1 /* $NetBSD: kern_exec.c,v 1.495 2020/04/06 08:20:05 kamil Exp $ */ 2 3 /*- 4 * Copyright (c) 2008, 2019, 2020 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Andrew Doran. 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 /*- 33 * Copyright (C) 1993, 1994, 1996 Christopher G. Demetriou 34 * Copyright (C) 1992 Wolfgang Solfrank. 35 * Copyright (C) 1992 TooLs GmbH. 36 * All rights reserved. 37 * 38 * Redistribution and use in source and binary forms, with or without 39 * modification, are permitted provided that the following conditions 40 * are met: 41 * 1. Redistributions of source code must retain the above copyright 42 * notice, this list of conditions and the following disclaimer. 43 * 2. Redistributions in binary form must reproduce the above copyright 44 * notice, this list of conditions and the following disclaimer in the 45 * documentation and/or other materials provided with the distribution. 46 * 3. All advertising materials mentioning features or use of this software 47 * must display the following acknowledgement: 48 * This product includes software developed by TooLs GmbH. 49 * 4. The name of TooLs GmbH may not be used to endorse or promote products 50 * derived from this software without specific prior written permission. 51 * 52 * THIS SOFTWARE IS PROVIDED BY TOOLS GMBH ``AS IS'' AND ANY EXPRESS OR 53 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 54 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. 55 * IN NO EVENT SHALL TOOLS GMBH BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, 56 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, 57 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; 58 * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, 59 * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR 60 * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF 61 * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 62 */ 63 64 #include <sys/cdefs.h> 65 __KERNEL_RCSID(0, "$NetBSD: kern_exec.c,v 1.495 2020/04/06 08:20:05 kamil Exp $"); 66 67 #include "opt_exec.h" 68 #include "opt_execfmt.h" 69 #include "opt_ktrace.h" 70 #include "opt_modular.h" 71 #include "opt_syscall_debug.h" 72 #include "veriexec.h" 73 #include "opt_pax.h" 74 75 #include <sys/param.h> 76 #include <sys/systm.h> 77 #include <sys/filedesc.h> 78 #include <sys/kernel.h> 79 #include <sys/proc.h> 80 #include <sys/ptrace.h> 81 #include <sys/mount.h> 82 #include <sys/kmem.h> 83 #include <sys/namei.h> 84 #include <sys/vnode.h> 85 #include <sys/file.h> 86 #include <sys/filedesc.h> 87 #include <sys/acct.h> 88 #include <sys/atomic.h> 89 #include <sys/exec.h> 90 #include <sys/ktrace.h> 91 #include <sys/uidinfo.h> 92 #include <sys/wait.h> 93 #include <sys/mman.h> 94 #include <sys/ras.h> 95 #include <sys/signalvar.h> 96 #include <sys/stat.h> 97 #include <sys/syscall.h> 98 #include <sys/kauth.h> 99 #include <sys/lwpctl.h> 100 #include <sys/pax.h> 101 #include <sys/cpu.h> 102 #include <sys/module.h> 103 #include <sys/syscallvar.h> 104 #include <sys/syscallargs.h> 105 #if NVERIEXEC > 0 106 #include <sys/verified_exec.h> 107 #endif /* NVERIEXEC > 0 */ 108 #include <sys/sdt.h> 109 #include <sys/spawn.h> 110 #include <sys/prot.h> 111 #include <sys/cprng.h> 112 113 #include <uvm/uvm_extern.h> 114 115 #include <machine/reg.h> 116 117 #include <compat/common/compat_util.h> 118 119 #ifndef MD_TOPDOWN_INIT 120 #ifdef __USE_TOPDOWN_VM 121 #define MD_TOPDOWN_INIT(epp) (epp)->ep_flags |= EXEC_TOPDOWN_VM 122 #else 123 #define MD_TOPDOWN_INIT(epp) 124 #endif 125 #endif 126 127 struct execve_data; 128 129 extern int user_va0_disable; 130 131 static size_t calcargs(struct execve_data * restrict, const size_t); 132 static size_t calcstack(struct execve_data * restrict, const size_t); 133 static int copyoutargs(struct execve_data * restrict, struct lwp *, 134 char * const); 135 static int copyoutpsstrs(struct execve_data * restrict, struct proc *); 136 static int copyinargs(struct execve_data * restrict, char * const *, 137 char * const *, execve_fetch_element_t, char **); 138 static int copyinargstrs(struct execve_data * restrict, char * const *, 139 execve_fetch_element_t, char **, size_t *, void (*)(const void *, size_t)); 140 static int exec_sigcode_map(struct proc *, const struct emul *); 141 142 #if defined(DEBUG) && !defined(DEBUG_EXEC) 143 #define DEBUG_EXEC 144 #endif 145 #ifdef DEBUG_EXEC 146 #define DPRINTF(a) printf a 147 #define COPYPRINTF(s, a, b) printf("%s, %d: copyout%s @%p %zu\n", __func__, \ 148 __LINE__, (s), (a), (b)) 149 static void dump_vmcmds(const struct exec_package * const, size_t, int); 150 #define DUMPVMCMDS(p, x, e) do { dump_vmcmds((p), (x), (e)); } while (0) 151 #else 152 #define DPRINTF(a) 153 #define COPYPRINTF(s, a, b) 154 #define DUMPVMCMDS(p, x, e) do {} while (0) 155 #endif /* DEBUG_EXEC */ 156 157 /* 158 * DTrace SDT provider definitions 159 */ 160 SDT_PROVIDER_DECLARE(proc); 161 SDT_PROBE_DEFINE1(proc, kernel, , exec, "char *"); 162 SDT_PROBE_DEFINE1(proc, kernel, , exec__success, "char *"); 163 SDT_PROBE_DEFINE1(proc, kernel, , exec__failure, "int"); 164 165 /* 166 * Exec function switch: 167 * 168 * Note that each makecmds function is responsible for loading the 169 * exec package with the necessary functions for any exec-type-specific 170 * handling. 171 * 172 * Functions for specific exec types should be defined in their own 173 * header file. 174 */ 175 static const struct execsw **execsw = NULL; 176 static int nexecs; 177 178 u_int exec_maxhdrsz; /* must not be static - used by netbsd32 */ 179 180 /* list of dynamically loaded execsw entries */ 181 static LIST_HEAD(execlist_head, exec_entry) ex_head = 182 LIST_HEAD_INITIALIZER(ex_head); 183 struct exec_entry { 184 LIST_ENTRY(exec_entry) ex_list; 185 SLIST_ENTRY(exec_entry) ex_slist; 186 const struct execsw *ex_sw; 187 }; 188 189 #ifndef __HAVE_SYSCALL_INTERN 190 void syscall(void); 191 #endif 192 193 /* NetBSD autoloadable syscalls */ 194 #ifdef MODULAR 195 #include <kern/syscalls_autoload.c> 196 #endif 197 198 /* NetBSD emul struct */ 199 struct emul emul_netbsd = { 200 .e_name = "netbsd", 201 #ifdef EMUL_NATIVEROOT 202 .e_path = EMUL_NATIVEROOT, 203 #else 204 .e_path = NULL, 205 #endif 206 #ifndef __HAVE_MINIMAL_EMUL 207 .e_flags = EMUL_HAS_SYS___syscall, 208 .e_errno = NULL, 209 .e_nosys = SYS_syscall, 210 .e_nsysent = SYS_NSYSENT, 211 #endif 212 #ifdef MODULAR 213 .e_sc_autoload = netbsd_syscalls_autoload, 214 #endif 215 .e_sysent = sysent, 216 .e_nomodbits = sysent_nomodbits, 217 #ifdef SYSCALL_DEBUG 218 .e_syscallnames = syscallnames, 219 #else 220 .e_syscallnames = NULL, 221 #endif 222 .e_sendsig = sendsig, 223 .e_trapsignal = trapsignal, 224 .e_sigcode = NULL, 225 .e_esigcode = NULL, 226 .e_sigobject = NULL, 227 .e_setregs = setregs, 228 .e_proc_exec = NULL, 229 .e_proc_fork = NULL, 230 .e_proc_exit = NULL, 231 .e_lwp_fork = NULL, 232 .e_lwp_exit = NULL, 233 #ifdef __HAVE_SYSCALL_INTERN 234 .e_syscall_intern = syscall_intern, 235 #else 236 .e_syscall = syscall, 237 #endif 238 .e_sysctlovly = NULL, 239 .e_vm_default_addr = uvm_default_mapaddr, 240 .e_usertrap = NULL, 241 .e_ucsize = sizeof(ucontext_t), 242 .e_startlwp = startlwp 243 }; 244 245 /* 246 * Exec lock. Used to control access to execsw[] structures. 247 * This must not be static so that netbsd32 can access it, too. 248 */ 249 krwlock_t exec_lock __cacheline_aligned; 250 251 static kmutex_t sigobject_lock __cacheline_aligned; 252 253 /* 254 * Data used between a loadvm and execve part of an "exec" operation 255 */ 256 struct execve_data { 257 struct exec_package ed_pack; 258 struct pathbuf *ed_pathbuf; 259 struct vattr ed_attr; 260 struct ps_strings ed_arginfo; 261 char *ed_argp; 262 const char *ed_pathstring; 263 char *ed_resolvedname; 264 size_t ed_ps_strings_sz; 265 int ed_szsigcode; 266 size_t ed_argslen; 267 long ed_argc; 268 long ed_envc; 269 }; 270 271 /* 272 * data passed from parent lwp to child during a posix_spawn() 273 */ 274 struct spawn_exec_data { 275 struct execve_data sed_exec; 276 struct posix_spawn_file_actions 277 *sed_actions; 278 struct posix_spawnattr *sed_attrs; 279 struct proc *sed_parent; 280 kcondvar_t sed_cv_child_ready; 281 kmutex_t sed_mtx_child; 282 int sed_error; 283 volatile uint32_t sed_refcnt; 284 }; 285 286 static struct vm_map *exec_map; 287 static struct pool exec_pool; 288 289 static void * 290 exec_pool_alloc(struct pool *pp, int flags) 291 { 292 293 return (void *)uvm_km_alloc(exec_map, NCARGS, 0, 294 UVM_KMF_PAGEABLE | UVM_KMF_WAITVA); 295 } 296 297 static void 298 exec_pool_free(struct pool *pp, void *addr) 299 { 300 301 uvm_km_free(exec_map, (vaddr_t)addr, NCARGS, UVM_KMF_PAGEABLE); 302 } 303 304 static struct pool_allocator exec_palloc = { 305 .pa_alloc = exec_pool_alloc, 306 .pa_free = exec_pool_free, 307 .pa_pagesz = NCARGS 308 }; 309 310 static void 311 exec_path_free(struct execve_data *data) 312 { 313 pathbuf_stringcopy_put(data->ed_pathbuf, data->ed_pathstring); 314 pathbuf_destroy(data->ed_pathbuf); 315 if (data->ed_resolvedname) 316 PNBUF_PUT(data->ed_resolvedname); 317 } 318 319 static int 320 exec_resolvename(struct lwp *l, struct exec_package *epp, struct vnode *vp, 321 char **rpath) 322 { 323 int error; 324 char *p; 325 326 KASSERT(rpath != NULL); 327 328 *rpath = PNBUF_GET(); 329 error = vnode_to_path(*rpath, MAXPATHLEN, vp, l, l->l_proc); 330 if (error) { 331 DPRINTF(("%s: can't resolve name for %s, error %d\n", 332 __func__, epp->ep_kname, error)); 333 PNBUF_PUT(*rpath); 334 *rpath = NULL; 335 return error; 336 } 337 epp->ep_resolvedname = *rpath; 338 if ((p = strrchr(*rpath, '/')) != NULL) 339 epp->ep_kname = p + 1; 340 return 0; 341 } 342 343 344 /* 345 * check exec: 346 * given an "executable" described in the exec package's namei info, 347 * see what we can do with it. 348 * 349 * ON ENTRY: 350 * exec package with appropriate namei info 351 * lwp pointer of exec'ing lwp 352 * NO SELF-LOCKED VNODES 353 * 354 * ON EXIT: 355 * error: nothing held, etc. exec header still allocated. 356 * ok: filled exec package, executable's vnode (unlocked). 