1 /* $NetBSD: linux_misc.c,v 1.234 2017/01/02 16:32:10 manu Exp $ */ 2 3 /*- 4 * Copyright (c) 1995, 1998, 1999, 2008 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Frank van der Linden and Eric Haszlakiewicz; by Jason R. Thorpe 9 * of the Numerical Aerospace Simulation Facility, NASA Ames Research Center. 10 * 11 * Redistribution and use in source and binary forms, with or without 12 * modification, are permitted provided that the following conditions 13 * are met: 14 * 1. Redistributions of source code must retain the above copyright 15 * notice, this list of conditions and the following disclaimer. 16 * 2. Redistributions in binary form must reproduce the above copyright 17 * notice, this list of conditions and the following disclaimer in the 18 * documentation and/or other materials provided with the distribution. 19 * 20 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 21 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 22 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 23 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 24 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 25 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 26 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 27 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 28 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 29 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 30 * POSSIBILITY OF SUCH DAMAGE. 31 */ 32 33 /* 34 * Linux compatibility module. Try to deal with various Linux system calls. 35 */ 36 37 /* 38 * These functions have been moved to multiarch to allow 39 * selection of which machines include them to be 40 * determined by the individual files.linux_<arch> files. 41 * 42 * Function in multiarch: 43 * linux_sys_break : linux_break.c 44 * linux_sys_alarm : linux_misc_notalpha.c 45 * linux_sys_getresgid : linux_misc_notalpha.c 46 * linux_sys_nice : linux_misc_notalpha.c 47 * linux_sys_readdir : linux_misc_notalpha.c 48 * linux_sys_setresgid : linux_misc_notalpha.c 49 * linux_sys_time : linux_misc_notalpha.c 50 * linux_sys_utime : linux_misc_notalpha.c 51 * linux_sys_waitpid : linux_misc_notalpha.c 52 * linux_sys_old_mmap : linux_oldmmap.c 53 * linux_sys_oldolduname : linux_oldolduname.c 54 * linux_sys_oldselect : linux_oldselect.c 55 * linux_sys_olduname : linux_olduname.c 56 * linux_sys_pipe : linux_pipe.c 57 */ 58 59 #include <sys/cdefs.h> 60 __KERNEL_RCSID(0, "$NetBSD: linux_misc.c,v 1.234 2017/01/02 16:32:10 manu Exp $"); 61 62 #include <sys/param.h> 63 #include <sys/systm.h> 64 #include <sys/namei.h> 65 #include <sys/proc.h> 66 #include <sys/dirent.h> 67 #include <sys/file.h> 68 #include <sys/stat.h> 69 #include <sys/filedesc.h> 70 #include <sys/ioctl.h> 71 #include <sys/kernel.h> 72 #include <sys/malloc.h> 73 #include <sys/mbuf.h> 74 #include <sys/mman.h> 75 #include <sys/mount.h> 76 #include <sys/poll.h> 77 #include <sys/prot.h> 78 #include <sys/reboot.h> 79 #include <sys/resource.h> 80 #include <sys/resourcevar.h> 81 #include <sys/select.h> 82 #include <sys/signal.h> 83 #include <sys/signalvar.h> 84 #include <sys/socket.h> 85 #include <sys/time.h> 86 #include <sys/times.h> 87 #include <sys/vnode.h> 88 #include <sys/uio.h> 89 #include <sys/wait.h> 90 #include <sys/utsname.h> 91 #include <sys/unistd.h> 92 #include <sys/vfs_syscalls.h> 93 #include <sys/swap.h> /* for SWAP_ON */ 94 #include <sys/sysctl.h> /* for KERN_DOMAINNAME */ 95 #include <sys/kauth.h> 96 97 #include <sys/ptrace.h> 98 #include <machine/ptrace.h> 99 100 #include <sys/syscall.h> 101 #include <sys/syscallargs.h> 102 103 #include <compat/sys/resource.h> 104 105 #include <compat/linux/common/linux_machdep.h> 106 #include <compat/linux/common/linux_types.h> 107 #include <compat/linux/common/linux_signal.h> 108 #include <compat/linux/common/linux_ipc.h> 109 #include <compat/linux/common/linux_sem.h> 110 111 #include <compat/linux/common/linux_fcntl.h> 112 #include <compat/linux/common/linux_mmap.h> 113 #include <compat/linux/common/linux_dirent.h> 114 #include <compat/linux/common/linux_util.h> 115 #include <compat/linux/common/linux_misc.h> 116 #include <compat/linux/common/linux_statfs.h> 117 #include <compat/linux/common/linux_limit.h> 118 #include <compat/linux/common/linux_ptrace.h> 119 #include <compat/linux/common/linux_reboot.h> 120 #include <compat/linux/common/linux_emuldata.h> 121 #include <compat/linux/common/linux_sched.h> 122 123 #include <compat/linux/linux_syscallargs.h> 124 125 const int linux_ptrace_request_map[] = { 126 LINUX_PTRACE_TRACEME, PT_TRACE_ME, 127 LINUX_PTRACE_PEEKTEXT, PT_READ_I, 128 LINUX_PTRACE_PEEKDATA, PT_READ_D, 129 LINUX_PTRACE_POKETEXT, PT_WRITE_I, 130 LINUX_PTRACE_POKEDATA, PT_WRITE_D, 131 LINUX_PTRACE_CONT, PT_CONTINUE, 132 LINUX_PTRACE_KILL, PT_KILL, 133 LINUX_PTRACE_ATTACH, PT_ATTACH, 134 LINUX_PTRACE_DETACH, PT_DETACH, 135 # ifdef PT_STEP 136 LINUX_PTRACE_SINGLESTEP, PT_STEP, 137 # endif 138 LINUX_PTRACE_SYSCALL, PT_SYSCALL, 139 -1 140 }; 141 142 const struct linux_mnttypes linux_fstypes[] = { 143 { MOUNT_FFS, LINUX_DEFAULT_SUPER_MAGIC }, 144 { MOUNT_NFS, LINUX_NFS_SUPER_MAGIC }, 145 { MOUNT_MFS, LINUX_DEFAULT_SUPER_MAGIC }, 146 { MOUNT_MSDOS, LINUX_MSDOS_SUPER_MAGIC }, 147 { MOUNT_LFS, LINUX_DEFAULT_SUPER_MAGIC }, 148 { MOUNT_FDESC, LINUX_DEFAULT_SUPER_MAGIC }, 149 { MOUNT_NULL, LINUX_DEFAULT_SUPER_MAGIC }, 150 { MOUNT_OVERLAY, LINUX_DEFAULT_SUPER_MAGIC }, 151 { MOUNT_UMAP, LINUX_DEFAULT_SUPER_MAGIC }, 152 { MOUNT_KERNFS, LINUX_DEFAULT_SUPER_MAGIC }, 153 { MOUNT_PROCFS, LINUX_PROC_SUPER_MAGIC }, 154 { MOUNT_AFS, LINUX_DEFAULT_SUPER_MAGIC }, 155 { MOUNT_CD9660, LINUX_ISOFS_SUPER_MAGIC }, 156 { MOUNT_UNION, LINUX_DEFAULT_SUPER_MAGIC }, 157 { MOUNT_ADOSFS, LINUX_ADFS_SUPER_MAGIC }, 158 { MOUNT_EXT2FS, LINUX_EXT2_SUPER_MAGIC }, 159 { MOUNT_CFS, LINUX_DEFAULT_SUPER_MAGIC }, 160 { MOUNT_CODA, LINUX_CODA_SUPER_MAGIC }, 161 { MOUNT_FILECORE, LINUX_DEFAULT_SUPER_MAGIC }, 162 { MOUNT_NTFS, LINUX_DEFAULT_SUPER_MAGIC }, 163 { MOUNT_SMBFS, LINUX_SMB_SUPER_MAGIC }, 164 { MOUNT_PTYFS, LINUX_DEVPTS_SUPER_MAGIC }, 165 { MOUNT_TMPFS, LINUX_TMPFS_SUPER_MAGIC } 166 }; 167 const int linux_fstypes_cnt = sizeof(linux_fstypes) / sizeof(linux_fstypes[0]); 168 169 # ifdef DEBUG_LINUX 170 #define DPRINTF(a) uprintf a 171 # else 172 #define DPRINTF(a) 173 # endif 174 175 /* Local linux_misc.c functions: */ 176 static void linux_to_bsd_mmap_args(struct sys_mmap_args *, 177 const struct linux_sys_mmap_args *); 178 static int linux_mmap(struct lwp *, const struct linux_sys_mmap_args *, 179 register_t *, off_t); 180 181 182 /* 183 * The information on a terminated (or stopped) process needs 184 * to be converted in order for Linux binaries to get a valid signal 185 * number out of it. 186 */ 187 int 188 bsd_to_linux_wstat(int st) 189 { 190 191 int sig; 192 193 if (WIFSIGNALED(st)) { 194 sig = WTERMSIG(st); 195 if (sig >= 0 && sig < NSIG) 196 st= (st & ~0177) | native_to_linux_signo[sig]; 197 } else if (WIFSTOPPED(st)) { 198 sig = WSTOPSIG(st); 199 if (sig >= 0 && sig < NSIG) 200 st = (st & ~0xff00) | 201 (native_to_linux_signo[sig] << 8); 202 } 203 return st; 204 } 205 206 /* 207 * wait4(2). Passed on to the NetBSD call, surrounded by code to 208 * reserve some space for a NetBSD-style wait status, and converting 209 * it to what Linux wants. 210 */ 211 int 212 linux_sys_wait4(struct lwp *l, const struct linux_sys_wait4_args *uap, register_t *retval) 213 { 214 /* { 215 syscallarg(int) pid; 216 syscallarg(int *) status; 217 syscallarg(int) options; 218 syscallarg(struct rusage50 *) rusage; 219 } */ 220 int error, status, options, linux_options, pid = SCARG(uap, pid); 221 struct rusage50 ru50; 222 struct rusage ru; 223 proc_t *p; 224 225 linux_options = SCARG(uap, options); 226 if (linux_options & ~(LINUX_WAIT4_KNOWNFLAGS)) 227 return (EINVAL); 228 229 options = 0; 230 if (linux_options & LINUX_WAIT4_WNOHANG) 231 options |= WNOHANG; 232 if (linux_options & LINUX_WAIT4_WUNTRACED) 233 options |= WUNTRACED; 234 if (linux_options & LINUX_WAIT4_WCONTINUED) 235 options |= WCONTINUED; 236 if (linux_options & LINUX_WAIT4_WALL) 237 options |= WALLSIG; 238 if (linux_options & LINUX_WAIT4_WCLONE) 239 options |= WALTSIG; 240 # ifdef DIAGNOSTIC 241 if (linux_options & LINUX_WAIT4_WNOTHREAD) 242 printf("WARNING: %s: linux process %d.%d called " 243 "waitpid with __WNOTHREAD set!", 244 __FILE__, l->l_proc->p_pid, l->l_lid); 245 246 # endif 247 248 error = do_sys_wait(&pid, &status, options, 249 SCARG(uap, rusage) != NULL ? &ru : NULL); 250 251 retval[0] = pid; 252 if (pid == 0) 253 return error; 254 255 p = curproc; 256 mutex_enter(p->p_lock); 257 sigdelset(&p->p_sigpend.sp_set, SIGCHLD); /* XXXAD ksiginfo leak */ 258 mutex_exit(p->p_lock); 259 260 if (SCARG(uap, rusage) != NULL) { 261 rusage_to_rusage50(&ru, &ru50); 262 error = copyout(&ru, SCARG(uap, rusage), sizeof(ru)); 263 } 264 265 if (error == 0 && SCARG(uap, status) != NULL) { 266 status = bsd_to_linux_wstat(status); 267 error = copyout(&status, SCARG(uap, status), sizeof status); 268 } 269 270 return error; 271 } 272 273 /* 274 * Linux brk(2). Like native, but always return the new break value. 275 */ 276 int 277 linux_sys_brk(struct lwp *l, const struct linux_sys_brk_args *uap, register_t *retval) 278 { 279 /* { 280 syscallarg(char *) nsize; 281 } */ 282 struct proc *p = l->l_proc; 283 struct vmspace *vm = p->p_vmspace; 284 struct sys_obreak_args oba; 285 286 SCARG(&oba, nsize) = SCARG(uap, nsize); 287 288 (void) sys_obreak(l, &oba, retval); 289 retval[0] = (register_t)((char *)vm->vm_daddr + ptoa(vm->vm_dsize)); 290 return 0; 291 } 292 293 /* 294 * Implement the fs stat functions. Straightforward. 295 */ 296 int 297 linux_sys_statfs(struct lwp *l, const struct linux_sys_statfs_args *uap, register_t *retval) 298 { 299 /* { 300 syscallarg(const char *) path; 301 syscallarg(struct linux_statfs *) sp; 302 } */ 303 struct statvfs *sb; 304 struct linux_statfs ltmp; 305 int error; 306 307 sb = STATVFSBUF_GET(); 308 error = do_sys_pstatvfs(l, SCARG(uap, path), ST_WAIT, sb); 309 if (error == 0) { 310 bsd_to_linux_statfs(sb, <mp); 311 error = copyout(<mp, SCARG(uap, sp), sizeof ltmp); 312 } 313 STATVFSBUF_PUT(sb); 314 315 return error; 316 } 317 318 int 319 linux_sys_fstatfs(struct lwp *l, const struct linux_sys_fstatfs_args *uap, register_t *retval) 320 { 321 /* { 322 syscallarg(int) fd; 323 syscallarg(struct linux_statfs *) sp; 324 } */ 325 struct statvfs *sb; 326 struct linux_statfs ltmp; 327 int error; 328 329 sb = STATVFSBUF_GET(); 330 error = do_sys_fstatvfs(l, SCARG(uap, fd), ST_WAIT, sb); 331 if (error == 0) { 332 bsd_to_linux_statfs(sb, <mp); 333 error = copyout(<mp, SCARG(uap, sp), sizeof ltmp); 334 } 335 STATVFSBUF_PUT(sb); 336 337 return error; 338 } 339 340 /* 341 * uname(). Just copy the info from the various strings stored in the 342 * kernel, and put it in the Linux utsname structure. That structure 343 * is almost the same as the NetBSD one, only it has fields 65 characters 344 * long, and an extra domainname field. 