fi_setup(7)

🌸 NAME 🌸

fi_setup(7) — libfabric setup and initialization. This is the Libfabric v2.5.1 page, and it’s basically your starter guide to getting a libfabric app off the ground. 💖

✨ OVERVIEW ✨

Okay so the full libfabric API? It lives in the relevant man pages, periodt. This section is just an introduction to select interfaces and how they may be used. It does not try to capture every subtlety or use case, and it does not describe every possible data structure or field. But if you’re a new developer trying to kick-start using libfabric? This is useful, fr fr. 🫶

🌈 fi_getinfo() 🌈

fi_getinfo() is one of the first calls an application invokes. It’s designed to be easy to use for simple applications, but extensible enough to configure a network for optimal performance. It serves several purposes:

• First, it abstracts away network implementation and addressing details.
• Second, it allows an application to specify which features they require of the network.
• Last, it provides a mechanism for a provider to report how an application can use the network in order to achieve the best performance.

fi_getinfo() is loosely based on the getaddrinfo() call. 👑

/* API prototypes */
struct fi_info *fi_allocinfo(void);

int fi_getinfo(int version, const char *node, const char *service,
    uint64_t flags, struct fi_info *hints, struct fi_info **info);

/* Sample initialization code flow */
struct fi_info *hints, *info;

hints = fi_allocinfo();

/* hints will point to a cleared fi_info structure
 * Initialize hints here to request specific network capabilities
 */

fi_getinfo(FI_VERSION(1, 16), NULL, NULL, 0, hints, &info);
fi_freeinfo(hints);

/* Use the returned info structure to allocate fabric resources */

The hints parameter is the key for requesting fabric services. The fi_info structure contains several data fields, plus pointers to a wide variety of attributes. The fi_allocinfo() call simplifies the creation of an fi_info structure and is strongly recommended for use. In this example, the application is merely attempting to get a list of what providers are available in the system and the features that they support. Note that the API is designed to be extensible. Versioning information is provided as part of the fi_getinfo() call. The version is used by libfabric to determine what API features the application is aware of. In this case, the application indicates that it can properly handle any feature that was defined for the 1.16 release (or earlier).

Applications should always hard code the version that they are written for into the fi_getinfo() call. This ensures that newer versions of libfabric will provide backwards compatibility with that used by the application. Newer versions of libfabric must support applications that were compiled against an older version of the library. It must also support applications written against header files from an older library version, but re-compiled against newer header files. Among other things, the version parameter allows libfabric to determine if an application is aware of new fields that may have been added to structures, or if the data in those fields may be uninitialized. 😭 (this is why you don’t freestyle the version, bestie.)

Typically, an application will initialize the hints parameter to list the features that it will use.

/* Taking a peek at the contents of fi_info */
struct fi_info {
    struct fi_info *next;
    uint64_t caps;
    uint64_t mode;
    uint32_t addr_format;
    size_t src_addrlen;
    size_t dest_addrlen;
    void *src_addr;
    void *dest_addr;
    fid_t handle;
    struct fi_tx_attr *tx_attr;
    struct fi_rx_attr *rx_attr;
    struct fi_ep_attr *ep_attr;
    struct fi_domain_attr *domain_attr;
    struct fi_fabric_attr *fabric_attr;
    struct fid_nic *nic;
};

The fi_info structure references several different attributes, which correspond to the different libfabric objects that an application allocates. For basic applications, modifying or accessing most attribute fields is unnecessary. Many applications will only need to deal with a few fields of fi_info, most notably the endpoint type, capability (caps) bits, and mode bits. These are defined in more detail below. 🌸

On success, the fi_getinfo() function returns a linked list of fi_info structures. Each entry in the list will meet the conditions specified through the hints parameter. The returned entries may come from different network providers, or may differ in the returned attributes. For example, if hints does not specify a particular endpoint type, there may be an entry for each of the three endpoint types. As a general rule, libfabric attempts to return the list of fi_info structures in order from most desirable to least. High-performance network providers are listed before more generic providers. ⭐

🦋 Capabilities (fi_info::caps) 🦋

The fi_info caps field is used to specify the features and services that the application requires of the network. This field is a bit-mask of desired capabilities. There are capability bits for each of the data transfer services previously mentioned: FI_MSG, FI_TAGGED, FI_RMA, FI_ATOMIC, and FI_COLLECTIVE. Applications should set each bit for each set of operations that it will use. These bits are often the only caps bits set by an application. 💅

Capabilities are grouped into three general categories: primary, secondary, and primary modifiers. Primary capabilities must explicitly be requested by an application, and a provider must enable support for only those primary capabilities which were selected. This is required for both performance and security reasons. Primary modifiers are used to limit a primary capability, such as restricting an endpoint to being send-only.

Secondary capabilities may optionally be requested by an application. If requested, a provider must support a capability if it is asked for or fail the fi_getinfo request. A provider may optionally report non-requested secondary capabilities if doing so would not compromise performance or security. That is, a provider may grant an application a secondary capability, regardless of whether the application requested it. The most commonly accessed secondary capability bits indicate if provider communication is restricted to the local node (for example, the shared memory provider only supports local communication) and/or remote nodes (which can be the case for NICs that lack loopback support). Other secondary capability bits mostly deal with features targeting highly-scalable applications, but may not be commonly supported across multiple providers.

Because different providers support different sets of capabilities, applications that desire optimal network performance may need to code for a capability being either present or absent. When present, such capabilities can offer a scalability or performance boost. When absent, an application may prefer to adjust its protocol or implementation to work around the network limitations. Although providers can often emulate features, doing so can impact overall performance, including the performance of data transfers that otherwise appear unrelated to the feature in use. For example, if a provider needs to insert protocol headers into the message stream in order to implement a given capability, the insertion of that header could negatively impact the performance of all transfers. By exposing such limitations to the application, the application developer has better control over how to best emulate the feature or work around its absence.

