环境:Atlas 800 9000服务器,910A NPU卡
背景:
- 原环境Ubuntu 18.04, CANN 5.1,linux kernel 4.15.0-156-generic
- 现在要升级到Ubuntu 20.04, CANN 6.3
问题:
Ubuntu 20.04升级成功,当前linux kernel版本 5.4.0-163-generic
执行以下命令升级NPU Driver:
./Ascend-hdk-910-npu-driver_23.0.rc2_linux-aarch64.run --full
报错
相应错误的log在附件中,报错信息如下
cmdline 错误信息:
Verifying archive integrity... 100% SHA256 checksums are OK. All good.
Uncompressing ASCEND DRIVER RUN PACKAGE 100%
[Driver] [2023-09-20 23:17:44] [INFO]Start time: 2023-09-20 23:17:44
[Driver] [2023-09-20 23:17:44] [INFO]LogFile: /var/log/ascend_seclog/ascend_install.log
[Driver] [2023-09-20 23:17:44] [INFO]OperationLogFile: /var/log/ascend_seclog/operation.log
[Driver] [2023-09-20 23:17:44] [INFO]base version is none.
[Driver] [2023-09-20 23:17:44] [WARNING]Do not power off or restart the system during the installation/upgrade
[Driver] [2023-09-20 23:17:45] [INFO]set username and usergroup, HwHiAiUser:HwHiAiUser
/usr/local/Ascend/driver/tools/upgrade-tool: error while loading shared libraries: libdrvdsmi_host.so: cannot open shared object file: No such file or directory
[Driver] [2023-09-20 23:18:32] [INFO]driver install type: DKMS
[Driver] [2023-09-20 23:18:32] [INFO]upgradePercentage:10%
[Driver] [2023-09-20 23:18:35] [INFO]upgradePercentage:30%
[Driver] [2023-09-20 23:18:35] [INFO]upgradePercentage:40%
[Driver] [2023-09-20 23:18:53] [ERROR]Dkms install failed, details in : /var/log/ascend_seclog/ascend_install.log
[Driver] [2023-09-20 23:18:53] [ERROR]Driver_ko_install failed, details in : /var/log/ascend_seclog/ascend_install.log
[Driver] [2023-09-20 23:18:53] [INFO]Failed to install driver package, please retry after uninstall and reboot!
/var/log/ascend_seclog/ascend_install.log 错误信息:
[Driver] [2023-09-20 23:18:35] [uid=0(root)] [INFO]driver dkms log >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
[Driver] [2023-09-20 23:18:35] [uid=0(root)] [INFO]run dkms add command...
Creating symlink /var/lib/dkms/davinci_ascend/1.0/source ->
/usr/src/davinci_ascend-1.0
DKMS: add completed.
[Driver] [2023-09-20 23:18:35] [uid=0(root)] [INFO]run dkms build command...
Kernel preparation unnecessary for this kernel. Skipping...
Building module:
cleaning build area...
'make' all KERNEL_UNAME=5.4.0-163-generic DAVINCI_HIAI_DKMS=y TARGET_PRODUCT=cloud TARGET_CHIP_ID=hi1980 Driver_Install_Mode=normal.....(bad exit status: 2)
ERROR (dkms apport): binary package for davinci_ascend: 1.0 not found
Error! Bad return status for module build on kernel: 5.4.0-163-generic (aarch64)
Consult /var/lib/dkms/davinci_ascend/1.0/build/make.log for more information.
[Driver] [2023-09-20 23:18:44] [uid=0(root)] [INFO]run dkms install command...
Kernel preparation unnecessary for this kernel. Skipping...
Building module:
cleaning build area...
'make' all KERNEL_UNAME=5.4.0-163-generic DAVINCI_HIAI_DKMS=y TARGET_PRODUCT=cloud TARGET_CHIP_ID=hi1980 Driver_Install_Mode=normal.....(bad exit status: 2)
ERROR (dkms apport): binary package for davinci_ascend: 1.0 not found
Error! Bad return status for module build on kernel: 5.4.0-163-generic (aarch64)
Consult /var/lib/dkms/davinci_ascend/1.0/build/make.log for more information.
