Description of problem:

All mvapich2 benchmarks fail with the following error when run in CXBG4 IW device.

[src/mpid/ch3/channels/mrail/src/gen2/rdma_iba_priv.c:2118] Could not modify qpto RTR
[src/mpid/ch3/channels/mrail/src/gen2/rdma_iba_priv.c:2118] Could not modify qpto RTR

Version-Release number of selected component (if applicable):

Clients: rdma-dev-13
Servers: rdma-dev-12

DISTRO=RHEL-8.7.0-20220609.0

+ [22-06-10 09:37:35] cat /etc/redhat-release
Red Hat Enterprise Linux release 8.7 Beta (Ootpa)

+ [22-06-10 09:37:35] uname -a
Linux rdma-dev-13.rdma.lab.eng.rdu2.redhat.com 4.18.0-398.el8.x86_64 #1 SMP Fri Jun 3 16:53:45 EDT 2022 x86_64 x86_64 x86_64 GNU/Linux

+ [22-06-10 09:37:35] cat /proc/cmdline
BOOT_IMAGE=(hd0,msdos1)/vmlinuz-4.18.0-398.el8.x86_64 root=/dev/mapper/rhel_rdma-dev13-root ro intel_idle.max_cstate=0 intremap=no_x2apic_optout processor.max_cstate=0 console=tty0 rd_NO_PLYMOUTH crashkernel=auto resume=/dev/mapper/rhel_rdmadev-13-swap rd.lvm.lv=rhel_rdma-dev-13/root rd.lvm.lv=rhel_rdma-dev-13/swap console=ttyS1,115200n81

+ [22-06-10 09:37:35] rpm -q rdma-core linux-firmware
rdma-core-37.2-1.el8.x86_64
linux-firmware-20220210-107.git6342082c.el8.noarch

+ [22-06-10 09:37:35] tail /sys/class/infiniband/cxgb4_0/fw_ver
1.26.6.0
+ [22-06-10 09:37:35] lspci
+ [22-06-10 09:37:35] grep -i -e ethernet -e infiniband -e omni -e ConnectX
03:00.0 Ethernet controller: Broadcom Inc. and subsidiaries NetXtreme BCM5717 Gigabit Ethernet PCIe (rev 10)
03:00.1 Ethernet controller: Broadcom Inc. and subsidiaries NetXtreme BCM5717 Gigabit Ethernet PCIe (rev 10)
05:00.0 Ethernet controller: Chelsio Communications Inc T62100-LP-CR Unified Wire Ethernet Controller
05:00.1 Ethernet controller: Chelsio Communications Inc T62100-LP-CR Unified Wire Ethernet Controller
05:00.2 Ethernet controller: Chelsio Communications Inc T62100-LP-CR Unified Wire Ethernet Controller
05:00.3 Ethernet controller: Chelsio Communications Inc T62100-LP-CR Unified Wire Ethernet Controller
05:00.4 Ethernet controller: Chelsio Communications Inc T62100-LP-CR Unified Wire Ethernet Controller

Installed:
mpitests-mvapich2-5.8-1.el8.x86_64 mvapich2-2.3.6-1.el8.x86_64

How reproducible:

100%

Steps to Reproduce:
1. bring up the RDMA hosts mentioned above with RHEL9.1 build
2. set up RDMA hosts for mvapich2 benchamrk tests
3. run one of the mvapich2 benchmark with "mpirun" command, as the following:

timeout --preserve-status --kill-after=5m 3m mpirun -hostfile /root/hfile_one_core -np 2 mpitests-IMB-MPI1 PingPong -time 1.5

Actual results:

[rdma-dev-12.rdma.lab.eng.rdu2.redhat.com:mpi_rank_0][rdma_param_handle_heterogeneity] All nodes involved in the job were detected to be homogeneous in terms of processors and interconnects. Setting MV2_HOMOGENEOUS_CLUSTER=1 can improve job startup performance on such systems. The following link has more details on enhancing job startup performance. http://mvapich.cse.ohio-state.edu/performance/job-startup/.
[rdma-dev-12.rdma.lab.eng.rdu2.redhat.com:mpi_rank_0][rdma_param_handle_heterogeneity] To suppress this warning, please set MV2_SUPPRESS_JOB_STARTUP_PERFORMANCE_WARNING to 1

[src/mpid/ch3/channels/mrail/src/gen2/rdma_iba_priv.c:2118] Could not modify qpto RTR
[src/mpid/ch3/channels/mrail/src/gen2/rdma_iba_priv.c:2118] Could not modify qpto RTR <<<=============================
#----------------------------------------------------------------

1. Intel(R) MPI Benchmarks 2021.3, MPI-1 part
   #----------------------------------------------------------------
2. Date : Fri Jun 10 09:37:50 2022
3. Machine : x86_64
4. System : Linux
5. Release : 4.18.0-398.el8.x86_64
6. Version : #1 SMP Fri Jun 3 16:53:45 EDT 2022
7. MPI Version : 3.1
8. MPI Thread Environment:

1. Calling sequence was:

1. mpitests-IMB-MPI1 PingPong -time 1.5

1. Minimum message length in bytes: 0
2. Maximum message length in bytes: 4194304
   #
3. MPI_Datatype : MPI_BYTE
4. MPI_Datatype for reductions : MPI_FLOAT
5. MPI_Op : MPI_SUM

1. List of Benchmarks to run:

1. PingPong

Expected results:

Normal execution of the benchmarks with complete stats output

Additional info: