enami
parent
551a5b2934
commit
2903a6278e
|
|
@ -1,4 +1,4 @@
|
|||
.\" $NetBSD: uvm.9,v 1.11 2000/04/05 03:21:31 enami Exp $
|
||||
.\" $NetBSD: uvm.9,v 1.12 2000/04/05 03:29:06 enami Exp $
|
||||
.\"
|
||||
.\" Copyright (c) 1998 Matthew R. Green
|
||||
.\" All rights reserved.
|
||||
|
|
@ -69,10 +69,7 @@ swapper process swaps in runnable processes that are currently swapped out,
|
|||
if there is room.
|
||||
.Pp
|
||||
There are also several miscellaneous functions.
|
||||
|
||||
.Pp
|
||||
.Sh INITIALISATION
|
||||
|
||||
.nr nS 1
|
||||
.Pp
|
||||
.Ft void
|
||||
|
|
@ -84,7 +81,6 @@ There are also several miscellaneous functions.
|
|||
.Ft void
|
||||
.Fn uvm_swap_init
|
||||
.nr nS 0
|
||||
|
||||
.Pp
|
||||
.Fn uvm_init
|
||||
sets up the UVM system at system boot time, after the
|
||||
|
|
@ -93,13 +89,11 @@ global state, the page, map, kernel virtual memory state,
|
|||
machine-dependent physical map, kernel memory allocator,
|
||||
pager and anonymous memory sub-systems, and then enables
|
||||
paging of kernel objects.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_init_limits
|
||||
initialises process limits for the named process. This is for use by
|
||||
the system startup for process zero, before any other processes are
|
||||
created.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_setpagesize
|
||||
initialises the uvmexp members pagesize (if not already done by
|
||||
|
|
@ -107,14 +101,10 @@ machine-dependent code), pageshift and pagemask. It should be called by
|
|||
machine-dependent code early in the
|
||||
.Xr pmap_init 9
|
||||
call.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_swap_init
|
||||
initialises the swap sub-system.
|
||||
|
||||
.Pp
|
||||
.Sh VIRTUAL ADDRESS SPACE MANAGEMENT
|
||||
|
||||
.Pp
|
||||
.nr nS 1
|
||||
.Ft int
|
||||
|
|
@ -141,7 +131,6 @@ initialises the swap sub-system.
|
|||
.Ft void
|
||||
.Fn uvmspace_unshare "struct proc *p"
|
||||
.nr nS 0
|
||||
|
||||
.Pp
|
||||
.Fn uvm_map
|
||||
establishes a valid mapping in map
|
||||
|
|
@ -262,7 +251,6 @@ macros return the protection, inheritance, maximum protection and advice,
|
|||
respectively.
|
||||
.Fn uvm_map
|
||||
returns a standard UVM return value.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_map_pageable
|
||||
changes the pageability of the pages in the range from
|
||||
|
|
@ -275,7 +263,6 @@ to
|
|||
.Fa new_pageable .
|
||||
.Fn uvm_map_pageable
|
||||
returns a standard UVM return value.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_map_checkprot
|
||||
checks the protection of the range from
|
||||
|
|
@ -290,7 +277,6 @@ This returns either
|
|||
.Dv TRUE
|
||||
or
|
||||
.Dv FALSE .
|
||||
|
||||
.Pp
|
||||
.Fn uvm_map_protect
|
||||
changes the protection
|
||||
|
|
@ -306,7 +292,6 @@ also setting the maximum protection to the region to
|
|||
if
|
||||
.Fa set_max
|
||||
is non-zero. This function returns a standard UVM return value.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_deallocate
|
||||
deallocates kernel memory in map
|
||||
|
|
@ -315,7 +300,6 @@ from address
|
|||
.Fa start
|
||||
to
|
||||
.Fa start + size .
|
||||
|
||||
.Pp
|
||||
.Fn uvmspace_alloc
|
||||
allocates and returns a new address space, with ranges from
|
||||
|
|
@ -324,7 +308,6 @@ to
|
|||
.Fa max ,
|
||||
setting the pageability of the address space to
|
||||
.Fa pageable.
|
||||
|
||||
.Pp
|
||||
.Fn uvmspace_exec
|
||||
either reuses the address space of process
|
||||
|
|
@ -332,27 +315,23 @@ either reuses the address space of process
|
|||
if there are no other references to it, or creates
|
||||
a new one with
|
||||
.Fn uvmspace_alloc .