It is recommended that applications code for only those capabilities required to achieve the best performance. If a capability would have little to no effect on overall performance, developers should avoid using such features as part of an initial implementation. This will allow the application to work well across the widest variety of hardware. Application optimizations can then add support for less common features. To see which features are supported by which providers, see the libfabric Provider Feature Maxtrix for the relevant release. 🔥

🧚 Mode Bits (fi_info::mode) 🧚

Where capability bits represent features desired by applications, mode bits correspond to behavior needed by the provider. That is, capability bits are top down requests, whereas mode bits are bottom up restrictions. Mode bits are set by the provider to request that the application use the API in a specific way in order to achieve optimal performance. Mode bits often imply that the additional work to implement certain communication semantics needed by the application will be less if implemented by the application than forcing that same implementation down into the provider. Mode bits arise as a result of hardware implementation restrictions.

An application developer decides which mode bits they want to or can easily support as part of their development process. Each mode bit describes a particular behavior that the application must follow to use various interfaces. Applications set the mode bits that they support when calling fi_getinfo(). If a provider requires a mode bit that isn’t set, that provider will be skipped by fi_getinfo(). If a provider does not need a mode bit that is set, it will respond to the fi_getinfo() call, with the mode bit cleared. This indicates that the application does not need to perform the action required by the mode bit. 🫶

One of common mode bit needed by providers is FI_CONTEXT (and FI_CONTEXT2). This mode bit requires that applications pass in a libfabric defined data structure (struct fi_context) into any data transfer function. That structure must remain valid and unused by the application until the data transfer operation completes. The purpose behind this mode bit is that the struct fi_context provides “scratch” space that the provider can use to track the request. For example, it may need to insert the request into a linked list while it is pending, or track the number of times that an outbound transfer has been retried. Since many applications already track outstanding operations with their own data structure, by embedding the struct fi_context into that same structure, overall performance can be improved. This avoids the provider needing to allocate and free internal structures for each request. ✨

Okay, continuing with this example: if an app doesn’t already track outstanding requests, then it would leave the FI_CONTEXT mode bit unset. That would indicate that the provider needs to get and release its own structure for tracking purposes. In this case, the costs would essentially be the same whether it were done by the application or the provider.

For the broadest support of different network technologies, applications should attempt to support as many mode bits as feasible. It is recommended that providers support applications that cannot support any mode bits, with as small an impact as possible. However, implementation of mode bit avoidance in the provider can still impact performance, even when the mode bit is disabled. As a result, some providers may always require specific mode bits be set.

🦋 FIDs (fid_t) 🦋

FID stands for fabric identifier. It is the base object type assigned to all libfabric API objects. All fabric resources are represented by a fid structure, and all fid’s are derived from a base fid type. In object-oriented terms, a fid would be the parent class. The contents of a fid are visible to the application.

/* Base FID definition */
enum {
    FI_CLASS_UNSPEC,
    FI_CLASS_FABRIC,
    FI_CLASS_DOMAIN,
    ...
};

struct fi_ops {
    size_t size;
    int (*close)(struct fid *fid);
    ...
};

/* All fabric interface descriptors must start with this structure */
struct fid {
    size_t fclass;
    void *context;
    struct fi_ops *ops;
};

The fid structure is designed as a trade-off between minimizing memory footprint versus software overhead. Each fid is identified as a specific object class, which helps with debugging. Examples are given above (e.g. FI_CLASS_FABRIC). The context field is an application defined data value, assigned to an object during its creation. The use of the context field is application specific, but it is meant to be read by applications. Applications often set context to a corresponding structure that it’s allocated. The context field is the only field that applications are recommended to access directly. Access to other fields should be done using defined function calls (for example, the close() operation).

The ops field points to a set of function pointers. The fi_ops structure defines the operations that apply to that class. The size field in the fi_ops structure is used for extensibility, and allows the fi_ops structure to grow in a backward compatible manner as new operations are added. The fid deliberately points to the fi_ops structure, rather than embedding the operations directly. This allows multiple fids to point to the same set of ops, which minimizes the memory footprint of each fid. (Internally, providers usually set ops to a static data structure, with the fid structure dynamically allocated.)

Although it’s possible for applications to access function pointers directly, it is strongly recommended that the static inline functions defined in the man pages be used instead. This is required by applications that may be built using the FABRIC_DIRECT library feature. (FABRIC_DIRECT is a compile time option that allows for highly optimized builds by tightly coupling an application with a specific provider.)

Other OFI classes are derived from this structure, adding their own set of operations.

/* Example of deriving a new class for a fabric object */
struct fi_ops_fabric {
    size_t size;
    int (*domain)(struct fid_fabric *fabric, struct fi_info *info,
        struct fid_domain **dom, void *context);
    ...
};

struct fid_fabric {
    struct fid fid;
    struct fi_ops_fabric *ops;
};

Other fid classes follow a similar pattern as that shown for fid_fabric. The base fid structure is followed by zero or more pointers to operation sets.

🌸 Fabric (fid_fabric) 🌸

The top-level object that applications open is the fabric identifier. The fabric can mostly be viewed as a container object by applications, though it does identify which provider(s) applications use.

Opening a fabric is usually a straightforward call after calling fi_getinfo().

int fi_fabric(struct fi_fabric_attr *attr, struct fid_fabric **fabric, void *context);

The fabric attributes can be directly accessed from struct fi_info. The newly opened fabric is returned through the fabric parameter. The context parameter appears in many operations. It is a user-specified value that is associated with the fabric. It may be used to point to an application specific structure and is retrievable from struct fid_fabric.

