From f2c129b281fc8fa30eccb9b34df534f954ee3c47 Mon Sep 17 00:00:00 2001 From: patacongo Date: Fri, 8 Mar 2013 22:01:50 +0000 Subject: up_addregion should use kmm_addregion; move garbage kmm*.c file to mm/. for now git-svn-id: svn://svn.code.sf.net/p/nuttx/code/trunk@5721 42af7a65-404d-4744-a932-0658087f49c3 --- nuttx/mm/Kconfig | 2 +- nuttx/mm/Makefile | 72 ++++++++++++++++++++---- nuttx/mm/README.txt | 14 +---- nuttx/mm/kmm_addregion.c | 113 +++++++++++++++++++++++++++++++++++++ nuttx/mm/kmm_initialize.c | 113 +++++++++++++++++++++++++++++++++++++ nuttx/mm/kmm_kfree.c | 110 ++++++++++++++++++++++++++++++++++++ nuttx/mm/kmm_kmalloc.c | 110 ++++++++++++++++++++++++++++++++++++ nuttx/mm/kmm_krealloc.c | 111 ++++++++++++++++++++++++++++++++++++ nuttx/mm/kmm_kzalloc.c | 110 ++++++++++++++++++++++++++++++++++++ nuttx/mm/kmm_semaphore.c | 140 ++++++++++++++++++++++++++++++++++++++++++++++ nuttx/mm/mm_initialize.c | 75 ++++--------------------- 11 files changed, 881 insertions(+), 89 deletions(-) create mode 100644 nuttx/mm/kmm_addregion.c create mode 100644 nuttx/mm/kmm_initialize.c create mode 100644 nuttx/mm/kmm_kfree.c create mode 100644 nuttx/mm/kmm_kmalloc.c create mode 100644 nuttx/mm/kmm_krealloc.c create mode 100644 nuttx/mm/kmm_kzalloc.c create mode 100644 nuttx/mm/kmm_semaphore.c (limited to 'nuttx/mm') diff --git a/nuttx/mm/Kconfig b/nuttx/mm/Kconfig index 911ae5545..9724eb407 100644 --- a/nuttx/mm/Kconfig +++ b/nuttx/mm/Kconfig @@ -37,7 +37,7 @@ config MM_REGIONS If the architecture includes multiple, non-contiguous regions of memory to allocate from, this specifies the number of memory regions that the memory manager must handle and enables the API - mm_addregion(start, end); + mm_addregion(heap, start, end); NOTE: If MM_MULTIHEAP is selected, then this maximum number of regions applies to all heaps. diff --git a/nuttx/mm/Makefile b/nuttx/mm/Makefile index da41f9f57..4b7412b43 100644 --- a/nuttx/mm/Makefile +++ b/nuttx/mm/Makefile @@ -1,7 +1,7 @@ ############################################################################ # mm/Makefile # -# Copyright (C) 2007, 2012 Gregory Nutt. All rights reserved. +# Copyright (C) 2007, 2012, 2013 Gregory Nutt. All rights reserved. # Author: Gregory Nutt # # Redistribution and use in source and binary forms, with or without @@ -35,22 +35,29 @@ -include $(TOPDIR)/Make.defs -ASRCS = -CSRCS = mm_initialize.c mm_sem.c mm_addfreechunk.c mm_size2ndx.c mm_shrinkchunk.c \ - mm_malloc.c mm_zalloc.c mm_calloc.c mm_realloc.c \ - mm_memalign.c mm_free.c mm_mallinfo.c +ASRCS = +CSRCS = mm_initialize.c mm_sem.c mm_addfreechunk.c mm_size2ndx.c +CSRCS += mm_shrinkchunk.c mm_malloc.c mm_zalloc.c mm_calloc.c mm_realloc.c +CSRCS += mm_memalign.c mm_free.c mm_mallinfo.c + +ifeq ($(CONFIG_NUTTX_KERNEL),y) +CSRCS += kmm_initialize.c kmm_addregion.c kmm_semaphore.c +CSRCS += kmm_kmalloc.c kmm_kzalloc.c kmm_krealloc.c kmm_kfree.c +endif ifeq ($(CONFIG_GRAN),y) -CSRCS += mm_graninit.c mm_granalloc.c mm_granfree.c mm_grancritical.c +CSRCS += mm_graninit.c mm_granalloc.c mm_granfree.c mm_grancritical.c endif -AOBJS = $(ASRCS:.S=$(OBJEXT)) -COBJS = $(CSRCS:.c=$(OBJEXT)) +AOBJS = $(ASRCS:.S=$(OBJEXT)) +COBJS = $(CSRCS:.c=$(OBJEXT)) -SRCS = $(ASRCS) $(CSRCS) -OBJS = $(AOBJS) $(COBJS) +SRCS = $(ASRCS) $(CSRCS) +OBJS = $(AOBJS) $(COBJS) -BIN = libmm$(LIBEXT) +UBIN = libumm$(LIBEXT) +KBIN = libkmm$(LIBEXT) +BIN = libmm$(LIBEXT) all: $(BIN) @@ -63,14 +70,57 @@ $(COBJS): %$(OBJEXT): %.c $(BIN): $(OBJS) $(call ARCHIVE, $@, $(OBJS)) +ifneq ($(BIN),$(UBIN)) +.userlib: + $(Q) $(MAKE) $(UBIN) BIN=$(UBIN) TOPDIR=$(TOPDIR) EXTRADEFINES=$(EXTRADEFINES) + $(Q) touch .userlib + +$(UBIN): kclean .userlib +endif + +ifneq ($(BIN),$(KBIN)) +.kernlib: + $(Q) $(MAKE) $(KBIN) BIN=$(KBIN) TOPDIR=$(TOPDIR) EXTRADEFINES=$(EXTRADEFINES) + $(Q) touch .kernlib + +$(KBIN): uclean .kernlib +endif + .depend: Makefile $(SRCS) $(Q) $(MKDEP) "$(CC)" -- $(CFLAGS) -- $(SRCS) >Make.dep $(Q) touch $@ depend: .depend +# Clean Targets: +# Clean user-mode temporary files (retaining the UBIN binary) + +uclean: +ifneq ($(OBJEXT),) +ifeq ($(CONFIG_WINDOWS_NATIVE),y) + $(Q) if exist .userlib ]; then del *$(OBJEXT) +else + $(Q) ( if [ -f .userlib ]; then rm -f *$(OBJEXT); fi ) +endif +endif + $(call DELFILE, .userlib) + +# Clean kernel-mode temporary files (retaining the KBIN binary) + +kclean: +ifneq ($(OBJEXT),) +ifeq ($(CONFIG_WINDOWS_NATIVE),y) + $(Q) if exist .kernlib ]; then del *$(OBJEXT) +else + $(Q) ( if [ -f .kernlib ]; then rm -f *$(OBJEXT); fi ) +endif +endif + $(call DELFILE, .kernlib) + clean: $(call DELFILE, $(BIN)) + $(call DELFILE, $(UBIN)) + $(call DELFILE, $(KBIN)) $(call CLEAN) distclean: clean diff --git a/nuttx/mm/README.txt b/nuttx/mm/README.txt index 2668432e3..d2f307849 100644 --- a/nuttx/mm/README.txt +++ b/nuttx/mm/README.txt @@ -32,19 +32,7 @@ This directory contains the NuttX memory management logic. This include: o Alignment: All