kzytaruk
commented
5 years ago Is this consistently reproducible or random?
Thanks,
Kelly
From: flumm [mailto:notifications@github.com] Sent: Friday, October 12, 2018 6:36 AM To: GPUOpen-LibrariesAndSDKs/MxGPU-Virtualization MxGPU-Virtualization@noreply.github.com Cc: Subscribed subscribed@noreply.github.com Subject: [GPUOpen-LibrariesAndSDKs/MxGPU-Virtualization] Host Crash qemu/kvm (#13)
uname -a: Linux myhost 4.15.18-7-pve #1 https://github.com/GPUOpen-LibrariesAndSDKs/MxGPU-Virtualization/issues/1 SMP PVE 4.15.18-26 (Thu, 04 Oct 2018 11:03:06 +0200) x86_64 GNU/Linux
even though i use kernel 4.15, compiling went without problems a modprobe produces following output though:
[ 230.259767] gim info:(gim_init:149) Start AMD open source GIM initialization [ 230.259770] gim info:(gim_init:152) GPU IOV MODULE - version 0.0 [ 230.259771] gim info:(gim_init:154) Copyright (c) 2014-2018 AMD Corporation. [ 230.260618] gim info:(parse_config_file:219) AMD GIM fb_option = 0 [ 230.260619] gim info:(parse_config_file:219) AMD GIM sched_option = 1 [ 230.260621] gim info:(parse_config_file:219) AMD GIM vf_num = 1 [ 230.260622] gim info:(parse_config_file:219) AMD GIM pf_fb = 256 [ 230.260623] gim info:(parse_config_file:219) AMD GIM vf_fb = 256 [ 230.260625] gim info:(parse_config_file:219) AMD GIM sched_interval = 0 [ 230.260626] gim info:(parse_config_file:219) AMD GIM sched_interval_us = 0 [ 230.260627] gim info:(parse_config_file:219) AMD GIM fb_clear = 0 [ 230.260629] gim info:(init_config:341) INIT CONFIG [ 230.268677] gim info:(set_new_adapter:572) curr allocated at 0000000007ecc9e8 [ 230.268678] gim info:(set_new_adapter:579) SRIOV is supported [ 230.268683] gim info:(set_new_adapter:587) found PCI bridge device [ 230.268684] gim info:(set_new_adapter:591) found: 00:3.0 [ 230.268718] gim info:(set_new_adapter:608) mmio_base = 0000000073a47b6a [ 230.268726] gim info:(set_new_adapter:610) doorbell = 00000000b86d6846 [ 230.268737] gim info:(set_new_adapter:612) pf.fb_va = 000000006a5f6588 [ 230.268746] gim info:(sriov_is_ari_enabled:164) PCI_SRIOV_CAP = 0x00000002 [ 230.268748] gim info:(sriov_is_ari_enabled:174) PCI_SRIOV_CTRL = 0x00000010 [ 230.268749] gim info:(sriov_is_ari_enabled:177) PCI_SRIOV_CTRL_ARI is set --> ARI is supported [ 230.268751] gim info:(program_ari_mode:441) Read bif_strap8 = 0x00200004 [ 230.268752] gim info:(program_ari_mode:446) program_ari_mode - Set ARI_Mode = PF_BUS [ 230.268753] gim info:(program_ari_mode:456) Write bif_strap8 = 0x00000004 [ 230.268754] gim info:(gim_read_rom_from_reg:181) Reading VBios from ROM [ 230.268903] gim info:(gim_read_vbios:243) VBIOS starts: 0x55, 0xaa [ 230.268904] gim info:(gim_read_vbios:246) VBios size is 0x10000 [ 230.268915] gim info:(gim_read_vbios:249) vbios allocated at 0000000011c744ba [ 230.268915] gim info:(gim_read_rom_from_reg:181) Reading VBios from ROM [ 230.441512] gim info:(gim_read_vbios:257) BIOS Version Major 0xF Minor 0x31 [ 230.441561] gim info:(gim_read_vbios:270) Valid video BIOS image, [ 230.441563] gim