Commit Graph

4223 Commits

Author SHA1 Message Date
losfair
2eb11f5052 Update Cargo.lock 2019-12-03 01:46:13 +08:00
losfair
04d8df0a46 Add comments and remove unneeded statics. 2019-12-03 01:46:06 +08:00
losfair
063fa211a7 Merge remote-tracking branch 'origin/master' into feature/singlepass-cache 2019-11-29 01:33:29 +08:00
losfair
c3f2481ced Update feature matrix. 2019-11-29 01:32:08 +08:00
losfair
513427be93 Remove singlepass codegen dependence on runtime memory. 2019-11-29 01:30:08 +08:00
bors[bot]
19dfdec236
Merge #1004
1004: Adapt backend usage depending on wasm file executed r=MarkMcCaskey a=d0iasm

Adapt backend usage depending on wasm file executed in issue #998.
Close #998 

# Description
Add `auto` backend into a runtime-core and use it as a default backend.
The `auto` backend is equivalent to: 
* singlepass if singlepass is enabled and the wasm file size is larger than 10MiB, or singlepass is enable and the target architecture is aarch64.
* cranelift otherwise.


Co-authored-by: Asami Doi <doiasami1219@gmail.com>
2019-11-27 19:37:04 +00:00
Heyang Zhou
24ead779d6
Remove extra '.'.
Co-Authored-By: nlewycky <nick@wasmer.io>
2019-11-28 03:25:05 +08:00
losfair
c7f4ca5dd3 Enable cache tests for singlepass. 2019-11-28 03:01:09 +08:00
losfair
70bc382843 Add changelog entry. 2019-11-28 02:54:53 +08:00
losfair
cfa0635026 Add caching support for singlepass backend. 2019-11-28 02:49:52 +08:00
bors[bot]
b6d8fedced
Merge #1019
1019: Update llvm backend to work with latest inkwell. r=nlewycky a=nlewycky

cargo update to pick up latest inkwell branch commit.

Add lifetime annotations to Module which now takes a lifetime.

Add <'ctx> across CtxType and Intrinsics. Remove it from reference bindings.

Use ManuallyDrop to ensure that context's members are dropped before the Context.

Co-authored-by: Mark McCaskey <mark@wasmer.io>

Co-authored-by: Nick Lewycky <nick@wasmer.io>
2019-11-27 18:49:36 +00:00
Asami Doi
fd0df9946b Add "auto" backend to change the backend depending on the wasm file size 2019-11-27 20:04:35 +09:00
Nick Lewycky
31a77b0eb7 Fix llvm backend to work with latest inkwell.
cargo update to pick up latest inkwell branch commit.

Add lifetime annotations to Module which now takes a lifetime, and more lifetime annotations across intrinsics.rs.

Add <'ctx> to missing places in CtxType and Intrinsics. Remove it from reference bindings.

Use ManuallyDrop to ensure that context's members are dropped before the Context.

Co-authored-by: Mark McCaskey <mark@wasmer.io>
2019-11-26 19:59:03 -08:00
bors[bot]
dfc7163b71
Merge #1020
1020: Turn a few more assert!s that should never fire into debug_assert!s. r=nlewycky a=nlewycky

These are here to protect against errors when refactoring more than anything else.


Co-authored-by: Nick Lewycky <nick@wasmer.io>
2019-11-27 01:16:34 +00:00
Nick Lewycky
5a49fe76df Turn a few more assert!s that should never fire into debug_assert!s.
These are here to protect against errors when refactoring more than anything else.
2019-11-26 16:43:17 -08:00
bors[bot]
f6f8780934
Merge #1006
1006: fix 1005 panic sub overflow r=MarkMcCaskey a=pventuzelo

# Description

Fix issue https://github.com/wasmerio/wasmer/issues/1005

# Review

- [x] Add a short description of the the change to the CHANGELOG.md file


Co-authored-by: Patrick Ventuzelo <ventuzelo.patrick@gmail.com>
Co-authored-by: Patrick Ventuzelo <9038181+pventuzelo@users.noreply.github.com>
2019-11-27 00:36:29 +00:00
bors[bot]
53f0a9cecf
Merge #934
934: In LLVM backend, track which floats are guaranteed to be arithmetic, which makes the canonicalization a no-op. r=nlewycky a=nlewycky

# Description
This is a reimplementation of the patch in PR #651.

Extend state.rs ExtraInfo to track more information about floats. In addition to tracking whether the value has a pending canonicalization of NaNs, also track whether the value is known to be arithmetic (which includes infinities, regular values, and non-signalling NaNs (aka. "arithmetic NaNs" in the webassembly spec)). When the value is arithmetic, the correct sequence of operations to canonicalize the value is a no-op. Therefore, we create a lattice where pending+arithmetic=arithmetic.

