summaryrefslogtreecommitdiff
path: root/drivers/infiniband/hw/mlx5
AgeCommit message (Collapse)AuthorFilesLines
2021-11-02RDMA/mlx5: Set user priority for DCTPatrisious Haddad1-0/+2
commit 1ab52ac1e9bc9391f592c9fa8340a6e3e9c36286 upstream. Currently, the driver doesn't set the PCP-based priority for DCT, hence DCT response packets are transmitted without user priority. Fix it by setting user provided priority in the eth_prio field in the DCT context, which in turn sets the value in the transmitted packet. Fixes: 776a3906b692 ("IB/mlx5: Add support for DC target QP") Link: https://lore.kernel.org/r/5fd2d94a13f5742d8803c218927322257d53205c.1633512672.git.leonro@nvidia.com Signed-off-by: Patrisious Haddad <phaddad@nvidia.com> Reviewed-by: Maor Gottlieb <maorg@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2021-09-18RDMA/mlx5: Delete not-available udata checkLeon Romanovsky1-3/+0
[ Upstream commit 5f6bb7e32283b8e3339b7adc00638234ac199cc4 ] XRC_TGT QPs are created through kernel verbs and don't have udata at all. Fixes: 6eefa839c4dd ("RDMA/mlx5: Protect from kernel crash if XRC_TGT doesn't have udata") Fixes: e383085c2425 ("RDMA/mlx5: Set ECE options during QP create") Link: https://lore.kernel.org/r/b68228597e730675020aa5162745390a2d39d3a2.1628014762.git.leonro@nvidia.com Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-08-18net/mlx5: Synchronize correct IRQ when destroying CQShay Drory2-5/+2
[ Upstream commit 563476ae0c5e48a028cbfa38fa9d2fc0418eb88f ] The CQ destroy is performed based on the IRQ number that is stored in cq->irqn. That number wasn't set explicitly during CQ creation and as expected some of the API users of mlx5_core_create_cq() forgot to update it. This caused to wrong synchronization call of the wrong IRQ with a number 0 instead of the real one. As a fix, set the IRQ number directly in the mlx5_core_create_cq() and update all users accordingly. Fixes: 1a86b377aa21 ("vdpa/mlx5: Add VDPA driver for supported mlx5 devices") Fixes: ef1659ade359 ("IB/mlx5: Add DEVX support for CQ events") Signed-off-by: Shay Drory <shayd@nvidia.com> Reviewed-by: Tariq Toukan <tariqt@nvidia.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-08-12RDMA/mlx5: Delay emptying a cache entry when a new MR is added to it recentlyAharon Landau1-2/+2
[ Upstream commit d6793ca97b76642b77629dd0783ec64782a50bdb ] Fixing a typo that causes a cache entry to shrink immediately after adding to it new MRs if the entry size exceeds the high limit. In doing so, the cache misses its purpose to prevent the creation of new mkeys on the runtime by using the cached ones. Fixes: b9358bdbc713 ("RDMA/mlx5: Fix locking in MR cache work queue") Link: https://lore.kernel.org/r/fcb546986be346684a016f5ca23a0567399145fa.1627370131.git.leonro@nvidia.com Signed-off-by: Aharon Landau <aharonl@nvidia.com> Reviewed-by: Maor Gottlieb <maorg@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-07-14RDMA/mlx5: Don't access NULL-cleared mpi pointerLeon Romanovsky1-1/+1
[ Upstream commit 4a754d7637026b42b0c9ba5787ad5ee3bc2ff77f ] The "dev->port[i].mp.mpi" is set to NULL during mlx5_ib_unbind_slave_port() execution, however that field is needed to add device to unaffiliated list. Such flow causes to the following kernel panic while unloading mlx5_ib module in multi-port mode, hence the device should be added to the list prior to unbind call. RPC: Unregistered rdma transport module. RPC: Unregistered rdma backchannel transport module. BUG: kernel NULL pointer dereference, address: 0000000000000000 #PF: supervisor write access in kernel mode #PF: error_code(0x0002) - not-present page PGD 0 P4D 0 Oops: 0002 [#1] SMP NOPTI CPU: 4 PID: 1904 Comm: modprobe Not tainted 5.13.0-rc7_for_upstream_min_debug_2021_06_24_12_08 #1 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014 RIP: 0010:mlx5_ib_cleanup_multiport_master+0x18b/0x2d0 [mlx5_ib] Code: 00 04 0f 85 c4 00 00 00 48 89 df e8 ef fa ff ff 48 8b 83 40 0d 00 00 48 8b 15 b9 e8 05 00 4a 8b 44 28 20 48 89 05 ad e8 05 00 <48> c7 00 d0 57 c5 a0 48 89 50 08 48 89 02 39 ab 88 0a 00 00 0f 86 RSP: 0018:ffff888116ee3df8 EFLAGS: 00010296 RAX: 0000000000000000 RBX: ffff8881154f6000 RCX: 0000000000000080 RDX: ffffffffa0c557d0 RSI: ffff88810b69d200 RDI: 000000000002d8a0 RBP: 0000000000000002 R08: ffff888110780408 R09: 0000000000000000 R10: ffff88812452e1c0 R11: fffffffffff7e028 R12: 0000000000000000 R13: 0000000000000080 