ubuntu 10.04 LTS安装通过NachOS-4.1

/* coff2noff.c * * This program reads in a COFF format file, and outputs a NOFF format file. * The NOFF format is essentially just a simpler version of the COFF file, * recording where each segment is in the NOFF file, and where it is to * go in the virtual address space. * * Assumes coff file is linked with either * gld with -N -Ttext 0 * ld with -N -T 0 * to make sure the object file has no shared text. * * Also assumes that the COFF file has at most 3 segments: * .text -- read-only executable instructions * .data -- initialized data * .bss/.sbss -- uninitialized data (should be zero'd on program startup) #ifdef RDATA * .rdata -- read-only data (e.g., string literals). * mark this segment readonly to prevent it from being modified #endif * * * Copyright (c) 1992-1993 The Regents of the University of California. * All rights reserved. See copyright.h for copyright notice and limitation * of liability and disclaimer of warranty provisions. */ /* * Modified at UW by KMS, August, 1997 * The modified program always writes the NOFF header in little-endian * format, rather than host format. This is to avoid the problem * that user programs run through coff2noff on a big-endian host * would not run properly on Nachos machines running on little-endian * hosts. * * Note that the Nachos address space loading code * (in AddrSpace::Load) on big-endian hosts converts the header * to big-endian format when it is read in. * Thus, the little-endian header NOFF * header should work OK whether Nachos is running on a little-endian * host or a big-endian host. */ #define _POSIX_SOURCE 1 #ifdef HAVE_CONFIG_H #include "config.h" #endif #define MAIN #include "copyright.h" #undef MAIN #include "coff.h" #include "noff.h" #include <stdarg.h> #include <unistd.h> #include <sys/types.h> #include <sys/stat.h> #include <fcntl.h> #include <limits.h> #include <stdio.h> /****************************************************************/ /* Routines for converting words and short words to and from the * simulated machine's format of little endian. These end up * being NOPs when the host machine is little endian. */ unsigned int WordToHost(unsigned int word) { #ifdef HOST_IS_BIG_ENDIAN register unsigned long result; result = (word >> 24) & 0x000000ff; result |= (word >> 8) & 0x0000ff00; result |= (word << 8) & 0x00ff0000; result |= (word << 24) & 0xff000000; return result; #else return word; #endif /* HOST_IS_BIG_ENDIAN */ } unsigned short ShortToHost(unsigned short shortword) { #if HOST_IS_BIG_ENDIAN register unsigned short result; result = (shortword << 8) & 0xff00; result |= (shortword >> 8) & 0x00ff; return result; #else return shortword; #endif /* HOST_IS_BIG_ENDIAN */ } unsigned int WordToMachine(unsigned int word) { return WordToHost(word); } unsigned short ShortToMachine(unsigned short shortword) { return ShortToHost(shortword); } // this routine was borrowed from userprog/addrspace.cc // on a big-endian machine, it converts all fields of // the NOFF header to little-endian format // on a little-endian machine, where the header is already // in little-endian format, it does nothing static void SwapHeader (NoffHeader *noffH) { noffH->noffMagic = WordToHost(noffH->noffMagic); noffH->code.size = WordToHost(noffH->code.size); noffH->code.virtualAddr = WordToHost(noffH->code.virtualAddr); noffH->code.inFileAddr = WordToHost(noffH->code.inFileAddr); #ifdef RDATA noffH->readonlyData.size = WordToHost(noffH->readonlyData.size); noffH->readonlyData.virtualAddr = WordToHost(noffH->readonlyData.virtualAddr); noffH->readonlyData.inFileAddr = WordToHost(noffH->readonlyData.inFileAddr); #endif noffH->initData.size = WordToHost(noffH->initData.size); noffH->initData.virtualAddr = WordToHost(noffH->initData.virtualAddr); noffH->initData.inFileAddr = WordToHost(noffH->initData.inFileAddr); noffH->uninitData.size = WordToHost(noffH->uninitData.size); noffH->uninitData.virtualAddr = WordToHost(noffH->uninitData.virtualAddr); noffH->uninitData.inFileAddr = WordToHost(noffH->uninitData.inFileAddr); } /****************************************************************/ #define ReadStruct(f,s) Read(f,(char *)&s,sizeof(s)) extern char *malloc(); char *noffFileName = NULL; /* read and check for error */ void Read(int fd, char *buf, int nBytes) { if (read(fd, buf, nBytes) != nBytes) { fprintf(stderr, "File is too short\n"); unlink(noffFileName); exit(1); } } /* write and check for error */ void Write(int fd, char *buf, int nBytes) { if (write(fd, buf, nBytes) != nBytes) { fprintf(stderr, "Unable to write file\n"); unlink(noffFileName); exit(1); } } main (int argc, char **argv) { int fdIn, fdOut, numsections, i, inNoffFile; struct filehdr fileh; struct aouthdr systemh; struct scnhdr *sections; char *buffer; NoffHeader noffH; if (argc < 2) { fprintf(stderr, "Usage: %s <coffFileName> <noffFileName>\n", argv[0]); exit(1); } /* open the COFF file (input) */ fdIn = open(argv[1], O_RDONLY, 0); if (fdIn == -1) { perror(argv[1]); exit(1); } /* open the NOFF file (output) */ fdOut = open(argv[2], O_WRONLY|O_CREAT|O_TRUNC , 0666); if (fdIn == -1) { perror(argv[2]); exit(1); } noffFileName = argv[2]; /* Read in the file header and check the magic number. */ ReadStruct(fdIn,fileh); fileh.f_magic = ShortToHost(fileh.f_magic); fileh.f_nscns = ShortToHost(fileh.f_nscns); if (fileh.f_magic != MIPSELMAGIC) { fprintf(stderr, "File is not a MIPSEL COFF file\n"); unlink(noffFileName); exit(1); } /* Read in the system header and check the magic number */ ReadStruct(fdIn,systemh); systemh.magic = ShortToHost(systemh.magic); if (systemh.magic != OMAGIC) { fprintf(stderr, "File is not a OMAGIC file\n"); unlink(noffFileName); exit(1); } /* Read in the section headers. */ numsections = fileh.f_nscns; printf("numsections %d \n",numsections); sections = (struct scnhdr *)malloc(numsections * sizeof(struct scnhdr)); Read(fdIn, (char *) sections, numsections * sizeof(struct scnhdr)); for (i = 0; i < numsections; i++) { sections[i].s_paddr = WordToHost(sections[i].s_paddr); sections[i].s_size = WordToHost(sections[i].s_size); sections[i].s_scnptr = WordToHost(sections[i].s_scnptr); } /* initialize the NOFF header, in case not all the segments are defined * in the COFF file */ noffH.noffMagic = NOFFMAGIC; noffH.code.size = 0; noffH.initData.size = 0; noffH.uninitData.size = 0; #ifdef RDATA noffH.readonlyData.size = 0; #endif /* Copy the segments in */ inNoffFile = sizeof(NoffHeader); lseek(fdOut, inNoffFile, 0); printf("Loading %d sections:\n", numsections); for (i = 0; i < numsections; i++) { printf("\t\"%s\", filepos 0x%x, mempos 0x%x, size 0x%x\n", sections[i].s_name, sections[i].s_scnptr, sections[i].s_paddr, sections[i].s_size); if (sections[i].s_size == 0) { /* do nothing! */ } else if (!strcmp(sections[i].s_name, ".text")) { noffH.code.virtualAddr = sections[i].s_paddr; noffH.code.inFileAddr = inNoffFile; noffH.code.size = sections[i].s_size; lseek(fdIn, sections[i].s_scnptr, 0); buffer = malloc(sections[i].s_size); Read(fdIn, buffer, sections[i].s_size); Write(fdOut, buffer, sections[i].s_size); free(buffer); inNoffFile += sections[i].s_size; } else if (!strcmp(sections[i].s_name, ".data")){ noffH.initData.virtualAddr = sections[i].s_paddr; noffH.initData.inFileAddr = inNoffFile; noffH.initData.size = sections[i].s_size; lseek(fdIn, sections[i].s_scnptr, 0); buffer = malloc(sections[i].s_size); Read(fdIn, buffer, sections[i].s_size);