Also, this extends the tracking to track all values, including non-SIMD integers. That's why there are more places where pending canonicalizations are applied.

Looking at c-wasm-simd128-example, this provides no performance change to the non-SIMD case (takes 58s on my noisy dev machine). The SIMD case drops from 46s to 29s.

# Review

- [ ] Add a short description of the the change to the CHANGELOG.md file


Co-authored-by: Nick Lewycky <nick@wasmer.io>
2019-11-26 20:29:31 +00:00
Nick Lewycky
ff73c5d71b Address review feedback from Mark.
Fix a bug in Operator::Select and add a comment to explain the intention.

Use derived default for ExtraInfo.

Make ExtraInfo associated functions const.

Turn two asserts into debug_asserts.
2019-11-26 12:25:03 -08:00
Nick Lewycky
d3fabe576b Add changelog entry. 2019-11-26 12:20:44 -08:00
Nick Lewycky
d1ce8ee20d Give that panic! a message. Also, make it an unreachable!. 2019-11-26 12:20:44 -08:00
Nick Lewycky
fa576093c2 Fix ExtraInfo on F32Add and similar.
We want to ignore the incoming pending NaN state (since the pending will propagate to the output if there was one on the input), and we want to add a new pending NaN state if we can (that is to say, if it isn't cancelled out by both inputs having arithmetic state). Do this by discarding the pending states on the inputs, intersecting them (to keep only the arithmetic state), then union in a pending nan state (which might do nothing, if it's arithmetic).

If the above sounds confusing, keep in mind that when a value is arithmetic, the act of performing the "NaN canonicalization" is a no-op. Thus, being arithmetic cancels out pending NaN states.
2019-11-26 12:20:44 -08:00
Nick Lewycky
a06c858087 Make ExtraInfo bitand check for pending validity.
Unfortunately, this is quite buggy. For something as simple as F32Sub, to combine two ExtraInfos, we want to add a new pending_f32_nan(), unless both of the inputs are arithmetic_f32(). In this commit, we incorrectly calculate that we don't need a pending_f32_nan if either one of the inputs was arithmetic_f32().
2019-11-26 12:20:43 -08:00
Nick Lewycky
284948b6d4 Refactor so as to convert ExtraInfo when potentially canonicalizing.
It seemed like a good idea at the time, but in practice we discard the extra info all or almost all of the time.

This also introduces a new bug. In an operation like multiply, it's valid to multiply two values, one with a pending NaN and one without. As written, in the SIMD case (because of the two kinds of pending in play), we assert.
2019-11-26 12:20:43 -08:00
Nick Lewycky
26c8fd52c8 Initial implementation of "known to be arithmetic NaN / not NaN". 2019-11-26 12:20:43 -08:00
Nick Lewycky
fafc7ad38c Add "known to not contain non-arithmetic NaNs" to ExtraInfo in LLVM backend.
Not wired up yet.
2019-11-26 12:20:43 -08:00
Patrick Ventuzelo
bb16547567 Merge branch 'ventuzelo/fix-1005-panic-sub-overflow' of github.com:pventuzelo/wasmer into ventuzelo/fix-1005-panic-sub-overflow 2019-11-26 09:18:05 +01:00
Patrick Ventuzelo
49665d5797 use checked_sub for peekn_extra 2019-11-26 09:17:13 +01:00
bors[bot]
7715be9ccb
Merge #1017
1017: Add a 'clear' method to map. r=nlewycky a=nlewycky

# Description
Motivated by work on the `nlewycky/inkwell-lifetimes` branch. No callers yet on master, but we use it on the branch and I'd like to merge in pieces.


Co-authored-by: Nick Lewycky <nick@wasmer.io>
2019-11-25 22:54:12 +00:00
Nick Lewycky
4e2440daec Add a 'clear' method to map. 2019-11-25 14:53:25 -08:00
Syrus Akbary
95ccd99809
Added ARM reference 2019-11-25 11:25:05 -08:00
bors[bot]
a69c029c21
Merge #1016
1016: [fix 962] Activate pre_validation when llvm backend selected r=nlewycky a=pventuzelo

# Description

After analysis of issue https://github.com/wasmerio/wasmer/issues/962 by me and @nlewycky, it seems to be better to validate the wasm module before calling the llvm compiler.

This change will imply potential overhead in term of execution speed using llvm backend.

I will also suggest to set pre_validation to true for Singlepass by default:
5c1c786e35/lib/runtime-core/src/codegen.rs (L263-L269)

Additional improvement proposal for wasmer API & cli command options:
* add an API `compile_without_validation`.
* add a flag option in wasmer binary to deactivate `pre_validation` for all backend.

