From: Kaike Wan kaike.wan@intel.com

When a TID RDMA ACK to RESYNC request is received, the flow PSNs for pending TID RDMA WRITE segments will be adjusted with the next flow generation number, based on the resync_psn value extracted from the flow PSN of the TID RDMA ACK packet. The resync_psn value indicates the last flow PSN for which a TID RDMA WRITE DATA packet has been received by the responder and the requester should resend TID RDMA WRITE DATA packets, starting from the next flow PSN. However, if resync_psn points to the last flow PSN for a segment and the next segment flow PSN starts with a new generation number, use of the old resync_psn to adjust the flow PSN for the next segment will lead to miscalculation, resulting in WARN_ON and sge rewinding errors: [2419460.492485] WARNING: CPU: 4 PID: 146961 at /nfs/site/home/phcvs2/gitrepo/ifs-all/components/Drivers/tmp/rpmbuild/BUILD/ifs-kernel-updates-3.10.0_957.el7.x86_64/hfi1/tid_rdma.c:4764 hfi1_rc_rcv_tid_rdma_ack+0x8f6/0xa90 [hfi1] [2419460.514565] Modules linked in: ib_ipoib(OE) hfi1(OE) rdmavt(OE) rpcsec_gss_krb5 auth_rpcgss nfsv4 dns_resolver nfsv3 nfs_acl nfs lockd grace fscache iTCO_wdt iTCO_vendor_support skx_edac intel_powerclamp coretemp intel_rapl iosf_mbi kvm irqbypass crc32_pclmul ghash_clmulni_intel ib_isert iscsi_target_mod target_core_mod aesni_intel lrw gf128mul glue_helper ablk_helper cryptd rpcrdma sunrpc opa_vnic ast ttm ib_iser libiscsi drm_kms_helper scsi_transport_iscsi ipmi_ssif syscopyarea sysfillrect sysimgblt fb_sys_fops drm joydev ipmi_si pcspkr sg drm_panel_orientation_quirks ipmi_devintf lpc_ich i2c_i801 ipmi_msghandler wmi rdma_ucm ib_ucm ib_uverbs acpi_cpufreq acpi_power_meter ib_umad rdma_cm ib_cm iw_cm ip_tables ext4 mbcache jbd2 sd_mod crc_t10dif crct10dif_generic crct10dif_pclmul i2c_algo_bit crct10dif_common [2419460.594432] crc32c_intel e1000e ib_core ahci libahci ptp libata pps_core nfit libnvdimm [last unloaded: rdmavt] [2419460.605645] CPU: 4 PID: 146961 Comm: kworker/4:0H Kdump: loaded Tainted: G W OE ------------ 3.10.0-957.el7.x86_64 #1 [2419460.619424] Hardware name: Intel Corporation S2600WFT/S2600WFT, BIOS SE5C620.86B.0X.02.0117.040420182310 04/04/2018 [2419460.631062] Workqueue: hfi0_0 _hfi1_do_tid_send [hfi1] [2419460.637423] Call Trace: [2419460.641044] <IRQ> [<ffffffff9e361dc1>] dump_stack+0x19/0x1b [2419460.647980] [<ffffffff9dc97648>] __warn+0xd8/0x100 [2419460.654023] [<ffffffff9dc9778d>] warn_slowpath_null+0x1d/0x20 [2419460.661025] [<ffffffffc05d28c6>] hfi1_rc_rcv_tid_rdma_ack+0x8f6/0xa90 [hfi1] [2419460.669333] [<ffffffffc05c21cc>] hfi1_kdeth_eager_rcv+0x1dc/0x210 [hfi1] [2419460.677295] [<ffffffffc05c23ef>] ? hfi1_kdeth_expected_rcv+0x1ef/0x210 [hfi1] [2419460.685693] [<ffffffffc0574f15>] kdeth_process_eager+0x35/0x90 [hfi1] [2419460.693394] [<ffffffffc0575b5a>] handle_receive_interrupt_nodma_rtail+0x17a/0x2b0 [hfi1] [2419460.702745] [<ffffffffc056a623>] receive_context_interrupt+0x23/0x40 [hfi1] [2419460.710963] [<ffffffff9dd4a294>] __handle_irq_event_percpu+0x44/0x1c0 [2419460.718659] [<ffffffff9dd4a442>] handle_irq_event_percpu+0x32/0x80 [2419460.726086] [<ffffffff9dd4a4cc>] handle_irq_event+0x3c/0x60 [2419460.732903] [<ffffffff9dd4d27f>] handle_edge_irq+0x7f/0x150 [2419460.739710] [<ffffffff9dc2e554>] handle_irq+0xe4/0x1a0 [2419460.746091] [<ffffffff9e3795dd>] do_IRQ+0x4d/0xf0 [2419460.752040] [<ffffffff9e36b362>] common_interrupt+0x162/0x162 [2419460.759029] <EOI> [<ffffffff9dfa0f79>] ? swiotlb_map_page+0x49/0x150 [2419460.766758] [<ffffffffc05c2ed1>] hfi1_verbs_send_dma+0x291/0xb70 [hfi1] [2419460.774637] [<ffffffffc05c2c40>] ? hfi1_wait_kmem+0xf0/0xf0 [hfi1] [2419460.782080] [<ffffffffc05c3f26>] hfi1_verbs_send+0x126/0x2b0 [hfi1] [2419460.789606] [<ffffffffc05ce683>] _hfi1_do_tid_send+0x1d3/0x320 [hfi1] [2419460.797298] [<ffffffff9dcb9d4f>] process_one_work+0x17f/0x440 [2419460.804292] [<ffffffff9dcbade6>] worker_thread+0x126/0x3c0 [2419460.811025] [<ffffffff9dcbacc0>] ? manage_workers.isra.25+0x2a0/0x2a0 [2419460.818710] [<ffffffff9dcc1c31>] kthread+0xd1/0xe0 [2419460.824751] [<ffffffff9dcc1b60>] ? insert_kthread_work+0x40/0x40 [2419460.832013] [<ffffffff9e374c1d>] ret_from_fork_nospec_begin+0x7/0x21 [2419460.839611] [<ffffffff9dcc1b60>] ? insert_kthread_work+0x40/0x40