|
||||
|
||||
.Pp
|
||||
.Fn uvmspace_fork
|
||||
creates and returns a new address space based upon the
|
||||
.Fa vm1
|
||||
address space, typically used when allocating an address space for a
|
||||
child process.
|
||||
|
||||
.Pp
|
||||
.Fn uvmspace_free
|
||||
lowers the reference count on the address space
|
||||
.Fa vm ,
|
||||
freeing the data structures if there are no other references.
|
||||
|
||||
.Pp
|
||||
.Fn uvmspace_share
|
||||
causes process
|
||||
.Pa p2
|
||||
to share the address space of
|
||||
.Fa p1 .
|
||||
|
||||
.Pp
|
||||
.Fn uvmspace_unshare
|
||||
ensures that process
|
||||
|
|
@ -360,10 +339,7 @@ ensures that process
|
|||
has its own, unshared address space, by creating a new one if
|
||||
necessary by calling
|
||||
.Fn uvmspace_fork .
|
||||
|
||||
.Pp
|
||||
.Sh PAGE FAULT HANDLING
|
||||
|
||||
.Pp
|
||||
.nr nS 1
|
||||
.Ft int
|
||||
|
|
@ -382,10 +358,7 @@ describing the type of fault, and
|
|||
describing the type of access requested.
|
||||
.Fn uvm_fault
|
||||
returns a standard UVM return value.
|
||||
|
||||
.Pp
|
||||
.Sh MEMORY MAPPING FILES AND DEVICES
|
||||
|
||||
.Pp
|
||||
.nr nS 1
|
||||
.Ft struct uvm_object *
|
||||
|
|
@ -399,13 +372,11 @@ returns a standard UVM return value.
|
|||
.Ft boolean_t
|
||||
.Fn uvm_vnp_uncache "struct vnode *vp"
|
||||
.nr nS 0
|
||||
|
||||
.Pp
|
||||
.Fn uvn_attach
|
||||
attaches a UVM object to vnode
|
||||
.Fa arg ,
|
||||
creating the object if necessary. The object is returned.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_vnp_setsize
|
||||
sets the size of vnode
|
||||
|
|
@ -415,7 +386,6 @@ to
|
|||
Caller must hold a reference to the vnode. If the vnode shrinks, pages
|
||||
no longer used are discarded. This function will be removed when the
|
||||
filesystem and VM buffer caches are merged.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_vnp_sync
|
||||
flushes dirty vnodes from either the mount point passed in
|
||||
|
|
@ -426,7 +396,6 @@ is
|
|||
.Dv NULL .
|
||||
This function will be removed when the filesystem and VM buffer caches
|
||||
are merged.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_vnp_terminate
|
||||
frees all VM resources allocated to vnode
|
||||
|
|
@ -434,7 +403,6 @@ frees all VM resources allocated to vnode
|
|||
If the vnode still has references, it will not be destroyed; however
|
||||
all future operations using this vnode will fail. This function will be
|
||||
removed when the filesystem and VM buffer caches are merged.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_vnp_uncache
|
||||
disables vnode
|
||||
|
|
@ -442,26 +410,19 @@ disables vnode
|
|||
from persisting when all references are freed. This function will be
|
||||
removed when the file-system and UVM caches are unified. Returns
|
||||
true if there is no active vnode.
|
||||
|
||||
.Pp
|
||||
.Sh VIRTUAL MEMORY I/O
|
||||
|
||||
.Pp
|
||||
.nr nS 1
|
||||
.Ft int
|
||||
.Fn uvm_io "vm_map_t map" "struct uio *uio"
|
||||
.nr nS 0
|
||||
|
||||
.Pp
|
||||
.Fn uvm_io
|
||||
performs the I/O described in
|
||||
.Fa uio
|
||||
on the memory described in
|
||||
.Fa map .
|
||||
|
||||
.Pp
|
||||
.Sh ALLOCATION OF KERNEL MEMORY
|
||||
|
||||
.Pp
|
||||
.nr nS 1
|
||||
.Ft vaddr_t
|
||||
|
|
@ -483,7 +444,6 @@ on the memory described in
|
|||
.Ft void
|
||||
.Fn uvm_km_free_wakeup "vm_map_t map" "vaddr_t addr" "vsize_t size"
|
||||
.nr nS 0
|
||||
|
||||
.Pp
|
||||
.Fn uvm_km_alloc
|
||||
and
|
||||
|
|
@ -499,7 +459,6 @@ terms of
|
|||
.Fn uvm_km_alloc1 ,
|
||||
and should almost always be used in preference to
|
||||
.Fn uvm_km_alloc1 .