357 * 358 * EXEC SWITCH ENTRY: 359 * Locked vnode to check, exec package, proc. 360 * 361 * EXEC SWITCH EXIT: 362 * ok: return 0, filled exec package, executable's vnode (unlocked). 363 * error: destructive: 364 * everything deallocated execept exec header. 365 * non-destructive: 366 * error code, executable's vnode (unlocked), 367 * exec header unmodified. 368 */ 369 int 370 /*ARGSUSED*/ 371 check_exec(struct lwp *l, struct exec_package *epp, struct pathbuf *pb, 372 char **rpath) 373 { 374 int error, i; 375 struct vnode *vp; 376 size_t resid; 377 378 if (epp->ep_resolvedname) { 379 struct nameidata nd; 380 381 // grab the absolute pathbuf here before namei() trashes it. 382 pathbuf_copystring(pb, epp->ep_resolvedname, PATH_MAX); 383 NDINIT(&nd, LOOKUP, FOLLOW | LOCKLEAF | TRYEMULROOT, pb); 384 385 /* first get the vnode */ 386 if ((error = namei(&nd)) != 0) 387 return error; 388 389 epp->ep_vp = vp = nd.ni_vp; 390 #ifdef DIAGNOSTIC 391 /* paranoia (take this out once namei stuff stabilizes) */ 392 memset(nd.ni_pnbuf, '~', PATH_MAX); 393 #endif 394 } else { 395 struct file *fp; 396 397 if ((error = fd_getvnode(epp->ep_xfd, &fp)) != 0) 398 return error; 399 epp->ep_vp = vp = fp->f_vnode; 400 vref(vp); 401 fd_putfile(epp->ep_xfd); 402 if ((error = exec_resolvename(l, epp, vp, rpath)) != 0) 403 return error; 404 vn_lock(vp, LK_EXCLUSIVE | LK_RETRY); 405 } 406 407 /* check access and type */ 408 if (vp->v_type != VREG) { 409 error = EACCES; 410 goto bad1; 411 } 412 if ((error = VOP_ACCESS(vp, VEXEC, l->l_cred)) != 0) 413 goto bad1; 414 415 /* get attributes */ 416 /* XXX VOP_GETATTR is the only thing that needs LK_EXCLUSIVE here */ 417 if ((error = VOP_GETATTR(vp, epp->ep_vap, l->l_cred)) != 0) 418 goto bad1; 419 420 /* Check mount point */ 421 if (vp->v_mount->mnt_flag & MNT_NOEXEC) { 422 error = EACCES; 423 goto bad1; 424 } 425 if (vp->v_mount->mnt_flag & MNT_NOSUID) 426 epp->ep_vap->va_mode &= ~(S_ISUID | S_ISGID); 427 428 /* try to open it */ 429 if ((error = VOP_OPEN(vp, FREAD, l->l_cred)) != 0) 430 goto bad1; 431 432 /* now we have the file, get the exec header */ 433 error = vn_rdwr(UIO_READ, vp, epp->ep_hdr, epp->ep_hdrlen, 0, 434 UIO_SYSSPACE, IO_NODELOCKED, l->l_cred, &resid, NULL); 435 if (error) 436 goto bad1; 437 438 /* unlock vp, since we need it unlocked from here on out. */ 439 VOP_UNLOCK(vp); 440 441 #if NVERIEXEC > 0 442 error = veriexec_verify(l, vp, 443 epp->ep_resolvedname ? epp->ep_resolvedname : epp->ep_kname, 444 epp->ep_flags & EXEC_INDIR ? VERIEXEC_INDIRECT : VERIEXEC_DIRECT, 445 NULL); 446 if (error) 447 goto bad2; 448 #endif /* NVERIEXEC > 0 */ 449 450 #ifdef PAX_SEGVGUARD 451 error = pax_segvguard(l, vp, epp->ep_resolvedname, false); 452 if (error) 453 goto bad2; 454 #endif /* PAX_SEGVGUARD */ 455 456 epp->ep_hdrvalid = epp->ep_hdrlen - resid; 457 458 /* 459 * Set up default address space limits. Can be overridden 460 * by individual exec packages. 461 */ 462 epp->ep_vm_minaddr = exec_vm_minaddr(VM_MIN_ADDRESS); 463 epp->ep_vm_maxaddr = VM_MAXUSER_ADDRESS; 464 465 /* 466 * set up the vmcmds for creation of the process 467 * address space 468 */ 469 error = ENOEXEC; 470 for (i = 0; i < nexecs; i++) { 471 int newerror; 472 473 epp->ep_esch = execsw[i]; 474 newerror = (*execsw[i]->es_makecmds)(l, epp); 475 476 if (!newerror) { 477 /* Seems ok: check that entry point is not too high */ 478 if (epp->ep_entry >= epp->ep_vm_maxaddr) { 479 #ifdef DIAGNOSTIC 480 printf("%s: rejecting %p due to " 481 "too high entry address (>= %p)\n", 482 __func__, (void *)epp->ep_entry, 483 (void *)epp->ep_vm_maxaddr); 484 #endif 485 error = ENOEXEC; 486 break; 487 } 488 /* Seems ok: check that entry point is not too low */ 489 if (epp->ep_entry < epp->ep_vm_minaddr) { 490 #ifdef DIAGNOSTIC 491 printf("%s: rejecting %p due to " 492 "too low entry address (< %p)\n", 493 __func__, (void *)epp->ep_entry, 494 (void *)epp->ep_vm_minaddr); 495 #endif 496 error = ENOEXEC; 497 break; 498 } 499 500 /* check limits */ 501 if ((epp->ep_tsize > MAXTSIZ) || 502 (epp->ep_dsize > (u_quad_t)l->l_proc->p_rlimit 503 [RLIMIT_DATA].rlim_cur)) { 504 #ifdef DIAGNOSTIC 505 printf("%s: rejecting due to " 506 "limits (t=%llu > %llu || d=%llu > %llu)\n", 507 __func__, 508 (unsigned long long)epp->ep_tsize, 509 (unsigned long long)MAXTSIZ, 510 (unsigned long long)epp->ep_dsize, 511 (unsigned long long) 512 l->l_proc->p_rlimit[RLIMIT_DATA].rlim_cur); 513 #endif 514 error = ENOMEM; 515 break; 516 } 517 return 0; 518 } 519 520 /* 521 * Reset all the fields that may have been modified by the 522 * loader. 523 */ 524 KASSERT(epp->ep_emul_arg == NULL); 525 if (epp->ep_emul_root != NULL) { 526 vrele(epp->ep_emul_root); 527 epp->ep_emul_root = NULL; 528 } 529 if (epp->ep_interp != NULL) { 530 vrele(epp->ep_interp); 531 epp->ep_interp = NULL; 532 } 533 epp->ep_pax_flags = 0; 534 535 /* make sure the first "interesting" error code is saved. */ 536 if (error == ENOEXEC) 537 error = newerror; 538 539 if (epp->ep_flags & EXEC_DESTR) 540 /* Error from "#!" code, tidied up by recursive call */ 541 return error; 542 } 543 544 /* not found, error */ 545 546 /* 547 * free any vmspace-creation commands, 548 * and release their references 549 */ 550 kill_vmcmds(&epp->ep_vmcmds); 551 552 #if NVERIEXEC > 0 || defined(PAX_SEGVGUARD) 553 bad2: 554 #endif 555 /* 556 * close and release the vnode, restore the old one, free the 557 * pathname buf, and punt. 558 */ 559 vn_lock(vp, LK_EXCLUSIVE | LK_RETRY); 560 VOP_CLOSE(vp, FREAD, l->l_cred); 561 vput(vp); 562 return error; 563 564 bad1: 565 /* 566 * free the namei pathname buffer, and put the vnode 567 * (which we don't yet have open). 568 */ 569 vput(vp); /* was still locked */ 570 return error; 571 } 572 573 #ifdef __MACHINE_STACK_GROWS_UP 574 #define STACK_PTHREADSPACE NBPG 575 #else 576 #define STACK_PTHREADSPACE 0 577 #endif 578 579 static int 580 execve_fetch_element(char * const *array, size_t index, char **value) 581 { 582 return copyin(array + index, value, sizeof(*value)); 583 } 584 585 /* 586 * exec system call 587 */ 588 int 589 sys_execve(struct lwp *l, const struct sys_execve_args *uap, register_t *retval) 590 { 591 /* { 592 syscallarg(const char *) path; 593 syscallarg(char * const *) argp; 594 syscallarg(char * const *) envp; 595 } */ 596 597 return execve1(l, true, SCARG(uap, path), -1, SCARG(uap, argp), 598 SCARG(uap, envp), execve_fetch_element); 599 } 600 601 int 602 sys_fexecve(struct lwp *l, const struct sys_fexecve_args *uap, 603 register_t *retval) 604 { 605 /* { 606 syscallarg(int) fd; 607 syscallarg(char * const *) argp; 608 syscallarg(char * const *) envp; 609 } */ 610 611 return execve1(l, false, NULL, SCARG(uap, fd), SCARG(uap, argp), 612 SCARG(uap, envp), execve_fetch_element); 613 } 614 615 /* 616 * Load modules to try and execute an image that we do not understand. 617 * If no execsw entries are present, we load those likely to be needed 618 * in order to run native images only. Otherwise, we autoload all 619 * possible modules that could let us run the binary. XXX lame 620 */ 621 static void 622 exec_autoload(void) 623 { 624 #ifdef MODULAR 625 static const char * const native[] = { 626 "exec_elf32", 627 "exec_elf64", 628 "exec_script", 629 NULL 630 }; 631 static const char * const compat[] = { 632 "exec_elf32", 633 "exec_elf64", 634 "exec_script", 635 "exec_aout", 636 "exec_coff", 637 "exec_ecoff", 638 "compat_aoutm68k", 639 "compat_netbsd32", 640 #if 0 641 "compat_linux", 642 "compat_linux32", 643 #endif 644 "compat_sunos", 645 "compat_sunos32", 646 "compat_ultrix", 647 NULL 648 }; 649 char const * const *list; 650 int i; 651 652 list = nexecs == 0 ? native : compat; 653 for (i = 0; list[i] != NULL; i++) { 654 if (module_autoload(list[i], MODULE_CLASS_EXEC) != 0) { 655 continue; 656 } 657 yield(); 658 } 659 #endif 660 } 661 662 /* 663 * Copy the user or kernel supplied upath to the allocated pathbuffer pbp 664 * making it absolute in the process, by prepending the current working 665 * directory if it is not. If offs is supplied it will contain the offset 666 * where the original supplied copy of upath starts. 667 */ 668 int 669 exec_makepathbuf(struct lwp *l, const char *upath, enum uio_seg seg, 670 struct pathbuf **pbp, size_t *offs) 671 { 672 char *path, *bp; 673 size_t len, tlen; 674 int error; 675 struct vnode *dvp; 676 677 path = PNBUF_GET(); 678 if (seg == UIO_SYSSPACE) { 679 error = copystr(upath, path, MAXPATHLEN, &len); 680 } else { 681 error = copyinstr(upath, path, MAXPATHLEN, &len); 682 } 683 if (error) 684 goto err; 685 686 if (path[0] == '/') { 687 if (offs) 688 *offs = 0; 689 goto out; 690 } 691 692 len++; 693 if (len + 1 >= MAXPATHLEN) { 694 error = ENAMETOOLONG; 695 goto err; 696 } 697 bp = path + MAXPATHLEN - len; 698 memmove(bp, path, len); 699 *(--bp) = '/'; 700 701 dvp = cwdcdir(); 702 error = getcwd_common(dvp, NULL, &bp, path, MAXPATHLEN / 2, 703 GETCWD_CHECK_ACCESS, l); 704 vrele(dvp); 705 706 if (error) 707 goto err; 708 tlen = path + MAXPATHLEN - bp; 709 710 memmove(path, bp, tlen); 711 path[tlen - 1] = '\0'; 712 if (offs) 713 *offs = tlen - len; 714 out: 715 *pbp = pathbuf_assimilate(path); 716 return 0; 717 err: 718 PNBUF_PUT(path); 719 return error; 720 } 721 722 vaddr_t 723 exec_vm_minaddr(vaddr_t va_min) 724 { 725 /* 726 * Increase va_min if we don't want NULL to be mappable by the 727 * process. 728 */ 729 #define VM_MIN_GUARD PAGE_SIZE 730 if (user_va0_disable && (va_min < VM_MIN_GUARD)) 731 return VM_MIN_GUARD; 732 return va_min; 733 } 734 735 static int 736 execve_loadvm(struct lwp *l, bool has_path, const char *path, int fd, 737 char * const *args, char * const *envs, 738 execve_fetch_element_t fetch_element, 739 struct execve_data * restrict data) 740 { 741 struct exec_package * const epp = &data->ed_pack; 742 int error; 743 struct proc *p; 744 char *dp; 745 u_int modgen; 746 747 KASSERT(data != NULL); 748 749 p = l->l_proc; 750 modgen = 0; 751 752 SDT_PROBE(proc, kernel, , exec, path, 0, 0, 0, 0); 753 754 /* 755 * Check if we have exceeded our number of processes limit. 