R14: ffff888102c58000 R15: 0000000000000000 FS: 00007f884393a740(0000) GS:ffff8882f5a00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000000000 CR3: 00000001249f6004 CR4: 0000000000370ea0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: mlx5_ib_stage_init_cleanup+0x16/0xd0 [mlx5_ib] __mlx5_ib_remove+0x33/0x90 [mlx5_ib] mlx5r_remove+0x22/0x30 [mlx5_ib] auxiliary_bus_remove+0x18/0x30 __device_release_driver+0x177/0x220 driver_detach+0xc4/0x100 bus_remove_driver+0x58/0xd0 auxiliary_driver_unregister+0x12/0x20 mlx5_ib_cleanup+0x13/0x897 [mlx5_ib] __x64_sys_delete_module+0x154/0x230 ? exit_to_user_mode_prepare+0x104/0x140 do_syscall_64+0x3f/0x80 entry_SYSCALL_64_after_hwframe+0x44/0xae RIP: 0033:0x7f8842e095c7 Code: 73 01 c3 48 8b 0d d9 48 2c 00 f7 d8 64 89 01 48 83 c8 ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 b8 b0 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d a9 48 2c 00 f7 d8 64 89 01 48 RSP: 002b:00007ffc68f6e758 EFLAGS: 00000206 ORIG_RAX: 00000000000000b0 RAX: ffffffffffffffda RBX: 00005638207929c0 RCX: 00007f8842e095c7 RDX: 0000000000000000 RSI: 0000000000000800 RDI: 0000563820792a28 RBP: 00005638207929c0 R08: 00007ffc68f6d701 R09: 0000000000000000 R10: 00007f8842e82880 R11: 0000000000000206 R12: 0000563820792a28 R13: 0000000000000001 R14: 0000563820792a28 R15: 00007ffc68f6fb40 Modules linked in: xt_MASQUERADE nf_conntrack_netlink nfnetlink iptable_nat xt_addrtype xt_conntrack nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 br_netfilter overlay rdma_ucm ib_iser libiscsi scsi_transport_iscsi rdma_cm iw_cm ib_ipoib ib_cm ib_umad mlx5_ib(-) mlx4_ib ib_uverbs ib_core mlx4_en mlx4_core mlx5_core ptp pps_core [last unloaded: rpcrdma] CR2: 0000000000000000 ---[ end trace a0bb7e20804e9e9b ]--- Fixes: 7ce6095e3bff ("RDMA/mlx5: Don't add slave port to unaffiliated list") Link: https://lore.kernel.org/r/899ac1b33a995be5ec0e16a4765c4e43c2b1ba5b.1624956444.git.leonro@nvidia.com Reviewed-by: Itay Aveksis <itayav@nvidia.com> Reviewed-by: Maor Gottlieb <maorg@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-07-14RDMA/mlx5: Don't add slave port to unaffiliated listLeon Romanovsky1-2/+2
[ Upstream commit 7ce6095e3bff8e20ce018b050960b527e298f7df ] The mlx5_ib_bind_slave_port() doesn't remove multiport device from the unaffiliated list, but mlx5_ib_unbind_slave_port() did it. This unbalanced flow caused to the situation where mlx5_ib_unaffiliated_port_list was changed during iteration. Fixes: 32f69e4be269 ("{net, IB}/mlx5: Manage port association for multiport RoCE") Link: https://lore.kernel.org/r/2726e6603b1e6ecfe76aa5a12a063af72173bcf7.1622477058.git.leonro@nvidia.com Reported-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-07-14RDMA/core: Sanitize WQ state received from the userspaceLeon Romanovsky1-4/+2
[ Upstream commit f97442887275d11c88c2899e720fe945c1f61488 ] The mlx4 and mlx5 implemented differently the WQ input checks. Instead of duplicating mlx4 logic in the mlx5, let's prepare the input in the central place. The mlx5 implementation didn't check for validity of state input. It is not real bug because our FW checked that, but still worth to fix. Fixes: f213c0527210 ("IB/uverbs: Add WQ support") Link: https://lore.kernel.org/r/ac41ad6a81b095b1a8ad453dcf62cf8d3c5da779.1621413310.git.leonro@nvidia.com Reported-by: Jiapeng Chong <jiapeng.chong@linux.alibaba.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-07-07RDMA/mlx5: Block FDB rules when not in switchdev modeMark Bloch1-0/+7
commit edc0b0bccc9c80d9a44d3002dcca94984b25e7cf upstream. Allow creating FDB steering rules only when in switchdev mode. The only software model where a userspace application can manipulate FDB entries is when it manages the eswitch. This is only possible in switchdev mode where we expose a single RDMA device with representors for all the vports that are connected to the eswitch. Fixes: 52438be44112 ("RDMA/mlx5: Allow inserting a steering rule to the FDB") Link: https://lore.kernel.org/r/e928ae7c58d07f104716a2a8d730963d1bd01204.1623052923.git.leonro@nvidia.com Reviewed-by: Maor Gottlieb <maorg@nvidia.com> Signed-off-by: Mark Bloch <mbloch@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> [sudip: use old mlx5_eswitch_mode] Signed-off-by: Sudip Mukherjee <sudipm.mukherjee@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2021-06-16IB/mlx5: Fix initializing CQ fragments bufferAlaa Hleihel1-5/+4