|
||||
|
||||
.Pp
|
||||
.Fn uvm_km_alloc1
|
||||
allocates and returns
|
||||
|
|
@ -507,7 +466,6 @@ allocates and returns
|
|||
bytes of wired memory in the kernel map, zeroing the memory if the
|
||||
.Fa zeroit
|
||||
argument is non-zero.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_km_kmemalloc
|
||||
allocates and returns
|
||||
|
|
@ -533,7 +491,6 @@ causes
|
|||
to use
|
||||
.Fn simple_lock_try
|
||||
when locking maps.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_km_valloc
|
||||
and
|
||||
|
|
@ -543,10 +500,7 @@ return a newly allocated zero-filled address in the kernel map of size
|
|||
.Fn uvm_km_valloc_wait
|
||||
will also wait for kernel memory to become available, if there is a
|
||||
memory shortage.
|
||||
|
||||
.Pp
|
||||
.Sh ALLOCATION OF PHYSICAL MEMORY
|
||||
|
||||
.Pp
|
||||
.nr nS 1
|
||||
.Ft struct vm_page *
|
||||
|
|
@ -562,7 +516,6 @@ memory shortage.
|
|||
.Ft void
|
||||
.Fn uvm_page_physload "vaddr_t start" "vaddr_t end" "vaddr_t avail_start" "vaddr_t avail_end"
|
||||
.nr nS 0
|
||||
|
||||
.Pp
|
||||
.Fn uvm_pagealloc
|
||||
allocates a page of memory at virtual address
|
||||
|
|
@ -580,7 +533,6 @@ can be non
|
|||
Returns
|
||||
.Dv NULL
|
||||
when no page can be found.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_pagerealloc
|
||||
reallocates page
|
||||
|
|
@ -589,12 +541,10 @@ to a new object
|
|||
.Fa newobj ,
|
||||
at a new offset
|
||||
.Fa newoff .
|
||||
|
||||
.Pp
|
||||
.Fn uvm_pagefree
|
||||
free's the physical page
|
||||
.Fa pg .
|
||||
|
||||
.Pp
|
||||
.Fn uvm_pglistalloc
|
||||
allocates a list of pages for size
|
||||
|
|
@ -615,12 +565,10 @@ The
|
|||
and
|
||||
.Fa waitok
|
||||
arguments are currently ignored.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_pglistfree
|
||||
frees the list of pages pointed to by
|
||||
.Fa list .
|
||||
|
||||
.Pp
|
||||
.Fn uvm_page_physload
|
||||
loads physical memory segments into VM space. It must be called at system
|
||||
|
|
@ -632,7 +580,6 @@ and
|
|||
of the physical addresses of the segment, and the available start and end
|
||||
addresses of pages not already in use.
|
||||
.\" XXX expand on "system boot time"!
|
||||
|
||||
.Pp
|
||||
.Fn uvm_km_suballoc
|
||||
allocates submap from
|
||||
|
|
@ -658,7 +605,6 @@ and
|
|||
If
|
||||
.Fa pageable
|
||||
is non-zero, entries in the map may be paged out.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_km_free
|
||||
and
|
||||
|
|
@ -671,10 +617,7 @@ bytes of memory in the kernal map, starting at address
|
|||
calls
|
||||
.Fn thread_wakeup
|
||||
on the map before unlocking the map.
|
||||
|
||||
.Pp
|
||||
.Sh PROCESSES
|
||||
|
||||
.Pp
|
||||
.nr nS 1
|
||||
.Ft void
|
||||
|
|
@ -684,25 +627,19 @@ on the map before unlocking the map.
|
|||
.Ft void
|
||||
.Fn uvm_swapin "struct proc *p"
|
||||
.nr nS 0
|
||||
|
||||
.Pp
|
||||
.Fn uvm_pageout
|
||||
is the main loop for the page daemon.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_scheduler
|
||||
is the process zero main loop, which is to be called after the
|
||||
system has finished starting other processes. It handles the
|
||||
swapping in of runnable, swapped out processes in priority
|
||||
order.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_swapin
|
||||
swaps in the named process.
|
||||
|
||||
.Pp
|
||||
.Sh MISCELLANEOUS FUNCTIONS
|
||||
|
||||
.nr nS 1
|
||||
.Pp
|
||||
.Ft struct uvm_object *
|
||||
|
|
@ -736,7 +673,6 @@ swaps in the named process.
|
|||
.Ft int
|
||||
.Fn uvm_coredump "struct proc *p" "struct vnode *vp" "struct ucred *cred" "struct core *chdr"
|
||||
.nr nS 0
|
||||
|
||||
.Pp
|
||||
The
|
||||
.Fn uao_create ,
|
||||
|
|
@ -760,7 +696,6 @@ creates an additional reference to the named anonymous memory object.
|
|||
.Fn uao_detach
|
||||
removes a reference from the named anonymous memory object, destroying
|
||||
it if removing the last reference.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_chgkprot
|
||||
changes the protection of kernel memory from
|
||||
|
|
@ -772,7 +707,6 @@ to the value of
|
|||
This is primarily useful for debuggers, for setting breakpoints.
|
||||
This function is only available with options
|
||||
.Dv KGDB .
|
||||
|
||||
.Pp
|
||||
.Fn uvm_kernacc
|
||||
and
|
||||
|
|
@ -785,7 +719,6 @@ for
|
|||
.Fa rw
|
||||
access, in the kernel address space, and the current process'
|
||||
address space respectively.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_vslock
|
||||
and
|
||||
|
|
@ -797,11 +730,9 @@ from
|
|||
to
|
||||
.Fa addr + len .
|
||||
These functions are normally used to wire memory for I/O.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_meter
|
||||
calculates the load average and wakes up the swapper if necessary.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_sysctl
|
||||
provides support for the
|
||||
|
|
@ -902,7 +833,6 @@ int pdpageouts; /* number of times daemon started a pageout */
|
|||
int pdpending; /* number of times daemon got a pending pagout */
|
||||
int pddeact; /* number of pages daemon deactivates */
|
||||
.Ed
|
||||
|
||||
.Pp
|
||||
.Fn uvm_fork
|
||||
forks a virtual address space for process' (old)
|
||||
|
|
@ -916,14 +846,12 @@ otherwise a new address space is created. This function
|
|||
currently has no return value, and thus cannot fail. In
|
||||
the future, this function will changed to allowed it to
|
||||
fail in low memory conditions.
|
||||
|
||||
.Pp
|
||||
.Fn uvm_grow
|
||||
increases the stack segment of process
|
||||
.Fa p
|
||||
to include
|
||||
.Fa sp .
|
||||
|
||||
.Pp
|
||||
.Fn uvm_coredump
|
||||
generates a coredump on vnode
|
||||
|
|
@ -934,7 +862,6 @@ with credentials
|
|||
.Fa cred
|
||||
and core header description in
|
||||
.Fa chdr .
|
||||
|
||||
.Sh STANDARD UVM RETURN VALUES
|
||||
This section documents the standard return values that callers of UVM
|
||||
functions can expect. They are derived from the Mach VM values
|
||||
|
|
@ -957,7 +884,6 @@ Note that
|
|||
and
|
||||
.Dv KERN_PAGES_LOCKED
|
||||
values are not actually returned by the UVM code.
|
||||
|
||||
.Sh NOTES
|
||||
These functions are only available with options
|
||||
.Dv UVM .
|
||||
|
|
@ -974,7 +900,6 @@ and types whose names begin with ``vm_'' will be renamed to ``uvm_''.
|
|||
The
|
||||
.Xr pmap 9
|
||||
manual page is not yet written.
|
||||
|
||||
.Sh HISTORY
|
||||
UVM is a new VM system developed at Washington University in St. Louis
|
||||
(Missouri). UVM's roots lie partly in the Mach-based
|
||||
|
|
@ -992,7 +917,6 @@ further documented in August 1998 in a dissertation by Charles D. Cranor.
|
|||
.Pp
|
||||
UVM appeared in
|
||||
.Nx 1.4 .
|
||||
|
||||
.Sh AUTHOR
|
||||
Charles D. Cranor <chuck@ccrc.wustl.edu> designed and implemented UVM.
|
||||
.Pp
|
||||
|
|
@ -1002,7 +926,6 @@ NetBSD source tree.
|
|||
.Pp
|
||||
Chuck Silvers <chuq@chuq.com> implemented the aobj pager, thus allowing
|
||||
UVM to support System V shared memory and process swapping.
|
||||
|
||||
.Sh SEE ALSO
|
||||
.Xr getloadavg 3 ,
|
||||
.Xr kvm 3 ,
|
||||
|
|
|
|||
Loading…
Reference in New Issue