linux-stable.git/drivers/infiniband, branch linux-6.14.y

RDMA/core: Fix best page size finding when it can cross SG entries

2025-05-29T09:13:16+00:00

[ Upstream commit 486055f5e09df959ad4e3aa4ee75b5c91ddeec2e ]

A single scatter-gather entry is limited by a 32 bits "length" field
that is practically 4GB - PAGE_SIZE. This means that even when the
memory is physically contiguous, we might need more than one entry to
represent it. Additionally when using dmabuf, the sg_table might be
originated outside the subsystem and optimized for other needs.

For instance an SGT of 16GB GPU continuous memory might look like this:
(a real life example)

dma_address 34401400000, length fffff000
dma_address 345013ff000, length fffff000
dma_address 346013fe000, length fffff000
dma_address 347013fd000, length fffff000
dma_address 348013fc000, length 4000

Since ib_umem_find_best_pgsz works within SG entries, in the above case
we will result with the worst possible 4KB page size.

Fix this by taking into consideration only the alignment of addresses of
real discontinuity points rather than treating SG entries as such, and
adjust the page iterator to correctly handle cross SG entry pages.

There is currently an assumption that drivers do not ask for pages
bigger than maximal DMA size supported by their devices.

Reviewed-by: Firas Jahjah 
Reviewed-by: Yonatan Nachum 
Signed-off-by: Michael Margolin 
Link: https://patch.msgid.link/20250217141623.12428-1-mrgolin@amazon.com
Signed-off-by: Leon Romanovsky 
Signed-off-by: Sasha Levin

RDMA/uverbs: Propagate errors from rdma_lookup_get_uobject()

2025-05-29T09:12:44+00:00

[ Upstream commit 81f8f7454ad9e0bf95efdec6542afdc9a6ab1e24 ]

Currently, the IB uverbs API calls uobj_get_uobj_read(), which in turn
uses the rdma_lookup_get_uobject() helper to retrieve user objects.
In case of failure, uobj_get_uobj_read() returns NULL, overriding the
error code from rdma_lookup_get_uobject(). The IB uverbs API then
translates this NULL to -EINVAL, masking the actual error and
complicating debugging. For example, applications calling ibv_modify_qp
that fails with EBUSY when retrieving the QP uobject will see the
overridden error code EINVAL instead, masking the actual error.

Furthermore, based on rdma-core commit:
"2a22f1ced5f3 ("Merge pull request #1568 from jakemoroni/master")"
Kernel's IB uverbs return values are either ignored and passed on as is
to application or overridden with other errnos in a few cases.

Thus, to improve error reporting and debuggability, propagate the
original error from rdma_lookup_get_uobject() instead of replacing it
with EINVAL.

Signed-off-by: Maher Sanalla 
Link: https://patch.msgid.link/64f9d3711b183984e939962c2f83383904f97dfb.1740577869.git.leon@kernel.org
Signed-off-by: Leon Romanovsky 
Signed-off-by: Sasha Levin

RDMA/core: Fix "KASAN: slab-use-after-free Read in ib_register_device" problem

2025-05-22T12:31:42+00:00

[ Upstream commit d0706bfd3ee40923c001c6827b786a309e2a8713 ]

Call Trace:

 __dump_stack lib/dump_stack.c:94 [inline]
 dump_stack_lvl+0x116/0x1f0 lib/dump_stack.c:120
 print_address_description mm/kasan/report.c:408 [inline]
 print_report+0xc3/0x670 mm/kasan/report.c:521
 kasan_report+0xe0/0x110 mm/kasan/report.c:634
 strlen+0x93/0xa0 lib/string.c:420
 __fortify_strlen include/linux/fortify-string.h:268 [inline]
 get_kobj_path_length lib/kobject.c:118 [inline]
 kobject_get_path+0x3f/0x2a0 lib/kobject.c:158
 kobject_uevent_env+0x289/0x1870 lib/kobject_uevent.c:545
 ib_register_device drivers/infiniband/core/device.c:1472 [inline]
 ib_register_device+0x8cf/0xe00 drivers/infiniband/core/device.c:1393
 rxe_register_device+0x275/0x320 drivers/infiniband/sw/rxe/rxe_verbs.c:1552
 rxe_net_add+0x8e/0xe0 drivers/infiniband/sw/rxe/rxe_net.c:550
 rxe_newlink+0x70/0x190 drivers/infiniband/sw/rxe/rxe.c:225
 nldev_newlink+0x3a3/0x680 drivers/infiniband/core/nldev.c:1796
 rdma_nl_rcv_msg+0x387/0x6e0 drivers/infiniband/core/netlink.c:195
 rdma_nl_rcv_skb.constprop.0.isra.0+0x2e5/0x450
 netlink_unicast_kernel net/netlink/af_netlink.c:1313 [inline]
 netlink_unicast+0x53a/0x7f0 net/netlink/af_netlink.c:1339
 netlink_sendmsg+0x8d1/0xdd0 net/netlink/af_netlink.c:1883
 sock_sendmsg_nosec net/socket.c:712 [inline]
 __sock_sendmsg net/socket.c:727 [inline]
 ____sys_sendmsg+0xa95/0xc70 net/socket.c:2566
 ___sys_sendmsg+0x134/0x1d0 net/socket.c:2620
 __sys_sendmsg+0x16d/0x220 net/socket.c:2652
 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
 do_syscall_64+0xcd/0x260 arch/x86/entry/syscall_64.c:94
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

