NOTICE: The Processors Wiki will End-of-Life in December of 2020. It is recommended to download any files or other content you may need that are hosted on processors.wiki.ti.com. The site is now set to read only.
- 1 Introduction
- 2 Development
- 3 Configuration
- 4 Debugging Techniques
- 5 General Purpose Heaps
- 6 Application Cleanup
- 7 Linux CMA Support
- 8 Android CMA Support
- 9 CMEM FAQ
- 10 Licensing
CMEM is an API (Reference Guide) and library for managing one or more blocks of physically contiguous memory. It also provides address translation services (e.g. virtual to physical translation) and user-mode cache management APIs. This physically contiguous memory is useful as data buffers that will be shared with another processor (e.g. for the DSP on an OMAP3) or a hardware accelerator/DMA (e.g. used by codecs on a DM365)
Using its pool-based configuration, CMEM enables users to avoid memory fragmentation, and ensures large physically contiguous memory blocks are available even after a system has been running for very long periods of time.
It was originally developed for the DM644x, and has been ported to several Operating Systems (e.g. Linux, WinCE, QNX, Nucleus, Green Hills Integrity, and others). Although generally associated with Codec Engine, it has no dependency on Codec Engine and can be used on its own.
CMEM is a component of Linux Utils, and is actively being developed in the publicly maintained, TI-hosted 'ludev' git repository - http://git.ti.com/ipc/ludev. The Linux Utils development process is documented here, patches are welcome!
Linux configuration is done when installing the cmemk.ko driver, typically done using the insmod command. The cmemk.ko driver accepts command line parameters for configuring the physical memory to reserve and how to carve it up.
The following is an example of installing the cmem kernel module:
/sbin/insmod cmemk.ko pools=4x30000,2x500000 phys_start=0x0 phys_end=0x3000000
- phys_start and phys_end must be specified in hexadecimal format
- pools must be specified using decimal format (for both number and size), since using hexadecimal format would visually clutter the specification due to the use of "x" as a token separator
This particular command creates 2 pools. The first pool is created with 4 buffers of size 30000 bytes and the second pool is created with 2 buffers of size 500000 bytes. The CMEM pool buffers start at 0x0 and end at 0x3000000 (max).
Pool buffers are aligned on a module-dependent boundary, and their sizes are rounded up to this same boundary. This applies to each buffer within a pool. The total space used by an individual pool will therefore be greater than (or equal to) the exact amount requested in the installation of the module.
The poolid used in the driver calls would be 0 for the first pool and 1 for the second pool.
Pool allocations can be requested explicitly by pool number, or more generally by just a size. For size-based allocations, the pool which best fits the requested size is automatically chosen.
Configuration of CMEM in WinCE-based environments is typically done via the registry and/or statically built into the driver (for closed systems). Here is an example for a line to be added to the MEMORY section of 'config.bib' of your BSP:
CMEM_DSP 89000000 02800000 RESERVED ; 40 MB
That reserves 40MB of memory for CMEM, DSPLINK, DSP code as well as DSP heap usage starting at virtual address 0x89000000. There is no distinction here between the different modules memory usage. Obviously all of them need to be configured accordingly. Registry settings for CMEM use physical start and end addresses for any defined block of pools.
