[AMD] Enable FA2 fwd on AMD MI300X (#1406)

* enable FA2 on AMD MI300X

* make lint happy

docker/Dockerfile.rocm

@@ -9,25 +9,43 @@ ENV DEBIAN_FRONTEND=noninteractive
 RUN apt-get update && apt-get install -y --no-install-recommends \
   build-essential git wget \
   libgtest-dev libprotobuf-dev protobuf-compiler libgflags-dev libsqlite3-dev llvm-dev \
+  rocm-dev rocm-libs hip-dev hipblas-dev rocblas-dev \
   && apt-get clean autoclean && rm -rf /var/lib/apt/lists/{cache,log} /tmp/* /var/tmp/*
 
 ENV PATH="/opt/conda/bin:${PATH}"
 ENV LIBGL_ALWAYS_INDIRECT=1
+ENV USE_ROCM=1
+ENV USE_CUDA=0
+ENV ROCM_HOME=/opt/rocm
+ENV HIP_PLATFORM=amd
+ENV PYTORCH_ROCM_ARCH="gfx90a;gfx942"
 
 
 RUN conda run -n py_3.10 conda install pip cmake -y && \
     conda run -n py_3.10 conda install -c conda-forge libstdcxx-ng=12 -y && \
     conda clean --all
 
-RUN apt-get install -y python3 python3-dev python3-setuptools gcc libtinfo-dev zlib1g-dev build-essential cmake libedit-dev libxml2-dev
+RUN apt-get update && apt-get install -y python3 python3-dev python3-setuptools gcc libtinfo-dev zlib1g-dev build-essential cmake libedit-dev libxml2-dev && \
+    apt-get clean autoclean && rm -rf /var/lib/apt/lists/{cache,log} /tmp/* /var/tmp/*
 
-RUN git clone https://github.com/tile-ai/tilelang.git --recursive -b main tilelang && \
-    mv /opt/conda/envs/py_3.10/compiler_compat /opt/conda/envs/py_3.10/compiler_compat.bak || true && \
-    conda run -n py_3.10 bash -c "pip install 'numpy<2.0' --force-reinstall && cd tilelang && USE_ROCM=1 pip install -e . -v"
+# Copy local tilelang directory instead of cloning from git
+# Build from tilelang root: docker build -f docker/Dockerfile.rocm -t mi300:latest .
+COPY . /root/tilelang
+
+RUN mv /opt/conda/envs/py_3.10/compiler_compat /opt/conda/envs/py_3.10/compiler_compat.bak || true && \
+    conda run -n py_3.10 bash -c "export USE_ROCM=1 USE_CUDA=0 && pip install 'numpy<2.0' --force-reinstall" && \
+    conda run -n py_3.10 bash -c "cd /root/tilelang && \
+        # Backup and modify pyproject.toml to remove torch from dependencies \
+        cp pyproject.toml pyproject.toml.bak && \
+        sed -i '/^[[:space:]]*\"torch/d' pyproject.toml && \
+        # Install tilelang with all dependencies except torch \
+        USE_ROCM=1 USE_CUDA=0 pip install -e . -v && \
+        # Restore original pyproject.toml \
+        mv pyproject.toml.bak pyproject.toml"
 
 RUN conda init bash && \
     echo "conda activate py_3.10" >> /root/.bashrc
 
 SHELL ["/bin/bash", "-l", "-c"]
 
-ENTRYPOINT ["/bin/bash", "--login", "-i"]
\ No newline at end of file
+ENTRYPOINT ["/bin/bash", "--login", "-i"]

examples/amd/example_amd_flash_attn_fwd.py

@@ -8,6 +8,21 @@ import argparse
 from functools import partial
 
 
+# Custom supply function to ensure tensors are created on GPU
+def supply_tensors_gpu(params):
+    """Supply function that creates tensors on GPU for ROCm/HIP."""
+    tensors = []
+    for param in params:
+        if hasattr(param, 'shape') and hasattr(param, 'dtype'):
+            # Force creation on GPU device
+            shape = [int(s) for s in param.shape]
+            tensor = torch.randn(shape, dtype=param.dtype, device='cuda')
+            tensors.append(tensor)
+        else:
+            tensors.append(param)
+    return tensors
+
+
 def ref_program(Q, K, V, is_causal, groups=1):
     assert Q.size(
         2) == K.size(2) * groups, f"Q heads {Q.size(2)} K heads {K.size(2)} groups {groups}"
@@ -63,7 +78,7 @@ def get_configs():
     return valid_configs
 
 
-@tilelang.autotune(configs=get_configs(), cache_input_tensors=True)
+@tilelang.autotune(configs=get_configs(), cache_input_tensors=True, supply_prog=supply_tensors_gpu)
 @tilelang.jit(out_idx=[3])
 def fast_flashattn(
     batch,

src/op/math.cc

@@ -33,6 +33,7 @@ TVM_REGISTER_OP("tl.pow_of_int")
     .set_attr<TCallEffectKind>("TCallEffectKind",
                                Integer(CallEffectKind::kPure))
     .set_attr<TScriptPrinterName>("TScriptPrinterName", "pow_of_int")
+    .set_attr<FLowerIntrinsic>("hip.FLowerIntrinsic", pow_of_int_op)
     .set_attr<FLowerIntrinsic>("cuda.FLowerIntrinsic", pow_of_int_op);
 
 PrimExpr infinity_op(PrimExpr args) {
@@ -59,7 +60,8 @@ TVM_REGISTER_OP("tl.infinity")
     .set_attr<TCallEffectKind>("TCallEffectKind",
                                Integer(CallEffectKind::kPure))
     .set_attr<TScriptPrinterName>("TScriptPrinterName", "infinity")
-    .set_attr<FLowerIntrinsic>("cuda.FLowerIntrinsic", infinity_op);
+    .set_attr<FLowerIntrinsic>("cuda.FLowerIntrinsic", infinity_op)
+    .set_attr<FLowerIntrinsic>("hip.FLowerIntrinsic", infinity_op);
 
 } // namespace tl
 } // namespace tvm

src/target/codegen_hip.cc

@@ -1190,9 +1190,9 @@ inline void PrintConst(const FloatImmNode *op, std::ostream &os,
       if (op->value < 0) {
         temp << "-";
       }
-      temp << ((op->dtype.bits() == 32) ? "HIPRT_INF_F" : "HIPRT_INF");
+      temp << ((op->dtype.bits() == 32) ? "HUGE_VALF" : "HUGE_VAL");
     } else if (std::isnan(op->value)) {
-      temp << ((op->dtype.bits() == 32) ? "HIPRT_NAN_F" : "HIPRT_NAN");
+      temp << ((op->dtype.bits() == 32) ? "NAN" : "NAN");
     } else {
       temp << std::scientific << op->value;
       if (op->dtype.bits() == 32)