* Contiguous Memory Allocator
The Contiguous Memory Allocator (CMA) is a framework, which allows
setting up a machine-specific configuration for physically-contiguous
memory management. Memory for devices is then allocated according
to that configuration.
The main role of the framework is not to allocate memory, but to
parse and manage memory configurations, as well as to act as an
in-between between device drivers and pluggable allocators. It is
thus not tied to any memory allocation method or strategy.
** Why is it needed?
Various devices on embedded systems have no scatter-getter and/or
IO map support and as such require contiguous blocks of memory to
operate. They include devices such as cameras, hardware video
decoders and encoders, etc.
Such devices often require big memory buffers (a full HD frame is,
for instance, more then 2 mega pixels large, i.e. more than 6 MB
of memory), which makes mechanisms such as kmalloc() ineffective.
Some embedded devices impose additional requirements on the
buffers, e.g. they can operate only on buffers allocated in
particular location/memory bank (if system has more than one
memory bank) or buffers aligned to a particular memory boundary.
Development of embedded devices have seen a big rise recently
(especially in the V4L area) and many such drivers include their
own memory allocation code. Most of them use bootmem-based methods.
CMA framework is an attempt to unify contiguous memory allocation
mechanisms and provide a simple API for device drivers, while
staying as customisable and modular as possible.
The main design goal for the CMA was to provide a customisable and
modular framework, which could be configured to suit the needs of
individual systems. Configuration specifies a list of memory
regions, which then are assigned to devices. Memory regions can
be shared among many device drivers or assigned exclusively to
one. This has been achieved in the following ways:
1. The core of the CMA does not handle allocation of memory and
management of free space. Dedicated allocators are used for
This way, if the provided solution does not match demands
imposed on a given system, one can develop a new algorithm and
easily plug it into the CMA framework.
The presented solution includes an implementation of a best-fit
2. CMA allows a run-time configuration of the memory regions it
will use to allocate chunks of memory from. The set of memory
regions is given on command line so it can be easily changed
without the need for recompiling the kernel.
Each region has it's own size, alignment demand, a start
address (physical address where it should be placed) and an
allocator algorithm assigned to the region.
This means that there can be different algorithms running at
the same time, if different devices on the platform have
distinct memory usage characteristics and different algorithm
match those the best way.
3. When requesting memory, devices have to introduce themselves.
This way CMA knows who the memory is allocated for. This
allows for the system architect to specify which memory regions
each device should use.
3a. Devices can also specify a "kind" of memory they want.
This makes it possible to configure the system in such
a way, that a single device may get memory from different
memory regions, depending on the "kind" of memory it
requested. For example, a video codec driver might want to
allocate some shared buffers from the first memory bank and
the other from the second to get the highest possible
** Use cases
Lets analyse some imaginary system that uses the CMA to see how
the framework can be used and configured.
We have a platform with a hardware video decoder and a camera each
needing 20 MiB of memory in worst case. Our system is written in
such a way though that the two devices are never used at the same
time and memory for them may be shared. In such a system the
following two command line arguments would be used:
The first instructs CMA to allocate a region of 20 MiB and use the
first available memory allocator on it. The second, that drivers
named "video" and "camera" are to be granted memory from the
previously defined region.
We can see, that because the devices share the same region of
memory, we save 20 MiB of memory, compared to the situation when
each of the devices would reserve 20 MiB of memory for itself.
However, after some development of the system, it can now run
video decoder and camera at the same time. The 20 MiB region is
no longer enough for the two to share. A quick fix can be made to
grant each of those devices separate regions:
This solution also shows how with CMA you can assign private pools
of memory to each device if that is required.
Allocation mechanisms can be replaced dynamically in a similar
manner as well. Let's say that during testing, it has been
discovered that, for a given shared region of 40 MiB,
fragmentation has become a problem. It has been observed that,
after some time, it becomes impossible to allocate buffers of the
required sizes. So to satisfy our requirements, we would have to
reserve a larger shared region beforehand.
