1 /* $NetBSD: linux_syscall.c,v 1.50 2009/11/21 03:11:00 rmind Exp $ */ 2 3 /*- 4 * Copyright (c) 1998, 2000 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Charles M. Hannum. 9 * 10 * Redistribution and use in source and binary forms, with or without 11 * modification, are permitted provided that the following conditions 12 * are met: 13 * 1. Redistributions of source code must retain the above copyright 14 * notice, this list of conditions and the following disclaimer. 15 * 2. Redistributions in binary form must reproduce the above copyright 16 * notice, this list of conditions and the following disclaimer in the 17 * documentation and/or other materials provided with the distribution. 18 * 19 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 20 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 21 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 22 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 23 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 24 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 25 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 26 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 27 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 28 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 29 * POSSIBILITY OF SUCH DAMAGE. 30 */ 31 32 #include <sys/cdefs.h> 33 __KERNEL_RCSID(0, "$NetBSD: linux_syscall.c,v 1.50 2009/11/21 03:11:00 rmind Exp $"); 34 35 #if defined(_KERNEL_OPT) 36 #include "opt_vm86.h" 37 #endif 38 39 #include <sys/param.h> 40 #include <sys/systm.h> 41 #include <sys/proc.h> 42 #include <sys/signal.h> 43 #include <sys/syscall.h> 44 #include <sys/syscallvar.h> 45 46 #include <uvm/uvm_extern.h> 47 48 #include <machine/cpu.h> 49 #include <machine/psl.h> 50 #include <machine/userret.h> 51 52 #include <compat/linux/common/linux_types.h> 53 #include <compat/linux/common/linux_errno.h> 54 #include <compat/linux/linux_syscall.h> 55 #include <compat/linux/common/linux_signal.h> 56 #include <compat/linux/arch/i386/linux_machdep.h> 57 58 static void linux_syscall(struct trapframe *); 59 extern struct sysent linux_sysent[]; 60 61 void 62 linux_syscall_intern(struct proc *p) 63 { 64 65 p->p_md.md_syscall = linux_syscall; 66 } 67 68 /* 69 * syscall(frame): 70 * System call request from POSIX system call gate interface to kernel. 71 * Like trap(), argument is call by reference. 72 */ 73 void 74 linux_syscall(struct trapframe *frame) 75 { 76 register const struct sysent *callp; 77 struct lwp *l; 78 int error; 79 register_t code, args[6], rval[2]; 80 81 l = curlwp; 82 LWP_CACHE_CREDS(l, l->l_proc); 83 84 code = frame->tf_eax & (LINUX_SYS_NSYSENT - 1); 85 callp = linux_sysent; 86 87 callp += code; 88 /* 89 * Linux passes the args in ebx, ecx, edx, esi, edi, ebp, in 90 * increasing order. 91 */ 92 args[0] = frame->tf_ebx; 93 args[1] = frame->tf_ecx; 94 args[2] = frame->tf_edx; 95 args[3] = frame->tf_esi; 96 args[4] = frame->tf_edi; 97 args[5] = frame->tf_ebp; 98 99 rval[0] = 0; 100 rval[1] = 0; 101 102 if (__predict_false(l->l_proc->p_trace_enabled)) { 103 error = trace_enter(code, args, callp->sy_narg); 104 if (__predict_true(error == 0)) { 105 error = sy_call(callp, l, args, rval); 106 code = frame->tf_eax & (LINUX_SYS_NSYSENT - 1); 107 trace_exit(code, rval, error); 108 } 109 } else 110 error = sy_call(callp, l, args, rval); 111 112 if (__predict_true(error == 0)) { 113 frame->tf_eax = rval[0]; 114 /* 115 * XXX The linux libc code I (dsl) looked at doesn't use the 116 * carry bit. 117 * Values above 0xfffff000 are assumed to be errno values and 118 * not result codes! 119 */ 120 frame->tf_eflags &= ~PSL_C; /* carry bit */ 121 } else { 122 switch (error) { 123 case ERESTART: 124 /* 125 * The offset to adjust the PC by depends on whether 126 * we entered the kernel through the trap or call gate. 127 * We save the instruction size in tf_err on entry. 128 */ 129 frame->tf_eip -= frame->tf_err; 130 break; 131 case EJUSTRETURN: 132 /* nothing to do */ 133 break; 134 default: 135 error = native_to_linux_errno[error]; 136 frame->tf_eax = error; 137 frame->tf_eflags |= PSL_C; /* carry bit */ 138 break; 139 } 140 } 141 142 userret(l); 143 } 144