- provos@cvs.openbsd.org 2002/03/18 17:50:31

     [auth-bsdauth.c auth-options.c auth-rh-rsa.c auth-rsa.c auth-skey.c auth.h
      auth1.c auth2-chall.c auth2.c kex.c kex.h kexdh.c kexgex.c servconf.c
      session.h servconf.h serverloop.c session.c sshd.c]
     integrate privilege separated openssh; its turned off by default for now.
     work done by me and markus@

applied, but outside of ensure that smaller code bits migrated with
their owners.. no work was tried to 'fix' it to work. =)  Later project!

1 files changed, 329 insertions, 0 deletions

diff --git a/monitor_mm.c b/monitor_mm.c
new file mode 100644
index 00000000..bce98c93
--- /dev/null
+++ b/monitor_mm.c
@@ -0,0 +1,329 @@
+/*
+ * Copyright 2002 Niels Provos <provos@citi.umich.edu>
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
+ * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
+ * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
+ * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
+ * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include "includes.h"
+RCSID("$OpenBSD: monitor_mm.c,v 1.3 2002/03/19 10:41:32 markus Exp $");
+
+#include <sys/mman.h>
+
+#include "ssh.h"
+#include "xmalloc.h"
+#include "log.h"
+#include "monitor_mm.h"
+
+static int
+mm_compare(struct mm_share *a, struct mm_share *b)
+{
+	return ((char *)a->address - (char *)b->address);
+}
+
+RB_GENERATE(mmtree, mm_share, next, mm_compare)
+
+static struct mm_share *
+mm_make_entry(struct mm_master *mm, struct mmtree *head,
+    void *address, size_t size)
+{
+	struct mm_share *tmp, *tmp2;
+
+	if (mm->mmalloc == NULL)
+		tmp = xmalloc(sizeof(struct mm_share));
+	else
+		tmp = mm_xmalloc(mm->mmalloc, sizeof(struct mm_share));
+	tmp->address = address;
+	tmp->size = size;
+
+	tmp2 = RB_INSERT(mmtree, head, tmp);
+	if (tmp2 != NULL)
+		fatal("mm_make_entry(%p): double address %p->%p(%d)",
+		    mm, tmp2, address, size);
+
+	return (tmp);
+}
+
+/* Creates a shared memory area of a certain size */
+
+struct mm_master *
+mm_create(struct mm_master *mmalloc, size_t size)
+{
+	void *address;
+	struct mm_master *mm;
+
+	if (mmalloc == NULL)
+		mm = xmalloc(sizeof(struct mm_master));
+	else
+		mm = mm_xmalloc(mmalloc, sizeof(struct mm_master));
+
+	/*
+	 * If the memory map has a mm_master it can be completely
+	 * shared including authentication between the child
+	 * and the client.
+	 */
+	mm->mmalloc = mmalloc;
+
+	address = mmap(NULL, size, PROT_WRITE|PROT_READ, MAP_ANON|MAP_SHARED,
+	    -1, 0);
+	if (address == MAP_FAILED)
+		fatal("mmap(%d)", size);
+
+	mm->address = address;
+	mm->size = size;
+
+	RB_INIT(&mm->rb_free);
+	RB_INIT(&mm->rb_allocated);
+
+	mm_make_entry(mm, &mm->rb_free, address, size);
+
+	return (mm);
+}
+
+/* Frees either the allocated or the free list */
+
+static void
+mm_freelist(struct mm_master *mmalloc, struct mmtree *head)
+{
+	struct mm_share *mms, *next;
+
+	for (mms = RB_ROOT(head); mms; mms = next) {
+		next = RB_NEXT(mmtree, head, mms);
+		RB_REMOVE(mmtree, head, mms);
+		if (mmalloc == NULL)
+			xfree(mms);
+		else
+			mm_free(mmalloc, mms);
+	}
+}
+
+/* Destroys a memory mapped area */
+
+void
+mm_destroy(struct mm_master *mm)
+{
+	mm_freelist(mm->mmalloc, &mm->rb_free);
+	mm_freelist(mm->mmalloc, &mm->rb_allocated);
+
+	if (munmap(mm->address, mm->size) == -1)
+		fatal("munmap(%p, %d)", mm->address, mm->size);
+	if (mm->mmalloc == NULL)
+		xfree(mm);
+	else
+		mm_free(mm->mmalloc, mm);
+}
+
+void *
+mm_xmalloc(struct mm_master *mm, size_t size)
+{
+	void *address;
+
+	address = mm_malloc(mm, size);
+	if (address == NULL)
+		fatal("%s: mm_malloc(%d)", __FUNCTION__, size);
+	return (address);
+}
+
+
+/* Allocates data from a memory mapped area */
+
+void *
+mm_malloc(struct mm_master *mm, size_t size)
+{
+	struct mm_share *mms, *tmp;
+
+	if (size == 0)
+		fatal("mm_malloc: try to allocate 0 space");
+
+	size = ((size + MM_MINSIZE - 1) / MM_MINSIZE) * MM_MINSIZE;
+
+	RB_FOREACH(mms, mmtree, &mm->rb_free) {
+		if (mms->size >= size)
+			break;
+	}
+
+	if (mms == NULL)
+		return (NULL);
+
+	/* Debug */
+	memset(mms->address, 0xd0, size);
+