🎀 Attributes (fi_fabric_attr) 🎀

The fabric attributes are straightforward.

struct fi_fabric_attr {
    struct fid_fabric *fabric;
    char *name;
    char *prov_name;
    uint32_t prov_version;
    uint32_t api_version;
};

The only field that applications are likely to use directly is the prov_name. This is a string value that can be used by hints to select a specific provider for use. On most systems, there will be multiple providers available. Only one is likely to represent the high-performance network attached to the system. Others are generic providers that may be available on any system, such as the TCP socket and UDP providers.

The fabric field is used to help applications manage opened fabric resources. If an application has already opened a fabric that can support the returned fi_info structure, this will be set to that fabric.

Domains (fid_domain)

Okay so domains — these are a whole vibe. 💅 A domain is usually gonna map to one specific local network interface adapter, but like… don't get it twisted. A domain could mean the entire NIC, or just one port on a multi-port NIC, or a virtual device a NIC exposes, or even several NICs being used in a multi-rail kinda way, and so on and so on. It's tempting to be like "domain = NIC, periodt," but libfabric lowkey does not expect that association. From the app's point of view? A domain just identifies a set of resources that get used together. That's it, that's the tweet. 🦋

Same energy as fabric, opening a domain is literally no big deal once you've called fi_getinfo().

int fi_domain(struct fid_fabric *fabric, struct fi_info *info,
    struct fid_domain **domain, void *context);

The fi_info structure that fi_getinfo() hands you can be passed directly into fi_domain() to open a brand-new domain. Slay. ✨

Attributes (fi_domain_attr)

So one of the whole goals of a domain is defining the relationship between data transfer services (endpoints) and completion services (completion queues and counters). A bunch of the domain attributes are literally just describing that relationship and how it impacts your application.

struct fi_domain_attr {
    struct fid_domain *domain;
    char *name;
    enum fi_threading threading;
    enum fi_progress progress;
    enum fi_resource_mgmt resource_mgmt;
    enum fi_av_type av_type;
    enum fi_mr_mode mr_mode;
    size_t mr_key_size;
    size_t cq_data_size;
    size_t cq_cnt;
    size_t ep_cnt;
    size_t tx_ctx_cnt;
    size_t rx_ctx_cnt;
    ...

Full deets on the domain attributes and what they mean live over in the fi_domain man page — go read that bestie, it's giving thorough. 👑 But below we're gonna cover some select attributes and how they hit your application.

Threading (fi_threading)

libfabric defines a unique threading model, and ngl the whole design is heavily influenced by object-oriented programming concepts. A multi-threaded application has to figure out how libfabric objects (domains, endpoints, completion queues, etc.) get allocated among its threads — or whether any thread can just access any object, yolo style. Like, an application might spawn a new thread to handle each new connected endpoint. The domain's threading field is the mechanism for an app to identify which objects may be accessed simultaneously by different threads. Which in turn lets a provider optimize — or, in some cases, straight up eliminate — internal synchronization and locking around those objects. 🔥

Threading defines where providers could optimize synchronization primitives. But heads up: providers may still implement more serialization than the app actually needs. (That's usually just a side effect of keeping the provider implementation simpler.) 🙈

It's recommended that apps target either FI_THREAD_SAFE (full thread safety, implemented by the provider) or FI_THREAD_DOMAIN (objects associated with a single domain will only be accessed by a single thread). Pick your fighter, both are valid. 💜

Progress (fi_progress)

Progress models are a result of using the host processor to do some portion of the transport protocol. To keep development simple, libfabric defines two progress models: automatic or manual. It does not try to identify which specific interface features may be offloaded, or which operations need extra processing by the application's thread. Not its job, not its problem. 💅

Automatic progress means an operation initiated by the application will eventually complete, even if the application makes no further calls into the libfabric API. The op is either offloaded entirely onto hardware, or the provider uses an internal thread, or the operating system kernel may perform the task. In those last two cases, using automatic progress may increase system overhead and latency. For control operations, like connection setup, that's usually acceptable. But the impact to data transfers might be measurable — especially if internal threads are required to provide automatic progress. 😭

The manual progress model can dodge that overhead for providers that don't offload all transport features into hardware. With manual progress, the provider implementation handles transport operations as part of specific libfabric functions. For example, a call to fi_cq_read() — which retrieves an array of completed operations — may also be responsible for sending ack messages to notify peers that a message has been received. Since reading the completion queue is part of an application's normal operation, there's minimal impact to the app and you avoid the extra threads. 🎀

Applications need to take care when using manual progress, particularly if they link into libfabric multiple times through different code paths or library dependencies. If application threads are used to drive progress — like responding to received data with ACKs — then it is critical that the application thread call into libfabric in a timely manner. Do not ghost libfabric, bestie. 🫶

Memory Registration (fid_mr)

RMA, atomic, and collective operations can read and write memory that's owned by a peer process, and neither one requires the involvement of the target processor. Which is iconic, but also means: because that memory can be modified over the network, an application has to opt in to exposing its memory to peers. That's what the memory registration process handles. Registered memory regions associate memory buffers with permissions granted for access by fabric resources. A memory buffer must be registered before it can be used as the target of a remote RMA, atomic, or collective data transfer. On top of that, a fabric provider may require data buffers be registered before being used even for local transfers — that latter bit is necessary to make sure the virtual-to-physical page mappings don't change while network hardware is doing the transfer. 💀

To handle all the diverse hardware requirements, there's a set of mr_mode bits associated with memory registration. The mr_mode bits behave similar to fi_info mode bits. Applications indicate which types of restrictions they can support, and the providers clear those bits that aren't needed. Clean. ✨

