summaryrefslogtreecommitdiffstats
path: root/drivers
AgeCommit message (Collapse)AuthorFilesLines
2019-11-06Revert "RDMA/srpt: Postpone HCA removal until after configfs directory removal"Bart Van Assche1-5/+1
Although the mentioned patch fixes a use-after-free bug, it introduces a hang during shutdown. Since the latter is worse, revert this patch. Link: https://lore.kernel.org/r/20191101204756.182162-1-bvanassche@acm.org Reported-by: Honggang Li <honli@redhat.com> Fixes: 9b64f7d0bb0a ("RDMA/srpt: Postpone HCA removal until after configfs directory removal") Signed-off-by: Bart Van Assche <bvanassche@acm.org> Acked-by: Honggang Li <honli@redhat.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-11-06RDMA/qedr: Remove unsupported modify_port callbackKamal Heib3-9/+0
There is no need to return always zero for function which is not supported. Fixes: ac1b36e55a51 ("qedr: Add support for user context verbs") Link: https://lore.kernel.org/r/20191028155931.1114-5-kamalheib1@gmail.com Signed-off-by: Kamal Heib <kamalheib1@gmail.com> Reviewed-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-11-06RDMA/ocrdma: Remove unsupported modify_port callbackKamal Heib3-9/+0
There is no need to return always zero for function which is not supported. Fixes: fe2caefcdf58 ("RDMA/ocrdma: Add driver for Emulex OneConnect IBoE RDMA adapter") Link: https://lore.kernel.org/r/20191028155931.1114-4-kamalheib1@gmail.com Signed-off-by: Kamal Heib <kamalheib1@gmail.com> Reviewed-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-11-06RDMA/hns: Remove unsupported modify_port callbackKamal Heib1-7/+0
There is no need to return always zero for function which is not supported. Fixes: 9a4435375cd1 ("IB/hns: Add driver files for hns RoCE driver") Link: https://lore.kernel.org/r/20191028155931.1114-3-kamalheib1@gmail.com Signed-off-by: Kamal Heib <kamalheib1@gmail.com> Reviewed-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-11-06RDMA/core: Fix return code when modify_port isn't supportedKamal Heib1-1/+5
Improve return code from ib_modify_port() by doing the following: - Use "-EOPNOTSUPP" instead "-ENOSYS" which is the proper return code - Allow only fake IB_PORT_CM_SUP manipulation for RoCE providers that didn't implement the modify_port callback, otherwise return "-EOPNOTSUPP" Fixes: 61e0962d5221 ("IB: Avoid ib_modify_port() failure for RoCE devices") Link: https://lore.kernel.org/r/20191028155931.1114-2-kamalheib1@gmail.com Signed-off-by: Kamal Heib <kamalheib1@gmail.com> Reviewed-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-11-06RDMA/qedr: Add iWARP doorbell recovery supportMichal Kalderon2-6/+43
This patch adds the iWARP specific doorbells to the doorbell recovery mechanism. Link: https://lore.kernel.org/r/20191030094417.16866-9-michal.kalderon@marvell.com Signed-off-by: Ariel Elior <ariel.elior@marvell.com> Signed-off-by: Michal Kalderon <michal.kalderon@marvell.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-11-06RDMA/qedr: Add doorbell overflow recovery supportMichal Kalderon2-50/+275
Use the doorbell recovery mechanism to register rdma related doorbells that will be restored in case there is a doorbell overflow attention. Link: https://lore.kernel.org/r/20191030094417.16866-8-michal.kalderon@marvell.com Signed-off-by: Ariel Elior <ariel.elior@marvell.com> Signed-off-by: Michal Kalderon <michal.kalderon@marvell.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-11-06RDMA/qedr: Use the common mmap APIMichal Kalderon4-121/+98
Remove all functions related to mmap from qedr and use the common API. Link: https://lore.kernel.org/r/20191030094417.16866-7-michal.kalderon@marvell.com Signed-off-by: Ariel Elior <ariel.elior@marvell.com> Signed-off-by: Michal Kalderon <michal.kalderon@marvell.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-11-06RDMA/siw: Use the common mmap_xa helpersMichal Kalderon4-102/+114
Remove the functions related to managing the mmap_xa database. This code is now common in ib_core. Link: https://lore.kernel.org/r/20191030094417.16866-6-michal.kalderon@marvell.com Signed-off-by: Ariel Elior <ariel.elior@marvell.com> Signed-off-by: Michal Kalderon <michal.kalderon@marvell.com> Signed-off-by: Bernard Metzler <bmt@zurich.ibm.com> Reviewed-by: Bernard Metzler <bmt@zurich.ibm.com> Tested-by: Bernard Metzler <bmt@zurich.ibm.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-11-06RDMA/efa: Use the common mmap_xa helpersMichal Kalderon3-194/+153
