Compare commits

..

39 Commits

Author SHA1 Message Date
Cyrill Leutwiler 52bf16a383 wip
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
2025-04-17 12:43:26 +02:00
Cyrill Leutwiler cf2fd2f1e8 pass assignment ptr
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
2025-04-17 09:10:30 +02:00
xermicus 5003f3e9ac llvm-context: alloca at the function entry if possible (#283)
Closes  #48

Change the code size test to no longer emit debug info as to get a more
accurate picture.

---------

Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
2025-04-15 15:22:24 +02:00
xermicus 431b5a2ce5 llvm-context: lazy handling of function arguments and immutable data (#282)
- Lazily load function arguments so that they can be passed as pointers.
- Lazily call the immutable store function to avoid storing zero sized
immutable data.

---------

Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
2025-04-14 15:54:59 +02:00
xermicus ad3315346c release resolc-0.1.0-dev.13 (#279) 2025-04-08 09:10:13 +02:00
dependabot[bot] 516f79ee0f Bump tokio from 1.43.0 to 1.44.2 (#281)
Bumps [tokio](https://github.com/tokio-rs/tokio) from 1.43.0 to 1.44.2.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a
href="https://github.com/tokio-rs/tokio/releases">tokio's
releases</a>.</em></p>
<blockquote>
<h2>Tokio v1.44.2</h2>
<p>This release fixes a soundness issue in the broadcast channel. The
channel
accepts values that are <code>Send</code> but <code>!Sync</code>.
Previously, the channel called
<code>clone()</code> on these values without synchronizing. This release
fixes the channel
by synchronizing calls to <code>.clone()</code> (Thanks Austin Bonander
for finding and
reporting the issue).</p>
<h3>Fixed</h3>
<ul>
<li>sync: synchronize <code>clone()</code> call in broadcast channel (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7232">#7232</a>)</li>
</ul>
<p><a
href="https://redirect.github.com/tokio-rs/tokio/issues/7232">#7232</a>:
<a
href="https://redirect.github.com/tokio-rs/tokio/pull/7232">tokio-rs/tokio#7232</a></p>
<h2>Tokio v1.44.1</h2>
<h1>1.44.1 (March 13th, 2025)</h1>
<h3>Fixed</h3>
<ul>
<li>rt: skip defer queue in <code>block_in_place</code> context (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7216">#7216</a>)</li>
</ul>
<p><a
href="https://redirect.github.com/tokio-rs/tokio/issues/7216">#7216</a>:
<a
href="https://redirect.github.com/tokio-rs/tokio/pull/7216">tokio-rs/tokio#7216</a></p>
<h2>Tokio v1.44.0</h2>
<h1>1.44.0 (March 7th, 2025)</h1>
<p>This release changes the <code>from_std</code> method on sockets to
panic if a blocking socket is provided. We determined this change is not
a breaking change as Tokio is not intended to operate using blocking
sockets. Doing so results in runtime hangs and should be considered a
bug. Accidentally passing a blocking socket to Tokio is one of the most
common user mistakes. If this change causes an issue for you, please
comment on <a
href="https://redirect.github.com/tokio-rs/tokio/issues/7172">#7172</a>.</p>
<h3>Added</h3>
<ul>
<li>coop: add <code>task::coop</code> module (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7116">#7116</a>)</li>
<li>process: add <code>Command::get_kill_on_drop()</code> (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7086">#7086</a>)</li>
<li>sync: add <code>broadcast::Sender::closed</code> (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/6685">#6685</a>,
<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7090">#7090</a>)</li>
<li>sync: add <code>broadcast::WeakSender</code> (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7100">#7100</a>)</li>
<li>sync: add <code>oneshot::Receiver::is_empty()</code> (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7153">#7153</a>)</li>
<li>sync: add <code>oneshot::Receiver::is_terminated()</code> (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7152">#7152</a>)</li>
</ul>
<h3>Fixed</h3>
<ul>
<li>fs: empty reads on <code>File</code> should not start a background
read (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7139">#7139</a>)</li>
<li>process: calling <code>start_kill</code> on exited child should not
fail (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7160">#7160</a>)</li>
<li>signal: fix <code>CTRL_CLOSE</code>, <code>CTRL_LOGOFF</code>,
<code>CTRL_SHUTDOWN</code> on windows (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7122">#7122</a>)</li>
<li>sync: properly handle panic during mpsc drop (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7094">#7094</a>)</li>
</ul>
<h3>Changes</h3>
<ul>
<li>runtime: clean up magic number in registration set (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7112">#7112</a>)</li>
<li>coop: make coop yield using waker defer strategy (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7185">#7185</a>)</li>
<li>macros: make <code>select!</code> budget-aware (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7164">#7164</a>)</li>
<li>net: panic when passing a blocking socket to <code>from_std</code>
(<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7166">#7166</a>)</li>
<li>io: clean up buffer casts (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7142">#7142</a>)</li>
</ul>
<!-- raw HTML omitted -->
</blockquote>
<p>... (truncated)</p>
</details>
<details>
<summary>Commits</summary>
<ul>
<li><a
href="https://github.com/tokio-rs/tokio/commit/ec4b1d7215a3e1e91797ad3fb6ba0f7c7f3d2566"><code>ec4b1d7</code></a>
chore: forward port 1.43.x</li>
<li><a
href="https://github.com/tokio-rs/tokio/commit/e3c3a56718d201fb7bb430567f05fbb64b2ef082"><code>e3c3a56</code></a>
Merge branch 'tokio-1.43.x' into forward-port-1.43.x</li>
<li><a
href="https://github.com/tokio-rs/tokio/commit/a7b658c35bd40f6811e557aeb97cbb361b612c56"><code>a7b658c</code></a>
chore: prepare Tokio v1.43.1 release</li>
<li><a
href="https://github.com/tokio-rs/tokio/commit/c1c8d1033d637d7027fdc137ec8008c5801cbc0d"><code>c1c8d10</code></a>
Merge remote-tracking branch 'origin/tokio-1.38.x' into
forward-port-1.38.x</li>
<li><a
href="https://github.com/tokio-rs/tokio/commit/aa303bc2051f7c21b48bb7bfcafe8fd4f39afd21"><code>aa303bc</code></a>
chore: prepare Tokio v1.38.2 release</li>
<li><a
href="https://github.com/tokio-rs/tokio/commit/7b6ccb515ff067151ed62db835f735e5653f8784"><code>7b6ccb5</code></a>
chore: backport CI fixes</li>
<li><a
href="https://github.com/tokio-rs/tokio/commit/4b174ce2c95fe1d1a217917db93fcc935e17e0da"><code>4b174ce</code></a>
sync: fix cloning value when receiving from broadcast channel</li>
<li><a
href="https://github.com/tokio-rs/tokio/commit/d413c9c02af8f2b4fea14b769b86484b12f46595"><code>d413c9c</code></a>
chore: prepare Tokio v1.44.1 (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7217">#7217</a>)</li>
<li><a
href="https://github.com/tokio-rs/tokio/commit/addbfb9204be25a8621feb3f20b44a7c1f00edbd"><code>addbfb9</code></a>
rt: skip defer queue in <code>block_in_place</code> context (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7216">#7216</a>)</li>
<li><a
href="https://github.com/tokio-rs/tokio/commit/8182ecf2628d5e80dac52b8ed1ea466dbb0925b9"><code>8182ecf</code></a>
chore: prepare Tokio v1.44.0 (<a
href="https://redirect.github.com/tokio-rs/tokio/issues/7202">#7202</a>)</li>
<li>Additional commits viewable in <a
href="https://github.com/tokio-rs/tokio/compare/tokio-1.43.0...tokio-1.44.2">compare
view</a></li>
</ul>
</details>
<br />


[![Dependabot compatibility
score](https://dependabot-badges.githubapp.com/badges/compatibility_score?dependency-name=tokio&package-manager=cargo&previous-version=1.43.0&new-version=1.44.2)](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)

Dependabot will resolve any conflicts with this PR as long as you don't
alter it yourself. You can also trigger a rebase manually by commenting
`@dependabot rebase`.

[//]: # (dependabot-automerge-start)
[//]: # (dependabot-automerge-end)

---

<details>
<summary>Dependabot commands and options</summary>
<br />

You can trigger Dependabot actions by commenting on this PR:
- `@dependabot rebase` will rebase this PR
- `@dependabot recreate` will recreate this PR, overwriting any edits
that have been made to it
- `@dependabot merge` will merge this PR after your CI passes on it
- `@dependabot squash and merge` will squash and merge this PR after
your CI passes on it
- `@dependabot cancel merge` will cancel a previously requested merge
and block automerging
- `@dependabot reopen` will reopen this PR if it is closed
- `@dependabot close` will close this PR and stop Dependabot recreating
it. You can achieve the same result by closing it manually
- `@dependabot show <dependency name> ignore conditions` will show all
of the ignore conditions of the specified dependency
- `@dependabot ignore this major version` will close this PR and stop
Dependabot creating any more for this major version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this minor version` will close this PR and stop
Dependabot creating any more for this minor version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this dependency` will close this PR and stop
Dependabot creating any more for this dependency (unless you reopen the
PR or upgrade to it yourself)
You can disable automated security fix PRs for this repo from the
[Security Alerts
page](https://github.com/paritytech/revive/network/alerts).

