[PATCH 19/19] IB/mad: Implement Intel Omni-Path Architecture MAD processing

2015-01-26 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com For devices which support OPA MADs OPA SMP packets must carry a valid pkey process wc.pkey_index returned by agents for response. Handle variable length OPA MADs by: * Adjusting the 'fake' WC for locally routed SMP's to represent

[PATCH 16/19] IB/mad: Add Intel Omni-Path Architecture defines

2015-01-26 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com OPA_SMP_CLASS_VERSION -- Defined at 0x80 OPA_MGMT_BASE_VERSION -- Defined at 0x80 Increase max management version to accommodate OPA Signed-off-by: Ira Weiny ira.we...@intel.com --- Changes since V2: Remove OPA_MIN_CLASS_VERSION drivers/infiniband

[PATCH 18/19] IB/mad: Implement Intel Omni-Path Architecture SMP processing

2015-01-26 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Define the new OPA SMP format, create support functions for this format using the previously defined helper functions as appropriate. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad_priv.h | 2 + drivers/infiniband/core

[PATCH 08/19] IB/mad: Add helper function for smi_handle_dr_smp_send

2015-01-26 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This helper function will be used for processing both IB and OPA SMP sends. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/smi.c | 81 +-- 1 file changed, 47 insertions(+), 34 deletions

[PATCH 11/19] IB/mad: Add helper function for SMI processing

2015-01-26 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This helper function will be used for processing both IB and OPA SMPs. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad.c | 85 +-- 1 file changed, 49 insertions(+), 36 deletions(-) diff

[PATCH 06/19] IB/core: Add max_mad_size to ib_device_attr

2015-01-26 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Change all IB drivers to report the max MAD size. Add check to verify that all devices support at least IB_MGMT_MAD_SIZE Signed-off-by: Ira Weiny ira.we...@intel.com --- Changes from V2: attributes - cached_dev_attrs Change variable from

[PATCH 05/19] IB/mad: Change cast in rcv_has_same_class

2015-01-26 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com rcv_has_same_class only needs access to the MAD header and can be used for both IB and Jumbo MADs. Signed-off-by: Ira Weiny ira.we...@intel.com --- Changes from v2: reword commit message drivers/infiniband/core/mad.c | 2 +- 1 file changed, 1

[PATCH 14/19] IB/core: Add IB_DEVICE_OPA_MAD_SUPPORT device cap flag

2015-01-26 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com OPA MADs share a common header with IBTA MADs but with a different base version and an extended length. These jumbo MADs increase the performance of management traffic. Sharing a common header with IBTA MADs allows us to share most of the MAD processing code

[PATCH 17/19] IB/mad: Implement support for Intel Omni-Path Architecture base version MADs in ib_create_send_mad

2015-01-26 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com If the device supports OPA MADs process the OPA base version. Set MAD size and sg lengths as appropriate Split RMPP MADs as appropriate Signed-off-by: Ira Weiny ira.we...@intel.com --- Changes from V2: attributes - cached_dev_attrs

[PATCH 15/19] IB/mad: Create jumbo_mad data structures

2015-01-26 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Define jumbo_mad and jumbo_rmpp_mad. Jumbo MAD structures are 2K versions of ib_mad and ib_rmpp_mad structures. Currently only OPA base version MADs are of this type. Create an RMPP Base header to share between ib_rmpp_mad and jumbo_rmpp_mad Update existing

[PATCH 10/19] IB/mad: Add helper function for smi_check_forward_dr_smp

2015-01-26 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This helper function will be used for processing both IB and OPA SMPs. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/smi.c | 26 +- 1 file changed, 17 insertions(+), 9 deletions(-) diff --git a/drivers

[PATCH 09/19] IB/mad: Add helper function for smi_handle_dr_smp_recv

2015-01-26 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This helper function will be used for processing both IB and OPA SMP recvs. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/smi.c | 80 +-- 1 file changed, 47 insertions(+), 33 deletions