commit 2ba0aa2feebda680ecfc3c552e867cf4d1b05a3a upstream. The function init_cq_frag_buf() can be called to initialize the current CQ fragments buffer cq->buf, or the temporary cq->resize_buf that is filled during CQ resize operation. However, the offending commit started to use function get_cqe() for getting the CQEs, the issue with this change is that get_cqe() always returns CQEs from cq->buf, which leads us to initialize the wrong buffer, and in case of enlarging the CQ we try to access elements beyond the size of the current cq->buf and eventually hit a kernel panic. [exception RIP: init_cq_frag_buf+103] [ffff9f799ddcbcd8] mlx5_ib_resize_cq at ffffffffc0835d60 [mlx5_ib] [ffff9f799ddcbdb0] ib_resize_cq at ffffffffc05270df [ib_core] [ffff9f799ddcbdc0] llt_rdma_setup_qp at ffffffffc0a6a712 [llt] [ffff9f799ddcbe10] llt_rdma_cc_event_action at ffffffffc0a6b411 [llt] [ffff9f799ddcbe98] llt_rdma_client_conn_thread at ffffffffc0a6bb75 [llt] [ffff9f799ddcbec8] kthread at ffffffffa66c5da1 [ffff9f799ddcbf50] ret_from_fork_nospec_begin at ffffffffa6d95ddd Fix it by getting the needed CQE by calling mlx5_frag_buf_get_wqe() that takes the correct source buffer as a parameter. Fixes: 388ca8be0037 ("IB/mlx5: Implement fragmented completion queue (CQ)") Link: https://lore.kernel.org/r/90a0e8c924093cfa50a482880ad7e7edb73dc19a.1623309971.git.leonro@nvidia.com Signed-off-by: Alaa Hleihel <alaa@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2021-05-26RDMA/mlx5: Fix query DCT via DEVXMaor Gottlieb1-4/+2
[ Upstream commit cfa3b797118eda7d68f9ede9b1a0279192aca653 ] When executing DEVX command to query QP object, we need to take the QP type from the mlx5_ib_qp struct which hold the driver specific QP types as well, such as DC. Fixes: 34613eb1d2ad ("IB/mlx5: Enable modify and query verbs objects via DEVX") Link: https://lore.kernel.org/r/6eee15d63f09bb70787488e0cf96216e2957f5aa.1621413654.git.leonro@nvidia.com Reviewed-by: Yishai Hadas <yishaih@nvidia.com> Signed-off-by: Maor Gottlieb <maorg@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-05-26RDMA/mlx5: Recover from fatal event in dual port modeMaor Gottlieb1-0/+1
[ Upstream commit 97f30d324ce6645a4de4ffb71e4ae9b8ca36ff04 ] When there is fatal event on the slave port, the device is marked as not active. We need to mark it as active again when the slave is recovered to regain full functionality. Fixes: d69a24e03659 ("IB/mlx5: Move IB event processing onto a workqueue") Link: https://lore.kernel.org/r/8906754455bb23019ef223c725d2c0d38acfb80b.1620711734.git.leonro@nvidia.com Signed-off-by: Maor Gottlieb <maorg@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-05-14RDMA/mlx5: Fix drop packet rule in egress tableMaor Gottlieb1-4/+5
[ Upstream commit c73700806d4e430d182c2be069d230076818a99a ] Initial drop action support missed that drop action can be added to egress flow tables as well. Add the missing support. This requires making sure that dest_type isn't set to PORT which in turn exposes a possibility of passing dst while indicating number of dsts as zero. Explicitly check for number of dsts and pass the appropriate pointer. Fixes: f29de9eee782 ("RDMA/mlx5: Add support for drop action in DV steering") Link: https://lore.kernel.org/r/20210318135123.680759-1-leon@kernel.org Reviewed-by: Mark Bloch <markb@nvidia.com> Signed-off-by: Maor Gottlieb <maorg@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-05-14RDMA/mlx5: Fix mlx5 rates to IB rates mapMark Zhang1-1/+14
[ Upstream commit 6fe6e568639859db960c8fcef19a2ece1c2d7eae ] Correct the map between mlx5 rates and corresponding ib rates, as they don't always have a fixed offset between them. Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters") Link: https://lore.kernel.org/r/20210304124517.1100608-4-leon@kernel.org Signed-off-by: Mark Zhang <markzhang@nvidia.com> Reviewed-by: Maor Gottlieb <maorg@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-03-25RDMA/mlx5: Allow creating all QPs even when non RDMA profile is usedMark Bloch1-6/+20