This problem is similar to the problem that the
commit 1d6a9e7449e2 ("RDMA/core: Fix use-after-free when rename device name")
fixes.

The root cause is: the function ib_device_rename() renames the name with
lock. But in the function kobject_uevent(), this name is accessed without
lock protection at the same time.

The solution is to add the lock protection when this name is accessed in
the function kobject_uevent().

Fixes: 779e0bf47632 ("RDMA/core: Do not indicate device ready when device enablement fails")
Link: https://patch.msgid.link/r/20250506151008.75701-1-yanjun.zhu@linux.dev
Reported-by: syzbot+e2ce9e275ecc70a30b72@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=e2ce9e275ecc70a30b72
Signed-off-by: Zhu Yanjun 
Signed-off-by: Jason Gunthorpe 
Signed-off-by: Sasha Levin

RDMA/rxe: Fix slab-use-after-free Read in rxe_queue_cleanup bug

2025-05-22T12:31:41+00:00

[ Upstream commit f81b33582f9339d2dc17c69b92040d3650bb4bae ]

Call Trace:
 
 __dump_stack lib/dump_stack.c:94 [inline]
 dump_stack_lvl+0x7d/0xa0 lib/dump_stack.c:120
 print_address_description mm/kasan/report.c:378 [inline]
 print_report+0xcf/0x610 mm/kasan/report.c:489
 kasan_report+0xb5/0xe0 mm/kasan/report.c:602
 rxe_queue_cleanup+0xd0/0xe0 drivers/infiniband/sw/rxe/rxe_queue.c:195
 rxe_cq_cleanup+0x3f/0x50 drivers/infiniband/sw/rxe/rxe_cq.c:132
 __rxe_cleanup+0x168/0x300 drivers/infiniband/sw/rxe/rxe_pool.c:232
 rxe_create_cq+0x22e/0x3a0 drivers/infiniband/sw/rxe/rxe_verbs.c:1109
 create_cq+0x658/0xb90 drivers/infiniband/core/uverbs_cmd.c:1052
 ib_uverbs_create_cq+0xc7/0x120 drivers/infiniband/core/uverbs_cmd.c:1095
 ib_uverbs_write+0x969/0xc90 drivers/infiniband/core/uverbs_main.c:679
 vfs_write fs/read_write.c:677 [inline]
 vfs_write+0x26a/0xcc0 fs/read_write.c:659
 ksys_write+0x1b8/0x200 fs/read_write.c:731
 do_syscall_x64 arch/x86/entry/common.c:52 [inline]
 do_syscall_64+0xaa/0x1b0 arch/x86/entry/common.c:83
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

In the function rxe_create_cq, when rxe_cq_from_init fails, the function
rxe_cleanup will be called to handle the allocated resources. In fact,
some memory resources have already been freed in the function
rxe_cq_from_init. Thus, this problem will occur.

The solution is to let rxe_cleanup do all the work.

Fixes: 8700e3e7c485 ("Soft RoCE driver")
Link: https://paste.ubuntu.com/p/tJgC42wDf6/
Tested-by: liuyi 
Signed-off-by: Zhu Yanjun 
Link: https://patch.msgid.link/20250412075714.3257358-1-yanjun.zhu@linux.dev
Reviewed-by: Daisuke Matsuda 
Signed-off-by: Leon Romanovsky 
Signed-off-by: Sasha Levin

qibfs: fix _another_ leak

2025-05-02T06:02:03+00:00

[ Upstream commit bdb43af4fdb39f844ede401bdb1258f67a580a27 ]

failure to allocate inode => leaked dentry...

this one had been there since the initial merge; to be fair,
if we are that far OOM, the odds of failing at that particular
allocation are low...

Signed-off-by: Al Viro 
Signed-off-by: Sasha Levin

RDMA/cma: Fix workqueue crash in cma_netevent_work_handler

2025-04-25T08:51:09+00:00

commit 45f5dcdd049719fb999393b30679605f16ebce14 upstream.

struct rdma_cm_id has member "struct work_struct net_work"
that is reused for enqueuing cma_netevent_work_handler()s
onto cma_wq.