</details>

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: xermicus <cyrill@parity.io>
2025-04-08 08:26:06 +02:00
Alexander Samusev 9f5443b6d6 Remove path from release json and add sha256 (#280)
PR addresses
https://github.com/paritytech/revive/issues/162#issuecomment-2783173447
2025-04-08 08:18:35 +02:00
Alexander Samusev 0dafc779ce ci: update release flow and publish list.json (#276)
This pull request adds changes described in **this comment**:

- `.github/scripts/json_generator.py` - a small script that generates
several json files for different platforms.
-
[generate_versions.yml](https://github.com/paritytech/revive/compare/as-release-json?expand=1#diff-2aee05b96020ac60943e6dfcb30597e53898f31542aeb570468b970d9a13a5a6)
- the workflow that runs when release is published, creates info.json
files and pushes them into resolc-bin repo
- `js/build.js` is adjusted in order to set `RESOLC_WASM_URI` from env
variable
- ⚠️ Release workflow is changed: 
  - In PRs and main branch it'll only build artifacts
- Release will happen automatically only on the `v*` tag push. This is
needed for revive_web.js to have the necessary `RESOLC_WASM_URI`
- workflow will check that version in Cargo.toml is the same as the tag
when the new tag is pushed

cc https://github.com/paritytech/revive/issues/162
cc https://github.com/paritytech/devops/issues/3890
2025-04-07 13:39:56 +02:00
xermicus 7d8fa75a0f storage keys and values should be big endian (#277)
Storage keys and values are big endian. Keeping them LE was a pre-mature
optimization because for the contract itself it this is a no-op and thus
not observable. However we should consider the storage layout as part of
the contract ABI. The endianness of transient storage values are still
kept as-is.

---------

Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
2025-04-07 10:34:44 +02:00
dependabot[bot] 80f94b5c76 Bump openssl from 0.10.71 to 0.10.72 (#278)
Bumps [openssl](https://github.com/sfackler/rust-openssl) from 0.10.71
to 0.10.72.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a
href="https://github.com/sfackler/rust-openssl/releases">openssl's
releases</a>.</em></p>
<blockquote>
<h2>openssl-v0.10.72</h2>
<h2>What's Changed</h2>
<ul>
<li>make set_rsa_oaep_md visible to boringssl config by <a
href="https://github.com/frncs-rss"><code>@​frncs-rss</code></a> in <a
href="https://redirect.github.com/sfackler/rust-openssl/pull/2372">sfackler/rust-openssl#2372</a></li>
<li>Fix typo in openssl-sys build script by <a
href="https://github.com/rushilmehra"><code>@​rushilmehra</code></a> in
<a
href="https://redirect.github.com/sfackler/rust-openssl/pull/2375">sfackler/rust-openssl#2375</a></li>
<li>Unify the two BoringSSL codepaths a bit and simplify init by <a
href="https://github.com/davidben"><code>@​davidben</code></a> in <a
href="https://redirect.github.com/sfackler/rust-openssl/pull/2377">sfackler/rust-openssl#2377</a></li>
<li>pkey_ctx: Fix link to the corresponding OpenSSL function by <a
href="https://github.com/Jakuje"><code>@​Jakuje</code></a> in <a
href="https://redirect.github.com/sfackler/rust-openssl/pull/2378">sfackler/rust-openssl#2378</a></li>
<li>fix test on MSRV by <a
href="https://github.com/alex"><code>@​alex</code></a> in <a
href="https://redirect.github.com/sfackler/rust-openssl/pull/2383">sfackler/rust-openssl#2383</a></li>
<li>Add support for AWS-LC to openssl and openssl-sys crates by <a
href="https://github.com/skmcgrail"><code>@​skmcgrail</code></a> in <a
href="https://redirect.github.com/sfackler/rust-openssl/pull/1805">sfackler/rust-openssl#1805</a></li>
<li>Enable additional capabilities for AWS-LC by <a
href="https://github.com/skmcgrail"><code>@​skmcgrail</code></a> in <a
href="https://redirect.github.com/sfackler/rust-openssl/pull/2386">sfackler/rust-openssl#2386</a></li>
<li>Use --experimental with bindgen-cli with aws-lc build by <a
href="https://github.com/skmcgrail"><code>@​skmcgrail</code></a> in <a
href="https://redirect.github.com/sfackler/rust-openssl/pull/2389">sfackler/rust-openssl#2389</a></li>
<li>Fixed two UAFs and bumped versions for release by <a
href="https://github.com/alex"><code>@​alex</code></a> in <a
href="https://redirect.github.com/sfackler/rust-openssl/pull/2390">sfackler/rust-openssl#2390</a></li>
</ul>
<h2>New Contributors</h2>
<ul>
<li><a href="https://github.com/Jakuje"><code>@​Jakuje</code></a> made
their first contribution in <a
href="https://redirect.github.com/sfackler/rust-openssl/pull/2378">sfackler/rust-openssl#2378</a></li>
<li><a href="https://github.com/skmcgrail"><code>@​skmcgrail</code></a>
made their first contribution in <a
href="https://redirect.github.com/sfackler/rust-openssl/pull/1805">sfackler/rust-openssl#1805</a></li>
</ul>
<p><strong>Full Changelog</strong>: <a
href="https://github.com/sfackler/rust-openssl/compare/openssl-v0.10.71...openssl-v0.10.72">https://github.com/sfackler/rust-openssl/compare/openssl-v0.10.71...openssl-v0.10.72</a></p>
</blockquote>
</details>
<details>
<summary>Commits</summary>
<ul>
<li><a
href="https://github.com/sfackler/rust-openssl/commit/87085bd67896b7f92e6de35d081f607a334beae4"><code>87085bd</code></a>
Merge pull request <a
href="https://redirect.github.com/sfackler/rust-openssl/issues/2390">#2390</a>
from alex/uaf-fix</li>
<li><a
href="https://github.com/sfackler/rust-openssl/commit/d1a12e21573e95727b2e38b8b65273cb389be7e4"><code>d1a12e2</code></a>
Fixed two UAFs and bumped versions for release</li>
<li><a
href="https://github.com/sfackler/rust-openssl/commit/7c7b2e6c9f95e77e56ab37af70b16de75beff387"><code>7c7b2e6</code></a>
Merge pull request <a
href="https://redirect.github.com/sfackler/rust-openssl/issues/2389">#2389</a>
from skmcgrail/aws-lc-follow-up</li>
<li><a
href="https://github.com/sfackler/rust-openssl/commit/34a477bff20cbe43492915338d3c12597430c345"><code>34a477b</code></a>
Use --experimental with bindgen-cli with aws-lc build</li>
<li><a
href="https://github.com/sfackler/rust-openssl/commit/d4bf0710640e4725b8b237968040aef3e5f4ab9a"><code>d4bf071</code></a>
Merge pull request <a
href="https://redirect.github.com/sfackler/rust-openssl/issues/2386">#2386</a>
from skmcgrail/aws-lc-follow-up</li>
<li><a
href="https://github.com/sfackler/rust-openssl/commit/a86bf670c4cba3ee5531838b52419356791d966e"><code>a86bf67</code></a>
Remove comment</li>
<li><a
href="https://github.com/sfackler/rust-openssl/commit/705dbfb2ee3f2d7151ff313d840bf558435d4379"><code>705dbfb</code></a>
Fix test</li>
<li><a
href="https://github.com/sfackler/rust-openssl/commit/e0df413d46a89303c42e15bf7d4566193b242466"><code>e0df413</code></a>
Skip final call for LibreSSL 4.1.0 for CCM mode</li>
<li><a
href="https://github.com/sfackler/rust-openssl/commit/2f1164b5e838d3665dd10a9fac19e22174289ea3"><code>2f1164b</code></a>
Enable additional capabilities for AWS-LC</li>
<li><a
href="https://github.com/sfackler/rust-openssl/commit/dde9ffb36071249ff98474eec853fd830aea44f5"><code>dde9ffb</code></a>
Merge pull request <a
href="https://redirect.github.com/sfackler/rust-openssl/issues/1805">#1805</a>
from skmcgrail/aws-lc-support-final</li>
<li>Additional commits viewable in <a
href="https://github.com/sfackler/rust-openssl/compare/openssl-v0.10.71...openssl-v0.10.72">compare
view</a></li>
</ul>
</details>
<br />


[![Dependabot compatibility
score](https://dependabot-badges.githubapp.com/badges/compatibility_score?dependency-name=openssl&package-manager=cargo&previous-version=0.10.71&new-version=0.10.72)](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)

Dependabot will resolve any conflicts with this PR as long as you don't
alter it yourself. You can also trigger a rebase manually by commenting
`@dependabot rebase`.

