[GitHub] [tvm] tvm-bot commented on pull request #13389: [TECompiler] Remove static qualifer in constant index

tvm-bot commented on PR #13389: URL: https://github.com/apache/tvm/pull/13389#issuecomment-1314909082 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] masahi opened a new pull request, #13389: [TECompiler] Remove static qualifer in constant index

masahi opened a new pull request, #13389: URL: https://github.com/apache/tvm/pull/13389 Trying to see if we can pass CI with this change. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the

[GitHub] [tvm] echuraev commented on a diff in pull request #13362: [OpenCL] Introduce OpenCL wrapper to TVM

echuraev commented on code in PR #13362: URL: https://github.com/apache/tvm/pull/13362#discussion_r1022429074 ## src/runtime/opencl/opencl_wrapper/opencl_wrapper.cc: ## @@ -0,0 +1,551 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor

[GitHub] [tvm] masahi closed issue #13387: [Bug] some bugs in tvm.

masahi closed issue #13387: [Bug] some bugs in tvm. URL: https://github.com/apache/tvm/issues/13387 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm] masahi commented on issue #13387: [Bug] some bugs in tvm.

masahi commented on issue #13387: URL: https://github.com/apache/tvm/issues/13387#issuecomment-1314899181 You are welcome to send a PR if you find a bug. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above

[tvm] branch fix_gpu_other_build created (now 8822cd5aa0)

This is an automated email from the ASF dual-hosted git repository. driazati pushed a change to branch fix_gpu_other_build in repository https://gitbox.apache.org/repos/asf/tvm.git at 8822cd5aa0 fix rocm build bug No new revisions were added by this update.

[GitHub] [tvm] tvm-bot commented on pull request #13388: [Hexagon] Use VTCM while scheduling conv2d op

tvm-bot commented on PR #13388: URL: https://github.com/apache/tvm/pull/13388#issuecomment-1314820841 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] abhikran-quic opened a new pull request, #13388: [Hexagon] Use VTCM while scheduling conv2d op

abhikran-quic opened a new pull request, #13388: URL: https://github.com/apache/tvm/pull/13388 cc: @csullivan @masahi -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To

[GitHub] [tvm] honghuichao opened a new issue, #13387: [Bug] some bugs in tvm.

honghuichao opened a new issue, #13387: URL: https://github.com/apache/tvm/issues/13387 I tested the tvm and found some bugs, as follows： （1）in mxnet frontend. def _mx_pad(inputs, attrs): pad_mode = attrs.get_str("mode", None) if pad_mode is None: raise

[GitHub] [tvm] tvm-bot commented on pull request #13386: feat: fix pylint issues for conftest

tvm-bot commented on PR #13386: URL: https://github.com/apache/tvm/pull/13386#issuecomment-1314807250 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] qzylalala opened a new pull request, #13386: feat: fix pylint issues for conftest

qzylalala opened a new pull request, #13386: URL: https://github.com/apache/tvm/pull/13386 This commit fixes pylint issues with conftest. tests reported in https://github.com/apache/tvm/issues/11414 cc: @Mousius @areusch for code review. -- This is an automated message from the

[tvm] branch main updated: [MetaSchedule] Add `from-target` Defaults for LLVM VNNI Targets (#13383)

This is an automated email from the ASF dual-hosted git repository. xiyou pushed a commit to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git The following commit(s) were added to refs/heads/main by this push: new 647be2b425 [MetaSchedule] Add `from-target` Defaults

[GitHub] [tvm] zxybazh merged pull request #13383: [MetaSchedule] Add `from-target` Defaults for LLVM VNNI Targets

zxybazh merged PR #13383: URL: https://github.com/apache/tvm/pull/13383 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm] junrushao commented on a diff in pull request #13195: [MetaSchedule] Refactor ScheduleRule Attributes

junrushao commented on code in PR #13195: URL: https://github.com/apache/tvm/pull/13195#discussion_r1022254178 ## src/te/operation/create_primfunc.cc: ## @@ -107,24 +108,30 @@ class LayoutFreePlaceholdersNormalizer : public StmtMutator { Stmt VisitStmt_(const BlockNode*