[ Upstream commit 2614488d1f3cd5989375042286b11424208e20c8 ] The cited commit disallowed creating any QP which isn't raw ethernet, reg umr or the special UD qp for testing WC, this proved too strict. While modify can't be done (no GIDS/GID table for example) just creating a QP is okay. This patch partially reverts the bellow mentioned commit and places the restriction at the modify QP stage and not at the creation. DEVX commands should be used to manipulate such QPs. Fixes: 42caf9cb5937 ("RDMA/mlx5: Allow only raw Ethernet QPs when RoCE isn't enabled") Link: https://lore.kernel.org/r/20210125120709.836718-1-leon@kernel.org Signed-off-by: Mark Bloch <mbloch@nvidia.com> Reviewed-by: Maor Gottlieb <maorg@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-03-09IB/mlx5: Add missing error codeYueHaibing1-1/+3
[ Upstream commit 3a9b3d4536e0c25bd3906a28c1f584177e49dd0f ] Set err to -ENOMEM if kzalloc fails instead of 0. Fixes: 759738537142 ("IB/mlx5: Enable subscription for device events over DEVX") Link: https://lore.kernel.org/r/20210222122343.19720-1-yuehaibing@huawei.com Signed-off-by: YueHaibing <yuehaibing@huawei.com> Acked-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-03-04IB/mlx5: Return appropriate error code instead of ENOMEMParav Pandit1-2/+1
[ Upstream commit d286ac1d05210695c312b9018b3aa7c2048e9aca ] When mlx5_ib_stage_init_init() fails, return the error code related to failure instead of -ENOMEM. Fixes: 16c1975f1032 ("IB/mlx5: Create profile infrastructure to add and remove stages") Link: https://lore.kernel.org/r/20210127150010.1876121-8-leon@kernel.org Signed-off-by: Parav Pandit <parav@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-03-04IB/mlx5: Add mutex destroy call to cap_mask_mutex mutexParav Pandit1-6/+5
[ Upstream commit ab40530a2e0a7aca9a5187824c4fb072f3916e85 ] mutex_destroy() call for device's cap_mask_mutex mutex is missing, let's add it to annotate destruction. Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters") Link: https://lore.kernel.org/r/20210113121703.559778-4-leon@kernel.org Signed-off-by: Parav Pandit <parav@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-03-04RDMA/mlx5: Use the correct obj_id upon DEVX TIR creationYishai Hadas1-1/+3
[ Upstream commit 8798e4ad0abe0ba1221928a46561981c510be0c6 ] Use the correct obj_id upon DEVX TIR creation by strictly taking the tirn 24 bits and not the general obj_id which is 32 bits. Fixes: 7efce3691d33 ("IB/mlx5: Add obj create and destroy functionality") Link: https://lore.kernel.org/r/20201230130121.180350-2-leon@kernel.org Signed-off-by: Yishai Hadas <yishaih@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-02-04Revert "RDMA/mlx5: Fix devlink deadlock on net namespace deletion"Parav Pandit1-4/+2
commit de641d74fb00f5b32f054ee154e31fb037e0db88 upstream. This reverts commit fbdd0049d98d44914fc57d4b91f867f4996c787b. Due to commit in fixes tag, netdevice events were received only in one net namespace of mlx5_core_dev. Due to this when netdevice events arrive in net namespace other than net namespace of mlx5_core_dev, they are missed. This results in empty GID table due to RDMA device being detached from its net device. Hence, revert back to receive netdevice events in all net namespaces to restore back RDMA functionality in non init_net net namespace. The deadlock will have to be addressed in another patch. Fixes: fbdd0049d98d ("RDMA/mlx5: Fix devlink deadlock on net namespace deletion") Link: https://lore.kernel.org/r/20210117092633.10690-1-leon@kernel.org Signed-off-by: Parav Pandit <parav@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2021-01-19IB/mlx5: Fix error unwinding when set_has_smi_cap failsParav Pandit1-1/+1
commit 2cb091f6293df898b47f4e0f2e54324e2bbaf816 upstream. When set_has_smi_cap() fails, multiport master cleanup is missed. Fix it by doing the correct error unwinding goto. Fixes: a989ea01cb10 ("RDMA/mlx5: Move SMI caps logic") Link: https://lore.kernel.org/r/20210113121703.559778-3-leon@kernel.org Signed-off-by: Parav Pandit <parav@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2021-01-19RDMA/mlx5: Fix wrong free of blue flame register on errorMark Bloch1-1/+1
commit 1c3aa6bd0b823105c2030af85d92d158e815d669 upstream. If the allocation of the fast path blue flame register fails, the driver should free the regular blue flame register allocated a statement above, not the one that it just failed to allocate. Fixes: 16c1975f1032 ("IB/mlx5: Create profile infrastructure to add and remove stages") Link: https://lore.kernel.org/r/20210113121703.559778-6-leon@kernel.org Reported-by: Hans Petter Selasky <hanss@nvidia.com> Signed-off-by: Mark Bloch <mbloch@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-12-30RDMA/mlx5: Fix MR cache memory leakMaor Gottlieb1-0/+1
