xref: /openbsd-src/sys/kern/exec_elf.c (revision 7da957a54b2ce3a1a6ccaa9268622ac16a8a07aa)
1 /*	$OpenBSD: exec_elf.c,v 1.117 2015/09/28 20:32:59 deraadt Exp $	*/
2 
3 /*
4  * Copyright (c) 1996 Per Fogelstrom
5  * All rights reserved.
6  *
7  * Copyright (c) 1994 Christos Zoulas
8  * All rights reserved.
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  * 3. The name of the author may not be used to endorse or promote products
19  *    derived from this software without specific prior written permission
20  *
21  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
22  * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
23  * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
24  * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
25  * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
26  * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
27  * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
28  * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
29  * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
30  * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
31  *
32  */
33 
34 /*
35  * Copyright (c) 2001 Wasabi Systems, Inc.
36  * All rights reserved.
37  *
38  * Written by Jason R. Thorpe for Wasabi Systems, Inc.
39  *
40  * Redistribution and use in source and binary forms, with or without
41  * modification, are permitted provided that the following conditions
42  * are met:
43  * 1. Redistributions of source code must retain the above copyright
44  *    notice, this list of conditions and the following disclaimer.
45  * 2. Redistributions in binary form must reproduce the above copyright
46  *    notice, this list of conditions and the following disclaimer in the
47  *    documentation and/or other materials provided with the distribution.
48  * 3. All advertising materials mentioning features or use of this software
49  *    must display the following acknowledgement:
50  *	This product includes software developed for the NetBSD Project by
51  *	Wasabi Systems, Inc.
52  * 4. The name of Wasabi Systems, Inc. may not be used to endorse
53  *    or promote products derived from this software without specific prior
54  *    written permission.
55  *
56  * THIS SOFTWARE IS PROVIDED BY WASABI SYSTEMS, INC. ``AS IS'' AND
57  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
58  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
59  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL WASABI SYSTEMS, INC
60  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
61  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
62  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
63  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
64  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
65  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
66  * POSSIBILITY OF SUCH DAMAGE.
67  */
68 
69 #include <sys/param.h>
70 #include <sys/systm.h>
71 #include <sys/kernel.h>
72 #include <sys/proc.h>
73 #include <sys/malloc.h>
74 #include <sys/pool.h>
75 #include <sys/mount.h>
76 #include <sys/namei.h>
77 #include <sys/vnode.h>
78 #include <sys/core.h>
79 #include <sys/exec.h>
80 #include <sys/exec_elf.h>
81 #include <sys/file.h>
82 #include <sys/ptrace.h>
83 #include <sys/syscall.h>
84 #include <sys/signalvar.h>
85 #include <sys/stat.h>
86 
87 #include <sys/mman.h>
88 
89 #include <uvm/uvm_extern.h>
90 
91 #include <machine/reg.h>
92 #include <machine/exec.h>
93 
94 #ifdef COMPAT_LINUX
95 #include <compat/linux/linux_exec.h>
96 #endif
97 
98 struct ELFNAME(probe_entry) {
99 	int (*func)(struct proc *, struct exec_package *, char *,
100 	    u_long *);
101 } ELFNAME(probes)[] = {
102 	/* XXX - bogus, shouldn't be size independent.. */
103 #ifdef COMPAT_LINUX
104 	{ linux_elf_probe },
105 #endif
106 	{ NULL }
107 };
108 
109 int ELFNAME(load_file)(struct proc *, char *, struct exec_package *,
110 	struct elf_args *, Elf_Addr *);
111 int ELFNAME(check_header)(Elf_Ehdr *);
112 int ELFNAME(read_from)(struct proc *, struct vnode *, u_long, caddr_t, int);
113 void ELFNAME(load_psection)(struct exec_vmcmd_set *, struct vnode *,
114 	Elf_Phdr *, Elf_Addr *, Elf_Addr *, int *, int);
115 int ELFNAMEEND(coredump)(struct proc *, void *);
116 
117 extern char sigcode[], esigcode[];
118 #ifdef SYSCALL_DEBUG
119 extern char *syscallnames[];
120 #endif
121 
122 /* round up and down to page boundaries. */
123 #define ELF_ROUND(a, b)		(((a) + (b) - 1) & ~((b) - 1))
124 #define ELF_TRUNC(a, b)		((a) & ~((b) - 1))
125 
126 /*
127  * We limit the number of program headers to 32, this should
128  * be a reasonable limit for ELF, the most we have seen so far is 12
129  */
130 #define ELF_MAX_VALID_PHDR 32
131 
132 /*
133  * This is the basic elf emul. elf_probe_funcs may change to other emuls.
134  */
135 struct emul ELFNAMEEND(emul) = {
136 	"native",
137 	NULL,
138 	sendsig,
139 	SYS_syscall,
140 	SYS_MAXSYSCALL,
141 	sysent,
142 #ifdef SYSCALL_DEBUG
143 	syscallnames,
144 #else
145 	NULL,
146 #endif
147 	(sizeof(AuxInfo) * ELF_AUX_ENTRIES / sizeof(char *)),
148 	ELFNAME(copyargs),
149 	setregs,
150 	ELFNAME2(exec,fixup),
151 	ELFNAMEEND(coredump),
152 	sigcode,
153 	esigcode,
154 	EMUL_ENABLED | EMUL_NATIVE,
155 };
156 
157 /*
158  * Copy arguments onto the stack in the normal way, but add some
159  * space for extra information in case of dynamic binding.
160  */
161 void *
162 ELFNAME(copyargs)(struct exec_package *pack, struct ps_strings *arginfo,
163 		void *stack, void *argp)
164 {
165 	stack = copyargs(pack, arginfo, stack, argp);
166 	if (!stack)
167 		return (NULL);
168 
169 	/*
170 	 * Push space for extra arguments on the stack needed by
171 	 * dynamically linked binaries.
172 	 */
173 	if (pack->ep_emul_arg != NULL) {
174 		pack->ep_emul_argp = stack;
175 		stack = (char *)stack + ELF_AUX_ENTRIES * sizeof (AuxInfo);
176 	}
177 	return (stack);
178 }
179 
180 /*
181  * Check header for validity; return 0 for ok, ENOEXEC if error
182  */
183 int
184 ELFNAME(check_header)(Elf_Ehdr *ehdr)
185 {
186 	/*
187 	 * We need to check magic, class size, endianess, and version before
188 	 * we look at the rest of the Elf_Ehdr structure. These few elements
189 	 * are represented in a machine independent fashion.