[GitHub] [tvm] masahi commented on a diff in pull request #13384: [MetaSchedule] Support schedules with cache read in RewriteLayout

masahi commented on code in PR #13384: URL: https://github.com/apache/tvm/pull/13384#discussion_r1022244412 ## src/te/operation/create_primfunc.cc: ## @@ -110,13 +110,20 @@ class LayoutFreePlaceholdersNormalizer : public StmtMutator { Block block =

[GitHub] [tvm] masahi commented on a diff in pull request #13383: [MetaSchedule] Add `from-target` Defaults for LLVM VNNI Targets

masahi commented on code in PR #13383: URL: https://github.com/apache/tvm/pull/13383#discussion_r1022238969 ## src/meta_schedule/postproc/postproc.cc: ## @@ -59,6 +59,14 @@ Array Postproc::DefaultLLVM() { }; } +Array Postproc::DefaultVNNI() { + return Array{ +

[GitHub] [tvm] vinx13 commented on a diff in pull request #13383: [MetaSchedule] Add `from-target` Defaults for LLVM VNNI Targets

vinx13 commented on code in PR #13383: URL: https://github.com/apache/tvm/pull/13383#discussion_r1022235387 ## src/meta_schedule/postproc/postproc.cc: ## @@ -59,6 +59,14 @@ Array Postproc::DefaultLLVM() { }; } +Array Postproc::DefaultVNNI() { + return Array{ +

[GitHub] [tvm] comaniac closed issue #13379: [Bug] no suitable user-defined conversion from "int2" to "longlong2" exists

comaniac closed issue #13379: [Bug] no suitable user-defined conversion from "int2" to "longlong2" exists URL: https://github.com/apache/tvm/issues/13379 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to

[tvm] branch main updated: [Codegen] Fix CUDA codegen for int64 Ramp (#13382)

This is an automated email from the ASF dual-hosted git repository. comaniac pushed a commit to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git The following commit(s) were added to refs/heads/main by this push: new 3aa16f72dd [Codegen] Fix CUDA codegen for int64 Ramp

[GitHub] [tvm] comaniac merged pull request #13382: [Codegen] Fix CUDA codegen for int64 Ramp

comaniac merged PR #13382: URL: https://github.com/apache/tvm/pull/13382 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm] comaniac commented on pull request #13382: [Codegen] Fix CUDA codegen for int64 Ramp

comaniac commented on PR #13382: URL: https://github.com/apache/tvm/pull/13382#issuecomment-1314643531 Thanks @vinx13 @junrushao -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific

[GitHub] [tvm] tvm-bot commented on pull request #13385: [TOPI] Add padding for dense/batch matmul for x86 vnni

tvm-bot commented on PR #13385: URL: https://github.com/apache/tvm/pull/13385#issuecomment-1314641665 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] vinx13 opened a new pull request, #13385: [TOPI] Add padding for dense/batch matmul for x86 vnni

vinx13 opened a new pull request, #13385: URL: https://github.com/apache/tvm/pull/13385 This added padding to make the shape of dense/batch matmul compatible with VNNI instructions. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to

[GitHub] [tvm] wzh99 commented on issue #12569: [Bug] no divide by zero report while te.floordiv(a, b-b)

wzh99 commented on issue #12569: URL: https://github.com/apache/tvm/issues/12569#issuecomment-1314637714 @cxx122 This bug is caused by a mistake in constant folding. It is fixed by #13348. -- This is an automated message from the Apache Git Service. To respond to the message, please log

[GitHub] [tvm] masahi commented on a diff in pull request #13195: [MetaSchedule] Refactor ScheduleRule Attributes

masahi commented on code in PR #13195: URL: https://github.com/apache/tvm/pull/13195#discussion_r106540 ## src/te/operation/create_primfunc.cc: ## @@ -107,24 +108,30 @@ class LayoutFreePlaceholdersNormalizer : public StmtMutator { Stmt VisitStmt_(const BlockNode*

[GitHub] [tvm] masahi commented on a diff in pull request #13195: [MetaSchedule] Refactor ScheduleRule Attributes

masahi commented on code in PR #13195: URL: https://github.com/apache/tvm/pull/13195#discussion_r106540 ## src/te/operation/create_primfunc.cc: ## @@ -107,24 +108,30 @@ class LayoutFreePlaceholdersNormalizer : public StmtMutator { Stmt VisitStmt_(const BlockNode*