345 */ 346 int 347 linux_sys_uname(struct lwp *l, const struct linux_sys_uname_args *uap, register_t *retval) 348 { 349 /* { 350 syscallarg(struct linux_utsname *) up; 351 } */ 352 struct linux_utsname luts; 353 354 strlcpy(luts.l_sysname, linux_sysname, sizeof(luts.l_sysname)); 355 strlcpy(luts.l_nodename, hostname, sizeof(luts.l_nodename)); 356 strlcpy(luts.l_release, linux_release, sizeof(luts.l_release)); 357 strlcpy(luts.l_version, linux_version, sizeof(luts.l_version)); 358 strlcpy(luts.l_machine, LINUX_UNAME_ARCH, sizeof(luts.l_machine)); 359 strlcpy(luts.l_domainname, domainname, sizeof(luts.l_domainname)); 360 361 return copyout(&luts, SCARG(uap, up), sizeof(luts)); 362 } 363 364 /* Used directly on: alpha, mips, ppc, sparc, sparc64 */ 365 /* Used indirectly on: arm, i386, m68k */ 366 367 /* 368 * New type Linux mmap call. 369 * Only called directly on machines with >= 6 free regs. 370 */ 371 int 372 linux_sys_mmap(struct lwp *l, const struct linux_sys_mmap_args *uap, register_t *retval) 373 { 374 /* { 375 syscallarg(unsigned long) addr; 376 syscallarg(size_t) len; 377 syscallarg(int) prot; 378 syscallarg(int) flags; 379 syscallarg(int) fd; 380 syscallarg(linux_off_t) offset; 381 } */ 382 383 if (SCARG(uap, offset) & PAGE_MASK) 384 return EINVAL; 385 386 return linux_mmap(l, uap, retval, SCARG(uap, offset)); 387 } 388 389 /* 390 * Guts of most architectures' mmap64() implementations. This shares 391 * its list of arguments with linux_sys_mmap(). 392 * 393 * The difference in linux_sys_mmap2() is that "offset" is actually 394 * (offset / pagesize), not an absolute byte count. This translation 395 * to pagesize offsets is done inside glibc between the mmap64() call 396 * point, and the actual syscall. 397 */ 398 int 399 linux_sys_mmap2(struct lwp *l, const struct linux_sys_mmap2_args *uap, register_t *retval) 400 { 401 /* { 402 syscallarg(unsigned long) addr; 403 syscallarg(size_t) len; 404 syscallarg(int) prot; 405 syscallarg(int) flags; 406 syscallarg(int) fd; 407 syscallarg(linux_off_t) offset; 408 } */ 409 410 return linux_mmap(l, uap, retval, 411 ((off_t)SCARG(uap, offset)) << PAGE_SHIFT); 412 } 413 414 /* 415 * Massage arguments and call system mmap(2). 416 */ 417 static int 418 linux_mmap(struct lwp *l, const struct linux_sys_mmap_args *uap, register_t *retval, off_t offset) 419 { 420 struct sys_mmap_args cma; 421 int error; 422 size_t mmoff=0; 423 424 linux_to_bsd_mmap_args(&cma, uap); 425 SCARG(&cma, pos) = offset; 426 427 if (SCARG(uap, flags) & LINUX_MAP_GROWSDOWN) { 428 /* 429 * Request for stack-like memory segment. On linux, this 430 * works by mmap()ping (small) segment, which is automatically 431 * extended when page fault happens below the currently 432 * allocated area. We emulate this by allocating (typically 433 * bigger) segment sized at current stack size limit, and 434 * offsetting the requested and returned address accordingly. 435 * Since physical pages are only allocated on-demand, this 436 * is effectively identical. 437 */ 438 rlim_t ssl = l->l_proc->p_rlimit[RLIMIT_STACK].rlim_cur; 439 440 if (SCARG(&cma, len) < ssl) { 441 /* Compute the address offset */ 442 mmoff = round_page(ssl) - SCARG(uap, len); 443 444 if (SCARG(&cma, addr)) 445 SCARG(&cma, addr) = (char *)SCARG(&cma, addr) - mmoff; 446 447 SCARG(&cma, len) = (size_t) ssl; 448 } 449 } 450 451 error = sys_mmap(l, &cma, retval); 452 if (error) 453 return (error); 454 455 /* Shift the returned address for stack-like segment if necessary */ 456 retval[0] += mmoff; 457 458 return (0); 459 } 460 461 static void 462 linux_to_bsd_mmap_args(struct sys_mmap_args *cma, const struct linux_sys_mmap_args *uap) 463 { 464 int flags = MAP_TRYFIXED, fl = SCARG(uap, flags); 465 466 flags |= cvtto_bsd_mask(fl, LINUX_MAP_SHARED, MAP_SHARED); 467 flags |= cvtto_bsd_mask(fl, LINUX_MAP_PRIVATE, MAP_PRIVATE); 468 flags |= cvtto_bsd_mask(fl, LINUX_MAP_FIXED, MAP_FIXED); 469 flags |= cvtto_bsd_mask(fl, LINUX_MAP_ANON, MAP_ANON); 470 flags |= cvtto_bsd_mask(fl, LINUX_MAP_LOCKED, MAP_WIRED); 471 /* XXX XAX ERH: Any other flags here? There are more defined... */ 472 473 SCARG(cma, addr) = (void *)SCARG(uap, addr); 474 SCARG(cma, len) = SCARG(uap, len); 475 SCARG(cma, prot) = SCARG(uap, prot); 476 if (SCARG(cma, prot) & VM_PROT_WRITE) /* XXX */ 477 SCARG(cma, prot) |= VM_PROT_READ; 478 SCARG(cma, flags) = flags; 479 SCARG(cma, fd) = flags & MAP_ANON ? -1 : SCARG(uap, fd); 480 SCARG(cma, PAD) = 0; 481 } 482 483 #define LINUX_MREMAP_MAYMOVE 1 484 #define LINUX_MREMAP_FIXED 2 485 486 int 487 linux_sys_mremap(struct lwp *l, const struct linux_sys_mremap_args *uap, register_t *retval) 488 { 489 /* { 490 syscallarg(void *) old_address; 491 syscallarg(size_t) old_size; 492 syscallarg(size_t) new_size; 493 syscallarg(u_long) flags; 494 } */ 495 496 struct proc *p; 497 struct vm_map *map; 498 vaddr_t oldva; 499 vaddr_t newva; 500 size_t oldsize; 501 size_t newsize; 502 int flags; 503 int uvmflags; 504 int error; 505 506 flags = SCARG(uap, flags); 507 oldva = (vaddr_t)SCARG(uap, old_address); 508 oldsize = round_page(SCARG(uap, old_size)); 509 newsize = round_page(SCARG(uap, new_size)); 510 if ((flags & ~(LINUX_MREMAP_FIXED|LINUX_MREMAP_MAYMOVE)) != 0) { 511 error = EINVAL; 512 goto done; 513 } 514 if ((flags & LINUX_MREMAP_FIXED) != 0) { 515 if ((flags & LINUX_MREMAP_MAYMOVE) == 0) { 516 error = EINVAL; 517 goto done; 518 } 519 #if 0 /* notyet */ 520 newva = SCARG(uap, new_address); 521 uvmflags = MAP_FIXED; 522 #else /* notyet */ 523 error = EOPNOTSUPP; 524 goto done; 525 #endif /* notyet */ 526 } else if ((flags & LINUX_MREMAP_MAYMOVE) != 0) { 527 uvmflags = 0; 528 } else { 529 newva = oldva; 530 uvmflags = MAP_FIXED; 531 } 532 p = l->l_proc; 533 map = &p->p_vmspace->vm_map; 534 error = uvm_mremap(map, oldva, oldsize, map, &newva, newsize, p, 535 uvmflags); 536 537 done: 538 *retval = (error != 0) ? 