[PATCH v4 01/19] IB/mad: Rename is_data_mad to is_rmpp_data_mad

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com is_rmpp_data_mad is more descriptive for this function. Reviewed-by: Sean Hefty sean.he...@intel.com Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad.c | 7 --- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git

[PATCH v4 16/19] IB/mad: Add Intel Omni-Path Architecture defines

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com OPA_SMP_CLASS_VERSION -- Defined at 0x80 OPA_MGMT_BASE_VERSION -- Defined at 0x80 Increase max management version to accommodate OPA Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad_priv.h | 2 +- include/rdma/ib_mad.h

[PATCH v4 00/19] IB/mad: Add support for Intel Omni-Path Architecture (OPA) MAD processing.

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com The following patch series modifies the kernel MAD processing (ib_mad/ib_umad) and related interfaces to send and receive Intel Omni-Path Architecture MADs on devices which support them. In addition to supporting some IBTA management classes, OPA devices use

[PATCH v4 15/19] IB/mad: Create jumbo_mad data structures

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Define jumbo_mad and jumbo_rmpp_mad. Jumbo MAD structures are 2K versions of ib_mad and ib_rmpp_mad structures. Currently only OPA base version MADs are of this type. Create an RMPP Base header to share between ib_rmpp_mad and jumbo_rmpp_mad Update existing

[PATCH v4 11/19] IB/mad: Add helper function for SMI processing

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This helper function will be used for processing both IB and OPA SMPs. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad.c | 85 +-- 1 file changed, 49 insertions(+), 36 deletions(-) diff

[PATCH v4 14/19] IB/core: Add IB_DEVICE_OPA_MAD_SUPPORT device cap flag

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com OPA MADs share a common header with IBTA MADs but with a different base version and an extended length. These jumbo MADs increase the performance of management traffic. Sharing a common header with IBTA MADs allows us to share most of the MAD processing code

[PATCH v4 10/19] IB/mad: Add helper function for smi_check_forward_dr_smp

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This helper function will be used for processing both IB and OPA SMPs. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/smi.c | 26 +- 1 file changed, 17 insertions(+), 9 deletions(-) diff --git a/drivers

[PATCH v4 08/19] IB/mad: Add helper function for smi_handle_dr_smp_send

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This helper function will be used for processing both IB and OPA SMP sends. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/smi.c | 81 +-- 1 file changed, 47 insertions(+), 34 deletions

[PATCH v4 17/19] IB/mad: Implement support for Intel Omni-Path Architecture base version MADs in ib_create_send_mad

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com If the device supports OPA MADs process the OPA base version. Set MAD size and sg lengths as appropriate Split RMPP MADs as appropriate Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad.c | 38

[PATCH v4 18/19] IB/mad: Implement Intel Omni-Path Architecture SMP processing

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Define the new OPA SMP format, create support functions for this format using the previously defined helper functions as appropriate. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad_priv.h | 2 + drivers/infiniband/core

[PATCH v4 19/19] IB/mad: Implement Intel Omni-Path Architecture MAD processing

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com For devices which support OPA MADs OPA SMP packets must carry a valid pkey process wc.pkey_index returned by agents for response. Handle variable length OPA MADs by: * Adjusting the 'fake' WC for locally routed SMP's to represent

[PATCH v4 02/19] IB/core: Cache device attributes for use by upper level drivers

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Upper level drivers can access these cached device attributes rather than caching them on their own. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/device.c | 2 ++ include/rdma/ib_verbs.h | 1 + 2 files changed, 3

[PATCH v4 05/19] IB/mad: Change cast in rcv_has_same_class

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com rcv_has_same_class only needs access to the MAD header and can be used for both IB and Jumbo MADs. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers

[PATCH v4 12/19] IB/mad: Add MAD size parameters to process_mad

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com In support of variable length MADs add in and out MAD size parameters to the process_mad call. The out MAD size parameter is passed by reference such that it can be updated by the agent to indicate the proper response length to be sent by the MAD stack

[PATCH v4 07/19] IB/mad: Convert ib_mad_private allocations from kmem_cache to kmalloc