info:(gim_read_vbios:271) size = 0x10000, check sum is 0x53b100 [ 230.441565] gim info:(set_new_adapter:661) Scheduler Time interval is per-vf from XL [ 230.441566] gim info:(set_new_adapter:662) config file [ 230.441567] gim info:(enable_sriov:299) Enable SRIOV [ 230.441568] gim info:(enable_sriov:300) Enable SRIOV vfs count = 2 [ 230.548564] gim info:(enumerate_vfs:123) vf found: 05:2.0 [ 230.548577] gim info:(enumerate_vfs:123) vf found: 05:2.1 [ 230.548611] gim info:(pci_disable_error_reporting:735) Disable error reporting for device: 05:2.0 [ 230.548613] gim info:(pci_disable_error_reporting:740) Mask before -> corr = 0x00000000, uncorr = 0x00000000 [ 230.548617] gim info:(pci_disable_error_reporting:751) Mask after -> corr = 0x00000000, uncorr = 0x00000000 [ 230.548633] gim info:(pci_disable_error_reporting:735) Disable error reporting for device: 05:2.1 [ 230.548635] gim info:(pci_disable_error_reporting:740) Mask before -> corr = 0x00000000, uncorr = 0x00000000 [ 230.548639] gim info:(pci_disable_error_reporting:751) Mask after -> corr = 0x00000000, uncorr = 0x00000000 [ 230.548648] gim info:(pci_gpu_iov_init:87) total_fb_available = 8190 [ 230.548649] gim info:(pci_gpu_iov_init:88) AMD GIM pci_gpu_iov_init pos = 400 [ 230.548650] gim info:(pci_gpu_iov_init:90) AMD GIM pci_gpu_iov_init total_fb_available = 1ffe [ 230.548652] gim info:(init_frame_buffer_partition:218) PCI defined PF FB size = 256 MB [ 230.548653] gim info:(init_frame_buffer_partition:222) PCI defined VF FB size = 256 MB [ 230.548654] gim info:(init_frame_buffer_partition:227) Total FB Available = 8190 MB, CSA = 8 MB [ 230.548655] gim info:(init_frame_buffer_partition:228) Max Remaining FB Size = 8160 [ 230.548656] gim info:(init_frame_buffer_partition:241) PF FB size after checking limits from config file = 256MB [ 230.548657] gim info:(init_frame_buffer_partition:244) PF rounded down to nearest 16MB boundary = 256 [ 230.548658] gim info:(init_pf_fb:60) total framebuffer available = 1ffe [ 230.548659] gim info:(init_pf_fb:61) pf framebuffer = 100 [ 230.548660] gim info:(init_pf_fb:63) total framebuffer consumed = 1efe [ 230.548661] gim info:(init_frame_buffer_partition:251) CSA starts at offset 256MB [ 230.548663] gim info:(init_context_save_area:42) AMD GIM init_context_save_area: base =100 size=1. [ 230.548664] gim info:(init_frame_buffer_partition:258) VF FB base = 272MB (256 + 16) [ 230.548666] gim info:(init_frame_buffer_partition:262) VF FB Size = 7904MB (8160 - 256) [ 230.548667] gim info:(init_fb_static:146) AMD GIM init_fb_static: num_vf = 2, base= 272, total_size=7904, min_size=256 [ 230.548669] gim info:(init_fb_static:158) VF FB size specified as 256MB, min_size = 256 [ 230.548670] gim info:(init_fb_static:167) AMD GIM init_fb_static: vf_fb_size = 256, base= 272 [ 230.548671] gim info:(init_fb_static:178) AMD GIM init_fb_static: partition 0 base =272,size= 256 [ 230.548673] gim info:(init_fb_static:178) AMD GIM init_fb_static: partition 1 base =528,size= 256 [ 230.548694] gim info:(set_new_adapter:707) enable MSI [ 230.548735] gim info:(ih_iv_ring_disable:383) disable iv ring successfully [ 230.548736] gim info:(alloc_iv_ring:99) ih->ivr_num_entries = 256 [ 230.548737] gim info:(alloc_iv_ring:102) ih->ivr_size_in_bytes = 4096 [ 230.548738] gim info:(alloc_iv_ring:108) ih->ivr_alloc_size_in_bytes = 4100 [ 230.548739] gim info:(alloc_iv_ring:110) iv ring page_cnt = 2 [ 230.548742] gim info:(alloc_iv_ring:141) ih->ivr_va = 0000000006512d73 [ 230.548862] gim info:(alloc_iv_ring:147) ih->ivr_ma.quad_part = 0xfffff000 [ 230.548863] gim info:(alloc_iv_ring:151) ih->ivr_wptr_wb = 000000004c4df8ef [ 230.548867] gim info:(alloc_iv_ring:158) ih->ivr_wptr_wa.quad_part = 0xffffe000 [ 230.548868] gim info:(alloc_iv_ring:163) update rptr via doorbell [ 230.548869] gim info:(ih_iv_ring_init:291) ih->rptr_doorbell = 0000000052363f63 [ 230.548870] gim info:(ih_iv_ring_init:292) ih->rptr_doorbell_offset = 0x1e8 [ 230.548872] gim info:(ih_iv_ring_hw_init:185) the physical address of ring buffer: 0xfffff0 [ 230.548878] gim info:(ih_iv_ring_setup_rptr:451) write mmBIF_DOORBELL_APER_EN: 0x1 [ 230.548880] gim info:(ih_iv_ring_enable:350) ih->ivr_wptr_reg = 0x0 [ 230.548881] gim info:(ih_iv_ring_enable:352) ih->ivr_wptr = 0 [ 230.548882] gim info:(ih_iv_ring_enable:354) ih->ivr_rptr_reg = 0x0 [ 230.548882] gim info:(ih_iv_ring_enable:356) ih->ivr_rptr = 0 [ 230.548884] gim info:(ih_iv_ring_enable:358) *(ih->rptr_doorbell) = 0x0 [ 230.548885] gim info:(ih_iv_ring_init:299) init iv ring successfully [ 230.548886] gim info:(set_new_adapter:720) init work [ 230.548887] gim info:(set_new_adapter:726) register interrupt [ 230.548923] gim info:(ih_irq_source_enable:584) IH: read 0x00000000 from mask_reg 0x14d1 [ 230.548925] gim info:(ih_irq_source_enable:590) IH: write 0x00000001 to mask_reg 0x14d1 [ 230.548926] gim info:(ih_irq_source_enable:593) irq sourceID 0x89 get enabled [ 230.548928] gim info:(ih_irq_source_enable:584) IH: read 0x00000001 from mask_reg 0x14d1 [ 230.548929] gim info:(ih_irq_source_enable:590) IH: write 0x00000003 to mask_reg 0x14d1 [ 230.548930] gim info:(ih_irq_source_enable:593) irq sourceID 0x88 get enabled [ 233.549138] gim error:(wait_cmd_complete:1681) wait_cmd_complete -- time out after 3.000038872 sec [ 233.549181] gim error:(wait_cmd_complete:1688) Cmd = 0x17, Status = 0x0 [ 233.549210] gim error:(dump_gpu_status:1420) dump gpu status begin for struct adapter 5:00.00 [ 233.549248] gim info:(check_base_addrs:1408) CP_MQD_BASE_ADDR = 0x0:00000000 [ 233.549271] gim error:(dump_gpu_status:1457) mmGRBM_STATUS = 0x3028 [ 233.549297] gim error:(dump_gpu_status:1460) mmGRBM_STATUS2 = 0x8 [ 233.549323] gim error:(dump_gpu_status:1463) mmSRBM_STATUS = 0x20000040 [ 233.549351] gim error:(dump_gpu_status:1466) mmSRBM_STATUS2 = 0x0 [ 233.549377] gim error:(dump_gpu_status:1469) mmSDMA0_STATUS_REG = 0x46dee557 [ 233.549406] gim error:(dump_gpu_status:1472) mmSDMA1_STATUS_REG = 0x46dee557 [ 233.549439] gim info:(check_me_cntl:1386) CP_ME_CNTL = 0x15000000 GPU dump [ 233.549440] gim error:(check_me_cntl:1388) ME HALTED! [ 