.\" $NetBSD: uvm.9,v 1.49 2003/11/03 08:07:36 wiz Exp $ .\" .\" Copyright (c) 1998 Matthew R. Green .\" 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. .\" 3. The name of the author may not be used to endorse or promote products .\" derived from this software without specific prior written permission. .\" .\" 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. .\" .Dd November 3, 2003 .Dt UVM 9 .Os .Sh NAME .Nm uvm .Nd virtual memory system external interface .Sh SYNOPSIS .In sys/param.h .In uvm/uvm.h .Sh DESCRIPTION The UVM virtual memory system manages access to the computer's memory resources. User processes and the kernel access these resources through UVM's external interface. UVM's external interface includes functions that: .Pp .Bl -hyphen -compact .It initialise UVM sub-systems .It manage virtual address spaces .It resolve page faults .It memory map files and devices .It perform uio-based I/O to virtual memory .It allocate and free kernel virtual memory .It allocate and free physical memory .El .Pp In addition to exporting these services, UVM has two kernel-level processes: pagedaemon and swapper. The pagedaemon process sleeps until physical memory becomes scarce. When that happens, pagedaemon is awoken. It scans physical memory, paging out and freeing memory that has not been recently used. The swapper process swaps in runnable processes that are currently swapped out, if there is room. .Pp There are also several miscellaneous functions. .Sh INITIALISATION .Ft void .br .Fn uvm_init "void" ; .Pp .Ft void .br .Fn uvm_init_limits "struct proc *p" ; .Pp .Ft void .br .Fn uvm_setpagesize "void" ; .Pp .Ft void .br .Fn uvm_swap_init "void" ; .Pp .Fn uvm_init sets up the UVM system at system boot time, after the copyright has been printed. It initialises 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 machine-dependent code), pageshift and pagemask. It should be called by machine-dependent code early in the .Fn pmap_init call (see .Xr pmap 9 ) . .Pp .Fn uvm_swap_init initialises the swap sub-system. .Sh VIRTUAL ADDRESS SPACE MANAGEMENT .Ft int .br .Fn uvm_map "struct vm_map *map" "vaddr_t *startp" "vsize_t size" "struct uvm_object *uobj" "voff_t uoffset" "vsize_t align" "uvm_flag_t flags" ; .Pp .Ft int .br .Fn uvm_map_pageable "struct vm_map *map" "vaddr_t start" "vaddr_t end" "boolean_t new_pageable" "int lockflags" ; .Pp .Ft boolean_t .br .Fn uvm_map_checkprot "struct vm_map *map" "vaddr_t start" "vaddr_t end" "vm_prot_t protection" ; .Pp .Ft int .br .Fn uvm_map_protect "struct vm_map *map" "vaddr_t start" "vaddr_t end" "vm_prot_t new_prot" "boolean_t set_max" ; .Pp .Ft int .br .Fn uvm_deallocate "struct vm_map *map" "vaddr_t start" "vsize_t size" ; .Pp .Ft struct vmspace * .br .Fn uvmspace_alloc "vaddr_t min" "vaddr_t max" "int pageable" ; .Pp .Ft void .br .Fn uvmspace_exec "struct proc *p" "vaddr_t start" "vaddr_t end" ; .Pp .Ft struct vmspace * .br .Fn uvmspace_fork "struct vmspace *vm" ; .Pp .Ft void .br .Fn uvmspace_free "struct vmspace *vm1" ; .Pp .Ft void .br .Fn uvmspace_share "struct proc *p1" "struct proc *p2" ; .Pp .Ft void .br .Fn uvmspace_unshare "struct proc *p" ; .Pp .Ft boolean_t .br .Fn uvm_uarea_alloc "vaddr_t *uaddrp" ; .Pp .Ft void .br .Fn uvm_uarea_free "vaddr_t uaddr" ; .Pp .Fn uvm_map establishes a valid mapping in map .Fa map , which must be unlocked. The new mapping has size .Fa size , which must a multiple of .Dv