756 * This is so that we handle the case where a root daemon 757 * forked, ran setuid to become the desired user and is trying 758 * to exec. The obvious place to do the reference counting check 759 * is setuid(), but we don't do the reference counting check there 760 * like other OS's do because then all the programs that use setuid() 761 * must be modified to check the return code of setuid() and exit(). 762 * It is dangerous to make setuid() fail, because it fails open and 763 * the program will continue to run as root. If we make it succeed 764 * and return an error code, again we are not enforcing the limit. 765 * The best place to enforce the limit is here, when the process tries 766 * to execute a new image, because eventually the process will need 767 * to call exec in order to do something useful. 768 */ 769 retry: 770 if (p->p_flag & PK_SUGID) { 771 if (kauth_authorize_process(l->l_cred, KAUTH_PROCESS_RLIMIT, 772 p, KAUTH_ARG(KAUTH_REQ_PROCESS_RLIMIT_BYPASS), 773 &p->p_rlimit[RLIMIT_NPROC], 774 KAUTH_ARG(RLIMIT_NPROC)) != 0 && 775 chgproccnt(kauth_cred_getuid(l->l_cred), 0) > 776 p->p_rlimit[RLIMIT_NPROC].rlim_cur) 777 return EAGAIN; 778 } 779 780 /* 781 * Drain existing references and forbid new ones. The process 782 * should be left alone until we're done here. This is necessary 783 * to avoid race conditions - e.g. in ptrace() - that might allow 784 * a local user to illicitly obtain elevated privileges. 785 */ 786 rw_enter(&p->p_reflock, RW_WRITER); 787 788 if (has_path) { 789 size_t offs; 790 /* 791 * Init the namei data to point the file user's program name. 792 * This is done here rather than in check_exec(), so that it's 793 * possible to override this settings if any of makecmd/probe 794 * functions call check_exec() recursively - for example, 795 * see exec_script_makecmds(). 796 */ 797 if ((error = exec_makepathbuf(l, path, UIO_USERSPACE, 798 &data->ed_pathbuf, &offs)) != 0) 799 goto clrflg; 800 data->ed_pathstring = pathbuf_stringcopy_get(data->ed_pathbuf); 801 epp->ep_kname = data->ed_pathstring + offs; 802 data->ed_resolvedname = PNBUF_GET(); 803 epp->ep_resolvedname = data->ed_resolvedname; 804 epp->ep_xfd = -1; 805 } else { 806 data->ed_pathbuf = pathbuf_assimilate(strcpy(PNBUF_GET(), "/")); 807 data->ed_pathstring = pathbuf_stringcopy_get(data->ed_pathbuf); 808 epp->ep_kname = "*fexecve*"; 809 data->ed_resolvedname = NULL; 810 epp->ep_resolvedname = NULL; 811 epp->ep_xfd = fd; 812 } 813 814 815 /* 816 * initialize the fields of the exec package. 817 */ 818 epp->ep_hdr = kmem_alloc(exec_maxhdrsz, KM_SLEEP); 819 epp->ep_hdrlen = exec_maxhdrsz; 820 epp->ep_hdrvalid = 0; 821 epp->ep_emul_arg = NULL; 822 epp->ep_emul_arg_free = NULL; 823 memset(&epp->ep_vmcmds, 0, sizeof(epp->ep_vmcmds)); 824 epp->ep_vap = &data->ed_attr; 825 epp->ep_flags = (p->p_flag & PK_32) ? EXEC_FROM32 : 0; 826 MD_TOPDOWN_INIT(epp); 827 epp->ep_emul_root = NULL; 828 epp->ep_interp = NULL; 829 epp->ep_esch = NULL; 830 epp->ep_pax_flags = 0; 831 memset(epp->ep_machine_arch, 0, sizeof(epp->ep_machine_arch)); 832 833 rw_enter(&exec_lock, RW_READER); 834 835 /* see if we can run it. */ 836 if ((error = check_exec(l, epp, data->ed_pathbuf, 837 &data->ed_resolvedname)) != 0) { 838 if (error != ENOENT && error != EACCES && error != ENOEXEC) { 839 DPRINTF(("%s: check exec failed for %s, error %d\n", 840 __func__, epp->ep_kname, error)); 841 } 842 goto freehdr; 843 } 844 845 /* allocate an argument buffer */ 846 data->ed_argp = pool_get(&exec_pool, PR_WAITOK); 847 KASSERT(data->ed_argp != NULL); 848 dp = data->ed_argp; 849 850 if ((error = copyinargs(data, args, envs, fetch_element, &dp)) != 0) { 851 goto bad; 852 } 853 854 /* 855 * Calculate the new stack size. 856 */ 857 858 #ifdef __MACHINE_STACK_GROWS_UP 859 /* 860 * copyargs() fills argc/argv/envp from the lower address even on 861 * __MACHINE_STACK_GROWS_UP machines. Reserve a few words just below the SP 862 * so that _rtld() use it. 863 */ 864 #define RTLD_GAP 32 865 #else 866 #define RTLD_GAP 0 867 #endif 868 869 const size_t argenvstrlen = (char *)ALIGN(dp) - data->ed_argp; 870 871 data->ed_argslen = calcargs(data, argenvstrlen); 872 873 const size_t len = calcstack(data, pax_aslr_stack_gap(epp) + RTLD_GAP); 874 875 if (len > epp->ep_ssize) { 876 /* in effect, compare to initial limit */ 877 DPRINTF(("%s: stack limit exceeded %zu\n", __func__, len)); 878 error = ENOMEM; 879 goto bad; 880 } 881 /* adjust "active stack depth" for process VSZ */ 882 epp->ep_ssize = len; 883 884 return 0; 885 886 bad: 887 /* free the vmspace-creation commands, and release their references */ 888 kill_vmcmds(&epp->ep_vmcmds); 889 /* kill any opened file descriptor, if necessary */ 890 if (epp->ep_flags & EXEC_HASFD) { 891 epp->ep_flags &= ~EXEC_HASFD; 892 fd_close(epp->ep_fd); 893 } 894 /* close and put the exec'd file */ 895 vn_lock(epp->ep_vp, LK_EXCLUSIVE | LK_RETRY); 896 VOP_CLOSE(epp->ep_vp, FREAD, l->l_cred); 897 vput(epp->ep_vp); 898 pool_put(&exec_pool, data->ed_argp); 899 900 freehdr: 901 kmem_free(epp->ep_hdr, epp->ep_hdrlen); 902 if (epp->ep_emul_root != NULL) 903 vrele(epp->ep_emul_root); 904 if (epp->ep_interp != NULL) 905 vrele(epp->ep_interp); 906 907 rw_exit(&exec_lock); 908 909 exec_path_free(data); 910 911 clrflg: 912 rw_exit(&p->p_reflock); 913 914 if (modgen != module_gen && error == ENOEXEC) { 915 modgen = module_gen; 916 exec_autoload(); 917 goto retry; 918 } 919 920 SDT_PROBE(proc, kernel, , exec__failure, error, 0, 0, 0, 0); 921 return error; 922 } 923 924 static int 925 execve_dovmcmds(struct lwp *l, struct execve_data * restrict data) 926 { 927 struct exec_package * const epp = &data->ed_pack; 928 struct proc *p = l->l_proc; 929 struct exec_vmcmd *base_vcp; 930 int error = 0; 931 size_t i; 932 933 /* record proc's vnode, for use by procfs and others */ 934 if (p->p_textvp) 935 vrele(p->p_textvp); 936 vref(epp->ep_vp); 937 p->p_textvp = epp->ep_vp; 938 939 /* create the new process's VM space by running the vmcmds */ 940 KASSERTMSG(epp->ep_vmcmds.evs_used != 0, "%s: no vmcmds", __func__); 941 942 #ifdef TRACE_EXEC 943 DUMPVMCMDS(epp, 0, 0); 944 #endif 945 946 base_vcp = NULL; 947 948 for (i = 0; i < epp->ep_vmcmds.evs_used && !error; i++) { 949 struct exec_vmcmd *vcp; 950 951 vcp = &epp->ep_vmcmds.evs_cmds[i]; 952 if (vcp->ev_flags & VMCMD_RELATIVE) { 953 KASSERTMSG(base_vcp != NULL, 954 "%s: relative vmcmd with no base", __func__); 955 KASSERTMSG((vcp->ev_flags & VMCMD_BASE) == 0, 956 "%s: illegal base & relative vmcmd", __func__); 957 vcp->ev_addr += base_vcp->ev_addr; 958 } 959 error = (*vcp->ev_proc)(l, vcp); 960 if (error) 961 DUMPVMCMDS(epp, i, error); 962 if (vcp->ev_flags & VMCMD_BASE) 963 base_vcp = vcp; 964 } 965 966 /* free the vmspace-creation commands, and release their references */ 967 kill_vmcmds(&epp->ep_vmcmds); 968 969 vn_lock(epp->ep_vp, LK_EXCLUSIVE | LK_RETRY); 970 VOP_CLOSE(epp->ep_vp, FREAD, l->l_cred); 971 vput(epp->ep_vp); 972 973 /* if an error happened, deallocate and punt */ 974 if (error != 0) { 975 DPRINTF(("%s: vmcmd %zu failed: %d\n", __func__, i - 1, error)); 976 } 977 return error; 978 } 979 980 static void 981 execve_free_data(struct execve_data *data) 982 { 983 struct exec_package * const epp = &data->ed_pack; 984 985 /* free the vmspace-creation commands, and release their references */ 986 kill_vmcmds(&epp->ep_vmcmds); 987 /* kill any opened file descriptor, if necessary */ 988 if (epp->ep_flags & EXEC_HASFD) { 989 epp->ep_flags &= ~EXEC_HASFD; 990 fd_close(epp->ep_fd); 991 } 992 993 /* close and put the exec'd file */ 994 vn_lock(epp->ep_vp, LK_EXCLUSIVE | LK_RETRY); 995 VOP_CLOSE(epp->ep_vp, FREAD, curlwp->l_cred); 996 vput(epp->ep_vp); 997 pool_put(&exec_pool, data->ed_argp); 998 999 kmem_free(epp->ep_hdr, epp->ep_hdrlen); 1000 if (epp->ep_emul_root != NULL) 1001 vrele(epp->ep_emul_root); 1002 if (epp->ep_interp != NULL) 1003 vrele(epp->ep_interp); 1004 1005 exec_path_free(data); 1006 } 1007 1008 static void 1009 pathexec(struct proc *p, const char *resolvedname) 1010 { 1011 /* set command name & other accounting info */ 1012 const char *cmdname; 1013 1014 if (resolvedname == NULL) { 1015 cmdname = "*fexecve*"; 1016 resolvedname = "/"; 1017 } else { 1018 cmdname = strrchr(resolvedname, '/') + 1; 1019 } 1020 KASSERTMSG(resolvedname[0] == '/', "bad resolvedname `%s'", 1021 resolvedname); 1022 1023 strlcpy(p->p_comm, cmdname, sizeof(p->p_comm)); 1024 1025 kmem_strfree(p->p_path); 1026 p->p_path = kmem_strdupsize(resolvedname, NULL, KM_SLEEP); 1027 } 1028 1029 /* XXX elsewhere */ 1030 static int 1031 credexec(struct lwp *l, struct vattr *attr) 1032 { 1033 struct proc *p = l->l_proc; 1034 int error; 1035 1036 /* 1037 * Deal with set[ug]id. MNT_NOSUID has already been used to disable 1038 * s[ug]id. It's OK to check for PSL_TRACED here as we have blocked 1039 * out additional references on the process for the moment. 1040 */ 1041 if ((p->p_slflag & PSL_TRACED) == 0 && 1042 1043 (((attr->va_mode & S_ISUID) != 0 && 1044 kauth_cred_geteuid(l->l_cred) != attr->va_uid) || 1045 1046 ((attr->va_mode & S_ISGID) != 0 && 1047 kauth_cred_getegid(l->l_cred) != attr->va_gid))) { 1048 /* 1049 * Mark the process as SUGID before we do 1050 * anything that might block. 1051 */ 1052 proc_crmod_enter(); 1053 proc_crmod_leave(NULL, NULL, true); 1054 1055 /* Make sure file descriptors 0..2 are in use. */ 1056 if ((error = fd_checkstd()) != 0) { 1057 DPRINTF(("%s: fdcheckstd failed %d\n", 1058 __func__, error)); 1059 return error; 1060 } 1061 1062 /* 1063 * Copy the credential so other references don't see our 1064 * changes. 1065 */ 1066 l->l_cred = kauth_cred_copy(l->l_cred); 1067 #ifdef KTRACE 1068 /* 1069 * If the persistent trace flag isn't set, turn off. 1070 */ 1071 if (p->p_tracep) { 1072 mutex_enter(&ktrace_lock); 1073 if (!