allocations are aligned to 8- or 4-bytes for large and small models, respectively. -2) Test Program. There is also a host-best test program that can be - used to verify the memory manager. These are the file: - - Makefile.test, mm_test.c, and mm_environment.h. - - Build instructions: - - make -f Makefile.test - - The executable will be built in the top-level directory as nuttx/mm_text - (or mm_test.exe under Cygwin). - -3) Granule Allocator. A non-standard granule allocator is also available +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. diff --git a/nuttx/mm/kmm_addregion.c b/nuttx/mm/kmm_addregion.c new file mode 100644 index 000000000..c53fe9cfe --- /dev/null +++ b/nuttx/mm/kmm_addregion.c @@ -0,0 +1,113 @@ +/************************************************************************ + * mm/kmm_addregion.c + * + * Copyright (C) 2011 Gregory Nutt. All rights reserved. + * Author: Gregory Nutt + * + * 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. Neither the name NuttX 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 COPYRIGHT HOLDERS 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 + * COPYRIGHT OWNER 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. + * + ************************************************************************/ + +/************************************************************************ + * Included Files + ************************************************************************/ + +#include +#include + +#if defined(CONFIG_NUTTX_KERNEL) && defined(__KERNEL__) + +/* This logic is all tentatively and, hopefully, will grow in usability. + * For now, the kernel-mode build uses the memory manager that is + * provided in the user-space build. That is awkward but reasonable for + * the current level of support: At present, only memory protection is + * provided. Kernel-mode code may call into user-mode code, but not + * vice-versa. So hosting the memory manager in user-space allows the + * memory manager to be shared in both kernel- and user-mode spaces. + * + * In the longer run, if an MMU is support that can provide virtualized + * memory, then some SLAB memory manager will be required in kernel-space + * with some kind of brk() system call to obtain mapped heap space. + * + * In the current build model, the user-space module is built first. The + * file user_map.h is generated in the first pass and contains the + * addresses of the memory manager needed in this file: + */ + +#include + +/************************************************************************ + * Pre-processor definition + ************************************************************************/ + +/* This value is obtained from user_map.h */ + +#define KADDREGION(h,s) ((kmaddregion_t)CONFIG_USER_MMADDREGION)(h,s) + +/************************************************************************ + * Private Types + ************************************************************************/ + +typedef void (*kmaddregion_t)(FAR void*, size_t); + +/************************************************************************ + * Private Functions + ************************************************************************/ + +/************************************************************************ + * Public Functions + ************************************************************************/ + +/************************************************************************ + * Name: kmm_addregion + * + * Description: + * This is a simple redirection to the user-space mm_addregion() + * function. + * + * Parameters: + * heap_start - Address of the beginning of the memory region + * heap_size - The size (in bytes) if the memory region. + * + * Return Value: + * None + * + * Assumptions: + * 1. mm_addregion() resides in user-space + * 2. The address of the user space mm_addregion() is provided in + * user_map.h + * 3. The user-space mm_addregion() is callable from kernel-space. + * + ************************************************************************/ + +void kmm_addregion(FAR void *heap_start, size_t heap_size) +{ + return KADDREGION(&g_kmmheap, heap_start, heap_size); +} + +#endif /* CONFIG_NUTTX_KERNEL && __KERNEL__ */ diff --git a/nuttx/mm/kmm_initialize.c b/nuttx/mm/kmm_initialize.c new file mode 100644 index 000000000..c5f5df05d --- /dev/null +++ b/nuttx/mm/kmm_initialize.c @@ -0,0 +1,113 @@ +/************************************************************************ + * mm/kmm_initialize.c + * + * Copyright (C) 2011 Gregory Nutt. All rights reserved. + * Author: Gregory Nutt + * + * 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. Neither the name NuttX 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 COPYRIGHT HOLDERS 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 + * COPYRIGHT OWNER 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. + * + ************************************************************************/ + +/************************************************************************ + * Included Files + ************************************************************************/ + +#include +#include + +#if defined(CONFIG_NUTTX_KERNEL) && defined(__KERNEL__) + +/* This logic is all tentatively and, hopefully, will grow in usability. + * For now, the kernel-mode build uses the memory manager that is + * provided in the user-space build. That is awkward but reasonable for + * the current level of support: At present, only memory protection is + * provided. Kernel-mode code may call into user-mode code, but not + * vice-versa. So hosting the memory manager in user-space allows the + * memory manager to be shared in both kernel- and user-mode spaces. + * + * In the longer run, if an MMU is support that can provide virtualized + * memory, then some SLAB memory manager will be required in kernel-space + * with some kind of brk() system call to obtain mapped heap space. + * + * In the current build model, the user-space module is built first. The + * file user_map.h is generated in the first pass and contains the + * addresses of the memory manager needed in this file: + */ + +#include + +/************************************************************************ + * Pre-processor definition + ************************************************************************/ + +/* This value is obtained from user_map.h */ + +#define KINITIALIZE(h,s) ((kminitialize_t)CONFIG_USER_MMINIT)(h,s) + +/************************************************************************ + * Private Types + ************************************************************************/ + +typedef void (*kminitialize_t)(FAR void*, size_t); + +/************************************************************************ + * Private Functions + ************************************************************************/ + +/************************************************************************ + * Public Functions + ************************************************************************/ + +/************************************************************************ + * Name: kmm_initialize + * + * Description: + * This is a simple redirection to the user-space mm_initialize() + * function. + * + * Parameters: + * heap_start - Address of the beginning of the (initial) memory region + * heap_size - The size (in bytes) if the (initial) memory region. + * + * Return Value: + * None + * + * Assumptions: + * 1. mm_initialize() resides in user-space + * 2. The address of the user space mm_initialize() is provided in + * user_map.h + * 3. The user-space mm_initialize() is callable from kernel-space. + * + ************************************************************************/ + +void kmm_initialize(FAR void *heap_start, size_t heap_size) +{ + return KINITIALIZE(heap_start, heap_size); +} + +#endif /* CONFIG_NUTTX_KERNEL && __KERNEL__ */ diff --git a/nuttx/mm/kmm_kfree.c b/nuttx/mm/kmm_kfree.c new file mode 100644 index 000000000..c4e31ebfa --- /dev/null +++ b/nuttx/mm/kmm_kfree.c @@ -0,0 +1,110 @@ +/************************************************************************ + * mm/kmm_kfree.c + * + * Copyright (C) 2011 Gregory Nutt. All rights reserved. + * Author: Gregory Nutt + * + * 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. Neither the name NuttX 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 COPYRIGHT HOLDERS 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 + * COPYRIGHT OWNER 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. + * + ************************************************************************/ + +/************************************************************************ + * Included Files + ************************************************************************/ + +#include +#include + +#if defined(CONFIG_NUTTX_KERNEL) && defined(__KERNEL__) + +/* This logic is all tentatively and, hopefully, will grow in usability. + * For now, the kernel-mode build uses the memory manager that is + * provided in the user-space build. That is awkward but reasonable for + * the current level of support: At present, only memory protection is + * provided. Kernel-mode code may call into user-mode code, but not + * vice-versa. So hosting the memory manager in user-space allows the + * memory manager to be shared in both kernel- and user-mode spaces. + * + * In the longer run, if an MMU is support that can provide virtualized + * memory, then some SLAB memory manager will be required in kernel-space + * with some kind of brk() system call to obtain mapped heap space. + * + * In the current build model, the user-space module is built first. The + * file user_map.h is generated in the first pass and contains the + * addresses of the memory manager needed in this file: + */ + +#include + +/************************************************************************ + * Pre-processor