- 11 Mar, 2016 40 commits
-
-
Mike Marciniszyn authored
Implement per-VL transmit counters. Not all errors can be attributed to a particular VL, so make a best attempt. o Extend the egress error bits used to count toward transmit discard. o When an egress error or send error occur, try to map back to a VL. o Implement a SDMA engine to VL (back) map. o Add per-VL port transmit counters Reviewed-by:
Mike Marciniszyn <mike.marciniszyn@intel.com> Reviewed-by:
Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by:
Dean Luick <dean.luick@intel.com> Signed-off-by:
Doug Ledford <dledford@redhat.com>
-
Dean Luick authored
The gen3 bump code must mark a firmware download failure as fatal. Otherwise a later load attempt will fail with a NULL dereference. Also: o Only do a firmware back-off for RTL. There are no alternates for FPGA or simulation. o Rearrange OS firmware request order to match what is actually loaded. This results in more coherent informational messages in the case of missing firmware. Reviewed-by:
-
Easwar Hariharan authored
This patch implements support for turning on and off the clock data recovery mechanisms implemented in QSFP cable on request by the DC 8051 on a per-lane basis. Reviewed-by:
Easwar Hariharan <easwar.hariharan@intel.com> Signed-off-by:
-
Easwar Hariharan authored
The current code employs a heuristic to guess the port type. The canonical location to identify the port type of the designed platform is from the platform configuration data. This patch uses the previously fetched port type from the platform configuration and removes the now obsolete heuristic routine and its associated defines. Reviewed-by:
Arthur Kepner <arthur.kepner@intel.com> Signed-off-by:
-
Easwar Hariharan authored
This patch qualifies and tunes active and optical cables for optimal bit error rate and signal integrity settings. These settings are fetched from the platform configuration data. Based on attributes of the QSFP cable as read from the SFF-8636 compliant memory map, we select the appropriate settings from the platform configuration data (examples: TX/RX equalization, enabling cable high power, enabling TX/RX clock data recovery mechanisms, and RX amplitude control) and apply them to the SERDES and QSFP cable. The platform configuration data also contains system parameters such as maximum power dissipation supported, and the cables are qualified based on these parameters. As part of qualifying the cables, the correct OfflineDisabledReasons are set for the appropriate scenarios. Reviewed-by:
Brent R Rothermel <brent.r.rothermel@intel.com> Signed-off-by:
Jubin John <jubin.john@intel.com> Signed-off-by:
-
Mike Marciniszyn authored
The QSFP memory cache reads both lower and upper page 0H in one shot, which leads to the address counter wrapping around to the beginning of lower page 00H at byte 128, as defined by SFF-8636. This patch fixes this by modifying the underlying QSFP read and writes to avoid this wrap around. Reviewed-by:
Ira Weiny <ira.weiny@intel.com> Signed-off-by:
-
Easwar Hariharan authored
The ":" in "%s:" adds no value. Reviewed-by:
-
Bryan Morgan authored
Removing QSFP cable should report 'No Local Media' instead of 'Transient' as reported by 'opaportinfo'. Workaround is to change the state to OPA_LINKDOWN_REASON_LOCAL_MEDIA_NOT_INSTALLED in cable handler. With cable still removed, 'opaportinfo bounce' should not cause a state change to Polling, as reported by 'opaportinfo'. Resolution is to prevent physical state change from Offline->Polling. Use a macro to mask lower nibble of OPA_LINKDOWN_REASON* as needed for offline_disabled_reason. Reviewed-by:
Todd Rimmer <todd.rimmer@intel.com> Signed-off-by:
Bryan Morgan <bryan.c.morgan@intel.com> Signed-off-by:
-
Jubin John authored
srq functionality is now in rdmavt. Remove it from the hfi1 driver. Reviewed-by:
Harish Chegondi <harish.chegondi@intel.com> Signed-off-by:
-
Harish Chegondi authored
Rely on rvt_query_qp function defined in rdmavt Reviewed-by:
-
Dennis Dalessandro authored
Get rid of create and free mad agent from the driver and use rdmavt version. Reviewed-by:
-
Dennis Dalessandro authored
No longer do drivers need to call into the IB core to allocate the verbs device. Use the functionality provided by rdmavt. Reviewed-by:
-
Dennis Dalessandro authored
Now that rdmavt has solidified in its design we can clean up the driver specific register device functions. This handles hfi1. Reviewed-by:
-
Dennis Dalessandro authored
This patch removes the simple post recv function in favor of using rdmavt. The packet receive processing still lives in the driver though. Reviewed-by:
-
Dennis Dalessandro authored
This removes the destroy qp verbs in favor of using rdmavt. Reviewed-by:
-
Dennis Dalessandro authored
In addition to removing the modify queue pair verb from hfi1 we also remove ancillary functions which existed only for modify queue pair and are also already present in hfi1. Reviewed-by:
-
Dennis Dalessandro authored
Multicast is now supported by rdmavt. Remove the verbs multicast functions and use that. Reviewed-by:
-
Dennis Dalessandro authored
This patch removes the post_send and post_one_send from the hfi1 driver. The "posting" of sends will be done by rdmavt which will walk a WQE and queue work. This patch will still provide the capability to schedule that work as well as kick the progress. These are provided to the rdmavt layer. Reviewed-by:
Edward Mascarenhas <edward.mascarenhas@intel.com> Signed-off-by:
-
Dennis Dalessandro authored
Return directly from rvt_resize_cq rather than use a goto/label. Reviewed-by:
-
Dennis Dalessandro authored
The completion queue is not a complex data structure and it can be removed at the same time as its functions. Unlike the more complicated queue pair which was done in multiple patches. This single patch removes all traces of hfi1 specific completeion queues from the hfi1 driver. Reviewed-by:
-
Harish Chegondi authored
Removed hfi1 query_device function to use rdmavt rvt_query_device function The rvt dev info device attributes still need to be filled in by the driver Reviewed-by:
-
Dennis Dalessandro authored
Rely on rdmavt to provide queue pair creation. Reviewed-by:
-
Dennis Dalessandro authored
Another change on the way to removing queue pair functionality from hfi1. This patch removes the private queue pair structure and the table which holds the queue pair numbers in favor of using what is provided by rdmavt. Reviewed-by:
-
Dennis Dalessandro authored
Use the definitions of the s_flags and r_flags which are now in rdmavt. Reviewed-by:
-
Harish Chegondi authored
IB user context alloc and dealloc functions have been added to rdmavt. This patch removes them from hfi1. Reviewed-by:
-
Dennis Dalessandro authored
No need to keep providing the query pkey function. This is now being done in rdmavt. Remove support from hfi1. The allocation and maintenance of the list still resides in the driver. Reviewed-by:
-
Dennis Dalessandro authored
Mmap data structure has already been moved to rdmavt and hfi1 supports it. Now that the mmap functionality has also been moved to rdmavt its time for hfi1 to use that as well. Reviewed-by:
-
Dennis Dalessandro authored
Remove most of the ibport members from hfi1 and use the rdmavt version. Also register the port with rdmavt. Reviewed-by:
-
Dennis Dalessandro authored
SRQ data structure has been moved to rdmavt. Make use of it. Reviewed-by:
-
Dennis Dalessandro authored
This patch does the actual removal of the queue pair from the hfi1 driver along with a number of dependent data structures. These were moved to rvt. It also removes the MR functions to use those in rdmavt. These two pieces can not reasonably be split apart becuase they depend on each other. Reviewed-by:
-
Dennis Dalessandro authored
For OPA devices additional work is required to create an AH. This patch adds support to set the VL correctly. Reviewed-by:
-
Dennis Dalessandro authored
Original patch from Kamal Heib <kamalh@mellanox.com>, split apart from original and modified to accomodate recent changes in rdmavt. Remove AH from hfi1 and use rdmavt version. Signed-off-by:
Kamal Heib <kamalh@mellanox.com> Signed-off-by:
-
Dennis Dalessandro authored
Rdmavt split the header files to be based on ibta object. This patch makes changes in hfi1 to account for the move. Reviewed-by:
-
Dennis Dalessandro authored
Implement get_card_name and get_pci_dev helper functions for rdmavt for hfi1. Reviewed-by:
-
Dennis Dalessandro authored
In preparation for moving the queue pair data structure to rdmavt the members of the driver specific queue pairs which are not common need to be pushed off to a private driver structure. This structure will be available in the queue pair once moved to rdmavt as a void pointer. This patch while not adding a lot of value in and of itself is a prerequisite to move the queue pair out of the drivers and into rdmavt. The driver specific, private queue pair data structure should condense as more of the send side code moves to rdmavt. Reviewed-by:
-
Dennis Dalessandro authored
Remove MR data structures from hfi1 and use the version in rdmavt Reviewed-by:
-
Dennis Dalessandro authored
Remove protection domain from hfi1 and use rdmavt's version. Reviewed-by:
-
Ira Weiny authored
Remove the dma.c file from hfi1 in favor of using that which is present in rdmavt. Reviewed-by:
-
Dennis Dalessandro authored
Most functionality is still being done in the driver, set flags so that rdmavt will let hfi1 continue to handle mr, qp, and cq init. Reviewed-by:
-
Dennis Dalessandro authored
This patch begins to make use of rdmavt by registering with it and providing access to the header files. This is just the beginning of rdmavt support in hfi1. Reviewed-by:
-