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Use the new max_mad_size specified by devices for the allocations and DMA maps. kmalloc is more flexible to support devices with different sized MADs and research and testing showed that the current use of kmem_cache does not provide performance benefits over

[PATCH v4 06/19] IB/core: Add max_mad_size to ib_device_attr

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Change all IB drivers to report the max MAD size. Add check to verify that all devices support at least IB_MGMT_MAD_SIZE Signed-off-by: Ira Weiny ira.we...@intel.com --- Changes from V3: Fix ehca compile found with 0-day build drivers/infiniband

[PATCH v4 04/19] IB/mad: Change ib_response_mad signature to take ib_mad_hdr rather than ib_mad

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com ib_response_mad only needs access to the MAD header and can be used for both IB and Jumbo MADs. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad.c | 20 ++-- drivers/infiniband/core/user_mad.c | 6

[PATCH v4 03/19] IB/mad: Change validate_mad signature to take ib_mad_hdr rather than ib_mad

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com validate_mad only needs access to the MAD header and can be used for both IB and Jumbo MADs. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad.c | 12 ++-- 1 file changed, 6 insertions(+), 6 deletions(-) diff --git

[PATCH v4 13/19] IB/mad: Add base version parameter to ib_create_send_mad

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com In preparation to support the new OPA MAD Base version, add a base version parameter to ib_create_send_mad and set it to IB_MGMT_BASE_VERSION for current users. Definition of the new base version and it's processing will occur in later patches. Signed-off

[PATCH v4 09/19] IB/mad: Add helper function for smi_handle_dr_smp_recv

2015-02-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This helper function will be used for processing both IB and OPA SMP recvs. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/smi.c | 80 +-- 1 file changed, 47 insertions(+), 33 deletions

[PATCH 05/20] IB/mad: Change cast in rcv_has_same_class

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Save dereference and clarifies that rcv_has_same_class can process both IB and OPA MADs. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad.c |2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/drivers

[PATCH 07/20] IB/mad: Convert ib_mad_private allocations from kmem_cache to kmalloc

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Use the new mad_size specified by devices for the allocations and DMA maps. kmalloc is more flexible to support devices with different sized MADs and research and testing showed that the current use of kmem_cache does not provide performance benefits over

[PATCH 16/20] IB/mad: Add Intel Omni-Path Architecture defines

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com OPA_MIN_CLASS_VERSION -- OPA Class versions are 0x80 OPA_SMP_CLASS_VERSION -- Defined at 0x80 OPA_MGMT_BASE_VERSION -- Defined at 0x80 Increase max management version to accommodate OPA Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband

[PATCH 09/20] IB/mad: Add helper function for smi_handle_dr_smp_recv

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This helper function will be used for processing both IB and OPA SMP recvs. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/smi.c | 80 - 1 files changed, 47 insertions(+), 33 deletions

[PATCH 15/20] IB/mad: Create jumbo_mad data structures

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Define jumbo_mad and jumbo_rmpp_mad Create an RMPP Base header to share between ib_rmpp_mad and jumbo_rmpp_mad Update code to use the new structures. Signed-off-by: Ira Weiny ira.we...@intel.com --- Changes from RFC: add JUMBO_MGMT_MAD_SIZE

[PATCH 11/20] IB/mad: Add helper function for SMI processing

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This helper function will be used for processing both IB and OPA SMPs. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad.c | 85 +++- 1 files changed, 49 insertions(+), 36 deletions(-) diff

[PATCH 03/20] IB/mad: Change validate_mad signature to take ib_mad_hdr rather than ib_mad

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com validate_mad only needs access to the MAD header Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad.c | 12 ++-- 1 files changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/infiniband/core/mad.c b/drivers

[PATCH 14/20] IB/core: Add IB_DEVICE_OPA_MAD_SUPPORT device cap flag

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Add a device capability flag to flag OPA MAD support on devices. Signed-off-by: Ira Weiny ira.we...@intel.com --- Changes from RFC: Changed flag to indicate OPA MAD base version support Added device_cap_flags2 as device_cap_flags bits

