-
### What happened?
I am seeing memory leaks due to the ROCm & HIP runtime backends in IREE. These issues can be reproduced simply by initializing and releasing a hip / rocm HAL device:
```
==168558…
-
### Problem Description
Hi, I am testing HIP on Intel CPU workstation with Nvidia GPU.
I could build hip/clr as shown in the AMD documentation (after running dos2unix for all sources) but testing sq…
-
apt-get install hip-runtime-nvidia hip-dev
When I run this command on my terminal I was getting following error message. Can someone help me fixing this issue ??
Reading package lists... Done
B…
-
### Description of errors
Minor Docs typo, looks like hipErrorInvalidDevicePointer got swapped to hipErrorNotFound at some point. Makes debugging a bit confusing...
Link to function in CLR:
https…
-
### Problem Description
OS:
NAME="Ubuntu"
VERSION="22.04.3 LTS (Jammy Jellyfish)"
CPU:
model name : AMD Ryzen 7 5800X3D 8-Core Processor
GPU:
Marketing Name: Radeon RX 7900 XTX …
-
Trying to run any app which uses HIP API with TSAN triggers a "thread leak" error at the end:
```console
$ hipcc tsan.cpp -g -fsanitize=thread -o tsan && ./tsan
clang-15: warning: ignoring '-fsan…
-
As we pull latest bits/dependencies into tests, our CI is prone to break automatically. For example, some recent ROCm deprecation seems to cause warnings (turned into errors). This may be wanted but i…
-
### Problem Description
I'm currently trying to port an NVIDIA code to ROCM/HIP using amdclang++
While compiling I've noticed that I'm getting some errors caused by extra space in some hip runtime i…
-
Right now we are seeing missing symbols and wrong validation regarding `hipDrvGraphAddMemcpyNode`. Need to have a build fixing those issues. And then enable SDXL on it to see how it performs.
-
### Problem Description
I'm trying to get started vllm on ROCM. I built and ran the docker image by following the instructions on this page https://docs.vllm.ai/en/latest/getting_started/amd-installa…