PAGE_SIZE . The .Fa uobj and .Fa uoffset arguments can have four meanings. When .Fa uobj is .Dv NULL and .Fa uoffset is .Dv UVM_UNKNOWN_OFFSET , .Fn uvm_map does not use the machine-dependent .Dv PMAP_PREFER function. If .Fa uoffset is any other value, it is used as the hint to .Dv PMAP_PREFER . When .Fa uobj is not .Dv NULL and .Fa uoffset is .Dv UVM_UNKNOWN_OFFSET , .Fn uvm_map finds the offset based upon the virtual address, passed as .Fa startp . If .Fa uoffset is any other value, we are doing a normal mapping at this offset. The start address of the map will be returned in .Fa startp . .Pp .Fa align specifies alignment of mapping unless .Dv UVM_FLAG_FIXED is specified in .Fa flags . .Fa align must be a power of 2. .Pp .Fa flags passed to .Fn uvm_map are typically created using the .Fn UVM_MAPFLAG "vm_prot_t prot" "vm_prot_t maxprot" "vm_inherit_t inh" "int advice" "int flags" macro, which uses the following values. The .Fa prot and .Fa maxprot can take are: .Bd -literal #define UVM_PROT_MASK 0x07 /* protection mask */ #define UVM_PROT_NONE 0x00 /* protection none */ #define UVM_PROT_ALL 0x07 /* everything */ #define UVM_PROT_READ 0x01 /* read */ #define UVM_PROT_WRITE 0x02 /* write */ #define UVM_PROT_EXEC 0x04 /* exec */ #define UVM_PROT_R 0x01 /* read */ #define UVM_PROT_W 0x02 /* write */ #define UVM_PROT_RW 0x03 /* read-write */ #define UVM_PROT_X 0x04 /* exec */ #define UVM_PROT_RX 0x05 /* read-exec */ #define UVM_PROT_WX 0x06 /* write-exec */ #define UVM_PROT_RWX 0x07 /* read-write-exec */ .Ed .Pp The values that .Fa inh can take are: .Bd -literal #define UVM_INH_MASK 0x30 /* inherit mask */ #define UVM_INH_SHARE 0x00 /* "share" */ #define UVM_INH_COPY 0x10 /* "copy" */ #define UVM_INH_NONE 0x20 /* "none" */ #define UVM_INH_DONATE 0x30 /* "donate" \*[Lt]\*[Lt] not used */ .Ed .Pp The values that .Fa advice can take are: .Bd -literal #define UVM_ADV_NORMAL 0x0 /* 'normal' */ #define UVM_ADV_RANDOM 0x1 /* 'random' */ #define UVM_ADV_SEQUENTIAL 0x2 /* 'sequential' */ #define UVM_ADV_MASK 0x7 /* mask */ .Ed .Pp The values that .Fa flags can take are: .Bd -literal #define UVM_FLAG_FIXED 0x010000 /* find space */ #define UVM_FLAG_OVERLAY 0x020000 /* establish overlay */ #define UVM_FLAG_NOMERGE 0x040000 /* don't merge map entries */ #define UVM_FLAG_COPYONW 0x080000 /* set copy_on_write flag */ #define UVM_FLAG_AMAPPAD 0x100000 /* for bss: pad amap to reduce malloc() */ #define UVM_FLAG_TRYLOCK 0x200000 /* fail if we can not lock map */ .Ed .Pp The .Dv UVM_MAPFLAG macro arguments can be combined with an or operator. There are several special purpose macros for checking protection combinations, e.g., the .Dv UVM_PROT_WX macro. There are also some additional macros to extract bits from the flags. The .Dv UVM_PROTECTION , .Dv UVM_INHERIT , .Dv UVM_MAXPROTECTION and .Dv UVM_ADVICE 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 .Fa start to .Fa end in map .Fa map 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 .Fa start to .Fa end in map .Fa map against .Fa protection . This returns either .Dv TRUE or .Dv FALSE . .Pp .Fn uvm_map_protect changes the protection .Fa start to .Fa end in map .Fa map to .Fa new_prot , also setting the maximum protection to the region to .Fa new_prot if .Fa set_max is non-zero. This function returns a standard UVM return value. .Pp .Fn uvm_deallocate deallocates kernel memory in map .Fa map from address .Fa start to .Fa start + size . .Pp .Fn uvmspace_alloc allocates and returns a new address space, with ranges from .Fa min to .Fa max , setting the pageability of the address space to .Fa pageable . .Pp .Fn uvmspace_exec either reuses the address space of process .Fa p if there are no other references to it, or creates a new one with .Fn uvmspace_alloc . The range of valid addresses in the address space is reset to .Fa start through .Fa end . .