]> git.saurik.com Git - apple/xnu.git/blobdiff - osfmk/i386/loose_ends.c
xnu-792.10.96.tar.gz
[apple/xnu.git] / osfmk / i386 / loose_ends.c
index 4ce9fd757987ac7d1d537ed9f9b9715955d62ecc..ada20c4acfba233633f3a6b8ec5869d6f2653072 100644 (file)
@@ -57,6 +57,7 @@
 #include <mach/i386/vm_param.h>
 #include <kern/kern_types.h>
 #include <kern/misc_protos.h>
+#include <sys/errno.h>
 #include <i386/param.h>
 #include <i386/misc_protos.h>
 #include <i386/cpu_data.h>
 #include <vm/vm_kern.h>
 #include <vm/vm_fault.h>
 
+#include <libkern/OSAtomic.h>
+#include <sys/kdebug.h>
+
+#if 0
+
+#undef KERNEL_DEBUG
+#define KERNEL_DEBUG KERNEL_DEBUG_CONSTANT
+#define KDEBUG 1
+
+#endif
+
 /* XXX - should be gone from here */
 extern void            invalidate_icache64(addr64_t addr, unsigned cnt, int phys);
 extern void            flush_dcache64(addr64_t addr, unsigned count, int phys);
 extern boolean_t       phys_page_exists(ppnum_t);
-extern pt_entry_t      *pmap_mapgetpte(vm_map_t, vm_offset_t);
 extern void            bcopy_no_overwrite(const char *from, char *to,vm_size_t bytes);
 extern void            pmap_set_reference(ppnum_t pn);
 extern void            mapping_set_mod(ppnum_t pa); 
 extern void            mapping_set_ref(ppnum_t pn);
-extern void            switch_to_serial_console(void);
-extern kern_return_t   copyp2p(vm_offset_t     source,
-                               vm_offset_t     dest,
-                               unsigned int    size,
-                               unsigned int    flush_action);
+
 extern void            fillPage(ppnum_t pa, unsigned int fill);
 extern void            ovbcopy(const char      *from,
                                char            *to,
@@ -97,71 +104,28 @@ bzero_phys(
           addr64_t src64,
           vm_size_t bytes)
 {
-  vm_offset_t src = low32(src64);
-  pt_entry_t save2;
-        mp_disable_preemption();
-       if (*(pt_entry_t *) CM2)
-               panic("bzero_phys: CMAP busy");
+        mapwindow_t *map;
+       pt_entry_t save;
 
-       *(pt_entry_t *) CM2 = INTEL_PTE_VALID | INTEL_PTE_RW | (src & PG_FRAME) | 
-         INTEL_PTE_REF | INTEL_PTE_MOD;
-       save2=*(pt_entry_t *)CM2;
-       invlpg((u_int)CA2);
-
-        bzero((void *)((unsigned int)CA2 | (src & INTEL_OFFMASK)), bytes);
-       if (save2 != *(pt_entry_t *)CM2)  panic("bzero_phys CMAP changed");
-       *(pt_entry_t *) CM2 = 0;
-       mp_enable_preemption();
-}
+        mp_disable_preemption();
+       map = pmap_get_mapwindow((pt_entry_t)(INTEL_PTE_VALID | INTEL_PTE_RW | ((pmap_paddr_t)src64 & PG_FRAME) | INTEL_PTE_REF | INTEL_PTE_MOD));
+       if (map == 0) {
+               panic("bzero_phys: CMAP busy");
+       }
+       save = *map->prv_CMAP;
 
-/*
- * copy 'size' bytes from physical to physical address
- * the caller must validate the physical ranges 
- *
- * if flush_action == 0, no cache flush necessary
- * if flush_action == 1, flush the source
- * if flush_action == 2, flush the dest
- * if flush_action == 3, flush both source and dest
- */
+       invlpg((uintptr_t)map->prv_CADDR);
 
-kern_return_t
-copyp2p(vm_offset_t    source,
-       vm_offset_t     dest,
-       unsigned int    size,
-       unsigned int    flush_action)
-{
-
-        switch(flush_action) {
-       case 1:
-               flush_dcache(source, size, 1);
-               break;
-       case 2:
-               flush_dcache(dest, size, 1);
-               break;
-       case 3:
-               flush_dcache(source, size, 1);
-               flush_dcache(dest, size, 1);
-               break;
+       bzero((void *)((uintptr_t)map->prv_CADDR | ((uint32_t)src64 & INTEL_OFFMASK)), bytes);
 
-       }
-        bcopy_phys((addr64_t)source, (addr64_t)dest, (vm_size_t)size); /* Do a physical copy */
-
-        switch(flush_action) {
-       case 1:
-               flush_dcache(source, size, 1);
-               break;
-       case 2:
-               flush_dcache(dest, size, 1);
-               break;
-       case 3:
-               flush_dcache(source, size, 1);
-               flush_dcache(dest, size, 1);
-               break;
+       if (save != *map->prv_CMAP)
+               panic("bzero_phys: CMAP changed");
+       *map->prv_CMAP = 0;
 
-       }
-       return KERN_SUCCESS;
+       mp_enable_preemption();
 }
 
+
 /*
  * bcopy_phys - like bcopy but copies from/to physical addresses.
  */
@@ -172,30 +136,38 @@ bcopy_phys(
           addr64_t dst64,
           vm_size_t bytes)
 {
-  vm_offset_t src = low32(src64);
-  vm_offset_t dst = low32(dst64);
-  pt_entry_t save1,save2;
-  /* ensure we stay within a page */
-  if ( (((src & (NBPG-1)) + bytes) > NBPG) ||
-       (((dst & (NBPG-1)) + bytes) > NBPG) ) panic("bcopy_phys");
-  mp_disable_preemption();
-       if (*(pt_entry_t *) CM1 || *(pt_entry_t *) CM2)
-               panic("bcopy_phys: CMAP busy");
-
-       *(pt_entry_t *) CM1 = INTEL_PTE_VALID | (src & PG_FRAME) | INTEL_PTE_REF;
-       *(pt_entry_t *) CM2 = INTEL_PTE_VALID | INTEL_PTE_RW | (dst & PG_FRAME) | 
-         INTEL_PTE_REF | INTEL_PTE_MOD;
-       save1 = *(pt_entry_t *)CM1;save2 = *(pt_entry_t *)CM2;
-       invlpg((u_int)CA1);
-       invlpg((u_int)CA2);
-
-        bcopy((void *) ((uintptr_t)CA1 | (src & INTEL_OFFMASK)),
-             (void *) ((uintptr_t)CA2 | (dst & INTEL_OFFMASK)), bytes);
-       if ( (save1 != *(pt_entry_t *)CM1) || (save2 != *(pt_entry_t *)CM2)) panic("bcopy_phys CMAP changed");
-       *(pt_entry_t *) CM1 = 0;
-       *(pt_entry_t *) CM2 = 0;
-       mp_enable_preemption();
+        mapwindow_t *src_map, *dst_map;
+       pt_entry_t save1, save2;
+
+       /* ensure we stay within a page */
+       if ( ((((uint32_t)src64 & (NBPG-1)) + bytes) > NBPG) || ((((uint32_t)dst64 & (NBPG-1)) + bytes) > NBPG) ) {
+               panic("bcopy_phys alignment");
+       }
+       mp_disable_preemption();
+
+       src_map = pmap_get_mapwindow((pt_entry_t)(INTEL_PTE_VALID | ((pmap_paddr_t)src64 & PG_FRAME) | INTEL_PTE_REF));
+       dst_map = pmap_get_mapwindow((pt_entry_t)(INTEL_PTE_VALID | INTEL_PTE_RW | ((pmap_paddr_t)dst64 & PG_FRAME) |
+                                                 INTEL_PTE_REF | INTEL_PTE_MOD));
+
+       if (src_map == 0 || dst_map == 0) {
+               panic("bcopy_phys: CMAP busy");
+       }
+       save1 = *src_map->prv_CMAP;
+       save2 = *dst_map->prv_CMAP;
+
+       invlpg((uintptr_t)src_map->prv_CADDR);
+       invlpg((uintptr_t)dst_map->prv_CADDR);
+
+       bcopy((void *) ((uintptr_t)src_map->prv_CADDR | ((uint32_t)src64 & INTEL_OFFMASK)),
+             (void *) ((uintptr_t)dst_map->prv_CADDR | ((uint32_t)dst64 & INTEL_OFFMASK)), bytes);
+
+       if ( (save1 != *src_map->prv_CMAP) || (save2 != *dst_map->prv_CMAP))
+               panic("bcopy_phys CMAP changed");
 