[ Upstream commit e89938902927a54abebccc9537991aca5237dfaf ] If the MR cache entry invalidation failed, then we detach this entry from the cache, therefore we must to free the memory as well. Allcation backtrace for the leaker: [<00000000d8e423b0>] alloc_cache_mr+0x23/0xc0 [mlx5_ib] [<000000001f21304c>] create_cache_mr+0x3f/0xf0 [mlx5_ib] [<000000009d6b45dc>] mlx5_ib_alloc_implicit_mr+0x41/0×210 [mlx5_ib] [<00000000879d0d68>] mlx5_ib_reg_user_mr+0x9e/0×6e0 [mlx5_ib] [<00000000be74bf89>] create_qp+0x2fc/0xf00 [ib_uverbs] [<000000001a532d22>] ib_uverbs_handler_UVERBS_METHOD_COUNTERS_READ+0x1d9/0×230 [ib_uverbs] [<0000000070f46001>] rdma_alloc_commit_uobject+0xb5/0×120 [ib_uverbs] [<000000006d8a0b38>] uverbs_alloc+0x2b/0xf0 [ib_uverbs] [<00000000075217c9>] ksysioctl+0x234/0×7d0 [<00000000eb5c120b>] __x64_sys_ioctl+0x16/0×20 [<00000000db135b48>] do_syscall_64+0x59/0×2e0 Fixes: 1769c4c57548 ("RDMA/mlx5: Always remove MRs from the cache before destroying them") Link: https://lore.kernel.org/r/20201213132940.345554-2-leon@kernel.org Signed-off-by: Maor Gottlieb <maorg@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2020-12-30RDMA/mlx5: Fix corruption of reg_pages in mlx5_ib_rereg_user_mr()Jason Gunthorpe1-9/+11
[ Upstream commit fc3325701a6353594083f08e297d4c1965c601aa ] reg_pages should always contain mr->npage since when the mr is finally de-reg'd it is always subtracted out. If there were any error exits then mlx5_ib_rereg_user_mr() would leave the reg_pages adjusted and this will cause it to be double subtracted eventually. The manipulation of reg_pages is inherently connected to the umem, so lift it out of set_mr_fields() and only adjust it around creating/destroying a umem. reg_pages is only used for diagnostics in sysfs. Fixes: 7d0cc6edcc70 ("IB/mlx5: Add MR cache for large UMR regions") Link: https://lore.kernel.org/r/20201026131936.1335664-3-leon@kernel.org Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
2020-10-27RDMA/mlx5: Fix devlink deadlock on net namespace deletionParav Pandit1-2/+4
When a mlx5 core devlink instance is reloaded in different net namespace, its associated IB device is deleted and recreated. Example sequence is: $ ip netns add foo $ devlink dev reload pci/0000:00:08.0 netns foo $ ip netns del foo mlx5 IB device needs to attach and detach the netdevice to it through the netdev notifier chain during load and unload sequence. A below call graph of the unload flow. cleanup_net() down_read(&pernet_ops_rwsem); <- first sem acquired ops_pre_exit_list() pre_exit() devlink_pernet_pre_exit() devlink_reload() mlx5_devlink_reload_down() mlx5_unload_one() [...] mlx5_ib_remove() mlx5_ib_unbind_slave_port() mlx5_remove_netdev_notifier() unregister_netdevice_notifier() down_write(&pernet_ops_rwsem);<- recurrsive lock Hence, when net namespace is deleted, mlx5 reload results in deadlock. When deadlock occurs, devlink mutex is also held. This not only deadlocks the mlx5 device under reload, but all the processes which attempt to access unrelated devlink devices are deadlocked. Hence, fix this by mlx5 ib driver to register for per net netdev notifier instead of global one, which operats on the net namespace without holding the pernet_ops_rwsem. Fixes: 4383cfcc65e7 ("net/mlx5: Add devlink reload") Link: https://lore.kernel.org/r/20201026134359.23150-1-parav@nvidia.com Signed-off-by: Parav Pandit <parav@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-10-16RDMA: Explicitly pass in the dma_device to ib_register_deviceJason Gunthorpe1-1/+1
The code in setup_dma_device has become rather convoluted, move all of this to the drivers. Drives now pass in a DMA capable struct device which will be used to setup DMA, or drivers must fully configure the ibdev for DMA and pass in NULL. Other than setting the masks in rvt all drivers were doing this already anyhow. mthca, mlx4 and mlx5 were already setting up maximum DMA segment size for DMA based on their hardweare limits in: __mthca_init_one() dma_set_max_seg_size (1G) __mlx4_init_one() dma_set_max_seg_size (1G) mlx5_pci_init() set_dma_caps() dma_set_max_seg_size (2G) Other non software drivers (except usnic) were extended to UINT_MAX [1, 2] instead of 2G as was before. [1] https://lore.kernel.org/linux-rdma/20200924114940.GE9475@nvidia.com/ [2] https://lore.kernel.org/linux-rdma/20200924114940.GE9475@nvidia.com/ Link: https://lore.kernel.org/r/20201008082752.275846-1-leon@kernel.org Link: https://lore.kernel.org/r/6b2ed339933d066622d5715903870676d8cc523a.1602590106.git.mchehab+huawei@kernel.org Suggested-by: Christoph Hellwig <hch@infradead.org> Signed-off-by: Parav Pandit <parav@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-10-02RDMA/core: Modify enum ib_gid_type and enum rdma_network_typeAvihai Horon2-3/+3