+	tmp = mm_make_entry(mm, &mm->rb_allocated, mms->address, size);
+
+	/* Does not change order in RB tree */
+	mms->size -= size;
+	mms->address = (u_char *)mms->address + size;
+
+	if (mms->size == 0) {
+		RB_REMOVE(mmtree, &mm->rb_free, mms);
+		if (mm->mmalloc == NULL)
+			xfree(mms);
+		else
+			mm_free(mm->mmalloc, mms);
+	}
+
+	return (tmp->address);
+}
+
+/* Frees memory in a memory mapped area */
+
+void
+mm_free(struct mm_master *mm, void *address)
+{
+	struct mm_share *mms, *prev, tmp;
+
+	tmp.address = address;
+	mms = RB_FIND(mmtree, &mm->rb_allocated, &tmp);
+	if (mms == NULL)
+		fatal("mm_free(%p): can not find %p", mm, address);
+
+	/* Debug */
+	memset(mms->address, 0xd0, mms->size);
+
+	/* Remove from allocated list and insert in free list */
+	RB_REMOVE(mmtree, &mm->rb_allocated, mms);
+	if (RB_INSERT(mmtree, &mm->rb_free, mms) != NULL)
+		fatal("mm_free(%p): double address %p", mm, address);
+
+	/* Find previous entry */
+	prev = mms;
+	if (RB_LEFT(prev, next)) {
+		prev = RB_LEFT(prev, next);
+		while (RB_RIGHT(prev, next))
+			prev = RB_RIGHT(prev, next);
+	} else {
+		if (RB_PARENT(prev, next) &&
+		    (prev == RB_RIGHT(RB_PARENT(prev, next), next)))
+			prev = RB_PARENT(prev, next);
+		else {
+			while (RB_PARENT(prev, next) &&
+			    (prev == RB_LEFT(RB_PARENT(prev, next), next)))
+				prev = RB_PARENT(prev, next);
+			prev = RB_PARENT(prev, next);
+		}
+	}
+
+	/* Check if range does not overlap */
+	if (prev != NULL && MM_ADDRESS_END(prev) > address)
+		fatal("mm_free: memory corruption: %p(%d) > %p",
+		    prev->address, prev->size, address);
+
+	/* See if we can merge backwards */
+	if (prev != NULL && MM_ADDRESS_END(prev) == address) {
+		prev->size += mms->size;
+		RB_REMOVE(mmtree, &mm->rb_free, mms);
+		if (mm->mmalloc == NULL)
+			xfree(mms);
+		else
+			mm_free(mm->mmalloc, mms);
+	} else
+		prev = mms;
+
+	if (prev == NULL)
+		return;
+
+	/* Check if we can merge forwards */
+	mms = RB_NEXT(mmtree, &mm->rb_free, prev);
+	if (mms == NULL)
+		return;
+
+	if (MM_ADDRESS_END(prev) > mms->address)
+		fatal("mm_free: memory corruption: %p < %p(%d)",
+		    mms->address, prev->address, prev->size);
+	if (MM_ADDRESS_END(prev) != mms->address)
+		return;
+
+	prev->size += mms->size;
+	RB_REMOVE(mmtree, &mm->rb_free, mms);
+
+	if (mm->mmalloc == NULL)
+		xfree(mms);
+	else
+		mm_free(mm->mmalloc, mms);
+}
+
+static void
+mm_sync_list(struct mmtree *oldtree, struct mmtree *newtree,
+    struct mm_master *mm, struct mm_master *mmold)
+{
+	struct mm_master *mmalloc = mm->mmalloc;
+	struct mm_share *mms, *new;
+
+	/* Sync free list */
+	RB_FOREACH(mms, mmtree, oldtree) {
+		/* Check the values */
+		mm_memvalid(mmold, mms, sizeof(struct mm_share));
+		mm_memvalid(mm, mms->address, mms->size);
+
+		new = mm_xmalloc(mmalloc, sizeof(struct mm_share));
+		memcpy(new, mms, sizeof(struct mm_share));
+		RB_INSERT(mmtree, newtree, new);
+	}
+}
+
+void
+mm_share_sync(struct mm_master **pmm, struct mm_master **pmmalloc)
+{
+	struct mm_master *mm;
+	struct mm_master *mmalloc;
+	struct mm_master *mmold;
+	struct mmtree rb_free, rb_allocated;
+
+	debug3("%s: Share sync", __FUNCTION__);
+
+	mm = *pmm;
+	mmold = mm->mmalloc;
+	mm_memvalid(mmold, mm, sizeof(*mm));
+
+	mmalloc = mm_create(NULL, mm->size);
+	mm = mm_xmalloc(mmalloc, sizeof(struct mm_master));
+	memcpy(mm, *pmm, sizeof(struct mm_master));
+	mm->mmalloc = mmalloc;
+
+	rb_free = mm->rb_free;
+	rb_allocated = mm->rb_allocated;
+
+	RB_INIT(&mm->rb_free);
+	RB_INIT(&mm->rb_allocated);
+
+	mm_sync_list(&rb_free, &mm->rb_free, mm, mmold);
+	mm_sync_list(&rb_allocated, &mm->rb_allocated, mm, mmold);
+
+	mm_destroy(mmold);
+
+	*pmm = mm;
+	*pmmalloc = mmalloc;
+
+	debug3("%s: Share sync end", __FUNCTION__);
+}
+
+void
+mm_memvalid(struct mm_master *mm, void *address, size_t size)
+{
+	void *end = (u_char *)address + size;
+
+	if (address < mm->address)
+		fatal("mm_memvalid: address too small: %p", address);
+	if (end < address)
+		fatal("mm_memvalid: end < address: %p < %p", end, address);
+	if (end > (void *)((u_char *)mm->address + mm->size))
+		fatal("mm_memvalid: address too large: %p", address);
+}