commit 512bedcf62985d2de92253b43057d6e4c46d0c58
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Fri Sep 11 13:37:21 2026 +0200

dev-python/quack-kernels: add 0.6.5

Quack now requires CUTLASS DSL >=4.7, while its TVM FFI <0.2 bound
remains unchanged. The overlay's 4.7.1 release satisfies the new
floor.

Tests remain restricted because they compile and execute CUDA kernels.

commit 452a057bc17c904a97cdfcb170cbbc971ba22800
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Thu Sep 10 05:59:54 2026 +0200

dev-python/quack-kernels: trim 0.5.0 comment prose

commit 0dd54ed69540467851dfc8f61470d92502d81544
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Thu Sep 10 05:36:18 2026 +0200

dev-python/quack-kernels: trim 0.6.1 comment prose

commit 5bbe249ee33b523610cb32dc09a106fbe4622ec1
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Thu Sep 10 05:30:16 2026 +0200

dev-python/quack-kernels: trim 0.6.4 comment prose

commit a22edd060838dbd1bf78fbb88c136aedf023e6e0
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Thu Sep 10 05:19:39 2026 +0200

dev-python/quack-kernels: trim 0.6.3 comment prose

commit 43d904215839cb3fe25c1f2eff4715e1f0b8602d
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sun Aug 23 23:17:13 2026 +0200

dev-python/quack-kernels: drop 0.6.2

commit 46ff5e87d8626b67fe9dd1cf666d567a94f8c2f6
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sun Aug 23 13:14:46 2026 +0200

dev-python/quack-kernels: depend on virtual/triton

commit 5f2d33e213eeabf0ec52f63961b3a84252627a45
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sat Aug 8 09:16:34 2026 +0200

dev-python/quack-kernels: add 0.6.4

0.6.4 pins nvidia-cutlass-dsl==4.6.2 (up from 4.6.1 in 0.6.2/0.6.3) and widens
the apache-tvm-ffi range to >=0.1.6,<0.2. Pin ~nvidia-cutlass-dsl-4.6.2 to match
the version the JIT kernels are ABI-locked to; needs the cutlass-dsl 4.6.2
cluster added alongside.

commit 8d816d8e7a8a37a6cfb88b942c879571b1d2f2cb
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Fri Aug 7 12:32:06 2026 +0200

dev-python/quack-kernels: restore 0.6.1

quack-kernels 0.6.2 raised its nvidia-cutlass-dsl pin from ~4.6.0 to ~4.6.1,
matching the newer JIT ABI its kernels are locked to. vllm-0.26.0 pins
~nvidia-cutlass-dsl-4.6.0 to match upstream requirements/cuda.txt and floors
quack-kernels uncapped at >=0.4.0, so on a vllm host -uDN tries to pull
quack-kernels-0.6.3, dragging in cutlass-dsl 4.6.1 and colliding with vllm's
~4.6.0. Dropping 0.6.1 removed the only 0.6-series version that pins ~4.6.0,
leaving the installed 0.6.1 with no backing ebuild. 0.6.1 remains the last
quack-kernels compatible with the cutlass-4.6.0 that the frozen vllm stack
requires; keep it so that stack stays resolvable and rollback-safe.

commit 9d8bb42dd7e37b8c710b9601f24c84359890b57f
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Thu Aug 6 18:21:06 2026 +0200

dev-python/quack-kernels: drop 0.6.1

commit 6a2b8b7209e9074fec90ec80ca0ef61505fbe54f
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Thu Aug 6 18:20:40 2026 +0200

dev-python/quack-kernels: add 0.6.3

commit b74337f5197d877d40d9f61617cc2370fb5228b6
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Wed Aug 5 13:29:18 2026 +0200

dev-python/quack-kernels: drop 0.6.0

commit 3d73f29f76e55f93320b4a05ee3ce6a00962d190
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Wed Aug 5 13:29:12 2026 +0200

dev-python/quack-kernels: add 0.6.2

0.6.2 ABI-locks to nvidia-cutlass-dsl==4.6.1 (was 4.6.0); update the pin. numpy,
cuda-bindings and triton stay declared (imported despite upstream under-declaring
them).