233.549462] gim error:(check_me_cntl:1391) PFP HALTED! [ 233.549484] gim error:(check_me_cntl:1394) CE HALTED! [ 233.549507] gim error:(dump_gpu_status:1588) dump gpu status end [ 233.549534] gim error:(init_register_init_state:3624) Failed to INIT PF for initial register 'init-state' [ 233.549541] gim info:(gim_clear_all_errors:357) PCIE cap pos 58 [ 233.549580] gim info:(gim_clear_all_errors:362) AER ext cap pos 150 [ 233.549581] gim info:(gim_clear_all_errors:369) DevStatus = 0x9 [ 233.549584] gim info:(gim_clear_all_errors:387) PCIE unrecoverable error = 0x2000 [ 233.549589] gim info:(add_func_to_run_list:2459) Add VF0 to the runlist [ 233.549591] gim info:(alloc_fn_list_node:3837) New Function List Node allocated at 00000000b55b812e index 0 [ 233.549592] gim info:(add_func_to_run_list:2459) Add VF1 to the runlist [ 233.549593] gim info:(alloc_fn_list_node:3837) New Function List Node allocated at 00000000ce52204e index 1 [ 233.549595] gim warning:(resume_scheduler:155) Restart the Scheduler for 7.000msec [ 233.549597] gim info:(gim_probe:91) AMD GIM probe: pf_count = 1
especially this:
[ 233.549534] gim error:(init_register_init_state:3624) Failed to INIT PF for initial register 'init-state'
after this, lspci shows the vgpu:
lspci | grep AMD
05:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Tonga XT GL [FirePro S7150] 05:02.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Tonga XTV GL [FirePro S7150V] 05:02.1 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Tonga XTV GL [FirePro S7150V]
even passing through 05:02.0 works, up to the point where i install the guest driver (win10 x64) then i am greeted with following output in dmesg and the host crashes:
[ 696.091017] gim info:(ih_irq_process:713) AMD ISR is being invoked [ 696.091019] gim info:(ih_iv_ring_get_pointers:482) ih_iv_ring_get_pointers [ 696.091021] gim info:(ih_iv_ring_get_pointers:483) ih->ivr_wptr_wb = 0x000000004c4df8ef [ 696.091023] gim info:(ih_iv_ring_get_pointers:485) write offset: *(ih->ivr_wptr_wb) = 0x10 [ 696.091024] gim info:(ih_iv_ring_get_pointers:486) read idx: ih->ivr_rptr = 0x0 [ 696.091025] gim info:(ih_iv_ring_get_pointers:488) Rx at entry 0 in the ring [ 696.091026] gim info:(ih_iv_ring_get_pointers:490) iv_ring_entry.source_id = 137 [ 696.091027] gim info:(ih_iv_ring_get_pointers:491) iv_ring_entry.source_data = 0 [ 696.091029] gim info:(ih_irq_process:734) received 1 irqs in one ISR [ 696.091031] gim info:(ih_irq_process:757) VF_PF_MSGBUF_VALID received(Received msg from VF) [ 696.091033] gim info:(mailbox_update_index:836) write mmMAILBOX_INDEX: 0x0 [ 696.091037] gim info:(mailbox_msg_rcv:882) read mmMAILBOX_MSGBUF_RCV_DW0: 0x1 [ 696.091038] gim info:(mailbox_msg_rcv:883) read mmMAILBOX_MSGBUF_RCV_DW1: 0 [ 696.091039] gim info:(mailbox_msg_rcv:884) read mmMAILBOX_MSGBUF_RCV_DW2: 0x0 [ 696.091040] gim info:(mailbox_msg_rcv:885) read mmMAILBOX_MSGBUF_RCV_DW3: 0x0 [ 696.091043] gim info:(mailbox_ack_receipt:1008) write mmMAILBOX_CONTROL: 0x300 to MAILBOX_INDEX 0x0 [ 696.091044] gim