Then, the main() function is called. The core initialization and launch is done in rte_eal_init() (see the API documentation).
It consist of calls to the pthread library (more specifically, pthread_self(), pthread_create(), and pthread_setaffinity_np()).
-.. _pg_figure_2:
+.. _figure_linuxapp_launch:
-**Figure 2. EAL Initialization in a Linux Application Environment**
+.. figure:: img/linuxapp_launch.*
-.. image3_png has been replaced
+ EAL Initialization in a Linux Application Environment
-|linuxapp_launch|
.. note::
~~~~~~~~~~~~~~~~~~~~~
The Linuxapp EAL allows a multi-process as well as a multi-threaded (pthread) deployment model.
-See chapter 2.20
+See chapter
:ref:`Multi-process Support <Multi-process_Support>` for more details.
Memory Mapping Discovery and Memory Reservation
.. note::
- Memory reservations done using the APIs provided by the rte_malloc library are also backed by pages from the hugetlbfs filesystem.
- However, physical address information is not available for the blocks of memory allocated in this way.
-
-Xen Dom0 support without hugetbls
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-
-The existing memory management implementation is based on the Linux kernel hugepage mechanism.
-However, Xen Dom0 does not support hugepages, so a new Linux kernel module rte_dom0_mm is added to workaround this limitation.
-
-The EAL uses IOCTL interface to notify the Linux kernel module rte_dom0_mm to allocate memory of specified size,
-and get all memory segments information from the module,
-and the EAL uses MMAP interface to map the allocated memory.
-For each memory segment, the physical addresses are contiguous within it but actual hardware addresses are contiguous within 2MB.
+ Memory reservations done using the APIs provided by rte_malloc are also backed by pages from the hugetlbfs filesystem.
PCI Access
~~~~~~~~~~
CPU Feature Identification
~~~~~~~~~~~~~~~~~~~~~~~~~~
-The EAL can query the CPU at runtime (using the rte_cpu_get_feature() function) to determine which CPU features are available.
+The EAL can query the CPU at runtime (using the rte_cpu_get_features() function) to determine which CPU features are available.
-User Space Interrupt and Alarm Handling
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+User Space Interrupt Event
+~~~~~~~~~~~~~~~~~~~~~~~~~~
+
++ User Space Interrupt and Alarm Handling in Host Thread
The EAL creates a host thread to poll the UIO device file descriptors to detect the interrupts.
Callbacks can be registered or unregistered by the EAL functions for a specific interrupt event
.. note::
- The only interrupts supported by the DPDK Poll-Mode Drivers are those for link status change,
- i.e. link up and link down notification.
+ In DPDK PMD, the only interrupts handled by the dedicated host thread are those for link status change
+ (link up and link down notification) and for sudden device removal.
+
+
++ RX Interrupt Event
+
+The receive and transmit routines provided by each PMD don't limit themselves to execute in polling thread mode.
+To ease the idle polling with tiny throughput, it's useful to pause the polling and wait until the wake-up event happens.
+The RX interrupt is the first choice to be such kind of wake-up event, but probably won't be the only one.
+
+EAL provides the event APIs for this event-driven thread mode.
+Taking linuxapp as an example, the implementation relies on epoll. Each thread can monitor an epoll instance
+in which all the wake-up events' file descriptors are added. The event file descriptors are created and mapped to
+the interrupt vectors according to the UIO/VFIO spec.
+From bsdapp's perspective, kqueue is the alternative way, but not implemented yet.
+
+EAL initializes the mapping between event file descriptors and interrupt vectors, while each device initializes the mapping
+between interrupt vectors and queues. In this way, EAL actually is unaware of the interrupt cause on the specific vector.
+The eth_dev driver takes responsibility to program the latter mapping.