# Review

- [x] Add a short description of the the change to the CHANGELOG.md file


Co-authored-by: Patrick Ventuzelo <ventuzelo.patrick@gmail.com>
2019-11-25 18:38:52 +00:00
Patrick Ventuzelo
7ca0430a40 activate pre_validation when llvm backend selected 2019-11-25 13:17:07 +01:00
Patrick Ventuzelo
d376b89ef0
Merge branch 'master' into ventuzelo/fix-1005-panic-sub-overflow 2019-11-25 12:05:33 +01:00
bors[bot]
5c1c786e35
Merge #1002
1002: Update the LLVM pass list. r=nlewycky a=nlewycky

# Description
Adds optimizations of loops, and inlinling and some simple interprocedural optimization.

Measured on the libsodium benchmarks, the new pass pipeline is a 2.35% geomean improvement. No major performance regressions known.

Co-authored-by: Nick Lewycky <nick@wasmer.io>
2019-11-23 05:11:44 +00:00
Syrus
9c0b6a01b2 Fixed Azure pipelines releases 2019-11-22 18:59:54 -08:00
bors[bot]
2f394c9f30
Merge #1009
1009: Fix LLVM crash on `return` with a float on the stack and a crash with gas metering. r=nlewycky a=nlewycky

# Description
This PR fixes some matters of testing. The return+float bug is caught by the LLVM verifier, which was accidentally unconditionally disabled even when running "make test check". Add a new cargo feature named "test" which is enabled by all the test crates. When the "test" feature is passed to the llvm-backend, run the LLVM verifier.

Add a new crate `llvm-backend-test` which has one test so far, to ensure that this bug with the return+float does not crash LLVM. Include this new crate in runs of `make llvm`.

Fix LLVM crash on `return` when there is a float with pending NaN canonicalization on the stack.

Now that we run the verifier in testing, we discover another bug where `internal_field()` incorrectly labels a GEP with a TBAA label instead of the intended load. Fix this by labeling the correct instruction. No new test is introduced since this is already caught with `make bench-llvm`.

# Review

- [ ] Add a short description of the the change to the CHANGELOG.md file


Co-authored-by: Nick Lewycky <nick@wasmer.io>
2019-11-23 02:18:07 +00:00
Nick Lewycky
91671d5347 Add changelog entry. 2019-11-22 17:16:54 -08:00
Nick Lewycky
3ef9f769df Fix TBAA crash with LLVM + the gas middleware. 2019-11-22 17:03:31 -08:00
Nick Lewycky
681219b06c Fix bug in LLVM lowering of 'return' when the stack has a float on it. 2019-11-22 17:03:31 -08:00
Nick Lewycky
85666fc522 Add new llvm-backend-test crate. 2019-11-22 17:03:23 -08:00
Nick Lewycky
2433d365af Add new test feature, enabled in test crates.
When test feature is enabled, turn on LLVM verifier. This was previously never enabled.
2019-11-22 16:51:44 -08:00
Syrus
8f50dab0f2 Trying to improve release process 2019-11-22 12:23:23 -08:00
Syrus Akbary
2fdf39f7fe
Merge pull request #1008 from wasmerio/release-0.11.0
Prepare for 0.11.0 release
2019-11-22 11:23:47 -08:00
Mark McCaskey
bb1e7a8d78 Prepare for 0.11.0 release 2019-11-22 11:18:06 -08:00
bors[bot]
bcd7bb4918
Merge #713
713: Add AArch64 support for singlepass. r=syrusakbary a=losfair

This PR includes:

- Support for AArch64 (ARM64) in Singlepass backend. Implemented with a combination of x86_64 instruction translation and native code generation.
- State tracing and backtraces on AArch64/Singlepass. (Tiering is not implemented for this pair because there's no other backend supporting AArch64 yet)
- Debugging tools: `BlockTrace` middleware, and support for reading states of previous WebAssembly stack frames from middlewares.

Co-authored-by: losfair <zhy20000919@hotmail.com>
2019-11-22 18:24:57 +00:00
Syrus
183dd328d6 Make travis work only on test branches 2019-11-22 09:55:31 -08:00
Syrus
a691d3454c Skip serializing on singlepass 2019-11-22 09:49:17 -08:00
Patrick Ventuzelo
cd0da74b33 add changelog entry 2019-11-22 17:25:17 +01:00
Patrick Ventuzelo
2261f8b449 cargo fmt 2019-11-22 17:21:15 +01:00
Patrick Ventuzelo
508004088e check stack len before getting value 2019-11-22 17:10:54 +01:00