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 .Fa p has its own, unshared address space, by creating a new one if necessary by calling .Fn uvmspace_fork . .Pp .Fn uvm_uarea_alloc allocates virtual space for a u-area (i.e., a kernel stack) and stores its virtual address in .Fa *uaddrp . The return value is .Dv TRUE if the u-area is already backed by wired physical memory, otherwise .Dv FALSE . .Pp .Fn uvm_uarea_free frees a u-area allocated with .Fn uvm_uarea_alloc , freeing both the virtual space and any physical pages which may have been allocated to back that virtual space later. .Sh PAGE FAULT HANDLING .Ft int .br .Fn uvm_fault "struct vm_map *orig_map" "vaddr_t vaddr" "vm_fault_t fault_type" "vm_prot_t access_type" ; .Pp .Fn uvm_fault is the main entry point for faults. It takes .Fa orig_map as the map the fault originated in, a .Fa vaddr offset into the map the fault occurred, .Fa fault_type describing the type of fault, and .Fa access_type describing the type of access requested. .Fn uvm_fault returns a standard UVM return value. .Sh MEMORY MAPPING FILES AND DEVICES .Ft struct uvm_object * .br .Fn uvn_attach "void *arg" "vm_prot_t accessprot" ; .Pp .Ft void .br .Fn uvm_vnp_setsize "struct vnode *vp" "voff_t newsize" ; .Pp .Ft void * .br .Fn ubc_alloc "struct uvm_object *uobj" "voff_t offset" "vsize_t *lenp" "int flags" ; .Pp .Ft void .br .Fn ubc_release "void *va" "int flags" ; .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 .Fa vp to .Fa newsize . Caller must hold a reference to the vnode. If the vnode shrinks, pages no longer used are discarded. .Pp .Fn ubc_alloc creates a kernel mappings of .Fa uobj starting at offset .Fa offset . the desired length of the mapping is pointed to by .Fa lenp , but the actual mapping may be smaller than this. .Fa lenp is updated to contain the actual length mapped. The flags must be one of .Bd -literal #define UBC_READ 0x01 /* mapping will be accessed for read */ #define UBC_WRITE 0x02 /* mapping will be accessed for write */ .Ed .Pp Currently, .Fa uobj must actually be a vnode object. Once the mapping is created, it must be accessed only by methods that can handle faults, such as .Fn uiomove or .Fn kcopy . Page faults on the mapping will result in the vnode's .Fn VOP_GETPAGES method being called to resolve the fault. .Pp .Fn ubc_release frees the mapping at .Fa va for reuse. The mapping may be cached to speed future accesses to the same region of the object. The flags are currently unused. .Sh VIRTUAL MEMORY I/O .Ft int .br .Fn uvm_io "struct vm_map *map" "struct uio *uio" ; .Pp .Fn uvm_io performs the I/O described in .Fa uio on the memory described in .Fa map . .Sh ALLOCATION OF KERNEL MEMORY .Ft vaddr_t .br .Fn uvm_km_alloc "struct vm_map *map" "vsize_t size" ; .Pp .Ft vaddr_t .br .Fn uvm_km_zalloc "struct vm_map *map" "vsize_t size" ; .Pp .Ft vaddr_t .br .Fn uvm_km_alloc1 "struct vm_map *map" "vsize_t size" "boolean_t zeroit" ; .Pp .Ft vaddr_t .br .Fn uvm_km_kmemalloc "struct vm_map *map" "struct uvm_object *obj" "vsize_t size" "int flags" ; .Pp .Ft vaddr_t .br .Fn uvm_km_valloc "struct vm_map *map" "vsize_t size" ; .Pp .Ft vaddr_t .br .Fn uvm_km_valloc_wait "struct vm_map *map" "vsize_t size" ; .Pp .Ft struct vm_map * .br .Fn uvm_km_suballoc "struct vm_map *map" "vaddr_t *min" "vaddr_t *max " "vsize_t size" "boolean_t pageable" "boolean_t fixed" "struct vm_map *submap" ; .Pp .Ft void .br .Fn uvm_km_free "struct vm_map *map" "vaddr_t addr" "vsize_t size" ; .Pp .Ft void .br .Fn uvm_km_free_wakeup "struct vm_map *map" "vaddr_t addr" "vsize_t size" ; .Pp .Fn uvm_km_alloc and .Fn uvm_km_zalloc allocate .Fa size bytes of wired kernel memory in map .Fa map . In addition to allocation, .Fn uvm_km_zalloc zeros the memory. Both of these functions are defined as macros in 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 .Fa size 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 .Fa size bytes of wired kernel memory into .Fa obj . The flags can be any of: .Bd -literal #define UVM_KMF_NOWAIT 0x1 /* matches M_NOWAIT */ #define UVM_KMF_VALLOC 0x2 /* allocate VA only */ #define UVM_KMF_CANFAIL 0x4 /* caller handles failure */ #define UVM_KMF_TRYLOCK UVM_FLAG_TRYLOCK /* try locking only */ .Ed .Pp .Dv UVM_KMF_NOWAIT causes .Fn uvm_km_kmemalloc to return immediately if no memory is available. .Dv UVM_KMF_VALLOC causes no physical pages to be allocated, only virtual space. .Dv UVM_KMF_TRYLOCK causes .Fn uvm_km_kmemalloc to use .Fn simple_lock_try when locking maps. .Dv UVM_KMF_CANFAIL indicates that .Fn uvm_km_kmemalloc can return 0 even if .Dv UVM_KMF_NOWAIT is not specified. (If neither .Dv UVM_KMF_NOWAIT nor .Dv UVM_KMF_CANFAIL are specified, .Fn uvm_km_kmemalloc will never fail, but rather sleep indefinitely until the allocation succeeds.) .Pp .Fn uvm_km_valloc and .Fn uvm_km_valloc_wait return a newly allocated zero-filled address in the kernel map of size .Fa size . .Fn uvm_km_valloc_wait will also wait for kernel memory to become available, if there is a memory shortage. .Pp .Fn uvm_km_free and .Fn uvm_km_free_wakeup free .Fa size bytes of memory in the kernel map, starting at address .Fa addr . .Fn uvm_km_free_wakeup calls .Fn wakeup on the map before unlocking the map. .Pp .Fn uvm_km_suballoc allocates submap from .Fa map , creating a new map if .Fa submap is .Dv NULL . The addresses of the submap can be specified exactly by setting the .Fa fixed argument to non-zero, which causes the .Fa min argument to specify the beginning of the address in the submap. If .Fa fixed is zero, any address of size .Fa size will be allocated from .Fa map and the start and end addresses returned in .Fa min and .Fa max . If .Fa pageable is non-zero, entries in the map may be paged out. .Sh ALLOCATION OF PHYSICAL MEMORY .Ft struct vm_page * .br .Fn uvm_pagealloc "struct uvm_object *uobj" "voff_t off" "struct vm_anon *anon" "int flags" ; .Pp .Ft void .br .Fn uvm_pagerealloc "struct vm_page *pg" "struct uvm_object *newobj" "voff_t newoff" ; .Pp .Ft void .br .Fn uvm_pagefree "struct vm_page *pg" ; .Pp .Ft int .br .Fn uvm_pglistalloc "psize_t size" "paddr_t low" "paddr_t high" "paddr_t alignment" "paddr_t boundary" "struct pglist *rlist" "int nsegs" "int waitok" ; .Pp .Ft void .br .Fn uvm_pglistfree "struct pglist *list" ; .Pp .Ft void .br .Fn uvm_page_physload "vaddr_t start" "vaddr_t end" "vaddr_t avail_start" "vaddr_t avail_end" "int free_list" ; .Pp .Fn uvm_pagealloc