190 	 */
191 	if (!IS_ELF(*ehdr) ||
192 	    ehdr->e_ident[EI_CLASS] != ELF_TARG_CLASS ||
193 	    ehdr->e_ident[EI_DATA] != ELF_TARG_DATA ||
194 	    ehdr->e_ident[EI_VERSION] != ELF_TARG_VER)
195 		return (ENOEXEC);
196 
197 	/* Now check the machine dependent header */
198 	if (ehdr->e_machine != ELF_TARG_MACH ||
199 	    ehdr->e_version != ELF_TARG_VER)
200 		return (ENOEXEC);
201 
202 	/* Don't allow an insane amount of sections. */
203 	if (ehdr->e_phnum > ELF_MAX_VALID_PHDR)
204 		return (ENOEXEC);
205 
206 	return (0);
207 }
208 
209 /*
210  * Load a psection at the appropriate address
211  */
212 void
213 ELFNAME(load_psection)(struct exec_vmcmd_set *vcset, struct vnode *vp,
214 	Elf_Phdr *ph, Elf_Addr *addr, Elf_Addr *size, int *prot, int flags)
215 {
216 	u_long msize, lsize, psize, rm, rf;
217 	long diff, offset, bdiff;
218 	Elf_Addr base;
219 
220 	/*
221 	 * If the user specified an address, then we load there.
222 	 */
223 	if (*addr != ELFDEFNNAME(NO_ADDR)) {
224 		if (ph->p_align > 1) {
225 			*addr = ELF_TRUNC(*addr, ph->p_align);
226 			diff = ph->p_vaddr - ELF_TRUNC(ph->p_vaddr, ph->p_align);
227 			/* page align vaddr */
228 			base = *addr + trunc_page(ph->p_vaddr)
229 			    - ELF_TRUNC(ph->p_vaddr, ph->p_align);
230 		} else {
231 			diff = 0;
232 			base = *addr + trunc_page(ph->p_vaddr) - ph->p_vaddr;
233 		}
234 	} else {
235 		*addr = ph->p_vaddr;
236 		if (ph->p_align > 1)
237 			*addr = ELF_TRUNC(*addr, ph->p_align);
238 		base = trunc_page(ph->p_vaddr);
239 		diff = ph->p_vaddr - *addr;
240 	}
241 	bdiff = ph->p_vaddr - trunc_page(ph->p_vaddr);
242 
243 	*prot |= (ph->p_flags & PF_R) ? PROT_READ : 0;
244 	*prot |= (ph->p_flags & PF_W) ? PROT_WRITE : 0;
245 	*prot |= (ph->p_flags & PF_X) ? PROT_EXEC : 0;
246 
247 	msize = ph->p_memsz + diff;
248 	offset = ph->p_offset - bdiff;
249 	lsize = ph->p_filesz + bdiff;
250 	psize = round_page(lsize);
251 
252 	/*
253 	 * Because the pagedvn pager can't handle zero fill of the last
254 	 * data page if it's not page aligned we map the last page readvn.
255 	 */
256 	if (ph->p_flags & PF_W) {
257 		psize = trunc_page(lsize);
258 		if (psize > 0)
259 			NEW_VMCMD2(vcset, vmcmd_map_pagedvn, psize, base, vp,
260 			    offset, *prot, flags);
261 		if (psize != lsize) {
262 			NEW_VMCMD2(vcset, vmcmd_map_readvn, lsize - psize,
263 			    base + psize, vp, offset + psize, *prot, flags);
264 		}
265 	} else {
266 		NEW_VMCMD2(vcset, vmcmd_map_pagedvn, psize, base, vp, offset,
267 		    *prot, flags);
268 	}
269 
270 	/*
271 	 * Check if we need to extend the size of the segment
272 	 */
273 	rm = round_page(*addr + ph->p_memsz + diff);
274 	rf = round_page(*addr + ph->p_filesz + diff);
275 
276 	if (rm != rf) {
277 		NEW_VMCMD2(vcset, vmcmd_map_zero, rm - rf, rf, NULLVP, 0,
278 		    *prot, flags);
279 	}
280 	*size = msize;
281 }
282 
283 /*
284  * Read from vnode into buffer at offset.
285  */
286 int
287 ELFNAME(read_from)(struct proc *p, struct vnode *vp, u_long off, caddr_t buf,
288 	int size)
289 {
290 	int error;
291 	size_t resid;
292 
293 	if ((error = vn_rdwr(UIO_READ, vp, buf, size, off, UIO_SYSSPACE,
294 	    0, p->p_ucred, &resid, p)) != 0)
295 		return error;
296 	/*
297 	 * See if we got all of it
298 	 */
299 	if (resid != 0)
300 		return (ENOEXEC);
301 	return (0);
302 }
303 
304 /*
305  * Load a file (interpreter/library) pointed to by path [stolen from
306  * coff_load_shlib()]. Made slightly generic so it might be used externally.
307  */
308 int
309 ELFNAME(load_file)(struct proc *p, char *path, struct exec_package *epp,
310 	struct elf_args *ap, Elf_Addr *last)
311 {
312 	int error, i;
313 	struct nameidata nd;
314 	Elf_Ehdr eh;
315 	Elf_Phdr *ph = NULL;
316 	u_long phsize;
317 	Elf_Addr addr;
318 	struct vnode *vp;
319 	Elf_Phdr *base_ph = NULL;
320 	struct interp_ld_sec {
321 		Elf_Addr vaddr;
322 		u_long memsz;
323 	} loadmap[ELF_MAX_VALID_PHDR];
324 	int nload, idx = 0;
325 	Elf_Addr pos = *last;
326 	int file_align;
327 	int loop;
328 	size_t randomizequota = ELF_RANDOMIZE_LIMIT;
329 
330 	NDINIT(&nd, LOOKUP, FOLLOW | LOCKLEAF, UIO_SYSSPACE, path, p);
331 	if ((error = namei(&nd)) != 0) {
332 		return (error);
333 	}
334 	vp = nd.ni_vp;
335 	if (vp->v_type != VREG) {
336 		error = EACCES;
337 		goto bad;
338 	}
339 	if ((error = VOP_GETATTR(vp, epp->ep_vap, p->p_ucred, p)) != 0)
340 		goto bad;
341 	if (vp->v_mount->mnt_flag & MNT_NOEXEC) {
342 		error = EACCES;
343 		goto bad;
344 	}
345 	if ((error = VOP_ACCESS(vp, VREAD, p->p_ucred, p)) != 0)
346 		goto bad1;
347 	if ((error = ELFNAME(read_from)(p, nd.ni_vp, 0,
348 				    (caddr_t)&eh, sizeof(eh))) != 0)
349 		goto bad1;
350 
351 	if (ELFNAME(check_header)(&eh) || eh.e_type != ET_DYN) {
352 		error = ENOEXEC;
353 		goto bad1;
354 	}
355 
356 	ph = mallocarray(eh.e_phnum, sizeof(Elf_Phdr), M_TEMP, M_WAITOK);
357 	phsize = eh.e_phnum * sizeof(Elf_Phdr);
358 
359 	if ((error = ELFNAME(read_from)(p, nd.ni_vp, eh.e_phoff, (caddr_t)ph,
360 	    phsize)) != 0)
361 		goto bad1;
362 
363 	for (i = 0; i < eh.e_phnum; i++) {
364 		if (ph[i].p_type == PT_LOAD) {
365 			if (ph[i].p_filesz > ph[i].p_memsz)
366 				goto bad1;
367 			loadmap[idx].vaddr = trunc_page(ph[i].p_vaddr);
368 			loadmap[idx].memsz = round_page (ph[i].p_vaddr +
369 			    ph[i].p_memsz - loadmap[idx].vaddr);
370 			file_align = ph[i].p_align;
371 			idx++;
372 		}
373 	}
374 	nload = idx;
375 
376 	/*
377 	 * If no position to load the interpreter was set by a probe
378 	 * function, pick the same address that a non-fixed mmap(0, ..)