commit dd62845ba6708723b4b845da7fd2fede4f1225cc
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Tue Aug 4 16:56:40 2026 +0200

dev-python/quack-kernels: fix dependencies in 0.6.1

commit 917ba4bfbc14021366473f363a39758108bac6fe
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Tue Aug 4 16:53:30 2026 +0200

dev-python/quack-kernels: fix dependencies in 0.6.0

commit 1c33cc76b9a330df09a9cca9f737c4f711ccc69e
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Tue Aug 4 16:51:23 2026 +0200

dev-python/quack-kernels: fix dependencies in 0.5.0

commit ba7a586a7c69849b1cb96a86ecc8b09f425c66e6
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Wed Jul 22 17:31:11 2026 +0200

dev-python/quack-kernels: restore 0.5.0

quack-kernels 0.6+ raised its nvidia-cutlass-dsl floor to >=4.6.0, which is
incompatible with the ~nvidia-cutlass-dsl-4.5.2 pin that vllm-0.25.1 carries to match
upstream requirements/cuda.txt. Dropping 0.5.0 left vllm's uncapped quack-kernels
dependency with no satisfiable version, aborting any @world resolution. 0.5.0 uses
>=nvidia-cutlass-dsl-4.5.2, so it remains installable alongside the pinned 4.5.2 and
is the last version compatible with the frozen cuda stack.

commit 941a362fa660e934619353651c1aa6ddce9ea8ee
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sun Jul 5 21:19:13 2026 +0200

dev-python/quack-kernels: drop 0.5.0

commit 4d11c3656b4bd272075e5e056f75da166db7da1a
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sun Jul 5 21:19:13 2026 +0200

dev-python/quack-kernels: add 0.6.1

commit c8276991087c43b54b527b7d3c57461da2a80e83
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sat Jul 4 18:43:27 2026 +0200

dev-python/quack-kernels: drop 0.4.1

commit b0bcd61b939311aa9ada467bc1f665f239d2ff30
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sat Jul 4 18:42:42 2026 +0200

dev-python/quack-kernels: add 0.6.0

Upstream 0.6.0 tightens its cute-DSL requirement to
nvidia-cutlass-dsl==4.6.0 (the JIT'd kernels are ABI-locked to the DSL
version), so the RDEPEND floor moves from 4.5.2 to 4.6.0 to match. vLLM
floor-pins quack-kernels>=0.3.3, so 0.6.0 satisfies every consumer.

commit b2d6a0447f0a118a3b74c4e004f3082645637e5e
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sat Jun 6 21:48:05 2026 +0200

dev-python/quack-kernels: drop 0.3.3

commit 0e21a503926f003a6b2646e7b57279b3c7ae3e65
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Fri May 29 12:04:16 2026 +0200

dev-python/quack-kernels: add 0.5.0

Track upstream's raised nvidia-cutlass-dsl floor (>=4.5.2).

commit f0feab8c9c0c499aeb19fb644d8b3d5d0cf5318f
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Wed May 13 14:35:28 2026 +0200

dev-python/quack-kernels: disable py3.11

commit 2136ab74575dae3399cde0adab1e4e5302491330
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sun May 10 15:24:45 2026 +0200

dev-python/quack-kernels: switch to DISTUTILS_SINGLE_IMPL

sci-ml/pytorch and the now-single-impl dev-python/torch-c-dlpack-ext
need $; the multi-impl helpers (cutlass-dsl,
apache-tvm-ffi, einops) move into python_gen_cond_dep.

commit 9b3f2a353ad27a8206075225b72edd97910c0390
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sun May 10 16:51:11 2026 +0200

dev-python/quack-kernels: switch to DISTUTILS_SINGLE_IMPL

sci-ml/pytorch and the now-single-impl dev-python/torch-c-dlpack-ext
need $; the multi-impl helpers (cutlass-dsl,
apache-tvm-ffi, einops) move into python_gen_cond_dep.

(Hand-port of the same conversion landing on 0.3.3 to our newer
0.4.1 ebuild.)

commit c4956a77759b043d3ae94a8b62afcbb4ed658d0c
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sun May 10 14:52:57 2026 +0200

dev-python/quack-kernels: add 0.4.1

commit 50dc7eccdb67c47628ae42212108f2e6678b22d1
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Thu May 7 18:28:27 2026 +0200

dev-python/quack-kernels: new package, 0.3.3

Tier 4 of the vllm CUDA target packaging cycle. QuACK attention
kernels — pure-Python wrappers that JIT-compile attention/elementwise
kernels through nvidia-cutlass-dsl + apache-tvm-ffi at first use.

All deps are already in this overlay's tracked stack. Upstream's
cu13 extra is folded into the unconditional nvidia-cutlass-dsl
dep — our nvidia-cutlass-dsl always pulls the cu13 path on this
amd64+CUDA-13 overlay, so no USE flag is needed.