allocates a page of memory at virtual address .Fa off in either the object .Fa uobj or the anonymous memory .Fa anon , which must be locked by the caller. Only one of .Fa uobj and .Fa anon can be non .Dv NULL . Returns .Dv NULL when no page can be found. The flags can be any of .Bd -literal #define UVM_PGA_USERESERVE 0x0001 /* ok to use reserve pages */ #define UVM_PGA_ZERO 0x0002 /* returned page must be zero'd */ .Ed .Pp .Dv UVM_PGA_USERESERVE means to allocate a page even if that will result in the number of free pages being lower than .Dv uvmexp.reserve_pagedaemon (if the current thread is the pagedaemon) or .Dv uvmexp.reserve_kernel (if the current thread is not the pagedaemon). .Dv UVM_PGA_ZERO causes the returned page to be filled with zeroes, either by allocating it from a pool of pre-zeroed pages or by zeroing it in-line as necessary. .Pp .Fn uvm_pagerealloc reallocates page .Fa pg to a new object .Fa newobj , at a new offset .Fa newoff . .Pp .Fn uvm_pagefree frees the physical page .Fa pg . If the content of the page is known to be zero-filled, caller should set .Dv PG_ZERO in pg-\*[Gt]flags so that the page allocator will use the page to serve future .Dv UVM_PGA_ZERO requests efficiently. .Pp .Fn uvm_pglistalloc allocates a list of pages for size .Fa size byte under various constraints. .Fa low and .Fa high describe the lowest and highest addresses acceptable for the list. If .Fa alignment is non-zero, it describes the required alignment of the list, in power-of-two notation. If .Fa boundary is non-zero, no segment of the list may cross this power-of-two boundary, relative to zero. .Fa nsegs is the maximum number of physically contigous segments. If .Fa waitok is non-zero, the function may sleep until enough memory is available. (It also may give up in some situations, so a non-zero .Fa waitok does not imply that .Fn uvm_pglistalloc cannot return an error.) The allocated memory is returned in the .Fa rlist list; the caller has to provide storage only, the list is initialized by .Fn uvm_pglistalloc . .Pp .Fn uvm_pglistfree frees the list of pages pointed to by .Fa list . If the content of the page is known to be zero-filled, caller should set .Dv PG_ZERO in pg-\*[Gt]flags so that the page allocator will use the page to serve future .Dv UVM_PGA_ZERO requests efficiently. .Pp .Fn uvm_page_physload loads physical memory segments into VM space on the specified .Fa free_list . It must be called at system boot time to set up physical memory management pages. The arguments describe the .Fa start and .Fa end 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"! .Sh PROCESSES .Ft void .br .Fn uvm_pageout "void" ; .Pp .Ft void .br .Fn uvm_scheduler "void" ; .Pp .Ft void .br .Fn uvm_swapin "struct proc *p" ; .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. .Sh PAGE LOAN .Ft int .br .Fn uvm_loan "struct vm_map *map" "vaddr_t start" "vsize_t len" "void *v" "int flags" ; .Pp .Ft void .br .Fn uvm_unloan "void *v" "int npages" "int flags" ; .Pp .Fn uvm_loan loans pages in a map out to anons or to the kernel. .Fa map should be unlocked , .Fa start and .Fa len should be multiples of .Dv PAGE_SIZE . Argument .Fa flags should be one of .Bd -literal #define UVM_LOAN_TOANON 0x01 /* loan to anons */ #define UVM_LOAN_TOPAGE 0x02 /* loan to kernel */ .Ed .Pp .Fa v should be pointer to array of pointers to .Li struct anon or .Li struct vm_page , as appropriate. The