Re: [apache/tvm] [release] v0.9.0 Release Schedule (Issue #11736)
@driazati / @areusch are we pushing the release schedule for #12022 ? -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm/issues/11736#issuecomment-1184613371 You are receiving this because you are subscribed to this thread. Message ID:
Re: [apache/tvm] [release] v0.9.0 Release Schedule (Issue #11736)
Yeah, along with some other things but #12022 is the only remaining item -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm/issues/11736#issuecomment-1184685291 You are receiving this because you are subscribed to this thread. Message ID:
Re: [apache/tvm-rfcs] Add Target Pre-processing RFC (PR #71)
@kparzysz-quic @junrushao1994 @tqchen @Lunderberg can you look at this one again? -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm-rfcs/pull/71#issuecomment-1184767286 You are receiving this because you are subscribed to this thread. Message ID:
Re: [apache/tvm-rfcs] [RFC] Name mangling in IRModules (PR #84)
we discussed this at the [community meeting](https://discuss.tvm.apache.org/t/next-tvm-community-meeting-july-13/13120/2) yesterday. here are notes: - @areusch: one of suggestions raised was to make it possible to reconstruct a NameSupply from an IRModule. we might need to make it possible to initialize it from multiple IRmodules since we currently lower Relay to many separate IRModules in the compiler. - @kparzysz-quic : what's the use of having multiple modules, unless they're built for separate devices. do we need `mod_name` as a field? - `mod_name` is not to distinguish different IRModules at compile time; more to distinguish separate models at runtime via namespace prefix. - @kparzysz-quic : what is the connection between the name that we attach to a global and the name that actually ends up in an object file? there is a strong need to have a degree of control over the names that end up in the object file and their visibility. how are we addressing this need? i think there is a lack of this in TVM now. we don't have a way - @tqchen: good point. we should document the linkage type. right now there is an implicit convention: the `name_hint` is not final (it's just an internal-to-compiler concept) and there's a GlobalVar which enforces that there is external linkage type with a fixed name. e.g. if we are splitting the compilation flow, then there is an expectation that when calling into a symbol, the name won't change. another rule: if a function has attribute `kGlobalSymbol`, you cannot change it or even delete it. - @manupa-arm : when we say we want to have control: do we want this to extend to Relay? if a Relay module contains a GlobalVar, should that be emitted with exactly that name? - @tqchen : if an IRModule has a function with attr `kGlobalSymbol` we should respect that. but as a rule of thumb we should delay attaching this attr til as late of possible, so we should try to avoid that symbol. - @areusch : we also don't have this ability in scheduling; it would be hard to attach that attribute to a Relay function e.g. main because we prefix lowered functions - @tqchen: we could special-case an AOT entry function when the compiler takes special care around a symbol. - filed https://github.com/apache/tvm/issues/12098 to document the convention. - @areusch some context for this RFC: in the C++ runtime, we previously mostly cared that a GlobalVar name could be passed to Module.GetFunction() and then that function would find that implemented PrimFunc. When we implemented AOT in the C runtime, we changed to expecting to be able to directly call those names as C symbols. This caused a bit of a problem when we moved to implement both the C and C++ AOTExecutor wrappers, where metadata lookups could be different because of name mangling. To solve that, we intended to move in a direction where the names of all symbols were fixed at birth, thus NameSupply. - @kparzysz-quic: who is typically the caller of these functions? - @areusch typically an executor - @kparzysz-quic can we modify the names of the global? shouldn't matter so long as everyone agrees -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm-rfcs/pull/84#issuecomment-1184817996 You are receiving this because you are subscribed to this thread. Message ID:
[apache/tvm] TVM v0.9.0.rc0 Release Candidate (Issue #12102)
**Please leave any comments or edit this issue directly to adjust the release notes. Also see the rc0 vote thread #tbd**. # Introduction The TVM community has worked since the v0.8 release to deliver many exciting features and improvements. v0.9.0 is the first release on the new [quarterly release](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0067-quarterly-releases.md) schedule and includes many highlights, such as: * [MetaSchedule's full implementation](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0005-meta-schedule-autotensorir.md) * [ARM cascading scheduler](https://github.com/apache/tvm-rfcs/pull/37) for Arm Ethos(TM)-U NPUs * [Collage](https://github.com/apache/tvm-rfcs/pull/62) which brings tuning to BYOC * Several microTVM improvements * [Dynamic shape optimization](https://github.com/apache/tvm-rfcs/pull/72) * New `tvm.relay.build` parameters: `runtime=`, `executor=`, * AOT: support for the C++ runtime (with `llvm` and `c` targets only) and support for host-driven AOT in the C runtime * Hexagon RPC support * Testing via Hexagon SDK simulator and on device via Snapdragon-based HDK boards and phones * AOT and USMP support * Threading * Initial op support * MLF: support for multiple modules in a single MLF artifact * And many more! See the list of RFCs and PRs included in v0.9.0 for a complete list, as well as [the full change list](https://github.com/apache/tvm/compare/v0.8.0...v0.9.0.rc0). ## RFCs These RFCs have been merged in [apache/tvm-rfcs](https://github.com/apache/tvm-rfcs) since the last release. * [[RFC] TUNIP: TVMScript Unified Printer (#74)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0074-tvmscript-unified-printer.md) ([`48d47c5`](https://github.com/apache/tvm-rfcs/commit/48d47c526fbb509115adf5bbaf8699129f4fe2be)) * [[RFC][Backend] RFC-CSI-NN2-Integration (#75)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0075_RISC-V_CSI-NN2_Intergration.md) ([`cfcf114`](https://github.com/apache/tvm-rfcs/commit/cfcf11464003af4c3201345a24ab380952fed944)) * [[RFC] Introducing DeclBuffer (#70)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0070-introducing-decl-buffer.md) ([`87ff1fa`](https://github.com/apache/tvm-rfcs/commit/87ff1facd55c0a7cef45efdf2b7548ee299e8e06)) * [[RFC][MLF] Model Library Format with Multiple Modules (#76)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0076_mlf_with_multiple_modules.md) ([`f47c6ad`](https://github.com/apache/tvm-rfcs/commit/f47c6ad660edf2c97e54279910e8b60b2bdf9ada)) * [[RFC] UMA Universal Modular Accelerator Interface (#60)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0060_UMA_Unified_Modular_Accelerator_Interface.md) ([`6990e13`](https://github.com/apache/tvm-rfcs/commit/6990e1363c96945b6c9d19dc2d331306d2be2a17)) * [[RFC] DietCode: An Auto-Scheduler for Dynamic Tensor Programs (#72)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0072-dynamic-autoscheduler.md) ([`a518000`](https://github.com/apache/tvm-rfcs/commit/a518000cbc82e53321f526d2090c7c8067607391)) * [[RFC] Quarterly Releases (#67)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0067-quarterly-releases.md) ([`70293c7`](https://github.com/apache/tvm-rfcs/commit/70293c7f910ad70f4744d26295d6c4f6fe0366c5)) * [RFC-BYOC-DNNL-Integration (#73)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0069-byoc-onednn-integration.md) ([`7aed0ca`](https://github.com/apache/tvm-rfcs/commit/7aed0ca0e1e636aebcfe7fd10585f4c7e1893674)) * [[RFC] Relay Next Roadmap (#69)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0069-relax-roadmap.md) ([`ac15f2a`](https://github.com/apache/tvm-rfcs/commit/ac15f2a7bd77b106636ae5e4d8357a05d7267ef5)) * [RFC: clarifying buffer declaration and access (#63)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0063-clarifying-buffer-declaration-and-access.md) ([`de4fe97`](https://github.com/apache/tvm-rfcs/commit/de4fe97ac0ab9d8fe9d4309a380a5ea278aa1493)) * [Inclusive Language RFC (#68) (#68)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0068-inclusive-language.md) ([`4203bd2`](https://github.com/apache/tvm-rfcs/commit/4203bd2da4ffc98eb70731a05fcadc478679a461)) * [[USMP] Adding U4 usecase (#65)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0009_Unified_Static_Memory_Planning.md) ([`b9e246f`](https://github.com/apache/tvm-rfcs/commit/b9e246f6d1715176f5de10fe1e7e382b88bacabe)) * [Collage RFC (#62)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0062-collage.md) ([`23250f5`](https://github.com/apache/tvm-rfcs/commit/23250f59cfc89673b7efe1d8f13192ed8381824d)) * [Replace codeowners with more relevant automation (#58)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0058-replace-codeowners.md) ([`540c1f8`](https://github.com/apache/tvm-rfcs/commit/540c1f80631b5d9ee2ae9ff827edc73240fe1b09)) * [[RFC][TIR] Layout transformations on buffer access (#39)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0039-buffer-physical-layout.md) ([`b675ef8`](https://gith
[apache/tvm] [VOTE] Release Apache TVM v0.9.0.rc0 (Issue #12103)
Dear TVM community, This is a call for vote to release Apache TVM version v0.9.0. This is a major release with many new features and improvements. All users of Apache TVM v0.8 are advised to upgrade. Please see #12102 for any edits you would like to see included in the release notes. Link to release notes: https://github.com/apache/tvm/releases/tag/v0.9.0.rc0 Link to release candidate: https://dist.apache.org/repos/dist/dev/tvm/tvm-v0.9.0-rc0/ The vote will be open for at least 72 hours. Everyone is welcomed to vote. Please vote by replying to this thread explicitly. +1 = approve +0 = no opinion -1 = disapprove (provide reason) NOTE: this thread is being mirrored in dev@ -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm/issues/12103 You are receiving this because you are subscribed to this thread. Message ID:
[apache/tvm] Pre-release v0.9.0.rc0 - Apache TVM v0.9.0.rc0
# Introduction The TVM community has worked since the v0.8 release to deliver many exciting features and improvements. v0.9.0 is the first release on the new [quarterly release](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0067-quarterly-releases.md) schedule and includes many highlights, such as: * [MetaSchedule's full implementation](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0005-meta-schedule-autotensorir.md) * [ARM cascading scheduler](https://github.com/apache/tvm-rfcs/pull/37) for Arm Ethos(TM)-U NPUs * [Collage](https://github.com/apache/tvm-rfcs/pull/62) which brings tuning to BYOC * Several microTVM improvements * [Dynamic shape optimization](https://github.com/apache/tvm-rfcs/pull/72) * New `tvm.relay.build` parameters: `runtime=`, `executor=`, * AOT: support for the C++ runtime (with `llvm` and `c` targets only) and support for host-driven AOT in the C runtime * Hexagon RPC support * Testing via Hexagon SDK simulator and on device via Snapdragon-based HDK boards and phones * AOT and USMP support * Threading * Initial op support * MLF: support for multiple modules in a single MLF artifact * And many more! See the list of RFCs and PRs included in v0.9.0 for a complete list, as well as [the full change list](https://github.com/apache/tvm/compare/v0.8.0...v0.9.0.rc0). ## RFCs These RFCs have been merged in [apache/tvm-rfcs](https://github.com/apache/tvm-rfcs) since the last release. * [[RFC] TUNIP: TVMScript Unified Printer (#74)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0074-tvmscript-unified-printer.md) ([`48d47c5`](https://github.com/apache/tvm-rfcs/commit/48d47c526fbb509115adf5bbaf8699129f4fe2be)) * [[RFC][Backend] RFC-CSI-NN2-Integration (#75)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0075_RISC-V_CSI-NN2_Intergration.md) ([`cfcf114`](https://github.com/apache/tvm-rfcs/commit/cfcf11464003af4c3201345a24ab380952fed944)) * [[RFC] Introducing DeclBuffer (#70)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0070-introducing-decl-buffer.md) ([`87ff1fa`](https://github.com/apache/tvm-rfcs/commit/87ff1facd55c0a7cef45efdf2b7548ee299e8e06)) * [[RFC][MLF] Model Library Format with Multiple Modules (#76)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0076_mlf_with_multiple_modules.md) ([`f47c6ad`](https://github.com/apache/tvm-rfcs/commit/f47c6ad660edf2c97e54279910e8b60b2bdf9ada)) * [[RFC] UMA Universal Modular Accelerator Interface (#60)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0060_UMA_Unified_Modular_Accelerator_Interface.md) ([`6990e13`](https://github.com/apache/tvm-rfcs/commit/6990e1363c96945b6c9d19dc2d331306d2be2a17)) * [[RFC] DietCode: An Auto-Scheduler for Dynamic Tensor Programs (#72)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0072-dynamic-autoscheduler.md) ([`a518000`](https://github.com/apache/tvm-rfcs/commit/a518000cbc82e53321f526d2090c7c8067607391)) * [[RFC] Quarterly Releases (#67)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0067-quarterly-releases.md) ([`70293c7`](https://github.com/apache/tvm-rfcs/commit/70293c7f910ad70f4744d26295d6c4f6fe0366c5)) * [RFC-BYOC-DNNL-Integration (#73)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0069-byoc-onednn-integration.md) ([`7aed0ca`](https://github.com/apache/tvm-rfcs/commit/7aed0ca0e1e636aebcfe7fd10585f4c7e1893674)) * [[RFC] Relay Next Roadmap (#69)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0069-relax-roadmap.md) ([`ac15f2a`](https://github.com/apache/tvm-rfcs/commit/ac15f2a7bd77b106636ae5e4d8357a05d7267ef5)) * [RFC: clarifying buffer declaration and access (#63)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0063-clarifying-buffer-declaration-and-access.md) ([`de4fe97`](https://github.com/apache/tvm-rfcs/commit/de4fe97ac0ab9d8fe9d4309a380a5ea278aa1493)) * [Inclusive Language RFC (#68) (#68)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0068-inclusive-language.md) ([`4203bd2`](https://github.com/apache/tvm-rfcs/commit/4203bd2da4ffc98eb70731a05fcadc478679a461)) * [[USMP] Adding U4 usecase (#65)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0009_Unified_Static_Memory_Planning.md) ([`b9e246f`](https://github.com/apache/tvm-rfcs/commit/b9e246f6d1715176f5de10fe1e7e382b88bacabe)) * [Collage RFC (#62)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0062-collage.md) ([`23250f5`](https://github.com/apache/tvm-rfcs/commit/23250f59cfc89673b7efe1d8f13192ed8381824d)) * [Replace codeowners with more relevant automation (#58)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0058-replace-codeowners.md) ([`540c1f8`](https://github.com/apache/tvm-rfcs/commit/540c1f80631b5d9ee2ae9ff827edc73240fe1b09)) * [[RFC][TIR] Layout transformations on buffer access (#39)](https://github.com/apache/tvm-rfcs/blob/main/rfcs/0039-buffer-physical-layout.md) ([`b675ef8`](https://github.com/apache/tvm-rfcs/commit/b675ef8e74351f298b8f3bc4ae944eaf3cbbe430)) * [Module Based Model Runtime for AOT (#46)](http
Re: [apache/tvm] [VOTE] Release Apache TVM v0.9.0.rc0 (Issue #12103)
+1 -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm/issues/12103#issuecomment-1185035185 You are receiving this because you are subscribed to this thread. Message ID:
Re: [apache/tvm] [VOTE] Release Apache TVM v0.9.0.rc0 (Issue #12103)
+1 -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm/issues/12103#issuecomment-1185206169 You are receiving this because you are subscribed to this thread. Message ID:
