Commit Graph

440 Commits

Author SHA1 Message Date
bors[bot]
04a4758df0 Merge #2656
2656: update(compiler) switched upstream Inkwell (for #2645) r=Amanieu a=ptitSeb

# Description
Use upstream Inkwell instead of custom version

Co-authored-by: ptitSeb <sebastien.chev@gmail.com>
2021-11-05 12:58:17 +00:00
ptitSeb
14431e5c61 fixed lint 2021-11-05 12:15:26 +01:00
ptitSeb
d7bb25de90 update(compiler) fixed new warning 2021-11-05 12:12:36 +01:00
ptitSeb
190d670eb9 fix linter 2021-11-05 09:59:58 +01:00
ptitSeb
7ccd6af929 update(compiler) switched upstream Inkwell (for #2645) 2021-11-05 09:54:47 +01:00
Amanieu d'Antras
7c67f7ce71 Remove unused nightly feature 2021-11-04 17:37:22 +00:00
ptitSeb
8c24ab48db Some last small changes 2021-10-18 11:55:42 +02:00
ptitSeb
eccc9dc77e Removed unused commented code 2021-10-12 13:30:57 +02:00
ptitSeb
b54780cdab feat(compiler) Refactor the new ARM Reloc and Trampoline to avoid a &mut 2021-10-12 10:11:36 +02:00
ptitSeb
0c6010ca34 feat(compiler) Use x17 as scratch instead of x16 on Aarch64 to help with Apple ABI 2021-10-08 16:19:31 +02:00
ptitSeb
204238c9c9 feat(compiler) Fixed linting 2021-10-07 11:27:08 +02:00
ptitSeb
9cf36055f2 feat(compiler) Don't try to use macOS Aarch64 specific ABI for now (at least coremark works now) 2021-10-07 11:24:32 +02:00
ptitSeb
20f0c66fe6 feat(compiler) Added Trampolines and more Relocations for Arm64 (llvm-universal on linux-aarch64 are OK now) 2021-10-05 14:08:55 +02:00
ptitSeb
21660e6d05 feat(compiler) Added preliminary support for Arm64Call relocation 2021-09-28 10:03:25 +02:00
ptitSeb
12135eb062 fix(compiler) macOS Aarch64 ABI is not SystemV 2021-09-28 10:01:16 +02:00
Syrus Akbary
f7dcb02a79 Updated version 2021-08-21 10:21:37 +01:00
Nathan Horrigan
6ddc2f6454 Add iOS support for Headless Wasmer 2021-08-21 10:21:37 +01:00
Syrus Akbary
005d129529 Update all packages 2021-08-10 11:40:38 -07:00
Syrus Akbary
98481f51fe Move ModuleInfo to wasmer-types 2021-07-16 15:42:39 -07:00
yukang
be70057830 add notes for llvm 2021-07-08 21:37:12 +08:00
Midas Lambrichts
15da6e8ce0 Added the 32 bit variant of x86 to generate a target machine for inkwell
Extended the match on X86_64 to include the X86_32 and all of its
architectures.
2021-06-20 21:01:38 +02:00
Mark McCaskey
0c0fdad3d2 Prepare for 2.0.0 release 2021-06-15 11:26:46 -07:00
Ivan Enderlin
2fbbf36d7a chore(llvm) Update object to 0.25. 2021-06-08 09:01:00 +02:00
Mark McCaskey
80a8658e5a Prepare for 2.0.0-rc2 release 2021-06-03 09:19:16 -07:00
dependabot[bot]
1b95c9c6f2 Bump semver from 0.11.0 to 1.0.2
Bumps [semver](https://github.com/dtolnay/semver) from 0.11.0 to 1.0.2.
- [Release notes](https://github.com/dtolnay/semver/releases)
- [Commits](https://github.com/dtolnay/semver/compare/0.11.0...1.0.2)

---
updated-dependencies:
- dependency-name: semver
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
2021-06-02 18:07:50 +00:00
Mark McCaskey
cfc1a5fc26 Fix typo 2.0.0-rc -> 2.0.0-rc1 2021-06-02 08:32:46 -07:00
Mark McCaskey
2897a5b05e Bump version number to 2.0.0-rc1 2021-06-02 08:21:34 -07:00
jubianchi
ec1430c51a cargo fmt 2021-05-31 15:21:10 +02:00
jubianchi
0b119a7045 cache into calls 2021-05-31 12:26:10 +02:00
jubianchi
bb73126880 Review 2021-05-31 11:15:30 +02:00
jubianchi
e5e51b9f07 Cleaning some bits 2021-05-28 18:31:36 +02:00
jubianchi
69cf276dfe Make the code compile with latest Inkwell changes 2021-05-28 18:14:06 +02:00
jubianchi
6a99268895 feat(llvm): Make NaN canonicalization configurable
compiler-llvm now uses the experimental.constrained intrinsics to ensure
correct behavior on FP operations when full-canonicalization is
disabled.

This patch requires TheDan64/inkwell#247
2021-05-28 11:40:10 +02:00
Ivan Enderlin
e5a784a006 feat: Rename wasmer-engine-jit to wasmer-engine-universal.
This (small) patch renames the `wasmer-engine-jit` crate into
`wasmer-engine-universal`.
2021-05-25 16:12:06 +02:00
Syrus Akbary
58b18189da Updated dependencies 2021-05-14 15:47:03 -07:00
Mark McCaskey
0d4fe293e5 Merge branch 'master' into feature/update-clif-small-simd-improvement 2021-05-13 15:38:40 -04:00
Mark McCaskey
37be884403 Clean up F32x4DemoteF64x2Zero op in compiler llvm 2021-05-13 08:20:00 -07:00
Mark McCaskey
f8b67c743a Remove REVIEW request comments 2021-05-12 10:14:58 -07:00
Mark McCaskey
9e04e2624b Update cranelift, impl low hanging fruit SIMD opcodes 2021-05-12 09:15:56 -07:00
Mark McCaskey
1082fe1692 Clean up LLVM SIMD changes 2021-05-12 08:10:54 -07:00
Mark McCaskey
88d2ac9626 Fix bug in I32x4TruncSatF64x2SZero
We need to use LEF64, not LEF32, when dealing with f64s
2021-05-11 14:47:44 -07:00
Mark McCaskey
20297858ed Implement I16x8Q15MulrSatS in compiler LLVM 2021-05-11 14:37:59 -07:00
Mark McCaskey
da86e69a45 Implement all but 2 new SIMD ops in compiler LLVM 2021-05-11 11:14:25 -07:00
Mark McCaskey
2f5f9d9e48 Implement most of the rest of the new SIMD in llvm 2021-05-11 09:49:48 -07:00
Mark McCaskey
17c27f1d15 Implement new SIMD instructions in compiler LLVM 2021-05-07 11:21:13 -07:00
Mark McCaskey
d699e61083 Implement bulk memory instructions in compiler-llvm 2021-05-06 09:07:51 -07:00
Nick Lewycky
41561a45d8 Move back to object 0.23.
There's nothing wrong with 0.24 but we have a dependency on 0.23 through 'backtrace' which is used by wasmer-engine, and we'd like to only have one copy.
2021-05-03 17:03:30 -07:00
Nick Lewycky
18732e4b65 cargo fmt 2021-05-03 16:45:03 -07:00
Nick Lewycky
3339b76414 Use symbolic name instead of number for relocation. 2021-05-03 16:39:49 -07:00
Nick Lewycky
bab4e53c58 Remove commented-out implementation of ElfSectionIndex. 2021-05-03 16:25:27 -07:00