--- name: "rds-rdma" section: "7" platform: "linux" package: "rds-tools" version: "1.4.1-OFED-1.4.2-1" release: "Debian 7 (Wheezy)" platforms: ["linux"] url: "https://new.carta.tech/rds-rdma" --- # rds-rdma(7) Zerocopy Interface for RDMA over RDS ## NAME RDS-rdma - Zerocopy Interface for RDMA over RDS ## DESCRIPTION This manual page describes the zerocopy interface of RDS, which was added in RDSv3. For a description of the basic RDS interface, please refer to [**rds**(7)](/rds/7.md). The principal mode of operation for RDS zerocopy is like this: one participant (the client) wishes to initiate a direct transfer to or from some area of memory in its process address space. This memory does not have to be aligned. The client obtains a handle for this region of memory, and passes it to the other participant (the server). This is called the RDMA cookie. To the application, the cookie is an opaque 64bit data type. The client sends this handle to the server application, along with other details of the RDMA request (such as which data to transfer to that memory area). Throughout the following discussion, we will refer to this message as the RDMA request. The server uses this RDMA cookie to initiate the requested RDMA transfer. The RDMA transfer is combined atomically with a normal RDS message, which is delivered to the client. This message is called the RDMA ACK throughout the following. Atomic in this context means that either both the RDMA succeeds and the RDMA ACK is delivered, or neither succeeds. Thus, when the client receives the RDMA ACK, it knows that the RDMA has completed successfully. It can then release the RDMA cookie for this memory region, if it wishes to. RDMA operations are not reliable, in the sense that unlike normal RDS messages, RDS RDMA operations may fail, and get dropped. ## INTERFACE The interface is currently based on control messages (ancillary data) sent or received via the [**sendmsg**(2)](/sendmsg/2.md) and [**recvmsg**(2)](/recvmsg/2.md) system calls. Optionally, an older interface can be used that is based on the [**setsockopt**(2)](/setsockopt/2.md) system call. However, we recommend using control messages, as this reduces the number of system calls required. ### Control message interface With the control message interface, the RDMA cookie is passed to the server out-of-band, included in an extension header attached to the RDS message. The following outlines the mode of operation; the data types used will be specified in details in a subsequent section. Initially, the client will send RDMA requests along with a **RDS\_CMSG\_RDMA\_MAP** control message. The control message contains the address and length of the memory region for which to obtain a handle, some flags, and a pointer to a memory location (in the caller's address space) where the kernel will store the RDMA cookie. Alternatively, if the application has already obtained a RDMA cookie for the memory range it wants to RDMA to/from, it can hand this cookie to the kernel using the **RDS\_CMSG\_RDMA\_DEST** control message. Either way, the kernel will include the resulting RDMA cookie in an extension header that is transmitted as part of the RDMA request to the server. When the server receives the RDMA request, the kernel will deliver the cookie wrapped inside a **RDS\_CMSG\_RDMA\_DEST** control message. The server then initiates the data transfer by sending the RDMA ACK message along with a **RDS\_CMSG\_RDMA\_ARGS** control message. This message contains the RDMA cookie, and the local memory to copy to or from. The server process may request a notification when an RDMA operation completes. Notifications are delivered as a **RDS\_CMSG\_RDMA\_STATUS** control messages. When an application calls [**recvmsg**(2)](/recvmsg/2.md), it will either receive a regular RDS message (possibly with other RDMA related control messages), or an empty message with one or more status control messages. In addition, applications When an RDMA operation fails for some reason and is discarded, the application can ask to receive notifications for failed messages as well, regardless of whether it asked for success notification of an individual message or not. This behavior is turned on by setting the **RDS\_RECVERR** socket option. ### Setsockopt interface In addition to the control message interface, RDS allows a process to register and release memory ranges