nuttx/mm
2013-09-29 11:46:10 -06:00
..
bin Directories where the same source files are used to build different objects need to keep those objects in separate directories; LPC17xx updates from Chris Taglia 2013-03-19 23:40:07 +00:00
kbin Directories where the same source files are used to build different objects need to keep those objects in separate directories; LPC17xx updates from Chris Taglia 2013-03-19 23:40:07 +00:00
ubin Directories where the same source files are used to build different objects need to keep those objects in separate directories; LPC17xx updates from Chris Taglia 2013-03-19 23:40:07 +00:00
.gitignore Dependency generation fix for directories that keep object files in a sub-directory 2013-09-29 11:46:10 -06:00
Kconfig Beginning updates of SAM3U header files o include support for the SAM4S: WDT, SUPC, EEFC, MATRIX, and PMC 2013-06-10 11:57:37 -06:00
Makefile Dependency generation fix for directories that keep object files in a sub-directory 2013-09-29 11:46:10 -06:00
mm_addfreechunk.c Add support for multiple heaps 2013-03-08 20:36:18 +00:00
mm_calloc.c Add support for multiple heaps 2013-03-08 20:36:18 +00:00
mm_free.c Add support for multiple heaps 2013-03-08 20:36:18 +00:00
mm_gran.h Fix MMC/SD support for Wildfire board; Granule allocator can now be used from intrrupt handler 2012-09-12 15:18:56 +00:00
mm_granalloc.c Misc fixes and optimizations for the granule allocator 2012-09-12 17:55:03 +00:00
mm_grancritical.c New file missed in last check-in 2012-09-12 15:21:26 +00:00
mm_granfree.c Misc fixes and optimizations for the granule allocator 2012-09-12 17:55:03 +00:00
mm_graninit.c Fix all occurrences of "the the" in documentation and comments 2013-08-27 09:40:19 -06:00
mm_initialize.c More changes for a kernel-mode allocator (more to be done) 2013-03-09 21:12:20 +00:00
mm_kernel.c Add kernel/user memalign functions. Not fully integrated 2013-08-20 13:04:49 -06:00
mm_mallinfo.c Add support for multiple heaps 2013-03-08 20:36:18 +00:00
mm_malloc.c Need to include debug.h or compile fails when memory management debug is enabled 2013-04-18 11:40:38 -06:00
mm_memalign.c Add support for multiple heaps 2013-03-08 20:36:18 +00:00
mm_realloc.c Fix a bug introduced in the last check-in 2013-03-11 00:04:50 +00:00
mm_sem.c Add support for multiple heaps 2013-03-08 20:36:18 +00:00
mm_shrinkchunk.c Add support for multiple heaps 2013-03-08 20:36:18 +00:00
mm_size2ndx.c Add support for multiple heaps 2013-03-08 20:36:18 +00:00
mm_user.c Fixe to kernel build and syscalls. 2013-03-13 17:32:37 +00:00
mm_zalloc.c Add support for multiple heaps 2013-03-08 20:36:18 +00:00
README.txt Add power management register defintions and clock control logic for the SAM4L 2013-06-05 13:35:19 -06:00

mm/README.txt
=============

This directory contains the NuttX memory management logic.  This include:

1) Standard Memory Management Functions:

     The standard memory management functions as prototyped in stdlib.h as
     specified in the  Base definitions volume of IEEE Std 1003.1-2001.  This
     include the files:

     o Standard Interfaces: mm_malloc.c, mm_calloc.c, mm_realloc.c,
       mm_memalign.c, mm_free.c
     o Less-Standard Interfaces: mm_zalloc.c, mm_mallinfo.c
     o Internal Implementation: mm_initialize.c mm_sem.c  mm_addfreechunk.c
       mm_size2ndx.c mm_shrinkchunk.c, mm_internal.h
     o Build and Configuration files: Kconfig, Makefile

   Memory Models:

     o Small Memory Model.  If the MCU supports only 16-bit data addressing
       then the small memory model is automatically used.  The maximum size
       of the heap is then 64K.  The small memory model can also be forced
       MCUs with wider addressing by defining CONFIG_SMALL_MEMORY in the
       NuttX configuration file.
     o Large Memory Model.  Otherwise, the allocator uses a model that
       supports a heap of up to 4G.

     This implementation uses a variable length allocator with the following
     properties:

     o Overhead:  Either 8- or 4-bytes per allocation for large and small
       models, respectively.
     o Alignment:  All allocations are aligned to 8- or 4-bytes for large
       and small models, respectively.

   Multiple Heaps:

     This allocator can be used to manage multiple heaps (albeit with some
     non-standard interfaces).  A heap is represented by struct mm_heap_s
     as defined in the file include/nuttx/mm.h.  To create another heap
     instance, you would allocate a heap structure, most likely statically
     in memory:

       include <nuttx/mm.h>
       static struct mm_heap_s g_myheap;

     Then initialize the heap using:

       mm_initialize(&g_myheap, myheap_start, myheap_size);

     Where mm_initialize() and all related interfaces are prototyped in the
     header file include/nuttx/mm.h.

     After the new heap instance has been initialized, it can then be used
     with these almost familiar interfaces: mm_malloc(), mm_realloc(), mm_free(),
     etc.  These are 'almost familiar' because they are analogous of the
     standard malloc(), realloc(), free(), etc. except that they expect a
     reference to the initialized heap structure as the first parameter.

     In fact, the standard malloc(), realloc(), free() use this same mechanism,
     but with a global heap structure called g_mmheap.

2) Granule Allocator.

     A non-standard granule allocator is also available in this directory  The
     granule allocator allocates memory in units of a fixed sized block ("granule").
     Allocations may be aligned to a user-provided address boundary.

     The granule allocator interfaces are defined in nuttx/include/nuttx/gran.h.
     The granule allocator consists of these files in this directory:

       mm_gran.h, mm_granalloc.c, mm_grancritical.c, mm_granfree.c
       mm_graninit.c

     The granule allocator is not used anywhere within the base NuttX code
     as of this writing.  The intent of the granule allocator is to provide
     a tool to support platform-specific management of aligned DMA memory.

     NOTE: Because each granule may be aligned and each allocation is in
     units of the granule size, selection of the granule size is important:
     Larger granules will give better performance and less overhead but more
     losses of memory due to quantization waste.  Additional memory waste
     can occur from alignment;  Of course, heap alignment should no be
     used unless (a) you are using the granule allocator to manage DMA memory
     and (b) your hardware has specific memory alignment requirements.

     The current implementation also restricts the maximum allocation size
     to 32 granules.  That restriction could be eliminated with some
     additional coding effort, but currently requires larger granule
     sizes for larger allocations.

   General Usage Example.

     This is an example using the GCC section attribute to position a DMA
     heap in memory (logic in the linker script would assign the section
     .dmaheap to the DMA memory.

        FAR uint32_t g_dmaheap[DMAHEAP_SIZE] __attribute__((section(.dmaheap)));

     The heap is created by calling gran_initialize.  Here the granule size
     is set to 64 bytes and the alignment to 16 bytes:

       GRAN_HANDLE handle = gran_initialize(g_dmaheap, DMAHEAP_SIZE, 6, 4);

     Then the GRAN_HANDLE can be used to allocate memory (There is no
     GRAN_HANDLE if CONFIG_GRAN_SINGLE=y):

       FAR uint8_t *dma_memory = (FAR uint8_t *)gran_alloc(handle, 47);

     The actual memory allocates will be 64 byte (wasting 17 bytes) and
     will be aligned at least to (1 << log2align).