Remove the functions related to managing the mmap_xa database. This code was replaced with common code in ib_core. Link: https://lore.kernel.org/r/20191030094417.16866-5-michal.kalderon@marvell.com Signed-off-by: Ariel Elior <ariel.elior@marvell.com> Signed-off-by: Michal Kalderon <michal.kalderon@marvell.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-11-06RDMA: Connect between the mmap entry and the umap_priv structureMichal Kalderon7-23/+68
The rdma_user_mmap_io interface created a common interface for drivers to correctly map hw resources and zap them once the ucontext is destroyed enabling the drivers to safely free the hw resources. However, this meant the drivers need to delay freeing the resource to the ucontext destroy phase to ensure they were no longer mapped. The new mechanism for a common way of handling user/driver address mapping enabled notifying the driver if all umap_priv mappings were removed, and enabled freeing the hw resources when they are done with and not delay it until ucontext destroy. Since not all drivers use the mechanism, NULL can be sent to the rdma_user_mmap_io interface to continue working as before. Drivers that use the mmap_xa interface can pass the entry being mapped to the rdma_user_mmap_io function to be linked together. Link: https://lore.kernel.org/r/20191030094417.16866-4-michal.kalderon@marvell.com Signed-off-by: Ariel Elior <ariel.elior@marvell.com> Signed-off-by: Michal Kalderon <michal.kalderon@marvell.com> Reviewed-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-11-06RDMA/core: Create mmap database and cookie helper functionsMichal Kalderon4-0/+240
Create some common API's for adding entries to a xa_mmap. Searching for an entry and freeing one. The general approach is copied from the EFA driver and improved to be more general and do more to help the drivers. Integration with the core allows a reference counted scheme with a free function so that the driver can know when its mmaps are all gone. This significant new functionality will be helpful for drivers to have the correct lifetime model for mmap objects. Link: https://lore.kernel.org/r/20191030094417.16866-3-michal.kalderon@marvell.com Signed-off-by: Ariel Elior <ariel.elior@marvell.com> Signed-off-by: Michal Kalderon <michal.kalderon@marvell.com> Reviewed-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-11-05net/mlx5: fix spelling mistake "metdata" -> "metadata"Colin Ian King1-1/+1
There is a spelling mistake in a esw_warn warning message. Fix it. Signed-off-by: Colin Ian King <colin.king@canonical.com> Reviewed-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-05net/mlx5: fix kvfree of uninitialized pointer specColin Ian King1-1/+1
Currently when a call to esw_vport_create_legacy_ingress_acl_group fails the error exit path to label 'out' will cause a kvfree on the uninitialized pointer spec. Fix this by ensuring pointer spec is initialized to NULL to avoid this issue. Addresses-Coverity: ("Uninitialized pointer read") Fixes: 10652f39943e ("net/mlx5: Refactor ingress acl configuration") Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-05RDMA/core: Move core content from ib_uverbs to ib_coreMichal Kalderon4-72/+86
Move functionality that is called by the driver, which is related to umap, to a new file that will be linked in ib_core. This is a first step in later enabling ib_uverbs to be optional. vm_ops is now initialized in ib_uverbs_mmap instead of priv_init to avoid having to move all the rdma_umap functions as well. Link: https://lore.kernel.org/r/20191030094417.16866-2-michal.kalderon@marvell.com Suggested-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Ariel Elior <ariel.elior@marvell.com> Signed-off-by: Michal Kalderon <michal.kalderon@marvell.com> Reviewed-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-11-01IB/mlx5: Introduce and use mlx5_core_is_vf()Parav Pandit1-1/+1
Instead of deciding a given device is virtual function or not based on a device is PF or not, use already defined MLX5_COREDEV_VF by introducing an helper API mlx5_core_is_vf(). This enables to clearly identify PF, VF and non virtual functions. Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Vu Pham <vuhuong@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: E-switch, Enable metadata on own vportParav Pandit3-23/+16
