NOTE: libkvm, w, ps, 'top', and any other utility which depends on struct

      proc or any VM system structure will have to be rebuilt!!!

Much needed overhaul of the VM system. Included in this first round of
changes:

1) Improved pager interfaces: init, alloc, dealloc, getpages, putpages,
   haspage, and sync operations are supported. The haspage interface now
   provides information about clusterability. All pager routines now take
   struct vm_object's instead of "pagers".

2) Improved data structures. In the previous paradigm, there is constant
   confusion caused by pagers being both a data structure ("allocate a
   pager") and a collection of routines. The idea of a pager structure has
   escentially been eliminated. Objects now have types, and this type is
   used to index the appropriate pager. In most cases, items in the pager
   structure were duplicated in the object data structure and thus were
   unnecessary. In the few cases that remained, a un_pager structure union
   was created in the object to contain these items.

3) Because of the cleanup of #1 & #2, a lot of unnecessary layering can now
   be removed. For instance, vm_object_enter(), vm_object_lookup(),
   vm_object_remove(), and the associated object hash list were some of the
   things that were removed.

4) simple_lock's removed. Discussion with several people reveals that the
   SMP locking primitives used in the VM system aren't likely the mechanism
   that we'll be adopting. Even if it were, the locking that was in the code
   was very inadequate and would have to be mostly re-done anyway. The
   locking in a uni-processor kernel was a no-op but went a long way toward
   making the code difficult to read and debug.

5) Places that attempted to kludge-up the fact that we don't have kernel
   thread support have been fixed to reflect the reality that we are really
   dealing with processes, not threads. The VM system didn't have complete
   thread support, so the comments and mis-named routines were just wrong.
   We now use tsleep and wakeup directly in the lock routines, for instance.

6) Where appropriate, the pagers have been improved, especially in the
   pager_alloc routines. Most of the pager_allocs have been rewritten and
   are now faster and easier to maintain.

7) The pagedaemon pageout clustering algorithm has been rewritten and
   now tries harder to output an even number of pages before and after
   the requested page. This is sort of the reverse of the ideal pagein
   algorithm and should provide better overall performance.

8) Unnecessary (incorrect) casts to caddr_t in calls to tsleep & wakeup
   have been removed. Some other unnecessary casts have also been removed.

9) Some almost useless debugging code removed.

10) Terminology of shadow objects vs. backing objects straightened out.
    The fact that the vm_object data structure escentially had this
    backwards really confused things. The use of "shadow" and "backing
    object" throughout the code is now internally consistent and correct
    in the Mach terminology.

11) Several minor bug fixes, including one in the vm daemon that caused
    0 RSS objects to not get purged as intended.

12) A "default pager" has now been created which cleans up the transition
    of objects to the "swap" type. The previous checks throughout the code
    for swp->pg_data != NULL were really ugly. This change also provides
    the rudiments for future backing of "anonymous" memory by something
    other than the swap pager (via the vnode pager, for example), and it
    allows the decision about which of these pagers to use to be made
    dynamically (although will need some additional decision code to do
    this, of course).

13) (dyson) MAP_COPY has been deprecated and the corresponding "copy
    object" code has been removed. MAP_COPY was undocumented and non-
    standard. It was furthermore broken in several ways which caused its
    behavior to degrade to MAP_PRIVATE. Binaries that use MAP_COPY will
    continue to work correctly, but via the slightly different semantics
    of MAP_PRIVATE.

14) (dyson) Sharing maps have been removed. It's marginal usefulness in a
    threads design can be worked around in other ways. Both #12 and #13
    were done to simplify the code and improve readability and maintain-
    ability. (As were most all of these changes)

TODO:

1) Rewrite most of the vnode pager to use VOP_GETPAGES/PUTPAGES. Doing
   this will reduce the vnode pager to a mere fraction of its current size.

2) Rewrite vm_fault and the swap/vnode pagers to use the clustering
   information provided by the new haspage pager interface. This will
   substantially reduce the overhead by eliminating a large number of
   VOP_BMAP() calls. The VOP_BMAP() filesystem interface should be
   improved to provide both a "behind" and "ahead" indication of
   contiguousness.

3) Implement the extended features of pager_haspage in swap_pager_haspage().
   It currently just says 0 pages ahead/behind.

4) Re-implement the swap device (swstrategy) in a more elegant way, perhaps
   via a much more general mechanism that could also be used for disk
   striping of regular filesystems.