definition + ************************************************************************/ + +/* This value is obtained from user_map.h */ + +#define KFREE(p) ((kfree_t)CONFIG_USER_FREE)(p) + +/************************************************************************ + * Private Types + ************************************************************************/ + +typedef void (*kfree_t)(FAR void *); + +/************************************************************************ + * Private Functions + ************************************************************************/ + +/************************************************************************ + * Public Functions + ************************************************************************/ + +/************************************************************************ + * Name: kfree + * + * Description: + * This is a simple redirection to the user-space free() function. + * + * Parameters: + * None + * + * Return Value: + * None + * + * Assumptions: + * 1. free() resides in user-space + * 2. The address of the user space free() is provided in user_map.h + * 3. The user-space free() is callable from kernel-space. + * + ************************************************************************/ + +void kfree(FAR void *mem) +{ + return KFREE(mem); +} + +#endif /* CONFIG_NUTTX_KERNEL && __KERNEL__ */ diff --git a/nuttx/mm/kmm_kmalloc.c b/nuttx/mm/kmm_kmalloc.c new file mode 100644 index 000000000..e2c347342 --- /dev/null +++ b/nuttx/mm/kmm_kmalloc.c @@ -0,0 +1,110 @@ +/************************************************************************ + * mm/kmm_kmalloc.c + * + * Copyright (C) 2011 Gregory Nutt. All rights reserved. + * Author: Gregory Nutt + * + * 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. Neither the name NuttX 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 COPYRIGHT HOLDERS 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 + * COPYRIGHT OWNER 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. + * + ************************************************************************/ + +/************************************************************************ + * Included Files + ************************************************************************/ + +#include +#include + +#if defined(CONFIG_NUTTX_KERNEL) && defined(__KERNEL__) + +/* This logic is all tentatively and, hopefully, will grow in usability. + * For now, the kernel-mode build uses the memory manager that is + * provided in the user-space build. That is awkward but reasonable for + * the current level of support: At present, only memory protection is + * provided. Kernel-mode code may call into user-mode code, but not + * vice-versa. So hosting the memory manager in user-space allows the + * memory manager to be shared in both kernel- and user-mode spaces. + * + * In the longer run, if an MMU is support that can provide virtualized + * memory, then some SLAB memory manager will be required in kernel-space + * with some kind of brk() system call to obtain mapped heap space. + * + * In the current build model, the user-space module is built first. The + * file user_map.h is generated in the first pass and contains the + * addresses of the memory manager needed in this file: + */ + +#include + +/************************************************************************ + * Pre-processor definition + ************************************************************************/ + +/* This value is obtained from user_map.h */ + +#define KMALLOC(s) ((kmalloc_t)CONFIG_USER_MALLOC)(s) + +/************************************************************************ + * Private Types + ************************************************************************/ + +typedef FAR void *(*kmalloc_t)(size_t); + +/************************************************************************ + * Private Functions + ************************************************************************/ + +/************************************************************************ + * Public Functions + ************************************************************************/ + +/************************************************************************ + * Name: kmalloc + * + * Description: + * This is a simple redirection to the user-space malloc() function. + * + * Parameters: + * size - Size (in bytes) of the memory region to be allocated. + * + * Return Value: + * The address of the allocated memory (NULL on failure to allocate) + * + * Assumptions: + * 1. malloc() resides in user-space + * 2. The address of the user space malloc() is provided in user_map.h + * 3. The user-space malloc() is callable from kernel-space. + * + ************************************************************************/ + +FAR void *kmalloc(size_t size) +{ + return KMALLOC(size); +} + +#endif /* CONFIG_NUTTX_KERNEL && __KERNEL__ */ diff --git a/nuttx/mm/kmm_krealloc.c