[//]: # (dependabot-automerge-start)
[//]: # (dependabot-automerge-end)

---

<details>
<summary>Dependabot commands and options</summary>
<br />

You can trigger Dependabot actions by commenting on this PR:
- `@dependabot rebase` will rebase this PR
- `@dependabot recreate` will recreate this PR, overwriting any edits
that have been made to it
- `@dependabot merge` will merge this PR after your CI passes on it
- `@dependabot squash and merge` will squash and merge this PR after
your CI passes on it
- `@dependabot cancel merge` will cancel a previously requested merge
and block automerging
- `@dependabot reopen` will reopen this PR if it is closed
- `@dependabot close` will close this PR and stop Dependabot recreating
it. You can achieve the same result by closing it manually
- `@dependabot show <dependency name> ignore conditions` will show all
of the ignore conditions of the specified dependency
- `@dependabot ignore this major version` will close this PR and stop
Dependabot creating any more for this major version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this minor version` will close this PR and stop
Dependabot creating any more for this minor version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this dependency` will close this PR and stop
Dependabot creating any more for this dependency (unless you reopen the
PR or upgrade to it yourself)
You can disable automated security fix PRs for this repo from the
[Security Alerts
page](https://github.com/paritytech/revive/network/alerts).

</details>

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2025-04-05 12:25:52 +02:00
xermicus dab29bc89b bugfix missing byte swap for the create2 salt value (#272)
Found in https://github.com/paritytech/contract-issues/issues/45, thanks
@albertov19 and @sekisamu

---------

Signed-off-by: xermicus <cyrill@parity.io>
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
2025-04-03 15:20:22 +02:00
xermicus 313c033261 fix Rust 1.86.0 clippy (#273)
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
2025-04-03 15:03:49 +02:00
xermicus 87c1d7a8be Suport passing arbitrary llvm arguments (#271)
- Support for passing LLVM command line options via the prcoess input or
providing one or more `--llvm-arg='..'` resolc CLI flag. This allows
more fine-grained control over the LLVM backend configuration.
- Make LLVM initialization idempotent.

---------

Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
2025-04-03 13:21:00 +02:00
xermicus 8a98346a9c Fix CI badge link (#270) 2025-04-01 07:52:39 +02:00
Alexander Samusev 1b4d2c6bb8 ci: modify release according to #162 (#269)
PR modifies release workflow according to changes requested in
https://github.com/paritytech/revive/issues/162#issuecomment-2751287953
2025-03-26 12:06:51 +01:00
xermicus 004b8ac16c Update release.yml (#265)
The Release workflow used to work with a fetch-depth: 0 - not sure why this is required but yolo-ing it in for a test
2025-03-20 20:07:04 +01:00
xermicus 497dae2494 factor out solc JSON interface crate (#264)
The differential testing framework will make a second consumer. There
seems to be no re-usable Rust crate for this. But we already have
everything here, just needs a small refactor to make it fully re-usable.

- Mostly decouple the solc JSON-input-output interface types from the
`solidity` frontend crate
- Expose the JSON-input-output interface types in a dedicated crate

---------

Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
2025-03-20 17:11:40 +01:00
Pavlo Khrystenko 36ea69b50f Add --supported-solc-versions CLI command (#260)
### Description 

Adds `--supported-solc-versions` cli command to return a semver range of
supported `solc` versions.
Is a hack until #162 is implemented.

---------

Co-authored-by: xermicus <cyrill@parity.io>
2025-03-19 16:07:48 +01:00
xermicus 2bbc5d713d remove support for webkit (#262)
Signed-off-by: xermicus <cyrill@parity.io>
2025-03-18 15:00:55 +01:00
xermicus 4a9b651235 Support solc v0.8.29 (#261)
No observable changes w.r.t. YUL compilation.
- Mark solc v0.8.29 as the latest supported version
- Add integration test for solc v0.8.29 specific feature
- Use the latest version on CI
- Drive-by fix linter complaints and a two types in the CI yaml files

---------

Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
Signed-off-by: xermicus <cyrill@parity.io>
2025-03-18 13:01:15 +01:00
Alexander Theißen 66f9a4d64f Release Windows Version (#251)
With LLVM working only minor changes were necessary to get resolc
running on Windows.

Release in my branch here:
https://github.com/paritytech/revive-alex-workflowtest/releases/tag/v0.1.0-dev.12
2025-03-01 13:49:04 +01:00
xermicus 76f4cf71d6 ci: re-enable the machete (#249)
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
2025-02-28 16:05:53 +01:00
xermicus 77e0344d80 llvm-context: remove dead code (#247)
- remove the __sha3 function symbol: this is provided by the pallet
- remove the storage address spaces: they are not mapped into memory

---------

Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
Co-authored-by: Alexander Theißen <alex.theissen@me.com>
2025-02-28 15:38:42 +01:00
Alexander Theißen 2fb8beee62 Build LLVM for Windows (#248)
This PR changes the CI build scripts to also build LLVM for windows.
**It doesn't build `revive` itself for windows**. This will come in a
follow up. But once we have a LLVM binary release the turn around time
will be much quicker for experimenting with the revive windows build.

I manually uploaded the release those changes produce
[here](https://github.com/paritytech/revive-alex-workflowtest/releases/tag/llvm-18.1.8-revive.22f3ceb).
This enables this PR's CI to find the proper release. This is necessary
because I am also making changes to the folder structure and artifact
naming that the other CI jobs are depending on.

Releases generated from this branch can be inspected here:
https://github.com/paritytech/revive-alex-workflowtest/releases/tag/v0.1.0-dev.12

Summary of changes:
- Change `llvm-builder` to use MSVC toolchain on windows
- Fix `llvm-builder` to work with `.exe` files
- Unify the llvm release jobs into a single one. This removed a lot of
copy pasted code and also speeds up the build by giving each their own
runner.
- Use the LLVM target triple to name the binary releases instead of an
ad-hoc naming convention
- Remove the nested folder hierarchy inside the llvm release. Its just
now a single folder `llvm-<target>` that contains the toolchain.
- Give jobs and workflows consistent names
- Replace all runners bei their `*-latest` counterpart
- Only use `parity-large` to build llvm now. All other jobs use github
runners
2025-02-28 15:06:03 +01:00
xermicus 93788e72e9 integration: additional mcopy test (#245)
Closes #237

Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
2025-02-27 11:39:31 +01:00
xermicus 296a226d0b integration: add delegate call corner case test (#243)
Closes #235

---------

Signed-off-by: xermicus <cyrill@parity.io>
2025-02-27 11:10:00 +01:00
xermicus 84deb3a29d integration: add function return type test (#244)
Closes #236
2025-02-27 10:37:37 +01:00
Evgeny Snitko ee064671e0 Release archive path fix (#241) 2025-02-26 16:40:49 +01:00
xermicus 63dfd046e5 Update RELEASE.md (#239) 2025-02-26 09:38:00 +01:00
xermicus 08f341ccc1 release resolc v0.1.0-dev.12 (#238)
Signed-off-by: xermicus <cyrill@parity.io>
2025-02-26 09:36:28 +01:00
xermicus a07968205b llvm-context: modularize compiler builtin functions (#234)
- Add the revive runtime function interface to minimize boiler plate
code.
- Outline heavily repeated code into dedicated functions to bring down
code size.
- The code size tests builds optimized for size.
- Function attributes are passed as slices.

This significantly brings down the code size for all OpenZeppelin wizard
contracts (using all possible features) compiled against OpenZeppelin
`v5.0.0` with size optimizations.

|contract|| `-Oz` main | `-Oz` PR || `-O3` main | `-O3` PR |
|-|-|-|-|-|-|-|
|erc1155.sol||100K|67K||114K|147K|
|erc20.sol||120K|90K||160K|191K|
|erc721.sol||128K|101K||178K|214K|
|governor.sol||226K|165K||293K|349K|
|rwa.sol||116K|85K||154K|185K|
|stable.sol||116K|86K||155K|192K|

On the flip side this introduces a heavy penalty for cycle optimized
builds. Setting the no-inline attributes for cycle optimized builds
helps a lot but heavily penalizes runtime speed (LLVM does not yet
inline everything properly - to be investigated later on).

Next steps:
- Modularize more functions
- Refactor the YUL function arguments to use pointers instead of values
- Afterwards check if LLVM still has trouble inline-ing properly on O3
or set the no-inline attribute if it does not penalize runtime
performance too bad.
2025-02-25 16:47:01 +01:00
Evgeny Snitko 7ffe64ed7c CI improvements (#230)
- llvm artifacts search, download and extract now in `get-llvm` action 
- `get-emsdk` action - clone, install, activate emsdk, so we don't need
to build `revive-builder` for this
- switch workflows to new actions
- `concurrency` for remaining workflows (cancel run if new run is
triggered)
- `target_commitish` for main release, fixes release commit 
- Run release workflow in PR without creating a release if PR is labeled
with `release-test`