Currently on ECPF, metadata is enabled on the ECPF vport = 0xfffe (manager vport). Metadata when supported, must be enabled on own vport which is used to pass metadata to vport of NIC Rx Flow Table. Due to this error, traffic tagged by ingress ACL is not processed correctly at NIC rx flow table level which is supposed to work on metadata tag. Hence, instead of working on eswitch manager vport, always working on eswitch own vport regardless of PF or ECPF. Given that mlx5_eswitch_query/modify_esw_vport_context() is used to access other vport in legacy mode and own vport settings in switchdev mode, extend low level API to explicitly specify other_vport. Fixes: c1286050cf47 ("net/mlx5: E-Switch, Pass metadata from FDB to eswitch manager") Signed-off-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: Refactor ingress acl configurationParav Pandit3-114/+200
Drop, untagged, spoof check and untagged spoof check flow groups are limited to legacy mode only. Therefore, following refactoring is done to (a) improve code readability (b) have better code split between legacy and offloads mode 1. Move legacy flow groups under legacy structure 2. Add validity check for group deletion 3. Restrict scope of esw_vport_disable_ingress_acl to legacy mode 4. Rename esw_vport_enable_ingress_acl() to esw_vport_create_ingress_acl_table() and limit its scope to table creation 5. Introduce legacy flow groups creation helper esw_legacy_create_ingress_acl_groups() and keep its scope to legacy mode 6. Reduce offloads ingress groups from 4 to just 1 metadata group per vport 7. Removed redundant IS_ERR_OR_NULL as entries are marked NULL on free. 8. Shortern error message to remove redundant 'E-switch' Signed-off-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: Restrict metadata disablement to offloads modeParav Pandit3-7/+6
Now that there is clear separation for acl setup/cleanup between legacy and offloads mode, limit metdata disablement to offloads mode. Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Vu Pham <vuhuong@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: E-switch, Offloads shift ACL programming during enable/disable vportVu Pham3-31/+24
Currently legacy mode enables ACL while enabling vport, while offloads mode enable ACL when moving to offloads mode. Bring consistency to both modes by enabling/disabling ACL when enabling/disabling a vport. It also eliminates creating ingress ACL table on unused ECPF vport in offloads mode. Signed-off-by: Vu Pham <vuhuong@mellanox.com> Signed-off-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: E-switch, Offloads introduce and use per vport acl tables APIsParav Pandit1-17/+32
Introduce and use per vport ACL tables creation and destroy APIs, so that subsequently patch can use them during enabling/disabling a vport. Signed-off-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: Move ACL drop counters life cycle close to ACL lifecycleParav Pandit1-39/+35
It is better to create/destroy ACL related drop counters where the actual drop rule ACLs are created/destroyed, so that ACL configuration is self contained for ingress and egress. Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Vu Pham <vuhuong@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: E-switch, Legacy introduce and use per vport acl tables APIsParav Pandit1-13/+60
Introduce and use per vport ACL tables creation and destroy APIs, so that subsequently patch can use them during enabling/disabling a vport in unified way for legacy vs offloads mode. Signed-off-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: E-switch, Prepare code to handle vport enable errorParav Pandit3-19/+50
In subsequent patch, esw_enable_vport() could fail and return error. Prepare code to handle such error. Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Vu Pham <vuhuong@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: Tide up state_lock and vport enabled flag usageParav Pandit1-9/+5
When eswitch is disabled, vport event handler is unregistered. This unregistration already synchronizes with running EQ event handler in below code flow. mlx5_eswitch_disable() mlx5_eswitch_event_handlers_unregister() mlx5_eq_notifier_unregister() atomic_notifier_chain_unregister() synchronize_rcu() notifier_callchain eswitch_vport_event() queue_work() Additionally vport->enabled flag is set under state_lock during esw_enable_vport() but is not read under state_lock in (a) esw_disable_vport() and (b) under atomic context eswitch_vport_event(). It is also necessary to synchronize with already scheduled vport event. This is already achieved using below sequence. mlx5_eswitch_event_handlers_unregister() [..] flush_workqueue() Hence, (a) Remove vport->enabled check in eswitch_vport_event() which doesn't make any sense. (b) Remove redundant flush_workqueue() on every vport disable. (c) Keep esw_disable_vport() symmetric with esw_enable_vport() for state_lock. Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Vu Pham <vuhuong@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: Move legacy drop counter and rule under legacy structureParav Pandit2-44/+50