For hardware that requires memory registration, managing registration is critical to getting good performance and scalability. The act of registering memory is costly, so it should be avoided on a per-data-transfer basis. libfabric has extensive internal support for managing memory registration — hiding registration from the user application, caching registration to reduce per-transfer overhead, and detecting when cached registrations are no longer valid. It's recommended that applications not natively designed to account for registering memory make use of libfabric's registration cache. You do that by simply not setting the relevant mr_mode bits. Easiest optimization of your life. 🌸

Memory Region APIs

These APIs highlight how to allocate and access a registered memory region. Not a complete list of memory region (MR) calls though — for full details on each API, readers should go directly to the fi_mr man page. iykyk. 📖

int fi_mr_reg(struct fid_domain *domain, const void *buf, size_t len,
    uint64_t access, uint64_t offset, uint64_t requested_key, uint64_t flags,
    struct fid_mr **mr, void *context);

void * fi_mr_desc(struct fid_mr *mr);
uint64_t fi_mr_key(struct fid_mr *mr);

By default, memory regions are associated with a domain. A MR is accessible by any endpoint opened on that domain. A region starts at the address specified by buf, and is len bytes long. The access parameter is permission flags OR'ed together — those permissions indicate which types of operations may be invoked against the region (e.g. FI_READ, FI_WRITE, FI_REMOTE_READ, FI_REMOTE_WRITE). The buf parameter typically references allocated virtual memory. 🧚

A MR is associated with local and remote protection keys. The local key is called a memory descriptor and can be retrieved by calling fi_mr_desc(). That call is only needed if the FI_MR_LOCAL mr_mode bit has been set. The memory descriptor gets passed directly into data transfer operations, like this:

/* fi_mr_desc() example using fi_send() */
fi_send(ep, buf, len, fi_mr_desc(mr), 0, NULL);

The remote key — or honestly just "MR key" — is what the peer uses when targeting the MR with an RMA or atomic operation. In many cases that key will need to be sent in a separate message to the initiating peer. The libfabric API uses a 64-bit key where one is used. The actual key size a provider uses is part of its domain attributes; support for larger key sizes, as required by some providers, is conveyed through an mr_mode bit, and requires using the extended MR API calls that map the larger size to a 64-bit value. Big key energy. 🔑

---

Endpoints

Endpoints are transport level communication portals. Opening an endpoint is trivial once you've called fi_getinfo(). Literally a walk in the park. 🌈

Active (fid_ep)

Active endpoints may be connection-oriented or connection-less. They're considered active 'cause they may be used to perform data transfers. All the data transfer interfaces — messages (fi_msg), tagged messages (fi_tagged), RMA (fi_rma), atomics (fi_atomic), and collectives (fi_collective) — are associated with active endpoints. Though an individual endpoint may not be enabled to use all data transfers. FOMO but make it capability-limited. 😭

In standard configurations, an active endpoint has one transmit and one receive queue. In general, operations that generate traffic on the fabric get posted to the transmit queue — that includes all RMA and atomic operations, along with sent messages and sent tagged messages. Operations that post buffers for receiving incoming data get submitted to the receive queue. Simple as. 💅

Active endpoints are created in the disabled state. The endpoint must first be configured before it's enabled. Endpoints must transition into an enabled state before accepting data transfer operations — including posting of receive buffers. The fi_enable() call is used to transition an active endpoint into an enabled state. The fi_connect() and fi_accept() calls will also transition an endpoint into the enabled state, if it isn't already enabled. Main character endpoint behavior, honestly. 👑

int fi_endpoint(struct fid_domain *domain, struct fi_info *info,
    struct fid_ep **ep, void *context);

Enabling (fi_enable)

In order to transition an endpoint into an enabled state, it must be bound to one or more fabric resources. That includes binding the endpoint to a completion queue and event queue. Unconnected endpoints must also be bound to an address vector. No shortcuts, bestie — bind the things. 🎀

/* Example to enable an unconnected endpoint */

              /* Allocate an address vector and associated it with the endpoint */
              fi_av_open(domain, &av_attr, &av, NULL);
              fi_ep_bind(ep, &av->fid, 0);

              /* Allocate and associate completion queues with the endpoint */
              fi_cq_open(domain, &cq_attr, &cq, NULL);
              fi_ep_bind(ep, &cq->fid, FI_TRANSMIT | FI_RECV);

              fi_enable(ep);

okay so in that little snippet we're allocating an AV and a CQ 💅 — and yeah, the attributes for the AV and the CQ are just... omitted here (we get into them further down, don't worry bestie). Then both of those get associated with the endpoint through the fi_ep_bind() call. And once all the necessary resources have been assigned to the endpoint, we enable it.

Enabling the endpoint is basically telling the provider "go off, allocate whatever hardware and software resources you need and finish initializing this endpoint" ✨ (and real talk: if the endpoint isn't bound to all the necessary resources, the fi_enable() call will fail. no cap.)

So, the fi_enable() call is always called for unconnected endpoints. Connected endpoints may be able to skip calling fi_enable(), because fi_connect() and fi_accept() will enable the endpoint automatically. But — applications may still call fi_enable() prior to calling fi_connect() or fi_accept() anyway. Why? Because doing that lets the application post receive buffers to the endpoint, which ensures they're available to receive data in case the peer endpoint sends messages immediately after it establishes the connection 🫶 iykyk.

