kernel/linux.git/drivers/infiniband, branch v6.1.124

RDMA/uverbs: Prevent integer overflow issue

2025-01-09T12:30:06+00:00

commit d0257e089d1bbd35c69b6c97ff73e3690ab149a9 upstream. In the expression "cmd.wqe_size * cmd.wr_count", both variables are u32 values that come from the user so the multiplication can lead to integer wrapping. Then we pass the result to uverbs_request_next_ptr() which also could potentially wrap. The "cmd.sge_count * sizeof(struct ib_uverbs_sge)" multiplication can also overflow on 32bit systems although it's fine on 64bit systems. This patch does two things. First, I've re-arranged the condition in uverbs_request_next_ptr() so that the use controlled variable "len" is on one side of the comparison by itself without any math. Then I've modified all the callers to use size_mul() for the multiplications. Fixes: 67cdb40ca444 ("[IB] uverbs: Implement more commands") Cc: stable@vger.kernel.org Signed-off-by: Dan Carpenter Link: https://patch.msgid.link/b8765ab3-c2da-4611-aae0-ddd6ba173d23@stanley.mountain Signed-off-by: Leon Romanovsky Signed-off-by: Greg Kroah-Hartman

RDMA/rtrs: Ensure 'ib_sge list' is accessible

2025-01-09T12:30:02+00:00

[ Upstream commit fb514b31395946022f13a08e06a435f53cf9e8b3 ] Move the declaration of the 'ib_sge list' variable outside the 'always_invalidate' block to ensure it remains accessible for use throughout the function. Previously, 'ib_sge list' was declared within the 'always_invalidate' block, limiting its accessibility, then caused a 'BUG: kernel NULL pointer dereference'[1]. ? __die_body.cold+0x19/0x27 ? page_fault_oops+0x15a/0x2d0 ? search_module_extables+0x19/0x60 ? search_bpf_extables+0x5f/0x80 ? exc_page_fault+0x7e/0x180 ? asm_exc_page_fault+0x26/0x30 ? memcpy_orig+0xd5/0x140 rxe_mr_copy+0x1c3/0x200 [rdma_rxe] ? rxe_pool_get_index+0x4b/0x80 [rdma_rxe] copy_data+0xa5/0x230 [rdma_rxe] rxe_requester+0xd9b/0xf70 [rdma_rxe] ? finish_task_switch.isra.0+0x99/0x2e0 rxe_sender+0x13/0x40 [rdma_rxe] do_task+0x68/0x1e0 [rdma_rxe] process_one_work+0x177/0x330 worker_thread+0x252/0x390 ? __pfx_worker_thread+0x10/0x10 This change ensures the variable is available for subsequent operations that require it. [1] https://lore.kernel.org/linux-rdma/6a1f3e8f-deb0-49f9-bc69-a9b03ecfcda7@fujitsu.com/ Fixes: 9cb837480424 ("RDMA/rtrs: server: main functionality") Signed-off-by: Li Zhijian Link: https://patch.msgid.link/20241231013416.1290920-1-lizhijian@fujitsu.com Signed-off-by: Leon Romanovsky Signed-off-by: Sasha Levin

RDMA/hns: Fix missing flush CQE for DWQE

2025-01-09T12:29:59+00:00

[ Upstream commit e3debdd48423d3d75b9d366399228d7225d902cd ] Flush CQE handler has not been called if QP state gets into errored mode in DWQE path. So, the new added outstanding WQEs will never be flushed. It leads to a hung task timeout when using NFS over RDMA: __switch_to+0x7c/0xd0 __schedule+0x350/0x750 schedule+0x50/0xf0 schedule_timeout+0x2c8/0x340 wait_for_common+0xf4/0x2b0 wait_for_completion+0x20/0x40 __ib_drain_sq+0x140/0x1d0 [ib_core] ib_drain_sq+0x98/0xb0 [ib_core] rpcrdma_xprt_disconnect+0x68/0x270 [rpcrdma] xprt_rdma_close+0x20/0x60 [rpcrdma] xprt_autoclose+0x64/0x1cc [sunrpc] process_one_work+0x1d8/0x4e0 worker_thread+0x154/0x420 kthread+0x108/0x150 ret_from_fork+0x10/0x18 Fixes: 01584a5edcc4 ("RDMA/hns: Add support of direct wqe") Signed-off-by: Chengchang Tang Signed-off-by: Junxian Huang Link: https://patch.msgid.link/20241220055249.146943-5-huangjunxian6@hisilicon.com Signed-off-by: Leon Romanovsky Signed-off-by: Sasha Levin

RDMA/hns: Fix warning storm caused by invalid input in IO path

2025-01-09T12:29:59+00:00

[ Upstream commit fa5c4ba8cdbfd2c2d6422e001311c8213283ebbf ] WARN_ON() is called in the IO path. And it could lead to a warning storm. Use WARN_ON_ONCE() instead of WARN_ON(). Fixes: 12542f1de179 ("RDMA/hns: Refactor process about opcode in post_send()") Signed-off-by: Chengchang Tang Signed-off-by: Junxian Huang Link: https://patch.msgid.link/20241220055249.146943-4-huangjunxian6@hisilicon.com Signed-off-by: Leon Romanovsky Signed-off-by: Sasha Levin

