xref: /freebsd/sys/amd64/linux/linux_machdep.c (revision 273c26a3c3bea87a241d6879abd4f991db180bf0)
1 /*-
2  * Copyright (c) 2013 Dmitry Chagin
3  * Copyright (c) 2004 Tim J. Robbins
4  * Copyright (c) 2002 Doug Rabson
5  * Copyright (c) 2000 Marcel Moolenaar
6  * All rights reserved.
7  *
8  * Redistribution and use in source and binary forms, with or without
9  * modification, are permitted provided that the following conditions
10  * are met:
11  * 1. Redistributions of source code must retain the above copyright
12  *    notice, this list of conditions and the following disclaimer
13  *    in this position and unchanged.
14  * 2. Redistributions in binary form must reproduce the above copyright
15  *    notice, this list of conditions and the following disclaimer in the
16  *    documentation and/or other materials provided with the distribution.
17  * 3. The name of the author may not be used to endorse or promote products
18  *    derived from this software without specific prior written permission.
19  *
20  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
21  * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
22  * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
23  * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
24  * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
25  * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26  * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27  * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28  * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
29  * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
30  */
31 
32 #include <sys/cdefs.h>
33 __FBSDID("$FreeBSD$");
34 
35 #include <sys/param.h>
36 #include <sys/kernel.h>
37 #include <sys/systm.h>
38 #include <sys/capsicum.h>
39 #include <sys/dirent.h>
40 #include <sys/file.h>
41 #include <sys/fcntl.h>
42 #include <sys/filedesc.h>
43 #include <sys/clock.h>
44 #include <sys/imgact.h>
45 #include <sys/ktr.h>
46 #include <sys/limits.h>
47 #include <sys/lock.h>
48 #include <sys/malloc.h>
49 #include <sys/mman.h>
50 #include <sys/mutex.h>
51 #include <sys/priv.h>
52 #include <sys/proc.h>
53 #include <sys/resource.h>
54 #include <sys/resourcevar.h>
55 #include <sys/sched.h>
56 #include <sys/syscallsubr.h>
57 #include <sys/sysproto.h>
58 #include <sys/vnode.h>
59 #include <sys/unistd.h>
60 #include <sys/wait.h>
61 
62 #include <security/mac/mac_framework.h>
63 
64 #include <ufs/ufs/extattr.h>
65 #include <ufs/ufs/quota.h>
66 #include <ufs/ufs/ufsmount.h>
67 
68 #include <machine/frame.h>
69 #include <machine/md_var.h>
70 #include <machine/pcb.h>
71 #include <machine/psl.h>
72 #include <machine/segments.h>
73 #include <machine/specialreg.h>
74 
75 #include <vm/vm.h>
76 #include <vm/pmap.h>
77 #include <vm/vm_extern.h>
78 #include <vm/vm_kern.h>
79 #include <vm/vm_map.h>
80 
81 #include <amd64/linux/linux.h>
82 #include <amd64/linux/linux_proto.h>
83 #include <compat/linux/linux_ipc.h>
84 #include <compat/linux/linux_file.h>
85 #include <compat/linux/linux_misc.h>
86 #include <compat/linux/linux_mmap.h>
87 #include <compat/linux/linux_signal.h>
88 #include <compat/linux/linux_util.h>
89 #include <compat/linux/linux_emul.h>
90 
91 
92 int
93 linux_execve(struct thread *td, struct linux_execve_args *args)
94 {
95 	struct image_args eargs;
96 	char *path;
97 	int error;
98 
99 	LCONVPATHEXIST(td, args->path, &path);
100 
101 	LINUX_CTR(execve);
102 
103 	error = exec_copyin_args(&eargs, path, UIO_SYSSPACE, args->argp,
104 	    args->envp);
105 	free(path, M_TEMP);
106 	if (error == 0)
107 		error = linux_common_execve(td, &eargs);
108 	return (error);
109 }
110 
111 int
112 linux_set_upcall_kse(struct thread *td, register_t stack)
113 {
114 
115 	if (stack)
116 		td->td_frame->tf_rsp = stack;
117 
118 	/*
119 	 * The newly created Linux thread returns
120 	 * to the user space by the same path that a parent do.