Below crash[1] can occur if more than one call to
cma_netevent_callback() occurs in quick succession,
which further enqueues cma_netevent_work_handler()s for the
same rdma_cm_id, overwriting any previously queued work-item(s)
that was just scheduled to run i.e. there is no guarantee
the queued work item may run between two successive calls
to cma_netevent_callback() and the 2nd INIT_WORK would overwrite
the 1st work item (for the same rdma_cm_id), despite grabbing
id_table_lock during enqueue.

Also drgn analysis [2] indicates the work item was likely overwritten.

Fix this by moving the INIT_WORK() to __rdma_create_id(),
so that it doesn't race with any existing queue_work() or
its worker thread.

[1] Trimmed crash stack:
=============================================
BUG: kernel NULL pointer dereference, address: 0000000000000008
kworker/u256:6 ... 6.12.0-0...
Workqueue:  cma_netevent_work_handler [rdma_cm] (rdma_cm)
RIP: 0010:process_one_work+0xba/0x31a
Call Trace:
 worker_thread+0x266/0x3a0
 kthread+0xcf/0x100
 ret_from_fork+0x31/0x50
 ret_from_fork_asm+0x1a/0x30
=============================================

[2] drgn crash analysis:

>>> trace = prog.crashed_thread().stack_trace()
>>> trace
(0)  crash_setup_regs (./arch/x86/include/asm/kexec.h:111:15)
(1)  __crash_kexec (kernel/crash_core.c:122:4)
(2)  panic (kernel/panic.c:399:3)
(3)  oops_end (arch/x86/kernel/dumpstack.c:382:3)
...
(8)  process_one_work (kernel/workqueue.c:3168:2)
(9)  process_scheduled_works (kernel/workqueue.c:3310:3)
(10) worker_thread (kernel/workqueue.c:3391:4)
(11) kthread (kernel/kthread.c:389:9)

Line workqueue.c:3168 for this kernel version is in process_one_work():
3168	strscpy(worker->desc, pwq->wq->name, WORKER_DESC_LEN);

>>> trace[8]["work"]
*(struct work_struct *)0xffff92577d0a21d8 = {
	.data = (atomic_long_t){
		.counter = (s64)536870912,    <=== Note
	},
	.entry = (struct list_head){
		.next = (struct list_head *)0xffff924d075924c0,
		.prev = (struct list_head *)0xffff924d075924c0,
	},
	.func = (work_func_t)cma_netevent_work_handler+0x0 = 0xffffffffc2cec280,
}

Suspicion is that pwq is NULL:
>>> trace[8]["pwq"]
(struct pool_workqueue *)

In process_one_work(), pwq is assigned from:
struct pool_workqueue *pwq = get_work_pwq(work);

and get_work_pwq() is:
static struct pool_workqueue *get_work_pwq(struct work_struct *work)
{
 	unsigned long data = atomic_long_read(&work->data);

 	if (data & WORK_STRUCT_PWQ)
 		return work_struct_pwq(data);
 	else
 		return NULL;
}

WORK_STRUCT_PWQ is 0x4:
>>> print(repr(prog['WORK_STRUCT_PWQ']))
Object(prog, 'enum work_flags', value=4)

But work->data is 536870912 which is 0x20000000.
So, get_work_pwq() returns NULL and we crash in process_one_work():
3168	strscpy(worker->desc, pwq->wq->name, WORKER_DESC_LEN);
=============================================

Fixes: 925d046e7e52 ("RDMA/core: Add a netevent notifier to cma")
Cc: stable@vger.kernel.org
Co-developed-by: Håkon Bugge 
Signed-off-by: Håkon Bugge 
Signed-off-by: Sharath Srinivasan 
Reviewed-by: Patrisious Haddad 
Link: https://patch.msgid.link/bf0082f9-5b25-4593-92c6-d130aa8ba439@oracle.com
Signed-off-by: Leon Romanovsky 
Signed-off-by: Greg Kroah-Hartman

RDMA/bnxt_re: Remove unusable nq variable

2025-04-25T08:50:49+00:00

[ Upstream commit ffc59e32c67e599cc473d6427a4aa584399d5b3c ]

Remove nq variable from bnxt_re_create_srq() and bnxt_re_destroy_srq()
as it generates the following compilation warnings:

>> drivers/infiniband/hw/bnxt_re/ib_verbs.c:1777:24: warning: variable
'nq' set but not used [-Wunused-but-set-variable]
    1777 |         struct bnxt_qplib_nq *nq = NULL;
         |                               ^
   drivers/infiniband/hw/bnxt_re/ib_verbs.c:1828:24: warning: variable
'nq' set but not used [-Wunused-but-set-variable]
    1828 |         struct bnxt_qplib_nq *nq = NULL;
         |                               ^
   2 warnings generated.