Here is an example CMEM configuration registry entry in platform.reg for TI EVM3530:
- -- CMEM --------------------------------------------------------------------
IF SYSGEN_CMEM [HKEY_LOCAL_MACHINE\Drivers\BuiltIn\CMEMK]
"Prefix"="CMK" "Dll"="cmemk.dll" "Index"=dword:1 ; Make 7 pools available for allocation for block 0 ; Make 1 pool available for allocation for block 1 "NumPools0"=dword:7 "NumPools1"=dword:0
"Block0_NumBuffers_Pool0"=dword:20 "Block0_PoolSize_Pool0"=dword:1000 ; size in bytes (hex) "Block0_NumBuffers_Pool1"=dword:8 "Block0_PoolSize_Pool1"=dword:20000 ; size in bytes (hex) "Block0_NumBuffers_Pool2"=dword:5 "Block0_PoolSize_Pool2"=dword:100000 ; size in bytes (hex)
"Block0_NumBuffers_Pool3"=dword:1 "Block0_PoolSize_Pool3"=dword:15cfc0 ; size in bytes (hex) "Block0_NumBuffers_Pool4"=dword:1 "Block0_PoolSize_Pool4"=dword:3e800 ; size in bytes (hex) "Block0_NumBuffers_Pool5"=dword:1 "Block0_PoolSize_Pool5"=dword:36ee80 ; size in bytes (hex)
"Block0_NumBuffers_Pool6"=dword:3 "Block0_PoolSize_Pool6"=dword:96000 ; size in bytes (hex)
;; "Block1_NumBuffers_Pool1"=dword:2 ;; "Block1_PoolSize_Pool1"=dword:4000 ; size in bytes (hex)
; Physical start + physical end can be use to ask CMEM to map a specific ; range of physical addresses. ; This is a potential security risk. If physical start == 0 then the code ; hits a special case. ; physical end - physical start == length of allocation. In the special ; case, memory is allocated via a call to AllocPhysMem() (as shown in ; this example). MmMapIoSpace() is used to map the normal case where ; physical start != 0. ; ; physical start and end for block 0 "PhysicalStart0"=dword:85000000 "PhysicalEnd0"=dword:86000000 ; physical start and end for block 1 "PhysicalStart1"=dword:0 "PhysicalEnd1"=dword:0
The CMEM driver information must also be added to the platform.bib file (or some other .bib file that gets put into ce.bib). Here is an example of the CMEM driver entry in platform.bib:
;-- CMEM ---------------------------------------------------------------------- IF SYSGEN_CMEM cmemk.dll $(_FLATRELEASEDIR)\cmemk.dll NK SHK ENDIF BSP_CMEM ;------------------------------------------------------------------------------
Linux users can execute "cat /proc/cmem" to get status on the buffers and pools managed by CMEM.
There is also a debug library provided that provides tracing diagnostics during execution. XDC Config users can link in this library by adding the following to their application's config script:
General Purpose Heaps
In CMEM 2.00, CMEM added support for a general purpose heap. Using the example above, in addition to the 2 pools, a general purpose heap block is created from which allocations of any size can be requested. Internally, allocation sizes are rounded up to a module-dependent boundary and allocation addresses are aligned either to this same boundary or to the requested alignment (whichever is greater).
The size of the heap block is the amount of CMEM memory remaining after all pool allocations. If more heap space is needed than is available after pool allocations, you must reduce the amount of CMEM memory granted to the pools.
The main disadvantage to using heap(s) over pools is fragmentation. After several sequences of codec creation/deletion, in different orders, with possibly different create() params, you may end up fragmenting your heap and being unable to acquire a requested memory block - possibly resulting in a codec creation failure.
Typically, during development, users will use CMEM with heap-based memory, as heap usage requires very little configuration, and users don't know how to configure pool memory(!). In a production system, however, it's strongly recommended that pool configuration be used to avoid memory fragmentation and confusing end user errors.
CMEM 2.23 introduced a facility to clean up unfreed buffers when an application exits, either prematurely or in a normal fashion. This facility is achieved by maintaining an "ownership" list for each allocated buffer that is inspected upon closing a device driver instance. During this inspection all allocated buffers are checked, and when it is determined that the closing process is on the ownership list of an allocated buffer, the process is removed from the list. If this causes the list to become empty the associated buffer is actually freed, otherwise it is maintained in the allocated state on behalf of other owners. A side-effect of this model is that only a buffer "owner" is allowed to free the buffer.
In order to facilitate multiple owners of an allocated buffer, a new set of APIs was introduced: <syntaxhighlight lang="c"> void *CMEM_registerAlloc(unsigned long physp); int CMEM_unregister(void *ptr, CMEM_AllocParams *params); </syntaxhighlight>
CMEM_registerAlloc() takes a buffer physical address as input (achieved through
CMEM_getPhys()) and returns a fresh virtual address that is mapped to that buffer, while also adding the calling process to the ownership list.
CMEM_unregister() is equivalent to
CMEM_free() and releases ownership of the buffer (as well as freeing it if all owners have released the buffer).
In CMEM 2.24, ownership is established on a per-process (and per-thread) basis. This detail becomes important when using CMEM in multiple threads of a given process - if one thread allocates a CMEM buffer and a separate thread of the same process is responsible for freeing that buffer, the "freeing" thread will not be allowed to free the buffer since it is not on the ownership list.