121 	 */
122 	td->td_frame->tf_rax = 0;
123 	return (0);
124 }
125 
126 int
127 linux_mmap2(struct thread *td, struct linux_mmap2_args *args)
128 {
129 
130 	return (linux_mmap_common(td, PTROUT(args->addr), args->len, args->prot,
131 		args->flags, args->fd, args->pgoff));
132 }
133 
134 int
135 linux_mprotect(struct thread *td, struct linux_mprotect_args *uap)
136 {
137 
138 	return (linux_mprotect_common(td, PTROUT(uap->addr), uap->len, uap->prot));
139 }
140 
141 int
142 linux_iopl(struct thread *td, struct linux_iopl_args *args)
143 {
144 	int error;
145 
146 	LINUX_CTR(iopl);
147 
148 	if (args->level > 3)
149 		return (EINVAL);
150 	if ((error = priv_check(td, PRIV_IO)) != 0)
151 		return (error);
152 	if ((error = securelevel_gt(td->td_ucred, 0)) != 0)
153 		return (error);
154 	td->td_frame->tf_rflags = (td->td_frame->tf_rflags & ~PSL_IOPL) |
155 	    (args->level * (PSL_IOPL / 3));
156 
157 	return (0);
158 }
159 
160 int
161 linux_rt_sigsuspend(struct thread *td, struct linux_rt_sigsuspend_args *uap)
162 {
163 	l_sigset_t lmask;
164 	sigset_t sigmask;
165 	int error;
166 
167 	LINUX_CTR2(rt_sigsuspend, "%p, %ld",
168 	    uap->newset, uap->sigsetsize);
169 
170 	if (uap->sigsetsize != sizeof(l_sigset_t))
171 		return (EINVAL);
172 
173 	error = copyin(uap->newset, &lmask, sizeof(l_sigset_t));
174 	if (error)
175 		return (error);
176 
177 	linux_to_bsd_sigset(&lmask, &sigmask);
178 	return (kern_sigsuspend(td, sigmask));
179 }
180 
181 int
182 linux_pause(struct thread *td, struct linux_pause_args *args)
183 {
184 	struct proc *p = td->td_proc;
185 	sigset_t sigmask;
186 
187 	LINUX_CTR(pause);
188 
189 	PROC_LOCK(p);
190 	sigmask = td->td_sigmask;
191 	PROC_UNLOCK(p);
192 	return (kern_sigsuspend(td, sigmask));
193 }
194 
195 int
196 linux_sigaltstack(struct thread *td, struct linux_sigaltstack_args *uap)
197 {
198 	stack_t ss, oss;
199 	l_stack_t lss;
200 	int error;
201 
202 	LINUX_CTR2(sigaltstack, "%p, %p", uap->uss, uap->uoss);
203 
204 	if (uap->uss != NULL) {
205 		error = copyin(uap->uss, &lss, sizeof(l_stack_t));
206 		if (error)
207 			return (error);
208 
209 		ss.ss_sp = PTRIN(lss.ss_sp);
210 		ss.ss_size = lss.ss_size;
211 		ss.ss_flags = linux_to_bsd_sigaltstack(lss.ss_flags);
212 	}
213 	error = kern_sigaltstack(td, (uap->uss != NULL) ? &ss : NULL,
214 	    (uap->uoss != NULL) ? &oss : NULL);
215 	if (!error && uap->uoss != NULL) {
216 		lss.ss_sp = PTROUT(oss.ss_sp);
217 		lss.ss_size = oss.ss_size;
218 		lss.ss_flags = bsd_to_linux_sigaltstack(oss.ss_flags);
219 		error = copyout(&lss, uap->uoss, sizeof(l_stack_t));
220 	}
221 
222 	return (error);
223 }
224 
225 int
226 linux_arch_prctl(struct thread *td, struct linux_arch_prctl_args *args)
227 {
228 	int error;
229 	struct pcb *pcb;
230 
231 	LINUX_CTR2(arch_prctl, "0x%x, %p", args->code, args->addr);
232 
233 	error = ENOTSUP;
234 	pcb = td->td_pcb;
235 
236 	switch (args->code) {
237 	case LINUX_ARCH_GET_GS:
238 		error = copyout(&pcb->pcb_gsbase, (unsigned long *)args->addr,
239 		    sizeof(args->addr));
240 		break;
241 	case LINUX_ARCH_SET_GS:
242 		if (args->addr >= VM_MAXUSER_ADDRESS)
243 			return(EPERM);
244 		break;
245 	case LINUX_ARCH_GET_FS:
246 		error = copyout(&pcb->pcb_fsbase, (unsigned long *)args->addr,
247 		    sizeof(args->addr));
248 		break;
249 	case LINUX_ARCH_SET_FS:
250 		error = linux_set_cloned_tls(td, (void *)args->addr);
251 		break;
252 	default:
253 		error = EINVAL;
254 	}
255 	return (error);
256 }
257 
258 int
259 linux_set_cloned_tls(struct thread *td, void *desc)
260 {
261 	struct pcb *pcb;
262 
263 	if ((uint64_t)desc >= VM_MAXUSER_ADDRESS)
264 		return (EPERM);
265 
266 	pcb = td->td_pcb;
267 	pcb->pcb_fsbase = (register_t)desc;
268 	td->td_frame->tf_fs = _ufssel;
269 
270 	return (0);
271 }
272