---------

Co-authored-by: cornholio <0@mcornholio.ru>
Co-authored-by: Alexander Theißen <alex.theissen@me.com>
Co-authored-by: xermicus <cyrill@parity.io>
Co-authored-by: xermicus <bigcyrill@hotmail.com>
2025-02-25 14:33:59 +01:00
Yuri Volkov f1bce4fe3f llvm-builder testing workflow (#164)
Co-authored-by: xermicus <cyrill@parity.io>
2025-02-25 12:39:03 +01:00
xermicus 9efbb4c0b4 ci: temporarily disable the machete (#233) 2025-02-25 09:40:40 +01:00
xermicus bcc8fa892c update dependencies (#229) 2025-02-24 12:45:05 +01:00
Pavlo Khrystenko 89ec25da65 Add per file output selection for --standard-json mode and some other small foundry support bits. (#228) 2025-02-24 12:44:22 +01:00
xermicus 75a83af4da ci: clippy (#227)
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
2025-02-21 16:43:09 +01:00
xermicus 9c330ef8fc llvm-builder: only run expensive tests on linux (#226)
Signed-off-by: Cyrill Leutwiler <bigcyrill@hotmail.com>
2025-02-21 16:38:44 +01:00
Yuri Volkov 687cec31ef Using released LLVM in revive build (#220) 2025-02-21 16:30:26 +01:00
132 changed files with 4528 additions and 2688 deletions
+19
View File
@@ -0,0 +1,19 @@
name: "Get Emscripten SDK"
inputs:
version:
description: ""
required: false
default: "3.1.64"
runs:
using: "composite"
steps:
- name: install emsdk
shell: bash
run: |
git clone https://github.com/emscripten-core/emsdk.git ./emsdk/
cd emsdk
git checkout tags/${{ inputs.version }}
./emsdk install ${{ inputs.version }}
./emsdk activate ${{ inputs.version }}
+70
View File
@@ -0,0 +1,70 @@
# example:
# - uses: ./.github/actions/get-llvm
# with:
# target: x86_64-unknown-linux-gnu
name: "Download LLVM"
inputs:
target:
required: true
runs:
using: "composite"
steps:
- name: find asset
id: find
uses: actions/github-script@v7
env:
target: ${{ inputs.target }}
with:
result-encoding: string
script: |
let page = 1;
let releases = [];
let releasePrefix = "llvm-"
let target = process.env.target
do {
const res = await github.rest.repos.listReleases({
owner: context.repo.owner,
repo: context.repo.repo,
per_page: 50,
page,
});
releases = res.data
releases.sort((a, b) => {
return (a.published_at < b.published_at) ? 1 : ((a.published_at > b.published_at) ? -1 : 0);
});
let llvmLatestRelease = releases.find(release => {
return release.tag_name.startsWith(releasePrefix);
});
if (llvmLatestRelease){
let asset = llvmLatestRelease.assets.find(asset =>{
return asset.name.includes(target);
});
if (!asset){
core.setFailed(`Artifact for '${target}' not found in release ${llvmLatestRelease.tag_name} (${llvmLatestRelease.html_url})`);
process.exit();
}
return asset.browser_download_url;
}
page++;
} while(releases.length > 0);
core.setFailed(`No LLVM releases with '${releasePrefix}' atifacts found! Please release LLVM before running this workflow.`);
process.exit();
- name: download
shell: bash
run: |
curl -sSLo llvm.tar.gz ${{ steps.find.outputs.result }}
- name: unpack
shell: bash
run: |
tar -xf llvm.tar.gz
rm llvm.tar.gz
+36
View File
@@ -0,0 +1,36 @@
name: "Install Solidity Compiler"
description: "Installs the Ethereum solc Solidity compiler frontend executable"
runs:
using: "composite"
steps:
- name: Figure out Solc Download URL
shell: bash
run: |
if [[ "${{ runner.os }}" == "Linux" ]]; then
echo "SOLC_NAME=solc-static-linux" >> $GITHUB_ENV
elif [[ "${{ runner.os }}" == "Windows" ]]; then
echo "SOLC_NAME=solc-windows.exe" >> $GITHUB_ENV
else
echo "SOLC_NAME=solc-macos" >> $GITHUB_ENV
fi
- name: Download Solc
shell: bash
run: |
mkdir -p solc
curl -sSL --output solc/solc https://github.com/ethereum/solidity/releases/download/v0.8.29/${SOLC_NAME}
- name: Make Solc Executable
if: ${{ runner.os == 'Windows' }}
shell: bash
run: |
echo "$(pwd -W)\\solc" >> $GITHUB_PATH
mv solc/solc solc/solc.exe
- name: Make Solc Executable
if: ${{ runner.os != 'Windows' }}
shell: bash
run: |
echo "$(pwd)/solc" >> $GITHUB_PATH
chmod +x solc/solc
+165
View File
@@ -0,0 +1,165 @@
import os
import sys
import json
import requests
def validate_github_token():
"""Validate that GITHUB_TOKEN environment variable is set."""
if 'GITHUB_TOKEN' not in os.environ:
print("Error: GITHUB_TOKEN environment variable is not set.")
sys.exit(1)
def fetch_release_data(repo, tag):
"""Fetch release data from GitHub API."""
url = f"https://api.github.com/repos/{repo}/releases/tags/{tag}"
headers = {
'Authorization': f"Bearer {os.environ['GITHUB_TOKEN']}",
'Accept': 'application/vnd.github+json',
'X-GitHub-Api-Version': '2022-11-28'
}
try:
response = requests.get(url, headers=headers)
response.raise_for_status()
return response.json()
except requests.RequestException as e:
print(f"Error fetching release data: {e}")
sys.exit(1)
def fetch_checksum_file(release_data):
"""
Fetch the checksum.txt file from the release assets
and parse it into a dictionary mapping file names to their SHA256 checksums.
"""
checksums = {}
# Find the checksum.txt asset URL
checksum_asset = None
for asset in release_data['assets']:
if asset['name'] == 'checksums.txt':
checksum_asset = asset
break
if not checksum_asset:
print("Warning: checksum.txt file not found in release assets.")
return checksums
# Download the checksum file
headers = {
'Authorization': f"Bearer {os.environ['GITHUB_TOKEN']}",
'Accept': 'application/octet-stream'
}
try:
response = requests.get(checksum_asset['browser_download_url'], headers=headers)
response.raise_for_status()
# Parse checksum file
for line in response.text.splitlines():
if line.strip():
checksum, filename = line.strip().split(None, 1)
checksums[filename] = checksum
return checksums
except requests.RequestException as e:
print(f"Error fetching checksum file: {e}")
return checksums
except Exception as e:
print(f"Error parsing checksum file: {e}")
return checksums
def extract_build_hash(target_commitish):
"""Extract the first 8 characters of the commit hash."""
return f"commit.{target_commitish[:8]}"
def generate_asset_json(release_data, asset, checksums):
"""Generate JSON for a specific asset."""
version = release_data['tag_name'].lstrip('v')
build = extract_build_hash(release_data['target_commitish'])
long_version = f"{version}+{build}"
# Get SHA256 checksum if available
sha256 = checksums.get(asset['name'], "")
return {
"name": asset['name'],
"version": version,
"build": build,
"longVersion": long_version,
"url": asset['browser_download_url'],
"sha256": sha256,
"firstSolcVersion": os.environ.get("FIRST_SOLC_VERSION", ""),
"lastSolcVersion": os.environ.get("LAST_SOLC_VERSION", "")
}
def save_platform_json(platform_folder, asset_json, tag):
"""Save asset JSON and update list.json for a specific platform."""
# Create platform folder if it doesn't exist
os.makedirs(platform_folder, exist_ok=True)
# Update or create list.json
list_file_path = os.path.join(platform_folder, "list.json")
if os.path.exists(list_file_path):
with open(list_file_path, 'r') as f:
try:
list_data = json.load(f)
except json.JSONDecodeError:
list_data = {"builds": [], "releases": {}, "latestRelease": ""}
else:
list_data = {"builds": [], "releases": {}, "latestRelease": ""}
# Remove any existing entry with the same path
list_data['builds'] = [
build for build in list_data['builds']
if build['version'] != asset_json['version']
]
# Add the new build
list_data['builds'].append(asset_json)
# Update releases
version = asset_json['version']
list_data['releases'][version] = f"{asset_json['name']}+{asset_json['longVersion']}"
# Update latest release
list_data['latestRelease'] = version
with open(list_file_path, 'w') as f:
json.dump(list_data, f, indent=4)
def main():
# Validate arguments
if len(sys.argv) != 3:
print("Usage: python script.py <repo> <tag>")
sys.exit(1)
repo, tag = sys.argv[1], sys.argv[2]
# Validate GitHub token
validate_github_token()
# Fetch release data
release_data = fetch_release_data(repo, tag)
# Fetch checksums
checksums = fetch_checksum_file(release_data)