But fortunately, you have also managed to develop a new allocation
algorithm -- Neat Allocation Algorithm or "na" for short -- which
satisfies the needs for both devices even on a 30 MiB region. The
configuration can be then quickly changed to:
This shows how you can develop your own allocation algorithms if
the ones provided with CMA do not suit your needs and easily
replace them, without the need to modify CMA core or even
recompiling the kernel.
** Technical Details
*** The command line parameters
As shown above, CMA is configured from command line via two
arguments: "cma" and "cma_map". The first one specifies regions
that are to be reserved for CMA. The second one specifies what
regions each device is assigned to.
The format of the "cma" parameter is as follows:
cma ::= "cma=" regions [ ';' ]
regions ::= region [ ';' regions ]
region ::= reg-name
[ '@' start ]
[ '/' alignment ]
[ ':' [ alloc-name ] [ '(' alloc-params ')' ] ]
reg-name ::= a sequence of letters and digits
// name of the region
size ::= memsize // size of the region
start ::= memsize // desired start address of
// the region
alignment ::= memsize // alignment of the start
// address of the region
alloc-name ::= a non-empty sequence of letters and digits
// name of an allocator that will be used
// with the region
alloc-params ::= a sequence of chars other then ')' and ';'
// optional parameters for the allocator
memsize ::= whatever memparse() accepts
The format of the "cma_map" parameter is as follows:
cma-map ::= "cma_map=" rules [ ';' ]
rules ::= rule [ ';' rules ]
rule ::= patterns '=' regions
patterns ::= pattern [ ',' patterns ]
regions ::= reg-name [ ',' regions ]
// list of regions to try to allocate memory
// from for devices that match pattern
pattern ::= dev-pattern [ '/' kind-pattern ]
| '/' kind-pattern
// pattern request must match for this rule to
// apply to it; the first rule that matches is
// applied; if dev-pattern part is omitted
// value identical to the one used in previous
// pattern is assumed
dev-pattern ::= pattern-str
// pattern that device name must match for the
// rule to apply.
kind-pattern ::= pattern-str
// pattern that "kind" of memory (provided by
// device) must match for the rule to apply.
pattern-str ::= a non-empty sequence of characters with '?'
meaning any character and possible '*' at
the end meaning to match the rest of the
Some examples (whitespace added for better readability):
cma = r1 = 64M // 64M region
@512M // starting at address 512M
// (or at least as near as possible)
/1M // make sure it's aligned to 1M
:foo(bar); // uses allocator "foo" with "bar"
// as parameters for it
r2 = 64M // 64M region
/1M; // make sure it's aligned to 1M
// uses the first available allocator
r3 = 64M // 64M region
@512M // starting at address 512M
:foo; // uses allocator "foo" with no parameters
cma_map = foo = r1;
// device foo with kind==NULL uses region r1
foo/quaz = r2; // OR:
/quaz = r2;
// device foo with kind == "quaz" uses region r2
foo/* = r3; // OR:
/* = r3;
// device foo with any other kind uses region r3
bar/* = r1,r2;
// device bar with any kind uses region r1 or r2
baz?/a* , baz?/b* = r3;
// devices named baz? where ? is any character
// with kind being a string starting with "a" or
// "b" use r3
*** The device and kind of memory
The name of the device is taken form the device structure. It is
not possible to use CMA if driver does not register a device
(actually this can be overcome if a fake device structure is
provided with at least the name set).
The kind of memory is an optional argument provided by the device
whenever it requests memory chunk. In many cases this can be
ignored but sometimes it may be required for some devices.
For instance, let say that there are two memory banks and for
performance reasons a device uses buffers in both of them. In
such case, the device driver would define two kinds and use it for
different buffers. Command line arguments could look as follows:
And whenever the driver allocated the memory it would specify the
kind of memory:
buffer1 = cma_alloc(dev, 1 << 20, 0, "a");
buffer2 = cma_alloc(dev, 1 << 20, 0, "b");
If it was needed to try to allocate from the other bank as well if
the dedicated one is full command line arguments could be changed
On the other hand, if the same driver was used on a system with
only one bank, the command line could be changed to:
without the need to change the driver at all.