---

🦋 Passive (fid_pep)

Passive endpoints are the ones that listen for incoming connection requests 👂✨ They're of type FI_EP_MSG, and heads up: they may not perform any data transfers. Periodt. An application that wants to create a passive endpoint typically calls fi_getinfo() using the FI_SOURCE flag, often only specifying a service address. And that service address corresponds to a TCP port number 🎀

Passive endpoints are associated with event queues. Event queues report connection requests from peers. Unlike active endpoints, passive endpoints are not associated with a domain. And that's actually iconic of them, because it lets an application listen for connection requests across multiple domains — though still restricted to a single provider 💅

              /* Example passive endpoint listen */
              fi_passive_ep(fabric, info, &pep, NULL);

              fi_eq_open(fabric, &eq_attr, &eq, NULL);
              fi_pep_bind(pep, &eq->fid, 0);

              fi_listen(pep);

A passive endpoint must be bound to an event queue before calling listen — that's what ensures connection requests can actually be reported to the application 🔥 To accept new connections, the app waits for a request, allocates a new active endpoint for it, and accepts the request. Like so:

              /* Example accepting a new connection */

              /* Wait for a CONNREQ event */
              fi_eq_sread(eq, &event, &cm_entry, sizeof cm_entry, -1, 0);
              assert(event == FI_CONNREQ);

              /* Allocate a new endpoint for the connection */
              if (!cm_entry.info->domain_attr->domain)
                  fi_domain(fabric, cm_entry.info, &domain, NULL);
              fi_endpoint(domain, cm_entry.info, &ep, NULL);

              fi_ep_bind(ep, &eq->fid, 0);
              fi_cq_open(domain, &cq_attr, &cq, NULL);
              fi_ep_bind(ep, &cq->fid, FI_TRANSMIT | FI_RECV);

              fi_enable(ep);
              fi_recv(ep, rx_buf, len, NULL, 0, NULL);

              fi_accept(ep, NULL, 0);
              fi_eq_sread(eq, &event, &cm_entry, sizeof cm_entry, -1, 0);
              assert(event == FI_CONNECTED);

The connection request event (FI_CONNREQ) includes info about the type of endpoint to allocate, including the default attributes to use 🌸 If a domain hasn't already been opened for that endpoint, you gotta open one. Then the endpoint and its related resources can be allocated. And ngl, this is where it differs from the unconnected endpoint example above: a connected endpoint does not have an AV, but it does need to be bound to an event queue. Here we're reusing the same EQ as the listening endpoint 🎀 Once the other EP resources (e.g. the CQ) have been allocated and bound, the EP can be enabled.

To accept the connection, the application calls fi_accept(). And listen — because of thread synchronization issues, it is possible for the active endpoint to receive data even before fi_accept() can return 😭 The posting of receive buffers prior to calling fi_accept() handles that condition, which avoids network flow control issues popping off immediately after connecting. Chef's kiss. 👑

The fi_eq_sread() calls are blocking (synchronous) read calls to the event queue. They wait until an event occurs — in this case, connection request and establishment events ⭐

---

💖 EP Attributes (fi_ep_attr)

The properties of an endpoint are specified using endpoint attributes. These describe the endpoint as a whole 🫶 (There are additional attributes that specifically relate to the transmit and receive contexts underpinning the endpoint — more on those just below, bestie.)

              struct fi_ep_attr {
                  enum fi_ep_type type;
                  uint32_t        protocol;
                  uint32_t        protocol_version;
                  size_t          max_msg_size;
                  ...
              };

A full description of each field is available in the fi_endpoint man page, with selected details listed below ✨

🌈 Endpoint Type (fi_ep_type)

This indicates the type of endpoint, and there are three flavors: reliable datagram (FI_EP_RDM), reliable-connected (FI_EP_MSG), or unreliable datagram (FI_EP_DGRAM) 🍓 Nearly all applications will want to specify the endpoint type as a hint passed into fi_getinfo, because most applications are only ever coded to support a single endpoint type. Just pick your lane, honestly.

💀 Maximum Message Size (max_msg_size)

This size is the maximum size for any data transfer operation that goes over the endpoint. For unreliable datagram endpoints, this is often the MTU of the underlying network. For reliable endpoints, this value is often a restriction of the underlying transport protocol. A common minimum maximum message size is 2GB, though some providers support an arbitrarily large size 👀 Applications that require transfers larger than the maximum reported size are required to break up a single, large transfer into multiple operations.

Providers expose their hardware or network limits to the applications rather than segmenting large transfers internally, and the reason is to minimize completion overhead. Like — for a provider to support large message segmentation internally, it would need to emulate all the completion mechanisms (queues and counters) in software, even if transfers larger than the transport-supported maximum were literally never used. Not worth it 💅

🧚 Message Order Size (max_order_xxx_size)

These fields specify data ordering. They define the delivery order of transport data into target memory for RMA and atomic operations. Data ordering requires message ordering — and if message ordering is not specified, these fields do not apply 🙈

So, example time: suppose an application issues two RMA write operations to the same target memory location. (Maybe it's writing a time stamp value every time a local condition is met, for instance.) Message ordering says the first write as initiated by the sender is the first write processed by the receiver. Data ordering says whether the data from the first write updates memory before the second write updates memory ✨

The max_order_xxx_size fields indicate how large a message may be while still achieving data ordering. If a field is 0, then no data ordering is guaranteed. If a field is the same as the max_msg_size, then data order is guaranteed for all messages 🎉

Providers may support data ordering up to max_msg_size for back-to-back operations that are the same. For example, an RMA write followed by an RMA write may have data ordering regardless of the size of the data transfer (max_order_waw_size = max_msg_size). Mixed operations, like a read followed by a write, are often restricted. Why? Because RMA read operations may require acknowledgments from the initiator, which impacts the re-transmission protocol 👀

Let's walk it through: consider an RMA read followed by a write. The target will process the read request, retrieve the data, and send a reply. While that's happening, a write is received that wants to update the same memory location accessed by the read. If the target processes the write, it will overwrite the memory used by the read. And if the read response is lost and the read is retried, the target will be unable to re-send the data 😭 To handle this, the target either needs to: defer handling the write until it receives an acknowledgment for the read response, buffer the read response so it can be re-transmitted, or indicate that data ordering is not guaranteed. Pick your struggle.