To improve code readability, move legacy drop counters and droup rule under legacy structure. While at it, (a) prefix drop flow counters helper with legacy_. (b) nullify the rule pointers only if they were valid. Signed-off-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: Move metdata fields under offloads structureParav Pandit2-18/+21
Metadata fields are offload mode specific. To improve code readability, move metadata under offloads structure. Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Vu Pham <vuhuong@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: Correct comment for legacy fieldsParav Pandit1-1/+1
fdb_table is used for both legacy and offloads mode. It was incorrect to comment that fdb_table is legacy specific. Hence, fix the comment to reflect that fdb_table is used in legacy and offloads mode. Fixes: 131ce7014043 ("net/mlx5: E-Switch, Remove redundant mc_promisc NULL check") Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Vu Pham <vuhuong@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: Introduce and use mlx5_esw_is_manager_vport()Parav Pandit2-6/+13
Currently esw_enable_vport() does vport check for zero to enable drop counters regardless of execution on ECPF/PF. While esw_disable_vport() considers such scenario. To keep consistency across code for checking for manager_vport, introduce and use mlx5_esw_is_manager_vport() to check if a specified vport is eswitch manager vport or not. Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Vu Pham <vuhuong@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: E-switch, Introduce and use vlan rule config helperParav Pandit3-68/+58
Between legacy mode and switchdev mode, only two fields are changed, vlan_tag and flow action. Hence to avoid duplicte code between two modes, introduce and and use helper function to configure allowed VLAN rule. While at it, get rid of duplicate debug message. Signed-off-by: Parav Pandit <parav@mellanox.com> Reviewed-by: Vu Pham <vuhuong@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: E-Switch, Rename ingress acl config in offloads modeVu Pham1-3/+3
Changing the function name esw_ingress_acl_common_config() to esw_ingress_acl_config() to be consistent with egress config function naming in offloads mode. Signed-off-by: Vu Pham <vuhuong@mellanox.com> Reviewed-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: E-Switch, Rename egress config to generic nameVu Pham1-24/+26
Refactor vport egress config in offloads mode Refactoring vport egress configuration in offloads mode that includes egress prio tag configuration. This makes code symmetric to ingress configuration. Signed-off-by: Vu Pham <vuhuong@mellanox.com> Reviewed-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-11-01net/mlx5: Fixed a typo in a comment in esw_del_uc_addr()Qing Huang1-1/+1
Changed "managerss" to "managers". Fixes: a1b3839ac4a4 ("net/mlx5: E-Switch, Properly refer to the esw manager vport") Signed-off-by: Qing Huang <qing.huang@oracle.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
2019-10-31IB/mlx5: Test write combining supportMichael Guralnik4-3/+223
Linux can run in all sorts of physical machines and VMs where write combining may or may not be supported. Currently there is no way to reliably tell if the system supports WC, or not. The driver uses WC to optimize posting work to the HCA, and getting this wrong in either direction can cause a significant performance loss. Add a test in mlx5_ib initialization process to test whether write-combining is supported on the machine. The test will run as part of the enable_driver callback to ensure that the test runs after the device is setup and can create and modify the QP needed, but runs before the device is exposed to the users. The test opens UD QP and posts NOP WQEs, the WQE written to the BlueFlame is different from the WQE in memory, requesting CQE only on the BlueFlame WQE. By checking whether we received a completion on one of these WQEs we can know if BlueFlame succeeded and this write-combining must be supported. Change reporting of BlueFlame support to be dependent on write-combining support instead of the FW's guess as to what the machine can do. Link: https://lore.kernel.org/r/20191027062234.10993-1-leon@kernel.org Signed-off-by: Michael Guralnik <michaelgur@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-31RDMA/mlx5: Return proper error valueLeon Romanovsky1-1/+1