CMEM 2.24.01 changes the ownership policy to be based on the calling process' file descriptor instead of the calling process' process descriptor. This facilitates thread-based sharing of buffers, allowing any thread within a process to free a buffer that was allocated by a different thread within the same process, since threads within a process all use the same file descriptor.
Linux CMA Support
CMEM 4.00 added the ability to leverage the Linux kernel's CMA feature. CMA supports a "global" memory pool, as well as device-specific memory - CMEM provides the facilities to allocate from either type of CMA pool.
Note that using CMEM to allocate from CMA-based memory is an additional feature. You can continue to use CMEM to manage memory carveouts as well.
Android CMA Support
Build Environment Setup
First download an unzip the latest Linux utils(4.00.01.08) zip file. The file products.mak (at the top level of this tree) contains two definitions used by the build subsystem:
KERNEL_INSTALL_DIR - The base directory of your Linux kernel source tree TOOLCHAIN_PREFIX - the 'prefix' for the GNU ARM codegen tools
The TOOLCHAIN_PREFIX can contain the full path of the codegen tools, ending with the tool prefix, i.e.:
or it can be just the tool prefix if your shell's $PATH contains your codegen's 'bin' directory:
where your $PATH contains:
For example, below is the setup environment which is validated
TOOLCHAIN_LONGNAME = arm-eabi TOOLCHAIN_INSTALL_DIR = /home/(user)/mydroid/prebuilts/gcc/linux-x86/arm/arm-eabi-4.7 KERNEL_INSTALL_DIR =/home/(user)/kernel/android-3.8
Now move to the src/cmem/module directory to run "make clean" and then "make".
Building Test Binaries
From the downloaded and installed linux utils base directory run the below commands,
Note: Any non-android toolchain should work and don't forget to export the toolchain path(until the bin folder) to PATH environment variable.
export ARCH=arm export CROSS_COMPILE=arm-linux-gnueabihf ./configure --disable-shared --host=arm-linux-gnueabihf --prefix=$PWD CFLAGS='--static'
Now run "make clean" and "make" to build the test binaries for android
Test Setup and Validation Process
For testing purpose we built the android kernel for mem=1200M.
Boot the system with android and then do adb push on the below mentioned files,
(linux utils base directory)/src/cmem/module/cmemk.ko to /system/lib/modules (linux utils base directory)/src/cmem/tests/apitest to /system/bin (linux utils base directory)/src/cmem/tests/multi_process to /system/bin (linux utils base directory)/src/cmem/tests/translate to /system/bin
The loadable kernel module 'cmemk.ko' can be installed into any running system. Out of the 3 tests mentioned below Multi_Process & Translate tests have been used to validate the CMEM module's usage of OCMC1 ram. OCMC1 ram range is 0x40300000 ~ 0x4033FFFF.
Multi Process Test
This app tries to use CMEM from multiple processes. It takes the number of processes to start as a parameter. Now load the kernel module 'cmemk.ko' with the below command:
% insmod cmemk.ko phys_start=0xcaf01000 phys_end=0xCB601000 pools=4x1000 phys_start_1=0xCB601000 phys_end_1=0xCB701000 pools_1=4x1000
% insmod cmemk.ko phys_start=0x40300000 phys_end=0x4033FFFF pools=4x500 phys_start_1=0x4033FFFF phys_end_1=0x4037ffff pools_1=4x500 allowOverlap=1
(Uses OCMC1, for this rebuild the Translate Test app with macro BUFFER_SIZE = 500 at line #49 in file (linuxutils)/src/cmem/tests/multi_process.c) Now run the Multi Process test,
% multi_process 3
where 3 is the number of processes to be spawned.
This app tests the address translation. Now load the kernel module 'cmemk.ko' with the below command:
% insmod cmemk.ko phys_start=0xcaf01000 phys_end=0xCB601000 pools=1x3145728
% insmod cmemk.ko phys_start=0x40300000 phys_end=0x4037ffff pools=1x20000 allowOverlap=1
(Uses OCMC1, for this rebuild the Translate Test app with macro BUFSIZE = 20000 at line #48 in file (linuxutils)/src/cmem/tests/translate.c) Now run the Translate test,
Tests basic API usage and memory allocation. This particular test has a limitation as it runs successfully only on kernel built with mem=120M. Now load the kernel module 'cmemk.ko' with the below command:
% insmod cmemk.ko phys_start=0x87800000 phys_end=0x87F00000 pools=4xBUFSIZE phys_start_1=0x87F00000 phys_end_1=0x88000000 pools_1=4xBUFSIZE
where BUFSIZE is the number of bytes you plan on passing as command line parameter to apitest. If in doubt, use a larger number as BUFSIZE denotes the maximum buffer you can allocate.Now run the Translate test, Now run the API test,
% apitest <BUFSIZE>
(e.g) With BUFSIZE=10240
% apitest 10240
Q: Why am I'm getting this error when loading the CMEM (or other!) driver: "insmod: error inserting 'cmemk.ko': -1 Invalid module format"?
