关于在LPC1752移植碰到的相关问题,总是死在rt_hw_hard_fault_exception,请教ffxz和aoz
硬件:LPC1752(zlg开发板) + CF卡 (多余的IO脚已经悬空)软件:RTT0.4.0RC
问题:跟踪发现每次都在rt_device_init_all中初始化串口1或者串口2或者串口3的Fdiv = ( pclk / 16 ) / UART0_BAUDRATE;处直接跳入fault_rvds的rt_hw_hard_fault PROC再跳入硬件异常函数。初始化函数没有问题已经脱离RTT在工程中用过,软件仿真的时候能正常通过rt_device_init_all,怀疑是和硬件那里冲突,一时找不到!下面是初始化函数和异常响应的。finsh在串口0,去掉所有串口驱动则不会进入硬件异常。
static rt_err_t rt_uart_init (rt_device_t dev)
{
rt_uint32_t Fdiv;
rt_uint32_t pclkdiv, pclk;
/* Init UART Hardware */
LPC_PINCON->PINSEL0 &= ~0x000000F0;
LPC_PINCON->PINSEL0 |= 0x00000050;/* RxD0 is P0.3 and TxD0 is P0.2 */
/* By default, the PCLKSELx value is zero, thus, the PCLK for
all the peripherals is 1/4 of the SystemFrequency. */
/* Bit 6~7 is for UART0 */
pclkdiv = (LPC_SC->PCLKSEL0 >> 6) & 0x03;
switch ( pclkdiv )
{
case 0x00:
default:
pclk = SystemCoreClock/4;
break;
case 0x01:
pclk = SystemCoreClock;
break;
case 0x02:
pclk = SystemCoreClock/2;
break;
case 0x03:
pclk = SystemCoreClock/8;
break;
}
LPC_UART0->LCR = 0x83; /* 8 bits, no Parity, 1 Stop bit */
Fdiv = ( pclk / 16 ) / UART0_BAUDRATE; /*baud rate */
LPC_UART0->DLM = Fdiv / 256; <==================== 此行跳入硬件异常,屏蔽后下面一行也进入到异常
LPC_UART0->DLL = Fdiv % 256;
LPC_UART0->LCR = 0x03; /* DLAB = 0 */
LPC_UART0->FCR = 0x07; /* Enable and reset TX and RX FIFO. */
/* Ensure a clean start, no data in either TX or RX FIFO. */
while (( LPC_UART0->LSR & (LSR_THRE|LSR_TEMT)) != (LSR_THRE|LSR_TEMT) );
while ( LPC_UART0->LSR & LSR_RDR )
{
Fdiv = LPC_UART0->RBR; /* Dump data from RX FIFO */
}
LPC_UART0->IER = IER_RBR | IER_THRE | IER_RLS; /* Enable UART interrupt */
return RT_EOK;
}
psr: 0x00000000
pc: 0x00000000
lr: 0x00000000
r12: 0x00000000
r03: 0x00000000
r02: 0x00000000
r01: 0x00000000
r00: 0x00000000
hard fault on thread: (NULL)
threadpristatus sp stack size max used left tickerror
-------- ---- ------- ---------- ---------- ---------- ---------- --- static rt_size_t rt_ide_read(rt_device_t dev, rt_off_t pos, void* buffer, rt_size_t size)
还有一个问题,CF卡驱动移植到RTT时,手册中也没有几个参数的含义?
size是指定扇区的地址还是要读取的个数?
pos是偏移量的位置还是偏移量?
已经实现了指定扇区地址上读写512B的底层函数
unsigned char WriteCompactFlashSector(unsigned long addr , unsigned char *buf)
unsigned char ReadCompactFlashSector(unsigned long addr , unsigned char *buf)
移植到rt_ide_read中是不是这样写
static rt_size_t rt_ide_read(rt_device_t dev, rt_off_t pos, void* buffer, rt_size_t size)
{
unsigned long count , offset , idx ;
unsigned char *pData ;
pData = ( unsigned char *)buffer ;
count= size / 512 ;
offset = (pos / 512) + part.offset ;
for(idx = 0 ; idx < count ; idx++)
{
ReadCompactFlashSector(size , pData) ;
}
ReadCompactFlashSector(offset , pData) ;
return count ;
} 楼主,这种涉及硬件的问题还是自己找原因来得快,我上次用RTT读写SDRAM也是碰到硬件问题,后来也只能自个解决。 是的,但是该找的都找了,现在没有思路了! Q:size是指定扇区的地址还是要读取的个数?
A:size直译为尺寸,肯定是指要读取的个数。以前单位是byte,为支持更大的设备,现在的单位是sector.
Q:pos是偏移量的位置还是偏移量?
A:rt_off_t pos, 即offset,单位也由原来的byte改为sector.
因此,驱动的 control 函数必须支持 RT_DEVICE_CTRL_BLK_GETGEOME 命令。
static rt_err_t rt_cf_control(rt_device_t dev, rt_uint8_t cmd, void *args)
{
RT_ASSERT(dev != RT_NULL);
if (cmd == RT_DEVICE_CTRL_BLK_GETGEOME)
{
struct rt_device_blk_geometry *geometry;
geometry = (struct rt_device_blk_geometry *)args;
if (geometry == RT_NULL) return -RT_ERROR;
geometry->bytes_per_sector = 512; /* 非512字节每扇区的设备需要根据实际情况确定 */
geometry->block_size = block_size;
geometry->sector_count = 500G / geometry->bytes_per_sector;
}
return RT_EOK;
}
关于 hard_fault,先检查一下 board.h中SRAM的大小有没不对应,然后不要仿真,直接下载让其冷脱机跑一下是否正常。
如果正常,且是用JLINK仿真器,试下MDK示例中那个deg.ini,让在断点时关掉timer。
如还不行,那就不知道了。 aozima读写函数是否有问题呢??在使用DFS文件系统时,CF卡的驱动需要实现控制函数吗
static rt_size_t rt_ide_read(rt_device_t dev, rt_off_t pos, void* buffer, rt_size_t size)
{
unsigned long count , offset , idx ;
unsigned char *pData ;
pData = ( unsigned char *)buffer ;
count= size / 512 ;
offset = (pos / 512) + part.offset ;
for(idx = 0 ; idx < count ; idx++)
{
ReadCompactFlashSector(size , pData) ;
}
ReadCompactFlashSector(offset , pData) ;
return count ;
}
ReadCompactFlashSector是在指定的地址读一个扇区512Byte ram我已经改了,因为是用的1752 Ram是8K,所以堆的定义应改到10004000,但是这样还是不行,同样的硬件异常@
#ifdef RT_USING_HEAP
#ifdef __CC_ARM
rt_system_heap_init((void*)&Image$$RW_IRAM1$$ZI$$Limit, (void*)0x10004000);
#elif __ICCARM__
rt_system_heap_init(__segment_end("HEAP"), (void*)0x10004000);
#else
/* init memory system */
rt_system_heap_init((void*)&__bss_end, (void*)0x10004000);
#endif
#endif 另外软件仿真没有出现问题,奇怪了!
页:
[1]