This patch fixes the issue by adjusting the resync_psn first if the flow generation has been advanced for a pending segment.

Fixes: 9e93e967f7b4 ("IB/hfi1: Add a function to receive TID RDMA ACK packet") Cc: stable@vger.kernel.org Reviewed-by: Mike Marciniszyn mike.marciniszyn@intel.com Signed-off-by: Kaike Wan kaike.wan@intel.com Signed-off-by: Dennis Dalessandro dennis.dalessandro@intel.com --- drivers/infiniband/hw/hfi1/tid_rdma.c | 9 +++++++++ 1 file changed, 9 insertions(+)

diff --git a/drivers/infiniband/hw/hfi1/tid_rdma.c b/drivers/infiniband/hw/hfi1/tid_rdma.c index e53f542..8a2e0d9 100644 --- a/drivers/infiniband/hw/hfi1/tid_rdma.c +++ b/drivers/infiniband/hw/hfi1/tid_rdma.c @@ -4633,6 +4633,15 @@ void hfi1_rc_rcv_tid_rdma_ack(struct hfi1_packet *packet) */ fpsn = full_flow_psn(flow, flow->flow_state.spsn); req->r_ack_psn = psn; + /* + * If resync_psn points to the last flow PSN for a + * segment and the new segment (likely from a new + * request) starts with a new generation number, we + * need to adjust resync_psn accordingly. + */ + if (flow->flow_state.generation != + (resync_psn >> HFI1_KDETH_BTH_SEQ_SHIFT)) + resync_psn = mask_psn(fpsn - 1); flow->resync_npkts += delta_psn(mask_psn(resync_psn + 1), fpsn); /*