379 	 * would (i.e. something safely out of the way).
380 	 */
381 	if (pos == ELFDEFNNAME(NO_ADDR)) {
382 		pos = uvm_map_hint(p->p_vmspace, PROT_EXEC,
383 		    VM_MIN_ADDRESS, VM_MAXUSER_ADDRESS);
384 	}
385 
386 	pos = ELF_ROUND(pos, file_align);
387 	*last = epp->ep_interp_pos = pos;
388 	loop = 0;
389 	for (i = 0; i < nload;/**/) {
390 		vaddr_t	addr;
391 		struct	uvm_object *uobj;
392 		off_t	uoff;
393 		size_t	size;
394 
395 #ifdef this_needs_fixing
396 		if (i == 0) {
397 			uobj = &vp->v_uvm.u_obj;
398 			/* need to fix uoff */
399 		} else {
400 #endif
401 			uobj = NULL;
402 			uoff = 0;
403 #ifdef this_needs_fixing
404 		}
405 #endif
406 
407 		addr = trunc_page(pos + loadmap[i].vaddr);
408 		size =  round_page(addr + loadmap[i].memsz) - addr;
409 
410 		/* CRAP - map_findspace does not avoid daddr+BRKSIZ */
411 		if ((addr + size > (vaddr_t)p->p_vmspace->vm_daddr) &&
412 		    (addr < (vaddr_t)p->p_vmspace->vm_daddr + BRKSIZ))
413 			addr = round_page((vaddr_t)p->p_vmspace->vm_daddr +
414 			    BRKSIZ);
415 
416 		if (uvm_map_mquery(&p->p_vmspace->vm_map, &addr, size,
417 		    (i == 0 ? uoff : UVM_UNKNOWN_OFFSET), 0) != 0) {
418 			if (loop == 0) {
419 				loop = 1;
420 				i = 0;
421 				*last = epp->ep_interp_pos = pos = 0;
422 				continue;
423 			}
424 			error = ENOMEM;
425 			goto bad1;
426 		}
427 		if (addr != pos + loadmap[i].vaddr) {
428 			/* base changed. */
429 			pos = addr - trunc_page(loadmap[i].vaddr);
430 			pos = ELF_ROUND(pos,file_align);
431 			epp->ep_interp_pos = *last = pos;
432 			i = 0;
433 			continue;
434 		}
435 
436 		i++;
437 	}
438 
439 	/*
440 	 * Load all the necessary sections
441 	 */
442 	for (i = 0; i < eh.e_phnum; i++) {
443 		Elf_Addr size = 0;
444 		int prot = 0;
445 		int flags;
446 
447 		switch (ph[i].p_type) {
448 		case PT_LOAD:
449 			if (base_ph == NULL) {
450 				flags = VMCMD_BASE;
451 				addr = *last;
452 				base_ph = &ph[i];
453 			} else {
454 				flags = VMCMD_RELATIVE;
455 				addr = ph[i].p_vaddr - base_ph->p_vaddr;
456 			}
457 			ELFNAME(load_psection)(&epp->ep_vmcmds, nd.ni_vp,
458 			    &ph[i], &addr, &size, &prot, flags);
459 			/* If entry is within this section it must be text */
460 			if (eh.e_entry >= ph[i].p_vaddr &&
461 			    eh.e_entry < (ph[i].p_vaddr + size)) {
462  				epp->ep_entry = addr + eh.e_entry -
463 				    ELF_TRUNC(ph[i].p_vaddr,ph[i].p_align);
464 				ap->arg_interp = addr;
465 			}
466 			addr += size;
467 			break;
468 
469 		case PT_DYNAMIC:
470 		case PT_PHDR:
471 		case PT_NOTE:
472 			break;
473 
474 		case PT_OPENBSD_RANDOMIZE:
475 			if (ph[i].p_memsz > randomizequota) {
476 				error = ENOMEM;
477 				goto bad1;
478 			}
479 			randomizequota -= ph[i].p_memsz;
480 			NEW_VMCMD(&epp->ep_vmcmds, vmcmd_randomize,
481 			    ph[i].p_memsz, ph[i].p_vaddr + pos, NULLVP, 0, 0);
482 			break;
483 
484 		default:
485 			break;
486 		}
487 	}
488 
489 	vn_marktext(nd.ni_vp);
490 
491 bad1:
492 	VOP_CLOSE(nd.ni_vp, FREAD, p->p_ucred, p);
493 bad:
494 	free(ph, M_TEMP, phsize);
495 
496 	*last = addr;
497 	vput(nd.ni_vp);
498 	return (error);
499 }
500 
501 /*
502  * Prepare an Elf binary's exec package
503  *
504  * First, set of the various offsets/lengths in the exec package.
505  *
506  * Then, mark the text image busy (so it can be demand paged) or error out if
507  * this is not possible.  Finally, set up vmcmds for the text, data, bss, and
508  * stack segments.
509  */
510 int
511 ELFNAME2(exec,makecmds)(struct proc *p, struct exec_package *epp)
512 {
513 	Elf_Ehdr *eh = epp->ep_hdr;
514 	Elf_Phdr *ph, *pp, *base_ph = NULL;
515 	Elf_Addr phdr = 0, exe_base = 0;
516 	int error, i, has_phdr = 0;
517 	char *interp = NULL;
518 	u_long pos = 0, phsize;
519 	size_t randomizequota = ELF_RANDOMIZE_LIMIT;
520 
521 	if (epp->ep_hdrvalid < sizeof(Elf_Ehdr))
522 		return (ENOEXEC);
523 
524 	if (ELFNAME(check_header)(eh) ||
525 	   (eh->e_type != ET_EXEC && eh->e_type != ET_DYN))
526 		return (ENOEXEC);
527 
528 	/*
529 	 * check if vnode is in open for writing, because we want to demand-
530 	 * page out of it.  if it is, don't do it, for various reasons.
531 	 */
532 	if (epp->ep_vp->v_writecount != 0) {
533 #ifdef DIAGNOSTIC
534 		if (epp->ep_vp->v_flag & VTEXT)
535 			panic("exec: a VTEXT vnode has writecount != 0");
536 #endif
537 		return (ETXTBSY);
538 	}
539 	/*
540 	 * Allocate space to hold all the program headers, and read them
541 	 * from the file
542 	 */
543 	ph = mallocarray(eh->e_phnum, sizeof(Elf_Phdr), M_TEMP, M_WAITOK);
544 	phsize = eh->e_phnum * sizeof(Elf_Phdr);
545 
546 	if ((error = ELFNAME(read_from)(p, epp->ep_vp, eh->e_phoff, (caddr_t)ph,
547 	    phsize)) != 0)
548 		goto bad;
549 
550 	epp->ep_tsize = ELFDEFNNAME(NO_ADDR);
551 	epp->ep_dsize = ELFDEFNNAME(NO_ADDR);
552 
553 	for (i = 0, pp = ph; i < eh->e_phnum; i++, pp++) {
554 		if (pp->p_type == PT_INTERP && !interp) {
555 			if (pp->p_filesz < 2 || pp->p_filesz > MAXPATHLEN)
556 				goto bad;
557 			interp = pool_get(&namei_pool, PR_WAITOK);
558 			if ((error = ELFNAME(read_from)(p, epp->ep_vp,
559 			    pp->p_offset, interp, pp->p_filesz)) != 0) {
560 				goto bad;
561 			}
562 			if (interp[pp->p_filesz - 1] != '\0')
563 				goto bad;
564 		} else if (pp->p_type == PT_LOAD) {
565 			if (pp->p_filesz > pp->p_memsz) {
566 				error = EINVAL;
567 				goto bad;
568 			}
569 			if (base_ph == NULL)
570 				base_ph = pp;
571 		} else if (pp->p_type == PT_PHDR) {
572 			has_phdr = 1;
573 		}
574 	}
575 
576 	if (eh->e_type == ET_DYN) {
577 		/* need phdr and load sections for PIE */
578 		if (!has_phdr || base_ph == NULL) {
579 			error = EINVAL;
580 			goto bad;
581 		}
582 		/* randomize exe_base for PIE */
583 		exe_base = uvm_map_pie(base_ph->p_align);
584 	}
585 
586 	/*
587 	 * OK, we want a slightly different twist of the
588 	 * standard emulation package for "real" elf.
589 	 */
590 	epp->ep_emul = &ELFNAMEEND(emul);
591 	pos = ELFDEFNNAME(NO_ADDR);
592 
593 	/*
594 	 * On the same architecture, we may be emulating different systems.
595 	 * See which one will accept this executable.
596 	 *
597 	 * Probe functions would normally see if the interpreter (if any)
598 	 * exists. Emulation packages may possibly replace the interpreter in
599 	 * *interp with a changed path (/emul/xxx/<path>), and also
600 	 * set the ep_emul field in the exec package structure.
601 	 */
602 	error = ENOEXEC;
603 	if (eh->e_ident[EI_OSABI] != ELFOSABI_OPENBSD &&
604 	    ELFNAME(os_pt_note)(p, epp, epp->ep_hdr, "OpenBSD", 8, 4) != 0) {
605 		for (i = 0; ELFNAME(probes)[i].func != NULL && error; i++)
606 			error = (*ELFNAME(probes)[i].func)(p, epp, interp, &pos);
607 		if (error)
608 			goto bad;
609 	}
610 
611 	/*
612 	 * Load all the necessary sections
613 	 */
614 	for (i = 0, pp = ph; i < eh->e_phnum; i++, pp++) {
615 		Elf_Addr addr, size = 0;
616 		int prot = 0;
617 		int flags = 0;
618 
619 		switch (pp->p_type) {
620 		case PT_LOAD:
621 			if (exe_base != 0) {
622 				if (pp == base_ph) {
623 					flags = VMCMD_BASE;
624 					addr = exe_base;
625 				} else {
626 					flags = VMCMD_RELATIVE;
627 					addr = pp->p_vaddr - base_ph->p_vaddr;
628 				}
629 			} else
630 				addr = ELFDEFNNAME(NO_ADDR);
631 
632 			/*
633 			 * Calculates size of text and data segments
634 			 * by starting at first and going to end of last.
635 			 * 'rwx' sections are treated as data.
636 			 * this is correct for BSS_PLT, but may not be
637 			 * for DATA_PLT, is fine for TEXT_PLT.
638 			 */
639 			ELFNAME(load_psection)(&epp->ep_vmcmds, epp->ep_vp,
640 			    pp, &addr, &size, &prot, flags);
641 
642 			/*
643 			 * Update exe_base in case alignment was off.
644 			 * For PIE, addr is relative to exe_base so
645 			 * adjust it (non PIE exe_base is 0 so no change).
646 			 */
647 			if (flags == VMCMD_BASE)
648 				exe_base = addr;
649 			else
650 				addr += exe_base;
651 
652 			/*
653 			 * Decide whether it's text or data by looking
654 			 * at the protection of the section
655 			 */
656 			if (prot & PROT_WRITE) {
657 				/* data section */
658 				if (epp->ep_dsize == ELFDEFNNAME(NO_ADDR)) {
659 					epp->ep_daddr = addr;
660 					epp->ep_dsize = size;
661 				} else {
662 					if (addr < epp->ep_daddr) {
663 						epp->ep_dsize =
664 						    epp->ep_dsize +
665 						    epp->ep_daddr -
666 						    addr;
667 						epp->ep_daddr = addr;
668 					} else
669 						epp->ep_dsize = addr+size -
670 						    epp->ep_daddr;
671 				}
672 			} else if (prot & PROT_EXEC) {
673 				/* text section */
674 				if (epp->ep_tsize == ELFDEFNNAME(NO_ADDR)) {
675 					epp->ep_taddr = addr;
676 					epp->ep_tsize = size;
677 				} else {
678 					if (addr < epp->ep_taddr) {
679 						epp->ep_tsize =
680 						    epp->ep_tsize +
681 						    epp->ep_taddr -
682 						    addr;
683 						epp->ep_taddr = addr;
684 					} else
685 						epp->ep_tsize = addr+size -
686 						    epp->ep_taddr;
687 				}
688 			}
689 			break;
690 
691 		case PT_SHLIB:
692 			error = ENOEXEC;
693 			goto bad;
694 
695 		case PT_INTERP:
696 			/* Already did this one */
697 		case PT_DYNAMIC:
698 		case PT_NOTE:
699 			break;
700 
701 		case PT_PHDR:
702 			/* Note address of program headers (in text segment) */
703 			phdr = pp->p_vaddr;
704 			break;
705 
706 		case PT_OPENBSD_RANDOMIZE:
707 			if (ph[i].p_memsz > randomizequota) {
708 				error = ENOMEM;
709 				goto bad;
710 			}
711 			randomizequota -= ph[i].p_memsz;
712 			NEW_VMCMD(&epp->ep_vmcmds, vmcmd_randomize,
713 			    ph[i].p_memsz, ph[i].p_vaddr + exe_base, NULLVP, 0, 0);
714 			break;
715 
716 		default:
717 			/*
718 			 * Not fatal, we don't need to understand everything
719 			 * :-)
720 			 */
721 			break;
722 		}
723 	}
724 
725 	phdr += exe_base;
726 
727 	/*
728 	 * Strangely some linux programs may have all load sections marked
729 	 * writeable, in this case, textsize is not -1, but rather 0;
730 	 */
731 	if (epp->ep_tsize == ELFDEFNNAME(NO_ADDR))
732 		epp->ep_tsize = 0;
733 	/*
734 	 * Another possibility is that it has all load sections marked
735 	 * read-only.  Fake a zero-sized data segment right after the
736 	 * text segment.
737 	 */
738 	if (epp->ep_dsize == ELFDEFNNAME(NO_ADDR)) {
739 		epp->ep_daddr = round_page(epp->ep_taddr + epp->ep_tsize);
740 		epp->ep_dsize = 0;
741 	}
742 
743 	epp->ep_interp = interp;
744 	epp->ep_entry = eh->e_entry + exe_base;
745 
746 	/*
747 	 * Check if we found a dynamically linked binary and arrange to load
748 	 * its interpreter when the exec file is released.
749 	 */
750 	if (interp || eh->e_type == ET_DYN) {
751 		struct elf_args *ap;
752 
753 		ap = malloc(sizeof(*ap), M_TEMP, M_WAITOK);
754 
755 		ap->arg_phaddr = phdr;
756 		ap->arg_phentsize = eh->e_phentsize;
757 		ap->arg_phnum = eh->e_phnum;
758 		ap->arg_entry = eh->e_entry + exe_base;
759 		ap->arg_interp = exe_base;
760 
761 		epp->ep_emul_arg = ap;
762 		epp->ep_emul_argsize = sizeof *ap;
763 		epp->ep_interp_pos = pos;
764 	}
765 
766 	free(ph, M_TEMP, phsize);
767 	vn_marktext(epp->ep_vp);
768 	return (exec_setup_stack(p, epp));
769 
770 bad:
771 	if (interp)
772 		pool_put(&namei_pool, interp);
773 	free(ph, M_TEMP, phsize);
774 	kill_vmcmds(&epp->ep_vmcmds);
775 	if (error == 0)
776 		return (ENOEXEC);
777 	return (error);
778 }
779 
780 /*
781  * Phase II of load. It is now safe to load the interpreter. Info collected
782  * when loading the program is available for setup of the interpreter.
783  */
784 int
785 ELFNAME2(exec,fixup)(struct proc *p, struct exec_package *epp)
786 {
787 	char	*interp;
788 	int	error = 0;
789 	struct	elf_args *ap;
790 	AuxInfo ai[ELF_AUX_ENTRIES], *a;
791 	Elf_Addr	pos = epp->ep_interp_pos;
792 
793 	if (epp->ep_emul_arg == NULL) {
794 		return (0);
795 	}
796 
797 	interp = epp->ep_interp;
798 	ap = epp->ep_emul_arg;
799 
800 	if (interp &&
801 	    (error = ELFNAME(load_file)(p, interp, epp, ap, &pos)) != 0) {
802 		free(ap, M_TEMP, epp->ep_emul_argsize);
803 		pool_put(&namei_pool, interp);
804 		kill_vmcmds(&epp->ep_vmcmds);
805 		return (error);
806 	}
807 	/*
808 	 * We have to do this ourselves...
809 	 */
810 	error = exec_process_vmcmds(p, epp);
811 
812 	/*
813 	 * Push extra arguments on the stack needed by dynamically
814 	 * linked binaries
815 	 */
816 	if (error == 0) {
817 		a = ai;
818 
819 		a->au_id = AUX_phdr;
820 		a->au_v = ap->arg_phaddr;
821 		a++;
822 
823 		a->au_id = AUX_phent;
824 		a->au_v = ap->arg_phentsize;
825 		a++;
826 
827 		a->au_id = AUX_phnum;
828 		a->au_v = ap->arg_phnum;
829 		a++;
830 
831 		a->au_id = AUX_pagesz;
832 		a->au_v = PAGE_SIZE;
833 		a++;
834 
835 		a->au_id = AUX_base;
836 		a->au_v = ap->arg_interp;
837 		a++;
838 
839 		a->au_id = AUX_flags;
840 		a->au_v = 0;
841 		a++;
842 
843 		a->au_id = AUX_entry;
844 		a->au_v = ap->arg_entry;
845 		a++;
846 
847 		a->au_id = AUX_null;
848 		a->au_v = 0;
849 		a++;
850 
851 		error = copyout(ai, epp->ep_emul_argp, sizeof ai);
852 	}
853 	free(ap, M_TEMP, epp->ep_emul_argsize);
854 	if (interp)
855 		pool_put(&namei_pool, interp);
856 	return (error);
857 }
858 
859 /*
860  * Older ELF binaries use EI_ABIVERSION (formerly EI_BRAND) to brand
861  * executables.  Newer ELF binaries use EI_OSABI instead.
862  */
863 char *
864 ELFNAME(check_brand)(Elf_Ehdr *eh)
865 {
866 	if (eh->e_ident[EI_ABIVERSION] == '\0')
867 		return (NULL);
868 	return (&eh->e_ident[EI_ABIVERSION]);
869 }
870 
871 int
872 ELFNAME(os_pt_note)(struct proc *p, struct exec_package *epp, Elf_Ehdr *eh,
873 	char *os_name, size_t name_size, size_t desc_size)
874 {
875 	Elf_Phdr *hph, *ph;
876 	Elf_Note *np = NULL;
877 	size_t phsize;
878 	int error;
879 
880 	hph = mallocarray(eh->e_phnum, sizeof(Elf_Phdr), M_TEMP, M_WAITOK);
881 	phsize = eh->e_phnum * sizeof(Elf_Phdr);
882 	if ((error = ELFNAME(read_from)(p, epp->ep_vp, eh->e_phoff,
883 	    (caddr_t)hph, phsize)) != 0)
884 		goto out1;
885 
886 	for (ph = hph;  ph < &hph[eh->e_phnum]; ph++) {
887 		if (ph->p_type != PT_NOTE ||
888 		    ph->p_filesz > 1024 ||
889 		    ph->p_filesz < sizeof(Elf_Note) + name_size)
890 			continue;
891 
892 		np = malloc(ph->p_filesz, M_TEMP, M_WAITOK);
893 		if ((error = ELFNAME(read_from)(p, epp->ep_vp, ph->p_offset,
894 		    (caddr_t)np, ph->p_filesz)) != 0)
895 			goto out2;
896 
897 #if 0
898 		if (np->type != ELF_NOTE_TYPE_OSVERSION) {
899 			free(np, M_TEMP, ph->p_filesz);
900 			np = NULL;
901 			continue;
902 		}
903 #endif
904 
905 		/* Check the name and description sizes. */
906 		if (np->namesz != name_size ||
907 		    np->descsz != desc_size)
908 			goto out3;
909 
910 		if (memcmp((np + 1), os_name, name_size))
911 			goto out3;
912 
913 		/* XXX: We could check for the specific emulation here */
914 		/* All checks succeeded. */
915 		error = 0;
916 		goto out2;
917 	}
918 
919 out3:
920 	error = ENOEXEC;
921 out2:
922 	free(np, M_TEMP, ph->p_filesz);
923 out1:
924 	free(hph, M_TEMP, phsize);
925 	return error;
926 }
927 
928 struct countsegs_state {
929 	int	npsections;
930 };
931 
932 int	ELFNAMEEND(coredump_countsegs)(struct proc *, void *,
933 	    struct uvm_coredump_state *);
934 
935 struct writesegs_state {
936 	Elf_Phdr *psections;
937 	off_t	secoff;
938 };
939 
940 int	ELFNAMEEND(coredump_writeseghdrs)(struct proc *, void *,
941 	    struct uvm_coredump_state *);
942 
943 int	ELFNAMEEND(coredump_notes)(struct proc *, void *, size_t *);
944 int	ELFNAMEEND(coredump_note)(struct proc *, void *, size_t *);
945 int	ELFNAMEEND(coredump_writenote)(struct proc *, void *, Elf_Note *,
946 	    const char *, void *);
947 
948 #define	ELFROUNDSIZE	4	/* XXX Should it be sizeof(Elf_Word)? */
949 #define	elfround(x)	roundup((x), ELFROUNDSIZE)
950 
951 int
952 ELFNAMEEND(coredump)(struct proc *p, void *cookie)
953 {
954 #ifdef SMALL_KERNEL
955 	return EPERM;
956 #else
957 	Elf_Ehdr ehdr;
958 	Elf_Phdr *psections = NULL;
959 	struct countsegs_state cs;
960 	struct writesegs_state ws;
961 	off_t notestart, secstart, offset;
962 	size_t notesize, psectionslen;
963 	int error, i;
964 
965 	/*
966 	 * We have to make a total of 3 passes across the map:
967 	 *
968 	 *	1. Count the number of map entries (the number of
969 	 *	   PT_LOAD sections).
970 	 *
971 	 *	2. Write the P-section headers.
972 	 *
973 	 *	3. Write the P-sections.
974 	 */
975 
976 	/* Pass 1: count the entries. */
977 	cs.npsections = 0;
978 	error = uvm_coredump_walkmap(p, NULL,
979 	    ELFNAMEEND(coredump_countsegs), &cs);
980 	if (error)
981 		goto out;
982 
983 	/* Count the PT_NOTE section. */
984 	cs.npsections++;
985 
986 	/* Get the size of the notes. */
987 	error = ELFNAMEEND(coredump_notes)(p, NULL, &notesize);
988 	if (error)
989 		goto out;
990 
991 	memset(&ehdr, 0, sizeof(ehdr));
992 	memcpy(ehdr.e_ident, ELFMAG, SELFMAG);
993 	ehdr.e_ident[EI_CLASS] = ELF_TARG_CLASS;
994 	ehdr.e_ident[EI_DATA] = ELF_TARG_DATA;
995 	ehdr.e_ident[EI_VERSION] = EV_CURRENT;
996 	/* XXX Should be the OSABI/ABI version of the executable. */
997 	ehdr.e_ident[EI_OSABI] = ELFOSABI_SYSV;
998 	ehdr.e_ident[EI_ABIVERSION] = 0;
999 	ehdr.e_type = ET_CORE;
1000 	/* XXX This should be the e_machine of the executable. */
1001 	ehdr.e_machine = ELF_TARG_MACH;
1002 	ehdr.e_version = EV_CURRENT;
1003 	ehdr.e_entry = 0;
1004 	ehdr.e_phoff = sizeof(ehdr);
1005 	ehdr.e_shoff = 0;
1006 	ehdr.e_flags = 0;
1007 	ehdr.e_ehsize = sizeof(ehdr);
1008 	ehdr.e_phentsize = sizeof(Elf_Phdr);
1009 	ehdr.e_phnum = cs.npsections;
1010 	ehdr.e_shentsize = 0;
1011 	ehdr.e_shnum = 0;
1012 	ehdr.e_shstrndx = 0;
1013 
1014 	/* Write out the ELF header. */
1015 	error = coredump_write(cookie, UIO_SYSSPACE, &ehdr, sizeof(ehdr));
1016 	if (error)
1017 		goto out;
1018 
1019 	psections = mallocarray(cs.npsections, sizeof(Elf_Phdr),
1020 	    M_TEMP, M_WAITOK|M_ZERO);
1021 	psectionslen = cs.npsections * sizeof(Elf_Phdr);
1022 
1023 	offset = sizeof(ehdr);
1024 	notestart = offset + psectionslen;
1025 	secstart = notestart + notesize;
1026 
1027 	/* Pass 2: now write the P-section headers. */
1028 	ws.secoff = secstart;
1029 	ws.psections = psections;
1030 	error = uvm_coredump_walkmap(p, cookie,
1031 	    ELFNAMEEND(coredump_writeseghdrs), &ws);
1032 	if (error)
1033 		goto out;
1034 
1035 	/* Write out the PT_NOTE header. */
1036 	ws.psections->p_type = PT_NOTE;
1037 	ws.psections->p_offset = notestart;
1038 	ws.psections->p_vaddr = 0;
1039 	ws.psections->p_paddr = 0;
1040 	ws.psections->p_filesz = notesize;
1041 	ws.psections->p_memsz = 0;
1042 	ws.psections->p_flags = PF_R;
1043 	ws.psections->p_align = ELFROUNDSIZE;
1044 
1045 	error = coredump_write(cookie, UIO_SYSSPACE, psections, psectionslen);
1046 	if (error)
1047 		goto out;
1048 
1049 #ifdef DIAGNOSTIC
1050 	offset += psectionslen;
1051 	if (offset != notestart)
1052 		panic("coredump: offset %lld != notestart %lld",
1053 		    (long long) offset, (long long) notestart);
1054 #endif
1055 
1056 	/* Write out the notes. */
1057 	error = ELFNAMEEND(coredump_notes)(p, cookie, &notesize);
1058 	if (error)
1059 		goto out;
1060 
1061 #ifdef DIAGNOSTIC
1062 	offset += notesize;
1063 	if (offset != secstart)
1064 		panic("coredump: offset %lld != secstart %lld",
1065 		    (long long) offset, (long long) secstart);
1066 #endif
1067 
1068 	/* Pass 3: finally, write the sections themselves. */
1069 	for (i = 0; i < cs.npsections - 1; i++) {
1070 		if (psections[i].p_filesz == 0)
1071 			continue;
1072 
1073 #ifdef DIAGNOSTIC
1074 		if (offset != psections[i].p_offset)
1075 			panic("coredump: offset %lld != p_offset[%d] %lld",
1076 			    (long long) offset, i,
1077 			    (long long) psections[i].p_filesz);
1078 #endif
1079 
1080 		error = coredump_write(cookie, UIO_USERSPACE,
1081 		    (void *)(vaddr_t)psections[i].p_vaddr,
1082 		    psections[i].p_filesz);
1083 		if (error)
1084 			goto out;
1085 
1086 		coredump_unmap(cookie, (vaddr_t)psections[i].p_vaddr,
1087 		    (vaddr_t)psections[i].p_vaddr + psections[i].p_filesz);
1088 
1089 #ifdef DIAGNOSTIC
1090 		offset += psections[i].p_filesz;
1091 #endif
1092 	}
1093 
1094 out:
1095 	free(psections, M_TEMP, psectionslen);
1096 	return (error);
1097 #endif
1098 }
1099 
1100 int
1101 ELFNAMEEND(coredump_countsegs)(struct proc *p, void *iocookie,
1102     struct uvm_coredump_state *us)
1103 {
1104 #ifndef SMALL_KERNEL
1105 	struct countsegs_state *cs = us->cookie;
1106 
1107 	cs->npsections++;
1108 #endif
1109 	return (0);
1110 }
1111 
1112 int
1113 ELFNAMEEND(coredump_writeseghdrs)(struct proc *p, void *iocookie,
1114     struct uvm_coredump_state *us)
1115 {
1116 #ifndef SMALL_KERNEL
1117 	struct writesegs_state *ws = us->cookie;
1118 	Elf_Phdr phdr;
1119 	vsize_t size, realsize;
1120 
1121 	size = us->end - us->start;
1122 	realsize = us->realend - us->start;
1123 
1124 	phdr.p_type = PT_LOAD;
1125 	phdr.p_offset = ws->secoff;
1126 	phdr.p_vaddr = us->start;
1127 	phdr.p_paddr = 0;
1128 	phdr.p_filesz = realsize;
1129 	phdr.p_memsz = size;
1130 	phdr.p_flags = 0;
1131 	if (us->prot & PROT_READ)
1132 		phdr.p_flags |= PF_R;
1133 	if (us->prot & PROT_WRITE)
1134 		phdr.p_flags |= PF_W;
1135 	if (us->prot & PROT_EXEC)
1136 		phdr.p_flags |= PF_X;
1137 	phdr.p_align = PAGE_SIZE;
1138 
1139 	ws->secoff += phdr.p_filesz;
1140 	*ws->psections++ = phdr;
1141 #endif
1142 
1143 	return (0);
1144 }
1145 
1146 int
1147 ELFNAMEEND(coredump_notes)(struct proc *p, void *iocookie, size_t *sizep)
1148 {
1149 #ifndef SMALL_KERNEL
1150 	struct ps_strings pss;
1151 	struct iovec iov;
1152 	struct uio uio;
1153 	struct elfcore_procinfo cpi;
1154 	Elf_Note nhdr;
1155 	struct process *pr = p->p_p;
1156 	struct proc *q;
1157 	size_t size, notesize;
1158 	int error;
1159 
1160 	size = 0;
1161 
1162 	/* First, write an elfcore_procinfo. */
1163 	notesize = sizeof(nhdr) + elfround(sizeof("OpenBSD")) +
1164 	    elfround(sizeof(cpi));
1165 	if (iocookie) {
1166 		memset(&cpi, 0, sizeof(cpi));
1167 
1168 		cpi.cpi_version = ELFCORE_PROCINFO_VERSION;
1169 		cpi.cpi_cpisize = sizeof(cpi);
1170 		cpi.cpi_signo = p->p_sisig;
1171 		cpi.cpi_sigcode = p->p_sicode;
1172 
1173 		cpi.cpi_sigpend = p->p_siglist;
1174 		cpi.cpi_sigmask = p->p_sigmask;
1175 		cpi.cpi_sigignore = pr->ps_sigacts->ps_sigignore;
1176 		cpi.cpi_sigcatch = pr->ps_sigacts->ps_sigcatch;
1177 
1178 		cpi.cpi_pid = pr->ps_pid;
1179 		cpi.cpi_ppid = pr->ps_pptr->ps_pid;
1180 		cpi.cpi_pgrp = pr->ps_pgid;
1181 		if (pr->ps_session->s_leader)
1182 			cpi.cpi_sid = pr->ps_session->s_leader->ps_pid;
1183 		else
1184 			cpi.cpi_sid = 0;
1185 
1186 		cpi.cpi_ruid = p->p_ucred->cr_ruid;
1187 		cpi.cpi_euid = p->p_ucred->cr_uid;
1188 		cpi.cpi_svuid = p->p_ucred->cr_svuid;
1189 
1190 		cpi.cpi_rgid = p->p_ucred->cr_rgid;
1191 		cpi.cpi_egid = p->p_ucred->cr_gid;
1192 		cpi.cpi_svgid = p->p_ucred->cr_svgid;
1193 
1194 		(void)strlcpy(cpi.cpi_name, p->p_comm, sizeof(cpi.cpi_name));
1195 
1196 		nhdr.namesz = sizeof("OpenBSD");
1197 		nhdr.descsz = sizeof(cpi);
1198 		nhdr.type = NT_OPENBSD_PROCINFO;
1199 
1200 		error = ELFNAMEEND(coredump_writenote)(p, iocookie, &nhdr,
1201 		    "OpenBSD", &cpi);
1202 		if (error)