[GitHub] [tvm] csullivan commented on a diff in pull request #13381: [Hexagon] Enable Hexagon User DMA bypass mode

csullivan commented on code in PR #13381: URL: https://github.com/apache/tvm/pull/13381#discussion_r1022211255 ## src/runtime/hexagon/hexagon_vtcm_pool.h: ## @@ -70,6 +70,16 @@ class HexagonVtcmPool { //! \brief Returns the total number of bytes in this pool size_t

[GitHub] [tvm] masahi commented on a diff in pull request #13383: [MetaSchedule] Add `from-target` Defaults for LLVM VNNI Targets

masahi commented on code in PR #13383: URL: https://github.com/apache/tvm/pull/13383#discussion_r1022209990 ## include/tvm/meta_schedule/mutator.h: ## @@ -131,6 +131,8 @@ class Mutator : public runtime::ObjectRef { FApply f_apply, FClone

[GitHub] [tvm] masahi commented on a diff in pull request #13383: [MetaSchedule] Add `from-target` Defaults for LLVM VNNI Targets

masahi commented on code in PR #13383: URL: https://github.com/apache/tvm/pull/13383#discussion_r1022209494 ## src/meta_schedule/postproc/postproc.cc: ## @@ -59,6 +59,14 @@ Array Postproc::DefaultLLVM() { }; } +Array Postproc::DefaultVNNI() { + return Array{ +

[GitHub] [tvm] zxybazh commented on pull request #13383: [MetaSchedule] Add `from-target` Defaults for LLVM VNNI Targets

zxybazh commented on PR #13383: URL: https://github.com/apache/tvm/pull/13383#issuecomment-1314607397 Thanks for the comments, the previous implementation was based on an earlier commit, now the schedule rules have been added. -- This is an automated message from the Apache Git Service.

[GitHub] [tvm] masahi opened a new pull request, #13384: [MetaSchedule] Support schedules with cache read in RewriteLayout

masahi opened a new pull request, #13384: URL: https://github.com/apache/tvm/pull/13384 Currently when `CacheRead` and `RewriteLayout` are used together, the index map is derived based on the cache read block, which leads to weird result. -- This is an automated message from the Apache

[GitHub] [tvm] tvm-bot commented on pull request #13384: [MetaSchedule] Support schedules with cache read in RewriteLayout

tvm-bot commented on PR #13384: URL: https://github.com/apache/tvm/pull/13384#issuecomment-1314607270 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] masahi commented on a diff in pull request #13383: [MetaSchedule] Add `from-target` Defaults for LLVM VNNI Targets

masahi commented on code in PR #13383: URL: https://github.com/apache/tvm/pull/13383#discussion_r1022206412 ## src/meta_schedule/schedule_rule/schedule_rule.cc: ## @@ -85,6 +85,39 @@ Array ScheduleRule::DefaultLLVM() { }; } +Array ScheduleRule::DefaultVNNI() { + return {

[GitHub] [tvm] masahi commented on a diff in pull request #13383: [MetaSchedule] Add `from-target` Defaults for LLVM VNNI Targets

masahi commented on code in PR #13383: URL: https://github.com/apache/tvm/pull/13383#discussion_r1022206036 ## src/meta_schedule/schedule_rule/schedule_rule.cc: ## @@ -85,6 +85,39 @@ Array ScheduleRule::DefaultLLVM() { }; } +Array ScheduleRule::DefaultVNNI() { + return {

[GitHub] [tvm] vinx13 commented on a diff in pull request #13383: [MetaSchedule] Add `from-target` Defaults for LLVM VNNI Targets

vinx13 commented on code in PR #13383: URL: https://github.com/apache/tvm/pull/13383#discussion_r1022205693 ## src/meta_schedule/schedule_rule/schedule_rule.cc: ## @@ -85,6 +85,39 @@ Array ScheduleRule::DefaultLLVM() { }; } +Array ScheduleRule::DefaultVNNI() { + return {