Separate IB_GID_TYPE_IB and IB_GID_TYPE_ROCE to two different values, so enum ib_gid_type will match the gid types of the new query GID table API which will be introduced in the following patches. This change in enum ib_gid_type requires to change also enum rdma_network_type by separating RDMA_NETWORK_IB and RDMA_NETWORK_ROCE_V1 values. Link: https://lore.kernel.org/r/20200923165015.2491894-3-leon@kernel.org Signed-off-by: Avihai Horon <avihaih@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-10-01RDMA/mlx5: Sync device with CPU pages upon ODP MR registrationYishai Hadas3-6/+32
Sync device with CPU pages upon ODP MR registration. mlx5 already has to zero the HW's version of the PAS list, may as well deliver a PAS list that matches the current CPU page tables configuration. Link: https://lore.kernel.org/r/20200930163828.1336747-5-leon@kernel.org Signed-off-by: Yishai Hadas <yishaih@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-10-01RDMA/mlx5: Extend advice MR to support non faulting modeYishai Hadas2-2/+8
Extend advice MR to support non faulting mode, this can improve performance by increasing the populated page tables in the device. Link: https://lore.kernel.org/r/20200930163828.1336747-4-leon@kernel.org Signed-off-by: Yishai Hadas <yishaih@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-10-01IB/core: Enable ODP sync without faultingYishai Hadas1-1/+1
Enable ODP sync without faulting, this improves performance by reducing the number of page faults in the system. The gain from this option is that the device page table can be aligned with the presented pages in the CPU page table without causing page faults. As of that, the overhead on data path from hardware point of view to trigger a fault which end-up by calling the driver to bring the pages will be dropped. Link: https://lore.kernel.org/r/20200930163828.1336747-3-leon@kernel.org Signed-off-by: Yishai Hadas <yishaih@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-10-01IB/core: Improve ODP to use hmm_range_fault()Yishai Hadas1-17/+7
Move to use hmm_range_fault() instead of get_user_pags_remote() to improve performance in a few aspects: This includes: - Dropping the need to allocate and free memory to hold its output - No need any more to use put_page() to unpin the pages - The logic to detect contiguous pages is done based on the returned order, no need to run per page and evaluate. In addition, moving to use hmm_range_fault() enables to reduce page faults in the system with it's snapshot mode, this will be introduced in next patches from this series. As part of this, cleanup some flows and use the required data structures to work with hmm_range_fault(). Link: https://lore.kernel.org/r/20200930163828.1336747-2-leon@kernel.org Signed-off-by: Yishai Hadas <yishaih@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-29RDMA/drivers: Remove udata check from special QPLeon Romanovsky1-12/+0
GSI QP can't be created from the user space, hence the udata check is always false (udata == NULL). Remove that check and simplify the flow. Link: https://lore.kernel.org/r/20200926102450.2966017-9-leon@kernel.org Reviewed-by: Maor Gottlieb <maorg@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-29RDMA/mlx5: Delete not needed GSI QP signal QP typeLeon Romanovsky2-8/+1
GSI QP doesn't need signal QP type because it is initialized statically to zero, which is IB_SIGNAL_ALL_WR also wr->send_flags isn't set too. This means that the GSI QP signal QP type can be removed. Link: https://lore.kernel.org/r/20200926102450.2966017-5-leon@kernel.org Reviewed-by: Maor Gottlieb <maorg@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-29RDMA/mlx5: Change GSI QP to have same creation flow like other QPsLeon Romanovsky3-46/+38
There is no reason to have separate create flow for the GSI QP, while general create_qp routine has all needed checks and ability to allocate and free the proper struct mlx5_ib_qp. Link: https://lore.kernel.org/r/20200926102450.2966017-4-leon@kernel.org Reviewed-by: Maor Gottlieb <maorg@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-29RDMA/mlx5: Reuse existing fields in parent QP storage objectLeon Romanovsky3-45/+31
