linux_dsm_epyc7002/drivers/infiniband/hw
Mark Bloch 42cea83f95 IB/mlx5: Fix cleanup order on unload
On load we create private CQ/QP/PD in order to be used by UMR, we create
those resources after we register ourself as an IB device, and we destroy
them after we unregister as an IB device. This was changed by commit
16c1975f10 ("IB/mlx5: Create profile infrastructure to add and remove
stages") which moved the destruction before we unregistration. This
allowed to trigger an invalid memory access when unloading mlx5_ib while
there are open resources:

BUG: unable to handle kernel paging request at 00000001002c012c
...
Call Trace:
 mlx5_ib_post_send_wait+0x75/0x110 [mlx5_ib]
 __slab_free+0x9a/0x2d0
 delay_time_func+0x10/0x10 [mlx5_ib]
 unreg_umr.isra.15+0x4b/0x50 [mlx5_ib]
 mlx5_mr_cache_free+0x46/0x150 [mlx5_ib]
 clean_mr+0xc9/0x190 [mlx5_ib]
 dereg_mr+0xba/0xf0 [mlx5_ib]
 ib_dereg_mr+0x13/0x20 [ib_core]
 remove_commit_idr_uobject+0x16/0x70 [ib_uverbs]
 uverbs_cleanup_ucontext+0xe8/0x1a0 [ib_uverbs]
 ib_uverbs_cleanup_ucontext.isra.9+0x19/0x40 [ib_uverbs]
 ib_uverbs_remove_one+0x162/0x2e0 [ib_uverbs]
 ib_unregister_device+0xd4/0x190 [ib_core]
 __mlx5_ib_remove+0x2e/0x40 [mlx5_ib]
 mlx5_remove_device+0xf5/0x120 [mlx5_core]
 mlx5_unregister_interface+0x37/0x90 [mlx5_core]
 mlx5_ib_cleanup+0xc/0x225 [mlx5_ib]
 SyS_delete_module+0x153/0x230
 do_syscall_64+0x62/0x110
 entry_SYSCALL_64_after_hwframe+0x21/0x86
...

We restore the original behavior by breaking the UMR stage into two parts,
pre and post IB registration stages, this way we can restore the original
functionality and maintain clean separation of logic between stages.

Fixes: 16c1975f10 ("IB/mlx5: Create profile infrastructure to add and remove stages")
Signed-off-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
2018-03-14 16:44:02 -04:00
..
bnxt_re RDMA/bnxt_re: Avoid Hard lockup during error CQE processing 2018-03-06 20:08:39 -07:00
cxgb3 Updates for 4.15 kernel merge window 2017-11-15 14:54:53 -08:00
cxgb4 iw_cxgb4: Change error/warn prints to pr_debug 2017-12-29 11:09:23 -07:00
hfi1 vfs: do bulk POLL* -> EPOLL* replacement 2018-02-11 14:34:03 -08:00
hns RDMA/hns: Fix the endian problem for hns 2018-02-05 10:48:48 -05:00
i40iw i40iw: Free IEQ resources 2018-01-16 20:38:18 -07:00
mlx4 IB/mlx4: Include GID type when deleting GIDs from HW table under RoCE 2018-03-06 20:08:38 -07:00
mlx5 IB/mlx5: Fix cleanup order on unload 2018-03-14 16:44:02 -04:00
mthca IB/mthca: remove mthca_user.h 2018-01-28 14:07:16 -07:00
nes nes: Change accelerated flag to bool 2017-12-22 13:33:30 -07:00
ocrdma IB/ocrdma: Use zeroing memory allocator than allocator/memset 2018-01-02 11:20:13 -07:00
qedr RDMA/qedr: Fix iWARP write and send with immediate 2018-03-06 19:57:37 -07:00
qib vfs: do bulk POLL* -> EPOLL* replacement 2018-02-11 14:34:03 -08:00
usnic drivers: infiniband: remove duplicate includes 2017-12-22 09:39:35 -07:00
vmw_pvrdma RDMA/vmw_pvrdma: Fix usage of user response structures in ABI file 2018-02-15 15:31:28 -07:00
Makefile License cleanup: add SPDX GPL-2.0 license identifier to files with no license 2017-11-02 11:10:55 +01:00