(p->p_traceflag & KTRFAC_PERSISTENT)) 1074 ktrderef(p); 1075 mutex_exit(&ktrace_lock); 1076 } 1077 #endif 1078 if (attr->va_mode & S_ISUID) 1079 kauth_cred_seteuid(l->l_cred, attr->va_uid); 1080 if (attr->va_mode & S_ISGID) 1081 kauth_cred_setegid(l->l_cred, attr->va_gid); 1082 } else { 1083 if (kauth_cred_geteuid(l->l_cred) == 1084 kauth_cred_getuid(l->l_cred) && 1085 kauth_cred_getegid(l->l_cred) == 1086 kauth_cred_getgid(l->l_cred)) 1087 p->p_flag &= ~PK_SUGID; 1088 } 1089 1090 /* 1091 * Copy the credential so other references don't see our changes. 1092 * Test to see if this is necessary first, since in the common case 1093 * we won't need a private reference. 1094 */ 1095 if (kauth_cred_geteuid(l->l_cred) != kauth_cred_getsvuid(l->l_cred) || 1096 kauth_cred_getegid(l->l_cred) != kauth_cred_getsvgid(l->l_cred)) { 1097 l->l_cred = kauth_cred_copy(l->l_cred); 1098 kauth_cred_setsvuid(l->l_cred, kauth_cred_geteuid(l->l_cred)); 1099 kauth_cred_setsvgid(l->l_cred, kauth_cred_getegid(l->l_cred)); 1100 } 1101 1102 /* Update the master credentials. */ 1103 if (l->l_cred != p->p_cred) { 1104 kauth_cred_t ocred; 1105 1106 kauth_cred_hold(l->l_cred); 1107 mutex_enter(p->p_lock); 1108 ocred = p->p_cred; 1109 p->p_cred = l->l_cred; 1110 mutex_exit(p->p_lock); 1111 kauth_cred_free(ocred); 1112 } 1113 1114 return 0; 1115 } 1116 1117 static void 1118 emulexec(struct lwp *l, struct exec_package *epp) 1119 { 1120 struct proc *p = l->l_proc; 1121 struct cwdinfo *cwdi; 1122 1123 /* The emulation root will usually have been found when we looked 1124 * for the elf interpreter (or similar), if not look now. */ 1125 if (epp->ep_esch->es_emul->e_path != NULL && 1126 epp->ep_emul_root == NULL) 1127 emul_find_root(l, epp); 1128 1129 /* Any old emulation root got removed by fdcloseexec */ 1130 KASSERT(p == curproc); 1131 cwdi = cwdenter(RW_WRITER); 1132 cwdi->cwdi_edir = epp->ep_emul_root; 1133 cwdexit(cwdi); 1134 epp->ep_emul_root = NULL; 1135 if (epp->ep_interp != NULL) 1136 vrele(epp->ep_interp); 1137 1138 /* 1139 * Call emulation specific exec hook. This can setup per-process 1140 * p->p_emuldata or do any other per-process stuff an emulation needs. 1141 * 1142 * If we are executing process of different emulation than the 1143 * original forked process, call e_proc_exit() of the old emulation 1144 * first, then e_proc_exec() of new emulation. If the emulation is 1145 * same, the exec hook code should deallocate any old emulation 1146 * resources held previously by this process. 1147 */ 1148 if (p->p_emul && p->p_emul->e_proc_exit 1149 && p->p_emul != epp->ep_esch->es_emul) 1150 (*p->p_emul->e_proc_exit)(p); 1151 1152 /* This is now LWP 1. Re-number the LWP if needed. */ 1153 if (l->l_lid != 1) 1154 lwp_renumber(l, 1); 1155 1156 /* 1157 * Call exec hook. Emulation code may NOT store reference to anything 1158 * from &pack. 1159 */ 1160 if (epp->ep_esch->es_emul->e_proc_exec) 1161 (*epp->ep_esch->es_emul->e_proc_exec)(p, epp); 1162 1163 /* update p_emul, the old value is no longer needed */ 1164 p->p_emul = epp->ep_esch->es_emul; 1165 1166 /* ...and the same for p_execsw */ 1167 p->p_execsw = epp->ep_esch; 1168 1169 #ifdef __HAVE_SYSCALL_INTERN 1170 (*p->p_emul->e_syscall_intern)(p); 1171 #endif 1172 ktremul(); 1173 } 1174 1175 static int 1176 execve_runproc(struct lwp *l, struct execve_data * restrict data, 1177 bool no_local_exec_lock, bool is_spawn) 1178 { 1179 struct exec_package * const epp = &data->ed_pack; 1180 int error = 0; 1181 struct proc *p; 1182 struct vmspace *vm; 1183 1184 /* 1185 * In case of a posix_spawn operation, the child doing the exec 1186 * might not hold the reader lock on exec_lock, but the parent 1187 * will do this instead. 1188 */ 1189 KASSERT(no_local_exec_lock || rw_lock_held(&exec_lock)); 1190 KASSERT(!no_local_exec_lock || is_spawn); 1191 KASSERT(data != NULL); 1192 1193 p = l->l_proc; 1194 1195 /* Get rid of other LWPs. */ 1196 if (p->p_nlwps > 1) { 1197 mutex_enter(p->p_lock); 1198 exit_lwps(l); 1199 mutex_exit(p->p_lock); 1200 } 1201 KDASSERT(p->p_nlwps == 1); 1202 1203 /* Destroy any lwpctl info. */ 1204 if (p->p_lwpctl != NULL) 1205 lwp_ctl_exit(); 1206 1207 /* Remove POSIX timers */ 1208 timers_free(p, TIMERS_POSIX); 1209 1210 /* Set the PaX flags. */ 1211 pax_set_flags(epp, p); 1212 1213 /* 1214 * Do whatever is necessary to prepare the address space 1215 * for remapping. Note that this might replace the current 1216 * vmspace with another! 1217 * 1218 * vfork(): do not touch any user space data in the new child 1219 * until we have awoken the parent below, or it will defeat 1220 * lazy pmap switching (on x86). 1221 */ 1222 if (is_spawn) 1223 uvmspace_spawn(l, epp->ep_vm_minaddr, 1224 epp->ep_vm_maxaddr, 1225 epp->ep_flags & EXEC_TOPDOWN_VM); 1226 else 1227 uvmspace_exec(l, epp->ep_vm_minaddr, 1228 epp->ep_vm_maxaddr, 1229 epp->ep_flags & EXEC_TOPDOWN_VM); 1230 vm = p->p_vmspace; 1231 1232 vm->vm_taddr = (void *)epp->ep_taddr; 1233 vm->vm_tsize = btoc(epp->ep_tsize); 1234 vm->vm_daddr = (void*)epp->ep_daddr; 1235 vm->vm_dsize = btoc(epp->ep_dsize); 1236 vm->vm_ssize = btoc(epp->ep_ssize); 1237 vm->vm_issize = 0; 1238 vm->vm_maxsaddr = (void *)epp->ep_maxsaddr; 1239 vm->vm_minsaddr = (void *)epp->ep_minsaddr; 1240 1241 pax_aslr_init_vm(l, vm, epp); 1242 1243 cwdexec(p); 1244 fd_closeexec(); /* handle close on exec */ 1245 1246 if (__predict_false(ktrace_on)) 1247 fd_ktrexecfd(); 1248 1249 execsigs(p); /* reset caught signals */ 1250 1251 mutex_enter(p->p_lock); 1252 l->l_ctxlink = NULL; /* reset ucontext link */ 1253 p->p_acflag &= ~AFORK; 1254 p->p_flag |= PK_EXEC; 1255 mutex_exit(p->p_lock); 1256 1257 error = credexec(l, &data->ed_attr); 1258 if (error) 1259 goto exec_abort; 1260 1261 #if defined(__HAVE_RAS) 1262 /* 1263 * Remove all RASs from the address space. 1264 */ 1265 ras_purgeall(); 1266 #endif 1267 1268 /* 1269 * Stop profiling. 1270 */ 1271 if ((p->p_stflag & PST_PROFIL) != 0) { 1272 mutex_spin_enter(&p->p_stmutex); 1273 stopprofclock(p); 1274 mutex_spin_exit(&p->p_stmutex); 1275 } 1276 1277 /* 1278 * It's OK to test PL_PPWAIT unlocked here, as other LWPs have 1279 * exited and exec()/exit() are the only places it will be cleared. 1280 * 1281 * Once the parent has been awoken, curlwp may teleport to a new CPU 1282 * in sched_vforkexec(), and it's then OK to start messing with user 1283 * data. See comment above. 1284 */ 1285 if ((p->p_lflag & PL_PPWAIT) != 0) { 1286 bool samecpu; 1287 lwp_t *lp; 1288 1289 mutex_enter(proc_lock); 1290 lp = p->p_vforklwp; 1291 p->p_vforklwp = NULL; 1292 l->l_lwpctl = NULL; /* was on loan from blocked parent */ 1293 cv_broadcast(&lp->l_waitcv); 1294 1295 /* Clear flags after cv_broadcast() (scheduler needs them). */ 1296 p->p_lflag &= ~PL_PPWAIT; 1297 lp->l_vforkwaiting = false; 1298 1299 /* If parent is still on same CPU, teleport curlwp elsewhere. */ 1300 samecpu = (lp->l_cpu == curlwp->l_cpu); 1301 mutex_exit(proc_lock); 1302 1303 /* Give the parent its CPU back - find a new home. */ 1304 KASSERT(!is_spawn); 1305 sched_vforkexec(l, samecpu); 1306 } 1307 1308 /* Now map address space. */ 1309 error = execve_dovmcmds(l, data); 1310 if (error != 0) 1311 goto exec_abort; 1312 1313 pathexec(p, epp->ep_resolvedname); 1314 1315 char * const newstack = STACK_GROW(vm->vm_minsaddr, epp->ep_ssize); 1316 1317 error = copyoutargs(data, l, newstack); 1318 if (error != 0) 1319 goto exec_abort; 1320 1321 doexechooks(p); 1322 1323 /* 1324 * Set initial SP at the top of the stack. 1325 * 1326 * Note that on machines where stack grows up (e.g. hppa), SP points to 1327 * the end of arg/env strings. Userland guesses the address of argc 1328 * via ps_strings::ps_argvstr. 1329 */ 1330 1331 /* Setup new registers and do misc. setup. */ 1332 (*epp->ep_esch->es_emul->e_setregs)(l, epp, (vaddr_t)newstack); 1333 if (epp->ep_esch->es_setregs) 1334 (*epp->ep_esch->es_setregs)(l, epp, (vaddr_t)newstack); 1335 1336 /* Provide a consistent LWP private setting */ 1337 (void)lwp_setprivate(l, NULL); 1338 1339 /* Discard all PCU state; need to start fresh */ 1340 pcu_discard_all(l); 1341 1342 /* map the process's signal trampoline code */ 1343 if ((error = exec_sigcode_map(p, epp->ep_esch->es_emul)) != 0) { 1344 DPRINTF(("%s: map sigcode failed %d\n", __func__, error)); 1345 goto exec_abort; 1346 } 1347 1348 pool_put(&exec_pool, data->ed_argp); 1349 1350 /* notify others that we exec'd */ 1351 KNOTE(&p->p_klist, NOTE_EXEC); 1352 1353 kmem_free(epp->ep_hdr, epp->ep_hdrlen); 1354 1355 SDT_PROBE(proc, kernel, , exec__success, epp->ep_kname, 0, 0, 0, 0); 1356 1357 emulexec(l, epp); 1358 1359 /* Allow new references from the debugger/procfs. */ 1360 rw_exit(&p->p_reflock); 1361 if (!no_local_exec_lock) 1362 rw_exit(&exec_lock); 1363 1364 mutex_enter(proc_lock); 1365 1366 /* posix_spawn(3) reports a single event with implied exec(3) */ 1367 if ((p->p_slflag & PSL_TRACED) && !is_spawn) { 1368 mutex_enter(p->p_lock); 1369 eventswitch(TRAP_EXEC, 0, 0); 1370 mutex_enter(proc_lock); 1371 } 1372 1373 if (p->p_sflag & PS_STOPEXEC) { 1374 ksiginfoq_t kq; 1375 1376 KERNEL_UNLOCK_ALL(l, &l->l_biglocks); 1377 p->p_pptr->p_nstopchild++; 1378 p->p_waited = 0; 1379 mutex_enter(p->p_lock); 1380 ksiginfo_queue_init(&kq); 1381 sigclearall(p, &contsigmask, &kq); 1382 lwp_lock(l); 1383 l->l_stat = LSSTOP; 1384 p->p_stat = SSTOP; 1385 p->p_nrlwps--; 1386 lwp_unlock(l); 1387 mutex_exit(p->p_lock); 1388 mutex_exit(proc_lock); 1389 lwp_lock(l); 1390 spc_lock(l->l_cpu); 1391 mi_switch(l); 1392 ksiginfo_queue_drain(&kq); 1393 } else { 1394 mutex_exit(proc_lock); 1395 } 1396 1397 exec_path_free(data); 1398 #ifdef TRACE_EXEC 1399 DPRINTF(("%s finished\n", __func__)); 1400 #endif 1401 return EJUSTRETURN; 1402 1403 exec_abort: 1404 SDT_PROBE(proc, kernel, , exec__failure, error, 0, 0, 0, 0); 1405 rw_exit(&p->p_reflock); 1406 if (!no_local_exec_lock) 1407 rw_exit(&exec_lock); 1408 1409 exec_path_free(data); 1410 1411 /* 1412 * the old process doesn't exist anymore. exit gracefully. 