[GitHub] [tvm] zxybazh commented on pull request #13383: [MetaSchedule] Add `from-target` Defaults for LLVM VNNI Targets

zxybazh commented on PR #13383: URL: https://github.com/apache/tvm/pull/13383#issuecomment-1314601703 CC @michalpiszczek @vinx13 @junrushao @masahi -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go

[GitHub] [tvm] tvm-bot commented on pull request #13383: [MetaSchedule] Add `from-target` Defaults for LLVM VNNI Targets

tvm-bot commented on PR #13383: URL: https://github.com/apache/tvm/pull/13383#issuecomment-1314598471 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] zxybazh opened a new pull request, #13383: [MetaSchedule] Add `from-target` Defaults for LLVM VNNI Targets

zxybazh opened a new pull request, #13383: URL: https://github.com/apache/tvm/pull/13383 This PR introduces default schedule rules, post processors and mutators for LLVM VNNI targets. The usage of default vnni classes will automatically be generated if using `from-target` as parameters.

[tvm] branch nightly-docker-update updated (6852420b0d -> 4ce2e82ea1)

This is an automated email from the ASF dual-hosted git repository. github-bot pushed a change to branch nightly-docker-update in repository https://gitbox.apache.org/repos/asf/tvm.git omit 6852420b0d [ci][docker] Nightly Docker image update add 5a767d053d [CI] Separate the ci scripts

[tvm] branch main updated: [CI] Do not merge before running CI on main (#13372)

This is an automated email from the ASF dual-hosted git repository. driazati pushed a commit to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git The following commit(s) were added to refs/heads/main by this push: new 41a2243260 [CI] Do not merge before running CI on

[GitHub] [tvm] driazati closed issue #12392: [ci] Don't merge before running CI on `main`

driazati closed issue #12392: [ci] Don't merge before running CI on `main` URL: https://github.com/apache/tvm/issues/12392 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To

[GitHub] [tvm] driazati merged pull request #13372: [CI] Do not merge before running CI on main

driazati merged PR #13372: URL: https://github.com/apache/tvm/pull/13372 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[tvm] branch main updated (b6fae9b35e -> 68f51e692b)

This is an automated email from the ASF dual-hosted git repository. masahi pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git from b6fae9b35e [TIR][Bugfix] Fix AXIS_SEPARATORS in tir.Schedule.transform_layout (#13326) add 68f51e692b [ci] Fix

[GitHub] [tvm] masahi merged pull request #13380: [ci] Fix Jenkins quoting

masahi merged PR #13380: URL: https://github.com/apache/tvm/pull/13380 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm] masahi commented on issue #13364: [Bug] Padded, quantized convolution operators give incorrect results on edges

masahi commented on issue #13364: URL: https://github.com/apache/tvm/issues/13364#issuecomment-1314512094 This is not a bug from QNN right? We already do pad by input zp https://github.com/apache/tvm/blob/main/src/relay/qnn/op/convolution.cc#L250 -- This is an automated message from the

[tvm] branch main updated: [TIR][Bugfix] Fix AXIS_SEPARATORS in tir.Schedule.transform_layout (#13326)

This is an automated email from the ASF dual-hosted git repository. wuwei pushed a commit to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git The following commit(s) were added to refs/heads/main by this push: new b6fae9b35e [TIR][Bugfix] Fix AXIS_SEPARATORS in

[GitHub] [tvm] vinx13 merged pull request #13326: [TIR][Bugfix] Fix AXIS_SEPARATORS in tir.Schedule.transform_layout

vinx13 merged PR #13326: URL: https://github.com/apache/tvm/pull/13326 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm] trahman-quic commented on pull request #13282: [TOPI][Hexagon] Implement quantized adaptive_avg_pool1d for hexagon

trahman-quic commented on PR #13282: URL: https://github.com/apache/tvm/pull/13282#issuecomment-1314507265 Hi @mehrdadh The lint issues have been fixed but there are some CI tests that are failing. Are these related to my tests? Can you please give me an idea on how to resolve the issues?