[PATCH 08/20] IB/mad: Add helper function for smi_handle_dr_smp_send

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This helper function will be used for processing both IB and OPA SMP sends. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/smi.c | 81 - 1 files changed, 47 insertions(+), 34 deletions

[PATCH 10/20] IB/mad: Add helper function for smi_check_forward_dr_smp

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This helper function will be used for processing both IB and OPA SMPs. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/smi.c | 26 +- 1 files changed, 17 insertions(+), 9 deletions(-) diff --git a/drivers

[PATCH 20/20] IB/mad: Implement Intel Omni-Path Architecture MAD processing

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com OPA SMP packets must carry a valid pkey process wc.pkey_index returned by agents for response. Handle variable length OPA MADs based on the Base Version Support is provided by: * Adjusting the 'fake' WC for locally routed SMP's to represent

[PATCH 13/20] IB/mad: Add base version parameter to ib_create_send_mad

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com In preparation to support the new OPA MAD Base version, add a base version parameter ib_create_send_mad and set it to IB_MGMT_BASE_VERSION for current users. Definition of the new base version and it's processing will occur in later patches. Signed-off

[PATCH 12/20] IB/mad: Add MAD size parameters to process_mad

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com In support of variable length MADs add in and out MAD size parameters to the process_mad call. The out MAD size parameter is passed by reference such that it can be updated by the agent to indicate the proper response length to be sent by the MAD stack

[PATCH 19/20] IB/mad: Implement Intel Omni-Path Architecture SMP processing

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Define the new OPA SMP format, create support functions for this format using the previously defined helper functions as appropriate. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad_priv.h |2 + drivers/infiniband/core

[PATCH 18/20] IB/mad: Implement support for Intel Omni-Path Architecture base version MADs in ib_create_send_mad

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com If the MAD is an OPA base version; verify the device supports Jumbo MAD sizes Set MAD size and sg lengths as appropriate Split RMPP MADs as appropriate Signed-off-by: Ira Weiny ira.we...@intel.com --- Changes from RFC: remove use

[PATCH 17/20] IB/mad: Add registration check for Intel Omni-Path Architecture MADs

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com If the registration specifies an OPA MAD class version and the device does not support OPA MADs, fail the MAD registration. Signed-off-by: Ira Weiny ira.we...@intel.com --- Changes from RFC: remove use of supports_jumbo_mads convert to use

[PATCH 04/20] IB/mad: Change ib_response_mad signature to take ib_mad_hdr rather than ib_mad

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com ib_response_mad only needs access to the MAD header Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad.c | 20 ++-- drivers/infiniband/core/user_mad.c |6 +++--- include/rdma/ib_mad.h |2

[PATCH 00/20] IB/mad: Add support for Intel Omni-Path Architecture (OPA) MAD processing.

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com The following patch series modifies the kernel MAD processing (ib_mad/ib_umad) and related interfaces to send and receive Intel Omni-Path Architecture MADs on devices which support them. In addition to supporting some IBTA management classes, OPA devices use

[PATCH 02/20] IB/core: Cache device attributes for use by upper level drivers

2015-01-12 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/device.c |2 ++ include/rdma/ib_verbs.h |1 + 2 files changed, 3 insertions(+), 0 deletions(-) diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core

[PATCH 0/2] infiniband-diags/rdma-ndd: Fix install/start up script

2015-03-16 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Fix the following issues reported by Dan: 1. I found an issue with the /etc/init.d/rdma-ndd script. The “stop” option (e.g. /etc/init.d/rdma-ndd stop) doesn’t kill the rdma-ndd. The problem is that the rdma-ndd doesn’t creates pidfile. Just

[PATCH 2/2] infiniband-diags/rdma-ndd: add --pidfile option

2015-03-16 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com rdma-ndd start up script works better with a pidfile. This was partially implemented in the start up script and not at all in the daemon. Fully implement this. Reported-by: Dan Ben Yosef da...@mellanox.com Signed-off-by: Ira Weiny ira.we...@intel.com

[RFC PATCH 5/5] IB/core: Add cap_opa_mad helper using RDMA_CORE_CAP_OPA_MAD flag

2015-05-04 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com OPA MADs share a common header with IBTA MADs but with a different base version and an extended length. These MADs increase the performance of management traffic on OPA devices. Sharing a common header with IBTA MADs allows us to share most of the MAD

[PATCH 4/4] IB/mad: Add const qualifiers to query only functions

2015-05-08 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com The following functions only need read access to the data passed to them. ib_mad_kernel_rmpp_agent is_rmpp_data_mad rcv_has_same_gid ib_find_send_mad Clarify with const specifiers Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core

[PATCH] IB/core: Change rdma_protcol_iboe to roce

2015-05-14 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com It has been decided that ROCE should be used within the kernel rather than IBOE as we used before. Change iboe to roce on the new rdma_protocol_* functions. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/cma.c| 12

[PATCH 13/14] IB/mad: Add partial Intel OPA MAD support

2015-05-20 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Add OPA SMP processing functionality. Define the new OPA SMP format, create support functions for this format using the previously defined helper functions as appropriate. These functions are defined in this patch and used in the final OPA MAD support patch

[PATCH 04/14] IB/mad: Add helper function for smi_handle_dr_smp_send

2015-05-20 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This helper function will be used later for processing both IB and OPA SMP sends. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/smi.c | 81 +-- 1 file changed, 47 insertions(+), 34 deletions

[PATCH 10/14] IB/mad: Add MAD size parameters to process_mad

2015-05-20 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com In support of variable length MADs add in and out MAD size parameters to the process_mad call. The out MAD size parameter is passed by reference such that it can be updated by the agent to indicate the proper response length to be sent by the MAD stack

[PATCH 00/14] IB/mad: Add support for OPA MAD processing.

2015-05-20 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com The following patch series modifies the kernel MAD processing (ib_mad/ib_umad) and related interfaces to send and receive Intel Omni-Path Architecture MADs on devices which support them. OPA MADs share the same common header with IBTA MADs which allows us

[PATCH 11/14] IB/core: Add rdma_cap_opa_mad helper

2015-05-20 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com OPA MADs share a common header with IBTA MADs but with a different base version and an extended length. These MADs increase the performance of management traffic on OPA devices. Sharing a common header with IBTA MADs allows us to share most of the MAD

[PATCH 07/14] IB/mad: Add base version to ib_create_send_mad

2015-05-20 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com In preparation to support the new OPA MAD Base version, add a base version parameter to ib_create_send_mad and set it to IB_MGMT_BASE_VERSION for current users. Definition of the new base version and it's processing will occur in later patches. Signed-off

[PATCH 01/14] IB/mad: Clean up ib_find_send_mad

2015-05-20 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com ib_find_send_mad only needs access to the MAD header. Change the local variable to ib_mad_hdr and change the corresponding cast. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad.c | 12 ++-- 1 file changed, 6 insertions

Re: [PATCH v7 5/6] xprtrdma, svcrdma: Switch to generic logging helpers

2015-06-08 Thread ira. weiny
On Mon, Jun 08, 2015 at 11:15:04AM +0300, Sagi Grimberg wrote: On 6/7/2015 9:00 AM, ira.weiny wrote: @@ -201,9 +202,10 @@ static void qp_event_handler(struct ib_event *event, void *context) case IB_EVENT_QP_ACCESS_ERR: case IB_EVENT_DEVICE_FATAL: default: -

[PATCH 07/14] IB/mad: Support alternate Base Versions when creating MADs

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com In preparation to support the new OPA MAD Base version, add a base version parameter to ib_create_send_mad and set it to IB_MGMT_BASE_VERSION for current users. Definition of the new base version and it's processing will occur in later patches. Signed-off