0 : (register_t)newva; 539 return error; 540 } 541 542 #ifdef USRSTACK 543 int 544 linux_sys_mprotect(struct lwp *l, const struct linux_sys_mprotect_args *uap, register_t *retval) 545 { 546 /* { 547 syscallarg(const void *) start; 548 syscallarg(unsigned long) len; 549 syscallarg(int) prot; 550 } */ 551 struct vm_map_entry *entry; 552 struct vm_map *map; 553 struct proc *p; 554 vaddr_t end, start, len, stacklim; 555 int prot, grows; 556 557 start = (vaddr_t)SCARG(uap, start); 558 len = round_page(SCARG(uap, len)); 559 prot = SCARG(uap, prot); 560 grows = prot & (LINUX_PROT_GROWSDOWN | LINUX_PROT_GROWSUP); 561 prot &= ~grows; 562 end = start + len; 563 564 if (start & PAGE_MASK) 565 return EINVAL; 566 if (end < start) 567 return EINVAL; 568 if (end == start) 569 return 0; 570 571 if (prot & ~(PROT_READ | PROT_WRITE | PROT_EXEC)) 572 return EINVAL; 573 if (grows == (LINUX_PROT_GROWSDOWN | LINUX_PROT_GROWSUP)) 574 return EINVAL; 575 576 p = l->l_proc; 577 map = &p->p_vmspace->vm_map; 578 vm_map_lock(map); 579 # ifdef notdef 580 VM_MAP_RANGE_CHECK(map, start, end); 581 # endif 582 if (!uvm_map_lookup_entry(map, start, &entry) || entry->start > start) { 583 vm_map_unlock(map); 584 return ENOMEM; 585 } 586 587 /* 588 * Approximate the behaviour of PROT_GROWS{DOWN,UP}. 589 */ 590 591 stacklim = (vaddr_t)p->p_limit->pl_rlimit[RLIMIT_STACK].rlim_cur; 592 if (grows & LINUX_PROT_GROWSDOWN) { 593 if (USRSTACK - stacklim <= start && start < USRSTACK) { 594 start = USRSTACK - stacklim; 595 } else { 596 start = entry->start; 597 } 598 } else if (grows & LINUX_PROT_GROWSUP) { 599 if (USRSTACK <= end && end < USRSTACK + stacklim) { 600 end = USRSTACK + stacklim; 601 } else { 602 end = entry->end; 603 } 604 } 605 vm_map_unlock(map); 606 return uvm_map_protect(map, start, end, prot, FALSE); 607 } 608 #endif /* USRSTACK */ 609 610 /* 611 * This code is partly stolen from src/lib/libc/compat-43/times.c 612 */ 613 614 #define CONVTCK(r) (r.tv_sec * hz + r.tv_usec / (1000000 / hz)) 615 616 int 617 linux_sys_times(struct lwp *l, const struct linux_sys_times_args *uap, register_t *retval) 618 { 619 /* { 620 syscallarg(struct times *) tms; 621 } */ 622 struct proc *p = l->l_proc; 623 struct timeval t; 624 int error; 625 626 if (SCARG(uap, tms)) { 627 struct linux_tms ltms; 628 struct rusage ru; 629 630 mutex_enter(p->p_lock); 631 calcru(p, &ru.ru_utime, &ru.ru_stime, NULL, NULL); 632 ltms.ltms_utime = CONVTCK(ru.ru_utime); 633 ltms.ltms_stime = CONVTCK(ru.ru_stime); 634 ltms.ltms_cutime = CONVTCK(p->p_stats->p_cru.ru_utime); 635 ltms.ltms_cstime = CONVTCK(p->p_stats->p_cru.ru_stime); 636 mutex_exit(p->p_lock); 637 638 if ((error = copyout(<ms, SCARG(uap, tms), sizeof ltms))) 639 return error; 640 } 641 642 getmicrouptime(&t); 643 644 retval[0] = ((linux_clock_t)(CONVTCK(t))); 645 return 0; 646 } 647 648 #undef CONVTCK 649 650 /* 651 * Linux 'readdir' call. This code is mostly taken from the 652 * SunOS getdents call (see compat/sunos/sunos_misc.c), though 653 * an attempt has been made to keep it a little cleaner (failing 654 * miserably, because of the cruft needed if count 1 is passed). 655 * 656 * The d_off field should contain the offset of the next valid entry, 657 * but in Linux it has the offset of the entry itself. We emulate 658 * that bug here. 659 * 660 * Read in BSD-style entries, convert them, and copy them out. 661 * 662 * Note that this doesn't handle union-mounted filesystems. 663 */ 664 int 665 linux_sys_getdents(struct lwp *l, const struct linux_sys_getdents_args *uap, register_t *retval) 666 { 667 /* { 668 syscallarg(int) fd; 669 syscallarg(struct linux_dirent *) dent; 670 syscallarg(unsigned int) count; 671 } */ 672 struct dirent *bdp; 673 struct vnode *vp; 674 char *inp, *tbuf; /* BSD-format */ 675 int len, reclen; /* BSD-format */ 676 char *outp; /* Linux-format */ 677 int resid, linux_reclen = 0; /* Linux-format */ 678 struct file *fp; 679 struct uio auio; 680 struct iovec aiov; 681 struct linux_dirent idb; 682 off_t off; /* true file offset */ 683 int buflen, error, eofflag, nbytes, oldcall; 684 struct vattr va; 685 off_t *cookiebuf = NULL, *cookie; 686 int ncookies; 687 688 /* fd_getvnode() will use the descriptor for us */ 689 if ((error = fd_getvnode(SCARG(uap, fd), &fp)) != 0) 690 return (error); 691 692 if ((fp->f_flag & FREAD) == 0) { 693 error = EBADF; 694 goto out1; 695 } 696 697 vp = (struct vnode *)fp->f_data; 698 if (vp->v_type != VDIR) { 699 error = ENOTDIR; 700 goto out1; 701 } 702 703 vn_lock(vp, LK_SHARED | LK_RETRY); 704 error = VOP_GETATTR(vp, &va, l->l_cred); 705 VOP_UNLOCK(vp); 706 if (error) 707 goto out1; 708 709 nbytes = SCARG(uap, count); 710 if (nbytes == 1) { /* emulating old, broken behaviour */ 711 nbytes = sizeof (idb); 712 buflen = max(va.va_blocksize, nbytes); 713 oldcall = 1; 714 } else { 715 