[GitHub] [tvm] janetsc commented on a diff in pull request #13349: [TIR][Hexagon] Add vtcm memory capacity verification for Hexagon target

janetsc commented on code in PR #13349: URL: https://github.com/apache/tvm/pull/13349#discussion_r1022119475 ## tests/python/contrib/test_hexagon/test_cache_read_write.py: ## @@ -222,5 +223,28 @@ def _visit(stmt): ) +def _raises_exception(f): +try: +f() +

[GitHub] [tvm] junrushao commented on pull request #13382: [Codegen] Fix CUDA codegen for int64 Ramp

junrushao commented on PR #13382: URL: https://github.com/apache/tvm/pull/13382#issuecomment-1314479720 Should fix #13379 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment.

[GitHub] [tvm] tvm-bot commented on pull request #13382: [Codegen] Fix CUDA codegen for int64 Ramp

tvm-bot commented on PR #13382: URL: https://github.com/apache/tvm/pull/13382#issuecomment-1314454753 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] vinx13 opened a new pull request, #13382: [Codegen] Fix CUDA codegen for int64 Ramp

vinx13 opened a new pull request, #13382: URL: https://github.com/apache/tvm/pull/13382 CUDA codegen fails to generate `T.ramp(T.int64(0), T.int64(1), 2)]` correctly. This is a legal IR that appears after a recent commit to simplify cast and ramp. Fixed

[GitHub] [tvm] csullivan commented on a diff in pull request #13130: [TIR][Arith] Implement basic data-flow analysis

csullivan commented on code in PR #13130: URL: https://github.com/apache/tvm/pull/13130#discussion_r1022004179 ## src/tir/analysis/control_flow_graph.h: ## @@ -0,0 +1,597 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license

[GitHub] [tvm] echuraev commented on a diff in pull request #13362: [OpenCL] Introduce OpenCL wrapper to TVM

echuraev commented on code in PR #13362: URL: https://github.com/apache/tvm/pull/13362#discussion_r1022057781 ## cmake/modules/OpenCL.cmake: ## @@ -49,12 +49,17 @@ else() endif(USE_AOCL) if(USE_OPENCL) - if (NOT OpenCL_FOUND) -find_package(OpenCL REQUIRED) - endif() -

[GitHub] [tvm] tvm-bot commented on pull request #13381: [Hexagon] Enable Hexagon User DMA bypass mode

tvm-bot commented on PR #13381: URL: https://github.com/apache/tvm/pull/13381#issuecomment-1314365255 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] adstraw opened a new pull request, #13381: [Hexagon] Enable Hexagon User DMA bypass mode

adstraw opened a new pull request, #13381: URL: https://github.com/apache/tvm/pull/13381 Enables Hexagon User DMA bypass mode based on user-specified `dma_bypass_cache` option for DMA copies between DDR and VTCM. The **upside** of this change is increased DMA bandwidth (up to 40

[GitHub] [tvm] echuraev commented on a diff in pull request #13362: [OpenCL] Introduce OpenCL wrapper to TVM

echuraev commented on code in PR #13362: URL: https://github.com/apache/tvm/pull/13362#discussion_r1022052445 ## apps/cpp_rpc/README.md: ## @@ -37,7 +37,15 @@ This folder contains a simple recipe to make RPC server in c++. # Path to the desired C++ cross compiler

[GitHub] [tvm] tvm-bot commented on pull request #13380: [ci] Fix Jenkins quoting

tvm-bot commented on PR #13380: URL: https://github.com/apache/tvm/pull/13380#issuecomment-1314345769 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] driazati opened a new pull request, #13380: [ci] Fix Jenkins quoting

driazati opened a new pull request, #13380: URL: https://github.com/apache/tvm/pull/13380 Merging #13368 caused CI to pass but run more than it needed to due to some failures in determination. This fixes the interpolation to use `"` which should correctly pass through the variables

[GitHub] [tvm] adstraw closed pull request #13147: [Hexagon] Enable Hexagon User DMA bypass mode

adstraw closed pull request #13147: [Hexagon] Enable Hexagon User DMA bypass mode URL: https://github.com/apache/tvm/pull/13147 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment.