Returned value from mlx5_mr_cache_alloc() is checked to be error or real pointer. Return proper error code instead of NULL which is not checked later. Fixes: 81713d3788d2 ("IB/mlx5: Add implicit MR support") Link: https://lore.kernel.org/r/20191029055721.7192-1-leon@kernel.org Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Reviewed-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-29RDMA/hns: Fix build error againArnd Bergmann2-5/+20
This is not the first attempt to fix building random configurations, unfortunately the attempt in commit a07fc0bb483e ("RDMA/hns: Fix build error") caused a new problem when CONFIG_INFINIBAND_HNS_HIP06=m and CONFIG_INFINIBAND_HNS_HIP08=y: drivers/infiniband/hw/hns/hns_roce_main.o:(.rodata+0xe60): undefined reference to `__this_module' Revert commits a07fc0bb483e ("RDMA/hns: Fix build error") and a3e2d4c7e766 ("RDMA/hns: remove obsolete Kconfig comment") to get back to the previous state, then fix the issues described there differently, by adding more specific dependencies: INFINIBAND_HNS can now only be built-in if at least one of HNS or HNS3 are built-in, and the individual back-ends are only available if that code is reachable from the main driver. Fixes: a07fc0bb483e ("RDMA/hns: Fix build error") Fixes: a3e2d4c7e766 ("RDMA/hns: remove obsolete Kconfig comment") Fixes: dd74282df573 ("RDMA/hns: Initialize the PCI device for hip08 RoCE") Fixes: 08805fdbeb2d ("RDMA/hns: Split hw v1 driver from hns roce driver") Link: https://lore.kernel.org/r/20191007211826.3361202-1-arnd@arndb.de Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28Merge branch 'odp_rework' into rdma.git for-nextJason Gunthorpe9-641/+624
Jason Gunthorpe says: ==================== In order to hoist the interval tree code out of the drivers and into the mmu_notifiers it is necessary for the drivers to not use the interval tree for other things. This series replaces the interval tree with an xarray and along the way re-aligns all the locking to use a sensible SRCU model where the 'update' step is done by modifying an xarray. The result is overall much simpler and with less locking in the critical path. Many functions were reworked for clarity and small details like using 'imr' to refer to the implicit MR make the entire code flow here more readable. This also squashes at least two race bugs on its own, and quite possibily more that haven't been identified. ==================== Merge conflicts with the odp statistics patch resolved. * branch 'odp_rework': RDMA/odp: Remove broken debugging call to invalidate_range RDMA/mlx5: Do not race with mlx5_ib_invalidate_range during create and destroy RDMA/mlx5: Do not store implicit children in the odp_mkeys xarray RDMA/mlx5: Rework implicit ODP destroy RDMA/mlx5: Avoid double lookups on the pagefault path RDMA/mlx5: Reduce locking in implicit_mr_get_data() RDMA/mlx5: Use an xarray for the children of an implicit ODP RDMA/mlx5: Split implicit handling from pagefault_mr RDMA/mlx5: Set the HW IOVA of the child MRs to their place in the tree RDMA/mlx5: Lift implicit_mr_alloc() into the two routines that call it RDMA/mlx5: Rework implicit_mr_get_data RDMA/mlx5: Delete struct mlx5_priv->mkey_table RDMA/mlx5: Use a dedicated mkey xarray for ODP RDMA/mlx5: Split sig_err MR data into its own xarray RDMA/mlx5: Use SRCU properly in ODP prefetch Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28RDMA/odp: Remove broken debugging call to invalidate_rangeJason Gunthorpe1-19/+19
invalidate_range() also obtains the umem_mutex which is being held at this point, so if this path were was ever called it would deadlock. Thus conclude the debugging never triggers and rework it into a simple WARN_ON and leave things as they are. While here add a note to explain how we could possibly get inconsistent page pointers. Link: https://lore.kernel.org/r/20191009160934.3143-16-jgg@ziepe.ca Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28RDMA/mlx5: Do not race with mlx5_ib_invalidate_range during create and destroyJason Gunthorpe3-59/+88