+       *src_map->prv_CMAP = 0;
+       *dst_map->prv_CMAP = 0;
+
+       mp_enable_preemption();
 }
 
 /* 
@@ -230,217 +202,242 @@ ovbcopy(
 
 
 static unsigned int
-ml_phys_read_data( vm_offset_t paddr, int size )
+ml_phys_read_data(pmap_paddr_t paddr, int size )
 {
-    unsigned int result;
-    pt_entry_t save;
-    mp_disable_preemption();
-       if (*(pt_entry_t *) CM3)
-               panic("ml_phys_read_data: CMAP busy");
+        mapwindow_t *map;
+       unsigned int result;
+       pt_entry_t save;
 
-       *(pt_entry_t *) CM3 = INTEL_PTE_VALID | (paddr & PG_FRAME) | INTEL_PTE_REF;
-        save = *(pt_entry_t *)CM3;
-       invlpg((u_int)CA3);
+       mp_disable_preemption();
+       map = pmap_get_mapwindow((pt_entry_t)(INTEL_PTE_VALID | (paddr & PG_FRAME) | INTEL_PTE_REF));
+       if (map == 0) {
+               panic("ml_phys_read_data: CMAP busy");
+       }
 
+       save = *map->prv_CMAP;
+       invlpg((uintptr_t)map->prv_CADDR);
 
         switch (size) {
             unsigned char s1;
             unsigned short s2;
         case 1:
-            s1 = *(unsigned char *)((unsigned int)CA3 | (paddr & INTEL_OFFMASK));
+            s1 = *(unsigned char *)((uintptr_t)map->prv_CADDR | ((uint32_t)paddr & INTEL_OFFMASK));
             result = s1;
             break;
         case 2:
-            s2 = *(unsigned short *)((unsigned int)CA3 | (paddr & INTEL_OFFMASK));
+            s2 = *(unsigned short *)((uintptr_t)map->prv_CADDR | ((uint32_t)paddr & INTEL_OFFMASK));
             result = s2;
             break;
         case 4:
         default:
-            result = *(unsigned int *)((unsigned int)CA3 | (paddr & INTEL_OFFMASK));
+            result = *(unsigned int *)((uintptr_t)map->prv_CADDR | ((uint32_t)paddr & INTEL_OFFMASK));
             break;
         }
 
-       if (save != *(pt_entry_t *)CM3) panic("ml_phys_read_data CMAP changed");
-       *(pt_entry_t *) CM3 = 0;
+       if (save != *map->prv_CMAP)
+               panic("ml_phys_read_data CMAP changed");
+        *map->prv_CMAP = 0;
        mp_enable_preemption();
+
         return result;
 }
 
 static unsigned long long
-ml_phys_read_long_long( vm_offset_t paddr )
+ml_phys_read_long_long(pmap_paddr_t paddr )
 {
-    unsigned long long result;
-    pt_entry_t save;
-    mp_disable_preemption();
-       if (*(pt_entry_t *) CM3)
-               panic("ml_phys_read_data: CMAP busy");
+        mapwindow_t *map;
+       unsigned long long result;
+       pt_entry_t save;
 
-       *(pt_entry_t *) CM3 = INTEL_PTE_VALID | (paddr & PG_FRAME) | INTEL_PTE_REF;
-        save = *(pt_entry_t *)CM3;
-       invlpg((u_int)CA3);
+       mp_disable_preemption();
+       map = pmap_get_mapwindow((pt_entry_t)(INTEL_PTE_VALID | (paddr & PG_FRAME) | INTEL_PTE_REF));
 
-        result = *(unsigned long long *)((unsigned int)CA3 | (paddr & INTEL_OFFMASK));
+       if (map == 0) {
+               panic("ml_phys_read_long_long: CMAP busy");
+       }
+
+       save = *map->prv_CMAP;
+       invlpg((uintptr_t)map->prv_CADDR);
 
-       if (save != *(pt_entry_t *)CM3) panic("ml_phys_read_data CMAP changed");
-       *(pt_entry_t *) CM3 = 0;
+       result = *(unsigned long long *)((uintptr_t)map->prv_CADDR | ((uint32_t)paddr & INTEL_OFFMASK));
+
+       if (save != *map->prv_CMAP)
+               panic("ml_phys_read_long_long CMAP changed");
+       *map->prv_CMAP = 0;
        mp_enable_preemption();
-        return result;
+
+       return result;
 }
 
-unsigned int ml_phys_read( vm_offset_t paddr)
+
+
+unsigned int ml_phys_read(vm_offset_t paddr)
 {
-    return ml_phys_read_data(paddr, 4);
+        return ml_phys_read_data((pmap_paddr_t)paddr, 4);
 }
 
 unsigned int ml_phys_read_word(vm_offset_t paddr) {
-    return ml_phys_read_data(paddr, 4);
+
+        return ml_phys_read_data((pmap_paddr_t)paddr, 4);
 }
 
 unsigned int ml_phys_read_64(addr64_t paddr64)
 {
-    return ml_phys_read_data(low32(paddr64), 4);
+        return ml_phys_read_data((pmap_paddr_t)paddr64, 4);
 }
 
 unsigned int ml_phys_read_word_64(addr64_t paddr64)
 {
-    return ml_phys_read_data(low32(paddr64), 4);
+        return ml_phys_read_data((pmap_paddr_t)paddr64, 4);
 }
 
 unsigned int ml_phys_read_half(vm_offset_t paddr)
 {
-    return ml_phys_read_data(paddr, 2);
+        return ml_phys_read_data((pmap_paddr_t)paddr, 2);
 }
 
 unsigned int ml_phys_read_half_64(addr64_t paddr64)
 {
-    return ml_phys_read_data(low32(paddr64), 2);
+        return ml_phys_read_data((pmap_paddr_t)paddr64, 2);
 }
 
 unsigned int ml_phys_read_byte(vm_offset_t paddr)
 {
-    return ml_phys_read_data(paddr, 1);
+        return ml_phys_read_data((pmap_paddr_t)paddr, 1);
 }
 
 unsigned int ml_phys_read_byte_64(addr64_t paddr64)
 {
-    return ml_phys_read_data(low32(paddr64), 1);
+        return ml_phys_read_data((pmap_paddr_t)paddr64, 1);
 }
 
 unsigned long long ml_phys_read_double(vm_offset_t paddr)
 {
-    return ml_phys_read_long_long(paddr);
+        return ml_phys_read_long_long((pmap_paddr_t)paddr);
 }
 