5) Do something to improve the architecture of vm_object_collapse(). The
   fact that it makes calls into the swap pager and knows too much about
   how the swap pager operates really bothers me. It also doesn't allow
   for collapsing of non-swap pager objects ("unnamed" objects backed by
   other pagers).

1 files changed, 74 insertions, 167 deletions

diff --git a/sys/vm/device_pager.c b/sys/vm/device_pager.c
index 8ce0bc16..a7b0a9b 100644
--- a/sys/vm/device_pager.c
+++ b/sys/vm/device_pager.c
@@ -36,11 +36,7 @@
  * SUCH DAMAGE.
  *
  *	@(#)device_pager.c	8.1 (Berkeley) 6/11/93
- * $Id: device_pager.c,v 1.10 1995/05/18 02:59:18 davidg Exp $
- */
-
-/*
- * Page to/from special files.
+ * $Id: device_pager.c,v 1.11 1995/05/30 08:15:46 rgrimes Exp $
  */
 
 #include <sys/param.h>
@@ -53,52 +49,35 @@
 #include <vm/vm.h>
 #include <vm/vm_kern.h>
 #include <vm/vm_page.h>
+#include <vm/vm_pager.h>
 #include <vm/device_pager.h>
 
-struct pagerlst dev_pager_list;	/* list of managed devices */
-struct pglist dev_pager_fakelist;	/* list of available vm_page_t's */
-
-#ifdef DEBUG
-int dpagerdebug;
+struct pagerlst dev_pager_object_list;	/* list of device pager objects */
+TAILQ_HEAD(, vm_page) dev_pager_fakelist;	/* list of available vm_page_t's */
 
-#define	DDB_FOLLOW	0x01
-#define DDB_INIT	0x02
-#define DDB_ALLOC	0x04
-#define DDB_FAIL	0x08
-#endif
-
-static vm_pager_t dev_pager_alloc __P((void *, vm_size_t, vm_prot_t, vm_offset_t));
-static void dev_pager_dealloc __P((vm_pager_t));
-static int dev_pager_getpage __P((vm_pager_t, vm_page_t, boolean_t));
-static boolean_t dev_pager_haspage __P((vm_pager_t, vm_offset_t));
-static void dev_pager_init __P((void));
-static int dev_pager_putpage __P((vm_pager_t, vm_page_t, boolean_t));
 static vm_page_t dev_pager_getfake __P((vm_offset_t));
 static void dev_pager_putfake __P((vm_page_t));
 
+static int dev_pager_alloc_lock, dev_pager_alloc_lock_want;
+
 struct pagerops devicepagerops = {
 	dev_pager_init,
 	dev_pager_alloc,
 	dev_pager_dealloc,
-	dev_pager_getpage,
-	0,
-	dev_pager_putpage,
-	0,
-	dev_pager_haspage
+	dev_pager_getpages,
+	dev_pager_putpages,
+	dev_pager_haspage,
+	NULL
 };
 
-static void
+void
 dev_pager_init()
 {
-#ifdef DEBUG
-	if (dpagerdebug & DDB_FOLLOW)
-		printf("dev_pager_init()\n");
-#endif
-	TAILQ_INIT(&dev_pager_list);
+	TAILQ_INIT(&dev_pager_object_list);
 	TAILQ_INIT(&dev_pager_fakelist);
 }
 
-static vm_pager_t
+vm_object_t
 dev_pager_alloc(handle, size, prot, foff)
 	void *handle;
 	vm_size_t size;
@@ -106,25 +85,10 @@ dev_pager_alloc(handle, size, prot, foff)
 	vm_offset_t foff;
 {
 	dev_t dev;
-	vm_pager_t pager;
 	int (*mapfunc) ();
 	vm_object_t object;
-	dev_pager_t devp;
 	unsigned int npages, off;
 
-#ifdef DEBUG
-	if (dpagerdebug & DDB_FOLLOW)
-		printf("dev_pager_alloc(%x, %x, %x, %x)\n",
-		    handle, size, prot, foff);
-#endif
-#ifdef DIAGNOSTIC
-	/*
-	 * Pageout to device, should never happen.
-	 */
-	if (handle == NULL)
-		panic("dev_pager_alloc called");
-#endif
-
 	/*
 	 * Make sure this device can be mapped.
 	 */
@@ -151,124 +115,74 @@ dev_pager_alloc(handle, size, prot, foff)
 			return (NULL);
 
