URL
https://opencores.org/ocsvn/or1k/or1k/trunk
Subversion Repositories or1k
[/] [or1k/] [trunk/] [linux/] [linux-2.4/] [include/] [asm-sh64/] [io.h] - Rev 1774
Go to most recent revision | Compare with Previous | Blame | View Log
#ifndef __ASM_SH64_IO_H #define __ASM_SH64_IO_H /* * This file is subject to the terms and conditions of the GNU General Public * License. See the file "COPYING" in the main directory of this archive * for more details. * * include/asm-sh64/io.h * * Copyright (C) 2000, 2001 Paolo Alberelli * Copyright (C) 2003 Paul Mundt * */ /* * Convention: * read{b,w,l}/write{b,w,l} are for PCI, * while in{b,w,l}/out{b,w,l} are for ISA * These may (will) be platform specific function. * * In addition, we have * ctrl_in{b,w,l}/ctrl_out{b,w,l} for SuperH specific I/O. * which are processor specific. Address should be the result of * onchip_remap(); */ #include <asm/cache.h> #include <asm/system.h> #define virt_to_bus virt_to_phys #define bus_to_virt phys_to_virt #define page_to_bus page_to_phys /* * Nothing overly special here.. instead of doing the same thing * over and over again, we just define a set of sh64_in/out functions * with an implicit size. The traditional read{b,w,l}/write{b,w,l} * mess is wrapped to this, as are the SH-specific ctrl_in/out routines. */ static inline unsigned char sh64_in8(unsigned long addr) { return *(volatile unsigned char *)addr; } static inline unsigned short sh64_in16(unsigned long addr) { return *(volatile unsigned short *)addr; } static inline unsigned long sh64_in32(unsigned long addr) { return *(volatile unsigned long *)addr; } static inline unsigned long long sh64_in64(unsigned long addr) { return *(volatile unsigned long long *)addr; } static inline void sh64_out8(unsigned char b, unsigned long addr) { *(volatile unsigned char *)addr = b; wmb(); } static inline void sh64_out16(unsigned short b, unsigned long addr) { *(volatile unsigned short *)addr = b; wmb(); } static inline void sh64_out32(unsigned long b, unsigned long addr) { *(volatile unsigned long *)addr = b; wmb(); } static inline void sh64_out64(unsigned long long b, unsigned long addr) { *(volatile unsigned long long *)addr = b; wmb(); } #define readb(addr) sh64_in8(addr) #define readw(addr) sh64_in16(addr) #define readl(addr) sh64_in32(addr) #define writeb(b, addr) sh64_out8(b, addr) #define writew(b, addr) sh64_out16(b, addr) #define writel(b, addr) sh64_out32(b, addr) #define ctrl_inb(addr) sh64_in8(addr) #define ctrl_inw(addr) sh64_in16(addr) #define ctrl_inl(addr) sh64_in32(addr) #define ctrl_outb(b, addr) sh64_out8(b, addr) #define ctrl_outw(b, addr) sh64_out16(b, addr) #define ctrl_outl(b, addr) sh64_out32(b, addr) unsigned long inb(unsigned long port); unsigned long inw(unsigned long port); unsigned long inl(unsigned long port); void outb(unsigned long value, unsigned long port); void outw(unsigned long value, unsigned long port); void outl(unsigned long value, unsigned long port); #ifdef __KERNEL__ #define IO_SPACE_LIMIT 0xffffffff /* * Change virtual addresses to physical addresses and vv. * These are trivial on the 1:1 Linux/SuperH mapping */ extern __inline__ unsigned long virt_to_phys(volatile void * address) { return __pa(address); } extern __inline__ void * phys_to_virt(unsigned long address) { return __va(address); } extern void * __ioremap(unsigned long phys_addr, unsigned long size, unsigned long flags); extern __inline__ void * ioremap(unsigned long phys_addr, unsigned long size) { return __ioremap(phys_addr, size, 1); } extern __inline__ void * ioremap_nocache (unsigned long phys_addr, unsigned long size) { return __ioremap(phys_addr, size, 0); } extern void iounmap(void *addr); unsigned long onchip_remap(unsigned long addr, unsigned long size, const char* name); extern void onchip_unmap(unsigned long vaddr); static __inline__ int check_signature(unsigned long io_addr, const unsigned char *signature, int length) { int retval = 0; do { if (readb(io_addr) != *signature) goto out; io_addr++; signature++; length--; } while (length); retval = 1; out: return retval; } /* * The caches on some architectures aren't dma-coherent and have need to * handle this in software. There are three types of operations that * can be applied to dma buffers. * * - dma_cache_wback_inv(start, size) makes caches and RAM coherent by * writing the content of the caches back to memory, if necessary. * The function also invalidates the affected part of the caches as * necessary before DMA transfers from outside to memory. * - dma_cache_inv(start, size) invalidates the affected parts of the * caches. Dirty lines of the caches may be written back or simply * be discarded. This operation is necessary before dma operations * to the memory. * - dma_cache_wback(start, size) writes back any dirty lines but does * not invalidate the cache. This can be used before DMA reads from * memory, */ /* * Implemented despite DMA is not yet supported on ST50. * * Also note that PCI DMA is supposed to be cache coherent, * therefore these should not be used by PCI device drivers. * */ static __inline__ void dma_cache_wback_inv (unsigned long start, unsigned long size) { unsigned long s = start & L1_CACHE_ALIGN_MASK; unsigned long e = (start + size) & L1_CACHE_ALIGN_MASK; for (; s <= e; s += L1_CACHE_BYTES) asm volatile ("ocbp %0, 0" : : "r" (s)); } static __inline__ void dma_cache_inv (unsigned long start, unsigned long size) { // Note that caller has to be careful with overzealous // invalidation should there be partial cache lines at the extremities // of the specified range unsigned long s = start & L1_CACHE_ALIGN_MASK; unsigned long e = (start + size) & L1_CACHE_ALIGN_MASK; for (; s <= e; s += L1_CACHE_BYTES) asm volatile ("ocbi %0, 0" : : "r" (s)); } static __inline__ void dma_cache_wback (unsigned long start, unsigned long size) { unsigned long s = start & L1_CACHE_ALIGN_MASK; unsigned long e = (start + size) & L1_CACHE_ALIGN_MASK; for (; s <= e; s += L1_CACHE_BYTES) asm volatile ("ocbwb %0, 0" : : "r" (s)); } #endif /* __KERNEL__ */ #endif /* __ASM_SH64_IO_H */
Go to most recent revision | Compare with Previous | Blame | View Log