And because a read or write operation may be gigabytes in size, deferring the write may add significant latency, and buffering the read response may be impractical 💀 The max_order_xxx_size fields indicate how large back-to-back operations may be with ordering still maintained. In many cases, read-after-write and write-and-read ordering may be significantly limited, but still usable for implementing specific algorithms — such as a global locking mechanism 👑

🎀 Rx/Tx Context Attributes (fi_rx_attr / fi_tx_attr)

The endpoint attributes define the overall abilities for the endpoint; however, attributes that apply specifically to receive or transmit contexts are defined by struct fi_rx_attr and fi_tx_attr, respectively:

              struct fi_rx_attr {
                  uint64_t caps;
                  uint64_t mode;
                  uint64_t op_flags;
                  uint64_t msg_order;
                  ...
              };

              struct fi_tx_attr {
                  uint64_t caps;
                  uint64_t mode;
                  uint64_t op_flags;
                  uint64_t msg_order;
                  size_t inject_size;
                  ...
              };

Rx/Tx context capabilities must be a subset of the endpoint capabilities 🔥 For many applications, the default attributes returned by the provider will be sufficient, with the application only needing to specify endpoint attributes. Slay, less work for you fr fr.

okay so bestie, both context attributes have an op_flags field, and it’s literally the vibe-setter for the default operation flags your app wants on any call. Like, if you set the transmit context’s op_flags to FI_INJECT, you’re telling the provider “hey, every transmit operation should assume inject behavior is what I want.” Which means the buffer you hand to the call has to be returned to you by the time the function returns. No cap. op_flags applies to every operation that doesn’t pass flags as part of the call itself (think fi_sendmsg). One iconic use is setting the default completion semantic you want (we’re getting to that next). Setting default op_flags at initialization time means you can avoid passing flags as arguments into some data transfer calls, avoid parsing those flags, and even prepare submitted commands ahead of time. Efficiency slay 💅.

also pls note: some attributes are dependent on the peer endpoint having supporting attributes, otherwise your app behavior won’t be correct. For example, message order must be compatible between the initiator’s transmit attributes and the target’s receive attributes. Any mismatch may result in incorrect behavior that could be difficult to debug. Consider yourself warned 🫶.

✨ Completions ✨

Data transfer operations complete asynchronously. Libfabric defines two mechanisms by which an application can be notified that an operation has completed: completion queues and counters. Regardless of which mechanism is used to notify the app that an operation is done, developers must be aware of what a completion indicates. Periodt.

In all cases, a completion indicates that it’s safe to reuse the buffer(s) associated with the data transfer. This completion mode is referred to as inject complete and corresponds to the operational flag FI_INJECT_COMPLETE. But a completion may also guarantee stronger semantics. 👑

Although libfabric does not define an implementation, a provider can meet the requirement for inject complete by copying the application’s buffer into a network buffer before generating the completion. Even if the transmit operation is lost and must be retried, the provider can re-send the original data from the copied location. For large transfers, a provider may not mark a request as inject complete until the data has been acknowledged by the target. Applications, however, should only infer that it is safe to reuse their data buffer for an inject complete operation. Don’t read more into it, bestie.

Transmit complete is a completion mode that provides slightly stronger guarantees to the application. The meaning of transmit complete depends on whether the endpoint is reliable or unreliable. For an unreliable endpoint (FI_EP_DGRAM), a transmit completion indicates that the request has been delivered to the network. That is, the message has been delivered at least as far as hardware queues on the local NIC. For reliable endpoints, a transmit complete occurs when the request has reached the target endpoint. Typically, this indicates that the target has acked the request. Transmit complete maps to the operation flag FI_TRANSMIT_COMPLETE. 🔥

A third completion mode is defined to provide guarantees beyond transmit complete. With transmit complete, an application knows that the message is no longer dependent on the local NIC or network (e.g. switches). However, the data may be buffered at the remote NIC and has not necessarily been written to the target memory. As a result, data sent in the request may not be visible to all processes. The third completion mode is delivery complete. ✨

Delivery complete indicates that the results of the operation are available to all processes on the fabric. The distinction between transmit and delivery complete is subtle, but important. It often deals with when the target endpoint generates an acknowledgment to a message. For providers that offload transport protocol to the NIC, support for transmit complete is common. Delivery complete guarantees are more easily met by providers that implement portions of their protocol on the host processor. Delivery complete corresponds to the FI_DELIVERY_COMPLETE operation flag. 💅

Applications can request a default completion mode when opening an endpoint by setting one of the above mentioned complete flags as an op_flags for the context’s attributes. However, it is usually recommended that application use the provider’s default flags for best performance, and amend its protocol to achieve its completion semantics. For example, many applications will perform a finalize or commit procedure as part of their operation, which synchronizes the processing of all peers and guarantees that all previously sent data has been received. That’s the move. 🫶

A full discussion of completion semantics is given in the fi_cq man page. Go read it, iykyk.

🎀 CQs (fid_cq) 🎀

Completion queues often map directly to provider hardware mechanisms, and libfabric is designed around minimizing the software impact of accessing those mechanisms. Unlike other objects discussed so far (fabrics, domains, endpoints), completion queues are not part of the fi_info structure or involved with the fi_getinfo() call. Not me explaining the whole family tree 💀.

All active endpoints must be bound with one or more completion queues. This is true even if completions will be suppressed by the application (e.g. using the FI_SELECTIVE_COMPLETION flag). Completion queues are needed to report operations that complete in error and help drive progress in the case of manual progress. No cap.

CQs are allocated separately from endpoints and are associated with endpoints through the fi_ep_bind() function. That’s the link-up. 🔗

🌸 CQ Format (fi_cq_format) 🌸

In order to minimize the amount of data that a provider must report, the type of completion data written back to the application is select-able. This limits the number of bytes the provider writes to memory, and allows necessary completion data to fit into a compact structure. Each CQ format maps to a specific completion structure. Developers should analyze each structure, select the smallest structure that contains all of the data it requires, and specify the corresponding enum value as the CQ format. Smallest-fit energy. 🎯

For example, if an application only needs to know which request completed, along with the size of a received message, it can select the following:

cq_attr->format = FI_CQ_FORMAT_MSG;

struct fi_cq_msg_entry {
    void      *op_context;
    uint64_t  flags;
    size_t    len;
};

Once the format has been selected, the underlying provider will assume that read operations against the CQ will pass in an array of the corresponding structure. The CQ data formats are designed such that a structure that reports more information can be cast to one that reports less. Casting slay, but make sure you’re going in the right direction. 💅

🦋 Reading Completions (fi_cq_read) 🦋

Completions may be read from a CQ by using one of the non-blocking calls, fi_cq_read / fi_cq_readfrom, or one of the blocking calls, fi_cq_sread / fi_cq_sreadfrom. Regardless of which call is used, applications pass in an array of completion structures based on the selected CQ format. The CQ interfaces are optimized for batch completion processing, allowing the application to retrieve multiple completions from a single read call. The difference between the read and readfrom calls is that readfrom returns source addressing data, if available. The readfrom derivative of the calls is only useful for unconnected endpoints, and only if the corresponding endpoint has been configured with the FI_SOURCE capability. Facts. 📚

FI_SOURCE requires that the provider use the source address available in the raw completion data, such as the packet’s source address, to retrieve a matching entry in the endpoint’s address vector. Applications that carry some sort of source identifier as part of their data packets can avoid the overhead associated with using FI_SOURCE. Work smarter, not harder. 🧠

💜 Retrieving Errors 💜

Because the selected completion structure is insufficient to report all data necessary to debug or handle an operation that completes in error, failed operations are reported using a separate fi_cq_readerr() function. This call takes as input a CQ error entry structure, which allows the provider to report more information regarding the reason for the failure. Error tea, served hot. 🍵

/* read error prototype */
fi_cq_readerr(struct fid_cq *cq, struct fi_cq_err_entry *buf, uint64_t flags);

/* error data structure */
struct fi_cq_err_entry {
    void      *op_context;
    uint64_t  flags;
    size_t    len;
    void      *buf;
    uint64_t  data;
    uint64_t  tag;
    size_t    olen;
    int       err;
    int       prov_errno;
    void      *err_data;
    size_t    err_data_size;
};

/* Sample error handling */
struct fi_cq_msg_entry entry;
struct fi_cq_err_entry err_entry;
char err_data[256];
int ret;

err_entry.err_data = err_data;
err_entry.err_data_size = 256;

ret = fi_cq_read(cq, &entry, 1);
if (ret == -FI_EAVAIL)
    ret = fi_cq_readerr(cq, &err_entry, 0);

As illustrated, if an error entry has been inserted into the completion queue, then attempting to read the CQ will result in the read call returning -FI_EAVAIL (error available). This indicates that the application must use the fi_cq_readerr() call to remove the failed operation’s completion information before other completions can be reaped from the CQ. Don’t skip this step, bestie. 😭

A fabric error code regarding the failure is reported as the err field in the fi_cq_err_entry structure. A provider specific error code is also available through the prov_errno field. This field can be decoded into a displayable string using the fi_cq_strerror() routine. The err_data field is provider specific data that assists the provider in decoding the reason for the failure. Decode the drama. 🔍

🎀 Address Vectors (fid_av) 🎀

Okay so address vectors have one main-character goal: let apps talk to thousands to millions of peers while storing as little peer-addressing data as possible. They push all the fabric-specific addressing details away from the app and hand them to the provider. That means the provider can optimize how it converts addresses into routing data and pull off data-compression tricks that an app would find hard to do without knowing low-level fabric addressing details. For example, providers may be able to algorithmically calculate addressing components instead of storing the data locally. They can also talk to resource management entities or fabric manager agents to get quality of service or other information about the fabric, all to improve network utilization. 💫

An equally important objective is making sure the resulting interfaces — especially data transfer operations — are fast and easy to use. Conceptually, an address vector converts an endpoint address into an fi_addr_t. The fi_addr_t (fabric interface address datatype) is a 64-bit value used in all fast-path operations — data transfers and completions. ✨

Address vectors are associated with domain objects. That lets providers implement portions of an address vector, like quality of service mappings, in hardware. 👑

🦋 AV Type (fi_av_type) 🦋

There are two types of address vectors. The type refers to the format of the returned fi_addr_t values for addresses that are inserted into the AV.

With type FI_AV_TABLE, returned addresses are simple indices, and developers may think of the AV as an array of addresses. Each address inserted into the AV is mapped to the index of the next free array slot. The advantage of FI_AV_TABLE is that apps can refer to peers using a simple index, eliminating the app’s need to store any addressing data. I.e. the app can generate the fi_addr_t values themselves. This type maps well to apps such as MPI, where a peer is referenced by rank. 🫶

The second type is FI_AV_MAP. This type does not define any specific format for the fi_addr_t value. Apps that use type map are required to provide the correct fi_addr_t for a given peer when issuing a data transfer operation. The advantage of FI_AV_MAP is that a provider can use the fi_addr_t to encode the target’s address, which avoids retrieving the data from memory. As a simple example, consider a fabric that uses TCP/IPv4 based addressing. An fi_addr_t is large enough to contain the address, which allows a provider to copy the data from the fi_addr_t directly into an outgoing packet. 💅

Sharing AVs Between Processes

Large scale parallel programs typically run with multiple processes allocated on each node. Because these processes communicate with the same set of peers, the addressing data needed by each process is the same. Libfabric defines a mechanism by which processes running on the same node may share their address vectors. This allows a system to maintain a single copy of addressing data, rather than one copy per process. 🌈