Fixes: 6b395d31146a ("RDMA/bnxt_re: Fix budget handling of notification queue")
Link: https://patch.msgid.link/r/8a4343e217d7d1c0a5a786b785c4ac57cb72a2a0.1744288299.git.leonro@nvidia.com
Reported-by: kernel test robot 
Closes: https://lore.kernel.org/oe-kbuild-all/202504091055.CzgXnk4C-lkp@intel.com/
Signed-off-by: Leon Romanovsky 
Signed-off-by: Jason Gunthorpe 
Signed-off-by: Sasha Levin

RDMA/core: Silence oversized kvmalloc() warning

2025-04-25T08:50:49+00:00

[ Upstream commit 9a0e6f15029e1a8a21e40f06fd05aa52b7f063de ]

syzkaller triggered an oversized kvmalloc() warning.
Silence it by adding __GFP_NOWARN.

syzkaller log:
 WARNING: CPU: 7 PID: 518 at mm/util.c:665 __kvmalloc_node_noprof+0x175/0x180
 CPU: 7 UID: 0 PID: 518 Comm: c_repro Not tainted 6.11.0-rc6+ #6
 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014
 RIP: 0010:__kvmalloc_node_noprof+0x175/0x180
 RSP: 0018:ffffc90001e67c10 EFLAGS: 00010246
 RAX: 0000000000000100 RBX: 0000000000000400 RCX: ffffffff8149d46b
 RDX: 0000000000000000 RSI: ffff8881030fae80 RDI: 0000000000000002
 RBP: 000000712c800000 R08: 0000000000000100 R09: 0000000000000000
 R10: ffffc90001e67c10 R11: 0030ae0601000000 R12: 0000000000000000
 R13: 0000000000000000 R14: 00000000ffffffff R15: 0000000000000000
 FS:  00007fde79159740(0000) GS:ffff88813bdc0000(0000) knlGS:0000000000000000
 CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
 CR2: 0000000020000180 CR3: 0000000105eb4005 CR4: 00000000003706b0
 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
 Call Trace:
  
  ib_umem_odp_get+0x1f6/0x390
  mlx5_ib_reg_user_mr+0x1e8/0x450
  ib_uverbs_reg_mr+0x28b/0x440
  ib_uverbs_write+0x7d3/0xa30
  vfs_write+0x1ac/0x6c0
  ksys_write+0x134/0x170
  ? __sanitizer_cov_trace_pc+0x1c/0x50
  do_syscall_64+0x50/0x110
  entry_SYSCALL_64_after_hwframe+0x76/0x7e

Fixes: 37824952dc8f ("RDMA/odp: Use kvcalloc for the dma_list and page_list")
Signed-off-by: Shay Drory 
Link: https://patch.msgid.link/c6cb92379de668be94894f49c2cfa40e73f94d56.1742388096.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky 
Signed-off-by: Sasha Levin

RDMA/hns: Fix wrong maximum DMA segment size

2025-04-25T08:50:48+00:00

[ Upstream commit 9beb2c91fb86e0be70a5833c6730441fa3c9efa8 ]

Set maximum DMA segment size to 2G instead of UINT_MAX due to HW limit.

Fixes: e0477b34d9d1 ("RDMA: Explicitly pass in the dma_device to ib_register_device")
Link: https://patch.msgid.link/r/20250327114724.3454268-3-huangjunxian6@hisilicon.com
Signed-off-by: Chengchang Tang 
Signed-off-by: Junxian Huang 
Signed-off-by: Jason Gunthorpe 
Signed-off-by: Sasha Levin

RDMA/usnic: Fix passing zero to PTR_ERR in usnic_ib_pci_probe()

2025-04-25T08:50:48+00:00

[ Upstream commit 95ba3850fed03e01b422ab5d7943aeba130c9723 ]

drivers/infiniband/hw/usnic/usnic_ib_main.c:590
 usnic_ib_pci_probe() warn: passing zero to 'PTR_ERR'

Make usnic_ib_device_add() return NULL on fail path, also remove
useless NULL check for usnic_ib_discover_pf()

Fixes: e3cf00d0a87f ("IB/usnic: Add Cisco VIC low-level hardware driver")
Link: https://patch.msgid.link/r/20250324123132.2392077-1-yuehaibing@huawei.com
Signed-off-by: Yue Haibing 
Reviewed-by: Zhu Yanjun 
Reviewed-by: Jason Gunthorpe 
Signed-off-by: Jason Gunthorpe 
Signed-off-by: Sasha Levin