RDMA/hns: Fix mapping error of zero-hop WQE buffer

2025-01-09T12:29:59+00:00

[ Upstream commit 8673a6c2d9e483dfeeef83a1f06f59e05636f4d1 ] Due to HW limitation, the three region of WQE buffer must be mapped and set to HW in a fixed order: SQ buffer, SGE buffer, and RQ buffer. Currently when one region is zero-hop while the other two are not, the zero-hop region will not be mapped. This violate the limitation above and leads to address error. Fixes: 38389eaa4db1 ("RDMA/hns: Add mtr support for mixed multihop addressing") Signed-off-by: wenglianfa Signed-off-by: Junxian Huang Link: https://patch.msgid.link/20241220055249.146943-2-huangjunxian6@hisilicon.com Signed-off-by: Leon Romanovsky Signed-off-by: Sasha Levin

RDMA/hns: Remove unused parameters and variables

2025-01-09T12:29:59+00:00

[ Upstream commit f4caa864af84f801a5821ea2ba6c1cc46f8252c1 ] Remove unused parameters and variables. Signed-off-by: Chengchang Tang Signed-off-by: Junxian Huang Link: https://lore.kernel.org/r/20240412091616.370789-3-huangjunxian6@hisilicon.com Signed-off-by: Leon Romanovsky Stable-dep-of: 8673a6c2d9e4 ("RDMA/hns: Fix mapping error of zero-hop WQE buffer") Signed-off-by: Sasha Levin

RDMA/hns: Refactor mtr find

2025-01-09T12:29:58+00:00

[ Upstream commit a4ca341080758d847db155b97887bff6f84016a4 ] hns_roce_mtr_find() is a collection of multiple functions, and the return value is also difficult to understand, which is not conducive to modification and maintenance. Separate the function of obtaining MTR root BA from this function. And some adjustments has been made to improve readability. Signed-off-by: Chengchang Tang Signed-off-by: Junxian Huang Link: https://lore.kernel.org/r/20240113085935.2838701-2-huangjunxian6@hisilicon.com Signed-off-by: Leon Romanovsky Stable-dep-of: 8673a6c2d9e4 ("RDMA/hns: Fix mapping error of zero-hop WQE buffer") Signed-off-by: Sasha Levin

RDMA/bnxt_re: Fix the locking while accessing the QP table

2025-01-09T12:29:58+00:00

[ Upstream commit 9272cba0ded71b5a2084da3004ec7806b8cb7fd2 ] QP table handling is synchronized with destroy QP and Async event from the HW. The same needs to be synchronized during create_qp also. Use the same lock in create_qp also. Fixes: 76d3ddff7153 ("RDMA/bnxt_re: synchronize the qp-handle table array") Fixes: f218d67ef004 ("RDMA/bnxt_re: Allow posting when QPs are in error") Fixes: 84cf229f4001 ("RDMA/bnxt_re: Fix the qp table indexing") Signed-off-by: Selvin Xavier Link: https://patch.msgid.link/20241217102649.1377704-6-kalesh-anakkur.purayil@broadcom.com Signed-off-by: Leon Romanovsky Signed-off-by: Sasha Levin

RDMA/bnxt_re: Fix max_qp_wrs reported

2025-01-09T12:29:58+00:00

[ Upstream commit 40be32303ec829ea12f9883e499bfd3fe9e52baf ] While creating qps, driver adds one extra entry to the sq size passed by the ULPs in order to avoid queue full condition. When ULPs creates QPs with max_qp_wr reported, driver creates QP with 1 more than the max_wqes supported by HW. Create QP fails in this case. To avoid this error, reduce 1 entry in max_qp_wqes and report it to the stack. Fixes: 1ac5a4047975 ("RDMA/bnxt_re: Add bnxt_re RoCE driver") Reviewed-by: Kalesh AP Signed-off-by: Selvin Xavier Link: https://patch.msgid.link/20241217102649.1377704-2-kalesh-anakkur.purayil@broadcom.com Signed-off-by: Leon Romanovsky Signed-off-by: Sasha Levin

RDMA/bnxt_re: Fix reporting hw_ver in query_device

2025-01-09T12:29:57+00:00

[ Upstream commit 7179fe0074a3c962e43a9e51169304c4911989ed ] Driver currently populates subsystem_device id in the "hw_ver" field of ib_attr structure in query_device. Updated to populate PCI revision ID. Fixes: 1ac5a4047975 ("RDMA/bnxt_re: Add bnxt_re RoCE driver") Reviewed-by: Preethi G Signed-off-by: Kalesh AP Signed-off-by: Selvin Xavier Link: https://patch.msgid.link/20241211083931.968831-6-kalesh-anakkur.purayil@broadcom.com Signed-off-by: Leon Romanovsky Signed-off-by: Sasha Levin