4.  Appendix A - Virtual Memory Interface

4.1.  Mapping pages

      The system supports sharing of data between processes by allowing pages to be mapped into memory. These mapped pages may be shared with other processes or private to the process. Protection and sharing options are defined in <sys/mman.h> as: 

/* protections are chosen from these bits, or-ed together */
#define	PROT_READ	0x04	/* pages can be read */
#define	PROT_WRITE	0x02	/* pages can be written */
#define	PROT_EXEC	0x01	/* pages can be executed */

/* flags contain mapping type, sharing type and options */
/* mapping type; choose one */
#define MAP_FILE	0x0001	/* mapped from a file or device */
#define MAP_ANON	0x0002	/* allocated from memory, swap space */
#define MAP_TYPE	0x000f	/* mask for type field */

/* sharing types; choose one */
#define	MAP_SHARED	0x0010	/* share changes */
#define	MAP_PRIVATE	0x0000	/* changes are private */

/* other flags */
#define MAP_FIXED	0x0020	/* map addr must be exactly as requested */
#define MAP_INHERIT	0x0040	/* region is retained after exec */
#define MAP_HASSEMAPHORE	0x0080	/* region may contain semaphores */
The cpu-dependent size of a page is returned by the getpagesize system call: 
pagesize = getpagesize();
result int pagesize;

The call: 

maddr = mmap(addr, len, prot, flags, fd, pos);
result caddr_t maddr; caddr_t addr; int *len, prot, flags, fd; off_t pos;
causes the pages starting at addr and continuing for at most len bytes to be mapped from the object represented by descriptor fd, starting at byte offset pos. The starting address of the region is returned; for the convenience of the system, it may differ from that supplied unless the MAP_FIXED flag is given, in which case the exact address will be used or the call will fail. The actual amount mapped is returned in len. The addr, len, and pos parameters must all be multiples of the pagesize. A successful mmap will delete any previous mapping in the allocated address range. The parameter prot specifies the accessibility of the mapped pages. The parameter flags specifies the type of object to be mapped, mapping options, and whether modifications made to this mapped copy of the page are to be kept private, or are to be shared with other references. Possible types include MAP_FILE, mapping a regular file or character-special device memory, and MAP_ANON, which maps memory not associated with any specific file. The file descriptor used for creating MAP_ANON regions is used only for naming, and may be given as -1 if no name is associated with the region.** The MAP_INHERIT flag allows a region to be inherited after an exec. The MAP_HASSEMAPHORE flag allows special handling for regions that may contain semaphores.

      A facility is provided to synchronize a mapped region with the file it maps; the call 

msync(addr, len);
caddr_t addr; int len;
writes any modified pages back to the filesystem and updates the file modification time. If len is 0, all modified pages within the region containing addr will be flushed; if len is non-zero, only the pages containing addr and len succeeding locations will be examined. Any required synchronization of memory caches will also take place at this time. Filesystem operations on a file that is mapped for shared modifications are unpredictable except after an msync.

      A mapping can be removed by the call 

munmap(addr, len);
caddr_t addr; int len;
This call deletes the mappings for the specified address range, and causes further references to addresses within the range to generate invalid memory references.

4.2.  Page protection control

      A process can control the protection of pages using the call 

mprotect(addr, len, prot);
caddr_t addr; int len, prot;
This call changes the specified pages to have protection prot. Not all implementations will guarantee protection on a page basis; the granularity of protection changes may be as large as an entire region.

4.3.  Giving and getting advice

      A process that has knowledge of its memory behavior may use the madvise call: 

madvise(addr, len, behav);
caddr_t addr; int len, behav;
Behav describes expected behavior, as given in <sys/mman.h>: 
#define	MADV_NORMAL	0	/* no further special treatment */
#define	MADV_RANDOM	1	/* expect random page references */
#define	MADV_SEQUENTIAL	2	/* expect sequential references */
#define	MADV_WILLNEED	3	/* will need these pages */
#define	MADV_DONTNEED	4	/* don't need these pages */
#define	MADV_SPACEAVAIL	5	/* insure that resources are reserved */
Finally, a process may obtain information about whether pages are core resident by using the call 
mincore(addr, len, vec)
caddr_t addr; int len; result char *vec;
Here the current core residency of the pages is returned in the character array vec, with a value of 1 meaning that the page is in-core.

4.4.  Synchronization primitives

      Primitives are provided for synchronization using semaphores in shared memory. Semaphores must lie within a MAP_SHARED region with at least modes PROT_READ and PROT_WRITE. The MAP_HASSEMAPHORE flag must have been specified when the region was created. To acquire a lock a process calls: 

value = mset(sem, wait)
result int value; semaphore *sem; int wait;
Mset indivisibly tests and sets the semaphore sem. If the previous value is zero, the process has acquired the lock and mset returns true immediately. Otherwise, if the wait flag is zero, failure is returned. If wait is true and the previous value is non-zero, mset relinquishes the processor until notified that it should retry.

To release a lock a process calls: 

mclear(sem)
semaphore *sem;
Mclear indivisibly tests and clears the semaphore sem. If the ``WANT'' flag is zero in the previous value, mclear returns immediately. If the ``WANT'' flag is non-zero in the previous value, mclear arranges for waiting processes to retry before returning.

      Two routines provide services analogous to the kernel sleep and wakeup functions interpreted in the domain of shared memory. A process may relinquish the processor by calling msleep with a set semaphore: 

msleep(sem)
semaphore *sem;
If the semaphore is still set when it is checked by the kernel, the process will be put in a sleeping state until some other process issues an mwakeup for the same semaphore within the region using the call: 
mwakeup(sem)
semaphore *sem;
An mwakeup may awaken all sleepers on the semaphore, or may awaken only the next sleeper on a queue.