-unsigned long long ml_phys_read_double_64(addr64_t paddr)
+unsigned long long ml_phys_read_double_64(addr64_t paddr64)
 {
-    return ml_phys_read_long_long(low32(paddr));
+        return ml_phys_read_long_long((pmap_paddr_t)paddr64);
 }
 
 
+
 /*
  *  Write data to a physical address. Memory should not be cache inhibited.
  */
 
 static void
-ml_phys_write_data( vm_offset_t paddr, unsigned long data, int size )
+ml_phys_write_data(pmap_paddr_t paddr, unsigned long data, int size)
 {
-    pt_entry_t save;
-    mp_disable_preemption();
-       if (*(pt_entry_t *) CM3)
+        mapwindow_t *map;
+       pt_entry_t save;
+
+       mp_disable_preemption();
+       map = pmap_get_mapwindow((pt_entry_t)(INTEL_PTE_VALID | INTEL_PTE_RW | (paddr & PG_FRAME) | 
+                                         INTEL_PTE_REF | INTEL_PTE_MOD));
+
+       if (map == 0) {
                panic("ml_phys_write_data: CMAP busy");
+       }
 
-       *(pt_entry_t *) CM3 = INTEL_PTE_VALID | INTEL_PTE_RW | (paddr & PG_FRAME) | 
-         INTEL_PTE_REF | INTEL_PTE_MOD;
-        save = *(pt_entry_t *)CM3;
-       invlpg((u_int)CA3);
+       save = *map->prv_CMAP;
+       invlpg((uintptr_t)map->prv_CADDR);
 
         switch (size) {
         case 1:
-            *(unsigned char *)((unsigned int)CA3 | (paddr & INTEL_OFFMASK)) = (unsigned char)data;
+           *(unsigned char *)((uintptr_t)map->prv_CADDR | ((uint32_t)paddr & INTEL_OFFMASK)) = (unsigned char)data;
             break;
         case 2:
-            *(unsigned short *)((unsigned int)CA3 | (paddr & INTEL_OFFMASK)) = (unsigned short)data;
+           *(unsigned short *)((uintptr_t)map->prv_CADDR | ((uint32_t)paddr & INTEL_OFFMASK)) = (unsigned short)data;
             break;
         case 4:
         default:
-            *(unsigned int *)((unsigned int)CA3 | (paddr & INTEL_OFFMASK)) = data;
+           *(unsigned int *)((uintptr_t)map->prv_CADDR | ((uint32_t)paddr & INTEL_OFFMASK)) = data;
             break;
         }
 
-       if (save != *(pt_entry_t *)CM3) panic("ml_phys_write_data CMAP changed");
-       *(pt_entry_t *) CM3 = 0;
+       if (save != *map->prv_CMAP)
+               panic("ml_phys_write_data CMAP changed");
+       *map->prv_CMAP = 0;
+
        mp_enable_preemption();
 }
 
 static void
-ml_phys_write_long_long( vm_offset_t paddr, unsigned long long data )
+ml_phys_write_long_long(pmap_paddr_t paddr, unsigned long long data)
 {
-    pt_entry_t save;
-    mp_disable_preemption();
-       if (*(pt_entry_t *) CM3)
+        mapwindow_t *map;
+       pt_entry_t save;
+
+       mp_disable_preemption();
+       map = pmap_get_mapwindow((pt_entry_t)(INTEL_PTE_VALID | INTEL_PTE_RW | (paddr & PG_FRAME) | 
+                                             INTEL_PTE_REF | INTEL_PTE_MOD));
+       if (map == 0) {
                panic("ml_phys_write_data: CMAP busy");
+       }
 
-       *(pt_entry_t *) CM3 = INTEL_PTE_VALID | INTEL_PTE_RW | (paddr & PG_FRAME) | 
-         INTEL_PTE_REF | INTEL_PTE_MOD;
-        save = *(pt_entry_t *)CM3;
-       invlpg((u_int)CA3);
+       save = *map->prv_CMAP;
+       invlpg((uintptr_t)map->prv_CADDR);
 
-        *(unsigned long long *)((unsigned int)CA3 | (paddr & INTEL_OFFMASK)) = data;
+       *(unsigned long long *)((uintptr_t)map->prv_CADDR | ((uint32_t)paddr & INTEL_OFFMASK)) = data;
 
-       if (save != *(pt_entry_t *)CM3) panic("ml_phys_write_data CMAP changed");
-       *(pt_entry_t *) CM3 = 0;
+       if (save != *map->prv_CMAP)
+               panic("ml_phys_write_data CMAP changed");
+       *map->prv_CMAP = 0;
        mp_enable_preemption();
 }
 
+
+
 void ml_phys_write_byte(vm_offset_t paddr, unsigned int data)
 {
-    ml_phys_write_data(paddr, data, 1);
+        ml_phys_write_data((pmap_paddr_t)paddr, data, 1);
 }
 
-void ml_phys_write_byte_64(addr64_t paddr, unsigned int data)
+void ml_phys_write_byte_64(addr64_t paddr64, unsigned int data)
 {
-    ml_phys_write_data(low32(paddr), data, 1);
+        ml_phys_write_data((pmap_paddr_t)paddr64, data, 1);
 }
 
 void ml_phys_write_half(vm_offset_t paddr, unsigned int data)
 {
-    ml_phys_write_data(paddr, data, 2);
+        ml_phys_write_data((pmap_paddr_t)paddr, data, 2);
 }
 
-void ml_phys_write_half_64(addr64_t paddr, unsigned int data)
+void ml_phys_write_half_64(addr64_t paddr64, unsigned int data)
 {
-    ml_phys_write_data(low32(paddr), data, 2);
+        ml_phys_write_data((pmap_paddr_t)paddr64, data, 2);
 }
 
 void ml_phys_write(vm_offset_t paddr, unsigned int data)
 {
-    ml_phys_write_data(paddr, data, 4);
+        ml_phys_write_data((pmap_paddr_t)paddr, data, 4);
 }
 
-void ml_phys_write_64(addr64_t paddr, unsigned int data)
+void ml_phys_write_64(addr64_t paddr64, unsigned int data)
 {
-    ml_phys_write_data(low32(paddr), data, 4);
+        ml_phys_write_data((pmap_paddr_t)paddr64, data, 4);
 }
 
 void ml_phys_write_word(vm_offset_t paddr, unsigned int data)
 {
-    ml_phys_write_data(paddr, data, 4);
+        ml_phys_write_data((pmap_paddr_t)paddr, data, 4);
 }
 