There are four calls provided by the CMA framework to devices. To
allocate a chunk of memory cma_alloc() function needs to be used:
unsigned long cma_alloc(const struct device *dev,
const char *kind,
unsigned long size,
unsigned long alignment);
If required, device may specify alignment that the chunk need to
satisfy. It have to be a power of two or zero. The chunks are
always aligned at least to a page.
The kind specifies the kind of memory as described to in the
previous subsection. If device driver does not use notion of
memory kinds it's safe to pass NULL as the kind.
The basic usage of the function is just a:
addr = cma_alloc(dev, NULL, size, 0);
The function returns physical address of allocated chunk or
a value that evaluated true if checked with IS_ERR_VALUE(), so the
correct way for checking for errors is:
unsigned long addr = cma_alloc(dev, size);
/* Allocated */
(Make sure to include <linux/err.h> which contains the definition
of the IS_ERR_VALUE() macro.)
Allocated chunk is freed via a cma_put() function:
int cma_put(unsigned long addr);
It takes physical address of the chunk as an argument and
decreases it's reference counter. If the counter reaches zero the
chunk is freed. Most of the time users do not need to think about
reference counter and simply use the cma_put() as a free call.
If one, however, were to share a chunk with others built in
reference counter may turn out to be handy. To increment it, one
needs to use cma_get() function:
int cma_put(unsigned long addr);
The last function is the cma_info() which returns information
about regions assigned to given (dev, kind) pair. Its syntax is:
int cma_info(struct cma_info *info,
const struct device *dev,
const char *kind);
On successful exit it fills the info structure with lower and
upper bound of regions, total size and number of regions assigned
to given (dev, kind) pair.
*** Allocator operations
Creating an allocator for CMA needs four functions to be
The first two are used to initialise an allocator far given driver
and clean up afterwards:
int cma_foo_init(struct cma_region *reg);
void cma_foo_done(struct cma_region *reg);
The first is called during platform initialisation. The
cma_region structure has saved starting address of the region as
well as its size. It has also alloc_params field with optional
parameters passed via command line (allocator is free to interpret
those in any way it pleases). Any data that allocate associated
with the region can be saved in private_data field.
The second call cleans up and frees all resources the allocator
has allocated for the region. The function can assume that all
chunks allocated form this region have been freed thus the whole
region is free.
The two other calls are used for allocating and freeing chunks.
struct cma_chunk *cma_foo_alloc(struct cma_region *reg,
unsigned long size,
unsigned long alignment);
void cma_foo_free(struct cma_chunk *chunk);
As names imply the first allocates a chunk and the other frees
a chunk of memory. It also manages a cma_chunk object
representing the chunk in physical memory.
Either of those function can assume that they are the only thread
accessing the region. Therefore, allocator does not need to worry
When allocator is ready, all that is left is register it by adding
a line to "mm/cma-allocators.h" file:
The first "foo" is a named that will be available to use with
command line argument. The second is the part used in function
*** Integration with platform
There is one function that needs to be called form platform
initialisation code. That is the cma_regions_allocate() function:
void cma_regions_allocate(int (*alloc)(struct cma_region *reg));
It traverses list of all of the regions given on command line and
reserves memory for them. The only argument is a callback
function used to reserve the region. Passing NULL as the argument
makes the function use cma_region_alloc() function which uses
bootmem for allocating.
Alternatively, platform code could traverse the cma_regions array
by itself but this should not be necessary.
The If cma_region_alloc() allocator is used, the
cma_regions_allocate() function needs to be allocated when bootmem
Platform has also a way of providing default cma and cma_map
parameters. cma_defaults() function is used for that purpose:
int cma_defaults(const char *cma, const char *cma_map)
It needs to be called after early params have been parsed but
prior to allocating regions. Arguments of this function are used
only if they are not-NULL and respective command line argument was
** Future work
In the future, implementation of mechanisms that would allow the
free space inside the regions to be used as page cache, filesystem
buffers or swap devices is planned. With such mechanisms, the
memory would not be wasted when not used.
Because all allocations and freeing of chunks pass the CMA
framework it can follow what parts of the reserved memory are
freed and what parts are allocated. Tracking the unused memory
would let CMA use it for other purposes such as page cache, I/O
buffers, swap, etc.
to post comments)