Remove duplication of mlx5_ib_qp and mlx5_ib_gsi_qp fields. This change returns the memory footprint of mlx5_ib QP to be as it was before embedding GSI QP. Link: https://lore.kernel.org/r/20200926102450.2966017-3-leon@kernel.org Reviewed-by: Maor Gottlieb <maorg@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-29RDMA/mlx5: Embed GSI QP into general mlx5_ib QPLeon Romanovsky3-33/+32
The GSI QPs have different create flow from the regular QPs, but it is not really needed. Update the code to use mlx5_ib_qp as a storage class for all outside of GSI calls. Link: https://lore.kernel.org/r/20200926102450.2966017-2-leon@kernel.org Reviewed-by: Maor Gottlieb <maorg@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-25RDMA/mlx5: Fix type warning of sizeof in __mlx5_ib_alloc_counters()Liu Shixin1-2/+2
sizeof() when applied to a pointer typed expression should give the size of the pointed data, even if the data is a pointer. Fixes: e1f24a79f424 ("IB/mlx5: Support congestion related counters") Link: https://lore.kernel.org/r/20200917081354.2083293-1-liushixin2@huawei.com Signed-off-by: Liu Shixin <liushixin2@huawei.com> Acked-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-23RDMA/mlx5: Don't call to restrack recursivelyLeon Romanovsky1-2/+14
The restrack is going to manage memory of all IB objects and must be called before object is created. GSI QP in the mlx5_ib separated between creating dummy interface and HW object beneath. This was achieved by double call to ib_create_qp(). In order to skip such reentry call to internal driver create_qp code. Link: https://lore.kernel.org/r/20200922091106.2152715-3-leon@kernel.org Reviewed-by: Mark Zhang <markz@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-18RDMA/mlx5: Clarify what the UMR is for when creating MRsJason Gunthorpe4-49/+97
Once a mkey is created it can be modified using UMR. This is desirable for performance reasons. However, different hardware has restrictions on what modifications are possible using UMR. Make sense of these checks: - mlx5_ib_can_reconfig_with_umr() returns true if the access flags can be altered. Most cases create MRs using 0 access flags (now made clear by consistent use of set_mkc_access_pd_addr_fields()), but the old logic here was tormented. Make it clear that this is checking if the current access_flags can be modified using UMR to different access_flags. It is always OK to use UMR to change flags that all HW supports. - mlx5_ib_can_load_pas_with_umr() returns true if UMR can be used to enable and update the PAS/XLT. Enabling requires updating the entity size, so UMR ends up completely disabled on this old hardware. Make it clear why it is disabled. FRWR, ODP and cache always requires mlx5_ib_can_load_pas_with_umr(). - mlx5_ib_pas_fits_in_mr() is used to tell if an existing MR can be resized to hold a new PAS list. This only works for cached MR's because we don't store the PAS list size in other cases. To be very clear, arrange things so any pre-created MR's in the cache check the newly requested access_flags before allowing the MR to leave the cache. If UMR cannot set the required access_flags the cache fails to create the MR. This in turn means relaxed ordering and atomic are now correctly blocked early for implicit ODP on older HW. Link: https://lore.kernel.org/r/20200914112653.345244-6-leon@kernel.org Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-18RDMA/mlx5: Disable IB_DEVICE_MEM_MGT_EXTENSIONS if IB_WR_REG_MR can't workJason Gunthorpe1-1/+3
set_reg_wr() always fails if !umr_modify_entity_size_disabled because mlx5_ib_can_use_umr() always fails. Without set_reg_wr() IB_WR_REG_MR doesn't work and that means the device should not advertise IB_DEVICE_MEM_MGT_EXTENSIONS. Fixes: 841b07f99a47 ("IB/mlx5: Block MR WR if UMR is not possible") Link: https://lore.kernel.org/r/20200914112653.345244-5-leon@kernel.org Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-18RDMA/mlx5: Make mkeys always owned by the kernel's PD when not enabledJason Gunthorpe1-25/+26