buflen = min(MAXBSIZE, nbytes); 716 if (buflen < va.va_blocksize) 717 buflen = va.va_blocksize; 718 oldcall = 0; 719 } 720 tbuf = malloc(buflen, M_TEMP, M_WAITOK); 721 722 vn_lock(vp, LK_EXCLUSIVE | LK_RETRY); 723 off = fp->f_offset; 724 again: 725 aiov.iov_base = tbuf; 726 aiov.iov_len = buflen; 727 auio.uio_iov = &aiov; 728 auio.uio_iovcnt = 1; 729 auio.uio_rw = UIO_READ; 730 auio.uio_resid = buflen; 731 auio.uio_offset = off; 732 UIO_SETUP_SYSSPACE(&auio); 733 /* 734 * First we read into the malloc'ed buffer, then 735 * we massage it into user space, one record at a time. 736 */ 737 error = VOP_READDIR(vp, &auio, fp->f_cred, &eofflag, &cookiebuf, 738 &ncookies); 739 if (error) 740 goto out; 741 742 inp = tbuf; 743 outp = (void *)SCARG(uap, dent); 744 resid = nbytes; 745 if ((len = buflen - auio.uio_resid) == 0) 746 goto eof; 747 748 for (cookie = cookiebuf; len > 0; len -= reclen) { 749 bdp = (struct dirent *)inp; 750 reclen = bdp->d_reclen; 751 if (reclen & 3) 752 panic("linux_readdir"); 753 if (bdp->d_fileno == 0) { 754 inp += reclen; /* it is a hole; squish it out */ 755 if (cookie) 756 off = *cookie++; 757 else 758 off += reclen; 759 continue; 760 } 761 linux_reclen = LINUX_RECLEN(&idb, bdp->d_namlen); 762 if (reclen > len || resid < linux_reclen) { 763 /* entry too big for buffer, so just stop */ 764 outp++; 765 break; 766 } 767 /* 768 * Massage in place to make a Linux-shaped dirent (otherwise 769 * we have to worry about touching user memory outside of 770 * the copyout() call). 771 */ 772 idb.d_ino = bdp->d_fileno; 773 /* 774 * The old readdir() call misuses the offset and reclen fields. 775 */ 776 if (oldcall) { 777 idb.d_off = (linux_off_t)linux_reclen; 778 idb.d_reclen = (u_short)bdp->d_namlen; 779 } else { 780 if (sizeof (idb.d_off) <= 4 && (off >> 32) != 0) { 781 compat_offseterr(vp, "linux_getdents"); 782 error = EINVAL; 783 goto out; 784 } 785 idb.d_off = (linux_off_t)off; 786 idb.d_reclen = (u_short)linux_reclen; 787 /* Linux puts d_type at the end of each record */ 788 *((char *)&idb + idb.d_reclen - 1) = bdp->d_type; 789 } 790 strcpy(idb.d_name, bdp->d_name); 791 if ((error = copyout((void *)&idb, outp, linux_reclen))) 792 goto out; 793 /* advance past this real entry */ 794 inp += reclen; 795 if (cookie) 796 off = *cookie++; /* each entry points to itself */ 797 else 798 off += reclen; 799 /* advance output past Linux-shaped entry */ 800 outp += linux_reclen; 801 resid -= linux_reclen; 802 if (oldcall) 803 break; 804 } 805 806 /* if we squished out the whole block, try again */ 807 if (outp == (void *)SCARG(uap, dent)) { 808 if (cookiebuf) 809 free(cookiebuf, M_TEMP); 810 cookiebuf = NULL; 811 goto again; 812 } 813 fp->f_offset = off; /* update the vnode offset */ 814 815 if (oldcall) 816 nbytes = resid + linux_reclen; 817 818 eof: 819 *retval = nbytes - resid; 820 out: 821 VOP_UNLOCK(vp); 822 if (cookiebuf) 823 free(cookiebuf, M_TEMP); 824 free(tbuf, M_TEMP); 825 out1: 826 fd_putfile(SCARG(uap, fd)); 827 return error; 828 } 829 830 /* 831 * Even when just using registers to pass arguments to syscalls you can 832 * have 5 of them on the i386. So this newer version of select() does 833 * this. 834 */ 835 int 836 linux_sys_select(struct lwp *l, const struct linux_sys_select_args *uap, register_t *retval) 837 { 838 /* { 839 syscallarg(int) nfds; 840 syscallarg(fd_set *) readfds; 841 syscallarg(fd_set *) writefds; 842 syscallarg(fd_set *) exceptfds; 843 syscallarg(struct timeval50 *) timeout; 844 } */ 845 846 return linux_select1(l, retval, SCARG(uap, nfds), SCARG(uap, readfds), 847 SCARG(uap, writefds), SCARG(uap, exceptfds), 848 (struct linux_timeval *)SCARG(uap, timeout)); 849 } 850 851 /* 852 * Common code for the old and new versions of select(). A couple of 853 * things are important: 854 * 1) return the amount of time left in the 'timeout' parameter 855 * 2) select never returns ERESTART on Linux, always return EINTR 856 */ 857 int 858 linux_select1(struct lwp *l, register_t *retval, int nfds, fd_set *readfds, 859 fd_set *writefds, fd_set *exceptfds, struct linux_timeval *timeout) 860 { 861 struct timespec ts0, ts1, uts, *ts = NULL; 862 struct linux_timeval ltv; 863 int error; 864 865 /* 866 * Store current time for computation of the amount of 867 * time left. 868 */ 869 if (timeout) { 870 if ((error = copyin(timeout, <v, sizeof(ltv)))) 871 return error; 872 uts.tv_sec = ltv.tv_sec; 873 uts.tv_nsec = ltv.tv_usec * 1000; 874 if (itimespecfix(&uts)) { 875 /* 876 * The timeval was invalid. Convert it to something 877 * valid that will act as it does under Linux. 878 */ 879 uts.tv_sec += uts.tv_nsec / 1000000000; 880 uts.tv_nsec %= 1000000000; 881 if (uts.tv_nsec < 0) { 882 uts.tv_sec -= 1; 883 uts.tv_nsec += 1000000000; 884 } 885 if (uts.tv_sec < 0) 886 timespecclear(&uts); 887 } 888 ts = &uts; 889 nanotime(&ts0); 890 } 891 892 error = selcommon(retval, nfds, readfds, writefds, exceptfds, ts, NULL); 893 894 if (error) { 895 /* 896 * See fs/select.c in the Linux kernel. Without this, 897 * Maelstrom doesn't work. 898 */ 899 if (error == ERESTART) 900 error = EINTR; 901 return error; 902 } 903 904 if (timeout) { 905 if (*retval) { 906 /* 907 * Compute how much time was left of the timeout, 908 * by subtracting the current time and the time 909 * before we started the call, and subtracting 910 * that result from the user-supplied value. 