+
+.. note::
+
+ Per queue RX interrupt event is only allowed in VFIO which supports multiple MSI-X vector. In UIO, the RX interrupt
+ together with other interrupt causes shares the same vector. In this case, when RX interrupt and LSC(link status change)
+ interrupt are both enabled(intr_conf.lsc == 1 && intr_conf.rxq == 1), only the former is capable.
+
+The RX interrupt are controlled/enabled/disabled by ethdev APIs - 'rte_eth_dev_rx_intr_*'. They return failure if the PMD
+hasn't support them yet. The intr_conf.rxq flag is used to turn on the capability of RX interrupt per device.
+
++ Device Removal Event
+
+This event is triggered by a device being removed at a bus level. Its
+underlying resources may have been made unavailable (i.e. PCI mappings
+unmapped). The PMD must make sure that on such occurrence, the application can
+still safely use its callbacks.
+
+This event can be subscribed to in the same way one would subscribe to a link
+status change event. The execution context is thus the same, i.e. it is the
+dedicated interrupt host thread.
+
+Considering this, it is likely that an application would want to close a
+device having emitted a Device Removal Event. In such case, calling
+``rte_eth_dev_close()`` can trigger it to unregister its own Device Removal Event
+callback. Care must be taken not to close the device from the interrupt handler
+context. It is necessary to reschedule such closing operation.
Blacklisting
~~~~~~~~~~~~
Multiple pthread
----------------
-DPDK usually pin one pthread per core to avoid task switch overhead. It gains
-performance a lot, but it's not flexible and not always efficient.
+DPDK usually pins one pthread per core to avoid the overhead of task switching.
+This allows for significant performance gains, but lacks flexibility and is not always efficient.
-Power management helps to improve the cpu efficient by limiting the cpu runtime frequency.
-But there's more reasonable motivation to utilize the ineffective idle cycles under the full capability of cpu.
+Power management helps to improve the CPU efficiency by limiting the CPU runtime frequency.
+However, alternately it is possible to utilize the idle cycles available to take advantage of
+the full capability of the CPU.
-By OS scheduing and cgroup, to each pthread on specified cpu, it can simply assign the cpu quota.
-It gives another way to improve the cpu efficiency. But the prerequisite is to run DPDK execution conext from multiple pthread on one core.
-
-For flexibility, it's also useful to allow the pthread affinity not only to a cpu but to a cpu set.
+By taking advantage of cgroup, the CPU utilization quota can be simply assigned.
+This gives another way to improve the CPU efficiency, however, there is a prerequisite;
+DPDK must handle the context switching between multiple pthreads per core.
+For further flexibility, it is useful to set pthread affinity not only to a CPU but to a CPU set.
EAL pthread and lcore Affinity
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-In terms of lcore, it stands for an EAL execution unit in the EAL pthread.
-EAL pthread indicates all the pthreads created/managed by EAL, they execute the tasks issued by *remote_launch*.
-In each EAL pthread, there's a TLS called *_lcore_id* for the unique identification.
-As EAL pthreads usually 1:1 bind to the physical cpu, *_lcore_id* typically equals to the cpu id.
+The term "lcore" refers to an EAL thread, which is really a Linux/FreeBSD pthread.
+"EAL pthreads" are created and managed by EAL and execute the tasks issued by *remote_launch*.
+In each EAL pthread, there is a TLS (Thread Local Storage) called *_lcore_id* for unique identification.
+As EAL pthreads usually bind 1:1 to the physical CPU, the *_lcore_id* is typically equal to the CPU ID.
-In multiple pthread case, EAL pthread is no longer always bind to one specific physical cpu.
-It may affinity to a cpuset. Then the *_lcore_id* won't always be the same as cpu id.
-So there's an EAL long option '--lcores' defined to assign the cpu affinity of lcores.
-For a specified lcore id or id group, it allows to set the cpuset for that EAL pthread.
+When using multiple pthreads, however, the binding is no longer always 1:1 between an EAL pthread and a specified physical CPU.
+The EAL pthread may have affinity to a CPU set, and as such the *_lcore_id* will not be the same as the CPU ID.