- A: This error indicates the CMEM kernel module was built with a different Linux kernel version than the version running on the target. You need to rebuild CMEM against the kernel running on your target.
CMEM_getPhys() be used to translate any virtual address to its physical address?
- A: In theory, "yes". However, sometime after Linux version 2.6.10 the CMEM kernel module
get_phys()function stopped working for kernel addresses. A new
get_phys()was provided to work with newer kernels, but it was discovered that this new one didn't correctly translate non-direct-mapped kernel addresses, so code was added to CMEM to save the lower/upper bounds of the CMEM blocks' kernel addresses, and manually look for those in
get_phys()before trying more general methods of translation.
- So, in short, CMEM's
get_phys()doesn't handle non-direct-mapped kernel addresses except the ones that correspond to CMEM's managed memory block(s).
Q: How does CMEM relate to DSPLink's POOL feature?
- A: Though they provide overlapping features, they are independent, and each has unique features.
- CMEM can be used on systems without a remote DSP slave (e.g. DM365 codecs require physically contiguous memory when using HW accelerators)
- CMEM buffers can be cached
- CMEM blocks support fixed size pools (no fragmentation) as well as heaps (easier to use)
- CMEM configuration doesn't require a rebuild (they're provided as insmod params)
- POOL buffers can be allocated on one processor and freed on another
Q: In Linux, how do I set aside the memory carveout that CMEM uses?
- A: The memory carveout used by CMEM must not be in use by Linux else an error will occur during module loading (i.e., insmod/modprobe). There are two simple methods for defining CMEM's memory carveout:
- 1) kernel command line
- This method involves the kernel command line issued from u-boot. When booting Linux, one may restrict the memory available to Linux by specifying physical memory blocks for Linux to use:
- mem=128M@0x80000000 mem=256M@0x90000000
- which grants the memory at 0x80000000 -> 0x88000000 and 0x90000000 -> 0xa0000000 to Linux, leaving the CMEM memory carveout as 128MB at 0x88000000 (0x88000000 -> 0x90000000). Without a "mem=" entry on the command line, Linux will use all available memory.
- 2) removal via machine's "
- This method involves modifying a machine's
.reservefunction to remove a block of memory from Linux. For example, for the Vayu architecture, the file arch/arm/mach-omap2/common.c contains a function named
dra7_reserve()which is assigned to the machine
.reservefunction in arch/arm/mach-omap2/board-generic.c. Adding the following C statement to
dra7_reserve()accomplishes the same memory carveout as specified in 1) above:
- The CMEM memory carveout can either precede, overlap, or succeed the Linux memory. For the case where it precedes or overlaps, don't forget to specify "allowOverlap=1" on the cmemk.ko insmod/modprobe command, else the module loading will fail.
- For both cases above, you would load cmemk.ko as follows:
- % modprobe cmemk.ko phys_start=0x88000000 phys_end=0x90000000 allowOverlap=1 pools=...
- The advantage for method 1) is that the CMEM memory carveout can be specified to be anywhere by the system integrator without changing the kernel, with a disadvantage of having to document this carveout specification along with potential error in doing so. The advantage of method 2) is that a given kernel image will always properly create the carveout for CMEM without any intervention by the system integrator, with a disadvantage of not being moveable without changing/rebuilding the kernel.
Q: Why CMEM failed in physical address > 32bits?
- A: The user space application need to be compiled with “–D_FILE_OFFSET_BITS=64" to allow physical addresses > 32 bits.
In CMEM 2.00, the CMEM Linux release is LGPL v2 for the user mode lib and GPL v2 for the kernel mode driver.
In CMEM 2.21, the Linux user mode library licensing changed from LGPL to BSD. The Linux kernel mode driver continued to be GPL v2.