 	/*
+	 * Lock to prevent object creation race contion.
+	 */
+	while (dev_pager_alloc_lock) {
+		dev_pager_alloc_lock_want++;
+		tsleep(&dev_pager_alloc_lock, PVM, "dvpall", 0);
+		dev_pager_alloc_lock_want--;
+	}
+	dev_pager_alloc_lock = 1;
+
+	/*
 	 * Look up pager, creating as necessary.
 	 */
-top:
-	pager = vm_pager_lookup(&dev_pager_list, handle);
-	if (pager == NULL) {
-		/*
-		 * Allocate and initialize pager structs
-		 */
-		pager = (vm_pager_t) malloc(sizeof *pager, M_VMPAGER, M_WAITOK);
-		if (pager == NULL)
-			return (NULL);
-		devp = (dev_pager_t) malloc(sizeof *devp, M_VMPGDATA, M_WAITOK);
-		if (devp == NULL) {
-			free((caddr_t) pager, M_VMPAGER);
-			return (NULL);
-		}
-		pager->pg_handle = handle;
-		pager->pg_ops = &devicepagerops;
-		pager->pg_type = PG_DEVICE;
-		pager->pg_data = (caddr_t) devp;
-		TAILQ_INIT(&devp->devp_pglist);
+	object = vm_pager_object_lookup(&dev_pager_object_list, handle);
+	if (object == NULL) {
 		/*
 		 * Allocate object and associate it with the pager.
 		 */
-		object = devp->devp_object = vm_object_allocate(foff + size);
-		object->flags &= ~OBJ_INTERNAL;
-		vm_object_enter(object, pager);
-		object->pager = pager;
-		/*
-		 * Finally, put it on the managed list so other can find it.
-		 * First we re-lookup in case someone else beat us to this
-		 * point (due to blocking in the various mallocs).  If so, we
-		 * free everything and start over.
-		 */
-		if (vm_pager_lookup(&dev_pager_list, handle)) {
-			free((caddr_t) devp, M_VMPGDATA);
-			free((caddr_t) pager, M_VMPAGER);
-			goto top;
-		}
-		TAILQ_INSERT_TAIL(&dev_pager_list, pager, pg_list);
-#ifdef DEBUG
-		if (dpagerdebug & DDB_ALLOC) {
-			printf("dev_pager_alloc: pager %x devp %x object %x\n",
-			    pager, devp, object);
-			vm_object_print(object, FALSE);
-		}
-#endif
+		object = vm_object_allocate(OBJT_DEVICE, foff + size);
+		object->handle = handle;
+		TAILQ_INIT(&object->un_pager.devp.devp_pglist);
+		TAILQ_INSERT_TAIL(&dev_pager_object_list, object, pager_object_list);
 	} else {
 		/*
 		 * Gain a reference to the object.
 		 */
-		object = vm_object_lookup(pager);
+		vm_object_reference(object);
 		if (foff + size > object->size)
 			object->size = foff + size;
-#ifdef DIAGNOSTIC
-		devp = (dev_pager_t) pager->pg_data;
-		if (object != devp->devp_object)
-			panic("dev_pager_setup: bad object");
-#endif
 	}
-	return (pager);
+
+	dev_pager_alloc_lock = 0;
+	if (dev_pager_alloc_lock_want)
+		wakeup(&dev_pager_alloc_lock);
+
+	return (object);
 }
 
-static void
-dev_pager_dealloc(pager)
-	vm_pager_t pager;
-{
-	dev_pager_t devp;
+void
+dev_pager_dealloc(object)
 	vm_object_t object;
+{
 	vm_page_t m;
 
-#ifdef DEBUG
-	if (dpagerdebug & DDB_FOLLOW)
-		printf("dev_pager_dealloc(%x)\n", pager);
-#endif
-	TAILQ_REMOVE(&dev_pager_list, pager, pg_list);
-	/*
-	 * Get the object. Note: cannot use vm_object_lookup since object has
-	 * already been removed from the hash chain.
-	 */
-	devp = (dev_pager_t) pager->pg_data;
-	object = devp->devp_object;
-#ifdef DEBUG
-	if (dpagerdebug & DDB_ALLOC)
-		printf("dev_pager_dealloc: devp %x object %x\n", devp, object);
-#endif
+	TAILQ_REMOVE(&dev_pager_object_list, object, pager_object_list);
 	/*
 	 * Free up our fake pages.
 	 */
-	while ((m = devp->devp_pglist.tqh_first) != 0) {
-		TAILQ_REMOVE(&devp->devp_pglist, m, pageq);
+	while ((m = object->un_pager.devp.devp_pglist.tqh_first) != 0) {
+		TAILQ_REMOVE(&object->un_pager.devp.devp_pglist, m, pageq);
 		dev_pager_putfake(m);
 	}
-	free((caddr_t) devp, M_VMPGDATA);
-	free((caddr_t) pager, M_VMPAGER);
 }
 