For creation, as soon as the umem_odp is created the notifier can be called, however the underlying MR may not have been setup yet. This would cause problems if mlx5_ib_invalidate_range() runs. There is some confusing/ulocked/racy code that might by trying to solve this, but without locks it isn't going to work right. Instead trivially solve the problem by short-circuiting the invalidation if there are not yet any DMA mapped pages. By definition there is nothing to invalidate in this case. The create code will have the umem fully setup before anything is DMA mapped, and npages is fully locked by the umem_mutex. For destroy, invalidate the entire MR at the HW to stop DMA then DMA unmap the pages before destroying the MR. This drives npages to zero and prevents similar racing with invalidate while the MR is undergoing destruction. Arguably it would be better if the umem was created after the MR and destroyed before, but that would require a big rework of the MR code. Fixes: 6aec21f6a832 ("IB/mlx5: Page faults handling infrastructure") Link: https://lore.kernel.org/r/20191009160934.3143-15-jgg@ziepe.ca Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28RDMA/mlx5: Do not store implicit children in the odp_mkeys xarrayJason Gunthorpe1-30/+6
These mkeys are entirely internal and are never used by the HW for page fault. They should also never be used by userspace for prefetch. Simplify & optimize things by not including them in the xarray. Since the prefetch path can now never see a child mkey there is no need for the second synchronize_srcu() during imr destroy. Link: https://lore.kernel.org/r/20191009160934.3143-14-jgg@ziepe.ca Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28RDMA/mlx5: Rework implicit ODP destroyJason Gunthorpe4-64/+120
Use SRCU in a sensible way by removing all MRs in the implicit tree from the two xarrays (the update operation), then a synchronize, followed by a normal single threaded teardown. This is only a little unusual from the normal pattern as there can still be some work pending in the unbound wq that may also require a workqueue flush. This is tracked with a single atomic, consolidating the redundant existing atomics and wait queue. For understand-ability the entire ODP implicit create/destroy flow now largely exists in a single pair of functions within odp.c, with a few support functions for tearing down an unused child. Link: https://lore.kernel.org/r/20191009160934.3143-13-jgg@ziepe.ca Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28RDMA/mlx5: Avoid double lookups on the pagefault pathJason Gunthorpe1-106/+80
Now that the locking is simplified combine pagefault_implicit_mr() with implicit_mr_get_data() so that we sweep over the idx range only once, and do the single xlt update at the end, after the child umems are setup. This avoids double iteration/xa_loads plus the sketchy failure path if the xa_load() fails. Link: https://lore.kernel.org/r/20191009160934.3143-12-jgg@ziepe.ca Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28RDMA/mlx5: Reduce locking in implicit_mr_get_data()Jason Gunthorpe1-12/+26
Now that the child MRs are stored in an xarray we can rely on the SRCU lock to protect the xa_load and use xa_cmpxchg on the slow allocation path to resolve races with concurrent page fault. This reduces the scope of the critical section of umem_mutex for implicit MRs to only cover mlx5_ib_update_xlt, and avoids taking a lock at all if the child MR is already in the xarray. This makes it consistent with the normal ODP MR critical section for umem_lock, and the locking approach used for destroying an unusued implicit child MR. The MLX5_IB_UPD_XLT_ATOMIC is no longer needed in implicit_get_child_mr() since it is no longer called with any locks. Link: https://lore.kernel.org/r/20191009160934.3143-11-jgg@ziepe.ca Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28RDMA/mlx5: Use an xarray for the children of an implicit ODPJason Gunthorpe2-133/+67
