= Verification = $ cat /proc/version Linux version 6.5.0-27-generic (buildd@lcy02-amd64-059) (x86_64-linux-gnu-gcc-13 (Ubuntu 13.2.0-4ubuntu3) 13.2.0, GNU ld (GNU Binutils for Ubuntu) 2.41) #28-Ubuntu SMP PREEMPT_DYNAMIC Thu Mar 7 18:21:00 UTC 2024
ubuntu@ubuntu:~/autotest-client-tests/ubuntu_performance_gpudirect_rdma/nvidia-peermem-test$ ./nvidia-peermem-test.sh -m peermem Repository: 'Types: deb URIs: https://ppa.launchpadcontent.net/canonical-nvidia/perftest+cuda/ubuntu/ Suites: mantic Components: main ' Description: Used internal for kernel regression testing More info: https://launchpad.net/~canonical-nvidia/+archive/ubuntu/perftest+cuda Adding repository. Found existing deb entry in /etc/apt/sources.list.d/canonical-nvidia-ubuntu-perftest_cuda-mantic.sources Hit:1 http://archive.ubuntu.com/ubuntu mantic InRelease Hit:2 http://archive.ubuntu.com/ubuntu mantic-updates InRelease Hit:3 http://archive.ubuntu.com/ubuntu mantic-security InRelease Hit:4 http://archive.ubuntu.com/ubuntu mantic-backports InRelease Hit:5 http://archive.ubuntu.com/ubuntu mantic-proposed InRelease Hit:6 https://ppa.launchpadcontent.net/canonical-nvidia/perftest+cuda/ubuntu mantic InRelease Hit:7 https://ppa.launchpadcontent.net/dannf/dannf/ubuntu mantic InRelease Reading package lists... Done Reading package lists... Done Building dependency tree... Done Reading state information... Done perftest is already the newest version (24.01.0+0.38-1+perftest+cuda.1~ubuntu23.10.1). 0 upgraded, 0 newly installed, 0 to remove and 10 not upgraded. Reading package lists... Done Building dependency tree... Done Reading state information... Done opensm is already the newest version (3.3.23-2). 0 upgraded, 0 newly installed, 0 to remove and 10 not upgraded. --use_cuda=<cuda device id> Use CUDA specific device for GPUDirect RDMA testing Perftest doesn't supports CUDA tests with inline messages: inline size set to 0 ************************************ * Waiting for client to connect... * ************************************ Perftest doesn't supports CUDA tests with inline messages: inline size set to 0 initializing CUDA initializing CUDA Listing all CUDA devices in system: CUDA device 0: PCIe address is 07:00 CUDA device 1: PCIe address is 0F:00 CUDA device 2: PCIe address is 47:00 CUDA device 3: PCIe address is 4E:00 CUDA device 4: PCIe address is 87:00 CUDA device 5: PCIe address is 90:00 CUDA device 6: PCIe address is B7:00 CUDA device 7: PCIe address is BD:00 Picking device No. 1 [pid = 15582, dev = 1] device name = [NVIDIA A100-SXM4-40GB] creating CUDA Ctx Listing all CUDA devices in system: CUDA device 0: PCIe address is 07:00 CUDA device 1: PCIe address is 0F:00 CUDA device 2: PCIe address is 47:00 CUDA device 3: PCIe address is 4E:00 CUDA device 4: PCIe address is 87:00 CUDA device 5: PCIe address is 90:00 CUDA device 6: PCIe address is B7:00 CUDA device 7: PCIe address is BD:00 Picking device No. 0 [pid = 15576, dev = 0] device name = [NVIDIA A100-SXM4-40GB] creating CUDA Ctx making it the current CUDA Ctx cuMemAlloc() of a 16777216 bytes GPU buffer allocated GPU buffer address at 00007c0146000000 pointer=0x7c0146000000 --------------------------------------------------------------------------------------- RDMA_Write BW Test Dual-port : OFF Device : mlx5_6 Number of qps : 1 Transport type : IB Connection type : RC Using SRQ : OFF PCIe relax order: ON making it the current CUDA Ctx cuMemAlloc() of a 16777216 bytes GPU buffer allocated GPU buffer address at 00007a08b4000000 pointer=0x7a08b4000000 --------------------------------------------------------------------------------------- RDMA_Write BW Test Dual-port : OFF Device : mlx5_2 Number of qps : 1 Transport type : IB Connection type : RC Using SRQ : OFF PCIe relax order: ON ibv_wr* API : ON TX depth : 128 CQ Moderation : 100 Mtu : 4096[B] Link type : IB Max inline data : 0[B] rdma_cm QPs : OFF Data ex. method : Ethernet --------------------------------------------------------------------------------------- ibv_wr* API : ON CQ Moderation : 100 Mtu : 4096[B] Link type : IB Max inline data : 0[B] rdma_cm QPs : OFF Data ex. method : Ethernet --------------------------------------------------------------------------------------- local address: LID 0x01 QPN 0x0029 PSN 0x6763b2 RKey 