[Driver] [2023-09-20 23:18:53] [uid=0(root)] [INFO]driver dkms log <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
[Driver] [2023-09-20 23:18:53] [uid=0(root)] [ERROR]There has no /ascend_dms_dtm.ko, dkms install failed
[Driver] [2023-09-20 23:18:53] [uid=0(root)] [ERROR]dkms install files check failed
[Driver] [2023-09-20 23:18:53] [uid=0(root)] [ERROR]dkms install failed
make.log错误信息:
CC [M] /var/lib/dkms/davinci_ascend/1.0/build/dms/smf/event/dms_event_dfx.o
/var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dma_pool_map.c: In function ‘dev_iommu_dma_sync_single_for_cpu’:
/var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dma_pool_map.c:99:27: error: passing argument 1 of ‘arch_sync_dma_for_cpu’ makes integer from pointer without a cast [-Werror=int-conversion]
99 | arch_sync_dma_for_cpu(dev, phys, size, dir);
| ^~~
| |
| struct device *
In file included from /var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dma_pool_map.c:24:
./include/linux/dma-noncoherent.h:88:40: note: expected ‘phys_addr_t’ {aka ‘long long unsigned int’} but argument is of type ‘struct device *’
88 | void arch_sync_dma_for_cpu(phys_addr_t paddr, size_t size,
| ~~~~~~~~~~~~^~~~~
/var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dma_pool_map.c:99:5: error: too many arguments to function ‘arch_sync_dma_for_cpu’
99 | arch_sync_dma_for_cpu(dev, phys, size, dir);
| ^~~~~~~~~~~~~~~~~~~~~
In file included from /var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dma_pool_map.c:24:
./include/linux/dma-noncoherent.h:88:6: note: declared here
88 | void arch_sync_dma_for_cpu(phys_addr_t paddr, size_t size,
| ^~~~~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
CC [M] /var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dvt_sysfs.o
CC [M] /var/lib/dkms/davinci_ascend/1.0/build/queue/host/common/queue_proc_fs.o
CC [M] /var/lib/dkms/davinci_ascend/1.0/build/dms/drv_devmng/drv_devmng_host/ascend910/devdrv_driver_pm.o
make[3]: *** [scripts/Makefile.build:270: /var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dma_pool_map.o] Error 1
make[3]: *** Waiting for unfinished jobs....
CC [M] /var/lib/dkms/davinci_ascend/1.0/build/vmng_host/soc_adapt/virtmng_res_drv.o
环境:Atlas 800 9000服务器,910A NPU卡
背景:
问题:
Ubuntu 20.04升级成功,当前linux kernel版本 5.4.0-163-generic
执行以下命令升级NPU Driver:
./Ascend-hdk-910-npu-driver_23.0.rc2_linux-aarch64.run --full
报错
相应错误的log在附件中,报错信息如下
cmdline 错误信息:
Verifying archive integrity... 100% SHA256 checksums are OK. All good.
Uncompressing ASCEND DRIVER RUN PACKAGE 100%
[Driver] [2023-09-20 23:17:44] [INFO]Start time: 2023-09-20 23:17:44
[Driver] [2023-09-20 23:17:44] [INFO]LogFile: /var/log/ascend_seclog/ascend_install.log
[Driver] [2023-09-20 23:17:44] [INFO]OperationLogFile: /var/log/ascend_seclog/operation.log
[Driver] [2023-09-20 23:17:44] [INFO]base version is none.
[Driver] [2023-09-20 23:17:44] [WARNING]Do not power off or restart the system during the installation/upgrade
[Driver] [2023-09-20 23:17:45] [INFO]set username and usergroup, HwHiAiUser:HwHiAiUser
/usr/local/Ascend/driver/tools/upgrade-tool: error while loading shared libraries: libdrvdsmi_host.so: cannot open shared object file: No such file or directory
[Driver] [2023-09-20 23:18:32] [INFO]driver install type: DKMS
[Driver] [2023-09-20 23:18:32] [INFO]upgradePercentage:10%
[Driver] [2023-09-20 23:18:35] [INFO]upgradePercentage:30%
[Driver] [2023-09-20 23:18:35] [INFO]upgradePercentage:40%
[Driver] [2023-09-20 23:18:53] [ERROR]Dkms install failed, details in : /var/log/ascend_seclog/ascend_install.log
[Driver] [2023-09-20 23:18:53] [ERROR]Driver_ko_install failed, details in : /var/log/ascend_seclog/ascend_install.log
[Driver] [2023-09-20 23:18:53] [INFO]Failed to install driver package, please retry after uninstall and reboot!