1413 * get rid of the (new) address space we have created, if any, get rid 1414 * of our namei data and vnode, and exit noting failure 1415 */ 1416 if (vm != NULL) { 1417 uvm_deallocate(&vm->vm_map, VM_MIN_ADDRESS, 1418 VM_MAXUSER_ADDRESS - VM_MIN_ADDRESS); 1419 } 1420 1421 exec_free_emul_arg(epp); 1422 pool_put(&exec_pool, data->ed_argp); 1423 kmem_free(epp->ep_hdr, epp->ep_hdrlen); 1424 if (epp->ep_emul_root != NULL) 1425 vrele(epp->ep_emul_root); 1426 if (epp->ep_interp != NULL) 1427 vrele(epp->ep_interp); 1428 1429 /* Acquire the sched-state mutex (exit1() will release it). */ 1430 if (!is_spawn) { 1431 mutex_enter(p->p_lock); 1432 exit1(l, error, SIGABRT); 1433 } 1434 1435 return error; 1436 } 1437 1438 int 1439 execve1(struct lwp *l, bool has_path, const char *path, int fd, 1440 char * const *args, char * const *envs, 1441 execve_fetch_element_t fetch_element) 1442 { 1443 struct execve_data data; 1444 int error; 1445 1446 error = execve_loadvm(l, has_path, path, fd, args, envs, fetch_element, 1447 &data); 1448 if (error) 1449 return error; 1450 error = execve_runproc(l, &data, false, false); 1451 return error; 1452 } 1453 1454 static size_t 1455 fromptrsz(const struct exec_package *epp) 1456 { 1457 return (epp->ep_flags & EXEC_FROM32) ? sizeof(int) : sizeof(char *); 1458 } 1459 1460 static size_t 1461 ptrsz(const struct exec_package *epp) 1462 { 1463 return (epp->ep_flags & EXEC_32) ? sizeof(int) : sizeof(char *); 1464 } 1465 1466 static size_t 1467 calcargs(struct execve_data * restrict data, const size_t argenvstrlen) 1468 { 1469 struct exec_package * const epp = &data->ed_pack; 1470 1471 const size_t nargenvptrs = 1472 1 + /* long argc */ 1473 data->ed_argc + /* char *argv[] */ 1474 1 + /* \0 */ 1475 data->ed_envc + /* char *env[] */ 1476 1; /* \0 */ 1477 1478 return (nargenvptrs * ptrsz(epp)) /* pointers */ 1479 + argenvstrlen /* strings */ 1480 + epp->ep_esch->es_arglen; /* auxinfo */ 1481 } 1482 1483 static size_t 1484 calcstack(struct execve_data * restrict data, const size_t gaplen) 1485 { 1486 struct exec_package * const epp = &data->ed_pack; 1487 1488 data->ed_szsigcode = epp->ep_esch->es_emul->e_esigcode - 1489 epp->ep_esch->es_emul->e_sigcode; 1490 1491 data->ed_ps_strings_sz = (epp->ep_flags & EXEC_32) ? 1492 sizeof(struct ps_strings32) : sizeof(struct ps_strings); 1493 1494 const size_t sigcode_psstr_sz = 1495 data->ed_szsigcode + /* sigcode */ 1496 data->ed_ps_strings_sz + /* ps_strings */ 1497 STACK_PTHREADSPACE; /* pthread space */ 1498 1499 const size_t stacklen = 1500 data->ed_argslen + 1501 gaplen + 1502 sigcode_psstr_sz; 1503 1504 /* make the stack "safely" aligned */ 1505 return STACK_LEN_ALIGN(stacklen, STACK_ALIGNBYTES); 1506 } 1507 1508 static int 1509 copyoutargs(struct execve_data * restrict data, struct lwp *l, 1510 char * const newstack) 1511 { 1512 struct exec_package * const epp = &data->ed_pack; 1513 struct proc *p = l->l_proc; 1514 int error; 1515 1516 memset(&data->ed_arginfo, 0, sizeof(data->ed_arginfo)); 1517 1518 /* remember information about the process */ 1519 data->ed_arginfo.ps_nargvstr = data->ed_argc; 1520 data->ed_arginfo.ps_nenvstr = data->ed_envc; 1521 1522 /* 1523 * Allocate the stack address passed to the newly execve()'ed process. 1524 * 1525 * The new stack address will be set to the SP (stack pointer) register 1526 * in setregs(). 1527 */ 1528 1529 char *newargs = STACK_ALLOC( 1530 STACK_SHRINK(newstack, data->ed_argslen), data->ed_argslen); 1531 1532 error = (*epp->ep_esch->es_copyargs)(l, epp, 1533 &data->ed_arginfo, &newargs, data->ed_argp); 1534 1535 if (error) { 1536 DPRINTF(("%s: copyargs failed %d\n", __func__, error)); 1537 return error; 1538 } 1539 1540 error = copyoutpsstrs(data, p); 1541 if (error != 0) 1542 return error; 1543 1544 return 0; 1545 } 1546 1547 static int 1548 copyoutpsstrs(struct execve_data * restrict data, struct proc *p) 1549 { 1550 struct exec_package * const epp = &data->ed_pack; 1551 struct ps_strings32 arginfo32; 1552 void *aip; 1553 int error; 1554 1555 /* fill process ps_strings info */ 1556 p->p_psstrp = (vaddr_t)STACK_ALLOC(STACK_GROW(epp->ep_minsaddr, 1557 STACK_PTHREADSPACE), data->ed_ps_strings_sz); 1558 1559 if (epp->ep_flags & EXEC_32) { 1560 aip = &arginfo32; 1561 arginfo32.ps_argvstr = (vaddr_t)data->ed_arginfo.ps_argvstr; 1562 arginfo32.ps_nargvstr = data->ed_arginfo.ps_nargvstr; 1563 arginfo32.ps_envstr = (vaddr_t)data->ed_arginfo.ps_envstr; 1564 arginfo32.ps_nenvstr = data->ed_arginfo.ps_nenvstr; 1565 } else 1566 aip = &data->ed_arginfo; 1567 1568 /* copy out the process's ps_strings structure */ 1569 if ((error = copyout(aip, (void *)p->p_psstrp, data->ed_ps_strings_sz)) 1570 != 0) { 1571 DPRINTF(("%s: ps_strings copyout %p->%p size %zu failed\n", 1572 __func__, aip, (void *)p->p_psstrp, data->ed_ps_strings_sz)); 1573 return error; 1574 } 1575 1576 return 0; 1577 } 1578 1579 static int 1580 copyinargs(struct execve_data * restrict data, char * const *args, 1581 char * const *envs, execve_fetch_element_t fetch_element, char **dpp) 1582 { 1583 struct exec_package * const epp = &data->ed_pack; 1584 char *dp; 1585 size_t i; 1586 int error; 1587 1588 dp = *dpp; 1589 1590 data->ed_argc = 0; 1591 1592 /* copy the fake args list, if there's one, freeing it as we go */ 1593 if (epp->ep_flags & EXEC_HASARGL) { 1594 struct exec_fakearg *fa = epp->ep_fa; 1595 1596 while (fa->fa_arg != NULL) { 1597 const size_t maxlen = ARG_MAX - (dp - data->ed_argp); 1598 size_t len; 1599 1600 len = strlcpy(dp, fa->fa_arg, maxlen); 1601 /* Count NUL into len. */ 1602 if (len < maxlen) 1603 len++; 1604 else { 1605 while (fa->fa_arg != NULL) { 1606 kmem_free(fa->fa_arg, fa->fa_len); 1607 fa++; 1608 } 1609 kmem_free(epp->ep_fa, epp->ep_fa_len); 1610 epp->ep_flags &= ~EXEC_HASARGL; 1611 return E2BIG; 1612 } 1613 ktrexecarg(fa->fa_arg, len - 1); 1614 dp += len; 1615 1616 kmem_free(fa->fa_arg, fa->fa_len); 1617 fa++; 1618 data->ed_argc++; 1619 } 1620 kmem_free(epp->ep_fa, epp->ep_fa_len); 1621 epp->ep_flags &= ~EXEC_HASARGL; 1622 } 1623 1624 /* 1625 * Read and count argument strings from user. 1626 */ 1627 1628 if (args == NULL) { 1629 DPRINTF(("%s: null args\n", __func__)); 1630 return EINVAL; 1631 } 1632 if (epp->ep_flags & EXEC_SKIPARG) 1633 args = (const void *)((const char *)args + fromptrsz(epp)); 1634 i = 0; 1635 error = copyinargstrs(data, args, fetch_element, &dp, &i, ktr_execarg); 1636 if (error != 0) { 1637 DPRINTF(("%s: copyin arg %d\n", __func__, error)); 1638 return error; 1639 } 1640 data->ed_argc += i; 1641 1642 /* 1643 * Read and count environment strings from user. 1644 */ 1645 1646 data->ed_envc = 0; 1647 /* environment need not be there */ 1648 if (envs == NULL) 1649 goto done; 1650 i = 0; 1651 error = copyinargstrs(data, envs, fetch_element, &dp, &i, ktr_execenv); 1652 if (error != 0) { 1653 DPRINTF(("%s: copyin env %d\n", __func__, error)); 1654 return error; 1655 } 1656 data->ed_envc += i; 1657 1658 done: 1659 *dpp = dp; 1660 1661 return 0; 1662 } 1663 1664 static int 1665 copyinargstrs(struct execve_data * restrict data, char * const *strs, 1666 execve_fetch_element_t fetch_element, char **dpp, size_t *ip, 1667 void (*ktr)(const void *, size_t)) 1668 { 1669 char *dp, *sp; 1670 size_t i; 1671 int error; 1672 1673 dp = *dpp; 1674 1675 i = 0; 1676 while (1) { 1677 const size_t maxlen = ARG_MAX - (dp - data->ed_argp); 1678 size_t len; 1679 1680 if ((error = (*fetch_element)(strs, i, &sp)) != 0) { 1681 return error; 1682 } 1683 if (!sp) 1684 break; 1685 if ((error = copyinstr(sp, dp, maxlen, &len)) != 0) { 1686 if (error == ENAMETOOLONG) 1687 error = E2BIG; 1688 return error; 1689 } 1690 if (__predict_false(ktrace_on)) 1691 (*ktr)(dp, len - 1); 1692 dp += len; 1693 i++; 1694 } 1695 1696 *dpp = dp; 1697 *ip = i; 1698 1699 return 0; 1700 } 1701 1702 /* 1703 * Copy argv and env strings from kernel buffer (argp) to the new stack. 1704 * Those strings are located just after auxinfo. 1705 */ 1706 int 1707 copyargs(struct lwp *l, struct exec_package *pack, struct ps_strings *arginfo, 1708 char **stackp, void *argp) 1709 { 1710 char **cpp, *dp, *sp; 1711 size_t len; 1712 void *nullp; 1713 long argc, envc; 1714 int error; 1715 1716 cpp = (char **)*stackp; 1717 nullp = NULL; 1718 argc = arginfo->ps_nargvstr; 1719 envc = arginfo->ps_nenvstr; 1720 1721 /* argc on stack is long */ 1722 CTASSERT(sizeof(*cpp) == sizeof(argc)); 1723 1724 dp = (char *)(cpp + 1725 1 + /* long argc */ 1726 argc + /* char *argv[] */ 1727 1 + /* \0 */ 1728 envc + /* char *env[] */ 1729 1) + /* \0 */ 1730 pack->ep_esch->es_arglen; /* auxinfo */ 1731 sp = argp; 1732 1733 if ((error = copyout(&argc, cpp++, sizeof(argc))) != 0) { 1734 COPYPRINTF("", cpp - 1, sizeof(argc)); 1735 return error; 1736 } 1737 1738 /* XXX don't copy them out, remap them! */ 1739 arginfo->ps_argvstr = cpp; /* remember location of argv for later */ 1740 1741 for (; --argc >= 0; sp += len, dp += len) { 1742 if ((error = copyout(&dp, cpp++, sizeof(dp))) != 0) { 1743 COPYPRINTF("", cpp - 1, sizeof(dp)); 1744 return error; 1745 } 1746 if ((error = copyoutstr(sp, dp, ARG_MAX, &len)) != 0) { 1747 COPYPRINTF("str", dp, (size_t)ARG_MAX); 1748 return error; 1749 } 1750 } 1751 1752 if ((error = copyout(&nullp, cpp++, sizeof(nullp))) != 0) { 1753 COPYPRINTF("", cpp - 1, sizeof(nullp)); 1754 return error; 1755 } 1756 1757 arginfo->ps_envstr = cpp; /* remember location of envp for later */ 1758 1759 for (; --envc >= 0; sp += len, dp += len) { 1760 if ((error = copyout(&dp, cpp++, sizeof(dp))) != 0) { 1761 COPYPRINTF("", cpp - 1, sizeof(dp)); 1762 return error; 1763 } 1764 if ((error = copyoutstr(sp, dp, ARG_MAX, &len)) != 0) { 1765 COPYPRINTF("str", dp, (size_t)ARG_MAX); 1766 return error; 1767 } 1768 1769 } 1770 1771 if ((error = copyout(&nullp, cpp++, sizeof(nullp))) != 0) { 1772 COPYPRINTF("", cpp - 1, sizeof(nullp)); 1773 return error; 1774 } 1775 1776 *stackp = (char *)cpp; 1777 return 0; 1778 } 1779 1780 1781 /* 1782 * Add execsw[] entries. 