0x180eef VAddr 0x007a08b4800000 local address: LID 0x02 QPN 0x36ef PSN 0x149b7b RKey 0x180efd VAddr 0x007c0146800000 remote address: LID 0x02 QPN 0x36ef PSN 0x149b7b RKey 0x180efd VAddr 0x007c0146800000 --------------------------------------------------------------------------------------- #bytes #iterations BW peak[MB/sec] BW average[MB/sec] MsgRate[Mpps] remote address: LID 0x01 QPN 0x0029 PSN 0x6763b2 RKey 0x180eef VAddr 0x007a08b4800000 --------------------------------------------------------------------------------------- #bytes #iterations BW peak[MB/sec] BW average[MB/sec] MsgRate[Mpps] Conflicting CPU frequency values detected: 1500.000000 != 1857.916000. CPU Frequency is not max. 2 5000 3.93 3.91 2.049159 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 4 5000 7.90 7.86 2.061297 Conflicting CPU frequency values detected: 1500.000000 != 1751.986000. CPU Frequency is not max. 8 5000 15.78 15.72 2.060780 Conflicting CPU frequency values detected: 1500.000000 != 3393.685000. CPU Frequency is not max. 16 5000 31.55 31.55 2.067723 Conflicting CPU frequency values detected: 1500.000000 != 3393.672000. CPU Frequency is not max. 32 5000 63.17 63.16 2.069580 Conflicting CPU frequency values detected: 1500.000000 != 3393.684000. CPU Frequency is not max. 64 5000 125.18 124.79 2.044571 Conflicting CPU frequency values detected: 1500.000000 != 3393.682000. CPU Frequency is not max. 128 5000 251.97 251.63 2.061392 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 256 5000 503.47 502.38 2.057737 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 512 5000 1007.86 1002.91 2.053960 Conflicting CPU frequency values detected: 1500.000000 != 1464.256000. CPU Frequency is not max. 1024 5000 2008.34 2007.01 2.055178 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 2048 5000 3710.86 3561.43 1.823450 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 4096 5000 4482.59 4311.86 1.103835 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 8192 5000 4579.76 4293.50 0.549568 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 16384 5000 4427.76 4284.51 0.274209 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 32768 5000 4431.66 4325.50 0.138416 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 65536 5000 4456.39 4428.15 0.070850 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 131072 5000 4508.29 4499.07 0.035993 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 262144 5000 4535.99 4529.94 0.018120 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 524288 5000 4561.30 4553.91 0.009108 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 1048576 5000 4556.28 4554.37 0.004554 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 2097152 5000 4552.12 4551.60 0.002276 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 4194304 5000 4553.83 4553.14 0.001138 Conflicting CPU frequency values detected: 1500.000000 != 2250.000000. CPU Frequency is not max. 8388608 5000 4553.65 4552.61 0.000569 --------------------------------------------------------------------------------------- 8388608 5000 4553.65 4552.61 0.000569 --------------------------------------------------------------------------------------- deallocating RX GPU buffer 00007a08b4000000 deallocating RX GPU buffer 00007c0146000000 destroying current CUDA Ctx destroying current CUDA Ctx ** Tags removed: verification-needed-mantic-linux ** Tags added: verification-done-mantic-linux -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/2055082 Title: IB peer memory feature regressed in 6.5 Status in linux package in Ubuntu: Won't Fix Status in linux source package in Mantic: Fix Committed Bug description: [Impact] The GPU Direct over Infiniband feature of NVIDIA GPUs no longer works on jammy/hwe kernels that have migrated to 6.5 *except* for the -nvidia kernel, which pulled in support via bug 2049537. We have carried this patch since 5.4 (see bug 1923104). We do not plan to carry this patch into 6.8 or later - we are working on a deprecation post for that to give users some time to migrate. [Test Case] https://git.launchpad.net/~canonical-kernel-team/+git/autotest-client-tests/tree/ubuntu_performance_gpudirect_rdma To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2055082/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