+For this reason, there is an EAL long option '--lcores' defined to assign the CPU affinity of lcores.
+For a specified lcore ID or ID group, the option allows setting the CPU set for that EAL pthread.
The format pattern:
--lcores='<lcore_set>[@cpu_set][,<lcore_set>[@cpu_set],...]'
A number is a "digit([0-9]+)"; a range is "<number>-<number>"; a group is "(<number|range>[,<number|range>,...])".
-If not supply a '\@cpu_set', the value of 'cpu_set' uses the same value as 'lcore_set'.
+If a '\@cpu_set' value is not supplied, the value of 'cpu_set' will default to the value of 'lcore_set'.
::
lcore 7 runs on cpuset 0x80 (cpu 7);
lcore 8 runs on cpuset 0x100 (cpu 8).
-By this option, for each given lcore id, the associated cpus can be assigned.
+Using this option, for each given lcore ID, the associated CPUs can be assigned.
It's also compatible with the pattern of corelist('-l') option.
non-EAL pthread support
~~~~~~~~~~~~~~~~~~~~~~~
-It allows to use DPDK execution context in any user pthread(aka. non-EAL pthread).
-
-In a non-EAL pthread, the *_lcore_id* is always LCORE_ID_ANY which means it's not an EAL thread along with a valid *_lcore_id*.
-Then the libraries won't take *_lcore_id* as unique id. Instead of it, some libraries use another alternative unique id(e.g. tid);
-some are totaly no impact; and some work with some limitation(e.g. timer, mempool).
+It is possible to use the DPDK execution context with any user pthread (aka. Non-EAL pthreads).
+In a non-EAL pthread, the *_lcore_id* is always LCORE_ID_ANY which identifies that it is not an EAL thread with a valid, unique, *_lcore_id*.
+Some libraries will use an alternative unique ID (e.g. TID), some will not be impacted at all, and some will work but with limitations (e.g. timer and mempool libraries).
All these impacts are mentioned in :ref:`known_issue_label` section.
Public Thread API
~~~~~~~~~~~~~~~~~
-There are two public API ``rte_thread_set_affinity()`` and ``rte_pthread_get_affinity()`` introduced for threads.
+There are two public APIs ``rte_thread_set_affinity()`` and ``rte_thread_get_affinity()`` introduced for threads.
When they're used in any pthread context, the Thread Local Storage(TLS) will be set/get.
Those TLS include *_cpuset* and *_socket_id*:
-* *_cpuset* stores the cpus bitmap to which the pthread affinity.
+* *_cpuset* stores the CPUs bitmap to which the pthread is affinitized.
-* *_socket_id* stores the NUMA node of the cpuset. If the cpus in cpuset belong to different NUMA node, the *_socket_id* set to SOCKTE_ID_ANY.
+* *_socket_id* stores the NUMA node of the CPU set. If the CPUs in CPU set belong to different NUMA node, the *_socket_id* will be set to SOCKET_ID_ANY.
.. _known_issue_label:
+ rte_mempool
- The rte_mempool uses a per-lcore cache inside mempool.
- For non-EAL pthread, ``rte_lcore_id()`` will not return a valid number.
- So for now, when rte_mempool is used in non-EAL pthread, the put/get operations will bypass the mempool cache.
- There's performance penalty if bypassing the mempool cache. The work for none-EAL mempool cache support is in progress.
-
- However, there's another problem. The rte_mempool is not preemptable. This comes from rte_ring.
+ The rte_mempool uses a per-lcore cache inside the mempool.
+ For non-EAL pthreads, ``rte_lcore_id()`` will not return a valid number.
+ So for now, when rte_mempool is used with non-EAL pthreads, the put/get operations will bypass the default mempool cache and there is a performance penalty because of this bypass.
