derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
51 lines
1.4 KiB
Markdown
51 lines
1.4 KiB
Markdown
# Adding a new GPU
|
|
|
|
This guide is not complete.
|
|
|
|
## Marking NPI tasks
|
|
|
|
Mark tasks that need to be done when introducing a new product with `\NPI` comment.
|
|
|
|
Single-line markers use any comment leader (e.g. `//`):
|
|
|
|
```c
|
|
// \NPI a one-line todo
|
|
```
|
|
|
|
Multi-line markers use a `/* */` block comment, ending each continued line with a
|
|
trailing `\` (a `//` comment ending in `\` is a line-continuation that GCC rejects
|
|
under `-Werror=comment`):
|
|
|
|
```c
|
|
/*
|
|
* \NPI I am a big complicated todo \
|
|
* that takes multiple lines to describe.
|
|
*/
|
|
```
|
|
|
|
`scripts/find-npi-tasks.sh` strips the comment leaders and joins the continued
|
|
lines into a single task.
|
|
|
|
## Find npi tasks
|
|
|
|
Use `scripts/find-npi-tasks.sh` to find tasks.
|
|
|
|
## Extra Steps
|
|
|
|
What the grep can't point at are the "create a new thing" steps:
|
|
|
|
- Construct a `configs/<gpu>.json` (plus the `_kmd` and any multi-GPU variant)
|
|
for the new device.
|
|
- For a brand-new ISA family: sync `shared/machine-readable-isa` with
|
|
`download.py`, regenerate the ISA/DBT sources per [codegen.md](codegen.md),
|
|
and author the hand-written per-arch files (`isa.h`, `insts.h`, `mma_exec.h`,
|
|
`addr_calc.h/.cpp`, ...) under
|
|
`lib/rocjitsu/src/rocjitsu/isa/arch/amdgpu/<isa>/`.
|
|
- Define its [ISA target provider](isa-target-providers.md) and select it in
|
|
each executable or shared object that should contain it.
|
|
- Add HIP kernel coverage under `tests/kernels/`.
|
|
|
|
## Reminder
|
|
|
|
Don't leak new products.
|