[GitHub] [tvm] elvin-n commented on a diff in pull request #13362: [OpenCL] Introduce OpenCL wrapper to TVM

elvin-n commented on code in PR #13362: URL: https://github.com/apache/tvm/pull/13362#discussion_r1021996120 ## apps/android_camera/app/src/main/jni/Application.mk: ## @@ -32,9 +32,9 @@ APP_ABI ?= all APP_STL := c++_shared APP_CPPFLAGS += -DTVM_LOG_STACK_TRACE=0

[GitHub] [tvm] elvin-n commented on a diff in pull request #13362: [OpenCL] Introduce OpenCL wrapper to TVM

elvin-n commented on code in PR #13362: URL: https://github.com/apache/tvm/pull/13362#discussion_r1021987579 ## cmake/modules/OpenCL.cmake: ## @@ -49,12 +49,17 @@ else() endif(USE_AOCL) if(USE_OPENCL) - if (NOT OpenCL_FOUND) -find_package(OpenCL REQUIRED) - endif() -

[GitHub] [tvm] joshherr-quic commented on pull request #12931: [Hexagon] 8 bit lut using tvm script for elemwise operators

joshherr-quic commented on PR #12931: URL: https://github.com/apache/tvm/pull/12931#issuecomment-1314287484 @tvm-bot rerun -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment.

[GitHub] [tvm] csullivan commented on a diff in pull request #13362: [OpenCL] Introduce OpenCL wrapper to TVM

csullivan commented on code in PR #13362: URL: https://github.com/apache/tvm/pull/13362#discussion_r1021953728 ## src/runtime/opencl/opencl_wrapper/opencl_wrapper.cc: ## @@ -0,0 +1,551 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor

[GitHub] [tvm] csullivan commented on a diff in pull request #13362: [OpenCL] Introduce OpenCL wrapper to TVM

csullivan commented on code in PR #13362: URL: https://github.com/apache/tvm/pull/13362#discussion_r1021953728 ## src/runtime/opencl/opencl_wrapper/opencl_wrapper.cc: ## @@ -0,0 +1,551 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor

[GitHub] [tvm] junrushao commented on pull request #13367: [Lint] Fix import in test_tvmscript_type.py

junrushao commented on PR #13367: URL: https://github.com/apache/tvm/pull/13367#issuecomment-1314206625 Perhaps it could be some caching issues -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to

[GitHub] [tvm] comaniac opened a new issue, #13379: [Bug] no suitable user-defined conversion from "int2" to "longlong2" exists

comaniac opened a new issue, #13379: URL: https://github.com/apache/tvm/issues/13379 PR #13317 introduces a bug when generating CUDA kernel for ops with index (e.g., gather). With some investigations, it depends on the input shape. Taking the unit tests shown below, error only happens

[GitHub] [tvm] gigiblender commented on pull request #13367: [Lint] Fix import in test_tvmscript_type.py

gigiblender commented on PR #13367: URL: https://github.com/apache/tvm/pull/13367#issuecomment-1314200276 > Would you mind sharing the error message from the linter? Thanks! It was a python module not found error. I ran the linter locally using the docker image and, strangely am no

[GitHub] [tvm] junrushao commented on pull request #13367: [Lint] Fix import in test_tvmscript_type.py

junrushao commented on PR #13367: URL: https://github.com/apache/tvm/pull/13367#issuecomment-1314195373 Would you mind sharing the error message from the linter? Thanks! -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use

[GitHub] [tvm] mehrdadh opened a new issue, #13378: [Bug][microTVM] Project API does not use projectOption default values

mehrdadh opened a new issue, #13378: URL: https://github.com/apache/tvm/issues/13378 Currently, if we set a default value for a projectOption but we don't pass the option when we generate the project, it will ignore the default value of the projectOption defined in the serve API. cc

[GitHub] [tvm] tkonolige commented on a diff in pull request #13375: [Metaschedule] Add partition primitive

tkonolige commented on code in PR #13375: URL: https://github.com/apache/tvm/pull/13375#discussion_r1021900519 ## src/tir/schedule/analysis/analysis.cc: ## @@ -890,7 +890,7 @@ IterVarType GetLoopIterType(const StmtSRef& loop_sref) { } }); } -

[GitHub] [tvm] jwfromm commented on a diff in pull request #13375: [Metaschedule] Add partition primitive

jwfromm commented on code in PR #13375: URL: https://github.com/apache/tvm/pull/13375#discussion_r1021898538 ## src/tir/schedule/analysis/analysis.cc: ## @@ -890,7 +890,7 @@ IterVarType GetLoopIterType(const StmtSRef& loop_sref) { } }); } -