info:(ih_irq_process:778) GPU access flag = 0x0 [ 696.091046] gim info:(idh_queue:656) new idh: task->event = 1, task->func_id = 0 [ 696.091048] gim info:(mailbox_update_index:836) write mmMAILBOX_INDEX: 0x0 [ 696.091053] gim info:(ih_iv_ring_update_rptr:546) update the new rptr: ih->ivr_rptr_reg = 0x10 [ 696.091054] gim info:(ih_iv_ring_update_rptr:554) update rptr via doorbell: 0x10 [ 696.091055] gim info:(ih_iv_ring_update_rptr:556) current wptr: 0x10 [ 696.091055] gim info:(ih_irq_process:823) AMD ISR is complete [ 696.091079] gim info:(signal_scheduler:1561) Invoked the task scheduler thread. Process IRQ activity [ 696.091080] gim info:(signal_scheduler:1573) Got a REQ_GPU_ACCESS task for VFindex 0 [ 696.091081] gim info:(signal_scheduler:1577) req_gpu_task --> Event = 1;FuncID = 0 [ 696.091082] gim info:(signal_scheduler:1663) IDH_REQ_GPU_INIT_ACCESS [ 696.091084] gim info:(handle_req_gpu_init_access:1184) handle req_gpu_init_access event for VF0, reset flag = 0 [ 696.091085] gim info:(alloc_new_vf:2321) alloc_new_vf with pid 0 [ 696.091087] gim info:(load_vbios:4031) FB.start = 0x43ee0000000, FB.size = 0x43eefffffff [ 696.091089] gim info:(load_vbios:4045) VBios -> 55 AA 80 E9 29 3 0 0 [ 696.091091] gim info:(map_vf_fb:381) Map region 0x43ee0000000 for length 268435456 [ 696.091148] gim info:(load_vbios:4048) FB VA = 00000000f4a04997 [ 696.091152] gim info:(load_vbios:4052) Copy VBios from 0000000011c744ba to 00000000f4a04997 for length 0x10000 [ 696.097027] gim info:(load_vbios:4055) UnMap FB VA 00000000f4a04997 [ 696.097079] gim info:(alloc_new_vf:2337) sav_res_list_size = -166651663 [ 697.030272] gim info:(alloc_new_vf:2427) Defer adding VF to runlist until Mailbox work queue [ 697.030274] gim info:(load_vbios:4031) FB.start = 0x43ee0000000, FB.size = 0x43eefffffff [ 697.030277] gim info:(load_vbios:4045) VBios -> 55 AA 80 E9 29 3 0 0 [ 697.030279] gim info:(map_vf_fb:381) Map region 0x43ee0000000 for length 268435456 [ 697.030315] gim info:(load_vbios:4048) FB VA = 0000000005ed60dd [ 697.030317] gim info:(load_vbios:4052) Copy VBios from 0000000011c744ba to 0000000005ed60dd for length 0x10000 [ 697.036194] gim info:(load_vbios:4055) UnMap FB VA 0000000005ed60dd [ 697.036229] gim info:(alloc_new_vf:2441) end of alloc_new_vf [ 697.036231] gim info:(pause_scheduler:124) Stop the Scheduler [ 697.036232] gim info:(stop_current_vf:2097) stop current vf[1] [ 697.036233] gim warning:(stop_current_vf:2121) runlist is not scheduled, ignore stop [ 697.036234] gim info:(handle_req_gpu_init_access:1229) init and run new VF [ 697.036236] gim info:(run_sdma:3696) Check SDMA for bdf 0x500 [ 697.036238] gim info:(run_sdma:3699) SDMA0 is HALTED, unhalt it [ 697.036239] gim info:(run_sdma:3709) SDMA1 is HALTED, unhalt it [ 697.036242] gim info:(run_sdma:3720) SDMA0_F32_CNTL = 0x0000, SDMA1_F32_CNTL = 0x0000 [ 697.036245] gim info:(run_sdma:3724) SDMA0 WPTR/RPTR = 0x0000 / 0x0000 [ 697.036246] gim info:(run_sdma:3696) Check SDMA for bdf 0x510 [ 697.036248] gim info:(run_sdma:3704) SDMA0 is already running;It