Any mkey that is not enabled and assigned to userspace should have the PD set to a kernel owned PD. When cache entries are created for the first time the PDN is set to 0, which is probably a kernel PD, but be explicit. When a MR is registered using the hybrid reg_create with UMR xlt & enable the disabled mkey is pointing at the user PD, keep it pointing at the kernel until a UMR enables it and sets the user PD. Fixes: 9ec4483a3f0f ("IB/mlx5: Move MRs to a kernel PD when freeing them to the MR cache") Link: https://lore.kernel.org/r/20200914112653.345244-4-leon@kernel.org Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-18RDMA/mlx5: Use set_mkc_access_pd_addr_fields() in reg_create()Jason Gunthorpe1-14/+1
reg_create() open codes this helper, use the shared code. Link: https://lore.kernel.org/r/20200914112653.345244-3-leon@kernel.org Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-18RDMA/mlx5: Remove dead check for EAGAIN after alloc_mr_from_cache()Jason Gunthorpe1-3/+1
alloc_mr_from_cache() no longer returns EAGAIN, this is just dead code now. Fixes: aad719dcf379 ("RDMA/mlx5: Allow MRs to be created in the cache synchronously") Link: https://lore.kernel.org/r/20200914112653.345244-2-leon@kernel.org Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-18Merge branch 'mlx_sw_owner_v2' into rdma.git for-nextJason Gunthorpe2-3/+7
Leon Romanovsky says: ==================== This series from Alex extends software steering interface to support devices with extra capability "sw_owner_2" which will replace existing "sw_owner". ==================== Based on the mlx5-next branch at git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux due to dependencies. * branch 'mlx5_sw_owner_v2: RDMA/mlx5: Expose TIR and QP ICM address for sw_owner_v2 devices RDMA/mlx5: Allow DM allocation for sw_owner_v2 enabled devices RDMA/mlx5: Add sw_owner_v2 bit capability
2020-09-18Merge branch 'mlx5_active_speed' into rdma.git for-nextJason Gunthorpe1-26/+15
Leon Romanovsky says: ==================== IBTA declares speed as 16 bits, but kernel stores it in u8. This series fixes in-kernel declaration while keeping external interface intact. ==================== Based on the mlx5-next branch at git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux due to dependencies. * branch 'mlx5_active_speed': RDMA: Fix link active_speed size RDMA/mlx5: Delete duplicated mlx5_ptys_width enum net/mlx5: Refactor query port speed functions
2020-09-18RDMA: Fix link active_speed sizeAharon Landau1-6/+2
According to the IB spec active_speed size should be u16 and not u8 as before. Changing it to allow further extensions in offered speeds. Link: https://lore.kernel.org/r/20200917090223.1018224-4-leon@kernel.org Signed-off-by: Aharon Landau <aharonl@mellanox.com> Reviewed-by: Michael Guralnik <michaelgur@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-18RDMA/mlx5: Expose TIR and QP ICM address for sw_owner_v2 devicesAlex Vesker1-2/+4
Expose the ICM address to access TIR and QP, this will allow sw_owned_v2 devices to steer traffic to TIRs and QPs same as done with sw_owner capability. Link: https://lore.kernel.org/r/20200903073857.1129166-4-leon@kernel.org Signed-off-by: Alex Vesker <valex@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-18RDMA/mlx5: Allow DM allocation for sw_owner_v2 enabled devicesAlex Vesker1-1/+3
sw_owner_v2 will replace sw_owner for future devices, this means that if sw_owner_v2 is set sw_owner should be ignored and DM allocation is required for sw_owner_v2 devices to function. Link: https://lore.kernel.org/r/20200903073857.1129166-3-leon@kernel.org Signed-off-by: Alex Vesker <valex@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-17RDMA: Convert RWQ table logic to ib_core allocation schemeLeon Romanovsky5-32/+24
Move struct ib_rwq_ind_table allocation to ib_core. Link: https://lore.kernel.org/r/20200902081623.746359-3-leon@kernel.org Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-17RDMA: Clean MW allocation and free flowsLeon Romanovsky3-27/+18
Move allocation and destruction of memory windows under ib_core responsibility and clean drivers to ensure that no updates to MW ib_core structures are done in driver layer. Link: https://lore.kernel.org/r/20200902081623.746359-2-leon@kernel.org Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
2020-09-17RDMA/mlx5: Delete duplicated mlx5_ptys_width enumAharon Landau1-14/+6
Combine two same enums to avoid duplication. Signed-off-by: Aharon Landau <aharonl@mellanox.com> Reviewed-by: Michael Guralnik <michaelgur@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
generated by cgit v1.2.3 (git 2.39.0) at 2025-01-18 13:38:06 +0300