disable sm100 for FlashMLA and fast-hadamard-transform in cuda12.6.1 (#11274)

docker/Dockerfile

@@ -6,7 +6,7 @@ ARG BUILD_TYPE=all
 ARG BRANCH_TYPE=remote
 ARG DEEPEP_COMMIT=9af0e0d0e74f3577af1979c9b9e1ac2cad0104ee
 ARG FLASHMLA_COMMIT=1408756a88e52a25196b759eaf8db89d2b51b5a1
-ARG FAST_HADAMARD_TRANSFORM_COMMIT=f3cdeed95b0f3284b5df3da9b3311d3d0600ce2b
+ARG FAST_HADAMARD_TRANSFORM_COMMIT=7fd811c2b47f63b0b08d2582619f939e14dad77c
 ARG CMAKE_BUILD_PARALLEL_LEVEL=2
 ARG SGL_KERNEL_VERSION=0.3.12
 ENV DEBIAN_FRONTEND=noninteractive \
@@ -145,6 +145,9 @@ RUN if [ "$TARGETARCH" = "amd64" ]; then \
       cd flash-mla && \
       git checkout ${FLASHMLA_COMMIT} && \
       git submodule update --init --recursive && \
+      if [ "$CUDA_VERSION" = "12.6.1" ]; then \
+        export FLASH_MLA_DISABLE_SM100=1; \
+      fi && \
       pip install -v . ; \
     fi