是否eeprom中有信息, //你可以用示波器量一下EEDI管脚,若为高电平,则EEPROM用到了,否则,他们根本没有用EEPROM. ///++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ unsigned char mac[]={0,0,'h','h','c','n'}; memcpy(dev->dev_addr,mac,6); } printk("%s: Setting MAC address to ", dev->name); for (i = 0; i < 6; i++) printk(" %2.2x", dev->dev_addr[i] = ((unsigned char *)addr)[i]); printk(".\n"); /* set the Ethernet address */ for (i=0; i < ETH_ALEN/2; i++) writereg(dev, PP_IA+i*2, dev->dev_addr[i*2] | (dev->dev_addr[i*2+1] << 8)); return 0; } LCD 与触摸屏 Modem拨号 Modem拨号程序步骤如下 1. off hook,发出摘机命令(ATH1\r), 等待modem返回状态,如果是"OK",则表示成功.进行第二步. 2. on hook 发出挂机命令(ATH\r), 等待modem返回状态,如果是"OK",则表示成功.进行第三步. 3. off hook,dialing,摘机拨号(ATDT....\r), 等待modem返回状态,如果是"CONNECT",则表示成功.进行第四步. 4. send data,发送数据. 最好按上述步骤一步一步调,并要等待正确的返回值后再进行下一步. 具体程序例子如下,拨号: // 清空接收FIFO REG(FCR, nsChan) |=FCR_RXCLR; // 等待,直到清空为止 while((REG(FCR, nsChan) & FCR_RXCLR) != 0x00); // 拨号 while(atdt1[i]!=0) { // 检查发送FIFO是否准备好,为空 while((REG(LSR, nsChan) & LSR_THRE) == 0x00); // 在发送FIFO中放入数据 REG(THR, nsChan) = atdt1[i]; i++; } // 等待是否有数据到达,即modem返回数据. while((REG(LSR, nsChan) & LSR_DR) == 0x00); // 接收数据并判断返回值 response[i-19]=REG(RBR,nsChan); if(response[0]!=0x43) .................. else .................. // 如果返回值是"connect", while(*data!=0) { while((REG(LSR, nsChan) & LSR_THRE) == 0x00); REG(THR, nsChan) = *data++; } 其他步骤基本一样. 这是在ARM7上做的modem驱动,硬件设计方法是modem集成在主板上,没有用串口,modem是用片选地址直接对硬件操作的. RTC

文章录入：fengfeiyi 责任编辑：fengfeiyi

上一篇文章： VxWork BSP 和启动过程

下一篇文章： VxWorks基本概念及常见问题

【发表评论】【加入收藏】【告诉好友】【打印此文】【关闭窗口】

最新热点		最新推荐		相关文章
				嵌入式Linux开发工具选择和应… 一步一步的制作arm-linux 交… 高可用性系统的硬件和软件设… 利用电子邮件实现与网络嵌入… 什么样的处理器会引领嵌入式… 配合Tornado使用的Wind Powe… 嵌入式Internet技术及其应用… 嵌入式Internet技术及其应用… 嵌入式Internet技术及其应用… 嵌入式Internet技术及其应用…

　网友评论：（只显示最新10条。评论内容只代表网友观点，与本站立场无关！）

\| 设为首页 \| 加入收藏 \| 联系站长 \| 友情链接 \| 版权申明 \| 管理登录 \|
	Copyright © 2005www.21control.com 21嵌入式控制技术研究室版权所有站长：康草科技