Skip to content

Actions: Xilinx/llvm-aie

PR Subscriber

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
85 workflow runs
85 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

[LLVM] Optimize G_SHUFFLE_VECTOR into more efficient generic opcodes
PR Subscriber #10: Pull request #41 labeled by ValentijnvdBeek
May 17, 2024 15:43 2s
May 17, 2024 15:43 2s
[AIE2] Implement simple (and inefficient) Shuffle Vector legalization
PR Subscriber #9: Pull request #40 labeled by ValentijnvdBeek
May 17, 2024 15:31 2s
May 17, 2024 15:31 2s
[AIE2] Implement simple (and inefficient) Shuffle Vector legalization
PR Subscriber #8: Pull request #40 labeled by ValentijnvdBeek
May 17, 2024 15:31 2s
May 17, 2024 15:31 2s
[AIE2] Implement simple (and inefficient) Shuffle Vector legalization
PR Subscriber #7: Pull request #40 labeled by ValentijnvdBeek
May 17, 2024 15:31 2s
May 17, 2024 15:31 2s
Support G_EXTRACT_VALUES for 16, 32 and 128-bit vectors
PR Subscriber #6: Pull request #39 labeled by ValentijnvdBeek
May 17, 2024 15:05 2s
May 17, 2024 15:05 2s
Support G_EXTRACT_VALUES for 16, 32 and 128-bit vectors
PR Subscriber #5: Pull request #39 labeled by ValentijnvdBeek
May 17, 2024 15:05 2s
May 17, 2024 15:05 2s
Support G_EXTRACT_VALUES for 16, 32 and 128-bit vectors
PR Subscriber #4: Pull request #39 labeled by ValentijnvdBeek
May 17, 2024 15:05 2s
May 17, 2024 15:05 2s
G_BUILD_VECTOR & G_UNMERGE_VALUES legalization for 32-bit vectors
PR Subscriber #3: Pull request #38 labeled by ValentijnvdBeek
May 17, 2024 14:59 3s
May 17, 2024 14:59 3s
G_BUILD_VECTOR & G_UNMERGE_VALUES legalization for 32-bit vectors
PR Subscriber #2: Pull request #38 labeled by ValentijnvdBeek
May 17, 2024 14:59 3s
May 17, 2024 14:59 3s
G_BUILD_VECTOR & G_UNMERGE_VALUES legalization for 32-bit vectors
PR Subscriber #1: Pull request #38 labeled by ValentijnvdBeek
May 17, 2024 14:59 3s
May 17, 2024 14:59 3s