Although libfabric does not require any implementation for how an address vector is shared, the interfaces map well to using shared memory. Address vectors which will be shared are given an application specific name. How an application selects a name that avoids conflicts with unrelated processes, or how it communicates the name with peer processes, is outside the scope of libfabric. 🙈

In addition to having a name, a shared AV also has a base map address — map_addr. Use of map_addr is only important for address vectors that are of type FI_AV_MAP, and it allows applications to share fi_addr_t values. From the viewpoint of the application, the map_addr is the base value for all fi_addr_t values. A common use for map_addr is for the process that creates the initial address vector to request a value from the provider, exchange the returned map_addr with its peers, and for the peers to open the shared AV using the same map_addr. This allows the fi_addr_t values to be stored in shared memory that is accessible by all peers. 🧚

Using Native Wait Objects: TryWait

There is an important difference between using libfabric completion objects versus sockets that may not be obvious from the discussions so far. With sockets, the object that is signaled is the same object that abstracts the queues — namely the file descriptor. When data is received on a socket, that data is placed in a queue associated directly with the fd. Reading from the fd retrieves that data. If an application wishes to block until data arrives on a socket, it calls select() or poll() on the fd. The fd is signaled when a message is received, which releases the blocked thread, allowing it to read the fd. 🍓

By associating the wait object with the underlying data queue, applications are exposed to an interface that is easy to use and race free. If data is available to read from the socket at the time select() or poll() is called, those calls simply return that the fd is readable. ⭐

There are a couple of significant disadvantages to this approach, which have been discussed previously but from different perspectives. The first is that every socket must be associated with its own fd. There is no way to share a wait object among multiple sockets. (This is a main reason for the development of epoll semantics.) The second is that the queue is maintained in the kernel, so that the select() and poll() calls can check them. 💀

Libfabric allows for the separation of the wait object from the data queues. For applications that use libfabric interfaces to wait for events, such as fi_cq_sread, this separation is mostly hidden from the application. The exception is that applications may receive a signal, but no events are retrieved when a queue is read. This separation allows the queues to reside in the application’s memory space, while wait objects may still use kernel components. A reason for the latter is that wait objects may be signaled as part of system interrupt processing, which would go through a kernel driver. 😭

Applications that want to use native wait objects (e.g. file descriptors) directly in operating system calls must perform an additional step in their processing. In order to handle race conditions that can occur between inserting an event into a completion or event object and signaling the corresponding wait object, libfabric defines an fi_trywait() function. The fi_trywait implementation is responsible for handling potential race conditions which could result in an application either losing events or hanging. The following example demonstrates the use of fi_trywait(). 🔥

              /* Get the native wait object -- an fd in this case */
              fi_control(&cq->fid, FI_GETWAIT, (void *) &fd);
              FD_ZERO(&fds);
              FD_SET(fd, &fds);

              while (1) {
                  ret = fi_trywait(fabric, &cq->fid, 1);
                  if (ret == FI_SUCCESS) {
                      /* It’s safe to block on the fd */
                      select(fd + 1, &fds, NULL, &fds, &timeout);
                  } else if (ret == -FI_EAGAIN) {
                      /* Read and process all completions from the CQ */
                      do {
                          ret = fi_cq_read(cq, &comp, 1);
                      } while (ret > 0);
                  } else {
                      /* something really bad happened */
                  }
              }

In this example, the application has allocated a CQ with an fd as its wait object. It calls select() on the fd. Before calling select(), the application must call fi_trywait() successfully (return code of FI_SUCCESS). Success indicates that a blocking operation can now be invoked on the native wait object without fear of the application hanging or events being lost. If fi_trywait() returns -FI_EAGAIN, it usually indicates that there are queued events to process. 💜

Environment Variables

Environment variables are used by providers to configure internal options for optimal performance or memory consumption. Libfabric provides an interface for querying which environment variables are usable, along with an application to display the information to a command window. Although environment variables are usually configured by an administrator, an application can query for variables programmatically. 🌸

              /* APIs to query for supported environment variables */
              enum fi_param_type {
                  FI_PARAM_STRING,
                  FI_PARAM_INT,
                  FI_PARAM_BOOL,
                  FI_PARAM_SIZE_T,
              };

              struct fi_param {
                  /* The name of the environment variable */
                  const char *name;
                  /* What type of value it stores */
                  enum fi_param_type type;
                  /* A description of how the variable is used */
                  const char *help_string;
                  /* The current value of the variable */
                  const char *value;
              };

              int fi_getparams(struct fi_param **params, int *count);
              void fi_freeparams(struct fi_param *params);

The modification of environment variables is typically a tuning activity done on larger clusters. However there are a few values that are useful for developers. These can be seen by executing the fi_info command. 🎀

              $ fi_info -e
              # FI_LOG_LEVEL: String
              # Specify logging level: warn, trace, info, debug (default: warn)

              # FI_LOG_PROV: String
              # Specify specific provider to log (default: all)

              # FI_PROVIDER: String
              # Only use specified provider (default: all available)

The fi_info application, which ships with libfabric, can be used to list all environment variables for all providers. The -e option will list all variables, and the -g option can be used to filter the output to only those variables with a matching substring. Variables are documented directly in code with the description available as the help_string output. 👑

The FI_LOG_LEVEL can be used to increase the debug output from libfabric and the providers. Note that in the release build of libfabric, debug output from data path operations (transmit, receive, and completion processing) may not be available. The FI_PROVIDER variable can be used to enable or disable specific providers. This is useful to ensure that a given provider will be used. ✨

🌸 AUTHORS 🌸

OpenFabrics.

---

💖 And that’s the whole fi_setup(7) glow-up, bestie — footer vibes: Libfabric Programmer’s Manual · 2024-12-31 · page fi_setup(7). 🎀