stable-diffusion-webui icon indicating copy to clipboard operation
stable-diffusion-webui copied to clipboard

[Bug]: AMD GPUs not running on with updated

Open seanthenry opened this issue 2 years ago • 7 comments

Is there an existing issue for this?

  • [X] I have searched the existing issues and checked the recent builds/commits

What happened?

When I try to generate an image I get the error below. It is related to the pytorch version as I can get it to work if I roll back the version.

MIOpen(HIP): Error [Compile] 'hiprtcCompileProgram(prog.get(), c_options.size(), c_options.data())' naive_conv.cpp: HIPRTC_ERROR_COMPILATION (6) MIOpen(HIP): Error [BuildHip] HIPRTC status = HIPRTC_ERROR_COMPILATION (6), source file: naive_conv.cpp MIOpen(HIP): Warning [BuildHip] /tmp/comgr-bafe6d/input/naive_conv.cpp:39:10: fatal error: 'limits' file not found #include // std::numeric_limits ^~~~~~~~ 1 error generated when compiling for gfx1030. terminate called after throwing an instance of 'miopen::Exception' what(): /long_pathname_so_that_rpms_can_package_the_debug_info/data/driver/MLOpen/src/hipoc/hipoc_program.cpp:304: Code object build failed. Source: naive_conv.cpp Aborted (core dumped)

Steps to reproduce the problem

Try to generate an image

What should have happened?

image is generated

Commit where the problem happens

1.3.2

What Python version are you running on ?

Python 3.10.x

What platforms do you use to access the UI ?

Linux

What device are you running WebUI on?

AMD GPUs (RX 6000 above)

What browsers do you use to access the UI ?

Google Chrome

Command Line Arguments

No

List of extensions

No

Console logs

MIOpen(HIP): Error [Compile] 'hiprtcCompileProgram(prog.get(), c_options.size(), c_options.data())' naive_conv.cpp: HIPRTC_ERROR_COMPILATION (6)
MIOpen(HIP): Error [BuildHip] HIPRTC status = HIPRTC_ERROR_COMPILATION (6), source file: naive_conv.cpp
MIOpen(HIP): Warning [BuildHip] /tmp/comgr-bafe6d/input/naive_conv.cpp:39:10: fatal error: 'limits' file not found
#include <limits> // std::numeric_limits
         ^~~~~~~~
1 error generated when compiling for gfx1030.
terminate called after throwing an instance of 'miopen::Exception'
  what():  /long_pathname_so_that_rpms_can_package_the_debug_info/data/driver/MLOpen/src/hipoc/hipoc_program.cpp:304: Code object build failed. Source: naive_conv.cpp
Aborted (core dumped)

Additional information

If I modify webui.sh to install the older version of pytorch I can get it to run

i``` f echo "$gpu_info" | grep -q "AMD" && [[ -z "${TORCH_COMMAND}" ]] then # AMD users will still use torch 1.13 because 2.0 does not seem to work.

export TORCH_COMMAND="pip install torch==2.0.1+rocm5.4.2 torchvision==0.15.2+rocm5.4.2 --index-url https://download.pytorch.org/whl/rocm5.4.2"

export TORCH_COMMAND="pip install torch==1.13.1+rocm5.2 torchvision==0.14.1+rocm5.2 --index-url https://download.pytorch.org/whl/rocm5.2"

GPU 
Name:                    gfx1030
Marketing Name:          AMD Radeon RX 6800

seanthenry avatar Jun 09 '23 15:06 seanthenry

Same issue. Running a 6700xt.

jmcconn18 avatar Jun 09 '23 22:06 jmcconn18

Same problem on an RX 6600, this appears to be an upstream issue in ROCm https://github.com/RadeonOpenCompute/ROCm/issues/1889.

> clinfo | rg gfx
  Device Name                                     gfx1032
    Device Name                                   gfx1032
    Device Name                                   gfx1032
    Device Name                                   gfx1032

LowYieldFire avatar Jun 14 '23 16:06 LowYieldFire

Same problem on RX 7900XTX. To even get my GPU noticed i have to run this script:

git clone --branch dev https://github.com/automatic1111/stable-diffusion-webui cd stable-diffusion-webui python3 -m venv venv . venv/bin/activate export HSA_OVERRIDE_GFX_VERSION=11.0.0 pip3 install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/rocm5.5 ./webui.sh --no-half

No errors occur during setup. I get the same error as the OP when trying to generate an image.

Linux Mint 21.2 Kernel 5.15.0-77-generic AMD Ryzen 7 7800X3D 32 GB RAM AMD RX 7900XTX ROCm 5.6.0

Daedalus1400 avatar Jul 13 '23 16:07 Daedalus1400

I think I know how to correct the issue. I found a similar post on pytorch forum.

There is a missing app run the following and let me know if it fixes it.

sudo apt install libstdc++-12-dev

seanthenry avatar Jul 15 '23 16:07 seanthenry

I think I know how to correct the issue. I found a similar post on pytorch forum.

There is a missing app run the following and let me know if it fixes it.

sudo apt install libstdc++-12-dev

I also hit this error, and that fix worked for me. Thanks!

Craig-Macomber avatar Jul 22 '23 19:07 Craig-Macomber

I think I know how to correct the issue. I found a similar post on pytorch forum.

There is a missing app run the following and let me know if it fixes it.

sudo apt install libstdc++-12-dev

Also worked for me, much appreciated

Gecked-Deck avatar Aug 08 '23 21:08 Gecked-Deck

On Ryzen 7 7700 the apt command didn't fis the issue

HSA_OVERRIDE_GFX_VERSION=10.3.0 ./webui.sh 

################################################################
Install script for stable-diffusion + Web UI
Tested on Debian 11 (Bullseye)
################################################################

################################################################
Running on g user
################################################################

################################################################
Repo already cloned, using it as install directory
################################################################

################################################################
Create and activate python venv
################################################################

################################################################
Launching launch.py...
################################################################
Using TCMalloc: libtcmalloc_minimal.so.4
Python 3.11.2 (main, Mar 13 2023, 12:18:29) [GCC 12.2.0]
Version: v1.5.1
Commit hash: 68f336bd994bed5442ad95bad6b6ad5564a5409a
Launching Web UI with arguments: --no-half --no-half-vae --skip-torch-cuda-test --listen
no module 'xformers'. Processing without...
no module 'xformers'. Processing without...
No module 'xformers'. Proceeding without it.
Loading weights [c5baa61ff4] from /home/g/Code/stable-diffusion-webui/stable-diffusion-webui/models/Stable-diffusion/sdxl/sd_xl_base_1.0.safetensors
MIOpen(HIP): Error [Compile] 'hiprtcCompileProgram(prog.get(), c_options.size(), c_options.data())' naive_conv.cpp: HIPRTC_ERROR_COMPILATION (6)
MIOpen(HIP): Error [BuildHip] HIPRTC status = HIPRTC_ERROR_COMPILATION (6), source file: naive_conv.cpp
MIOpen(HIP): Warning [BuildHip] /tmp/comgr-499674/input/naive_conv.cpp:39:10: fatal error: 'limits' file not found
#include <limits> // std::numeric_limits
         ^~~~~~~~
1 error generated when compiling for gfx1030.
terminate called after throwing an instance of 'miopen::Exception'
  what():  /long_pathname_so_that_rpms_can_package_the_debug_info/data/driver/MLOpen/src/hipoc/hipoc_program.cpp:304: Code object build failed. Source: naive_conv.cpp
./webui.sh: line 254: 780665 Aborted                 "${python_cmd}" "${LAUNCH_SCRIPT}" "$@"

grigio avatar Aug 18 '23 09:08 grigio