-void ml_phys_write_word_64(addr64_t paddr, unsigned int data)
+void ml_phys_write_word_64(addr64_t paddr64, unsigned int data)
 {
-    ml_phys_write_data(low32(paddr), data, 4);
+        ml_phys_write_data((pmap_paddr_t)paddr64, data, 4);
 }
 
-
 void ml_phys_write_double(vm_offset_t paddr, unsigned long long data)
 {
-    ml_phys_write_long_long(paddr, data);
+        ml_phys_write_long_long((pmap_paddr_t)paddr, data);
 }
 
-void ml_phys_write_double_64(addr64_t paddr, unsigned long long data)
+void ml_phys_write_double_64(addr64_t paddr64, unsigned long long data)
 {
-    ml_phys_write_long_long(low32(paddr), data);
+        ml_phys_write_long_long((pmap_paddr_t)paddr64, data);
 }
 
 
@@ -456,8 +453,9 @@ void ml_phys_write_double_64(addr64_t paddr, unsigned long long data)
 boolean_t
 ml_probe_read(vm_offset_t paddr, unsigned int *val)
 {
-    *val = ml_phys_read(paddr);
-    return TRUE;
+        *val = ml_phys_read((pmap_paddr_t)paddr);
+
+       return TRUE;
 }
 
 /*
@@ -467,10 +465,11 @@ ml_probe_read(vm_offset_t paddr, unsigned int *val)
  *  We assume that machine checks are enabled both in MSR and HIDs
  */
 boolean_t 
-ml_probe_read_64(addr64_t paddr, unsigned int *val)
+ml_probe_read_64(addr64_t paddr64, unsigned int *val)
 {
-    *val = ml_phys_read_64(paddr);
-    return TRUE;
+        *val = ml_phys_read_64((pmap_paddr_t)paddr64);
+
+       return TRUE;
 }
 
 
@@ -526,8 +525,6 @@ strlen(
        return string - 1 - ret;
 }
 
-#include <libkern/OSAtomic.h>
-
 uint32_t
 hw_atomic_add(
        uint32_t        *dest,
@@ -624,23 +621,26 @@ void machine_callstack(
 
 void fillPage(ppnum_t pa, unsigned int fill)
 {
-  pmap_paddr_t src;
-  int i;
-  int cnt = PAGE_SIZE/sizeof(unsigned int);
-  unsigned int *addr;
-  mp_disable_preemption();
-  if (*(pt_entry_t *) CM2)
+    mapwindow_t *map;
+    pmap_paddr_t src;
+    int i;
+    int cnt = PAGE_SIZE/sizeof(unsigned int);
+    unsigned int *addr;
+
+    mp_disable_preemption();
+    src = i386_ptob(pa);
+    map = pmap_get_mapwindow((pt_entry_t)(INTEL_PTE_VALID | INTEL_PTE_RW | (src & PG_FRAME) | 
+                                         INTEL_PTE_REF | INTEL_PTE_MOD));
+    if (map == 0) {
         panic("fillPage: CMAP busy");
-  src = (pmap_paddr_t)i386_ptob(pa);
-  *(pt_entry_t *) CM2 = INTEL_PTE_VALID | INTEL_PTE_RW | (src & PG_FRAME) | 
-      INTEL_PTE_REF | INTEL_PTE_MOD;
-      invlpg((u_int)CA2);
+    }
+    invlpg((uintptr_t)map->prv_CADDR);
 
-  for (i = 0, addr = (unsigned int *)CA2; i < cnt ; i++ )
-    *addr++ = fill;
+    for (i = 0, addr = (unsigned int *)map->prv_CADDR; i < cnt ; i++ )
+        *addr++ = fill;
 
-  *(pt_entry_t *) CM2 = 0;
-  mp_enable_preemption();
+    *map->prv_CMAP = 0;
+    mp_enable_preemption();
 }
 
 static inline void __sfence(void)
@@ -657,26 +657,30 @@ static inline void __wbinvd(void)
 }
 static inline void __clflush(void *ptr)
 {
-    __asm__ volatile(".byte 0x0F; .byte 0xae; .byte 0x38" : : "a" (ptr));
+       __asm__ volatile("clflush (%0)" : : "r" (ptr));
 }
 
 void dcache_incoherent_io_store64(addr64_t pa, unsigned int count)
 {
-    if (cpuid_features() & CPUID_FEATURE_CLFSH)
-    {
+        mapwindow_t *map;
         uint32_t  linesize = cpuid_info()->cache_linesize;
         addr64_t  addr;
         uint32_t  offset, chunk;
         boolean_t istate;
 
+       __mfence();
+
         istate = ml_set_interrupts_enabled(FALSE);
 
-        if (*(pt_entry_t *) CM2)
+        offset = pa & (linesize - 1);
+        addr   = pa - offset;
+
+        map = pmap_get_mapwindow((pt_entry_t)(i386_ptob(atop_64(addr)) | INTEL_PTE_VALID));
+        if (map == 0) {
                 panic("cache_flush_page_phys: CMAP busy");
+        }
 
-        offset = pa & (linesize - 1);
         count += offset;
-        addr   = pa - offset;
         offset = addr & ((addr64_t) (page_size - 1));
         chunk  = page_size - offset;
 
@@ -685,11 +689,11 @@ void dcache_incoherent_io_store64(addr64_t pa, unsigned int count)
             if (chunk > count)
                 chunk = count;
 
-            *(pt_entry_t *) CM2 = i386_ptob(atop_64(addr)) | INTEL_PTE_VALID;
-            invlpg((u_int)CA2);
+            *map->prv_CMAP = (pt_entry_t)(i386_ptob(atop_64(addr)) | INTEL_PTE_VALID);
+            invlpg((uintptr_t)map->prv_CADDR);
     
             for (; offset < chunk; offset += linesize)
-                __clflush((void *)(((u_int)CA2) + offset));
+                __clflush((void *)(((uintptr_t)map->prv_CADDR) + offset));
 
             count -= chunk;
             addr  += chunk;
@@ -698,13 +702,11 @@ void dcache_incoherent_io_store64(addr64_t pa, unsigned int count)
         }
         while (count);
 
-        *(pt_entry_t *) CM2 = 0;
+        *map->prv_CMAP = 0;
 
         (void) ml_set_interrupts_enabled(istate);
-    }
-    else
-        __wbinvd();
-    __sfence();
+
+       __mfence();
 }
 
 void dcache_incoherent_io_flush64(addr64_t pa, unsigned int count)
@@ -726,174 +728,586 @@ invalidate_icache64(__unused addr64_t addr,
 {
 }
 
-kern_return_t copypv(addr64_t          src64,
-                    addr64_t           snk64,
-                    unsigned int       size,
-                    int                which)
+
+addr64_t         vm_last_addr;
+
+void
+mapping_set_mod(ppnum_t pn)
 {
-       vm_map_t map;
-       kern_return_t ret;
-       vm_offset_t source, sink;
-       vm_offset_t vaddr;
-       vm_offset_t paddr;
-       spl_t s;
-       unsigned int lop, csize;
-       int needtran, bothphys;
-       vm_prot_t prot;
-       pt_entry_t *ptep;
-       
-       map = (which & cppvKmap) ? kernel_map : current_map_fast();
+  pmap_set_modify(pn);
+}
 