for RDMA through calls to [**setsockopt**(2)](/setsockopt/2.md). - **RDS\_GET\_MR** To obtain a RDMA cookie for a given memory range, the application can use **setsockopt** with **RDS\_GET\_MR**. This operates essentially the same way as the **RDS\_CMSG\_RDMA\_MAP** control message: the argument contains the address and length of the memory range to be registered, and a pointer to a RDMA cookie variable, in which the system call will store the cookie for the registered range. - **RDS\_FREE\_MR** Memory ranges can be released by calling **setsockopt** with **RDS\_FREE\_MR**, giving the RDMA cookie and additional flags as arguments. - **RDS\_RECVERR** This is a boolean option which can be set as well as queried (using **getsockopt**). When enabled, RDS will send RDMA notification messages to the application for any RDMA operation that fails. This option defaults to off. For all of these calls, the **level** argument to **setsockopt** is **SOL\_RDS**. ## RDMA MACROS AND TYPES - **RDMA cookie** ``` typedef u_int64_t rds_rdma_cookie_t ``` - This encapsulates a memory location in the client process. In the current implementation, it contains the R\_Key of the remote memory region, and the offset into it (so that the application does not have to worry about alignment. - The RDMA cookie is used in several struct types described below. The **RDS\_CMSG\_RDMA\_DEST** control message contains a rds\_rdma\_cookie\_t all by itself as payload. - **Mapping arguments** The following data type is used with **RDS\_CMSG\_RDMA\_MAP** control messages and with the **RDS\_GET\_MR** socket option: - ``` struct rds_iovec { u_int64_t addr; u_int64_t bytes; }; struct rds_get_mr_args { struct rds_iovec vec; u_int64_t cookie_addr; uint64_t flags; }; ``` - The **cookie\_addr** specifies a memory location where to store the RDMA cookie. - The **flags** value is a bitwise OR of any of the following flags: - **RDS\_RDMA\_USE\_ONCE** This tells the kernel that the allocated RDMA cookie is to be used exactly once. When the RDMA ACK message arrives, the kernel will automatically unbind the memory area and release any resources associated with the cookie. - If this flag is not set, it is the application's responsibility to release the memory region at a later time using the **RDS\_FREE\_MR** socket option. - **RDS\_RDMA\_INVALIDATE** Normally, RDMA memory mappings are invalidated lazily, as this requires some relatively costly synchronization with the HCA. However, this means that the server application can continue to access the registered memory for some indeterminate amount of time. If this flag is set, the RDS code will invalidate the mapping at the time it is released (either upon arrival of the RDMA ACK, if **USE\_ONCE** was specified; or when the application destroys it using **FREE\_MR**). - **RDMA Operation** RDMA operations are initiated by the server using the **RDS\_CMSG\_RDMA\_ARGS** control message, which takes the following data as payload: - ``` struct rds_rdma_args { rds_rdma_cookie_t cookie; struct rds_iovec remote_vec; u_int64_t local_vec_addr; u_int64_t nr_local; u_int64_t flags; u_int32_t user_token; }; ``` - The **cookie** argument contains the RDMA cookie received from the client. The local memory is given via an array of **rds\_iovec**s. The array address is given in **local\_vec\_addr**, and its number of elements is given in **nr\_local**. - The struct member **remote\_vec** specifies a location relative to the memory area identified by the cookie: **remote\_vec**.**addr** is an offset into that region, and **remote\_vec**.**bytes** is the length of the memory window to copy to/from. This length must match the size of the local memory area, i.e. the sum of bytes in all members of the local iovec. - The flags field contains the bitwise OR of any of the following flags: - **RDS\_RDMA\_READWRITE** If set, any RDMA WRITE is initiated from the server's memory to the client's. If not set, RDS will do a RDMA READ from the client's memory to the server's memory. - **RDS\_RDMA\_FENCE** By default, Infiniband makes no guarantee about the ordering of an RDMA READ with respect to subsequent SEND operations. Setting this flag asks that the RDMA READ should be fenced off the subsequent RDS ACK message. Setting this flag requires an additional round-trip of the IB fabric, but it is a good idea to use set this flag by default, unless