1203 			return (error);
1204 	}
1205 	size += notesize;
1206 
1207 	/* Second, write an NT_OPENBSD_AUXV note. */
1208 	notesize = sizeof(nhdr) + elfround(sizeof("OpenBSD")) +
1209 	    elfround(pr->ps_emul->e_arglen * sizeof(char *));
1210 	if (iocookie) {
1211 		iov.iov_base = &pss;
1212 		iov.iov_len = sizeof(pss);
1213 		uio.uio_iov = &iov;
1214 		uio.uio_iovcnt = 1;
1215 		uio.uio_offset = (off_t)pr->ps_strings;
1216 		uio.uio_resid = sizeof(pss);
1217 		uio.uio_segflg = UIO_SYSSPACE;
1218 		uio.uio_rw = UIO_READ;
1219 		uio.uio_procp = NULL;
1220 
1221 		error = uvm_io(&p->p_vmspace->vm_map, &uio, 0);
1222 		if (error)
1223 			return (error);
1224 
1225 		if (pss.ps_envstr == NULL)
1226 			return (EIO);
1227 
1228 		nhdr.namesz = sizeof("OpenBSD");
1229 		nhdr.descsz = pr->ps_emul->e_arglen * sizeof(char *);
1230 		nhdr.type = NT_OPENBSD_AUXV;
1231 
1232 		error = coredump_write(iocookie, UIO_SYSSPACE,
1233 		    &nhdr, sizeof(nhdr));
1234 		if (error)
1235 			return (error);
1236 
1237 		error = coredump_write(iocookie, UIO_SYSSPACE,
1238 		    "OpenBSD", elfround(nhdr.namesz));
1239 		if (error)
1240 			return (error);
1241 
1242 		error = coredump_write(iocookie, UIO_USERSPACE,
1243 		    pss.ps_envstr + pss.ps_nenvstr + 1, nhdr.descsz);
1244 		if (error)
1245 			return (error);
1246 	}
1247 	size += notesize;
1248 
1249 #ifdef PT_WCOOKIE
1250 	notesize = sizeof(nhdr) + elfround(sizeof("OpenBSD")) +
1251 	    elfround(sizeof(register_t));
1252 	if (iocookie) {
1253 		register_t wcookie;
1254 
1255 		nhdr.namesz = sizeof("OpenBSD");
1256 		nhdr.descsz = sizeof(register_t);
1257 		nhdr.type = NT_OPENBSD_WCOOKIE;
1258 
1259 		wcookie = process_get_wcookie(p);
1260 		error = ELFNAMEEND(coredump_writenote)(p, iocookie, &nhdr,
1261 		    "OpenBSD", &wcookie);
1262 		if (error)
1263 			return (error);
1264 	}
1265 	size += notesize;
1266 #endif
1267 
1268 	/*
1269 	 * Now write the register info for the thread that caused the
1270 	 * coredump.
1271 	 */
1272 	error = ELFNAMEEND(coredump_note)(p, iocookie, &notesize);
1273 	if (error)
1274 		return (error);
1275 	size += notesize;
1276 
1277 	/*
1278 	 * Now, for each thread, write the register info and any other
1279 	 * per-thread notes.  Since we're dumping core, all the other
1280 	 * threads in the process have been stopped and the list can't
1281 	 * change.
1282 	 */
1283 	TAILQ_FOREACH(q, &pr->ps_threads, p_thr_link) {
1284 		if (q == p)		/* we've taken care of this thread */
1285 			continue;
1286 		error = ELFNAMEEND(coredump_note)(q, iocookie, &notesize);
1287 		if (error)
1288 			return (error);
1289 		size += notesize;
1290 	}
1291 
1292 	*sizep = size;
1293 #endif
1294 	return (0);
1295 }
1296 
1297 int
1298 ELFNAMEEND(coredump_note)(struct proc *p, void *iocookie, size_t *sizep)
1299 {
1300 #ifndef SMALL_KERNEL
1301 	Elf_Note nhdr;
1302 	int size, notesize, error;
1303 	int namesize;
1304 	char name[64+ELFROUNDSIZE];
1305 	struct reg intreg;
1306 #ifdef PT_GETFPREGS
1307 	struct fpreg freg;
1308 #endif
1309 
1310 	size = 0;
1311 
1312 	snprintf(name, sizeof(name)-ELFROUNDSIZE, "%s@%d",
1313 	    "OpenBSD", p->p_pid);
1314 	namesize = strlen(name) + 1;
1315 	memset(name + namesize, 0, elfround(namesize) - namesize);
1316 
1317 	notesize = sizeof(nhdr) + elfround(namesize) + elfround(sizeof(intreg));
1318 	if (iocookie) {
1319 		error = process_read_regs(p, &intreg);
1320 		if (error)
1321 			return (error);
1322 
1323 		nhdr.namesz = namesize;
1324 		nhdr.descsz = sizeof(intreg);
1325 		nhdr.type = NT_OPENBSD_REGS;
1326 
1327 		error = ELFNAMEEND(coredump_writenote)(p, iocookie, &nhdr,
1328 		    name, &intreg);
1329 		if (error)
1330 			return (error);
1331 
1332 	}
1333 	size += notesize;
1334 
1335 #ifdef PT_GETFPREGS
1336 	notesize = sizeof(nhdr) + elfround(namesize) + elfround(sizeof(freg));
1337 	if (iocookie) {
1338 		error = process_read_fpregs(p, &freg);
1339 		if (error)
1340 			return (error);
1341 
1342 		nhdr.namesz = namesize;
1343 		nhdr.descsz = sizeof(freg);
1344 		nhdr.type = NT_OPENBSD_FPREGS;
1345 
1346 		error = ELFNAMEEND(coredump_writenote)(p, iocookie, &nhdr,
1347 		    name, &freg);
1348 		if (error)
1349 			return (error);
1350 	}
1351 	size += notesize;
1352 #endif
1353 
1354 	*sizep = size;
1355 	/* XXX Add hook for machdep per-LWP notes. */
1356 #endif
1357 	return (0);
1358 }
1359 
1360 int
1361 ELFNAMEEND(coredump_writenote)(struct proc *p, void *cookie, Elf_Note *nhdr,
1362     const char *name, void *data)
1363 {
1364 #ifdef SMALL_KERNEL
1365 	return EPERM;
1366 #else
1367 	int error;
1368 
1369 	error = coredump_write(cookie, UIO_SYSSPACE, nhdr, sizeof(*nhdr));
1370 	if (error)
1371 		return error;
1372 
1373 	error = coredump_write(cookie, UIO_SYSSPACE, name,
1374 	    elfround(nhdr->namesz));
1375 	if (error)
1376 		return error;
1377 
1378 	return coredump_write(cookie, UIO_SYSSPACE, data, nhdr->descsz);
1379 #endif
1380 }
1381