[GitHub] [tvm] junrushao commented on pull request #13374: [Hexagon] Change single dma alloc buffers to be vtcm.

junrushao commented on PR #13374: URL: https://github.com/apache/tvm/pull/13374#issuecomment-1314160451 Yep I tried it out and see tests failing :-( -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go

[GitHub] [tvm] tvm-bot commented on pull request #13377: [microTVM][Zephyr]Add serial_number test args

tvm-bot commented on PR #13377: URL: https://github.com/apache/tvm/pull/13377#issuecomment-1314158876 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] tvm-bot commented on pull request #13377: [microTVM][Zephyr]Add serial_number test args

tvm-bot commented on PR #13377: URL: https://github.com/apache/tvm/pull/13377#issuecomment-1314158891 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] mehrdadh opened a new pull request, #13377: [microTVM][Zephyr]Add serial_number test args

mehrdadh opened a new pull request, #13377: URL: https://github.com/apache/tvm/pull/13377 This PR does these changes: - add serial number to specify device serial number in Zephyr testing - change Zephyr server API to run tests with specific serial number - remove west_cmd in test

[GitHub] [tvm] joshherr-quic commented on a diff in pull request #12931: [Hexagon] 8 bit lut using tvm script for elemwise operators

joshherr-quic commented on code in PR #12931: URL: https://github.com/apache/tvm/pull/12931#discussion_r1021888230 ## python/tvm/topi/hexagon/lut.py: ## @@ -0,0 +1,74 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements.

[GitHub] [tvm] tvm-bot commented on pull request #13376: Bump cryptography from 37.0.2 to 38.0.3 in /docker/python

tvm-bot commented on PR #13376: URL: https://github.com/apache/tvm/pull/13376#issuecomment-1314128294 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] tvm-bot commented on pull request #13376: Bump cryptography from 37.0.2 to 38.0.3 in /docker/python

tvm-bot commented on PR #13376: URL: https://github.com/apache/tvm/pull/13376#issuecomment-1314128300 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] tvm-bot commented on pull request #13376: Bump cryptography from 37.0.2 to 38.0.3 in /docker/python

tvm-bot commented on PR #13376: URL: https://github.com/apache/tvm/pull/13376#issuecomment-1314128278 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[tvm] branch dependabot/pip/docker/python/cryptography-38.0.3 created (now 36f533ce51)

This is an automated email from the ASF dual-hosted git repository. github-bot pushed a change to branch dependabot/pip/docker/python/cryptography-38.0.3 in repository https://gitbox.apache.org/repos/asf/tvm.git at 36f533ce51 Bump cryptography from 37.0.2 to 38.0.3 in /docker/python No

[GitHub] [tvm] dependabot[bot] opened a new pull request, #13376: Bump cryptography from 37.0.2 to 38.0.3 in /docker/python

dependabot[bot] opened a new pull request, #13376: URL: https://github.com/apache/tvm/pull/13376 Bumps [cryptography](https://github.com/pyca/cryptography) from 37.0.2 to 38.0.3. Changelog Sourced from https://github.com/pyca/cryptography/blob/main/CHANGELOG.rst;>cryptography's

[tvm] branch main updated (b8384d105b -> 5a767d053d)

This is an automated email from the ASF dual-hosted git repository. driazati pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git from b8384d105b [ci] Assert some tests are not skipped in the CI (#12915) add 5a767d053d [CI] Separate the ci scripts

[GitHub] [tvm] driazati merged pull request #13368: [CI] Separate the ci scripts into Github and Jenkins scripts

driazati merged PR #13368: URL: https://github.com/apache/tvm/pull/13368 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm] tvm-bot commented on pull request #13375: [Metaschedule] Add partition primitive

tvm-bot commented on PR #13375: URL: https://github.com/apache/tvm/pull/13375#issuecomment-1314124649 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] tvm-bot commented on pull request #13375: [Metaschedule] Add partition primitive

tvm-bot commented on PR #13375: URL: https://github.com/apache/tvm/pull/13375#issuecomment-1314124651 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] tkonolige opened a new pull request, #13375: [Metaschedule] Add partition primitive

tkonolige opened a new pull request, #13375: URL: https://github.com/apache/tvm/pull/13375 Add partition scheduling primitive. Partition splits a loop into two sequential loops based on a factor. For example: ``` for i in range(10): x[i] = 0 ``` could be partitioned