-       source = low32(src64);
-       sink = low32(snk64);
+void
+mapping_set_ref(ppnum_t pn)
+{
+  pmap_set_reference(pn);
+}
 
-       if((which & (cppvPsrc | cppvPsnk)) == 0 ) {             /* Make sure that only one is virtual */
-               panic("copypv: no more than 1 parameter may be virtual\n");     /* Not allowed */
+void
+cache_flush_page_phys(ppnum_t pa)
+{
+        mapwindow_t     *map;
+       boolean_t       istate;
+       int             i;
+       unsigned char   *cacheline_addr;
+       int             cacheline_size = cpuid_info()->cache_linesize;
+       int             cachelines_in_page = PAGE_SIZE/cacheline_size;
+
+       __mfence();
+
+       istate = ml_set_interrupts_enabled(FALSE);
+
+        map = pmap_get_mapwindow((pt_entry_t)(i386_ptob(pa) | INTEL_PTE_VALID));
+       if (map == 0) {
+               panic("cache_flush_page_phys: CMAP busy");
+        }
+
+       invlpg((uintptr_t)map->prv_CADDR);
+
+       for (i = 0, cacheline_addr = (unsigned char *)map->prv_CADDR;
+            i < cachelines_in_page;
+            i++, cacheline_addr += cacheline_size) {
+               __clflush((void *) cacheline_addr);
        }
+
+        *map->prv_CMAP = 0;
+
+       (void) ml_set_interrupts_enabled(istate);
+
+       __mfence();
+}
+
+
+void exit_funnel_section(void)
+{
+        thread_t thread;
+
+       thread = current_thread();
+
+        if (thread->funnel_lock)
+               (void) thread_funnel_set(thread->funnel_lock, FALSE);
+}
+
+
+
+/*
+ * the copy engine has the following characteristics
+ *   - copyio handles copies to/from user or kernel space
+ *   - copypv deals with physical or virtual addresses
+ *
+ * implementation details as follows
+ *   - a cache of up to NCOPY_WINDOWS is maintained per thread for
+ *     access of user virutal space
+ *   - the window size is determined by the amount of virtual space
+ *     that can be mapped by a single page table
+ *   - the mapping is done by copying the page table pointer from
+ *     the user's directory entry corresponding to the window's
+ *     address in user space to the directory entry corresponding
+ *     to the window slot in the kernel's address space
+ *   - the set of mappings is preserved across context switches,
+ *     so the copy can run with pre-emption enabled
+ *   - there is a gdt entry set up to anchor the kernel window on
+ *     each processor
+ *   - the copies are done using the selector corresponding to the
+ *     gdt entry
+ *   - the addresses corresponding to the user virtual address are
+ *     relative to the beginning of the window being used to map
+ *     that region... thus the thread can be pre-empted and switched
+ *     to a different processor while in the midst of a copy
+ *   - the window caches must be invalidated if the pmap changes out
+ *     from under the thread... this can happen during vfork/exec...
+ *     inval_copy_windows is the invalidation routine to be used
+ *   - the copyio engine has 4 different states associated with it
+ *     that allows for lazy tlb flushes and the ability to avoid
+ *     a flush all together if we've just come from user space
+ *     the 4 states are as follows...
+ *
+ *     WINDOWS_OPENED - set by copyio to indicate to the context
+ *       switch code that it is necessary to do a tlbflush after
+ *       switching the windows since we're in the middle of a copy
+ *
+ *     WINDOWS_CLOSED - set by copyio to indicate that it's done
+ *       using the windows, so that the context switch code need
+ *       not do the tlbflush... instead it will set the state to...
+ *
+ *     WINDOWS_DIRTY - set by the context switch code to indicate
+ *       to the copy engine that it is responsible for doing a 
+ *       tlbflush before using the windows again... it's also
+ *       set by the inval_copy_windows routine to indicate the
+ *       same responsibility.
+ *
+ *     WINDOWS_CLEAN - set by the return to user path to indicate
+ *       that a tlbflush has happened and that there is no need
+ *       for copyio to do another when it is entered next...
+ *
+ *   - a window for mapping single physical pages is provided for copypv
+ *   - this window is maintained across context switches and has the
+ *     same characteristics as the user space windows w/r to pre-emption
+ */
+
+extern int copyout_user(const char *, vm_offset_t, vm_size_t);
+extern int copyout_kern(const char *, vm_offset_t, vm_size_t);
+extern int copyin_user(const vm_offset_t, char *, vm_size_t);
+extern int copyin_kern(const vm_offset_t, char *, vm_size_t);
+extern int copyoutphys_user(const char *, vm_offset_t, vm_size_t);
+extern int copyoutphys_kern(const char *, vm_offset_t, vm_size_t);
+extern int copyinphys_user(const vm_offset_t, char *, vm_size_t);
+extern int copyinphys_kern(const vm_offset_t, char *, vm_size_t);
+extern int copyinstr_user(const vm_offset_t, char *, vm_size_t, vm_size_t *);
+extern int copyinstr_kern(const vm_offset_t, char *, vm_size_t, vm_size_t *);
+
+static int copyio(int, user_addr_t, char *, vm_size_t, vm_size_t *, int);
+static int copyio_phys(addr64_t, addr64_t, vm_size_t, int);
+
+
+#define COPYIN         0
+#define COPYOUT                1
+#define COPYINSTR      2
+#define COPYINPHYS     3
+#define COPYOUTPHYS    4
+
+
+
+void inval_copy_windows(thread_t thread)
+{
+        int    i;
        
-       bothphys = 1;                                                                   /* Assume both are physical */
-       
-       if(!(which & cppvPsnk)) {                                               /* Is there a virtual page here? */
-               vaddr = sink;                                                           /* Sink side is virtual */
-               bothphys = 0;                                                           /* Show both aren't physical */
-               prot = VM_PROT_READ | VM_PROT_WRITE;            /* Sink always must be read/write */
-       } else /* if(!(which & cppvPsrc)) */ {                          /* Source side is virtual */
-               vaddr = source;                                                         /* Source side is virtual */
-               bothphys = 0;                                                           /* Show both aren't physical */
-               prot = VM_PROT_READ;                                            /* Virtual source is always read only */
+       for (i = 0; i < NCOPY_WINDOWS; i++) {
+                thread->machine.copy_window[i].user_base = -1;
        }
+       thread->machine.nxt_window = 0;
+       thread->machine.copyio_state = WINDOWS_DIRTY;
 