[PATCH 04/14] IB/mad: Create a generic helper for DR SMP Send processing

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com IB and OPA SMPs share the same processing algorithm but have different header formats and permissive LID detection. Add a helper function which is generic to processing DR SMP Send messages which can be used by both IB and OPA SMP code. Use this function

[PATCH 06/14] IB/mad: Create a generic helper for DR forwarding checks

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com IB and OPA SMPs share the same processing algorithm but have different header formats and permissive LID detection. Add a helper function which is generic to processing the DR forwarding checks which can be used by both IB and OPA SMP code. Use this function

[PATCH 05/14] IB/mad: Create a generic helper for DR SMP Recv processing

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com IB and OPA SMPs share the same processing algorithm but have different header formats and permissive LID detection. Add a helper function which is generic to processing DR SMP Recv messages which can be used by both IB and OPA SMP code. Use this function

[PATCH 11/14] IB/core: Add OPA MAD core capability flag

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Add OPA MAD support flags to the core capability immutable flags. In addition add the rdma_cap_opa_mad helper function for core functions to use to detect OPA MAD support. OPA MADs share a common header with IBTA MADs but with some differences for increased

[PATCH 13/14] IB/mad: Add partial Intel OPA MAD support

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Add OPA SMP processing functionality. Define the new OPA SMP format, create support functions for this format using the previously defined helper functions as appropriate. These functions are defined in this patch and used in the final OPA MAD support patch

[PATCH 14/14] IB/mad: Add final OPA MAD processing

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com For devices which support OPA MADs 1) Use previously defined SMP support functions. 2) Pass correct base version to ib_create_send_mad when processing OPA MADs. 3) Process out_mad_key_index returned by agents for a response. This is necessary

[PATCH 03/14] IB/mad: Split IB SMI handling from MAD Recv handler

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Make a helper function to process Directed Route SMPs to be called by the IB MAD Recv Handler, ib_mad_recv_done_handler. This cleans up the MAD receive handler code a bit and allows for us to better share the SMP processing code between IB and OPA SMPs. IB

[PATCH 01/14] IB/mad cleanup: Clean up function params -- find_mad_agent

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com find_mad_agent only needs read only access to the MAD header. Update the ib_mad pointer to be const ib_mad_hdr. Adjust call tree. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/mad.c | 36 ++-- 1

[PATCH 12/14] IB/mad: Add partial Intel OPA MAD support

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This patch is the first of 3 which adds processing of OPA MADs 1) Add Intel Omni-Path Architecture defines 2) Increase max management version to accommodate OPA 3) update ib_create_send_mad If the device supports OPA MADs and the MAD being sent

[PATCH 09/14] IB/mad: Convert allocations from kmem_cache to kzalloc

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com This patch implements allocating alternate receive MAD buffers within the MAD stack. Support for OPA to send/recv variable sized MADs is implemented later. 1) Convert MAD allocations from kmem_cache to kzalloc kzalloc is more flexible to support

[PATCH 10/14] IB/mad: Add support for additional MAD info to/from drivers

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com In order to support alternate sized MADs (and variable sized MADs on OPA devices) add in/out MAD size parameters to the process_mad core call. In addition, add an out_mad_pkey_index to communicate the pkey index the driver wishes the MAD stack to use when

[PATCH 08/14] IB/core: Add ability for drivers to report an alternate MAD size.

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Add max MAD size to the device immutable data set and have all drivers that support MADs report the current IB MAD size (IB_MGMT_MAD_SIZE) to the core. Verify MAD size data in both the MAD core and when reading the immutable data. OPA drivers will report

[PATCH 00/14] IB/mad: Add support for OPA MAD processing.