# Mapping of asset names to platform folders
platform_mapping = {
'resolc-x86_64-unknown-linux-musl': 'linux',
'resolc-universal-apple-darwin': 'macos',
'resolc-x86_64-pc-windows-msvc.exe': 'windows',
'resolc_web.js': 'wasm'
}
# Process each asset
for asset in release_data['assets']:
platform_name = platform_mapping.get(asset['name'])
if platform_name:
platform_folder = os.path.join(platform_name)
asset_json = generate_asset_json(release_data, asset, checksums)
save_platform_json(platform_folder, asset_json, tag)
print(f"Processed {asset['name']} for {platform_name}")
if __name__ == "__main__":
main()
-127
View File
@@ -1,127 +0,0 @@
name: Build revive-wasm
on:
push:
branches: ["main"]
pull_request:
branches: ["main"]
workflow_dispatch:
env:
CARGO_TERM_COLOR: always
REVIVE_WASM_INSTALL_DIR: ${{ github.workspace }}/target/wasm32-unknown-emscripten/release
jobs:
build-revive-wasm:
runs-on: parity-large
defaults:
run:
shell: bash
steps:
- uses: actions/checkout@v4
- name: Cache LLVM build
id: cache-llvm
uses: actions/cache@v3
with:
path: |
target-llvm/emscripten/target-final
target-llvm/gnu/target-final
# Use a unique key based on LLVM version or configuration files to avoid cache invalidation
key: llvm-build-${{ runner.os }}-${{ hashFiles('LLVM.lock', 'Cargo.toml', 'Cargo.lock', 'crates/llvm-builder/**', '.github/workflows/build-revive-wasm.yml') }}
- name: Install system dependencies
run: |
sudo apt-get update && sudo apt-get install -y cmake ninja-build curl git libssl-dev pkg-config clang lld
- name: Install Rust stable toolchain
uses: actions-rs/toolchain@v1
with:
profile: minimal
toolchain: stable
components: rust-src
target: wasm32-unknown-emscripten
- name: Install LLVM build dependencies
run: |
make install-llvm-builder
revive-llvm --target-env emscripten clone
- name: Setup revive environment variables
run: |
echo "LLVM_SYS_181_PREFIX=$(pwd)/target-llvm/gnu/target-final" >> $GITHUB_ENV
echo "REVIVE_LLVM_TARGET_PREFIX=$(pwd)/target-llvm/emscripten/target-final" >> $GITHUB_ENV
- name: Build host LLVM
if: steps.cache-llvm.outputs.cache-hit != 'true'
run: |
revive-llvm build --llvm-projects lld --llvm-projects clang
- name: Build target LLVM
if: steps.cache-llvm.outputs.cache-hit != 'true'
run: |
source emsdk/emsdk_env.sh
revive-llvm --target-env emscripten build --llvm-projects lld
- run: |
rustup show
cargo --version
rustup +nightly show
cargo +nightly --version
cmake --version
bash --version
- name: Use Cached LLVM
if: steps.cache-llvm.outputs.cache-hit == 'true'
run: |
echo "Using cached LLVM"
- name: Build revive
run: |
source emsdk/emsdk_env.sh
make install-wasm
- uses: actions/upload-artifact@v4
with:
name: revive-wasm
path: |
${{ env.REVIVE_WASM_INSTALL_DIR }}/resolc.js
${{ env.REVIVE_WASM_INSTALL_DIR }}/resolc.wasm
${{ env.REVIVE_WASM_INSTALL_DIR }}/resolc_web.js
retention-days: 1
test-revive-wasm:
needs: build-revive-wasm
strategy:
matrix:
os: ["ubuntu-24.04", "macos-14", "windows-2022"]
runs-on: ${{ matrix.os }}
steps:
- uses: actions/checkout@v4
- name: Create Target Directory
run: mkdir -p ${{ env.REVIVE_WASM_INSTALL_DIR }}
- name: Download Artifact
uses: actions/download-artifact@v4
with:
name: revive-wasm
path: ${{ env.REVIVE_WASM_INSTALL_DIR }}
- name: Set Up Node.js
uses: actions/setup-node@v3
with:
node-version: "20"
- name: Install packages
run: npm install
- name: Run Playwright tests
run: |
cd js
npx playwright install --with-deps
npx playwright test
- name: Test revive
run: |
echo "Running tests for ${{ matrix.os }}"
npm run test:wasm
+64
View File
@@ -0,0 +1,64 @@
name: Generate JSON for resolc-bin
on:
release:
types: [published]
jobs:
generateJson:
runs-on: ubuntu-latest
if: contains(github.event.release.tag_name, 'llvm') == false
environment: tags
env:
# the token is needed for json_generator.py
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
steps:
- name: Checkout
uses: actions/checkout@v4
with:
path: tmp
- name: Checkout
uses: actions/checkout@v4
with:
repository: paritytech/resolc-bin
path: resolc-bin
- uses: actions/create-github-app-token@v1
id: app-token
with:
app-id: ${{ secrets.REVIVE_JSON_APP_ID }}
private-key: ${{ secrets.REVIVE_JSON_APP_KEY }}
owner: paritytech
repositories: resolc-bin
- name: Generate json and push
env:
TOKEN: ${{ steps.app-token.outputs.token }}
APP_NAME: "paritytech-revive-json"
Green: "\e[32m"
NC: "\e[0m"
run: |
sudo apt-get update && sudo apt-get install -y wget
wget https://github.com/${GITHUB_REPOSITORY}/releases/download/${GITHUB_REF_NAME}/resolc-x86_64-unknown-linux-musl
chmod +x resolc-x86_64-unknown-linux-musl
export FIRST_SOLC_VERSION=$(./resolc-x86_64-unknown-linux-musl --supported-solc-versions | cut -f 1 -d "," | tr -d ">=")
export LAST_SOLC_VERSION=$(./resolc-x86_64-unknown-linux-musl --supported-solc-versions | cut -f 2 -d "," | tr -d "<=")
cd resolc-bin
python ../tmp/.github/scripts/json_generator.py ${GITHUB_REPOSITORY} ${{ github.event.release.tag_name }}
echo "${Green}Add new remote with gh app token${NC}"
git remote set-url origin $(git config remote.origin.url | sed "s/github.com/${APP_NAME}:${TOKEN}@github.com/g")
echo "${Green}Remove http section that causes issues with gh app auth token${NC}"
sed -i.bak '/\[http/d' ./.git/config
sed -i.bak '/extraheader/d' ./.git/config
git config user.email "ci@parity.io"
git config user.name "${APP_NAME}"
git add .
git commit -m "Update json"
git push origin main
echo "::notice::info.list files were successfully published to https://github.com/paritytech/resolc-bin"
+74 -109
View File
@@ -1,5 +1,4 @@
name: Release LLVM
on:
workflow_dispatch:
inputs:
@@ -8,12 +7,16 @@ on:
required: true
description: llvm version in "x.x.x" format, e.g. "18.1.8"
concurrency:
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
cancel-in-progress: true
env:
CARGO_TERM_COLOR: always
jobs:
create-release:
runs-on: ubuntu-latest
create-release-draft:
runs-on: ubuntu-24.04
permissions:
contents: write
outputs:
@@ -23,145 +26,107 @@ jobs:
run: |
echo "version=llvm-${{ inputs.llvm_version }}-revive.${GITHUB_SHA:0:7}" >> $GITHUB_OUTPUT
- name: create release
- name: Create Release
uses: softprops/action-gh-release@v2
with:
name: "LLVM binaries release: ${{ steps.resolve-version.outputs.version }}"
body: "This release includes binaries of LLVM, used to compile revive itself"
make_latest: "false"
name: ${{ steps.resolve-version.outputs.version }}
body: "LLVM is a dependency of revive. The LLVM releases are used by our CI to build revive."
draft: true
tag_name: ${{ steps.resolve-version.outputs.version }}
build-macos:
build:
strategy:
matrix:
os: [macos-14, macos-13]
target: [x86_64-unknown-linux-gnu, x86_64-unknown-linux-musl, wasm32-unknown-emscripten, aarch64-apple-darwin, x86_64-apple-darwin, x86_64-pc-windows-msvc]
include:
- os: macos-13
arch: x64
- os: macos-14
arch: arm64
needs: create-release
runs-on: ${{ matrix.os }}
name: "build-macos-${{ matrix.arch }}"
- target: x86_64-unknown-linux-gnu
builder-arg: gnu
host: linux
runner: parity-large
- target: x86_64-unknown-linux-musl
builder-arg: musl
host: linux
runner: parity-large
- target: wasm32-unknown-emscripten
builder-arg: emscripten
host: linux
runner: parity-large
- target: aarch64-apple-darwin
builder-arg: gnu
host: macos
runner: macos-14
- target: x86_64-apple-darwin
builder-arg: gnu
host: macos
runner: macos-13
- target: x86_64-pc-windows-msvc
builder-arg: gnu
host: windows
runner: windows-2022
needs: create-release-draft
runs-on: ${{ matrix.runner }}
env:
RUST_LOG: trace
permissions:
contents: write # for uploading assets to release
steps:
- uses: actions/checkout@v4
- name: install macos deps
run: |
brew install ninja
- name: versions
run: |
rustup show
cargo --version
cmake --version
echo "bash:" && bash --version
echo "ninja:" && ninja --version
echo "clang:" && clang --version
- name: Build LLVM
run: |
make install-llvm
- name: clean
# check removed files
run: |