1783 */ 1784 int 1785 exec_add(struct execsw *esp, int count) 1786 { 1787 struct exec_entry *it; 1788 int i; 1789 1790 if (count == 0) { 1791 return 0; 1792 } 1793 1794 /* Check for duplicates. */ 1795 rw_enter(&exec_lock, RW_WRITER); 1796 for (i = 0; i < count; i++) { 1797 LIST_FOREACH(it, &ex_head, ex_list) { 1798 /* assume unique (makecmds, probe_func, emulation) */ 1799 if (it->ex_sw->es_makecmds == esp[i].es_makecmds && 1800 it->ex_sw->u.elf_probe_func == 1801 esp[i].u.elf_probe_func && 1802 it->ex_sw->es_emul == esp[i].es_emul) { 1803 rw_exit(&exec_lock); 1804 return EEXIST; 1805 } 1806 } 1807 } 1808 1809 /* Allocate new entries. */ 1810 for (i = 0; i < count; i++) { 1811 it = kmem_alloc(sizeof(*it), KM_SLEEP); 1812 it->ex_sw = &esp[i]; 1813 LIST_INSERT_HEAD(&ex_head, it, ex_list); 1814 } 1815 1816 /* update execsw[] */ 1817 exec_init(0); 1818 rw_exit(&exec_lock); 1819 return 0; 1820 } 1821 1822 /* 1823 * Remove execsw[] entry. 1824 */ 1825 int 1826 exec_remove(struct execsw *esp, int count) 1827 { 1828 struct exec_entry *it, *next; 1829 int i; 1830 const struct proclist_desc *pd; 1831 proc_t *p; 1832 1833 if (count == 0) { 1834 return 0; 1835 } 1836 1837 /* Abort if any are busy. */ 1838 rw_enter(&exec_lock, RW_WRITER); 1839 for (i = 0; i < count; i++) { 1840 mutex_enter(proc_lock); 1841 for (pd = proclists; pd->pd_list != NULL; pd++) { 1842 PROCLIST_FOREACH(p, pd->pd_list) { 1843 if (p->p_execsw == &esp[i]) { 1844 mutex_exit(proc_lock); 1845 rw_exit(&exec_lock); 1846 return EBUSY; 1847 } 1848 } 1849 } 1850 mutex_exit(proc_lock); 1851 } 1852 1853 /* None are busy, so remove them all. */ 1854 for (i = 0; i < count; i++) { 1855 for (it = LIST_FIRST(&ex_head); it != NULL; it = next) { 1856 next = LIST_NEXT(it, ex_list); 1857 if (it->ex_sw == &esp[i]) { 1858 LIST_REMOVE(it, ex_list); 1859 kmem_free(it, sizeof(*it)); 1860 break; 1861 } 1862 } 1863 } 1864 1865 /* update execsw[] */ 1866 exec_init(0); 1867 rw_exit(&exec_lock); 1868 return 0; 1869 } 1870 1871 /* 1872 * Initialize exec structures. If init_boot is true, also does necessary 1873 * one-time initialization (it's called from main() that way). 1874 * Once system is multiuser, this should be called with exec_lock held, 1875 * i.e. via exec_{add|remove}(). 1876 */ 1877 int 1878 exec_init(int init_boot) 1879 { 1880 const struct execsw **sw; 1881 struct exec_entry *ex; 1882 SLIST_HEAD(,exec_entry) first; 1883 SLIST_HEAD(,exec_entry) any; 1884 SLIST_HEAD(,exec_entry) last; 1885 int i, sz; 1886 1887 if (init_boot) { 1888 /* do one-time initializations */ 1889 vaddr_t vmin = 0, vmax; 1890 1891 rw_init(&exec_lock); 1892 mutex_init(&sigobject_lock, MUTEX_DEFAULT, IPL_NONE); 1893 exec_map = uvm_km_suballoc(kernel_map, &vmin, &vmax, 1894 maxexec*NCARGS, VM_MAP_PAGEABLE, false, NULL); 1895 pool_init(&exec_pool, NCARGS, 0, 0, PR_NOALIGN|PR_NOTOUCH, 1896 "execargs", &exec_palloc, IPL_NONE); 1897 pool_sethardlimit(&exec_pool, maxexec, "should not happen", 0); 1898 } else { 1899 KASSERT(rw_write_held(&exec_lock)); 1900 } 1901 1902 /* Sort each entry onto the appropriate queue. */ 1903 SLIST_INIT(&first); 1904 SLIST_INIT(&any); 1905 SLIST_INIT(&last); 1906 sz = 0; 1907 LIST_FOREACH(ex, &ex_head, ex_list) { 1908 switch(ex->ex_sw->es_prio) { 1909 case EXECSW_PRIO_FIRST: 1910 SLIST_INSERT_HEAD(&first, ex, ex_slist); 1911 break; 1912 case EXECSW_PRIO_ANY: 1913 SLIST_INSERT_HEAD(&any, ex, ex_slist); 1914 break; 1915 case EXECSW_PRIO_LAST: 1916 SLIST_INSERT_HEAD(&last, ex, ex_slist); 1917 break; 1918 default: 1919 panic("%s", __func__); 1920 break; 1921 } 1922 sz++; 1923 } 1924 1925 /* 1926 * Create new execsw[]. Ensure we do not try a zero-sized 1927 * allocation. 1928 */ 1929 sw = kmem_alloc(sz * sizeof(struct execsw *) + 1, KM_SLEEP); 1930 i = 0; 1931 SLIST_FOREACH(ex, &first, ex_slist) { 1932 sw[i++] = ex->ex_sw; 1933 } 1934 SLIST_FOREACH(ex, &any, ex_slist) { 1935 sw[i++] = ex->ex_sw; 1936 } 1937 SLIST_FOREACH(ex, &last, ex_slist) { 1938 sw[i++] = ex->ex_sw; 1939 } 1940 1941 /* Replace old execsw[] and free used memory. */ 1942 if (execsw != NULL) { 1943 kmem_free(__UNCONST(execsw), 1944 nexecs * sizeof(struct execsw *) + 1); 1945 } 1946 execsw = sw; 1947 nexecs = sz; 1948 1949 /* Figure out the maximum size of an exec header. */ 1950 exec_maxhdrsz = sizeof(int); 1951 for (i = 0; i < nexecs; i++) { 1952 if (execsw[i]->es_hdrsz > exec_maxhdrsz) 1953 exec_maxhdrsz = execsw[i]->es_hdrsz; 1954 } 1955 1956 return 0; 1957 } 1958 1959 static int 1960 exec_sigcode_map(struct proc *p, const struct emul *e) 1961 { 1962 vaddr_t va; 1963 vsize_t sz; 1964 int error; 1965 struct uvm_object *uobj; 1966 1967 sz = (vaddr_t)e->e_esigcode - (vaddr_t)e->e_sigcode; 1968 1969 if (e->e_sigobject == NULL || sz == 0) { 1970 return 0; 1971 } 1972 1973 /* 1974 * If we don't have a sigobject for this emulation, create one. 1975 * 1976 * sigobject is an anonymous memory object (just like SYSV shared 1977 * memory) that we keep a permanent reference to and that we map 1978 * in all processes that need this sigcode. The creation is simple, 1979 * we create an object, add a permanent reference to it, map it in 1980 * kernel space, copy out the sigcode to it and unmap it. 1981 * We map it with PROT_READ|PROT_EXEC into the process just 1982 * the way sys_mmap() would map it. 1983 */ 1984 1985 uobj = *e->e_sigobject; 1986 if (uobj == NULL) { 1987 mutex_enter(&sigobject_lock); 1988 if ((uobj = *e->e_sigobject) == NULL) { 1989 uobj = uao_create(sz, 0); 1990 (*uobj->pgops->pgo_reference)(uobj); 1991 va = vm_map_min(kernel_map); 1992 if ((error = uvm_map(kernel_map, &va, round_page(sz), 1993 uobj, 0, 0, 1994 UVM_MAPFLAG(UVM_PROT_RW, UVM_PROT_RW, 1995 UVM_INH_SHARE, UVM_ADV_RANDOM, 0)))) { 1996 printf("kernel mapping failed %d\n", error); 1997 (*uobj->pgops->pgo_detach)(uobj); 1998 mutex_exit(&sigobject_lock); 1999 return error; 2000 } 2001 memcpy((void *)va, e->e_sigcode, sz); 2002 #ifdef PMAP_NEED_PROCWR 2003 pmap_procwr(&proc0, va, sz); 2004 #endif 2005 uvm_unmap(kernel_map, va, va + round_page(sz)); 2006 *e->e_sigobject = uobj; 2007 } 2008 mutex_exit(&sigobject_lock); 2009 } 2010 2011 /* Just a hint to uvm_map where to put it. */ 2012 va = e->e_vm_default_addr(p, (vaddr_t)p->p_vmspace->vm_daddr, 2013 round_page(sz), p->p_vmspace->vm_map.flags & VM_MAP_TOPDOWN); 2014 2015 #ifdef __alpha__ 2016 /* 2017 * Tru64 puts /sbin/loader at the end of user virtual memory, 2018 * which causes the above calculation to put the sigcode at 2019 * an invalid address. Put it just below the text instead. 2020 */ 2021 if (va == (vaddr_t)vm_map_max(&p->p_vmspace->vm_map)) { 2022 va = (vaddr_t)p->p_vmspace->vm_taddr - round_page(sz); 2023 } 2024 #endif 2025 2026 (*uobj->pgops->pgo_reference)(uobj); 2027 error = uvm_map(&p->p_vmspace->vm_map, &va, round_page(sz), 2028 uobj, 0, 0, 2029 UVM_MAPFLAG(UVM_PROT_RX, UVM_PROT_RX, UVM_INH_SHARE, 2030 UVM_ADV_RANDOM, 0)); 2031 if (error) { 2032 DPRINTF(("%s, %d: map %p " 2033 "uvm_map %#"PRIxVSIZE"@%#"PRIxVADDR" failed %d\n", 2034 __func__, __LINE__, &p->p_vmspace->vm_map, round_page(sz), 2035 va, error)); 2036 (*uobj->pgops->pgo_detach)(uobj); 2037 return error; 2038 } 2039 p->p_sigctx.ps_sigcode = (void *)va; 2040 return 0; 2041 } 2042 2043 /* 2044 * Release a refcount on spawn_exec_data and destroy memory, if this 2045 * was the last one. 2046 */ 2047 static void 2048 spawn_exec_data_release(struct spawn_exec_data *data) 2049 { 2050 if (atomic_dec_32_nv(&data->sed_refcnt) != 0) 2051 return; 2052 2053 cv_destroy(&data->sed_cv_child_ready); 2054 mutex_destroy(&data->sed_mtx_child); 2055 2056 if (data->sed_actions) 2057 posix_spawn_fa_free(data->sed_actions, 2058 data->sed_actions->len); 2059 if (data->sed_attrs) 2060 kmem_free(data->sed_attrs, 2061 sizeof(*data->sed_attrs)); 2062 kmem_free(data, sizeof(*data)); 2063 } 2064 2065 static int 2066 handle_posix_spawn_file_actions(struct posix_spawn_file_actions *actions) 2067 { 2068 struct lwp *l = curlwp; 2069 register_t retval; 2070 int error, newfd; 2071 2072 if (actions == NULL) 2073 return 0; 2074 2075 for (size_t i = 0; i < actions->len; i++) { 2076 const struct posix_spawn_file_actions_entry *fae = 2077 &actions->fae[i]; 2078 switch (fae->fae_action) { 2079 case FAE_OPEN: 2080 if (fd_getfile(fae->fae_fildes) != NULL) { 2081 error = fd_close(fae->fae_fildes); 2082 if (error) 2083 return error; 2084 } 2085 error = fd_open(fae->fae_path, fae->fae_oflag, 2086 fae->fae_mode, &newfd); 2087 if (error) 2088 return error; 2089 if (newfd != fae->fae_fildes) { 2090 error = dodup(l, newfd, 2091 fae->fae_fildes, 0, &retval); 2092 if (fd_getfile(newfd) != NULL) 2093 fd_close(newfd); 2094 } 2095 break; 2096 case FAE_DUP2: 2097 error = dodup(l, fae->fae_fildes, 2098 fae->fae_newfildes, 0, &retval); 2099 break; 2100 case FAE_CLOSE: 2101 if (fd_getfile(fae->fae_fildes) == NULL) { 2102 return EBADF; 2103 } 2104 error = fd_close(fae->fae_fildes); 2105 break; 2106 } 2107 if (error) 2108 return error; 2109 } 2110 return 0; 2111 } 2112 2113 static int 2114 handle_posix_spawn_attrs(struct posix_spawnattr *attrs, struct proc *parent) 2115 { 2116 struct sigaction sigact; 2117 int error; 2118 struct proc *p = curproc; 2119 struct lwp *l = curlwp; 2120 2121 if (attrs == NULL) 2122 return 0; 2123 2124 memset(&sigact, 0, sizeof(sigact)); 2125 sigact._sa_u._sa_handler = SIG_DFL; 2126 sigact.sa_flags = 0; 2127 2128 /* 2129 * set state to SSTOP so that this proc can be found by pid. 2130 * see proc_enterprp, do_sched_setparam below 2131 */ 2132 mutex_enter(proc_lock); 2133 /* 2134 * p_stat should be SACTIVE, so we need to adjust the 2135 * parent's p_nstopchild here. For safety, just make 2136 * we're on the good side of SDEAD before we adjust. 