b/nuttx/mm/kmm_krealloc.c new file mode 100644 index 000000000..d5bebff38 --- /dev/null +++ b/nuttx/mm/kmm_krealloc.c @@ -0,0 +1,111 @@ +/************************************************************************ + * mm/kmm_krealloc.c + * + * Copyright (C) 2011 Gregory Nutt. All rights reserved. + * Author: Gregory Nutt + * + * 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. Neither the name NuttX 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 COPYRIGHT HOLDERS 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 + * COPYRIGHT OWNER 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. + * + ************************************************************************/ + +/************************************************************************ + * Included Files + ************************************************************************/ + +#include +#include + +#if defined(CONFIG_NUTTX_KERNEL) && defined(__KERNEL__) + +/* This logic is all tentatively and, hopefully, will grow in usability. + * For now, the kernel-mode build uses the memory manager that is + * provided in the user-space build. That is awkward but reasonable for + * the current level of support: At present, only memory protection is + * provided. Kernel-mode code may call into user-mode code, but not + * vice-versa. So hosting the memory manager in user-space allows the + * memory manager to be shared in both kernel- and user-mode spaces. + * + * In the longer run, if an MMU is support that can provide virtualized + * memory, then some SLAB memory manager will be required in kernel-space + * with some kind of brk() system call to obtain mapped heap space. + * + * In the current build model, the user-space module is built first. The + * file user_map.h is generated in the first pass and contains the + * addresses of the memory manager needed in this file: + */ + +#include + +/************************************************************************ + * Pre-processor definition + ************************************************************************/ + +/* This value is obtained from user_map.h */ + +#define KREALLOC(p,s) ((krealloc_t)CONFIG_USER_REALLOC)(p,s) + +/************************************************************************ + * Private Types + ************************************************************************/ + +typedef FAR void *(*krealloc_t)(FAR void*, size_t); + +/************************************************************************ + * Private Functions + ************************************************************************/ + +/************************************************************************ + * Public Functions + ************************************************************************/ + +/************************************************************************ + * Name: krealloc + * + * Description: + * This is a simple redirection to the user-space realloc() function. + * + * Parameters: + * oldmem - The old memory allocated + * size - Size (in bytes) of the new memory region to be re-allocated. + * + * Return Value: + * The address of the re-allocated memory (NULL on failure to re-allocate) + * + * Assumptions: + * 1. realloc() resides in user-space + * 2. The address of the user space realloc() is provided in user_map.h + * 3. The user-space realloc() is callable from kernel-space. + * + ************************************************************************/ + +FAR void *krealloc(FAR void *oldmem, size_t size) +{ + return KREALLOC(oldmem, size); +} + +#endif /* CONFIG_NUTTX_KERNEL && __KERNEL__ */ diff --git a/nuttx/mm/kmm_kzalloc.c b/nuttx/mm/kmm_kzalloc.c new file mode 100644 index 000000000..4775312c6 --- /dev/null +++ b/nuttx/mm/kmm_kzalloc.c @@ -0,0 +1,110 @@ +/************************************************************************ + * mm/kmm_kzalloc.c + * + * Copyright (C) 2011 Gregory Nutt. All rights reserved. + * Author: Gregory Nutt + * + * 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. Neither the name NuttX 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 COPYRIGHT HOLDERS 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 + * COPYRIGHT OWNER 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. + * + ************************************************************************/ + +/************************************************************************ + * Included Files + ************************************************************************/ + +#include +#include + +#if defined(CONFIG_NUTTX_KERNEL) && defined(__KERNEL__) + +/* This logic is all tentatively and, hopefully, will grow in usability. + * For now, the kernel-mode build uses the memory manager that is + * provided in the user-space build. That is awkward but reasonable for + * the current level of support: At present, only memory protection is + * provided. Kernel-mode code may call into user-mode code, but not + * vice-versa. So