[GSD-13016] Sysman events are not supported with the release packages
还没有人认领这个 Issue。
评估
- 难度
- 3/5
- 预计耗时
- 1-2 天
- 新手友好度
- 58/100
- Issue 类型
- 缺陷
- 描述清晰度
- 基本清楚
- 活跃度
- 冷清
- 技术栈
- cmake, cpp, linux
调研方向
从 level_zero/tools/source/sysman/linux/udev/CMakeLists.txt 和 release 打包配置开始,然后在 Ubuntu 24.04 上使用 release 库构建 zello_sysman。比较 release 二进制文件和本地构建的二进制文件是否支持 libudev,并使用 ZELLO_SYSMAN_USE_ZESINIT=1 运行 zello_sysman;当 release 软件包支持 sysman 事件,且事件等待行为与本地构建一致时,即表示完成。
由索引模型根据 Issue 内容生成。
描述
Pre-submission Checklist
- I am using the latest GPU driver version (releases)
- I have searched for similar issues and found none
GPU Hardware
Intel Arc B580
DRI Devices Information
n/a
GPU Detailed Information (lspci output)
n/a
Driver Version
26.22.38646.4
Installed GPU Driver Packages
$ dpkg --list | grep -iE "libze"
ii libze-dev 1.31.0 amd64 oneAPI Level Zero
ii libze-intel-gpu1 26.22.38646.4-0 amd64 Intel(R) Graphics Compute Runtime for oneAPI Level Zero.
ii libze1 1.31.0 amd64 oneAPI Level Zero
Driver Installation Details
With dpkg/apt
Linux Distribution
Ubuntu 24.04 LTS
Other Linux Distribution
No response
Kernel Version & Boot Parameters
n/a
Actual Behavior
I installed the latest release packages to a 24.04 Ubuntu system (https://github.com/intel/compute-runtime/releases/tag/26.22.38646.4) and built the zello_sysman tool from the same sources.
When I run the tool against the release packages I get this behavior:
$ export PrintDebugMessages=1
$ export NEOReadDebugKeys=1
$ sudo -E ZELLO_SYSMAN_USE_ZESINIT=1 ./workspace/compute-runtime/build/bin/zello_sysman -E
ZES_ENABLE_SYSMAN environment variable Not Set
WARNING: Failed to request OCL Turbo Boost
WARNING: Failed to request OCL Turbo Boost
Sysman Initialization done via zesInit
libudev library instantiation failed
The key line being the last libudev library instantiation failed. The execution also takes only <1sec, while it should wait for events.
When I run the same tool against compute-runtime libraries that I built myself:
$ sudo -E LD_LIBRARY_PATH=./workspace/compute-runtime/build/bin ZELLO_SYSMAN_USE_ZESINIT=1 ./workspace/compute-runtime/build/bin/zello_sysman -E
ZES_ENABLE_SYSMAN environment variable Not Set
WARNING: Failed to request OCL Turbo Boost
WARNING: Failed to request OCL Turbo Boost
Sysman Initialization done via zesInit
No line about libudev and the execution takes 10 seconds.
When checking the binaries for udev strings:
$ strings /usr/lib/x86_64-linux-gnu/libze_intel_gpu.so.1 | grep -E udev
libudev library instantiation failed
:~/workspace/compute-runtime/build$ strings bin/libze_intel_gpu.so | grep -E "udev"
libudev.
libudev.
udev_monitor_new_from_netlink
udev_monitor_enable_receiving
udev_monitor_get_fd
udev_monitor_receive_device
udev_device_get_devnum
udev_device_get_action
udev_device_unref
libudev library instantiation failed
udev_monitor_filter_add_match_subsystem_devtype
udev_device_get_property_value
udev_new
udev_new
...
My binary has udev strings while the release binary does not. Looking at the level_zero/tools/source/sysman/linux/udev/CMakeLists.txt file, the compilation is conditional based on if libudev is found. Most likely the build env doesn't have libudev-dev thus the releases never support sysman events.
Expected Behavior
Release version of compute runtime should support sysman events.
Reproduction Rate
Always reproduces - 100%
Steps to Reproduce
- Install compute-runtime and its dependencies
- Build zello_sysman
- Run zello_sysman
$ export PrintDebugMessages=1
$ export NEOReadDebugKeys=1
$ sudo -E ZELLO_SYSMAN_USE_ZESINIT=1 ./zello_sysman -E
- The execution should wait for events (~10sec) or stop on an event.
Is this a regression?
- Yes, this is a regression - functionality that previously worked is now broken
Last Known Working Driver Version
No response
First Known Failing Driver Version
No response
API Call Logs
No response
strace Logs
No response
System Logs / dmesg Output
No response
Backtrace (if crash or hang occurred)
No response
Source Code / Reproducer
No response
Command Line / Application Details
No response
oneAPI Version (if applicable)
No response
Screenshots / Video
No response
Additional Notes
No response
- 主要语言
- C++
- 星标
- 1.5k
- 派生
- 303
- PR 合并指标
- 30 天内没有已合并 PR
环境准备
- 没有 Dockerfile 或 Docker Compose 文件
- 没有 Pull Request 模板
- 阅读贡献指南
从这里开始
- 先读完整个 Issue,再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库,在一个分支上完成修改。
- 提交 Pull Request,并在描述里引用这个 Issue 编号。
intel/compute-runtime 的其他 Issue
-
难度 2/5 1-3 小时 新手友好度 78/100
intel/compute-runtime#943 ·
-
难度 1/5 1 小时以内 新手友好度 1/100
intel/compute-runtime#1022 ·
-
难度 4/5 3-5 天 新手友好度 35/100
intel/compute-runtime#1019 ·
-
OS: Linux Type: Bug
难度 4/5 3-5 天 新手友好度 34/100
intel/compute-runtime#1017 ·
-
OS: Linux Type: Bug
难度 4/5 3-5 天 新手友好度 48/100
intel/compute-runtime#1016 · 3 条评论 ·
查看 intel/compute-runtime 的全部 Issue
相似的 Issue
-
难度 2/5 1-3 小时 新手友好度 70/100
FriendsofECCE/ECCE#336 · 1 条评论 ·
-
error not reported low priority semantics
难度 2/5 1-3 小时 新手友好度 70/100
维护者通常 1 天内回复
-
难度 2/5 1-3 小时 新手友好度 66/100
bitcoin-core/gui-qml#1030 ·
维护者通常 3 天内回复
-
难度 2/5 1-3 小时 新手友好度 80/100
-
难度 2/5 1-3 小时 新手友好度 82/100
维护者通常 1 天内回复