+ Only user-owned external caches can be used in a non-EAL context in conjunction with ``rte_mempool_generic_put()`` and ``rte_mempool_generic_get()`` that accept an explicit cache parameter.
+ rte_ring
- rte_ring supports multi-producer enqueue and multi-consumer dequeue. But it's non-preemptive.
+ rte_ring supports multi-producer enqueue and multi-consumer dequeue.
+ However, it is non-preemptive, this has a knock on effect of making rte_mempool non-preemptable.
.. note::
be preempted by another pthread doing a multi-consumer dequeue on
the same ring.
- Bypassing this constraints may cause the 2nd pthread to spin until the 1st one is scheduled again.
+ Bypassing this constraint may cause the 2nd pthread to spin until the 1st one is scheduled again.
Moreover, if the 1st pthread is preempted by a context that has an higher priority, it may even cause a dead lock.
- But it doesn't means we can't use. Just need to narrow down the situation when it's used by multi-pthread on the same core.
+ This does not mean it cannot be used, simply, there is a need to narrow down the situation when it is used by multi-pthread on the same core.
1. It CAN be used for any single-producer or single-consumer situation.
- 2. It MAY be used by multi-producer/consumer pthread whose scheduling policy are all SCHED_OTHER(cfs). User SHOULD aware of the performance penalty before using it.
-
- 3. It MUST not be used by multi-producer/consumer pthread, while some of their scheduling policies is SCHED_FIFO or SCHED_RR.
+ 2. It MAY be used by multi-producer/consumer pthread whose scheduling policy are all SCHED_OTHER(cfs). User SHOULD be aware of the performance penalty before using it.
- ``RTE_RING_PAUSE_REP_COUNT`` is defined for rte_ring to reduce contention. It's mainly for case 2, a yield is issued after number of times pause repeat.
-
- It adds a sched_yield() syscall if the thread spins for too long, waiting other thread to finish its operations on the ring.
- That gives pre-empted thread a chance to proceed and finish with ring enqnue/dequeue operation.
+ 3. It MUST not be used by multi-producer/consumer pthreads, whose scheduling policies are SCHED_FIFO or SCHED_RR.
+ rte_timer
- It's not allowed to run ``rte_timer_manager()`` on a non-EAL pthread. But it's all right to reset/stop the timer from a non-EAL pthread.
+ Running ``rte_timer_manager()`` on a non-EAL pthread is not allowed. However, resetting/stopping the timer from a non-EAL pthread is allowed.
+ rte_log
- In non-EAL pthread, there's no per thread loglevel and logtype. It uses the global loglevel.
+ In non-EAL pthreads, there is no per thread loglevel and logtype, global loglevels are used.
+ misc
cgroup control
~~~~~~~~~~~~~~
-Here's a simple example, there's two pthreads(t0 and t1) doing packet IO on the same core($cpu).
+The following is a simple example of cgroup control usage, there are two pthreads(t0 and t1) doing packet I/O on the same core ($CPU).
We expect only 50% of CPU spend on packet IO.
- .. code::
+ .. code-block:: console
mkdir /sys/fs/cgroup/cpu/pkt_io
mkdir /sys/fs/cgroup/cpuset/pkt_io
echo 50000 > pkt_io/cpu.cfs_quota_us
-.. |linuxapp_launch| image:: img/linuxapp_launch.*
+Malloc
+------
+
+The EAL provides a malloc API to allocate any-sized memory.
+
+The objective of this API is to provide malloc-like functions to allow
+allocation from hugepage memory and to facilitate application porting.
+The *DPDK API Reference* manual describes the available functions.
+
+Typically, these kinds of allocations should not be done in data plane
+processing because they are slower than pool-based allocation and make
+use of locks within the allocation and free paths.
+However, they can be used in configuration code.
+
+Refer to the rte_malloc() function description in the *DPDK API Reference*
+manual for more information.
+
+Cookies
+~~~~~~~
+
+When CONFIG_RTE_MALLOC_DEBUG is enabled, the allocated memory contains
+overwrite protection fields to help identify buffer overflows.
