Hello, I have RHEL 8.3 on Power9 machine and Quadro GV100 GPU. And during gcc libgomp tests it produces the kernel stack trace below. The bug is not caused by any particular test but appears on random test. Any advice are appreciated.
[ 97.864204] ------------[ cut here ]------------
[ 97.864207] unexpected DMA address compression (0x800200e2d600000, 0x800080e2d600000)
[ 97.864367] WARNING: CPU: 16 PID: 1437 at /var/lib/dkms/nvidia/460.32.03/build/nvidia/nv-dma.c:366 nv_dma_nvlink_addr_compress.isra.2+0xac/0x170 [nvidia]
[ 97.864368] Modules linked in: i2c_dev sunrpc ext4 mbcache jbd2 xts ofpart ipmi_powernv powernv_flash ipmi_devintf ipmi_msghandler vmx_crypto mtd ibmpowernv ses enclosure opal_prd scsi_transport_sas at24 uio_pdrv_genirq uio binfmt_misc ip_tables xfs libcrc32c nvidia_drm(POE) nvidia_modeset(POE) nvidia(POE) sr_mod cdrom amdgpu sd_mod sg mfd_core gpu_sched i2c_algo_bit ttm drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops drm ahci libahci libata nvme nvme_core aacraid tg3 drm_panel_orientation_quirks dm_mirror dm_region_hash dm_log dm_mod
[ 97.864393] CPU: 16 PID: 1437 Comm: nvidia-persiste Tainted: P OE --------- - - 4.18.0-240.el8.ppc64le #1
[ 97.864394] NIP: c00800000cc0ac54 LR: c00800000cc0ac50 CTR: 0000000000000000
[ 97.864396] REGS: c000200e49f1ac30 TRAP: 0700 Tainted: P OE --------- - - (4.18.0-240.el8.ppc64le)
[ 97.864396] MSR: 900000000282b033 <SF,HV,VEC,VSX,EE,FP,ME,IR,DR,RI,LE> CR: 42004888 XER: 20040000
[ 97.864400] CFAR: c000000000159884 IRQMASK: 0
GPR00: c00800000cc0ac50 c000200e49f1aeb0 c00800000d9c4c00 0000000000000049
GPR04: 0000000000000001 0000000000000001 00000000000003f1 c000000001cbeb44
GPR08: 0000000000000007 0000000000000007 0000000000000001 30202c3030303030
GPR12: 0000000000004000 c000200fff7fee80 0000000000000000 0000000000000000
GPR16: 0000000000000000 c000200e4ddd0008 c000200e4ddf4008 c000200e4b0f5808
GPR20: c000200e4ddb2b28 0000000000052000 0000000000000006 0000000000000000
GPR24: 0000000000000001 c000200e5e4034c8 0000000000000006 c000200e49f1b0e0
GPR28: 0000000000000000 c000200e5e406ff0 c000000fe5480728 0000000000000001
[ 97.864500] NIP [c00800000cc0ac54] nv_dma_nvlink_addr_compress.isra.2+0xac/0x170 [nvidia]
[ 97.864595] LR [c00800000cc0ac50] nv_dma_nvlink_addr_compress.isra.2+0xa8/0x170 [nvidia]
[ 97.864595] Call Trace:
[ 97.864685] [c000200e49f1aeb0] [c00800000cc0ac50] nv_dma_nvlink_addr_compress.isra.2+0xa8/0x170 [nvidia] (unreliable)
[ 97.864775] [c000200e49f1af10] [c00800000cc0b814] nv_dma_map_pages+0x35c/0x6b0 [nvidia]
[ 97.864865] [c000200e49f1afd0] [c00800000cc0bf70] nv_dma_map_alloc+0x208/0x330 [nvidia]
[ 97.864998] [c000200e49f1b080] [c00800000d839b0c] _nv032887rm+0x30c/0x630 [nvidia]
[ 97.865144] [c000200e49f1b140] [c00800000cccd410] _nv027783rm+0x240/0x490 [nvidia]
[ 97.865350] [c000200e49f1b1e0] [c00800000d323564] _nv028971rm+0xb4/0x390 [nvidia]
[ 97.865556] [c000200e49f1b280] [c00800000d323c2c] _nv008563rm+0x3ec/0x4e0 [nvidia]
[ 97.865761] [c000200e49f1b410] [c00800000d323e38] _nv028921rm+0x118/0x8a0 [nvidia]
[ 97.865943] [c000200e49f1b4c0] [c00800000cee8118] _nv040076rm+0x158/0x1e0 [nvidia]
[ 97.866124] [c000200e49f1b520] [c00800000ce85c80] _nv040123rm+0xc40/0x1390 [nvidia]
[ 97.866256] [c000200e49f1b650] [c00800000d849db4] _nv000680rm+0xe34/0x22c0 [nvidia]
[ 97.866392] [c000200e49f1b780] [c00800000d83f354] rm_init_adapter+0x104/0x120 [nvidia]
[ 97.866480] [c000200e49f1b840] [c00800000cc02740] nv_start_device+0x4e8/0x890 [nvidia]
[ 97.866568] [c000200e49f1b8f0] [c00800000cc02b8c] nv_open_device+0xa4/0x1e0 [nvidia]
[ 97.866658] [c000200e49f1b970] [c00800000cc03360] nvidia_open+0x178/0x520 [nvidia]
[ 97.866746] [c000200e49f1ba20] [c00800000cc005dc] nvidia_frontend_open+0xa4/0x120 [nvidia]
[ 97.866752] [c000200e49f1ba70] [c00000000052cb30] chrdev_open+0x180/0x3c0
[ 97.866755] [c000200e49f1bae0] [c000000000518bf8] do_dentry_open+0x268/0x4d0
[ 97.866757] [c000200e49f1bb30] [c000000000540368] do_last+0x1c8/0xb00
[ 97.866759] [c000200e49f1bc00] [c0000000005413c4] path_openat+0x124/0x410
[ 97.866760] [c000200e49f1bc90] [c000000000543ac0] do_filp_open+0x90/0x170
[ 97.866761] [c000200e49f1bdb0] [c00000000051c2c8] do_sys_open+0x278/0x380
[ 97.866764] [c000200e49f1be30] [c00000000000b408] system_call+0x5c/0x70
[ 97.866764] Instruction dump:
[ 97.866766] 892a0000 2f890000 4c9e0020 7c0802a6 f8010010 f821ffa1 39200001 3c620000
[ 97.866769] e8638540 992a0000 48d06d05 e8410018 <0fe00000> 38210060 e8010010 7c0803a6
[ 97.866772] —[ end trace 2bbb1e6d8ce0ec5a ]—
[ 98.024989] NVRM: 0033:01:00.0: DMA address not in addressable range of device (0x800200e57b60000-0x800200e57b6ffff, 0x800000000000000-0x80000ffffffffff)