-       needtran = 1;                                                                   /* Show we need to map the virtual the first time */
-       s = splhigh();                                                                  /* Don't bother me */
-
-       while(size) {
-
-               if(!bothphys && (needtran || !(vaddr & 4095LL))) {      /* If first time or we stepped onto a new page, we need to translate */
-                       needtran = 0;
-                       while(1) {
-                         ptep = pmap_mapgetpte(map, vaddr);
-                         if((0 == ptep) || ((*ptep & INTEL_PTE_VALID) == 0)) {
-                                       splx(s);                                                /* Restore the interrupt level */
-                                       ret = vm_fault(map, vm_map_trunc_page(vaddr), prot, FALSE, THREAD_UNINT, NULL, 0);      /* Didn't find it, try to fault it in... */
-                               
-                                       if(ret != KERN_SUCCESS)return KERN_FAILURE;     /* Didn't find any, return no good... */
-                                       
-                                       s = splhigh();                                  /* Don't bother me */
-                                       continue;                                               /* Go try for the map again... */
-       
-                               }
-               
-                               /* Note that we have to have the destination writable.  So, if we already have it, or we are mapping the source,
-                                       we can just leave.
-                               */              
-                               if((which & cppvPsnk) || (*ptep & INTEL_PTE_WRITE)) break;              /* We got it mapped R/W or the source is not virtual, leave... */
-                               splx(s);                                                        /* Restore the interrupt level */
-                               
-                               ret = vm_fault(map, vm_map_trunc_page(vaddr), VM_PROT_READ | VM_PROT_WRITE, FALSE, THREAD_UNINT, NULL, 0);      /* check for a COW area */
-                               if (ret != KERN_SUCCESS) return KERN_FAILURE;   /* We couldn't get it R/W, leave in disgrace... */
-                               s = splhigh();                                          /* Don't bother me */
-                       }
+       KERNEL_DEBUG(0xeff70058 | DBG_FUNC_NONE, (int)thread, (int)thread->map, 0, 0, 0);
+}
 
-                       paddr = pte_to_pa(*ptep) | (vaddr & 4095);
-       
-                       if(which & cppvPsrc) sink = paddr;              /* If source is physical, then the sink is virtual */
-                       else source = paddr;                                    /* Otherwise the source is */
+
+static int
+copyio(int copy_type, user_addr_t user_addr, char *kernel_addr, vm_size_t nbytes, vm_size_t *lencopied, int use_kernel_map)
+{
+        thread_t       thread;
+       pmap_t          pmap;
+       pt_entry_t      *updp;
+       pt_entry_t      *kpdp;
+       user_addr_t     user_base;
+       vm_offset_t     user_offset;
+       vm_offset_t     kern_vaddr;
+       vm_size_t       cnt;
+       vm_size_t       bytes_copied;
+       int             error = 0;
+       int             window_index;
+       int             copyio_state;
+        boolean_t      istate;
+#if KDEBUG
+       int             debug_type = 0xeff70010;
+       debug_type += (copy_type << 2);
+#endif
+
+       thread = current_thread();
+
+       KERNEL_DEBUG(debug_type | DBG_FUNC_START, (int)(user_addr >> 32), (int)user_addr, (int)nbytes, thread->machine.copyio_state, 0);
+
+       if (nbytes == 0) {
+               KERNEL_DEBUG(debug_type | DBG_FUNC_END, (int)user_addr, (int)kernel_addr, (int)nbytes, 0, 0);
+               return (0);
+       }
+        pmap = thread->map->pmap;
+
+        if (pmap == kernel_pmap || use_kernel_map) {
+
+               kern_vaddr = (vm_offset_t)user_addr;
+         
+               switch (copy_type) {
+
+               case COPYIN:
+                       error = copyin_kern(kern_vaddr, kernel_addr, nbytes);
+                       break;
+
+               case COPYOUT:
+                       error = copyout_kern(kernel_addr, kern_vaddr, nbytes);
+                       break;
+
+               case COPYINSTR:
+                       error = copyinstr_kern(kern_vaddr, kernel_addr, nbytes, lencopied);
+                       break;
+
+               case COPYINPHYS:
+                       error = copyinphys_kern(kern_vaddr, kernel_addr, nbytes);
+                       break;
+
+               case COPYOUTPHYS:
+                       error = copyoutphys_kern(kernel_addr, kern_vaddr, nbytes);
+                       break;
                }
-                       
-               lop = (unsigned int)(4096LL - (sink & 4095LL));         /* Assume sink smallest */
-               if(lop > (unsigned int)(4096LL - (source & 4095LL))) lop = (unsigned int)(4096LL - (source & 4095LL));  /* No, source is smaller */
-               
-               csize = size;                                                           /* Assume we can copy it all */
-               if(lop < size) csize = lop;                                     /* Nope, we can't do it all */
-               
-               if(which & cppvFsrc) flush_dcache64((addr64_t)source, csize, 1);        /* If requested, flush source before move */
-               if(which & cppvFsnk) flush_dcache64((addr64_t)sink, csize, 1);  /* If requested, flush sink before move */
+               KERNEL_DEBUG(debug_type | DBG_FUNC_END, (int)kern_vaddr, (int)kernel_addr, (int)nbytes, error | 0x80000000, 0);
 
-               bcopy_phys((addr64_t)source, (addr64_t)sink, csize);                    /* Do a physical copy, virtually */
-               
-               if(which & cppvFsrc) flush_dcache64((addr64_t)source, csize, 1);        /* If requested, flush source after move */
-               if(which & cppvFsnk) flush_dcache64((addr64_t)sink, csize, 1);  /* If requested, flush sink after move */
+               return (error);
+       }
+       user_base = user_addr & ~((user_addr_t)(NBPDE - 1));
+       user_offset = user_addr & (NBPDE - 1);
 
+       KERNEL_DEBUG(debug_type | DBG_FUNC_NONE, (int)(user_base >> 32), (int)user_base, (int)user_offset, 0, 0);
 
-/*
- *             Note that for certain ram disk flavors, we may be copying outside of known memory.
- *             Therefore, before we try to mark it modifed, we check if it exists.
- */
+       cnt = NBPDE - user_offset;
+
+       if (cnt > nbytes)
+               cnt = nbytes;
+
+       istate = ml_set_interrupts_enabled(FALSE);
+
+       copyio_state = thread->machine.copyio_state;
+       thread->machine.copyio_state = WINDOWS_OPENED;
+
+       (void) ml_set_interrupts_enabled(istate);
+
+
+       for (;;) {
 
-               if( !(which & cppvNoModSnk)) {
-                 if (phys_page_exists((ppnum_t)sink >> 12))
-                       mapping_set_mod((ppnum_t)(sink >> 12));         /* Make sure we know that it is modified */
+               for (window_index = 0; window_index < NCOPY_WINDOWS; window_index++) {
+                       if (thread->machine.copy_window[window_index].user_base == user_base)
+                               break;
                }
-               if( !(which & cppvNoRefSrc)) {
-                 if (phys_page_exists((ppnum_t)source >> 12))
-                       mapping_set_ref((ppnum_t)(source >> 12));               /* Make sure we know that it is modified */
+               if (window_index >= NCOPY_WINDOWS) {
+
+                       window_index = thread->machine.nxt_window;
+                       thread->machine.nxt_window++;
+
+                       if (thread->machine.nxt_window >= NCOPY_WINDOWS)
+                               thread->machine.nxt_window = 0;
+                       thread->machine.copy_window[window_index].user_base = user_base;
+
+                       /*
+                        * it's necessary to disable pre-emption
+                        * since I have to compute the kernel descriptor pointer
+                        * for the new window
+                        */
+                       istate = ml_set_interrupts_enabled(FALSE);
+
+                       updp = pmap_pde(pmap, user_base);
+
+                       kpdp = current_cpu_datap()->cpu_copywindow_pdp;
+                       kpdp += window_index;
+
+                       pmap_store_pte(kpdp, updp ? *updp : 0);
+
+                       (void) ml_set_interrupts_enabled(istate);
+
+                       copyio_state = WINDOWS_DIRTY;
+
+                       KERNEL_DEBUG(0xeff70040 | DBG_FUNC_NONE, window_index, (int)user_base, (int)updp, (int)kpdp, 0);
+
                }
+#if JOE_DEBUG
+               else {
+                       updp = pmap_pde(pmap, user_base);
 