2015-06-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com The following patch series modifies the kernel MAD processing (ib_mad/ib_umad) and related interfaces to send and receive Intel Omni-Path Architecture MADs on devices which support them. OPA MADs share the same common header with IBTA MADs which allows us

[PATCH V2] IB/mad: Remove improper use of BUG_ON

2015-06-25 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com We recently added BUG_ON's which were inappropriate for a condition which should never happen. Change these to be WARN_ON_ONCE as a debugging aid. Fixes: 4cd7c9479aff ('IB/mad: Add support for additional MAD info to/from drivers') Signed-off-by: Ira Weiny

[PATCH V2] IB/mad: Fix 0-day build

2015-06-25 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Commit 3d33f9d9fea0 IB/mad: Add final OPA MAD processing Added a 0-day build error which did not appear until building with the OPA driver sent to the list. Signed-off-by: Ira Weiny ira.we...@intel.com Reviewed-by: Mike Marciniszyn mike.marcinis...@intel.com

[PATCH V3] IB/mad: Fix 0-day build

2015-06-25 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com The define OPA_LID_PERMISSIVE is big endian and was compared to cpu value opa_drslid. 0-day build caught this while building with the OPA (hfi1) driver which was recently sent to the list. Fixes: 8e4349d13f33 (IB/mad: Add final OPA MAD processing) Signed-off

[PATCH] IB/mad: Fix 0-day build

2015-06-24 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com Commit 3d33f9d9fea0 IB/mad: Add final OPA MAD processing Added a 0-day build error which did not appear until building with the WFR driver sent to the list. Signed-off-by: Ira Weiny ira.we...@intel.com Reviewed-by: Mike Marciniszyn mike.marcinis...@intel.com

[PATCH] IB/mad: Remove improper use of BUG_ON

2015-06-24 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com commit 97f229a8515f932e5adc6cdfa18cc1440235b9fd IB/mad: Add support for additional MAD info to/from drivers added BUG_ON's which were inappropriate for a condition which should never happen. Change these to be WARN_ON_ONCE as a debugging aid. Signed-off

[RESEND PATCH v3 14/14] IB/mad: Add final OPA MAD processing

2015-06-10 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com For devices which support OPA MADs 1) Use previously defined SMP support functions. 2) Pass correct base version to ib_create_send_mad when processing OPA MADs. 3) Process out_mad_key_index returned by agents for a response. This is necessary

[PATCH 1/3] IB/core cleanup: Add const to RDMA helpers

2015-05-31 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com The ib_device passed to the new RDMA helpers is constant. Declare the ib_device as const in the following functions. rdma_protocol_ib rdma_protocol_roce rdma_protocol_iwarp rdma_ib_or_roce rdma_cap_ib_mad rdma_cap_ib_smi rdma_cap_ib_cm rdma_cap_iw_cm

[PATCH 3/3] IB/core cleanup: Add const to args - agent_send_response

2015-05-31 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com In order to support constant callers of agent_send_response we add const specifiers to the its pointer arguments. Adjust the call tree accordingly. Signed-off-by: Ira Weiny ira.we...@intel.com --- Changes from V1: Update commit message drivers

[PATCH 0/3] Add const to various core functions

2015-05-31 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com In order to support some of Jason's comments to add const to the functions I am adding for OPA the following patches clean up the call trees of the currently defined functions. They stand on their own as clean up patches. Therefore, I am submitting them

[PATCH 3/3] IB/core: Add const to agent_send_response

2015-05-21 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com In order to support constant callers of agent_send_response we add const specifiers to the its pointer arguments. Adjust the call tree accordingly. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/core/addr.c | 4 ++-- drivers/infiniband

[PATCH 2/3] IB/core: Add const to device process_mad

2015-05-21 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com The process_mad device function declares some parameters as in. Make those parameters const and adjust the call tree under process_mad in the various drivers accordingly. Signed-off-by: Ira Weiny ira.we...@intel.com --- drivers/infiniband/hw/amso1100

[PATCH 1/3] IB/core: Add const to RDMA helpers

2015-05-21 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com The ib_device passed to the new RDMA helpers is constant. Signed-off-by: Ira Weiny ira.we...@intel.com --- include/rdma/ib_verbs.h | 24 1 file changed, 12 insertions(+), 12 deletions(-) diff --git a/include/rdma/ib_verbs.h b/include