911 */ 912 nanotime(&ts1); 913 timespecsub(&ts1, &ts0, &ts1); 914 timespecsub(&uts, &ts1, &uts); 915 if (uts.tv_sec < 0) 916 timespecclear(&uts); 917 } else 918 timespecclear(&uts); 919 ltv.tv_sec = uts.tv_sec; 920 ltv.tv_usec = uts.tv_nsec / 1000; 921 if ((error = copyout(<v, timeout, sizeof(ltv)))) 922 return error; 923 } 924 925 return 0; 926 } 927 928 /* 929 * Derived from FreeBSD's sys/compat/linux/linux_misc.c:linux_pselect6() 930 * which was contributed by Dmitry Chagin 931 * https://svnweb.freebsd.org/base?view=revision&revision=283403 932 */ 933 int 934 linux_sys_pselect6(struct lwp *l, 935 const struct linux_sys_pselect6_args *uap, register_t *retval) 936 { 937 /* { 938 syscallarg(int) nfds; 939 syscallarg(fd_set *) readfds; 940 syscallarg(fd_set *) writefds; 941 syscallarg(fd_set *) exceptfds; 942 syscallarg(struct timespec *) timeout; 943 syscallarg(linux_sized_sigset_t *) ss; 944 } */ 945 struct timespec uts, ts0, ts1, *tsp; 946 linux_sized_sigset_t lsss; 947 struct linux_timespec lts; 948 linux_sigset_t lss; 949 sigset_t *ssp; 950 sigset_t ss; 951 int error; 952 953 ssp = NULL; 954 if (SCARG(uap, ss) != NULL) { 955 if ((error = copyin(SCARG(uap, ss), &lsss, sizeof(lsss))) != 0) 956 return (error); 957 if (lsss.ss_len != sizeof(lss)) 958 return (EINVAL); 959 if (lsss.ss != NULL) { 960 if ((error = copyin(lsss.ss, &lss, sizeof(lss))) != 0) 961 return (error); 962 linux_to_native_sigset(&ss, &lss); 963 ssp = &ss; 964 } 965 } 966 967 if (SCARG(uap, timeout) != NULL) { 968 error = copyin(SCARG(uap, timeout), <s, sizeof(lts)); 969 if (error != 0) 970 return (error); 971 linux_to_native_timespec(&uts, <s); 972 973 if (itimespecfix(&uts)) 974 return (EINVAL); 975 976 nanotime(&ts0); 977 tsp = &uts; 978 } else { 979 tsp = NULL; 980 } 981 982 error = selcommon(retval, SCARG(uap, nfds), SCARG(uap, readfds), 983 SCARG(uap, writefds), SCARG(uap, exceptfds), tsp, ssp); 984 985 if (error == 0 && tsp != NULL) { 986 if (retval != 0) { 987 /* 988 * Compute how much time was left of the timeout, 989 * by subtracting the current time and the time 990 * before we started the call, and subtracting 991 * that result from the user-supplied value. 992 */ 993 nanotime(&ts1); 994 timespecsub(&ts1, &ts0, &ts1); 995 timespecsub(&uts, &ts1, &uts); 996 if (uts.tv_sec < 0) 997 timespecclear(&uts); 998 } else { 999 timespecclear(&uts); 1000 } 1001 1002 native_to_linux_timespec(<s, &uts); 1003 error = copyout(<s, SCARG(uap, timeout), sizeof(lts)); 1004 } 1005 1006 return (error); 1007 } 1008 1009 int 1010 linux_sys_ppoll(struct lwp *l, 1011 const struct linux_sys_ppoll_args *uap, register_t *retval) 1012 { 1013 /* { 1014 syscallarg(struct pollfd *) fds; 1015 syscallarg(u_int) nfds; 1016 syscallarg(struct linux_timespec *) timeout; 1017 syscallarg(linux_sigset_t *) sigset; 1018 } */ 1019 struct linux_timespec lts0, *lts; 1020 struct timespec ts0, *ts = NULL; 1021 linux_sigset_t lsigmask0, *lsigmask; 1022 sigset_t sigmask0, *sigmask = NULL; 1023 int error; 1024 1025 lts = SCARG(uap, timeout); 1026 if (lts) { 1027 if ((error = copyin(lts, <s0, sizeof(lts0))) != 0) 1028 return error; 1029 linux_to_native_timespec(&ts0, <s0); 1030 ts = &ts0; 1031 } 1032 1033 lsigmask = SCARG(uap, sigset); 1034 if (lsigmask) { 1035 if ((error = copyin(lsigmask, &lsigmask0, sizeof(lsigmask0)))) 1036 return error; 1037 linux_to_native_sigset(&sigmask0, &lsigmask0); 1038 sigmask = &sigmask0; 1039 } 1040 1041 return pollcommon(retval, SCARG(uap, fds), SCARG(uap, nfds), 1042 ts, sigmask); 1043 } 1044 1045 /* 1046 * Set the 'personality' (emulation mode) for the current process. Only 1047 * accept the Linux personality here (0). This call is needed because 1048 * the Linux ELF crt0 issues it in an ugly kludge to make sure that 1049 * ELF binaries run in Linux mode, not SVR4 mode. 1050 */ 1051 int 1052 linux_sys_personality(struct lwp *l, const struct linux_sys_personality_args *uap, register_t *retval) 1053 { 1054 /* { 1055 syscallarg(unsigned long) per; 1056 } */ 1057 struct linux_emuldata *led; 1058 int per; 1059 1060 per = SCARG(uap, per); 1061 led = l->l_emuldata; 1062 if (per == LINUX_PER_QUERY) { 1063 retval[0] = led->led_personality; 1064 return 0; 1065 } 1066 1067 switch (per & LINUX_PER_MASK) { 1068 case LINUX_PER_LINUX: 1069 case LINUX_PER_LINUX32: 1070 led->led_personality = per; 1071 break; 1072 1073 default: 1074 return EINVAL; 1075 } 1076 1077 retval[0] = per; 1078 return 0; 1079 } 1080 1081 /* 1082 * We have nonexistent fsuid equal to uid. 1083 * If modification is requested, refuse. 1084 */ 1085 int 1086 linux_sys_setfsuid(struct lwp *l, const struct linux_sys_setfsuid_args *uap, register_t *retval) 1087 { 1088 /* { 1089 syscallarg(uid_t) uid; 1090 } */ 1091 uid_t uid; 1092 1093 uid = SCARG(uap, uid); 1094 if (kauth_cred_getuid(l->l_cred) != uid) 1095 return sys_nosys(l, uap, retval); 1096 1097 *retval = uid; 1098 return 0; 1099 } 1100 1101 int 1102 linux_sys_setfsgid(struct lwp *l, const struct linux_sys_setfsgid_args *uap, register_t *retval) 1103 { 1104 /* { 1105 syscallarg(gid_t) gid; 1106 } */ 1107 gid_t gid; 1108 1109 gid = SCARG(uap, gid); 1110 if (kauth_cred_getgid(l->l_cred) != gid) 1111 return sys_nosys(l, uap, retval); 1112 1113 *retval = gid; 1114 return 0; 1115 } 1116 1117 int 1118 linux_sys_setresuid(struct lwp *l, const struct linux_sys_setresuid_args *uap, register_t *retval) 1119 { 1120 /* { 1121 syscallarg(uid_t) ruid; 1122 syscallarg(uid_t) euid; 1123 syscallarg(uid_t) suid; 1124 } */ 1125 1126 /* 1127 * Note: These checks are a little different than the NetBSD 1128 * setreuid(2) call performs. This precisely follows the 1129 * behavior of the Linux kernel. 1130 */ 1131 1132 return do_setresuid(l, SCARG(uap, ruid), SCARG(uap, euid), 1133 SCARG(uap, suid), 1134 ID_R_EQ_R | ID_R_EQ_E | ID_R_EQ_S | 1135 ID_E_EQ_R | ID_E_EQ_E | ID_E_EQ_S | 1136 ID_S_EQ_R | ID_S_EQ_E | ID_S_EQ_S ); 1137 } 1138 1139 int 1140 linux_sys_getresuid(struct lwp *l, const struct linux_sys_getresuid_args *uap, register_t *retval) 1141 { 1142 /* { 1143 syscallarg(uid_t *) ruid; 1144 syscallarg(uid_t *) euid; 1145 syscallarg(uid_t *) suid; 1146 } */ 1147 kauth_cred_t pc = l->l_cred; 1148 int error; 1149 uid_t uid; 1150 1151 /* 1152 * Linux copies these values out to userspace like so: 1153 * 1154 * 1. Copy out ruid. 1155 * 2. If that succeeds, copy out euid. 1156 * 3. If both of those succeed, copy out suid. 1157 */ 1158 uid = kauth_cred_getuid(pc); 1159 if ((error = copyout(&uid, SCARG(uap, ruid), sizeof(uid_t))) != 0) 1160 return (error); 1161 1162 uid = kauth_cred_geteuid(pc); 1163 if ((error = copyout(&uid, SCARG(uap, euid), sizeof(uid_t))) != 0) 1164 return (error); 1165 1166 uid = kauth_cred_getsvuid(pc); 1167 1168 return (copyout(&uid, SCARG(uap, suid), sizeof(uid_t))); 1169 } 1170 1171 int 1172 linux_sys_ptrace(struct lwp *l, const struct linux_sys_ptrace_args *uap, register_t *retval) 1173 { 1174 /* { 1175 i386, m68k, powerpc: T=int 1176 alpha, amd64: T=long 1177 syscallarg(T) request; 1178 syscallarg(T) pid; 1179 syscallarg(T) addr; 1180 syscallarg(T) data; 1181 } */ 1182 const int *ptr; 1183 int request; 1184 int error; 1185 1186 ptr = linux_ptrace_request_map; 1187 request = SCARG(uap, request); 1188 while (*ptr != -1) 1189 if (*ptr++ == request) { 1190 struct sys_ptrace_args pta; 1191 1192 SCARG(&pta, req) = *ptr; 1193 SCARG(&pta, pid) = SCARG(uap, pid); 1194 SCARG(&pta, addr) = (void *)SCARG(uap, addr); 1195 SCARG(&pta, data) = SCARG(uap, data); 1196 1197 /* 1198 * Linux ptrace(PTRACE_CONT, pid, 0, 0) means actually 1199 * to continue where the process left off previously. 1200 * The same thing is achieved by addr == (void *) 1 1201 * on NetBSD, so rewrite 'addr' appropriately. 1202 */ 1203 if (request == LINUX_PTRACE_CONT && SCARG(uap, addr)==0) 1204 SCARG(&pta, addr) = (void *) 1; 1205 1206 error = sysent[SYS_ptrace].sy_call(l, &pta, retval); 1207 if (error) 1208 return error; 1209 switch (request) { 1210 case LINUX_PTRACE_PEEKTEXT: 1211 case LINUX_PTRACE_PEEKDATA: 1212 error = copyout (retval, 1213 (void *)SCARG(uap, data), 1214 sizeof *retval); 1215 *retval = SCARG(uap, data); 1216 break; 1217 default: 1218 break; 1219 } 1220 return error; 1221 } 1222 else 1223 ptr++; 1224 1225 return LINUX_SYS_PTRACE_ARCH(l, uap, retval); 1226 } 1227 1228 int 1229 linux_sys_reboot(struct lwp *l, const struct linux_sys_reboot_args *uap, register_t *retval) 1230 { 1231 /* { 1232 syscallarg(int) magic1; 1233 syscallarg(int) magic2; 1234 syscallarg(int) cmd; 1235 syscallarg(void *) arg; 1236 } */ 1237 struct sys_reboot_args /* { 1238 syscallarg(int) opt; 1239 syscallarg(char *) bootstr; 1240 } */ sra; 1241 int error; 1242 1243 if ((error = kauth_authorize_system(l->l_cred, 1244 KAUTH_SYSTEM_REBOOT, 0, NULL, NULL, NULL)) != 0) 1245 return(error); 1246 1247 if (SCARG(uap, magic1) != LINUX_REBOOT_MAGIC1) 1248 return(EINVAL); 1249 if (SCARG(uap, magic2) != LINUX_REBOOT_MAGIC2 && 1250 SCARG(uap, magic2) != LINUX_REBOOT_MAGIC2A && 1251 SCARG(uap, magic2) != LINUX_REBOOT_MAGIC2B) 1252 return(EINVAL); 1253 1254 switch ((unsigned long)SCARG(uap, cmd)) { 1255 case LINUX_REBOOT_CMD_RESTART: 1256 SCARG(&sra, opt) = RB_AUTOBOOT; 1257 break; 1258 case LINUX_REBOOT_CMD_HALT: 1259 SCARG(&sra, opt) = RB_HALT; 1260 break; 1261 case LINUX_REBOOT_CMD_POWER_OFF: 1262 SCARG(&sra, opt) = RB_HALT|RB_POWERDOWN; 1263 break; 1264 case LINUX_REBOOT_CMD_RESTART2: 1265 /* Reboot with an argument. */ 1266 SCARG(&sra, opt) = RB_AUTOBOOT|RB_STRING; 1267 SCARG(&sra, bootstr) = SCARG(uap, arg); 1268 break; 1269 case LINUX_REBOOT_CMD_CAD_ON: 1270 return(EINVAL); /* We don't implement ctrl-alt-delete */ 1271 case LINUX_REBOOT_CMD_CAD_OFF: 1272 return(0); 1273 default: 1274 return(EINVAL); 1275 } 1276 1277 return(sys_reboot(l, &sra, retval)); 1278 } 1279 1280 /* 1281 * Copy of compat_12_sys_swapon(). 1282 */ 1283 int 1284 linux_sys_swapon(struct lwp *l, const struct linux_sys_swapon_args *uap, register_t *retval) 1285 { 1286 /* { 1287 syscallarg(const char *) name; 1288 } */ 1289 struct sys_swapctl_args ua; 1290 1291 SCARG(&ua, cmd) = SWAP_ON; 1292 SCARG(&ua, arg) = (void *)__UNCONST(SCARG(uap, name)); 1293 SCARG(&ua, misc) = 0; /* priority */ 1294 return (sys_swapctl(l, &ua, retval)); 1295 } 1296 1297 /* 1298 * Stop swapping to the file or block device specified by path. 1299 */ 1300 int 1301 linux_sys_swapoff(struct lwp *l, const struct linux_sys_swapoff_args *uap, register_t *retval) 1302 { 1303 /* { 1304 syscallarg(const char *) path; 1305 } */ 1306 struct sys_swapctl_args ua; 1307 1308 SCARG(&ua, cmd) = SWAP_OFF; 1309 SCARG(&ua, arg) = __UNCONST(SCARG(uap, path)); /*XXXUNCONST*/ 1310 return (sys_swapctl(l, &ua, retval)); 1311 } 1312 1313 /* 1314 * Copy of compat_09_sys_setdomainname() 1315 */ 1316 /* ARGSUSED */ 1317 int 1318 linux_sys_setdomainname(struct lwp *l, const struct linux_sys_setdomainname_args *uap, register_t *retval) 1319 { 1320 /* { 1321 