.\" $OpenBSD: extent.9,v 1.8 2003/06/06 20:56:32 jmc Exp $ .\" $NetBSD: extent.9,v 1.15 1999/03/16 00:40:47 garbled Exp $ .\" .\" Copyright (c) 1996, 1998 The NetBSD Foundation, Inc. .\" All rights reserved. .\" .\" This code is derived from software contributed to The NetBSD Foundation .\" by Jason R. Thorpe and Greg Hudson. .\" .\" 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. All advertising materials mentioning features or use of this software .\" must display the following acknowledgement: .\" This product includes software developed by the NetBSD .\" Foundation, Inc. and its contributors. .\" 4. Neither the name of The NetBSD Foundation nor the names of its .\" contributors may be used to endorse or promote products derived .\" from this software without specific prior written permission. .\" .\" THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS .\" ``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 FOUNDATION OR CONTRIBUTORS .\" 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 September 23, 1996 .Dt EXTENT 9 .Os .Sh NAME .Nm extent_create , .Nm extent_destroy , .Nm extent_alloc , .Nm extent_alloc_subregion , .Nm extent_alloc_region , .Nm extent_free , .Nm extent_print .Nd general purpose extent manager .Sh SYNOPSIS .Fd #include .Fd #include .Ft struct extent * .Fn extent_create "char *name" "u_long start" "u_long end" "int mtype" "caddr_t storage" "size_t storagesize" "int flags" .Ft void .Fn extent_destroy "struct extent *ex" .Ft int .Fn extent_alloc "struct extent *ex" "u_long size" "u_long alignment" "u_long skew" "u_long boundary" "int flags" "u_long *result" .Ft int .\" too many arguments for a single .Fn .Fo extent_alloc_subregion .Fa "struct extent *ex" .Fa "u_long substart" .Fa "u_long subend" .Fa "u_long size" .Fa "u_long alignment" .Fa "u_long skew" .Fa "u_long boundary" .Fa "int flags" .Fa "u_long *result" .Fc .Ft int .Fn extent_alloc_region "struct extent *ex" "u_long start" "u_long size" "int flags" .Ft int .Fn extent_free "struct extent *ex" "u_long start" "u_long size" "int flags" .Ft void .Fn extent_print "struct extent *ex" .Sh DESCRIPTION The extent manager provides management of areas of memory or other enumerable spaces (such as I/O ports). An opaque structure called an .Nm extent map keeps track of allocated regions within the enumerable space. .Pp .Fn extent_create creates an extent map managing the space from .Fa start to .Fa end inclusive. All memory allocation will use the memory type .Fa mtype .Po see .Xr malloc 9 .Pc . The extent map will have the name .Fa name , used for identification in case of errors or in .Xr ddb 4 .Ic show extents . If the flag .Dv EX_NOCOALESCE is set, internal coalescing of regions is disabled, and only entire regions may be freed within the extent map, so that .Fn extent_free will never have to allocate a region descriptor. .Pp Some applications may want to use an extent map but can't use .Fn malloc and .Fn free . These applications may provide pre-allocated storage for all descriptor overhead with the arguments .Fa storage and .Fa storagesize . An extent of this type is called a .Nm fixed extent . If the application can safely use .Fn malloc and .Fn free , .Fa storage should be .Dv NULL . A fixed extent has a fixed number of region descriptors, so care should be taken to provide enough storage for them; alternatively, the flag .Dv EX_MALLOCOK may be passed to extent requests to indicate that a fixed extent map may be extended using a call to .Fn malloc . .Pp The caller should pass the flag .Dv EX_WAITOK or .Dv EX_NOWAIT to extent functions that have a memory overhead, to specify whether it is okay to wait. These functions are .Fn extent_create (non fixed extents), .Fn extent_free (unless .Dv EX_NOCOALESCE is set), .Fn extent_alloc , .Fn extent_alloc_subregion and .Fn extent_alloc_region . .Pp .Fn extent_destroy destroys the extent map .Fa ex , freeing all allocated regions. If the extent is not a fixed extent, the region and internal extent descriptors themselves are freed. This function always succeeds. .Pp .Fn extent_alloc allocates a region in the extent map .Fa ex of size .Fa size that fits the provided parameters. There are two distinct allocation policies, which are selected by the .Fa flags argument: .Bl -tag -offset indent -width "XXXXXXXXX" .It Dv EX_FAST Allocate the first region that fits the provided parameters, regardless of resulting extent fragmentation. .It default Allocate the smallest region that is capable of holding the request, thus minimizing fragmentation of the extent. .El .Pp The caller may specify that it is okay to wait for space to become free in the extent by setting the flag .Dv EX_WAITSPACE . If .Dv EX_WAITSPACE is not set, the allocation