Re: [PATCH] RDMA/irdma: Avoid free the non-cqp_request scratch

From: Shifeng Li
Date: Fri Nov 24 2023 - 09:29:44 EST


On 2023/11/22 1:25, Ismail, Mustafa wrote:
-----Original Message-----
From: Shifeng Li <lishifeng1992@xxxxxxx>
Sent: Monday, November 20, 2023 2:31 AM
To: Ismail, Mustafa <mustafa.ismail@xxxxxxxxx>
Cc: Saleem, Shiraz <shiraz.saleem@xxxxxxxxx>; jgg@xxxxxxxx;
leon@xxxxxxxxxx; linux-rdma@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx;
Ding, Hui <dinghui@xxxxxxxxxxxxxx>; Shifeng Li <lishifeng1992@xxxxxxx>
Subject: [PATCH] RDMA/irdma: Avoid free the non-cqp_request scratch

When creating ceq_0 during probing irdma, cqp.sc_cqp will be sent as a
cqp_request to cqp->sc_cqp.sq_ring. If the request is pending when removing
the irdma driver or unplugging its aux device, cqp.sc_cqp will be dereferenced
as wrong struct in irdma_free_pending_cqp_request().

crash> bt 3669
PID: 3669 TASK: ffff88aef892c000 CPU: 28 COMMAND: "kworker/28:0"
#0 [fffffe0000549e38] crash_nmi_callback at ffffffff810e3a34
#1 [fffffe0000549e40] nmi_handle at ffffffff810788b2
#2 [fffffe0000549ea0] default_do_nmi at ffffffff8107938f
#3 [fffffe0000549eb8] do_nmi at ffffffff81079582
#4 [fffffe0000549ef0] end_repeat_nmi at ffffffff82e016b4
[exception RIP: native_queued_spin_lock_slowpath+1291]
RIP: ffffffff8127e72b RSP: ffff88aa841ef778 RFLAGS: 00000046
RAX: 0000000000000000 RBX: ffff88b01f849700 RCX: ffffffff8127e47e
RDX: 0000000000000000 RSI: 0000000000000004 RDI: ffffffff83857ec0
RBP: ffff88afe3e4efc8 R8: ffffed15fc7c9dfa R9: ffffed15fc7c9dfa
R10: 0000000000000001 R11: ffffed15fc7c9df9 R12: 0000000000740000
R13: ffff88b01f849708 R14: 0000000000000003 R15: ffffed1603f092e1
ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0000
--- <NMI exception stack> ---
#5 [ffff88aa841ef778] native_queued_spin_lock_slowpath at ffffffff8127e72b
#6 [ffff88aa841ef7b0] _raw_spin_lock_irqsave at ffffffff82c22aa4
#7 [ffff88aa841ef7c8] __wake_up_common_lock at ffffffff81257363
#8 [ffff88aa841ef888] irdma_free_pending_cqp_request at ffffffffa0ba12cc
[irdma]
#9 [ffff88aa841ef958] irdma_cleanup_pending_cqp_op at ffffffffa0ba1469
[irdma]
#10 [ffff88aa841ef9c0] irdma_ctrl_deinit_hw at ffffffffa0b2989f [irdma]
#11 [ffff88aa841efa28] irdma_remove at ffffffffa0b252df [irdma]
#12 [ffff88aa841efae8] auxiliary_bus_remove at ffffffff8219afdb
#13 [ffff88aa841efb00] device_release_driver_internal at ffffffff821882e6
#14 [ffff88aa841efb38] bus_remove_device at ffffffff82184278
#15 [ffff88aa841efb88] device_del at ffffffff82179d23
#16 [ffff88aa841efc48] ice_unplug_aux_dev at ffffffffa0eb1c14 [ice]
#17 [ffff88aa841efc68] ice_service_task at ffffffffa0d88201 [ice]
#18 [ffff88aa841efde8] process_one_work at ffffffff811c589a
#19 [ffff88aa841efe60] worker_thread at ffffffff811c71ff
#20 [ffff88aa841eff10] kthread at ffffffff811d87a0
#21 [ffff88aa841eff50] ret_from_fork at ffffffff82e0022f

Fixes: 44d9e52977a1 ("RDMA/irdma: Implement device initialization
definitions")
Signed-off-by: Shifeng Li <lishifeng1992@xxxxxxx>
---
drivers/infiniband/hw/irdma/utils.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/infiniband/hw/irdma/utils.c
b/drivers/infiniband/hw/irdma/utils.c
index 445e69e86409..222ec1f761a1 100644
--- a/drivers/infiniband/hw/irdma/utils.c
+++ b/drivers/infiniband/hw/irdma/utils.c
@@ -541,7 +541,7 @@ void irdma_cleanup_pending_cqp_op(struct
irdma_pci_f *rf)
for (i = 0; i < pending_work; i++) {
cqp_request = (struct irdma_cqp_request *)(unsigned long)
cqp->scratch_array[wqe_idx];
- if (cqp_request)
+ if (cqp_request && cqp_request != (struct irdma_cqp_request
+*)&cqp->sc_cqp)
irdma_free_pending_cqp_request(cqp, cqp_request);
wqe_idx = (wqe_idx + 1) % IRDMA_RING_SIZE(cqp-
sc_cqp.sq_ring);
}
--
2.25.1

Hi Li,

Could you describe how you hit this issue? It seems like the probe would not have completed successfully if the create cceq request was still pending.

A better fix might be to set the scratch to 0 in this case: In irdma_create_ceq(), pass 0 for scratch in call to irdma_sc_cceq_create(&iwceq->sc_ceq, 0), since it doesn't appear to be used in this case. Is it possible to test this fix? Thanks!

1. The firmware of net card or the hardware may enter into some kind of error state. And I observed
that the sq_ring of sc_cqp is full through crash. This means that all the 2048 cqp_requests were pending.

crash> struct irdma_cqp.sc_cqp ffff88afe3e4eee8
sc_cqp = {
...
sq_ring = {
head = 91, (The queue head caught up with the queue tail.)
tail = 92,
size = 2048
},
...
}

2. The issue is not reproducible. But I have tested the solution that pass 0 for scratch in
call to irdma_sc_cceq_create(&iwceq->sc_ceq, 0) in irdma_create_ceq(). It works well.

Thanks!