aboutsummaryrefslogtreecommitdiff
path: root/libc/src/__support/GPU
AgeCommit message (Expand)AuthorFilesLines
2025-10-10[libc] Add a config option to disable slab reclaiming (#151599)Joseph Huber1-4/+8
2025-08-11[libc] Simplifiy slab waiting in GPU memory allocator (#152872)Joseph Huber1-12/+12
2025-08-08[libc] Cache old slabs when allocating GPU memory (#151866)Joseph Huber3-2/+154
2025-07-31[libc] Fix leader calculation when done in wave64 modeJoseph Huber1-1/+1
2025-07-30[libc] Start slab search at number of allocated bitsJoseph Huber1-36/+34
2025-07-29[libc] Fix incorrect count when initializing slabJoseph Huber1-3/+5
2025-07-28[libc] Reduce reference counter to a 32-bit integer (#150961)Joseph Huber1-23/+27
2025-07-28[libc] Increase the number of times we wait on a slabJoseph Huber1-1/+1
2025-07-28[libc] Rework match any use in hot allocate bitfield loopJoseph Huber1-12/+27
2025-07-28[libc] Small performance improvements to GPU allocatorJoseph Huber1-12/+16
2025-07-28[libc] Improve starting indices for GPU allocation (#150432)Joseph Huber1-6/+13
2025-07-23[libc][NFC] Add stdint.h proxy header to fix dependency issue with <stdint.h>...lntue2-1/+2
2025-07-23[libc] Search empty bits after failed allocation (#149910)Joseph Huber1-3/+13
2025-07-23[libc] Cache the most recently used slot for a chunk size (#149751)Joseph Huber1-59/+63
2025-07-02[libc] Fix internal alignment in allcoator (#146738)Joseph Huber1-3/+5
2025-07-02[libc] Efficiently implement `aligned_alloc` for AMDGPU (#146585)Joseph Huber2-0/+28
2025-07-02[libc] Use is aligned builtin instead of ptrtoint (#146402)Joseph Huber1-2/+2
2025-06-30[libc] Efficiently implement 'realloc' for AMDGPU devices (#145960)Joseph Huber2-0/+23
2025-06-27[libc] Add and use 'cpp::launder' to guard placement new (#146123)Joseph Huber1-2/+2
2025-06-11[libc] Perform bitfield zero initialization wave-parallel (#143607)Joseph Huber1-11/+35
2025-06-11[libc][NFC] Remove template from GPU allocator reference counterJoseph Huber1-16/+16
2025-06-04[libc] Coalesce bitfield access in GPU malloc (#142692)Joseph Huber1-12/+22
2025-05-28[libc] Implement efficient 'malloc' on the GPU (#140156)Joseph Huber2-7/+483
2025-05-13[libc][NFC] Add u64 shuffle helpers to GPUJoseph Huber1-0/+12
2025-02-17[Clang] Add handlers for 'match_any' and 'match_all' to `gpuintrin.h` (#127504)Joseph Huber1-0/+8
2025-02-05[Clang] Add width handling for <gpuintrin.h> shuffle helper (#125896)Joseph Huber1-2/+3
2025-01-07[libc] Switch to using the generic `<gpuintrin.h>` implementations (#121810)Joseph Huber8-479/+87
2024-12-02[libc][NFC] Rename RPC opcodes to better reflect their usageJoseph Huber1-2/+2
2024-10-15[libc] Remove dependency on `cpp::function` in `rpc.h` (#112422)Joseph Huber1-5/+6
2024-07-12[libc] Migrate to using LIBC_NAMESPACE_DECL for namespace declaration (#98597)Petr Hosek6-12/+17
2024-07-12Revert "[libc] Migrate to using LIBC_NAMESPACE_DECL for namespace declaration...Mehdi Amini6-17/+12
2024-07-11[libc] Migrate to using LIBC_NAMESPACE_DECL for namespace declaration (#98075)Petr Hosek6-12/+17
2024-05-10[libc] Add memory fence utility to the GPU utilities (#91756)Joseph Huber2-0/+9
2024-03-12[libc] Use __builtin_ffsll for RPC lane mask (#85000)Petr Hosek1-1/+1
2024-03-12[libc] Add utility functions for warp-level scan and reduction (#84866)Joseph Huber4-0/+35
2024-03-12[libc] Fix lane-id utility function not using built-in (#84902)Joseph Huber1-1/+1
2024-03-10[libc][NFC] Move GPU allocator implementation to common header (#84690)Joseph Huber3-0/+80
2024-03-04[libc][NFC] Clean up GPU utilities directory after rework (#83692)Joseph Huber1-5/+3
2024-02-28[libc] fix clang-tidy llvm-header-guard warnings (#82679)Nick Desaulniers2-6/+6
2024-02-22[libc] Rework the GPU build to be a regular target (#81921)Joseph Huber1-1/+1
2024-02-13[libc] Remove leftover target dependent intrinsicJoseph Huber1-8/+0
2024-02-13[libc] Remove remaining GPU architecture dependent instructions (#81612)Joseph Huber2-12/+3
2024-02-13[libc] Rework the RPC interface to accept runtime wave sizes (#80914)Joseph Huber3-13/+8
2024-02-06[libc] Remove CPU dependent AMDGPU instructions (#80707)Joseph Huber1-18/+4
2024-01-30[libc] Change the starting port index to use the SMID (#79200)Joseph Huber3-0/+14
2024-01-29[libc][NFC] Remove inline assembly for PTX instructions (#79913)Joseph Huber1-13/+4
2023-09-26[libc] Mass replace enclosing namespace (#67032)Guillaume Chatelet4-8/+8
2023-09-21[libc][clang-tidy] Add llvm-header-guard to get consistant naming and prevent...Guillaume Chatelet4-9/+9
2023-09-12[libc][NFC] Factor GPU exiting into a common function (#66093)Joseph Huber3-0/+11
2023-08-31[libc] Support 'assert.h' on the GPUJoseph Huber1-0/+15