Currently the child leaves are stored in the shared interval tree and every lookup for a child must be done under the interval tree rwsem. This is further complicated by dropping the rwsem during iteration (ie the odp_lookup(), odp_next() pattern), which requires a very tricky an difficult to understand locking scheme with SRCU. Instead reserve the interval tree for the exclusive use of the mmu notifier related code in umem_odp.c and give each implicit MR a xarray containing all the child MRs. Since the size of each child is 1GB of VA, a 1 level xarray will index 64G of VA, and a 2 level will index 2TB, making xarray a much better data structure choice than an interval tree. The locking properties of xarray will be used in the next patches to rework the implicit ODP locking scheme into something simpler. At this point, the xarray is locked by the implicit MR's umem_mutex, and read can also be locked by the odp_srcu. Link: https://lore.kernel.org/r/20191009160934.3143-10-jgg@ziepe.ca Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28RDMA/mlx5: Split implicit handling from pagefault_mrJason Gunthorpe1-49/+76
The single routine has a very confusing scheme to advance to the next child MR when working on an implicit parent. This scheme can only be used when working with an implicit parent and must not be triggered when working on a normal MR. Re-arrange things by directly putting all the single-MR stuff into one function and calling it in a loop for the implicit case. Simplify some of the error handling in the new pagefault_real_mr() to remove unneeded gotos. Link: https://lore.kernel.org/r/20191009160934.3143-9-jgg@ziepe.ca Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28RDMA/mlx5: Set the HW IOVA of the child MRs to their place in the treeJason Gunthorpe1-2/+3
Instead of rewriting all the IOVA's to 0 as things progress down the tree make the IOVA of the children equal to placement in the tree. This makes things easier to understand by keeping mmkey.iova == HW configuration. Link: https://lore.kernel.org/r/20191009160934.3143-8-jgg@ziepe.ca Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28RDMA/mlx5: Lift implicit_mr_alloc() into the two routines that call itJason Gunthorpe1-77/+74
This makes the routines easier to understand, particularly with respect the locking requirements of the entire sequence. The implicit_mr_alloc() had a lot of ifs specializing it to each of the callers, and only a very small amount of code was actually shared. Following patches will cause the flow in the two functions to diverge further. Link: https://lore.kernel.org/r/20191009160934.3143-7-jgg@ziepe.ca Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28RDMA/mlx5: Rework implicit_mr_get_dataJason Gunthorpe1-54/+69
This function is intended to loop across each MTT chunk in the implicit parent that intersects the range [io_virt, io_virt+bnct). But it is has a confusing construction, so: - Consistently use imr and odp_imr to refer to the implicit parent to avoid confusion with the normal mr and odp of the child - Directly compute the inclusive start/end indexes by shifting. This is clearer to understand the intent and avoids any errors from unaligned values of addr - Iterate directly over the range of MTT indexes, do not make a loop out of goto - Follow 'success oriented flow', with goto error unwind - Directly calculate the range of idx's that need update_xlt - Ensure that any leaf MR added to the interval tree always results in an update to the XLT Link: https://lore.kernel.org/r/20191009160934.3143-6-jgg@ziepe.ca Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28RDMA/mlx5: Delete struct mlx5_priv->mkey_tableJason Gunthorpe3-40/+1
No users are left, delete it. Link: https://lore.kernel.org/r/20191009160934.3143-5-jgg@ziepe.ca Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-10-28RDMA/mlx5: Use a dedicated mkey xarray for ODPJason Gunthorpe5-73/+83
There is a per device xarray storing mkeys that is used to store every mkey in the system. However, this xarray is now only read by ODP for certain ODP designated MRs (ODP, implicit ODP, MW, DEVX_INDIRECT). Create an xarray only for use by ODP, that only contains ODP related MKeys. This xarray is protected by SRCU and all erases are protected by a synchronize. This improves performance: - All MRs in the odp_mkeys xarray are ODP MRs, so some tests for is_odp() can be deleted. The xarray will also consume fewer nodes. - normal MR's are never mixed with ODP MRs in a SRCU data structure so performance sucking synchronize_srcu() on every MR destruction is not needed. - No smp_load_acquire(live) and xa_load() double barrier on read Due to the SRCU locking scheme care must be taken with the placement of the xa_store(). Once it completes the MR is immediately visible to other threads and only through a xa_erase() & synchronize_srcu() cycle could it be destroyed. Link: https://lore.kernel.org/r/20191009160934.3143-4-jgg@ziepe.ca Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-30 07:44:52 +0000