doesn't need to be unhalted [ 697.036249] gim info:(run_sdma:3714) SDMA1 is already running;It doesn't need to be unhalted [ 697.036252] gim info:(run_sdma:3720) SDMA0_F32_CNTL = 0x0000, SDMA1_F32_CNTL = 0x0000 [ 697.036254] gim info:(run_sdma:3724) SDMA0 WPTR/RPTR = 0x0000 / 0x0000 [ 699.036429] gim error:(wait_cmd_complete:1681) wait_cmd_complete -- time out after 2.000068956 sec [ 699.036471] gim error:(wait_cmd_complete:1688) Cmd = 0x13, Status = 0x0 [ 699.036500] gim error:(dump_gpu_status:1420) dump gpu status begin for struct adapter 5:00.00 [ 699.036538] gim info:(check_base_addrs:1408) CP_MQD_BASE_ADDR = 0x0:00000000 [ 699.036565] gim error:(dump_gpu_status:1457) mmGRBM_STATUS = 0x3028 [ 699.036592] gim error:(dump_gpu_status:1460) mmGRBM_STATUS2 = 0x8 [ 699.036617] gim error:(dump_gpu_status:1463) mmSRBM_STATUS = 0x20000040 [ 699.036645] gim error:(dump_gpu_status:1466) mmSRBM_STATUS2 = 0x0 [ 699.036671] gim error:(dump_gpu_status:1469) mmSDMA0_STATUS_REG = 0x46dee557 [ 699.036701] gim error:(dump_gpu_status:1472) mmSDMA1_STATUS_REG = 0x46dee557 [ 699.036736] gim info:(check_me_cntl:1386) CP_ME_CNTL = 0x15000000 GPU dump [ 699.036737] gim error:(check_me_cntl:1388) ME HALTED! [ 699.036759] gim error:(check_me_cntl:1391) PFP HALTED! [ 699.036781] gim error:(check_me_cntl:1394) CE HALTED! [ 699.036803] gim error:(dump_gpu_status:1588) dump gpu status end [ 699.036830] gim error:(load_register_init_state:3666) Failed to LOAD register 'init-state' [ 699.036832] gim info:(mark_func_scheduled:2614) Mark VF0 is scheduled [ 699.036865] gim warning:(add_func_to_run_list:2488) VF0 is already on the run_list. [ 702.037033] gim error:(wait_cmd_complete:1681) wait_cmd_complete -- time out after 3.000063346 sec [ 702.037108] gim error:(wait_cmd_complete:1688) Cmd = 0x17, Status = 0x0 [ 702.037186] gim error:(dump_gpu_status:1420) dump gpu status begin for struct adapter 5:00.00 [ 702.037258] gim info:(check_base_addrs:1408) CP_MQD_BASE_ADDR = 0x0:00000000 [ 702.037285] gim error:(dump_gpu_status:1457) mmGRBM_STATUS = 0x3028 [ 702.037335] gim error:(dump_gpu_status:1460) mmGRBM_STATUS2 = 0x8 [ 702.037383] gim error:(dump_gpu_status:1463) mmSRBM_STATUS = 0x20000040 [ 702.037435] gim error:(dump_gpu_status:1466) mmSRBM_STATUS2 = 0x0 [ 702.037484] gim error:(dump_gpu_status:1469) mmSDMA0_STATUS_REG = 0x46dee557 [ 702.037539] gim error:(dump_gpu_status:1472) mmSDMA1_STATUS_REG = 0x46dee557 [ 702.038959] gim info:(check_me_cntl:1386) CP_ME_CNTL = 0x15000000 GPU dump [ 702.038960] gim error:(check_me_cntl:1388) ME HALTED! [ 702.040385] gim error:(check_me_cntl:1391) PFP HALTED! [ 702.041795] gim error:(check_me_cntl:1394) CE HALTED! [ 702.043154] gim error:(dump_gpu_status:1588) dump gpu status end [ 702.044508] gim info:(load_vbios:4031) FB.start = 0x43ee0000000, FB.size = 0x43eefffffff [ 702.044511] gim info:(load_vbios:4045) VBios -> 55 AA 80 E9 29 3 0 0 [ 702.044513] gim info:(map_vf_fb:381) Map region 0x43ee0000000 for length 268435456 [ 702.044570] gim info:(load_vbios:4048) FB VA = 0000000005ed60dd [ 702.044573] gim info:(load_vbios:4052) Copy VBios from 0000000011c744ba to 0000000005ed60dd for length 0x10000 [ 702.050448] gim info:(load_vbios:4055) UnMap FB VA 0000000005ed60dd [ 702.050496] gim info:(run_sdma:3696) Check SDMA for bdf 0x500 [ 702.050498] gim info:(run_sdma:3704) SDMA0 is already running;It doesn't need to be unhalted [ 702.050499] gim info:(run_sdma:3714) SDMA1 is already running;It doesn't need to be unhalted [ 702.050502] gim info:(run_sdma:3720) SDMA0_F32_CNTL = 0x0000, SDMA1_F32_CNTL = 0x0000 [ 702.050505] gim info:(run_sdma:3724) SDMA0 WPTR/RPTR = 0x0000 / 0x0000 [ 702.050507] gim info:(run_sdma:3696) Check SDMA for bdf 0x510 [ 702.050509] gim info:(run_sdma:3704) SDMA0 is already running;It doesn't need to be unhalted [ 702.050511] gim info:(run_sdma:3714) SDMA1 is already running;It doesn't need to be unhalted [ 702.050513] gim info:(run_sdma:3720) SDMA0_F32_CNTL = 0x0000, SDMA1_F32_CNTL = 0x0000 [ 702.050517] gim info:(run_sdma:3724) SDMA0 WPTR/RPTR = 0x0000 / 0x0000 [ 704.050654] gim error:(wait_cmd_complete:1681) wait_cmd_complete -- time out after 2.000031965 sec [ 704.051388] gim error:(wait_cmd_complete:1688) Cmd = 0x14, Status = 0x0 [ 704.052108] gim error:(dump_gpu_status:1420) dump gpu status begin for struct adapter 5:00.00 [ 704.052853] gim info:(check_base_addrs:1408) CP_MQD_BASE_ADDR = 0x0:00000000 [ 704.052879] gim error:(dump_gpu_status:1457) mmGRBM_STATUS = 0x3028 [ 704.053639] gim error:(dump_gpu_status:1460) mmGRBM_STATUS2 = 0x8 [ 704.054405] gim error:(dump_gpu_status:1463) mmSRBM_STATUS = 0x20000040 [ 704.055115] gim error:(dump_gpu_status:1466) mmSRBM_STATUS2 = 0x0 [ 704.055792] gim error:(dump_gpu_status:1469) mmSDMA0_STATUS_REG = 0x46dee557 [ 704.056450] gim error:(dump_gpu_status:1472) mmSDMA1_STATUS_REG = 0x46dee557 [ 704.057092] gim info:(check_me_cntl:1386) CP_ME_CNTL = 0x15000000 GPU dump [ 704.057093] gim error:(check_me_cntl:1388) ME HALTED! [ 704.057731] gim error:(check_me_cntl:1391) PFP HALTED! [ 704.058409] gim error:(check_me_cntl:1394) CE HALTED! [ 704.059039] gim error:(dump_gpu_status:1588) dump gpu status end [ 706.059855] gim error:(wait_cmd_complete:1681) wait_cmd_complete -- time out after 2.000070472 sec [ 706.061148] gim error:(wait_cmd_complete:1688) Cmd = 0x18, Status = 0x0 [ 706.062436] gim error:(dump_gpu_status:1420) dump gpu status begin for struct adapter 5:00.00 [ 706.063755] gim info:(check_base_addrs:1408) CP_MQD_BASE_ADDR = 0x0:00000000 [ 706.063782] gim error:(dump_gpu_status:1457) mmGRBM_STATUS = 0x3028 [ 706.065101] gim error:(dump_gpu_status:1460) mmGRBM_STATUS2 = 0x8 [ 706.066439] gim error:(dump_gpu_status:1463) mmSRBM_STATUS = 0x20000040 [ 706.067178] gim error:(dump_gpu_status:1466) mmSRBM_STATUS2 = 0x0 [ 706.067845] gim error:(dump_gpu_status:1469) mmSDMA0_STATUS_REG = 0x46dee557 [ 706.068496] gim error:(dump_gpu_status:1472) mmSDMA1_STATUS_REG = 0x46dee557 [ 706.069153] gim info:(check_me_cntl:1386) CP_ME_CNTL = 0x15000000 GPU dump [ 706.069154] gim error:(check_me_cntl:1388) ME HALTED! [ 