2137 */ 2138 int ostat = p->p_stat; 2139 KASSERT(ostat < SSTOP); 2140 p->p_stat = SSTOP; 2141 p->p_waited = 0; 2142 p->p_pptr->p_nstopchild++; 2143 mutex_exit(proc_lock); 2144 2145 /* Set process group */ 2146 if (attrs->sa_flags & POSIX_SPAWN_SETPGROUP) { 2147 pid_t mypid = p->p_pid; 2148 pid_t pgrp = attrs->sa_pgroup; 2149 2150 if (pgrp == 0) 2151 pgrp = mypid; 2152 2153 error = proc_enterpgrp(parent, mypid, pgrp, false); 2154 if (error) 2155 goto out; 2156 } 2157 2158 /* Set scheduler policy */ 2159 if (attrs->sa_flags & POSIX_SPAWN_SETSCHEDULER) 2160 error = do_sched_setparam(p->p_pid, 0, attrs->sa_schedpolicy, 2161 &attrs->sa_schedparam); 2162 else if (attrs->sa_flags & POSIX_SPAWN_SETSCHEDPARAM) { 2163 error = do_sched_setparam(parent->p_pid, 0, 2164 SCHED_NONE, &attrs->sa_schedparam); 2165 } 2166 if (error) 2167 goto out; 2168 2169 /* Reset user ID's */ 2170 if (attrs->sa_flags & POSIX_SPAWN_RESETIDS) { 2171 error = do_setresuid(l, -1, kauth_cred_getgid(l->l_cred), -1, 2172 ID_E_EQ_R | ID_E_EQ_S); 2173 if (error) 2174 return error; 2175 error = do_setresuid(l, -1, kauth_cred_getuid(l->l_cred), -1, 2176 ID_E_EQ_R | ID_E_EQ_S); 2177 if (error) 2178 goto out; 2179 } 2180 2181 /* Set signal masks/defaults */ 2182 if (attrs->sa_flags & POSIX_SPAWN_SETSIGMASK) { 2183 mutex_enter(p->p_lock); 2184 error = sigprocmask1(l, SIG_SETMASK, &attrs->sa_sigmask, NULL); 2185 mutex_exit(p->p_lock); 2186 if (error) 2187 goto out; 2188 } 2189 2190 if (attrs->sa_flags & POSIX_SPAWN_SETSIGDEF) { 2191 /* 2192 * The following sigaction call is using a sigaction 2193 * version 0 trampoline which is in the compatibility 2194 * code only. This is not a problem because for SIG_DFL 2195 * and SIG_IGN, the trampolines are now ignored. If they 2196 * were not, this would be a problem because we are 2197 * holding the exec_lock, and the compat code needs 2198 * to do the same in order to replace the trampoline 2199 * code of the process. 2200 */ 2201 for (int i = 1; i <= NSIG; i++) { 2202 if (sigismember(&attrs->sa_sigdefault, i)) 2203 sigaction1(l, i, &sigact, NULL, NULL, 0); 2204 } 2205 } 2206 error = 0; 2207 out: 2208 mutex_enter(proc_lock); 2209 p->p_stat = ostat; 2210 p->p_pptr->p_nstopchild--; 2211 mutex_exit(proc_lock); 2212 return error; 2213 } 2214 2215 /* 2216 * A child lwp of a posix_spawn operation starts here and ends up in 2217 * cpu_spawn_return, dealing with all filedescriptor and scheduler 2218 * manipulations in between. 2219 * The parent waits for the child, as it is not clear whether the child 2220 * will be able to acquire its own exec_lock. If it can, the parent can 2221 * be released early and continue running in parallel. If not (or if the 2222 * magic debug flag is passed in the scheduler attribute struct), the 2223 * child rides on the parent's exec lock until it is ready to return to 2224 * to userland - and only then releases the parent. This method loses 2225 * concurrency, but improves error reporting. 2226 */ 2227 static void 2228 spawn_return(void *arg) 2229 { 2230 struct spawn_exec_data *spawn_data = arg; 2231 struct lwp *l = curlwp; 2232 struct proc *p = l->l_proc; 2233 int error; 2234 bool have_reflock; 2235 bool parent_is_waiting = true; 2236 2237 /* 2238 * Check if we can release parent early. 2239 * We either need to have no sed_attrs, or sed_attrs does not 2240 * have POSIX_SPAWN_RETURNERROR or one of the flags, that require 2241 * safe access to the parent proc (passed in sed_parent). 2242 * We then try to get the exec_lock, and only if that works, we can 2243 * release the parent here already. 2244 */ 2245 struct posix_spawnattr *attrs = spawn_data->sed_attrs; 2246 if ((!attrs || (attrs->sa_flags 2247 & (POSIX_SPAWN_RETURNERROR|POSIX_SPAWN_SETPGROUP)) == 0) 2248 && rw_tryenter(&exec_lock, RW_READER)) { 2249 parent_is_waiting = false; 2250 mutex_enter(&spawn_data->sed_mtx_child); 2251 cv_signal(&spawn_data->sed_cv_child_ready); 2252 mutex_exit(&spawn_data->sed_mtx_child); 2253 } 2254 2255 /* don't allow debugger access yet */ 2256 rw_enter(&p->p_reflock, RW_WRITER); 2257 have_reflock = true; 2258 2259 /* handle posix_spawn_file_actions */ 2260 error = handle_posix_spawn_file_actions(spawn_data->sed_actions); 2261 if (error) 2262 goto report_error; 2263 2264 /* handle posix_spawnattr */ 2265 error = handle_posix_spawn_attrs(attrs, spawn_data->sed_parent); 2266 if (error) 2267 goto report_error; 2268 2269 /* now do the real exec */ 2270 error = execve_runproc(l, &spawn_data->sed_exec, parent_is_waiting, 2271 true); 2272 have_reflock = false; 2273 if (error == EJUSTRETURN) 2274 error = 0; 2275 else if (error) 2276 goto report_error; 2277 2278 if (parent_is_waiting) { 2279 mutex_enter(&spawn_data->sed_mtx_child); 2280 cv_signal(&spawn_data->sed_cv_child_ready); 2281 mutex_exit(&spawn_data->sed_mtx_child); 2282 } 2283 2284 /* release our refcount on the data */ 2285 spawn_exec_data_release(spawn_data); 2286 2287 if (p->p_slflag & PSL_TRACED) 2288 eventswitchchild(p, TRAP_CHLD, PTRACE_POSIX_SPAWN); 2289 2290 /* and finally: leave to userland for the first time */ 2291 cpu_spawn_return(l); 2292 2293 /* NOTREACHED */ 2294 return; 2295 2296 report_error: 2297 if (have_reflock) { 2298 /* 2299 * We have not passed through execve_runproc(), 2300 * which would have released the p_reflock and also 2301 * taken ownership of the sed_exec part of spawn_data, 2302 * so release/free both here. 2303 */ 2304 rw_exit(&p->p_reflock); 2305 execve_free_data(&spawn_data->sed_exec); 2306 } 2307 2308 if (parent_is_waiting) { 2309 /* pass error to parent */ 2310 mutex_enter(&spawn_data->sed_mtx_child); 2311 spawn_data->sed_error = error; 2312 cv_signal(&spawn_data->sed_cv_child_ready); 2313 mutex_exit(&spawn_data->sed_mtx_child); 2314 } else { 2315 rw_exit(&exec_lock); 2316 } 2317 2318 /* release our refcount on the data */ 2319 spawn_exec_data_release(spawn_data); 2320 2321 /* done, exit */ 2322 mutex_enter(p->p_lock); 2323 /* 2324 * Posix explicitly asks for an exit code of 127 if we report 2325 * errors from the child process - so, unfortunately, there 2326 * is no way to report a more exact error code. 2327 * A NetBSD specific workaround is POSIX_SPAWN_RETURNERROR as 2328 * flag bit in the attrp argument to posix_spawn(2), see above. 2329 */ 2330 exit1(l, 127, 0); 2331 } 2332 2333 void 2334 posix_spawn_fa_free(struct posix_spawn_file_actions *fa, size_t len) 2335 { 2336 2337 for (size_t i = 0; i < len; i++) { 2338 struct posix_spawn_file_actions_entry *fae = &fa->fae[i]; 2339 if (fae->fae_action != FAE_OPEN) 2340 continue; 2341 kmem_strfree(fae->fae_path); 2342 } 2343 if (fa->len > 0) 2344 kmem_free(fa->fae, sizeof(*fa->fae) * fa->len); 2345 kmem_free(fa, sizeof(*fa)); 2346 } 2347 2348 static int 2349 posix_spawn_fa_alloc(struct posix_spawn_file_actions **fap, 2350 const struct posix_spawn_file_actions *ufa, rlim_t lim) 2351 { 2352 struct posix_spawn_file_actions *fa; 2353 struct posix_spawn_file_actions_entry *fae; 2354 char *pbuf = NULL; 2355 int error; 2356 size_t i = 0; 2357 2358 fa = kmem_alloc(sizeof(*fa), KM_SLEEP); 2359 error = copyin(ufa, fa, sizeof(*fa)); 2360 if (error || fa->len == 0) { 2361 kmem_free(fa, sizeof(*fa)); 2362 return error; /* 0 if not an error, and len == 0 */ 2363 } 2364 2365 if (fa->len > lim) { 2366 kmem_free(fa, sizeof(*fa)); 2367 return EINVAL; 2368 } 2369 2370 fa->size = fa->len; 2371 size_t fal = fa->len * sizeof(*fae); 2372 fae = fa->fae; 2373 fa->fae = kmem_alloc(fal, KM_SLEEP); 2374 error = copyin(fae, fa->fae, fal); 2375 if (error) 2376 goto out; 2377 2378 pbuf = PNBUF_GET(); 2379 for (; i < fa->len; i++) { 2380 fae = &fa->fae[i]; 2381 if (fae->fae_action != FAE_OPEN) 2382 continue; 2383 error = copyinstr(fae->fae_path, pbuf, MAXPATHLEN, &fal); 2384 if (error) 2385 goto out; 2386 fae->fae_path = kmem_alloc(fal, KM_SLEEP); 2387 memcpy(fae->fae_path, pbuf, fal); 2388 } 2389 PNBUF_PUT(pbuf); 2390 2391 *fap = fa; 2392 return 0; 2393 out: 2394 if (pbuf) 2395 PNBUF_PUT(pbuf); 2396 posix_spawn_fa_free(fa, i); 2397 return error; 2398 } 2399 2400 int 2401 check_posix_spawn(struct lwp *l1) 2402 { 2403 int error, tnprocs, count; 2404 uid_t uid; 2405 struct proc *p1; 2406 2407 p1 = l1->l_proc; 2408 uid = kauth_cred_getuid(l1->l_cred); 2409 tnprocs = atomic_inc_uint_nv(&nprocs); 2410 2411 /* 2412 * Although process entries are dynamically created, we still keep 2413 * a global limit on the maximum number we will create. 2414 */ 2415 if (__predict_false(tnprocs >= maxproc)) 2416 error = -1; 2417 else 2418 error = kauth_authorize_process(l1->l_cred, 2419 KAUTH_PROCESS_FORK, p1, KAUTH_ARG(tnprocs), NULL, NULL); 2420 2421 if (error) { 2422 atomic_dec_uint(&nprocs); 2423 return EAGAIN; 2424 } 2425 2426 /* 2427 * Enforce limits. 2428 */ 2429 count = chgproccnt(uid, 1); 2430 if (kauth_authorize_process(l1->l_cred, KAUTH_PROCESS_RLIMIT, 2431 p1, KAUTH_ARG(KAUTH_REQ_PROCESS_RLIMIT_BYPASS), 2432 &p1->p_rlimit[RLIMIT_NPROC], KAUTH_ARG(RLIMIT_NPROC)) != 0 && 2433 __predict_false(count > p1->p_rlimit[RLIMIT_NPROC].rlim_cur)) { 2434 (void)chgproccnt(uid, -1); 2435 atomic_dec_uint(&nprocs); 2436 return EAGAIN; 2437 } 2438 2439 return 0; 2440 } 2441 2442 int 2443 do_posix_spawn(struct lwp *l1, pid_t *pid_res, bool *child_ok, const char *path, 2444 struct posix_spawn_file_actions *fa, 2445 struct posix_spawnattr *sa, 2446 char *const *argv, char *const *envp, 2447 execve_fetch_element_t fetch) 2448 { 2449 2450 struct proc *p1, *p2; 2451 struct lwp *l2; 2452 int error; 2453 struct spawn_exec_data *spawn_data; 2454 vaddr_t uaddr; 2455 pid_t pid; 2456 bool have_exec_lock = false; 2457 2458 p1 = l1->l_proc; 2459 2460 /* Allocate and init spawn_data */ 2461 spawn_data = kmem_zalloc(sizeof(*spawn_data), KM_SLEEP); 2462 spawn_data->sed_refcnt = 1; /* only parent so far */ 2463 cv_init(&spawn_data->sed_cv_child_ready, "pspawn"); 2464 mutex_init(&spawn_data->sed_mtx_child, MUTEX_DEFAULT, IPL_NONE); 2465 mutex_enter(&spawn_data->sed_mtx_child); 2466 2467 /* 2468 * Do the first part of the exec now, collect state 2469 * in spawn_data. 2470 */ 2471 error = execve_loadvm(l1, true, path, -1, argv, 2472 envp, fetch, &spawn_data->sed_exec); 2473 if (error == EJUSTRETURN) 2474 error = 0; 2475 else if (error) 2476 goto error_exit; 2477 2478 have_exec_lock = true; 2479 2480 /* 2481 * Allocate virtual address space for the U-area now, while it 2482 * is still easy to abort the fork operation if we're out of 2483 * kernel virtual address space. 2484 */ 2485 uaddr = uvm_uarea_alloc(); 2486 if (__predict_false(uaddr == 0)) { 2487 error = ENOMEM; 2488 goto error_exit; 2489 } 2490 2491 /* 2492 * Allocate new proc. Borrow proc0 vmspace for it, we will 2493 * replace it with its own before returning to userland 2494 * in the child. 