cd target-llvm/gnu/target-final/bin/
rm diagtool llvm-libtool-darwin llvm-lipo llvm-pdbutil llvm-dwarfdump llvm-nm llvm-readobj llvm-cfi-verify \
sancov llvm-debuginfo-analyzer llvm-objdump llvm-profgen llvm-extract llvm-jitlink llvm-c-test llvm-gsymutil llvm-dwp \
dsymutil llvm-dwarfutil llvm-exegesis lli clang-rename bugpoint clang-extdef-mapping clang-refactor c-index-test \
llvm-reduce llvm-lto clang-linker-wrapper llc llvm-lto2
- name: package artifacts
run: |
tar -czf "${{ needs.create-release.outputs.version }}-macos-${{ matrix.arch }}.tar.gz" target-llvm/gnu/target-final
- name: upload archive to release
uses: softprops/action-gh-release@v2
- uses: actions-rust-lang/setup-rust-toolchain@v1
with:
make_latest: "false"
tag_name: ${{ needs.create-release.outputs.version }}
files: |
${{ needs.create-release.outputs.version }}-macos-${{ matrix.arch }}.tar.gz
# without this it will override our rust flags
rustflags: ""
cache-key: ${{ matrix.target }}
build-linux-all:
needs: create-release
runs-on: parity-large
env:
RUST_LOG: trace
permissions:
contents: write # for uploading assets to release
steps:
- uses: actions/checkout@v4
- name: install linux deps
- name: Install Dependencies
if: ${{ matrix.host == 'linux' }}
run: |
sudo apt-get update && sudo apt-get install -y cmake ninja-build curl git libssl-dev pkg-config clang lld musl
- uses: actions-rust-lang/setup-rust-toolchain@v1
with:
toolchain: stable
components: rust-src
target: wasm32-unknown-emscripten
rustflags: ""
- name: versions
- name: Install Dependencies
if: ${{ matrix.host == 'macos' }}
run: |
rustup show
cargo --version
cmake --version
echo "bash:" && bash --version
echo "ninja:" && ninja --version
echo "clang:" && clang --version
brew install ninja
- name: Build host LLVM
- name: Install LLVM Builder
run: |
make install-llvm
cargo install --path crates/llvm-builder
- name: Build gnu LLVM
- name: Clone LLVM
run: |
revive-llvm clone
revive-llvm build --llvm-projects lld --llvm-projects clang
revive-llvm --target-env ${{ matrix.builder-arg }} clone
- name: Build musl LLVM
- name: Build LLVM
if: ${{ matrix.target != 'wasm32-unknown-emscripten' }}
run: |
revive-llvm --target-env musl build --llvm-projects lld --llvm-projects clang
revive-llvm --target-env ${{ matrix.builder-arg }} build --llvm-projects lld --llvm-projects clang
- name: Build emscripten LLVM
- name: Build LLVM
if: ${{ matrix.target == 'wasm32-unknown-emscripten' }}
run: |
revive-llvm --target-env emscripten clone
source emsdk/emsdk_env.sh
revive-llvm --target-env emscripten build --llvm-projects lld
revive-llvm --target-env ${{ matrix.builder-arg }} build --llvm-projects lld
- name: clean
# check removed files
- name: Remove Unnecessary Binaries
shell: bash
run: |
for target in gnu emscripten musl; do
cd target-llvm/${target}/target-final/bin/
rm -rf diagtool llvm-libtool-darwin llvm-lipo llvm-pdbutil llvm-dwarfdump llvm-nm llvm-readobj llvm-cfi-verify \
sancov llvm-debuginfo-analyzer llvm-objdump llvm-profgen llvm-extract llvm-jitlink llvm-c-test llvm-gsymutil llvm-dwp \
dsymutil llvm-dwarfutil llvm-exegesis lli clang-rename bugpoint clang-extdef-mapping clang-refactor c-index-test \
llvm-reduce llvm-lto clang-linker-wrapper llc llvm-lto2 llvm-otool llvm-readelf
cd -
done
cd target-llvm/${{ matrix.builder-arg }}/target-final/bin/
rm -f diagtool* llvm-libtool-darwin* llvm-lipo* llvm-pdbutil* llvm-dwarfdump* llvm-nm* llvm-readobj* llvm-cfi-verify* \
sancov* llvm-debuginfo-analyzer* llvm-objdump* llvm-profgen* llvm-extract* llvm-jitlink* llvm-c-test* llvm-gsymutil* llvm-dwp* \
dsymutil* llvm-dwarfutil* llvm-exegesis* lli clang-rename* bugpoint* clang-extdef-mapping* clang-refactor* c-index-test* \
llvm-reduce* llvm-lto* clang-linker-wrapper* llc* llvm-lto2* llvm-otool* llvm-readelf* \
clang-repl* clang-check* clang-scan-deps*
cd -
- name: package artifacts
- name: Package Artifact
shell: bash
run: |
tar -czf "${{ needs.create-release.outputs.version }}-x86_64-linux-gnu-linux.tar.gz" target-llvm/gnu/target-final
tar -czf "${{ needs.create-release.outputs.version }}-x86_64-linux-musl.tar.gz" target-llvm/musl/target-final
tar -czf "${{ needs.create-release.outputs.version }}-wasm32-unknown-emscripten.tar.gz" target-llvm/emscripten/target-final
mv target-llvm/${{ matrix.builder-arg }}/target-final/ llvm-${{ matrix.target }}
tar -czf "${{ needs.create-release-draft.outputs.version }}-${{ matrix.target }}.tar.gz" llvm-${{ matrix.target }}
- name: upload archive to release
- name: Add Artifact to Release
uses: softprops/action-gh-release@v2
with:
make_latest: "false"
tag_name: ${{ needs.create-release.outputs.version }}
tag_name: ${{ needs.create-release-draft.outputs.version }}
draft: true
files: |
${{ needs.create-release.outputs.version }}-x86_64-linux-gnu-linux.tar.gz
${{ needs.create-release.outputs.version }}-x86_64-linux-musl.tar.gz
${{ needs.create-release.outputs.version }}-wasm32-unknown-emscripten.tar.gz
${{ needs.create-release-draft.outputs.version }}-${{ matrix.target }}.tar.gz
+288 -347
View File
@@ -1,347 +1,288 @@
name: Release
run-name: Release ${{ github.ref_name }}
on:
push:
branches:
- 'main'
concurrency:
group: ${{ github.ref }}-${{ github.workflow }}
cancel-in-progress: true
env:
#rust-musl-cross:x86_64-musl
RUST_MUSL_CROSS_IMAGE: messense/rust-musl-cross@sha256:68b86bc7cb2867259e6b233415a665ff4469c28b57763e78c3bfea1c68091561
jobs:
#
#
#
tag:
runs-on: ubuntu-24.04
permissions:
contents: write
outputs:
TAG: ${{ steps.versions.outputs.TAG }}
PKG_VER: ${{ steps.versions.outputs.PKG_VER }}
steps:
- name: Checkout
uses: actions/checkout@v4
with:
fetch-tags: 'true'
fetch-depth: 0
- name: Versions
id: versions
run: |
export CURRENT_TAG=$(git describe --tags --abbrev=0 --exclude "llvm-*")
export PKG_VER=v$(cat Cargo.toml | grep -A 5 package] | grep version | cut -d '=' -f 2 | tr -d '"' | tr -d " ")
echo "Current tag $CURRENT_TAG"
echo "Package version $PKG_VER"
#
echo "PKG_VER=$PKG_VER" >> $GITHUB_OUTPUT
if [ $CURRENT_TAG == $PKG_VER ];
then
echo "Tag is up to date. Nothing to do.";
export TAG=old;
else
echo "Tag was updated.";
export TAG=new;
fi
echo "TAG=$TAG" >> $GITHUB_OUTPUT
#
#
#
build-linux-all:
if: ${{ needs.tag.outputs.TAG == 'new' }}
runs-on: parity-large
needs: [tag]
steps:
- uses: actions/checkout@v4
- name: install linux deps
run: |
sudo apt-get update && sudo apt-get install -y cmake ninja-build \
curl git libssl-dev pkg-config clang lld musl
- name: Install Rust stable toolchain
uses: actions-rs/toolchain@v1
with:
profile: minimal
toolchain: stable
components: rust-src
target: wasm32-unknown-emscripten
- name: versions
run: |
rustup show
cargo --version
cmake --version
echo "bash:" && bash --version
echo "ninja:" && ninja --version
echo "clang:" && clang --version
- name: build revive-llvm
run: make install-llvm-builder
# musl LLVM
- name: llvm-musl-cache restore
id: llvm-musl-cache
uses: actions/cache/restore@v4
with:
path: target-llvm/musl/target-final
key: llvm-linux-musl-${{ hashFiles('crates/solidity/**') }}
- name: Build musl LLVM
if: steps.llvm-musl-cache.outputs.cache-hit != 'true'
run: |
revive-llvm --target-env musl clone
revive-llvm --target-env musl build --llvm-projects lld --llvm-projects clang
- name: llvm-musl-cache save
if: steps.llvm-musl-cache.outputs.cache-hit != 'true'
uses: actions/cache/save@v4
with:
path: target-llvm/musl/target-final
key: llvm-linux-musl-${{ hashFiles('crates/solidity/**') }}
# emscripten LLVM
- name: llvm-emscripten-cache restore
id: llvm-emscripten-cache
uses: actions/cache/restore@v4
with:
path: |
target-llvm/emscripten/target-final
emsdk
key: llvm-linux-emscripten-${{ hashFiles('crates/solidity/**') }}
- name: Build emscripten LLVM
if: steps.llvm-emscripten-cache.outputs.cache-hit != 'true'
run: |
revive-llvm --target-env emscripten clone
source emsdk/emsdk_env.sh
revive-llvm --target-env emscripten build --llvm-projects lld