caller has to allocate memory for the array and ensure it's big enough to hold .Fa len / PAGE_SIZE pointers. Returns 0 for success, or appropriate error number otherwise. .Pp .Fn uvm_unloan kills loans on pages or anons. The .Fa v must point to the array of pointers initialized by previous call to .Fn uvm_loan . .Fa npages should match number of pages allocated for loan, this also matches number of items in the array. Argument .Fa flags should be one of .Bd -literal #define UVM_LOAN_TOANON 0x01 /* loan to anons */ #define UVM_LOAN_TOPAGE 0x02 /* loan to kernel */ .Ed .Pp and should match what was used for previous call to .Fn uvm_loan . .Sh MISCELLANEOUS FUNCTIONS .Ft struct uvm_object * .br .Fn uao_create "vsize_t size" "int flags" ; .Pp .Ft void .br .Fn uao_detach "struct uvm_object *uobj" ; .Pp .Ft void .br .Fn uao_reference "struct uvm_object *uobj" ; .Pp .Ft boolean_t .br .Fn uvm_chgkprot "caddr_t addr" "size_t len" "int rw" ; .Pp .Ft void .br .Fn uvm_kernacc "caddr_t addr" "size_t len" "int rw" ; .Pp .Ft boolean_t .br .Fn uvm_useracc "caddr_t addr" "size_t len" "int rw" ; .Pp .Ft int .br .Fn uvm_vslock "struct proc *p" "caddr_t addr" "size_t len" "vm_prot_t prot" ; .Pp .Ft void .br .Fn uvm_vsunlock "struct proc *p" "caddr_t addr" "size_t len" ; .Pp .Ft void .br .Fn uvm_meter "void" ; .Pp .Ft int .br .Fn uvm_sysctl "int *name" "u_int namelen" "void *oldp" "size_t *oldlenp" "void *newp " "size_t newlen" "struct proc *p" ; .Pp .Ft void .br .Fn uvm_fork "struct proc *p1" "struct proc *p2" "boolean_t shared" ; .Pp .Ft int .br .Fn uvm_grow "struct proc *p" "vaddr_t sp" ; .Pp .Ft int .br .Fn uvm_coredump "struct proc *p" "struct vnode *vp" "struct ucred *cred" "struct core *chdr" ; .Pp .Ft void .br .Fn uvn_findpages "struct uvm_object *uobj" "voff_t offset" "int *npagesp" "struct vm_page **pps" "int flags" ; .Pp .Ft void .br .Fn uvm_swap_stats "int cmd" "struct swapent *sep" "int sec" "register_t *retval" ; .Pp The .Fn uao_create , .Fn uao_detach , and .Fn uao_reference functions operate on anonymous memory objects, such as those used to support System V shared memory. .Fn uao_create returns an object of size .Fa size with flags: .Bd -literal #define UAO_FLAG_KERNOBJ 0x1 /* create kernel object */ #define UAO_FLAG_KERNSWAP 0x2 /* enable kernel swap */ .Ed .Pp which can only be used once each at system boot time. .Fn uao_reference 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 .Fa addr to .Fa addr + len to the value of .Fa rw . This is primarily useful for debuggers, for setting breakpoints. This function is only available with options .Dv KGDB . .Pp .Fn uvm_kernacc and .Fn uvm_useracc check the access at address .Fa addr to .Fa addr + len for .Fa rw access, in the kernel address space, and the current process' address space respectively. .Pp .Fn uvm_vslock and .Fn uvm_vsunlock control the wiring and unwiring of pages for process .Fa p from .Fa addr 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 .Dv CTL_VM domain of the .Xr sysctl 3 hierarchy. .Fn uvm_sysctl handles the .Dv VM_LOADAVG , .Dv VM_METER and .Dv VM_UVMEXP calls, which return the current load averages, calculates current VM totals, and returns the uvmexp structure respectively. The load averages are accessed from userland using the .Xr getloadavg 3 function. The uvmexp structure has all global state of the UVM system, and has the following members: .Bd -literal /* vm_page constants */ int pagesize; /* size of a page (PAGE_SIZE): must be power of 2 */ int pagemask; /* page mask */ int pageshift; /* page shift */ /* vm_page counters */ int npages; /* number of pages we manage */ int free; /* number of free pages */ int active; /* number of active pages */ int inactive; /* number of pages that we free'd but may want back */ int paging; /* number of pages in the process of being paged out */ int wired; /* number of wired pages */ int reserve_pagedaemon; /* number of pages reserved for pagedaemon */ int reserve_kernel; /* number of pages reserved for kernel */ /* pageout params */ int freemin; /* min number of free pages */ int freetarg; /* target number of free pages */ int inactarg; /* target number of inactive pages */ int wiredmax; /* max number of wired pages */ /* swap */ int nswapdev; /* number of configured swap devices in system */ int swpages; /* number of PAGE_SIZE'ed swap pages */ int swpginuse; /* number of swap pages in use */ int nswget; /* number of times fault calls uvm_swap_get() */ int nanon; /* number total of anon's in system */ int nfreeanon; /* number of free anon's */ /* stat counters */ int faults; /* page fault count */ int traps; /* trap count */ int intrs; /* interrupt count */ int swtch; /* context switch count */ int softs; /* software interrupt count */ int syscalls; /* system calls */ int pageins; /* pagein operation count */ /* pageouts are in pdpageouts below */ int swapins; /* swapins */ int swapouts; /* swapouts */ int pgswapin; /* pages swapped in */ int pgswapout; /* pages swapped out */ int forks; /* forks */ int forks_ppwait; /* forks where parent waits */ int forks_sharevm; /* forks where vmspace is shared */ /* fault subcounters */ int fltnoram; /* number of times fault was out of ram */ int fltnoanon; /* number of times fault was out of anons */ int fltpgwait; /* number of times fault had to wait on a page */ int fltpgrele; /* number of times fault found a released page */ int fltrelck; /* number of times fault relock called */ int fltrelckok; /* number of times fault relock is a success */ int fltanget; /* number of times fault gets anon page */ int fltanretry; /* number of times fault retrys an anon get */ int fltamcopy; /* number of times fault clears "needs copy" */ int fltnamap; /* number of times fault maps a neighbor anon page */ int fltnomap; /* number of times fault maps a neighbor obj page */ int fltlget; /* number of times fault does a locked pgo_get */ int fltget; /* number of times fault does an unlocked get */ int flt_anon; /* number of times fault anon (case 1a) */ int flt_acow; /* number of times fault anon cow (case 1b) */ int flt_obj; /* number of times fault is on object page (2a) */ int flt_prcopy; /* number of times fault promotes with copy (2b) */ int flt_przero; /* number of times fault promotes with zerofill (2b) */ /* daemon counters */ int pdwoke; /* number of times daemon woke up */ int pdrevs; /* number of times daemon rev'd clock hand */ int pdswout; /* number of times daemon called for swapout */ int pdfreed; /* number of pages daemon freed since boot */ int pdscans; /* number of pages daemon scanned since boot */ int pdanscan; /* number of anonymous pages scanned by daemon */ int pdobscan; /* number of object pages scanned by daemon */ int pdreact; /* number of pages daemon reactivated since boot */ int pdbusy; /* number of times daemon found a busy page */ int pdpageouts; /* number