kernel/linux.git/drivers/infiniband/core/verbs.c, branch v6.2

RDMA/core: Fix ib block iterator counter overflow

2023-01-10T10:21:39+00:00

When registering a new DMA MR after selecting the best aligned page size for it, we iterate over the given sglist to split each entry to smaller, aligned to the selected page size, DMA blocks. In given circumstances where the sg entry and page size fit certain sizes and the sg entry is not aligned to the selected page size, the total size of the aligned pages we need to cover the sg entry is >= 4GB. Under this circumstances, while iterating page aligned blocks, the counter responsible for counting how much we advanced from the start of the sg entry is overflowed because its type is u32 and we pass 4GB in size. This can lead to an infinite loop inside the iterator function because the overflow prevents the counter to be larger than the size of the sg entry. Fix the presented problem by changing the advancement condition to eliminate overflow. Backtrace: [ 192.374329] efa_reg_user_mr_dmabuf [ 192.376783] efa_register_mr [ 192.382579] pgsz_bitmap 0xfffff000 rounddown 0x80000000 [ 192.386423] pg_sz [0x80000000] umem_length[0xc0000000] [ 192.392657] start 0x0 length 0xc0000000 params.page_shift 31 params.page_num 3 [ 192.399559] hp_cnt[3], pages_in_hp[524288] [ 192.403690] umem->sgt_append.sgt.nents[1] [ 192.407905] number entries: [1], pg_bit: [31] [ 192.411397] biter->__sg_nents [1] biter->__sg [0000000008b0c5d8] [ 192.415601] biter->__sg_advance [665837568] sg_dma_len[3221225472] [ 192.419823] biter->__sg_nents [1] biter->__sg [0000000008b0c5d8] [ 192.423976] biter->__sg_advance [2813321216] sg_dma_len[3221225472] [ 192.428243] biter->__sg_nents [1] biter->__sg [0000000008b0c5d8] [ 192.432397] biter->__sg_advance [665837568] sg_dma_len[3221225472] Fixes: a808273a495c ("RDMA/verbs: Add a DMA iterator to return aligned contiguous memory blocks") Signed-off-by: Yonatan Nachum Link: https://lore.kernel.org/r/20230109133711.13678-1-ynachum@amazon.com Signed-off-by: Leon Romanovsky

RDMA/core: Clean up a variable name in ib_create_srq_user()

2022-09-22T11:31:05+00:00

"&srq->pd->usecnt" and "&pd->usecnt" are different names for the same reference count. Use "&pd->usecnt" consistently for both the increment and decrement. Signed-off-by: Dan Carpenter Link: https://lore.kernel.org/r/YyxFe3Pm0uzRuBkQ@kili Signed-off-by: Leon Romanovsky

IB: Set IOVA/LENGTH on IB_MR in core/uverbs layers

2022-09-22T09:46:39+00:00

Set 'iova' and 'length' on ib_mr in ib_uverbs and ib_core layers to let all drivers have the members filled. Also, this commit removes redundancy in the respective drivers. Previously, commit 04c0a5fcfcf65 ("IB/uverbs: Set IOVA on IB MR in uverbs layer") changed to set 'iova', but seems to have missed 'length' and the ib_core layer at that time. Fixes: 04c0a5fcfcf65 ("IB/uverbs: Set IOVA on IB MR in uverbs layer") Signed-off-by: Daisuke Matsuda Link: https://lore.kernel.org/r/20220921080844.1616883-1-matsuda-daisuke@fujitsu.com Signed-off-by: Leon Romanovsky

RDMA: Split kernel-only global device caps from uverbs device caps

2022-04-06T18:02:13+00:00

Split out flags from ib_device::device_cap_flags that are only used internally to the kernel into kernel_cap_flags that is not part of the uapi. This limits the device_cap_flags to being the same bitmap that will be copied to userspace. This cleanly splits out the uverbs flags from the kernel flags to avoid confusion in the flags bitmap. Add some short comments describing which each of the kernel flags is connected to. Remove unused kernel flags. Link: https://lore.kernel.org/r/0-v2-22c19e565eef+139a-kern_caps_jgg@nvidia.com Reviewed-by: Christoph Hellwig Reviewed-by: Max Gurtovoy Signed-off-by: Jason Gunthorpe

Revert "RDMA/core: Fix ib_qp_usecnt_dec() called when error"

2022-03-14T23:39:21+00:00

This reverts commit 7c4a539ec38f4ce400a0f3fcb5ff6c940fcd67bb. which causes to the following error in mlx4. Destroy of kernel CQ shouldn't fail WARNING: CPU: 4 PID: 18064 at include/rdma/ib_verbs.h:3936 mlx4_ib_dealloc_xrcd+0x12e/0x1b0 [mlx4_ib] Modules linked in: bonding ib_ipoib ip_gre ipip tunnel4 geneve rdma_ucm nf_tables ib_umad mlx4_en mlx4_ib ib_uverbs mlx4_core ip6_gre gre ip6_tunnel tunnel6 iptable_raw openvswitch nsh rpcrdma ib_iser libiscsi scsi_transport_iscsi rdma_cm iw_cm ib_cm ib_core xt_conntrack xt_MASQUERADE nf_conntrack_netlink nfnetlink xt_addrtype iptable_nat nf_nat br_netfilter overlay fuse [last unloaded: mlx4_core] CPU: 4 PID: 18064 Comm: ibv_xsrq_pingpo Not tainted 5.17.0-rc7_master_62c6ecb #1 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014 RIP: 0010:mlx4_ib_dealloc_xrcd+0x12e/0x1b0 [mlx4_ib] Code: 1e 93 08 00 40 80 fd 01 0f 87 fa f1 04 00 83 e5 01 0f 85 2b ff ff ff 48 c7 c7 20 4f b6 a0 c6 05 fd 92 08 00 01 e8 47 c9 82 e2 <0f> 0b e9 11 ff ff ff 0f b6 2d eb 92 08 00 40 80 fd 01 0f 87 b1 f1 RSP: 0018:ffff8881a4957750 EFLAGS: 00010286 RAX: 0000000000000000 RBX: ffff8881ac4b6800 RCX: 0000000000000000 RDX: 0000000000000027 RSI: 0000000000000004 RDI: ffffed103492aedc RBP: 0000000000000000 R08: 0000000000000001 R09: ffff8884d2e378eb R10: ffffed109a5c6f1d R11: 0000000000000001 R12: ffff888132620000 R13: ffff8881a4957a90 R14: ffff8881aa2d4000 R15: ffff8881a4957ad0 FS: 00007f0401747740(0000) GS:ffff8884d2e00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055f8ae036118 CR3: 000000012fe94005 CR4: 0000000000370ea0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: ib_dealloc_xrcd_user+0xce/0x120 [ib_core] ib_uverbs_dealloc_xrcd+0xad/0x210 [ib_uverbs] uverbs_free_xrcd+0xe8/0x190 [ib_uverbs] destroy_hw_idr_uobject+0x7a/0x130 [ib_uverbs] uverbs_destroy_uobject+0x164/0x730 [ib_uverbs] uobj_destroy+0x72/0xf0 [ib_uverbs] ib_uverbs_cmd_verbs+0x19fb/0x3110 [ib_uverbs] ib_uverbs_ioctl+0x169/0x260 [ib_uverbs] __x64_sys_ioctl+0x856/0x1550 do_syscall_64+0x3d/0x90 entry_SYSCALL_64_after_hwframe+0x44/0xae Fixes: 7c4a539ec38f ("RDMA/core: Fix ib_qp_usecnt_dec() called when error") Link: https://lore.kernel.org/r/74c11029adaf449b3b9228a77cc82f39e9e892c8.1646851220.git.leonro@nvidia.com Signed-off-by: Leon Romanovsky Signed-off-by: Jason Gunthorpe

RDMA/core: Fix ib_qp_usecnt_dec() called when error

2022-03-04T21:30:31+00:00

ib_destroy_qp() would called by ib_create_qp_user() if error, the former contains ib_qp_usecnt_dec(), but ib_qp_usecnt_inc() was not called before. So move ib_qp_usecnt_inc() into create_qp(). Fixes: d2b10794fc13 ("RDMA/core: Create clean QP creations interface for uverbs") Link: https://lore.kernel.org/r/20220303024232.2847388-1-yajun.deng@linux.dev Signed-off-by: Yajun Deng Reviewed-by: Leon Romanovsky Signed-off-by: Jason Gunthorpe

RDMA/core: Remove unnecessary statements

2022-02-28T17:57:24+00:00

The rdma_zalloc_drv_obj() in __ib_alloc_pd() would zero pd, it unnecessary add NULL to the object in struct pd. The uverbs_free_pd() already return busy if pd->usecnt is true, there is no need to add a warning. Link: https://lore.kernel.org/r/20220223074901.201506-1-yajun.deng@linux.dev Signed-off-by: Yajun Deng Reviewed-by: Leon Romanovsky Signed-off-by: Jason Gunthorpe

RDMA/core: Set MR type in ib_reg_user_mr

2022-01-28T15:39:28+00:00

Add missing assignment of MR type to IB_MR_TYPE_USER. Fixes: 33006bd4f37f ("IB/core: Introduce ib_reg_user_mr") Link: https://lore.kernel.org/r/be2e91bcd6e52dc36be289ae92f30d3a5cc6dcb1.1642491047.git.leonro@nvidia.com Signed-off-by: Maor Gottlieb Signed-off-by: Leon Romanovsky Signed-off-by: Jason Gunthorpe

RDMA/core: Set send and receive CQ before forwarding to the driver

2021-11-16T17:16:50+00:00

Preset both receive and send CQ pointers prior to call to the drivers and overwrite it later again till the mlx4 is going to be changed do not overwrite ibqp properties. This change is needed for mlx5, because in case of QP creation failure, it will go to the path of QP destroy which relies on proper CQ pointers. BUG: KASAN: use-after-free in create_qp.cold+0x164/0x16e [mlx5_ib] Write of size 8 at addr ffff8880064c55c0 by task a.out/246 CPU: 0 PID: 246 Comm: a.out Not tainted 5.15.0+ #291 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014 Call Trace: dump_stack_lvl+0x45/0x59 print_address_description.constprop.0+0x1f/0x140 kasan_report.cold+0x83/0xdf create_qp.cold+0x164/0x16e [mlx5_ib] mlx5_ib_create_qp+0x358/0x28a0 [mlx5_ib] create_qp.part.0+0x45b/0x6a0 [ib_core] ib_create_qp_user+0x97/0x150 [ib_core] ib_uverbs_handler_UVERBS_METHOD_QP_CREATE+0x92c/0x1250 [ib_uverbs] ib_uverbs_cmd_verbs+0x1c38/0x3150 [ib_uverbs] ib_uverbs_ioctl+0x169/0x260 [ib_uverbs] __x64_sys_ioctl+0x866/0x14d0 do_syscall_64+0x3d/0x90 entry_SYSCALL_64_after_hwframe+0x44/0xae Allocated by task 246: kasan_save_stack+0x1b/0x40 __kasan_kmalloc+0xa4/0xd0 create_qp.part.0+0x92/0x6a0 [ib_core] ib_create_qp_user+0x97/0x150 [ib_core] ib_uverbs_handler_UVERBS_METHOD_QP_CREATE+0x92c/0x1250 [ib_uverbs] ib_uverbs_cmd_verbs+0x1c38/0x3150 [ib_uverbs] ib_uverbs_ioctl+0x169/0x260 [ib_uverbs] __x64_sys_ioctl+0x866/0x14d0 do_syscall_64+0x3d/0x90 entry_SYSCALL_64_after_hwframe+0x44/0xae Freed by task 246: kasan_save_stack+0x1b/0x40 kasan_set_track+0x1c/0x30 kasan_set_free_info+0x20/0x30 __kasan_slab_free+0x10c/0x150 slab_free_freelist_hook+0xb4/0x1b0 kfree+0xe7/0x2a0 create_qp.part.0+0x52b/0x6a0 [ib_core] ib_create_qp_user+0x97/0x150 [ib_core] ib_uverbs_handler_UVERBS_METHOD_QP_CREATE+0x92c/0x1250 [ib_uverbs] ib_uverbs_cmd_verbs+0x1c38/0x3150 [ib_uverbs] ib_uverbs_ioctl+0x169/0x260 [ib_uverbs] __x64_sys_ioctl+0x866/0x14d0 do_syscall_64+0x3d/0x90 entry_SYSCALL_64_after_hwframe+0x44/0xae Fixes: 514aee660df4 ("RDMA: Globally allocate and release QP memory") Link: https://lore.kernel.org/r/2dbb2e2cbb1efb188a500e5634be1d71956424ce.1636631035.git.leonro@nvidia.com Signed-off-by: Leon Romanovsky Signed-off-by: Jason Gunthorpe

RDMA/core: Fix missed initialization of rdma_hw_stats::lock

2021-10-28T11:58:26+00:00

alloc_and_bind() creates a new rdma_hw_stats structure but misses initializing the mutex lock. This causes debug kernel failures: DEBUG_LOCKS_WARN_ON(lock->magic != lock) WARNING: CPU: 4 PID: 64464 at kernel/locking/mutex.c:575 __mutex_lock+0x9c3/0x12b0 Call Trace: fill_res_counter_entry+0x6ee/0x1020 [ib_core] res_get_common_dumpit+0x907/0x10a0 [ib_core] nldev_stat_get_dumpit+0x20a/0x290 [ib_core] netlink_dump+0x451/0x1040 __netlink_dump_start+0x583/0x830 rdma_nl_rcv_msg+0x3f3/0x7c0 [ib_core] rdma_nl_rcv+0x264/0x410 [ib_core] netlink_unicast+0x433/0x700 netlink_sendmsg+0x707/0xbf0 sock_sendmsg+0xb0/0xe0 __sys_sendto+0x193/0x240 __x64_sys_sendto+0xdd/0x1b0 do_syscall_64+0x3d/0x90 entry_SYSCALL_64_after_hwframe+0x44/0xae Instead of requiring all users to open code initialization of the lock put it in the general rdma_alloc_hw_stats_struct() function and remove duplicates. Fixes: c4ffee7c9bdb ("RDMA/netlink: Implement counter dumpit calback") Link: https://lore.kernel.org/r/4a22986c4685058d2c735d91703ee7d865815bb9.1635237668.git.leonro@nvidia.com Signed-off-by: Mark Zhang Reviewed-by: Maor Gottlieb Signed-off-by: Leon Romanovsky Signed-off-by: Jason Gunthorpe