2495 * This is a point of no return, we will have to go through 2496 * the child proc to properly clean it up past this point. 2497 */ 2498 p2 = proc_alloc(); 2499 pid = p2->p_pid; 2500 2501 /* 2502 * Make a proc table entry for the new process. 2503 * Start by zeroing the section of proc that is zero-initialized, 2504 * then copy the section that is copied directly from the parent. 2505 */ 2506 memset(&p2->p_startzero, 0, 2507 (unsigned) ((char *)&p2->p_endzero - (char *)&p2->p_startzero)); 2508 memcpy(&p2->p_startcopy, &p1->p_startcopy, 2509 (unsigned) ((char *)&p2->p_endcopy - (char *)&p2->p_startcopy)); 2510 p2->p_vmspace = proc0.p_vmspace; 2511 2512 TAILQ_INIT(&p2->p_sigpend.sp_info); 2513 2514 LIST_INIT(&p2->p_lwps); 2515 LIST_INIT(&p2->p_sigwaiters); 2516 2517 /* 2518 * Duplicate sub-structures as needed. 2519 * Increase reference counts on shared objects. 2520 * Inherit flags we want to keep. The flags related to SIGCHLD 2521 * handling are important in order to keep a consistent behaviour 2522 * for the child after the fork. If we are a 32-bit process, the 2523 * child will be too. 2524 */ 2525 p2->p_flag = 2526 p1->p_flag & (PK_SUGID | PK_NOCLDWAIT | PK_CLDSIGIGN | PK_32); 2527 p2->p_emul = p1->p_emul; 2528 p2->p_execsw = p1->p_execsw; 2529 2530 mutex_init(&p2->p_stmutex, MUTEX_DEFAULT, IPL_HIGH); 2531 mutex_init(&p2->p_auxlock, MUTEX_DEFAULT, IPL_NONE); 2532 rw_init(&p2->p_reflock); 2533 rw_init(&p2->p_treelock); 2534 cv_init(&p2->p_waitcv, "wait"); 2535 cv_init(&p2->p_lwpcv, "lwpwait"); 2536 2537 p2->p_lock = mutex_obj_alloc(MUTEX_DEFAULT, IPL_NONE); 2538 2539 kauth_proc_fork(p1, p2); 2540 2541 p2->p_raslist = NULL; 2542 p2->p_fd = fd_copy(); 2543 2544 /* XXX racy */ 2545 p2->p_mqueue_cnt = p1->p_mqueue_cnt; 2546 2547 p2->p_cwdi = cwdinit(); 2548 2549 /* 2550 * Note: p_limit (rlimit stuff) is copy-on-write, so normally 2551 * we just need increase pl_refcnt. 2552 */ 2553 if (!p1->p_limit->pl_writeable) { 2554 lim_addref(p1->p_limit); 2555 p2->p_limit = p1->p_limit; 2556 } else { 2557 p2->p_limit = lim_copy(p1->p_limit); 2558 } 2559 2560 p2->p_lflag = 0; 2561 l1->l_vforkwaiting = false; 2562 p2->p_sflag = 0; 2563 p2->p_slflag = 0; 2564 p2->p_pptr = p1; 2565 p2->p_ppid = p1->p_pid; 2566 LIST_INIT(&p2->p_children); 2567 2568 p2->p_aio = NULL; 2569 2570 #ifdef KTRACE 2571 /* 2572 * Copy traceflag and tracefile if enabled. 2573 * If not inherited, these were zeroed above. 2574 */ 2575 if (p1->p_traceflag & KTRFAC_INHERIT) { 2576 mutex_enter(&ktrace_lock); 2577 p2->p_traceflag = p1->p_traceflag; 2578 if ((p2->p_tracep = p1->p_tracep) != NULL) 2579 ktradref(p2); 2580 mutex_exit(&ktrace_lock); 2581 } 2582 #endif 2583 2584 /* 2585 * Create signal actions for the child process. 2586 */ 2587 p2->p_sigacts = sigactsinit(p1, 0); 2588 mutex_enter(p1->p_lock); 2589 p2->p_sflag |= 2590 (p1->p_sflag & (PS_STOPFORK | PS_STOPEXEC | PS_NOCLDSTOP)); 2591 sched_proc_fork(p1, p2); 2592 mutex_exit(p1->p_lock); 2593 2594 p2->p_stflag = p1->p_stflag; 2595 2596 /* 2597 * p_stats. 2598 * Copy parts of p_stats, and zero out the rest. 2599 */ 2600 p2->p_stats = pstatscopy(p1->p_stats); 2601 2602 /* copy over machdep flags to the new proc */ 2603 cpu_proc_fork(p1, p2); 2604 2605 /* 2606 * Prepare remaining parts of spawn data 2607 */ 2608 spawn_data->sed_actions = fa; 2609 spawn_data->sed_attrs = sa; 2610 2611 spawn_data->sed_parent = p1; 2612 2613 /* create LWP */ 2614 lwp_create(l1, p2, uaddr, 0, NULL, 0, spawn_return, spawn_data, 2615 &l2, l1->l_class, &l1->l_sigmask, &l1->l_sigstk); 2616 l2->l_ctxlink = NULL; /* reset ucontext link */ 2617 2618 /* 2619 * Copy the credential so other references don't see our changes. 2620 * Test to see if this is necessary first, since in the common case 2621 * we won't need a private reference. 2622 */ 2623 if (kauth_cred_geteuid(l2->l_cred) != kauth_cred_getsvuid(l2->l_cred) || 2624 kauth_cred_getegid(l2->l_cred) != kauth_cred_getsvgid(l2->l_cred)) { 2625 l2->l_cred = kauth_cred_copy(l2->l_cred); 2626 kauth_cred_setsvuid(l2->l_cred, kauth_cred_geteuid(l2->l_cred)); 2627 kauth_cred_setsvgid(l2->l_cred, kauth_cred_getegid(l2->l_cred)); 2628 } 2629 2630 /* Update the master credentials. */ 2631 if (l2->l_cred != p2->p_cred) { 2632 kauth_cred_t ocred; 2633 2634 kauth_cred_hold(l2->l_cred); 2635 mutex_enter(p2->p_lock); 2636 ocred = p2->p_cred; 2637 p2->p_cred = l2->l_cred; 2638 mutex_exit(p2->p_lock); 2639 kauth_cred_free(ocred); 2640 } 2641 2642 *child_ok = true; 2643 spawn_data->sed_refcnt = 2; /* child gets it as well */ 2644 #if 0 2645 l2->l_nopreempt = 1; /* start it non-preemptable */ 2646 #endif 2647 2648 /* 2649 * It's now safe for the scheduler and other processes to see the 2650 * child process. 2651 */ 2652 mutex_enter(proc_lock); 2653 2654 if (p1->p_session->s_ttyvp != NULL && p1->p_lflag & PL_CONTROLT) 2655 p2->p_lflag |= PL_CONTROLT; 2656 2657 LIST_INSERT_HEAD(&p1->p_children, p2, p_sibling); 2658 p2->p_exitsig = SIGCHLD; /* signal for parent on exit */ 2659 2660 if ((p1->p_slflag & (PSL_TRACEPOSIX_SPAWN|PSL_TRACED)) == 2661 (PSL_TRACEPOSIX_SPAWN|PSL_TRACED)) { 2662 proc_changeparent(p2, p1->p_pptr); 2663 p2->p_oppid = p1->p_pid; 2664 } 2665 2666 LIST_INSERT_AFTER(p1, p2, p_pglist); 2667 LIST_INSERT_HEAD(&allproc, p2, p_list); 2668 2669 p2->p_trace_enabled = trace_is_enabled(p2); 2670 #ifdef __HAVE_SYSCALL_INTERN 2671 (*p2->p_emul->e_syscall_intern)(p2); 2672 #endif 2673 2674 /* 2675 * Make child runnable, set start time, and add to run queue except 2676 * if the parent requested the child to start in SSTOP state. 2677 */ 2678 mutex_enter(p2->p_lock); 2679 2680 getmicrotime(&p2->p_stats->p_start); 2681 2682 lwp_lock(l2); 2683 KASSERT(p2->p_nrlwps == 1); 2684 KASSERT(l2->l_stat == LSIDL); 2685 p2->p_nrlwps = 1; 2686 p2->p_stat = SACTIVE; 2687 setrunnable(l2); 2688 /* LWP now unlocked */ 2689 2690 mutex_exit(p2->p_lock); 2691 mutex_exit(proc_lock); 2692 2693 cv_wait(&spawn_data->sed_cv_child_ready, &spawn_data->sed_mtx_child); 2694 error = spawn_data->sed_error; 2695 mutex_exit(&spawn_data->sed_mtx_child); 2696 spawn_exec_data_release(spawn_data); 2697 2698 rw_exit(&p1->p_reflock); 2699 rw_exit(&exec_lock); 2700 have_exec_lock = false; 2701 2702 *pid_res = pid; 2703 2704 if (error) 2705 return error; 2706 2707 if (p1->p_slflag & PSL_TRACED) { 2708 /* Paranoid check */ 2709 mutex_enter(proc_lock); 2710 if ((p1->p_slflag & (PSL_TRACEPOSIX_SPAWN|PSL_TRACED)) != 2711 (PSL_TRACEPOSIX_SPAWN|PSL_TRACED)) { 2712 mutex_exit(proc_lock); 2713 return 0; 2714 } 2715 2716 mutex_enter(p1->p_lock); 2717 eventswitch(TRAP_CHLD, PTRACE_POSIX_SPAWN, pid); 2718 } 2719 return 0; 2720 2721 error_exit: 2722 if (have_exec_lock) { 2723 execve_free_data(&spawn_data->sed_exec); 2724 rw_exit(&p1->p_reflock); 2725 rw_exit(&exec_lock); 2726 } 2727 mutex_exit(&spawn_data->sed_mtx_child); 2728 spawn_exec_data_release(spawn_data); 2729 2730 return error; 2731 } 2732 2733 int 2734 sys_posix_spawn(struct lwp *l1, const struct sys_posix_spawn_args *uap, 2735 register_t *retval) 2736 { 2737 /* { 2738 syscallarg(pid_t *) pid; 2739 syscallarg(const char *) path; 2740 syscallarg(const struct posix_spawn_file_actions *) file_actions; 2741 syscallarg(const struct posix_spawnattr *) attrp; 2742 syscallarg(char *const *) argv; 2743 syscallarg(char *const *) envp; 2744 } */ 2745 2746 int error; 2747 struct posix_spawn_file_actions *fa = NULL; 2748 struct posix_spawnattr *sa = NULL; 2749 pid_t pid; 2750 bool child_ok = false; 2751 rlim_t max_fileactions; 2752 proc_t *p = l1->l_proc; 2753 2754 error = check_posix_spawn(l1); 2755 if (error) { 2756 *retval = error; 2757 return 0; 2758 } 2759 2760 /* copy in file_actions struct */ 2761 if (SCARG(uap, file_actions) != NULL) { 2762 max_fileactions = 2 * uimin(p->p_rlimit[RLIMIT_NOFILE].rlim_cur, 2763 maxfiles); 2764 error = posix_spawn_fa_alloc(&fa, SCARG(uap, file_actions), 2765 max_fileactions); 2766 if (error) 2767 goto error_exit; 2768 } 2769 2770 /* copyin posix_spawnattr struct */ 2771 if (SCARG(uap, attrp) != NULL) { 2772 sa = kmem_alloc(sizeof(*sa), KM_SLEEP); 2773 error = copyin(SCARG(uap, attrp), sa, sizeof(*sa)); 2774 if (error) 2775 goto error_exit; 2776 } 2777 2778 /* 2779 * Do the spawn 2780 */ 2781 error = do_posix_spawn(l1, &pid, &child_ok, SCARG(uap, path), fa, sa, 2782 SCARG(uap, argv), SCARG(uap, envp), execve_fetch_element); 2783 if (error) 2784 goto error_exit; 2785 2786 if (error == 0 && SCARG(uap, pid) != NULL) 2787 error = copyout(&pid, SCARG(uap, pid), sizeof(pid)); 2788 2789 *retval = error; 2790 return 0; 2791 2792 error_exit: 2793 if (!child_ok) { 2794 (void)chgproccnt(kauth_cred_getuid(l1->l_cred), -1); 2795 atomic_dec_uint(&nprocs); 2796 2797 if (sa) 2798 kmem_free(sa, sizeof(*sa)); 2799 if (fa) 2800 posix_spawn_fa_free(fa, fa->len); 2801 } 2802 2803 *retval = error; 2804 return 0; 2805 } 2806 2807 void 2808 exec_free_emul_arg(struct exec_package *epp) 2809 { 2810 if (epp->ep_emul_arg_free != NULL) { 2811 KASSERT(epp->ep_emul_arg != NULL); 2812 (*epp->ep_emul_arg_free)(epp->ep_emul_arg); 2813 epp->ep_emul_arg_free = NULL; 2814 epp->ep_emul_arg = NULL; 2815 } else { 2816 KASSERT(epp->ep_emul_arg == NULL); 2817 } 2818 } 2819 2820 #ifdef DEBUG_EXEC 2821 static void 2822 dump_vmcmds(const struct exec_package * const epp, size_t x, int error) 2823 { 2824 struct exec_vmcmd *vp = &epp->ep_vmcmds.evs_cmds[0]; 2825 size_t j; 2826 2827 if (error == 0) 2828 DPRINTF(("vmcmds %u\n", epp->ep_vmcmds.evs_used)); 2829 else 2830 DPRINTF(("vmcmds %zu/%u, error %d\n", x, 2831 epp->ep_vmcmds.evs_used, error)); 2832 2833 for (j = 0; j < epp->ep_vmcmds.evs_used; j++) { 2834 DPRINTF(("vmcmd[%zu] = vmcmd_map_%s %#" 2835 PRIxVADDR"/%#"PRIxVSIZE" fd@%#" 2836 PRIxVSIZE" prot=0%o flags=%d\n", j, 2837 vp[j].ev_proc == vmcmd_map_pagedvn ? 2838 "pagedvn" : 2839 vp[j].ev_proc == vmcmd_map_readvn ? 2840 "readvn" : 2841 vp[j].ev_proc == vmcmd_map_zero ? 2842 "zero" : "*unknown*", 2843 vp[j].ev_addr, vp[j].ev_len, 2844 vp[j].ev_offset, vp[j].ev_prot, 2845 vp[j].ev_flags)); 2846 if (error != 0 && j == x) 2847 DPRINTF((" ^--- failed\n")); 2848 } 2849 } 2850 #endif 2851