- name: llvm-emscripten-cache save
if: steps.llvm-emscripten-cache.outputs.cache-hit != 'true'
uses: actions/cache/save@v4
with:
path: |
target-llvm/emscripten/target-final
emsdk
key: llvm-linux-emscripten-${{ hashFiles('crates/solidity/**') }}
# Build revive
- name: build musl
run: |
mkdir resolc-out
docker run -v $PWD:/opt/revive $RUST_MUSL_CROSS_IMAGE /bin/bash -c "
cd /opt/revive
apt update && apt upgrade -y && apt install -y pkg-config
export LLVM_SYS_181_PREFIX=/opt/revive/target-llvm/musl/target-final
make install-bin
cp /root/.cargo/bin/resolc /opt/revive/resolc-out/resolc
"
- name: check musl
run: |
mkdir solc
curl -sSLo solc/solc https://github.com/ethereum/solidity/releases/download/v0.8.28/solc-static-linux
chmod +x solc/solc
PATH=$PWD/solc:$PATH
result=$(./resolc-out/resolc --bin crates/integration/contracts/flipper.sol)
echo $result
if [[ $result == *'0x50564d'* ]]; then exit 0; else exit 1; fi
- name: Set Up Node.js
uses: actions/setup-node@v3
with:
node-version: "20"
- name: build wasm
run: |
export LLVM_SYS_181_PREFIX=$PWD/target-llvm/musl/target-final
export REVIVE_LLVM_TARGET_PREFIX=$PWD/target-llvm/emscripten/target-final
source emsdk/emsdk_env.sh
rustup target add wasm32-unknown-emscripten
make install-wasm
- name: check wasm
run: |
curl -sSLo solc/soljson.js https://github.com/ethereum/solidity/releases/download/v0.8.28/soljson.js
node -e "
const soljson = require('solc/soljson');
const createRevive = require('./target/wasm32-unknown-emscripten/release/resolc.js');
const compiler = createRevive();
compiler.soljson = soljson;
const standardJsonInput =
{
language: 'Solidity',
sources: {
'MyContract.sol': {
content: 'pragma solidity ^0.8.0; contract MyContract { function greet() public pure returns (string memory) { return \'Hello\'; } }',
},
},
settings: { optimizer: { enabled: false } }
};
compiler.writeToStdin(JSON.stringify(standardJsonInput));
compiler.callMain(['--standard-json']);
// Collect output
const stdout = compiler.readFromStdout();
const stderr = compiler.readFromStderr();
if (stderr) { console.error(stderr); process.exit(1); }
let out = JSON.parse(stdout);
let bytecode = out.contracts['MyContract.sol']['MyContract'].evm.bytecode.object
console.log(bytecode);
if(!bytecode.startsWith('50564d')) { process.exit(1); }
"
- uses: actions/upload-artifact@v4
with:
name: revive-wasm
path: |
./target/wasm32-unknown-emscripten/release/resolc.js
./target/wasm32-unknown-emscripten/release/resolc.wasm
./target/wasm32-unknown-emscripten/release/resolc_web.js
retention-days: 1
- uses: actions/upload-artifact@v4
with:
name: revive-linux
path: |
./resolc-out/resolc
retention-days: 1
#
#
#
create-release:
needs: [tag, build-linux-all]
runs-on: ubuntu-24.04
permissions:
contents: write
outputs:
upload_url: ${{ steps.create_release.outputs.result }}
steps:
- uses: actions/checkout@v4
- name: Create/update tag
id: tag
uses: actions/github-script@v7
with:
result-encoding: string
script: |
await github.rest.git.createRef({
owner: context.repo.owner,
repo: context.repo.repo,
ref: 'refs/tags/${{ needs.tag.outputs.PKG_VER }}',
sha: context.sha
})
- name: get relese notes
id: get-notes
run: |
{
echo 'releaseNotes<<EOF'
sed '/^## ${{ needs.tag.outputs.PKG_VER }}/,/^## v/!d' CHANGELOG.md | sed -e '1d' -e '$d'
echo EOF
} >> "$GITHUB_OUTPUT"
- name: Create release
id: create_release
env:
releaseNotes: ${{ steps.get-notes.outputs.releaseNotes }}
version: ${{ needs.tag.outputs.PKG_VER }}
uses: actions/github-script@v7
with:
result-encoding: string
script: |
let response = await github.rest.repos.createRelease({
owner: context.repo.owner,
repo: context.repo.repo,
tag_name: process.env.version,
name: process.env.version,
body: process.env.releaseNotes,
draft: true,
prerelease: true
});
console.log(response);
return response.data.upload_url;
- name: Log
run: |
echo "tag result: ${{ needs.tag.outputs.TAG }}"
echo "pkg version: ${{ needs.tag.outputs.PKG_VER }}"
#
#
#
upload-assets:
runs-on: ubuntu-24.04
needs: [create-release]
steps:
- name: Download Artifact
uses: actions/download-artifact@v4
with:
name: revive-wasm
path: resolc/
- name: Download Artifact
uses: actions/download-artifact@v4
with:
name: revive-linux
path: resolc/
- name: upload resolc
uses: actions/upload-release-asset@v1
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
with:
upload_url: ${{ needs.create-release.outputs.upload_url }}
asset_path: ./resolc/resolc
asset_name: resolc-static-linux
asset_content_type: application/octet-stream
- name: upload resolc.js
uses: actions/upload-release-asset@v1
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
with:
upload_url: ${{ needs.create-release.outputs.upload_url }}
asset_path: ./resolc/resolc.js
asset_name: resolc.js
asset_content_type: application/octet-stream
- name: upload resolc.wasm
uses: actions/upload-release-asset@v1
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
with:
upload_url: ${{ needs.create-release.outputs.upload_url }}
asset_path: ./resolc/resolc.wasm
asset_name: resolc.wasm
asset_content_type: application/octet-stream
- name: upload resolc_web.js
uses: actions/upload-release-asset@v1
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
with:
upload_url: ${{ needs.create-release.outputs.upload_url }}
asset_path: ./resolc/resolc_web.js
asset_name: resolc_web.js
asset_content_type: application/octet-stream
name: Build & Release
on:
push:
branches: ["main"]
tags:
- "v*"
pull_request:
branches: ["main"]
types: [opened, synchronize, labeled, unlabeled]
concurrency:
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
cancel-in-progress: true
env:
CARGO_TERM_COLOR: always
RUST_MUSL_CROSS_IMAGE: messense/rust-musl-cross@sha256:68b86bc7cb2867259e6b233415a665ff4469c28b57763e78c3bfea1c68091561
jobs:
check-version-changed:
runs-on: ubuntu-24.04
permissions:
contents: write
env:
CURRENT_TAG: ${{ github.ref_name }}
outputs:
RELEASE_NOTES: ${{ steps.versions.outputs.RELEASE_NOTES }}
steps:
- uses: actions/checkout@v4
# Check that tag and version in Cargo.toml match
- name: Check versions
id: versions
run: |
if [[ $CURRENT_TAG == 'main' ]];
then
echo "::notice::Tag $CURRENT_TAG is not a release tag, skipping the check in the main branch";
exit 0
fi
if [[ $CURRENT_TAG != "v"* ]];
then
echo "::notice::Tag $CURRENT_TAG is not a release tag, skipping the check in a PR";
exit 0
fi
export PKG_VER=v$(cat Cargo.toml | grep -A 5 package] | grep version | cut -d '=' -f 2 | tr -d '"' | tr -d " ")
echo "Current tag $CURRENT_TAG"
echo "Package version $PKG_VER"
#
if [[ $CURRENT_TAG != $PKG_VER ]];
then
echo "::error::Tag $CURRENT_TAG doesn't match package version $PKG_VER in Cargo.toml, please fix";
exit 1
fi
# Generating release notes early, in order to avoid checkout at the last step
export RELEASE_NOTES="$(sed '/^## '${CURRENT_TAG}'/,/^## v/!d' CHANGELOG.md | sed -e '1d' -e '$d')"
echo "Release notes:"
echo "$RELEASE_NOTES"
echo 'RELEASE_NOTES<<EOF' >> $GITHUB_OUTPUT
echo "$RELEASE_NOTES" >> $GITHUB_OUTPUT
echo 'EOF' >> $GITHUB_OUTPUT
build:
strategy:
matrix:
target:
[
x86_64-unknown-linux-musl,
aarch64-apple-darwin,
x86_64-apple-darwin,
x86_64-pc-windows-msvc,
]
include:
- target: x86_64-unknown-linux-musl
type: musl
runner: ubuntu-24.04
- target: aarch64-apple-darwin
type: native
runner: macos-14
- target: x86_64-apple-darwin
type: native
runner: macos-13
- target: x86_64-pc-windows-msvc
type: native
runner: windows-2022
runs-on: ${{ matrix.runner }}
needs: [check-version-changed]
steps:
- uses: actions/checkout@v4
- uses: actions-rust-lang/setup-rust-toolchain@v1
with:
# without this it will override our rust flags
rustflags: ""
cache-key: ${{ matrix.target }}
- name: Download LLVM
uses: ./.github/actions/get-llvm
with:
target: ${{ matrix.target }}
- name: Build
if: ${{ matrix.type == 'native' }}
shell: bash
run: |
export LLVM_SYS_181_PREFIX=$PWD/llvm-${{ matrix.target }}
