- 23 Jun, 2016 40 commits
-
-
Doug Ledford authored
-
Ira Weiny authored
Export the firmware version through the core. Acked-by:
Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by:
Ira Weiny <ira.weiny@intel.com> Signed-off-by:
Doug Ledford <dledford@redhat.com>
-
Ira Weiny authored
Now that all the devices have stopped exporting their own sysfs entry points we can have the core export this on their behalf. Eventually this may be removed but this provides for backwards compatibility. Reviewed-by:
-
Ira Weiny authored
Using this allows for devices to specify the format of their firmware version rather than forcing a format. Reviewed-by:
-
Ira Weiny authored
And remove sysfs file in favor of the common core. Reviewed-by:
-
Ira Weiny authored
And remove sysfs in favor of the core support. Reviewed-by:
-
Ira Weiny authored
And remove the sysfs in favor of the core version. Reviewed-by:
-
Ira Weiny authored
And remove the sysfs entry in favor of the core support. Reviewed-by:
-
Ira Weiny authored
And remove sysfs entry in favor of the common code. Reviewed-by:
-
Ira Weiny authored
And remove the sysfs in favor of common core version. Reviewed-by:
-
Ira Weiny authored
And remove sysfs support in favor of the core version. Reviewed-by:
-
Ira Weiny authored
And remove sysfs fw_ver in favor of the core. Reviewed-by:
Steve Wise <swise@opengridcomputing.com> Reviewed-by:
-
Ira Weiny authored
Also remove fw_ver sysfs to be replaced by the common core one. Reviewed-by:
-
Ira Weiny authored
Allow for a common core function to get firmware version strings from the individual devices. In later patches this format can then then be used to pass a properly formated version string through the IPoIB layer. The problem with the current code in the IPoIB layer is that it is specific to certain hardware types. Furthermore, this gives us a common function through which the core can provide a common sysfs entry. Eventually we may want to remove the sysfs export but this provides for user space backwards compatibility. Reviewed-by:
-
Mark Bloch authored
Expose new counters using the get protocol stats callback. We expose the following counters: |------------------------------------------------------------------------| | Name | IB | EN | Description | |------------------------------------------------------------------------| |rx_write_requests | + | - | Number of received WRITE requests for | | | | | the associated QP. | |------------------------------------------------------------------------| |rx_read_requests | + | - | Number of received READ requests for | | | | | the associated QP. | |------------------------------------------------------------------------| |rx_atomic_requests | + | - | Number of received ATOMIC requests for | | | | | the associated QP. | |------------------------------------------------------------------------| |out_of_buffer | + | + | Number of drops occurred due to lack | | | | | of WQE for the associated QPs/RQs. | |------------------------------------------------------------------------| |out_of_sequence | + | - | Number of errors in the packet | | | | | transport sequence number | |------------------------------------------------------------------------| |duplicate_request | + | + | Number of received duplicated packets. | | | | | A request that previously executed is | | | | | named duplicated. | |------------------------------------------------------------------------| |rnr_nak_retry_err | + | + | Number of received RNR NAC packets. | | | | | The QP retry limit did not exceed. | |------------------------------------------------------------------------| |packet_seq_err | + | + | Number of received NAK - sequence error| | | | | packets. The QP retry limit did not | | | | | exceed. | |------------------------------------------------------------------------| |implied_nak_err | + | + | Number of times the requester detected | | | | | an ACK with a PSN larger than expected | | | | | PSN for RDMA READ or ATOMIC response | | | | | The QP retry limit did not exceed. | |------------------------------------------------------------------------| |local_ack_timeout_err| + | - | Number of NO ACK responses from | | | | | responder within timer interval. | | | | | The QP retry limit did not exceed. | |------------------------------------------------------------------------| Counters are available if all of them are supported. Signed-off-by:
Mark Bloch <markb@mellanox.com> Signed-off-by:
Leon Romanovsky <leonro@mellanox.com> Signed-off-by:
Leon Romanovsky <leon@kernel.org> Reviewed-by:
Christoph Lameter <cl@linux.com> Signed-off-by:
-
Mark Bloch authored
In order to support statistics for ports, we attach each QP to a counter set which is dedicate to this port. Signed-off-by:
-
Artemy Kovalyov authored
Currently, the SRQ API uses the obsolete mlx5_*_srq_mbox_{in,out} structs which limit the ability to pass the SRQ attributes between net and IB parts of the driver. This patch changes the SRQ API so as to use auto-generated structs and provides a better way to pass attributes which will be in use by coming features. Signed-off-by:Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by:
-
Artemy Kovalyov authored
Make MODIFY_QP command input structure compliant to specification Fixes: e126ba97 ('mlx5: Add driver for Mellanox Connect-IB adapters') Signed-off-by:
-
Bodong Wang authored
Enable mlx5 based hardware to report TCP segmentation offload (TSO) capabilities from kernel to user space. A TSO enabled NIC will accept big chunks of data with sizes greater than MTU for TCP traffic. The TSO engine will break the data into separate packets and will insert headers automatically. The capabilities are exposed to user space through query_device by uhw directly. The following capabilities are reported: 1. The maximum payload size in bytes supported for segmentation by TSO engine. 2. Bitmap showing which QP types are supported by TSO operation. The bitmap is built by members from 'enmu ib_qp_type'. For example, similar code should be performed if UD QP is supported: supported_qpts |= 1 << IB_QPT_UD; To make user-space library aware of whether kernel supports uhw or not, a new flag: cmds_supp_uhw will be returned back to user-space through alloc_ucontext. Signed-off-by:
Bodong Wang <bodong@mellanox.com> Signed-off-by:
-
Maor Gottlieb authored
Enable flow steering for IPv6 traffic by using an IPv6 spec. Signed-off-by:
Maor Gottlieb <maorg@mellanox.com> Signed-off-by:
-
Maor Gottlieb authored
Add IPv6 flow specification support. Signed-off-by:
-
Maor Gottlieb authored
The driver exposes interfaces that directly relate to HW state. Upon fatal error, consumers of these interfaces (ULPs) that rely on completion of all their posted work-request could hang, thereby introducing dependencies in shutdown order. To prevent this from happening, we manage the relevant resources (CQs, QPs) that are used by the device. Upon a fatal error, we now generate simulated completions for outstanding WQEs that were not completed at the time the HW was reset. It includes invoking the completion event handler for all involved CQs so that the ULPs will poll those CQs. When polled we return simulated CQEs with IB_WC_WR_FLUSH_ERR return code enabling ULPs to clean up their resources and not wait forever for completions upon receiving remove_one. The above change requires an extra check in the data path to make sure that when device is in error state, the simulated CQEs will be returned and no further WQEs will be posted. Signed-off-by:
-
Maor Gottlieb authored
Implements the IB core disassociate_ucontext API. The driver detaches the HW resources for a given user context to prevent a dependency between application termination and device disconnect. This is done by managing the VMAs that were mapped to the HW bars such as doorbell and blueflame. When need to detach, remap them to an arbitrary kernel page returned by the zap API. Signed-off-by:
-
Yishai Hadas authored
Add support for Raw Ethernet RX HASH QP. Currently, creation and destruction of such a QP are supported. This QP is implemented as a simple TIR object which points to the receive RQ indirection table. The given hashing configuration is used to configure the TIR and by that it chooses the right RQ from the RQ indirection table. Signed-off-by:
Yishai Hadas <yishaih@mellanox.com> Signed-off-by:
Matan Barak <matanb@mellanox.com> Reviewed-by:
Sagi Grimberg <sagi@grimberg.me> Signed-off-by:
-
Yishai Hadas authored
User applications that want to spread incoming traffic between several WQs should create a QP which contains an indirection table. When such a QP is created other receive side parameters are not valid and should not be given. Its send side is optional and assumed active based on max_send_wr capability value. Extend create QP to work accordingly. Signed-off-by:
-
Yishai Hadas authored
Extend create QP to get Receive Work Queue (WQ) indirection table. QP can be created with external Receive Work Queue indirection table, in that case it is ready to receive immediately. Signed-off-by:
-
Yishai Hadas authored
Some mlx5 based hardwares support a RQ table object. This RQ table points to a few RQ objects. We implement the receive work queue indirection table API (create and destroy) by using this hardware object. Signed-off-by:
-
Yishai Hadas authored
User applications that want to spread traffic on several WQs, need to create an indirection table, by using already created WQs. Adding uverbs API in order to create and destroy this table. Signed-off-by:
-
Yishai Hadas authored
Introduce Receive Work Queue (WQ) indirection table. This object can be used to spread incoming traffic to different receive Work Queues. A Receive WQ indirection table points to variable size of WQs. This table is given to a QP in downstream patches. Signed-off-by:
Sagi Grimberg <sagi@grimerg.me> Signed-off-by:
-
Yishai Hadas authored
A QP can be created without internal WQs "packaged" inside it, this QP can be configured to use "external" WQ object as its receive/send queue. WQ is a necessary component for RSS technology since RSS mechanism is supposed to distribute the traffic between multiple Receive Work Queues Receive WQs are implemented by RQs. Implement the WQ creation, modification and destruction verbs. Signed-off-by:
-
Yishai Hadas authored
User space applications which use RSS functionality need to create a work queue object (WQ). The lifetime of such an object is: * Create a WQ * Modify the WQ from reset to init state. * Use the WQ (by downstream patches). * Destroy the WQ. These commands are added to the uverbs API. Signed-off-by:
Sagi Grimberg <sagi@rimberg.me> Signed-off-by:
-
Yishai Hadas authored
Introduce Work Queue object and its create/destroy/modify verbs. QP can be created without internal WQs "packaged" inside it, this QP can be configured to use "external" WQ object as its receive/send queue. WQ is a necessary component for RSS technology since RSS mechanism is supposed to distribute the traffic between multiple Receive Work Queues. WQ associated (many to one) with Completion Queue and it owns WQ properties (PD, WQ size, etc.). WQ has a type, this patch introduces the IB_WQT_RQ (i.e.receive queue), it may be extend to others such as IB_WQT_SQ. (send queue). WQ from type IB_WQT_RQ contains receive work requests. PD is an attribute of a work queue (i.e. send/receive queue), it's used by the hardware for security validation before scattering to a memory region which is pointed by the WQ. For that, an external WQ object needs a PD, letting the hardware makes that validation. When accessing a memory region that is pointed by the WQ its PD is used and not the QP's PD, this behavior is similar to a SRQ and a QP. WQ context is subject to a well-defined state transitions done by the modify_wq verb. When WQ is created its initial state becomes IB_WQS_RESET. >From IB_WQS_RESET it can be modified to itself or to IB_WQS_RDY. >From IB_WQS_RDY it can be modified to itself, to IB_WQS_RESET or to IB_WQS_ERR. >From IB_WQS_ERR it can be modified to IB_WQS_RESET. Note: transition to IB_WQS_ERR might occur implicitly in case there was some HW error. Signed-off-by:
-
Yishai Hadas authored
In order to support RSS QPs, we need to create Ethernet based objects. This is done by create_rq, destroy_rq, create_rqt and destroy_rqt mlx5_core functions. We export these functions. Signed-off-by:
-
Hariprasad S authored
Pre-allocate buffers to deallocate completion queue, so that completion queue is deallocated during RDMA termination when system is running out of memory. Signed-off-by:
Hariprasad Shenai <hariprasad@chelsio.com> Signed-off-by:
-
Hariprasad S authored
Pre-allocate buffers for deregistering memory region and memory window during RDMA connection close, when system is running out of memory. Signed-off-by:
-
Hariprasad S authored
Pre-allocate buffers for sending various control messages to close connection, abort connection, etc so that we gracefully handle connections when system is running out of memory. Signed-off-by:
-
Hariprasad S authored
Signed-off-by:
-
Hariprasad S authored
Get rid of unneeded code, and refactor things a bit. For MPA version 0 we abort the connection. For > 0, we attempt to send an MPA_START/REJECT Reply, and then disconnect gracefully. If the send of the MPA message fails, then we abort the connection. We can ignore c4iw_ep_disconnect() errors here because it will clean up the endpoint if there are failures. Signed-off-by:
-
Hariprasad S authored
With IPv6 addresses, the "qps" debugfs is running out of space and truncating the output. Bump the required size accordingly. Signed-off-by:
-
Hariprasad S authored
markers_enabled should be read only once during MPA negotiation. The present code does read markers_enabled twice during negotiation which results in setting wrong recv/xmit markers if the markers_enabled is changed in the middle of negotiation. With this change the markers_enabled is read only once during MPA negotiation. recv markers are set based on markers enabled module parameter and xmit markers are set based on markers flag from the MPA_START_REQ/MPA_START_REP. Signed-off-by:
-