+
+Alignment and NUMA Constraints
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The rte_malloc() takes an align argument that can be used to request a memory
+area that is aligned on a multiple of this value (which must be a power of two).
+
+On systems with NUMA support, a call to the rte_malloc() function will return
+memory that has been allocated on the NUMA socket of the core which made the call.
+A set of APIs is also provided, to allow memory to be explicitly allocated on a
+NUMA socket directly, or by allocated on the NUMA socket where another core is
+located, in the case where the memory is to be used by a logical core other than
+on the one doing the memory allocation.
+
+Use Cases
+~~~~~~~~~
+
+This API is meant to be used by an application that requires malloc-like
+functions at initialization time.
+
+For allocating/freeing data at runtime, in the fast-path of an application,
+the memory pool library should be used instead.
+
+Internal Implementation
+~~~~~~~~~~~~~~~~~~~~~~~
+
+Data Structures
+^^^^^^^^^^^^^^^
+
+There are two data structure types used internally in the malloc library:
+
+* struct malloc_heap - used to track free space on a per-socket basis
+
+* struct malloc_elem - the basic element of allocation and free-space
+ tracking inside the library.
+
+Structure: malloc_heap
+""""""""""""""""""""""
+
+The malloc_heap structure is used to manage free space on a per-socket basis.
+Internally, there is one heap structure per NUMA node, which allows us to
+allocate memory to a thread based on the NUMA node on which this thread runs.
+While this does not guarantee that the memory will be used on that NUMA node,
+it is no worse than a scheme where the memory is always allocated on a fixed
+or random node.
+
+The key fields of the heap structure and their function are described below
+(see also diagram above):
+
+* lock - the lock field is needed to synchronize access to the heap.
+ Given that the free space in the heap is tracked using a linked list,
+ we need a lock to prevent two threads manipulating the list at the same time.
+
+* free_head - this points to the first element in the list of free nodes for
+ this malloc heap.
+
+.. note::
+
+ The malloc_heap structure does not keep track of in-use blocks of memory,
+ since these are never touched except when they are to be freed again -
+ at which point the pointer to the block is an input to the free() function.
+
+.. _figure_malloc_heap:
+
+.. figure:: img/malloc_heap.*
+
+ Example of a malloc heap and malloc elements within the malloc library
+
+
+.. _malloc_elem:
+
+Structure: malloc_elem
+""""""""""""""""""""""
+
+The malloc_elem structure is used as a generic header structure for various
+blocks of memory.
+It is used in three different ways - all shown in the diagram above:
+
+#. As a header on a block of free or allocated memory - normal case
+
+#. As a padding header inside a block of memory
+
+#. As an end-of-memseg marker
+
+The most important fields in the structure and how they are used are described below.
+
+.. note::
+
+ If the usage of a particular field in one of the above three usages is not
+ described, the field can be assumed to have an undefined value in that
+ situation, for example, for padding headers only the "state" and "pad"
+ fields have valid values.
+
+* heap - this pointer is a reference back to the heap structure from which
+ this block was allocated.
+ It is used for normal memory blocks when they are being freed, to add the
+ newly-freed block to the heap's free-list.
+
+* prev - this pointer points to the header element/block in the memseg
+ immediately behind the current one. When freeing a block, this pointer is
+ used to reference the previous block to check if that block is also free.
+ If so, then the two free blocks are merged to form a single larger block.
+
+* next_free - this pointer is used to chain the free-list of unallocated
+ memory blocks together.
+ It is only used in normal memory blocks; on ``malloc()`` to find a suitable
+ free block to allocate and on ``free()`` to add the newly freed element to
+ the free-list.
+
+* state - This field can have one of three values: ``FREE``, ``BUSY`` or
+ ``PAD``.