/var/log/ascend_seclog/ascend_install.log 错误信息:
[Driver] [2023-09-20 23:18:35] [uid=0(root)] [INFO]driver dkms log >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
[Driver] [2023-09-20 23:18:35] [uid=0(root)] [INFO]run dkms add command...
Creating symlink /var/lib/dkms/davinci_ascend/1.0/source ->
/usr/src/davinci_ascend-1.0
DKMS: add completed.
[Driver] [2023-09-20 23:18:35] [uid=0(root)] [INFO]run dkms build command...
Kernel preparation unnecessary for this kernel. Skipping...
Building module:
cleaning build area...
'make' all KERNEL_UNAME=5.4.0-163-generic DAVINCI_HIAI_DKMS=y TARGET_PRODUCT=cloud TARGET_CHIP_ID=hi1980 Driver_Install_Mode=normal.....(bad exit status: 2)
ERROR (dkms apport): binary package for davinci_ascend: 1.0 not found
Error! Bad return status for module build on kernel: 5.4.0-163-generic (aarch64)
Consult /var/lib/dkms/davinci_ascend/1.0/build/make.log for more information.
[Driver] [2023-09-20 23:18:44] [uid=0(root)] [INFO]run dkms install command...
Kernel preparation unnecessary for this kernel. Skipping...
Building module:
cleaning build area...
'make' all KERNEL_UNAME=5.4.0-163-generic DAVINCI_HIAI_DKMS=y TARGET_PRODUCT=cloud TARGET_CHIP_ID=hi1980 Driver_Install_Mode=normal.....(bad exit status: 2)
ERROR (dkms apport): binary package for davinci_ascend: 1.0 not found
Error! Bad return status for module build on kernel: 5.4.0-163-generic (aarch64)
Consult /var/lib/dkms/davinci_ascend/1.0/build/make.log for more information.
[Driver] [2023-09-20 23:18:53] [uid=0(root)] [INFO]driver dkms log <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
[Driver] [2023-09-20 23:18:53] [uid=0(root)] [ERROR]There has no /ascend_dms_dtm.ko, dkms install failed
[Driver] [2023-09-20 23:18:53] [uid=0(root)] [ERROR]dkms install files check failed
[Driver] [2023-09-20 23:18:53] [uid=0(root)] [ERROR]dkms install failed
make.log错误信息:
CC [M] /var/lib/dkms/davinci_ascend/1.0/build/dms/smf/event/dms_event_dfx.o
/var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dma_pool_map.c: In function ‘dev_iommu_dma_sync_single_for_cpu’:
/var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dma_pool_map.c:99:27: error: passing argument 1 of ‘arch_sync_dma_for_cpu’ makes integer from pointer without a cast [-Werror=int-conversion]
99 | arch_sync_dma_for_cpu(dev, phys, size, dir);
| ^~~
| |
| struct device *
In file included from /var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dma_pool_map.c:24:
./include/linux/dma-noncoherent.h:88:40: note: expected ‘phys_addr_t’ {aka ‘long long unsigned int’} but argument is of type ‘struct device *’
88 | void arch_sync_dma_for_cpu(phys_addr_t paddr, size_t size,
| ~~~~~~~~~~~~^~~~~
/var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dma_pool_map.c:99:5: error: too many arguments to function ‘arch_sync_dma_for_cpu’
99 | arch_sync_dma_for_cpu(dev, phys, size, dir);
| ^~~~~~~~~~~~~~~~~~~~~
In file included from /var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dma_pool_map.c:24:
./include/linux/dma-noncoherent.h:88:6: note: declared here
88 | void arch_sync_dma_for_cpu(phys_addr_t paddr, size_t size,
| ^~~~~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
CC [M] /var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dvt_sysfs.o
CC [M] /var/lib/dkms/davinci_ascend/1.0/build/queue/host/common/queue_proc_fs.o
CC [M] /var/lib/dkms/davinci_ascend/1.0/build/dms/drv_devmng/drv_devmng_host/ascend910/devdrv_driver_pm.o
make[3]: *** [scripts/Makefile.build:270: /var/lib/dkms/davinci_ascend/1.0/build/vascend_drv/dma_pool_map.o] Error 1
make[3]: *** Waiting for unfinished jobs....
CC [M] /var/lib/dkms/davinci_ascend/1.0/build/vmng_host/soc_adapt/virtmng_res_drv.o