of times daemon started a pageout */ int pdpending; /* number of times daemon got a pending pageout */ int pddeact; /* number of pages daemon deactivates */ .Ed .Pp .Fn uvm_fork forks a virtual address space for process' (old) .Fa p1 and (new) .Fa p2 . If the .Fa shared argument is non zero, p1 shares its address space with p2, otherwise a new address space is created. This function currently has no return value, and thus cannot fail. In the future, this function will be changed to allow 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 .Fa vp for process .Fa p with credentials .Fa cred and core header description in .Fa chdr . .Pp .Fn uvn_findpages looks up or creates pages in .Fa uobj at offset .Fa offset , marks them busy and returns them in the .Fa pps array. Currently .Fa uobj must be a vnode object. The number of pages requested is pointed to by .Fa npagesp , and this value is updated with the actual number of pages returned. The flags can be .Bd -literal #define UFP_ALL 0x00 /* return all pages requested */ #define UFP_NOWAIT 0x01 /* don't sleep */ #define UFP_NOALLOC 0x02 /* don't allocate new pages */ #define UFP_NOCACHE 0x04 /* don't return pages which already exist */ #define UFP_NORDONLY 0x08 /* don't return PG_READONLY pages */ .Ed .Pp .Dv UFP_ALL is a pseudo-flag meaning all requested pages should be returned. .Dv UFP_NOWAIT means that we must not sleep. .Dv UFP_NOALLOC causes any pages which do not already exist to be skipped. .Dv UFP_NOCACHE causes any pages which do already exist to be skipped. .Dv UFP_NORDONLY causes any pages which are marked PG_READONLY to be skipped. .Pp .Fn uvm_swap_stats implements the .Dv SWAP_STATS and .Dv SWAP_OSTATS operation of the .Xr swapctl 2 system call. .Fa cmd is the requested command, .Dv SWAP_STATS or .Dv SWAP_OSTATS . The function will copy no more than .Fa sec entries in the array pointed by .Fa sep . On return, .Fa retval holds the actual number of entries copied in the array. .Sh NOTES .Fn uvm_chgkprot is only available if the kernel has been compiled with options .Dv KGDB . .Pp All structure and types whose names begin with .Dq vm_ will be renamed to .Dq uvm_ . .Sh SEE ALSO .Xr swapctl 2 , .Xr getloadavg 3 , .Xr kvm 3 , .Xr sysctl 3 , .Xr ddb 4 , .Xr options 4 , .Xr pmap 9 .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 .Bx 4.4 VM system, the .Fx VM system, and the SunOS 4 VM system. UVM's basic structure is based on the .Bx 4.4 VM system. UVM's new anonymous memory system is based on the anonymous memory system found in the SunOS 4 VM (as described in papers published by Sun Microsystems, Inc.). UVM also includes a number of feature new to .Bx including page loanout, map entry passing, simplified copy-on-write, and clustered anonymous memory pageout. UVM is also further documented in an August 1998 dissertation by Charles D. Cranor. .Pp UVM appeared in .Nx 1.4 . .Sh AUTHORS Charles D. Cranor .Aq chuck@ccrc.wustl.edu designed and implemented UVM. .Pp Matthew Green .Aq mrg@eterna.com.au wrote the swap-space management code and handled the logistical issues involved with merging UVM into the .Nx source tree. .Pp Chuck Silvers .Aq chuq@chuq.com implemented the aobj pager, thus allowing UVM to support System V shared memory and process swapping. He also designed and implemented the UBC part of UVM, which uses UVM pages to cache vnode data rather than the traditional buffer cache buffers.