[PATCH 0/3] Add const to various core functions

2015-05-21 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com In order to support some of Jason's comments to add const to the functions I am adding for OPA the following patches clean up the call trees of the currently defined functions. They stand on their own as clean up patches. Therefore, I am submitting them

[PATCH] IB/sa: Restrict SA Netlink to admin users

2015-08-06 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com The recently added SA Netlink service requires admin privileges to receive kernel requests. This is only partially sufficient to protect the kernel from malicious users. This patch fixes two issues. 1) Path responses from user space could be spoofed

[PATCH V2] IB/sa: Restrict SA Netlink to admin users

2015-08-10 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com The recently added SA Netlink service requires admin privileges to receive kernel requests. This is only partially sufficient to protect the kernel from malicious users. This patch fixes two issues. 1) Path responses from user space could be spoofed

[PATCH] perftest: Fix rpm generation from dist tarball

2015-07-21 Thread ira . weiny
From: Ira Weiny ira.we...@intel.com The spec file within the git tree was not included in the dist tarball which would prevent rpmbuild from building it. Further errors are fixed by making the spec file autogenerated and adding missing files to the tarball. The following build sequence will now

[PATCH v3 14/23] staging/rdma/hfi1: Implement Expected Receive TID caching

2015-10-26 Thread ira . weiny
gned-off-by: Mitko Haralanov <mitko.harala...@intel.com> Signed-off-by: Ira Weiny <ira.we...@intel.com> --- Changes from V2: Fix random Kconfig 0-day build error Fix leak of random memory to user space caught by Dan Carpenter Separate out pointer bug fix into a pr

[PATCH v3 21/23] staging/rdma/hfi1: Add unit # to verbs txreq cache name

2015-10-26 Thread ira . weiny
niszyn <mike.marcinis...@intel.com> Signed-off-by: Jubin John <jubin.j...@intel.com> Signed-off-by: Ira Weiny <ira.we...@intel.com> --- drivers/staging/rdma/hfi1/verbs.c | 7 ++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/staging/rdma/hfi1/verbs.c b/drivers/s

[PATCH v3 09/23] staging/rdma/hfi1: Add a schedule in send thread

2015-10-26 Thread ira . weiny
tel.com> Signed-off-by: Ira Weiny <ira.we...@intel.com> --- drivers/staging/rdma/hfi1/chip.c | 10 ++ drivers/staging/rdma/hfi1/chip.h | 1 + drivers/staging/rdma/hfi1/ruc.c | 12 drivers/staging/rdma/hfi1/verbs.h | 1 + 4 files changed, 24 insertions(+) diff --gi

[PATCH v3 10/23] staging/rdma/hfi1: Fix port bounce issues with 0.22 DC firmware

2015-10-26 Thread ira . weiny
ane_tx register to all 4 lanes. Reviewed-by: Dean Luick <dean.lu...@intel.com> Signed-off-by: Easwar Hariharan <easwar.hariha...@intel.com> Signed-off-by: Ira Weiny <ira.we...@intel.com> --- drivers/staging/rdma/hfi1/chip.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/d

[PATCH v3 22/23] staging/rdma/hfi1: add additional rc traces

2015-10-26 Thread ira . weiny
From: Mike Marciniszyn <mike.marcinis...@intel.com> Add additional rc traces to aid in debugging rc retry logic. Reviewed-by: Dennis Dalessandro <dennis.dalessan...@intel.com> Signed-off-by: Mike Marciniszyn <mike.marcinis...@intel.com> Signed-off-by: Ira Weiny <ira.we...@in

[PATCH v3 17/23] staging/rdma/hfi1: Add irqsaves in the packet processing path

2015-10-26 Thread ira . weiny
locks in the packet processing path. Anything that did not have an obvious, "close" irqsave in its caller is a candidate. Reviewed-by: Mike Marciniszyn <mike.marcinis...@intel.com> Signed-off-by: Dean Luick <dean.lu...@intel.com> Signed-off-by: Ira Weiny <ira.we...@intel

<    2   3   4   5   6   7   8   9   >