486 Коммитов

Все ветки

Поиск

Автор	SHA1	Сообщение	Дата
ghostplant	0a6df04423	fix parsing code name for MI200/300	2023-09-05 15:29:49 +08:00
ghostplant	9fc5dcec3f	fallback using gcc atomic	2023-08-29 07:04:56 +00:00
ghostplant	daaa0204d7	update rocm dockerfiles	2023-08-10 18:33:32 +08:00
ghostplant	86317b0350	update container env for torch2	2023-08-09 17:02:29 +08:00
ghostplant	0246d68184	update compute compate list	2023-08-04 06:13:08 +08:00
ghostplant	2215a0f53e	add get_hash as builtin operation	2023-07-23 13:25:19 +00:00
Ziming Miao	1e76ea3152	support msvc build (#369)	2023-07-17 22:14:57 +08:00
ghostplant	e576e86ec3	input calibration in numpy format	2023-07-17 15:30:50 +08:00
ghostplant	f50649c15f	update GPU policy for non GPU devices	2023-07-17 12:20:35 +08:00
ghostplant	4ab9d09d59	update sycl shared memory parsing	2023-07-13 17:53:02 +08:00
ghostplant	d9adfbb604	add more atomic defines	2023-07-06 23:54:22 +08:00
ghostplant	6936e1b8bf	add compile support for xbox	2023-07-06 01:23:44 +08:00
ghostplant	1f0e700e33	parse CODE_DEBUG option when saving object	2023-07-03 10:38:05 +00:00
ghostplant	d592a86c6b	add custom defines & atomic macro	2023-07-01 15:32:15 +08:00
ghostplant	28218cbbe5	handle codegen method for __builtin_cond	2023-06-18 16:40:18 +08:00
ghostplant	142b419f64	parse LDS defined from subfunction	2023-06-17 15:10:41 +08:00
ghostplant	64b4644370	add namespace to call init function	2023-06-16 13:51:09 +08:00
Wei CUI	4100774b73	use uint as boolean buffer type	2023-06-14 17:43:04 +08:00
ghostplant	3abbbf2358	deal with persistent device context	2023-06-13 23:40:28 +08:00
ghostplant	c3762eaced	using newline before appending codes	2023-05-08 13:32:01 +08:00
ghostplant	5df9a1d2c5	avoid reconnecting socket due to multi-init	2023-04-19 00:00:27 +08:00
Wei CUI	458acf2b39	add d2d entry of HLSL backend	2023-04-18 13:07:27 +08:00
Wei CUI	edfec644b7	add custom renaming of kernel function	2023-04-15 23:52:06 +08:00
ghostplant	7880d49af2	support uploading file for filemgr.py	2023-04-11 03:31:31 +08:00
ghostplant	6880cab5ca	add filemgr.py	2023-04-09 08:29:45 +08:00
Wei CUI	e3650e6bd9	update binder to support upload	2023-04-09 08:23:45 +08:00
ghostplant	30308c6db7	aligning memsize for GPU memalloc	2023-04-07 23:41:23 +08:00
ghostplant	3df7c2f0d2	add -ldl to cuda	2023-04-07 20:15:39 +08:00
ghostplant	60aa00bc48	fix replacement of scatter IR	2023-04-02 15:48:58 +08:00
ghostplant	65a50439d2	fp16 x int32 -> fp16	2023-03-31 15:08:38 +08:00
Wei CUI	fe3f3e39ba	allow xbox client to work in standalone mode	2023-03-29 20:59:43 +08:00
ghostplant	c95ca32481	refactor for refresh codegen	2023-03-06 15:43:22 +08:00
ghostplant	450b3c2f56	define uint32_t to avoid undefined types	2023-02-26 13:58:11 +00:00
ghostplant	5b8ec8b63f	make moduleLoad compatible for src & bin	2023-02-24 19:05:42 +08:00
ghostplant	b3d4d3e182	set flags for dependency checking	2023-02-17 23:46:13 +08:00
ghostplant	1a102320bf	fix template typename sycl atomics	2023-02-13 00:16:30 +08:00
ghostplant	8d761ae973	less strict condition dtype	2023-02-10 23:08:47 +08:00
Wei CUI	6f7a1a73dc	handle the difference of memory management	2023-02-10 13:38:08 +08:00
Wei CUI	39b0ccf50e	deal with flags for remote backend	2023-02-10 00:20:15 +08:00
ghostplant	7f809d4e12	update kernel parsing	2023-02-09 23:15:54 +08:00
ghostplant	af839a1dfc	add binder script	2023-02-09 22:45:17 +08:00
ghostplant	5232f3b303	fix hlsl atomic	2023-02-09 21:21:53 +08:00
ghostplant	0bad75f7a2	fix aggregation max/min	2023-02-09 15:11:13 +08:00
ghostplant	d3dce9220c	synchronize common atomic funcs	2023-02-09 00:28:01 +08:00
ghostplant	ca07421e4e	sync to sycl-2023	2023-02-07 00:08:19 +08:00
Wei CUI	433f4c7c3e	add flags to distinguish xbox model	2023-02-01 23:06:19 +08:00
Wei CUI	30ed0dd261	apply ATOMIC_ADD symbol	2023-01-31 20:32:49 +08:00
Wei CUI	b347b3a562	upgrade antares_hlsl.dll to v0.3.4	2023-01-31 15:55:10 +08:00
Wei CUI	43938402b8	dxwarpper support fatbin compilation	2023-01-31 15:38:20 +08:00
Wei CUI	9361419404	add template flags / recommend mode + compat	2023-01-30 17:54:57 +08:00