+ The former two are to indicate the allocation state of a normal memory block
+ and the latter is to indicate that the element structure is a dummy structure
+ at the end of the start-of-block padding, i.e. where the start of the data
+ within a block is not at the start of the block itself, due to alignment
+ constraints.
+ In that case, the pad header is used to locate the actual malloc element
+ header for the block.
+ For the end-of-memseg structure, this is always a ``BUSY`` value, which
+ ensures that no element, on being freed, searches beyond the end of the
+ memseg for other blocks to merge with into a larger free area.
+
+* pad - this holds the length of the padding present at the start of the block.
+ In the case of a normal block header, it is added to the address of the end
+ of the header to give the address of the start of the data area, i.e. the
+ value passed back to the application on a malloc.
+ Within a dummy header inside the padding, this same value is stored, and is
+ subtracted from the address of the dummy header to yield the address of the
+ actual block header.
+
+* size - the size of the data block, including the header itself.
+ For end-of-memseg structures, this size is given as zero, though it is never
+ actually checked.
+ For normal blocks which are being freed, this size value is used in place of
+ a "next" pointer to identify the location of the next block of memory that
+ in the case of being ``FREE``, the two free blocks can be merged into one.
+
+Memory Allocation
+^^^^^^^^^^^^^^^^^
+
+On EAL initialization, all memsegs are setup as part of the malloc heap.
+This setup involves placing a dummy structure at the end with ``BUSY`` state,
+which may contain a sentinel value if ``CONFIG_RTE_MALLOC_DEBUG`` is enabled,
+and a proper :ref:`element header<malloc_elem>` with ``FREE`` at the start
+for each memseg.
+The ``FREE`` element is then added to the ``free_list`` for the malloc heap.
+
+When an application makes a call to a malloc-like function, the malloc function
+will first index the ``lcore_config`` structure for the calling thread, and
+determine the NUMA node of that thread.
+The NUMA node is used to index the array of ``malloc_heap`` structures which is
+passed as a parameter to the ``heap_alloc()`` function, along with the
+requested size, type, alignment and boundary parameters.
+
+The ``heap_alloc()`` function will scan the free_list of the heap, and attempt
+to find a free block suitable for storing data of the requested size, with the
+requested alignment and boundary constraints.
+
+When a suitable free element has been identified, the pointer to be returned
+to the user is calculated.
+The cache-line of memory immediately preceding this pointer is filled with a
+struct malloc_elem header.
+Because of alignment and boundary constraints, there could be free space at
+the start and/or end of the element, resulting in the following behavior:
+
+#. Check for trailing space.
+ If the trailing space is big enough, i.e. > 128 bytes, then the free element
+ is split.
+ If it is not, then we just ignore it (wasted space).
+
+#. Check for space at the start of the element.
+ If the space at the start is small, i.e. <=128 bytes, then a pad header is
+ used, and the remaining space is wasted.
+ If, however, the remaining space is greater, then the free element is split.
+
+The advantage of allocating the memory from the end of the existing element is
+that no adjustment of the free list needs to take place - the existing element
+on the free list just has its size pointer adjusted, and the following element
+has its "prev" pointer redirected to the newly created element.
+
+Freeing Memory
+^^^^^^^^^^^^^^
+
+To free an area of memory, the pointer to the start of the data area is passed
+to the free function.
+The size of the ``malloc_elem`` structure is subtracted from this pointer to get
+the element header for the block.
+If this header is of type ``PAD`` then the pad length is further subtracted from
+the pointer to get the proper element header for the entire block.
+
+From this element header, we get pointers to the heap from which the block was
+allocated and to where it must be freed, as well as the pointer to the previous
+element, and via the size field, we can calculate the pointer to the next element.
+These next and previous elements are then checked to see if they are also
+``FREE``, and if so, they are merged with the current element.
+This means that we can never have two ``FREE`` memory blocks adjacent to one
+another, as they are always merged into a single block.
git.droids-corp.org / dpdk.git / blobdiff