hosting the memory manager in user-space allows the + * memory manager to be shared in both kernel- and user-mode spaces. + * + * In the longer run, if an MMU is support that can provide virtualized + * memory, then some SLAB memory manager will be required in kernel-space + * with some kind of brk() system call to obtain mapped heap space. + * + * In the current build model, the user-space module is built first. The + * file user_map.h is generated in the first pass and contains the + * addresses of the memory manager needed in this file: + */ + +#include + +/************************************************************************ + * Pre-processor definition + ************************************************************************/ + +/* This value is obtained from user_map.h */ + +#define KZALLOC(s) ((kzalloc_t)CONFIG_USER_ZALLOC)(s) + +/************************************************************************ + * Private Types + ************************************************************************/ + +typedef FAR void *(*kzalloc_t)(size_t); + +/************************************************************************ + * Private Functions + ************************************************************************/ + +/************************************************************************ + * Public Functions + ************************************************************************/ + +/************************************************************************ + * Name: kzalloc + * + * Description: + * This is a simple redirection to the user-space zalloc() function. + * + * Parameters: + * size - Size (in bytes) of the memory region to be allocated. + * + * Return Value: + * The address of the allocated memory (NULL on failure to allocate) + * + * Assumptions: + * 1. zalloc() resides in user-space + * 2. The address of the user space zalloc() is provided in user_map.h + * 3. The user-space zalloc() is callable from kernel-space. + * + ************************************************************************/ + +FAR void *kzalloc(size_t size) +{ + return KZALLOC(size); +} + +#endif /* CONFIG_NUTTX_KERNEL && __KERNEL__ */ diff --git a/nuttx/mm/kmm_semaphore.c b/nuttx/mm/kmm_semaphore.c new file mode 100644 index 000000000..b391e5151 --- /dev/null +++ b/nuttx/mm/kmm_semaphore.c @@ -0,0 +1,140 @@ +/************************************************************************ + * mm/kmm_semaphore.c + * + * Copyright (C) 2011 Gregory Nutt. All rights reserved. + * Author: Gregory Nutt + * + * 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. Neither the name NuttX 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 COPYRIGHT HOLDERS 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 + * COPYRIGHT OWNER 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. + * + ************************************************************************/ + +/************************************************************************ + * Included Files + ************************************************************************/ + +#include +#include + +#if defined(CONFIG_NUTTX_KERNEL) && defined(__KERNEL__) + +/* This logic is all tentatively and, hopefully, will grow in usability. + * For now, the kernel-mode build uses the memory manager that is + * provided in the user-space build. That is awkward but reasonable for + * the current level of support: At present, only memory protection is + * provided. Kernel-mode code may call into user-mode code, but not + * vice-versa. So hosting the memory manager in user-space allows the + * memory manager to be shared in both kernel- and user-mode spaces. + * + * In the longer run, if an MMU is support that can provide virtualized + * memory, then some SLAB memory manager will be required in kernel-space + * with some kind of brk() system call to obtain mapped heap space. + * + * In the current build model, the user-space module is built first. The + * file user_map.h is generated in the first pass and contains the + * addresses of the memory manager needed in this file: + */ + +#include + +/************************************************************************ + * Pre-processor definition + ************************************************************************/ + +/* These values are obtained from user_map.h */ + +#define KTRYSEMAPHORE() ((kmtrysemaphore_t) CONFIG_USER_MMTRYSEM )() +#define KGIVESEMAPHORE() ((kmgivesemaphore_t)CONFIG_USER_MMGIVESEM)() + +/************************************************************************ + * Private Types + ************************************************************************/ + +typedef int (*kmtrysemaphore_t)(void); +typedef void (*kmgivesemaphore_t)(void); + +/************************************************************************ + * Private Functions + ************************************************************************/ + +/************************************************************************ + * Public Functions + ************************************************************************/ + +/************************************************************************ + * Name: kmm_trysemaphore + * + * Description: + * This is a simple redirection to the user-space mm_trysemaphore() + * function. + * + * Parameters: + * None + * + * Return Value: + * OK on success; a negated errno on failure + * + * Assumptions: + * 1. mm_trysemaphore() resides in user-space + * 2. The address of the user space mm_trysemaphore() is provided in + * user_map.h + * 3. The user-space mm_semaphore() is callable from kernel-space. + * + ************************************************************************/ + +int kmm_trysemaphore(void) +{ + return KTRYSEMAPHORE(); +} + +/************************************************************************ + * Name: kmm_givesemaphore + * + * Description: + * This is a simple redirection to the user-space mm_givesemaphore() + * function. + * + * Parameters: + * None + * + * Return Value: + * OK on success; a negated errno on failure + * + * Assumptions: + * 1. mm_givesemaphore() resides in user-space + * 2. The address of the user space mm_givesemaphore() is provided in + * user_map.h + * 3. The user-space mm_semaphore() is callable from kernel-space. + * + ************************************************************************/ + +void kmm_givesemaphore(void) +{ + KGIVESEMAPHORE(); +} + +#endif /* CONFIG_NUTTX_KERNEL && __KERNEL__ */ diff --git a/nuttx/mm/mm_initialize.c b/nuttx/mm/mm_initialize.c index 37df3ab5f..8930f6190 100644 --- a/nuttx/mm/mm_initialize.c +++ b/nuttx/mm/mm_initialize.c @@ -65,7 +65,11 @@ struct mm_heap_s g_mmheap; ****************************************************************************/ /**************************************************************************** - * Name: _mm_addregion + * Public Functions + ****************************************************************************/ + +/**************************************************************************** + * Name: mm_addregion * * Description: * This function adds a region of contiguous memory to the selected heap. @@ -82,8 +86,8 @@ struct mm_heap_s g_mmheap; * ****************************************************************************/ -static inline void _mm_addregion(FAR struct mm_heap_s *heap, - FAR void *heapstart, size_t heapsize) +void mm_addregion(FAR struct mm_heap_s *heap, FAR void *heapstart, + size_t heapsize) { FAR struct mm_freenode_s *node; uintptr_t heapbase; @@ -149,7 +153,7 @@ static inline void _mm_addregion(FAR struct mm_heap_s *heap, } /**************************************************************************** - * Name: _mm_initialize + * Name: mm_initialize * * Description: * Initialize the selected heap data structures, providing the initial @@ -167,8 +171,8 @@ static inline void _mm_addregion(FAR struct mm_heap_s *heap, * ****************************************************************************/ -static inline void _mm_initialize(FAR struct mm_heap_s *heap, - FAR void *heapstart, size_t heapsize) +void mm_initialize(FAR struct mm_heap_s *heap, FAR void *heapstart, + size_t heapsize) { int i; @@ -209,62 +213,5 @@ static inline void _mm_initialize(FAR struct mm_heap_s *heap, /* Add the initial region of memory to the heap */ - _mm_addregion(heap, heapstart, heapsize); -} - -/**************************************************************************** - * Public Functions - ****************************************************************************/ - -/**************************************************************************** - * Name: mm_initialize - * - * Description: - * This this function is called during initialization to initialize the - * user heap. - * - * Parameters: - * heapstart - Start of the initial heap region - * heapsize - Size of the initial heap region - * - * Return Value: - * None - * - * Assumptions: - * - ****************************************************************************/ - -#if !defined(CONFIG_NUTTX_KERNEL) || !defined(__KERNEL__) - -void mm_initialize(FAR void *heapstart, size_t heapsize) -{ - _mm_initialize(&g_mmheap, heapstart, heapsize); + mm_addregion(heap, heapstart, heapsize); } - -#endif - -/**************************************************************************** - * Name: mm_addregion - * - * Description: - * This function adds a region of contiguous memory to the user heap. - * - * Parameters: - * heapstart - Start of the heap region - * heapsize - Size of the heap region - * - * Return Value: - * None - * - * Assumptions: - * - ****************************************************************************/ - -#if !defined(CONFIG_NUTTX_KERNEL) || !defined(__KERNEL__) - -void mm_addregion(FAR void *heapstart, size_t heapsize) -{ - _mm_addregion(&g_mmheap, heapstart, heapsize); -} - -#endif -- cgit v1.2.3