[GitHub] [tvm] tvm-bot commented on pull request #13374: [Hexagon] Change single dma alloc buffers to be vtcm.

tvm-bot commented on PR #13374: URL: https://github.com/apache/tvm/pull/13374#issuecomment-1314124367 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] nverke opened a new pull request, #13374: [Hexagon] Change single dma alloc buffers to be vtcm.

nverke opened a new pull request, #13374: URL: https://github.com/apache/tvm/pull/13374 These need to be global.vtcm scope in order to properly test the single dma schedule. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub

[GitHub] [tvm] nverke commented on a diff in pull request #13301: [TIR] Update ReductionIterNotIndexOutputBuffer to check BlockRealizeN…

nverke commented on code in PR #13301: URL: https://github.com/apache/tvm/pull/13301#discussion_r1021841923 ## tests/python/unittest/test_tir_transform_lower_cross_thread_reduction.py: ## @@ -487,6 +487,149 @@ def lowered_single_reduction_loop_with_block_predicate(

[GitHub] [tvm] tvm-bot commented on pull request #13373: [TIR] Add test to cover specific case of reducer match buffer checking

tvm-bot commented on PR #13373: URL: https://github.com/apache/tvm/pull/13373#issuecomment-1314102973 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] nverke opened a new pull request, #13373: [TIR] Add test to cover specific case of reducer match buffer checking

nverke opened a new pull request, #13373: URL: https://github.com/apache/tvm/pull/13373 Adding in a test that covers another case that can happen when running this check. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and

[GitHub] [tvm] tmoreau89 commented on issue #13364: [Bug] Padded, quantized convolution operators give incorrect results on edges

tmoreau89 commented on issue #13364: URL: https://github.com/apache/tvm/issues/13364#issuecomment-1314063821 CC @ibsidorenko as an FYI -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the

[GitHub] [tvm] Lunderberg commented on a diff in pull request #13130: [TIR][Arith] Implement basic data-flow analysis

Lunderberg commented on code in PR #13130: URL: https://github.com/apache/tvm/pull/13130#discussion_r1021747479 ## src/tir/analysis/control_flow_graph.h: ## @@ -0,0 +1,458 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license

[GitHub] [tvm] tvm-bot commented on pull request #13372: [CI] Do not merge before running CI on main

tvm-bot commented on PR #13372: URL: https://github.com/apache/tvm/pull/13372#issuecomment-1313965901 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] gigiblender opened a new pull request, #13372: [CI] Do not merge before running CI on main

gigiblender opened a new pull request, #13372: URL: https://github.com/apache/tvm/pull/13372 This PR does not merge `main` if CI is running already on `main`. It aims to avoid a case where a race happens between two subsequent commits, and one of them merges the other. Fixes #12392.

[GitHub] [tvm] guberti commented on issue #13364: [Bug] [microTVM] Native schedules for microTVM have worse accuracy than expected

guberti commented on issue #13364: URL: https://github.com/apache/tvm/issues/13364#issuecomment-1313903539 # How does quantization work in general? Consider a **fused, quantized conv2d** operator, like those in the MLPerf Tiny (or any quantized MobileNetV1 model). In addition to

[GitHub] [tvm] tvm-bot commented on pull request #13371: This fix adds check for time result presence.

tvm-bot commented on PR #13371: URL: https://github.com/apache/tvm/pull/13371#issuecomment-1313896475 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] shtinsa opened a new pull request, #13371: This fix adds check for time result presence.

shtinsa opened a new pull request, #13371: URL: https://github.com/apache/tvm/pull/13371 The cost mode can fail if run_secs value is not exist. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to

[GitHub] [tvm] leandron commented on pull request #13369: [usmp] Hill Climb greedy layout size check relaxed

leandron commented on PR #13369: URL: https://github.com/apache/tvm/pull/13369#issuecomment-1313867387 cc @Alex-grovety @sergey-grovety @ilyag-grovety for reviews -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the

  1   2   >