+                       istate = ml_set_interrupts_enabled(FALSE);
 
-               size = size - csize;                                            /* Calculate what is left */
-               vaddr = vaddr + csize;                                          /* Move to next sink address */
-               source = source + csize;                                        /* Bump source to next physical address */
-               sink = sink + csize;                                            /* Bump sink to next physical address */
+                       kpdp = current_cpu_datap()->cpu_copywindow_pdp;
+
+                       kpdp += window_index;
+
+                       if ((*kpdp & PG_FRAME) != (*updp & PG_FRAME)) {
+                               panic("copyio: user pdp mismatch - kpdp = 0x%x,  updp = 0x%x\n", kpdp, updp);
+                       }
+                       (void) ml_set_interrupts_enabled(istate);
+               }
+#endif
+               if (copyio_state == WINDOWS_DIRTY) {
+                       flush_tlb();
+
+                       copyio_state = WINDOWS_CLEAN;
+
+                       KERNEL_DEBUG(0xeff70054 | DBG_FUNC_NONE, window_index, 0, 0, 0, 0);
+               }
+               user_offset += (window_index * NBPDE);
+
+               KERNEL_DEBUG(0xeff70044 | DBG_FUNC_NONE, (int)user_offset, (int)kernel_addr, cnt, 0, 0);
+
+               switch (copy_type) {
+
+               case COPYIN:
+                       error = copyin_user(user_offset, kernel_addr, cnt);
+                       break;
+                       
+               case COPYOUT:
+                       error = copyout_user(kernel_addr, user_offset, cnt);
+                       break;
+
+               case COPYINPHYS:
+                       error = copyinphys_user(user_offset, kernel_addr, cnt);
+                       break;
+                       
+               case COPYOUTPHYS:
+                       error = copyoutphys_user(kernel_addr, user_offset, cnt);
+                       break;
+
+               case COPYINSTR:
+                       error = copyinstr_user(user_offset, kernel_addr, cnt, &bytes_copied);
+
+                       /*
+                        * lencopied should be updated on success
+                        * or ENAMETOOLONG...  but not EFAULT
+                        */
+                       if (error != EFAULT)
+                               *lencopied += bytes_copied;
+
+                       /*
+                        * if we still have room, then the ENAMETOOLONG
+                        * is just an artifact of the buffer straddling
+                        * a window boundary and we should continue
+                        */
+                       if (error == ENAMETOOLONG && nbytes > cnt)
+                               error = 0;
+
+                       if (error) {
+#if KDEBUG
+                               nbytes = *lencopied;
+#endif
+                               break;
+                       }
+                       if (*(kernel_addr + bytes_copied - 1) == 0) {
+                               /*
+                                * we found a NULL terminator... we're done
+                                */
+#if KDEBUG
+                               nbytes = *lencopied;
+#endif
+                               goto done;
+                       }
+                       if (cnt == nbytes) {
+                               /*
+                                * no more room in the buffer and we haven't
+                                * yet come across a NULL terminator
+                                */
+#if KDEBUG
+                               nbytes = *lencopied;
+#endif
+                               error = ENAMETOOLONG;
+                               break;
+                       }
+                       assert(cnt == bytes_copied);
+
+                       break;
+               }
+               if (error)
+                       break;
+               if ((nbytes -= cnt) == 0)
+                       break;
+
+               kernel_addr += cnt;
+               user_base += NBPDE;
+               user_offset = 0;
+
+               if (nbytes > NBPDE)
+                       cnt = NBPDE;
+               else
+                       cnt = nbytes;
        }
-       
-       splx(s);                                                                                /* Open up for interrupts */
+done:
+       thread->machine.copyio_state = WINDOWS_CLOSED;
 
-       return KERN_SUCCESS;
+       KERNEL_DEBUG(debug_type | DBG_FUNC_END, (int)user_addr, (int)kernel_addr, (int)nbytes, error, 0);
+
+       return (error);
 }
 
-void switch_to_serial_console(void)
+
+static int
+copyio_phys(addr64_t source, addr64_t sink, vm_size_t csize, int which)
 {
+        pmap_paddr_t paddr;
+       user_addr_t  vaddr;
+       char        *window_offset;
+       pt_entry_t  pentry;
+       int         ctype;
+       int         retval;
+       boolean_t   istate;
+
+       if (which & cppvPsnk) {
+               paddr  = (pmap_paddr_t)sink;
+               vaddr  = (user_addr_t)source;
+               ctype  = COPYINPHYS;
+               pentry = (pt_entry_t)(INTEL_PTE_VALID | (paddr & PG_FRAME) | INTEL_PTE_RW);
+       } else {
+               paddr  = (pmap_paddr_t)source;
+               vaddr  = (user_addr_t)sink;
+               ctype  = COPYOUTPHYS;
+               pentry = (pt_entry_t)(INTEL_PTE_VALID | (paddr & PG_FRAME));
+       }
+       window_offset = (char *)((uint32_t)paddr & (PAGE_SIZE - 1));
+
+       if (current_thread()->machine.physwindow_busy) {
+               pt_entry_t      old_pentry;
+
+               KERNEL_DEBUG(0xeff70048 | DBG_FUNC_NONE, paddr, csize, 0, -1, 0);
+               /*
+                * we had better be targeting wired memory at this point
+                * we will not be able to handle a fault with interrupts
+                * disabled... we disable them because we can't tolerate
+                * being preempted during this nested use of the window
+                */
+               istate = ml_set_interrupts_enabled(FALSE);
+
+               old_pentry = *(current_cpu_datap()->cpu_physwindow_ptep);
+               pmap_store_pte((current_cpu_datap()->cpu_physwindow_ptep), pentry);
+
+               invlpg((uintptr_t)current_cpu_datap()->cpu_physwindow_base);
+
+               retval = copyio(ctype, vaddr, window_offset, csize, NULL, which & cppvKmap);
+
+               pmap_store_pte((current_cpu_datap()->cpu_physwindow_ptep), old_pentry);
+
+               invlpg((uintptr_t)current_cpu_datap()->cpu_physwindow_base);
+
+               (void) ml_set_interrupts_enabled(istate);
+       } else {
+               /*
+                * mark the window as in use... if an interrupt hits while we're
+                * busy, or we trigger another coyppv from the fault path into
+                * the driver on a user address space page fault due to a copyin/out
+                * then we need to save and restore the current window state instead
+                * of caching the window preserving it across context switches
+                */
+               current_thread()->machine.physwindow_busy = 1;
+
+               if (current_thread()->machine.physwindow_pte != pentry) {
+                       KERNEL_DEBUG(0xeff70048 | DBG_FUNC_NONE, paddr, csize, 0, 0, 0);
+
+                       current_thread()->machine.physwindow_pte = pentry;
+                       
+                       /*
+                        * preemption at this point would be bad since we
+                        * could end up on the other processor after we grabbed the
+                        * pointer to the current cpu data area, but before we finished
+                        * using it to stuff the page table entry since we would
+                        * be modifying a window that no longer belonged to us
+                        * the invlpg can be done unprotected since it only flushes
+                        * this page address from the tlb... if it flushes the wrong
+                        * one, no harm is done, and the context switch that moved us
+                        * to the other processor will have already take care of 
+                        * flushing the tlb after it reloaded the page table from machine.physwindow_pte
+                        */
+                       istate = ml_set_interrupts_enabled(FALSE);
+                       *(current_cpu_datap()->cpu_physwindow_ptep) = pentry;
+                       (void) ml_set_interrupts_enabled(istate);
+
+                       invlpg((uintptr_t)current_cpu_datap()->cpu_physwindow_base);
+               }
+#if JOE_DEBUG
+               else {
+                       if (pentry !=
+                           (*(current_cpu_datap()->cpu_physwindow_ptep) & (INTEL_PTE_VALID | PG_FRAME | INTEL_PTE_RW)))
+                               panic("copyio_phys: pentry != *physwindow_ptep");
+               }
+#endif
+               retval = copyio(ctype, vaddr, window_offset, csize, NULL, which & cppvKmap);
+
+               current_thread()->machine.physwindow_busy = 0;
+       }
+       return (retval);
 }
 
-addr64_t         vm_last_addr;
 
-void
-mapping_set_mod(ppnum_t pn)
+
+int
+copyinmsg(const user_addr_t user_addr, char *kernel_addr, vm_size_t nbytes)
 {
-  pmap_set_modify(pn);
-}
+        return (copyio(COPYIN, user_addr, kernel_addr, nbytes, NULL, 0));
+}    
 
-void
-mapping_set_ref(ppnum_t pn)
+int
+copyin(const user_addr_t user_addr, char *kernel_addr, vm_size_t nbytes)
 {
-  pmap_set_reference(pn);
+        return (copyio(COPYIN, user_addr, kernel_addr, nbytes, NULL, 0));
 }
 
-void
-cache_flush_page_phys(ppnum_t pa)
+int
+copyinstr(const user_addr_t user_addr,  char *kernel_addr, vm_size_t nbytes, vm_size_t *lencopied)
 {
-       boolean_t       istate;
-       int             i;
-       unsigned int    *cacheline_addr;
-       int             cacheline_size = cpuid_info()->cache_linesize;
-       int             cachelines_in_page = PAGE_SIZE/cacheline_size;
+       *lencopied = 0;
 
-       /*
-        * If there's no clflush instruction, we're sadly forced to use wbinvd.
-        */
-       if (!(cpuid_features() & CPUID_FEATURE_CLFSH)) {
-               asm volatile("wbinvd" : : : "memory");
-               return;
-       } 
+        return (copyio(COPYINSTR, user_addr, kernel_addr, nbytes, lencopied, 0));
+}
 
-       istate = ml_set_interrupts_enabled(FALSE);
+int
+copyoutmsg(const char *kernel_addr, user_addr_t user_addr, vm_size_t nbytes)
+{
+        return (copyio(COPYOUT, user_addr, (char *)kernel_addr, nbytes, NULL, 0));
+}
 
-       if (*(pt_entry_t *) CM2)
-               panic("cache_flush_page_phys: CMAP busy");
+int
+copyout(const char *kernel_addr, user_addr_t user_addr, vm_size_t nbytes)
+{
+        return (copyio(COPYOUT, user_addr, (char *)kernel_addr, nbytes, NULL, 0));
+}
 
-       *(pt_entry_t *) CM2 = i386_ptob(pa) | INTEL_PTE_VALID;
-       invlpg((u_int)CA2);
 
-       for (i = 0, cacheline_addr = (unsigned int *)CA2;
-            i < cachelines_in_page;
-            i++, cacheline_addr += cacheline_size) {
-               asm volatile("clflush %0" : : "m" (cacheline_addr));
-       }
+kern_return_t copypv(addr64_t src64, addr64_t snk64, unsigned int size, int which)
+{
+       unsigned int lop, csize;
+       int bothphys = 0;
+       
 
-       *(pt_entry_t *) CM2 = 0;
+       KERNEL_DEBUG(0xeff7004c | DBG_FUNC_START, (int)src64, (int)snk64, size, which, 0);
 
-       (void) ml_set_interrupts_enabled(istate);
+       if ((which & (cppvPsrc | cppvPsnk)) == 0 )                              /* Make sure that only one is virtual */
+               panic("copypv: no more than 1 parameter may be virtual\n");     /* Not allowed */
 
-}
+       if ((which & (cppvPsrc | cppvPsnk)) == (cppvPsrc | cppvPsnk))
+               bothphys = 1;                                                   /* both are physical */
+
+       while (size) {
+         
+               if (bothphys) {
+                       lop = (unsigned int)(PAGE_SIZE - (snk64 & (PAGE_SIZE - 1)));            /* Assume sink smallest */
+
+                       if (lop > (unsigned int)(PAGE_SIZE - (src64 & (PAGE_SIZE - 1))))
+                               lop = (unsigned int)(PAGE_SIZE - (src64 & (PAGE_SIZE - 1)));    /* No, source is smaller */
+               } else {
+                       /*
+                        * only need to compute the resid for the physical page
+                        * address... we don't care about where we start/finish in
+                        * the virtual since we just call the normal copyin/copyout
+                        */
+                       if (which & cppvPsrc)
+                               lop = (unsigned int)(PAGE_SIZE - (src64 & (PAGE_SIZE - 1)));
+                       else
+                               lop = (unsigned int)(PAGE_SIZE - (snk64 & (PAGE_SIZE - 1)));
+               }
+               csize = size;                                           /* Assume we can copy it all */
+               if (lop < size)
+                       csize = lop;                                    /* Nope, we can't do it all */
+#if 0          
+               /*
+                * flush_dcache64 is currently a nop on the i386... 
+                * it's used when copying to non-system memory such
+                * as video capture cards... on PPC there was a need
+                * to flush due to how we mapped this memory... not
+                * sure if it's needed on i386.
+                */
+               if (which & cppvFsrc)
+                       flush_dcache64(src64, csize, 1);                /* If requested, flush source before move */
+               if (which & cppvFsnk)
+                       flush_dcache64(snk64, csize, 1);                /* If requested, flush sink before move */
+#endif
+               if (bothphys)
+                       bcopy_phys(src64, snk64, csize);                /* Do a physical copy, virtually */
+               else {
+                       if (copyio_phys(src64, snk64, csize, which))
+                               return (KERN_FAILURE);
+               }
+#if 0
+               if (which & cppvFsrc)
+                       flush_dcache64(src64, csize, 1);        /* If requested, flush source after move */
+               if (which & cppvFsnk)
+                       flush_dcache64(snk64, csize, 1);        /* If requested, flush sink after move */
+#endif
+               size  -= csize;                                 /* Calculate what is left */
+               snk64 += csize;                                 /* Bump sink to next physical address */
+               src64 += csize;                                 /* Bump source to next physical address */
+       }
+       KERNEL_DEBUG(0xeff7004c | DBG_FUNC_END, (int)src64, (int)snk64, size, which, 0);
 
+       return KERN_SUCCESS;
+}