706.069798] gim error:(check_me_cntl:1391) PFP HALTED! [ 706.070411] gim error:(check_me_cntl:1394) CE HALTED! [ 706.071025] gim error:(dump_gpu_status:1588) dump gpu status end [ 706.071653] gim info:(save_rlcv_state:1970) SAVE_RLCV_STATE failed on VF0 [ 706.071655] gim info:(gim_save_vddgfx_state:116) RLCV responced SAVE_RLCV_STATE [ 706.072015] gim info:(gim_save_vddgfx_state:123) mmRLC_GPU_IOV_SCRATCH_ADDR = 0 [ 706.072222] gim info:(gim_save_vddgfx_state:130) RLCV scratch saved [ 706.072223] gim info:(handle_req_gpu_init_access:1277) amdgpuv_save_vddgfx_state: VF 0 vddgfx state is saved [ 706.072226] gim info:(mailbox_update_index:836) write mmMAILBOX_INDEX: 0x0 [ 706.072230] gim info:(mailbox_msg_trn:936) write mmMAILBOX_MSGBUF_TRN_DW0: 0x1 to MAILBOX_INDEX 0x0 [ 706.072232] gim info:(mailbox_msg_trn:952) write mmMAILBOX_CONTROL: 0x1 [ 706.072241] gim info:(ih_irq_process:713) AMD ISR is being invoked [ 706.072242] gim info:(ih_iv_ring_get_pointers:482) ih_iv_ring_get_pointers [ 706.072244] gim info:(ih_iv_ring_get_pointers:483) ih->ivr_wptr_wb = 0x000000004c4df8ef [ 706.072245] gim info:(ih_iv_ring_get_pointers:485) write offset: *(ih->ivr_wptr_wb) = 0x20 [ 706.072246] gim info:(ih_iv_ring_get_pointers:486) read idx: ih->ivr_rptr = 0x1 [ 706.072247] gim info:(ih_iv_ring_get_pointers:488) Rx at entry 1 in the ring [ 706.072248] gim info:(ih_iv_ring_get_pointers:490) iv_ring_entry.source_id = 136 [ 706.072249] gim info:(ih_iv_ring_get_pointers:491) iv_ring_entry.source_data = 0 [ 706.072250] gim info:(ih_irq_process:734) received 1 irqs in one ISR [ 706.072252] gim info:(ih_irq_process:792) PF_VF_MSGBUF_ACK received(VF has ACK'd the msg) [ 706.072254] gim info:(mailbox_update_index:836) write mmMAILBOX_INDEX: 0x0 [ 706.072255] gim info:(mailbox_clear_msg_valid:1022) write mmMAILBOX_CONTROL: 0x2 [ 706.072257] gim info:(mailbox_update_index:836) write mmMAILBOX_INDEX: 0x0 [ 706.072259] gim info:(ih_iv_ring_update_rptr:546) update the new rptr: ih->ivr_rptr_reg = 0x20 [ 706.072260] gim info:(ih_iv_ring_update_rptr:554) update rptr via doorbell: 0x20 [ 706.072261] gim info:(ih_iv_ring_update_rptr:556) current wptr: 0x20 [ 706.072262] gim info:(ih_irq_process:823) AMD ISR is complete
any ideas?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/GPUOpen-LibrariesAndSDKs/MxGPU-Virtualization/issues/13 , or mute the thread https://github.com/notifications/unsubscribe-auth/ASfsU7ng3L80xwmTP2IfBuSqocJw7RAYks5ukHCWgaJpZM4XZOWr . https://github.com/notifications/beacon/ASfsU0BYe544S1exUz5VHEykBpuih-82ks5ukHCWgaJpZM4XZOWr.gif
flumm
vigchand2705
uname -a: Linux myhost 4.15.18-7-pve #1 SMP PVE 4.15.18-26 (Thu, 04 Oct 2018 11:03:06 +0200) x86_64 GNU/Linux
even though i use kernel 4.15, compiling went without problems a modprobe produces following output though:
especially this:
after this, lspci shows the vgpu:
lspci | grep AMD
even passing through 05:02.0 works, up to the point where i install the guest driver (win10 x64) then i am greeted with following output in dmesg and the host crashes:
any ideas?