syscallarg(char *) domainname; 1322 syscallarg(int) len; 1323 } */ 1324 int name[2]; 1325 1326 name[0] = CTL_KERN; 1327 name[1] = KERN_DOMAINNAME; 1328 return (old_sysctl(&name[0], 2, 0, 0, SCARG(uap, domainname), 1329 SCARG(uap, len), l)); 1330 } 1331 1332 /* 1333 * sysinfo() 1334 */ 1335 /* ARGSUSED */ 1336 int 1337 linux_sys_sysinfo(struct lwp *l, const struct linux_sys_sysinfo_args *uap, register_t *retval) 1338 { 1339 /* { 1340 syscallarg(struct linux_sysinfo *) arg; 1341 } */ 1342 struct linux_sysinfo si; 1343 struct loadavg *la; 1344 1345 si.uptime = time_uptime; 1346 la = &averunnable; 1347 si.loads[0] = la->ldavg[0] * LINUX_SYSINFO_LOADS_SCALE / la->fscale; 1348 si.loads[1] = la->ldavg[1] * LINUX_SYSINFO_LOADS_SCALE / la->fscale; 1349 si.loads[2] = la->ldavg[2] * LINUX_SYSINFO_LOADS_SCALE / la->fscale; 1350 si.totalram = ctob((u_long)physmem); 1351 si.freeram = (u_long)uvmexp.free * uvmexp.pagesize; 1352 si.sharedram = 0; /* XXX */ 1353 si.bufferram = (u_long)uvmexp.filepages * uvmexp.pagesize; 1354 si.totalswap = (u_long)uvmexp.swpages * uvmexp.pagesize; 1355 si.freeswap = 1356 (u_long)(uvmexp.swpages - uvmexp.swpginuse) * uvmexp.pagesize; 1357 si.procs = nprocs; 1358 1359 /* The following are only present in newer Linux kernels. */ 1360 si.totalbig = 0; 1361 si.freebig = 0; 1362 si.mem_unit = 1; 1363 1364 return (copyout(&si, SCARG(uap, arg), sizeof si)); 1365 } 1366 1367 int 1368 linux_sys_getrlimit(struct lwp *l, const struct linux_sys_getrlimit_args *uap, register_t *retval) 1369 { 1370 /* { 1371 syscallarg(int) which; 1372 # ifdef LINUX_LARGEFILE64 1373 syscallarg(struct rlimit *) rlp; 1374 # else 1375 syscallarg(struct orlimit *) rlp; 1376 # endif 1377 } */ 1378 # ifdef LINUX_LARGEFILE64 1379 struct rlimit orl; 1380 # else 1381 struct orlimit orl; 1382 # endif 1383 int which; 1384 1385 which = linux_to_bsd_limit(SCARG(uap, which)); 1386 if (which < 0) 1387 return -which; 1388 1389 bsd_to_linux_rlimit(&orl, &l->l_proc->p_rlimit[which]); 1390 1391 return copyout(&orl, SCARG(uap, rlp), sizeof(orl)); 1392 } 1393 1394 int 1395 linux_sys_setrlimit(struct lwp *l, const struct linux_sys_setrlimit_args *uap, register_t *retval) 1396 { 1397 /* { 1398 syscallarg(int) which; 1399 # ifdef LINUX_LARGEFILE64 1400 syscallarg(struct rlimit *) rlp; 1401 # else 1402 syscallarg(struct orlimit *) rlp; 1403 # endif 1404 } */ 1405 struct rlimit rl; 1406 # ifdef LINUX_LARGEFILE64 1407 struct rlimit orl; 1408 # else 1409 struct orlimit orl; 1410 # endif 1411 int error; 1412 int which; 1413 1414 if ((error = copyin(SCARG(uap, rlp), &orl, sizeof(orl))) != 0) 1415 return error; 1416 1417 which = linux_to_bsd_limit(SCARG(uap, which)); 1418 if (which < 0) 1419 return -which; 1420 1421 linux_to_bsd_rlimit(&rl, &orl); 1422 return dosetrlimit(l, l->l_proc, which, &rl); 1423 } 1424 1425 # if !defined(__mips__) && !defined(__amd64__) 1426 /* XXX: this doesn't look 100% common, at least mips doesn't have it */ 1427 int 1428 linux_sys_ugetrlimit(struct lwp *l, const struct linux_sys_ugetrlimit_args *uap, register_t *retval) 1429 { 1430 return linux_sys_getrlimit(l, (const void *)uap, retval); 1431 } 1432 # endif 1433 1434 /* 1435 * This gets called for unsupported syscalls. The difference to sys_nosys() 1436 * is that process does not get SIGSYS, the call just returns with ENOSYS. 1437 * This is the way Linux does it and glibc depends on this behaviour. 1438 */ 1439 int 1440 linux_sys_nosys(struct lwp *l, const void *v, register_t *retval) 1441 { 1442 return (ENOSYS); 1443 } 1444 1445 int 1446 linux_sys_getpriority(struct lwp *l, const struct linux_sys_getpriority_args *uap, register_t *retval) 1447 { 1448 /* { 1449 syscallarg(int) which; 1450 syscallarg(int) who; 1451 } */ 1452 struct sys_getpriority_args bsa; 1453 int error; 1454 1455 SCARG(&bsa, which) = SCARG(uap, which); 1456 SCARG(&bsa, who) = SCARG(uap, who); 1457 1458 if ((error = sys_getpriority(l, &bsa, retval))) 1459 return error; 1460 1461 *retval = NZERO - *retval; 1462 1463 return 0; 1464 } 1465 1466 int 1467 linux_do_sys_utimensat(struct lwp *l, int fd, const char *path, struct timespec *tsp, int flags, register_t *retval) 1468 { 1469 int follow, error; 1470 1471 follow = (flags & LINUX_AT_SYMLINK_NOFOLLOW) ? NOFOLLOW : FOLLOW; 1472 1473 if (path == NULL && fd != AT_FDCWD) { 1474 file_t *fp; 1475 1476 /* fd_getvnode() will use the descriptor for us */ 1477 if ((error = fd_getvnode(fd, &fp)) != 0) 1478 return error; 1479 error = do_sys_utimensat(l, AT_FDCWD, fp->f_data, NULL, 0, 1480 tsp, UIO_SYSSPACE); 1481 fd_putfile(fd); 1482 return error; 1483 } 1484 1485 return do_sys_utimensat(l, fd, NULL, path, follow, tsp, UIO_SYSSPACE); 1486 } 1487 1488 int 1489 linux_sys_utimensat(struct lwp *l, const struct linux_sys_utimensat_args *uap, 1490 register_t *retval) 1491 { 1492 /* { 1493 syscallarg(int) fd; 1494 syscallarg(const char *) path; 1495 syscallarg(const struct linux_timespec *) times; 1496 syscallarg(int) flag; 1497 } */ 1498 int error; 1499 struct linux_timespec lts[2]; 1500 struct timespec *tsp = NULL, ts[2]; 1501 1502 if (SCARG(uap, times)) { 1503 error = copyin(SCARG(uap, times), <s, sizeof(lts)); 1504 if (error != 0) 1505 return error; 1506 linux_to_native_timespec(&ts[0], <s[0]); 1507 linux_to_native_timespec(&ts[1], <s[1]); 1508 tsp = ts; 1509 } 1510 1511 return linux_do_sys_utimensat(l, SCARG(uap, fd), SCARG(uap, path), 1512 tsp, SCARG(uap, flag), retval); 1513 } 1514