you are really sure you do not want it. - **RDS\_RDMA\_NOTIFY\_ME** This flag requests a notification upon completion of the RDMA operation (successful or otherwise). The noticiation will contain the value of the **user\_token** field passed in by the application. This allows the application to release resources (such as buffers) assosicated with the RDMA transfer. - The **user\_token** can be used to pass an application specific identifier to the kernel. This token is returned to the application when a status notification is generated (see the following section). - **RDMA Notification** The RDS kernel code is able to notify the server application when an RDMA operation completes. These notifications are delivered via **RDS\_CMSG\_RDMA\_STATUS** control messages. - By default, no notifications are generated. There are two ways an application can request them. On one hand, status notifications can be enabled on a per-operation basis by setting the **RDS\_RDMA\_NOTIFY\_ME** flag in the RDMA arguments. On the other hand, the application can request notifications for all RDMA operations that fail by setting the **RDS\_RECVERR** socket option (see below). In both cases, the format of the notification is the same; and at most one notification will be sent per completed operation. - The message format is this: - ``` struct rds_rdma_notify { u_int32_t user_token; int32_t status; }; ``` - The **user\_token** field contains the value previously given to the kernel in the **RDS\_CMSG\_RDMA\_ARGS** control message. The **status** field contains a status value, with 0 indicating success, and non-zero indicating an error. - The following status codes are currently defined: - **RDS\_RDMA\_SUCCESS** The RDMA operation succeeded. - **RDS\_RDMA\_REMOTE\_ERROR** The RDMA operation failed due to a remote access error. This is usually due to an invalid R\_key, offset or transfer size. - **RDS\_RDMA\_CANCELED** The RDMA operation was canceled by the application. (This error code is not yet generated). - **RDS\_RDMA\_DROPPED** RDMA operations were discarded after the connection broke and was re-established. The RDMA operation may have been processed partially. - **RDS\_RDMA\_OTHER\_ERROR** Any other failure. - **RDMA setsockopt arguments** When using the **RDS\_GET\_MR** socket option to register a memory range, the application passes a pointer to a **struct rds\_get\_mr\_args** variable, described above. - The **RDS\_FREE\_MR** call takes an argument of type **struct rds\_free\_mr\_args**: - ``` struct rds_free_mr_args { rds_rdma_cookie_t cookie; u_int64_t flags; }; ``` - **cookie** specifies the RDMA cookie to be released. RDMA access to the memory range will usually not be invoked instantly, because the operation is rather costly. However, if the **flags** argument contains **RDS\_RDMA\_INVALIDATE**, RDS will invalidate the indicated mapping immediately, as described in section **Mapping arguments** above. - If the **cookie** argument is 0, and **RDS\_RDMA\_INVALIDATE** is set, RDS will invalidate old memory mappings on all devices. ## ERRORS In addition to the usual error codes returned by **sendmsg**, **recvmsg** and **setsockopt**, RDS returns the following error codes: - **EAGAIN** RDS was unable to map a memory range because the limit was exceeded (returned by **RDS\_CMSG\_RDMA\_MAP** and **RDS\_GET\_MR**). - **EINVAL** When sending a message, there were were conflicting control messages (e.g. two **RDMA\_MAP** messages, or a **RDMA\_MAP and a RDMA\_DEST** message). - In a **RDS\_CMSG\_RDMA\_MAP** or **RDS\_GET\_MR** operation, the application specified memory range greater than the maximum size supported. - When setting up an RDMA operation with **RDS\_CMSG\_RDMA\_ARGS**, the size of the local memory (given in the **rds\_iovec**) did not match the size of the remote memory range. - **EBUSY** RDS was unable to obtain a DMA mapping for the indicated memory. ## LIMITS Currently, the following limits apply - • The maximum size of a zerocopy transfer is 1MB. This can be adjusted via the **fmr\_message\_size** module parameter. - • The maximum number of memory ranges that can be mapped is limited to 2048 at the moment. This can be adjusted via the **fmr\_pool\_size** module parameter. However, the actual limit imposed by the hardware may in fact be lower. ## AUTHORS RDS was written and is Copyright (C) 2007-2008 by Oracle, Inc.