-static int
-dev_pager_getpage(pager, m, sync)
-	vm_pager_t pager;
-	vm_page_t m;
-	boolean_t sync;
+int
+dev_pager_getpages(object, m, count, reqpage)
+	vm_object_t object;
+	vm_page_t *m;
+	int count;
+	int reqpage;
 {
-	register vm_object_t object;
 	vm_offset_t offset, paddr;
 	vm_page_t page;
 	dev_t dev;
-	int s;
+	int i, s;
 	int (*mapfunc) (), prot;
 
-#ifdef DEBUG
-	if (dpagerdebug & DDB_FOLLOW)
-		printf("dev_pager_getpage(%x, %x)\n", pager, m);
-#endif
-
-	object = m->object;
-	dev = (dev_t) (u_long) pager->pg_handle;
-	offset = m->offset + object->paging_offset;
+	dev = (dev_t) (u_long) object->handle;
+	offset = m[reqpage]->offset + object->paging_offset;
 	prot = PROT_READ;	/* XXX should pass in? */
 	mapfunc = cdevsw[major(dev)].d_mmap;
 
@@ -281,49 +195,44 @@ dev_pager_getpage(pager, m, sync)
 		panic("dev_pager_getpage: map function returns error");
 #endif
 	/*
-	 * Replace the passed in page with our own fake page and free up the
-	 * original.
+	 * Replace the passed in reqpage page with our own fake page and free up the
+	 * all of the original pages.
 	 */
 	page = dev_pager_getfake(paddr);
-	TAILQ_INSERT_TAIL(&((dev_pager_t) pager->pg_data)->devp_pglist,
-	    page, pageq);
-	vm_object_lock(object);
-	vm_page_lock_queues();
-	PAGE_WAKEUP(m);
-	vm_page_free(m);
-	vm_page_unlock_queues();
+	TAILQ_INSERT_TAIL(&object->un_pager.devp.devp_pglist, page, pageq);
+	for (i = 0; i < count; i++) {
+		PAGE_WAKEUP(m[i]);
+		vm_page_free(m[i]);
+	}
 	s = splhigh();
 	vm_page_insert(page, object, offset);
 	splx(s);
-	vm_object_unlock(object);
 
 	return (VM_PAGER_OK);
 }
 
-static int
-dev_pager_putpage(pager, m, sync)
-	vm_pager_t pager;
-	vm_page_t m;
+int
+dev_pager_putpages(object, m, count, sync, rtvals)
+	vm_object_t object;
+	vm_page_t *m;
+	int count;
 	boolean_t sync;
+	int *rtvals;
 {
-#ifdef DEBUG
-	if (dpagerdebug & DDB_FOLLOW)
-		printf("dev_pager_putpage(%x, %x)\n", pager, m);
-#endif
-	if (pager == NULL)
-		return 0;
 	panic("dev_pager_putpage called");
 }
 
-static boolean_t
-dev_pager_haspage(pager, offset)
-	vm_pager_t pager;
+boolean_t
+dev_pager_haspage(object, offset, before, after)
+	vm_object_t object;
 	vm_offset_t offset;
+	int *before;
+	int *after;
 {
-#ifdef DEBUG
-	if (dpagerdebug & DDB_FOLLOW)
-		printf("dev_pager_haspage(%x, %x)\n", pager, offset);
-#endif
+	if (before != NULL)
+		*before = 0;
+	if (after != NULL)
+		*after = 0;
 	return (TRUE);
 }
 
@@ -345,8 +254,8 @@ dev_pager_getfake(paddr)
 	TAILQ_REMOVE(&dev_pager_fakelist, m, pageq);
 
 	m->flags = PG_BUSY | PG_FICTITIOUS;
-	m->dirty = 0;
 	m->valid = VM_PAGE_BITS_ALL;
+	m->dirty = 0;
 	m->busy = 0;
 	m->bmapped = 0;
 
@@ -360,9 +269,7 @@ static void
 dev_pager_putfake(m)
 	vm_page_t m;
 {
-#ifdef DIAGNOSTIC
 	if (!(m->flags & PG_FICTITIOUS))
 		panic("dev_pager_putfake: bad page");
-#endif
 	TAILQ_INSERT_TAIL(&dev_pager_fakelist, m, pageq);
 }