will fail if the request can not be satisfied without sleeping. .Pp The request will be aligned to a multiple of .Fa alignment . That value must be a power of 2. If no alignment is necessary, the value .Dv EX_NOALIGN should be specified. If .Fa skew is non-zero, it modifies the requested alignment result in the following way: the value .Po Fa result \& - .Fa skew .Pc is aligned to .Fa alignment boundaries. .Fa skew must be a smaller number than .Fa alignment . If .Fa boundary is not .Dv EX_NOBOUNDARY , the allocated region will not cross any boundary lines, spaced .Fa boundary apart. If the caller specifies the .Dv EX_BOUNDZERO flag, boundary lines begin at zero. Otherwise, boundary lines begin at the beginning of the extent. The allocated region may begin on a boundary line, but the end of the region will not touch nor cross a boundary line. A .Fa boundary argument smaller than the sum of the requested skew and the size of the request is invalid. Upon successful completion, .Fa *result will contain the start of the allocated region. .Pp .Fn extent_alloc_subregion is a generalized version of .Fn extent_alloc that also allows the caller to specify that the allocated region must fall within the subregion from .Fa substart to .Fa subend inclusive. .Pp .Fn extent_alloc_region allocates the specific region in the extent map .Fa ex beginning at .Fa start with the size .Fa size . The caller may specify that it is okay to wait for the indicated region to be free by setting the flag .Dv EX_WAITSPACE . If .Dv EX_WAITSPACE is not set, the allocation will fail if the request can not be satisfied without sleeping. .Pp .Fn extent_free frees a region of .Fa size bytes starting at .Fa start in the extent map .Fa ex . If the extent has the .Dv EX_NOCOALESCE property, only entire regions may be freed. If the extent has the .Dv EX_NOCOALESCE property and the caller attempts to free a partial region, behavior is undefined. If called on an extent without the .Dv EX_NOCOALESCE property, this function can fail with error codes listed below, otherwise this function will always succeed. .Pp .Fn extent_print Prints out information about extent .Fa ex . This function always succeeds. .Sh RETURN VALUES The behavior of all extent manager functions is undefined if given invalid arguments. .Fn extent_create returns the extent map on success, or .Dv NULL if it fails to allocate storage for the extent map. It always succeeds when creating a fixed extent or when given the flag .Dv EX_WAITOK . .Fn extent_alloc , .Fn extent_alloc_region , .Fn extent_alloc_subregion , and .Fn extent_free return one of the following values: .Bl -tag -offset indent -width "XXXXXXXX" .It Dv 0 Operation was successful. .It Dv ENOMEM If .Dv EX_NOWAIT is specified, the extent manager was not able to allocate a region descriptor for the new region or to split a region when freeing a partial region. .It Dv EAGAIN Requested region is not available and .Dv EX_WAITSPACE was not specified. .It Dv EINTR Process received a signal while waiting for the requested region to become available in the extent. .El .Sh EXAMPLES Here is an example of a (useless) function that uses several of the extent manager routines. .Bd -literal void func() { struct extent *foo_ex; u_long region_start; int error; /* * Extent "foo" manages a 256k region starting at 0x0 and * only allows complete regions to be freed so that * extent_free() never needs to allocate memory. */ foo_ex = extent_create("foo", 0x0, 0x3ffff, M_DEVBUF, NULL, 0, EX_WAITOK | EX_NOCOALESCE); /* * Allocate an 8k region, aligned to a 4k boundary, which * does not cross any of the 3 64k boundaries (at 64k, * 128k, and 192k) within the extent. */ error = extent_alloc(foo_ex, 0x2000, 0x1000, 0x10000, EX_NOWAIT, ®ion_start); if (error) panic("you lose"); /* * Give up the extent. */ extent_destroy(foo_ex); } .Ed .\" .\" Yeah, right... document EX_CATCH first... .\" .\" .Sh LIMITATIONS .\" The flag .\" .Dv EX_CATCH .\" cannot be used to catch signals in all circumstances since .\" .Xr malloc 9 .\" does not provide such a functionality. .Sh CODE REFERENCES The extent manager itself is implemented within the file .Pa sys/kern/subr_extent.c . .Pp The i386 bus management code uses the extent manager for managing I/O ports and I/O memory. See .Pa sys/arch/i386/i386/machdep.c . .Sh SEE ALSO .Xr ddb 4 , .Xr malloc 9 .Sh HISTORY The extent manager appeared in .Nx 1.3 . .Sh AUTHORS The extent manager was designed and implemented by .An Jason R. Thorpe Aq thorpej@NetBSD.ORG . .An Matthias Drochner Aq drochner@zelux6.zel.kfa-juelich.de contributed to the initial testing and optimization of the implementation. .An Chris Demetriou Aq cgd@NetBSD.ORG contributed many architectural suggestions.