make install-bin
mv target/release/resolc resolc-${{ matrix.target }} || mv target/release/resolc.exe resolc-${{ matrix.target }}.exe
- name: Build
if: ${{ matrix.type == 'musl' }}
run: |
docker run -v $PWD:/opt/revive $RUST_MUSL_CROSS_IMAGE /bin/bash -c "
cd /opt/revive
chown -R root:root .
apt update && apt upgrade -y && apt install -y pkg-config
export LLVM_SYS_181_PREFIX=/opt/revive/llvm-${{ matrix.target }}
make install-bin
mv target/${{ matrix.target }}/release/resolc resolc-${{ matrix.target }}
"
sudo chown -R $(id -u):$(id -g) .
- name: Install Solc
uses: ./.github/actions/get-solc
- name: Basic Sanity Check
shell: bash
run: |
result=$(./resolc-${{ matrix.target }} --bin crates/integration/contracts/flipper.sol)
echo $result
if [[ $result == *'0x50564d'* ]]; then exit 0; else exit 1; fi
- uses: actions/upload-artifact@v4
with:
name: resolc-${{ matrix.target }}
path: resolc-${{ matrix.target }}*
retention-days: 1
build-wasm:
runs-on: ubuntu-24.04
needs: [check-version-changed]
env:
RELEASE_RESOLC_WASM_URI: https://github.com/paritytech/revive-workflow-test/releases/download/${{ github.ref_name }}/resolc.wasm
steps:
- uses: actions/checkout@v4
- uses: actions-rust-lang/setup-rust-toolchain@v1
with:
target: wasm32-unknown-emscripten
# without this it will override our rust flags
rustflags: ""
- name: Download Host LLVM
uses: ./.github/actions/get-llvm
with:
target: x86_64-unknown-linux-gnu
- name: Download Wasm LLVM
uses: ./.github/actions/get-llvm
with:
target: wasm32-unknown-emscripten
- name: Download EMSDK
uses: ./.github/actions/get-emsdk
- name: Build
run: |
export LLVM_SYS_181_PREFIX=$PWD/llvm-x86_64-unknown-linux-gnu
export REVIVE_LLVM_TARGET_PREFIX=$PWD/llvm-wasm32-unknown-emscripten
source emsdk/emsdk_env.sh
make install-wasm
chmod -x ./target/wasm32-unknown-emscripten/release/resolc.wasm
- name: Set Up Node.js
uses: actions/setup-node@v3
with:
node-version: "20"
- name: Basic Sanity Check
run: |
mkdir -p solc
curl -sSLo solc/soljson.js https://github.com/ethereum/solidity/releases/download/v0.8.29/soljson.js
node -e "
const soljson = require('solc/soljson');
const createRevive = require('./target/wasm32-unknown-emscripten/release/resolc.js');
const compiler = createRevive();
compiler.soljson = soljson;
const standardJsonInput =
{
language: 'Solidity',
sources: {
'MyContract.sol': {
content: 'pragma solidity ^0.8.0; contract MyContract { function greet() public pure returns (string memory) { return \'Hello\'; } }',
},
},
settings: { optimizer: { enabled: false } }
};
compiler.writeToStdin(JSON.stringify(standardJsonInput));
compiler.callMain(['--standard-json']);
// Collect output
const stdout = compiler.readFromStdout();
const stderr = compiler.readFromStderr();
if (stderr) { console.error(stderr); process.exit(1); }
let out = JSON.parse(stdout);
let bytecode = out.contracts['MyContract.sol']['MyContract'].evm.bytecode.object
console.log(bytecode);
if(!bytecode.startsWith('50564d')) { process.exit(1); }
"
- name: Compress Artifact
run: |
mkdir -p resolc-wasm32-unknown-emscripten
mv ./target/wasm32-unknown-emscripten/release/resolc.js ./resolc-wasm32-unknown-emscripten/
mv ./target/wasm32-unknown-emscripten/release/resolc.wasm ./resolc-wasm32-unknown-emscripten/
mv ./target/wasm32-unknown-emscripten/release/resolc_web.js ./resolc-wasm32-unknown-emscripten/
- uses: actions/upload-artifact@v4
with:
name: resolc-wasm32-unknown-emscripten
path: resolc-wasm32-unknown-emscripten/*
retention-days: 1
create-release:
if: startsWith(github.ref_name, 'v')
needs: [check-version-changed, build-wasm]
runs-on: macos-14
environment: tags
steps:
- name: Download Artifacts
uses: actions/download-artifact@v4
with:
merge-multiple: true
- name: Create macOS Fat Binary
run: |
lipo resolc-aarch64-apple-darwin resolc-x86_64-apple-darwin -create -output resolc-universal-apple-darwin
- name: Make Executable
run: |
chmod +x resolc-x86_64-unknown-linux-musl
chmod +x resolc-universal-apple-darwin
- name: Create sha-256 checksum
run: |
shasum -a 256 resolc-x86_64-unknown-linux-musl > checksums.txt
shasum -a 256 resolc-universal-apple-darwin >> checksums.txt
shasum -a 256 resolc-x86_64-pc-windows-msvc.exe >> checksums.txt
shasum -a 256 resolc.js >> checksums.txt
shasum -a 256 resolc.wasm >> checksums.txt
shasum -a 256 resolc_web.js >> checksums.txt
- uses: actions/create-github-app-token@v1
id: app-token
with:
app-id: ${{ secrets.REVIVE_RELEASE_APP_ID }}
private-key: ${{ secrets.REVIVE_RELEASE_APP_KEY }}
- name: create-release
uses: softprops/action-gh-release@v2
with:
body: |
## Changelog
${{ needs.check-version-changed.outputs.RELEASE_NOTES }}
## Note for macOS Users
The macOS binary is unsigned and it needs to be made runnable using `xattr -c resolc-universal-apple-darwin`.
tag_name: ${{ github.ref_name }}
name: ${{ github.ref_name }}
prerelease: true
token: ${{ steps.app-token.outputs.token }}
target_commitish: ${{ github.sha }}
files: |
resolc-x86_64-unknown-linux-musl
resolc-universal-apple-darwin
resolc-x86_64-pc-windows-msvc.exe
resolc.js
resolc.wasm
resolc_web.js
checksums.txt
-58
View File
@@ -1,58 +0,0 @@
name: Build
on:
push:
branches: ["main"]
pull_request:
branches: ["main"]
env:
CARGO_TERM_COLOR: always
jobs:
build-ubuntu-x86:
runs-on: ubuntu-24.04
steps:
- uses: actions/checkout@v4
- name: Install solc
run: |
mkdir -p solc
curl -sSL --output solc/solc https://github.com/ethereum/solidity/releases/download/v0.8.28/solc-static-linux
chmod +x solc/solc
echo "$(pwd)/solc/" >> $GITHUB_PATH
- name: Install LLVM
run: |
curl -sSL --output llvm.tar.xz https://github.com/paritytech/revive/releases/download/v0.1.0-dev.7/clang+llvm-18.1.8-x86_64-linux-gnu-ubuntu-24.04.tar.xz
mkdir llvm18
tar Jxf llvm.tar.xz -C llvm18/
echo "LLVM_SYS_181_PREFIX=$(pwd)/llvm18" >> $GITHUB_ENV
- name: Install geth
run: |
sudo add-apt-repository -y ppa:ethereum/ethereum
sudo apt update
sudo apt install -y ethereum
- name: Machete
uses: bnjbvr/cargo-machete@main
- name: Format
run: make format
- name: Clippy
run: make clippy
- name: Test cargo workspace
run: make test-workspace
- name: Test CLI
run: make test-cli
- uses: actions/upload-artifact@v4
with:
name: ${{ github.job }}-resolc
path: ./target/release/resolc
retention-days: 1
+45
View File
@@ -0,0 +1,45 @@
name: Test LLVM Builder
on:
pull_request:
branches: ["main"]
types: [opened, synchronize]
paths:
- 'LLVM.lock'
- 'crates/llvm-builder/**'
- '.github/workflows/test-llvm-builder.yml'
concurrency:
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
cancel-in-progress: true
env:
CARGO_TERM_COLOR: always
jobs:
test:
strategy:
matrix:
runner: [parity-large, macos-14, windows-2022]
runs-on: ${{ matrix.runner }}
steps:
- uses: actions/checkout@v4
- uses: actions-rust-lang/setup-rust-toolchain@v1
with:
# without this it will override our rust flags
rustflags: ""
cache-key: ${{ matrix.runner }}
- name: Install Dependencies
if: matrix.runner == 'parity-large'
run: |
sudo apt update && sudo apt-get install -y cmake ninja-build curl git libssl-dev pkg-config clang lld musl
- name: Install Dependencies
if: matrix.runner == 'macos-14'
run: |
brew install ninja
- name: Test
run: make test-llvm-builder
env:
RUST_LOG: trace
+98
View File
@@ -0,0 +1,98 @@
name: Test Wasm Version
on:
push:
branches: ["main"]
pull_request:
branches: ["main"]
types: [opened, synchronize]
concurrency:
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
cancel-in-progress: true
env:
CARGO_TERM_COLOR: always
REVIVE_WASM_INSTALL_DIR: ${{ github.workspace }}/target/wasm32-unknown-emscripten/release
jobs:
build:
runs-on: ubuntu-24.04
defaults:
run:
shell: bash
steps:
- uses: actions/checkout@v4
- uses: actions-rust-lang/setup-rust-toolchain@v1
with:
target: wasm32-unknown-emscripten
# without this it will override our rust flags
rustflags: ""
- name: Download Host LLVM
uses: ./.github/actions/get-llvm
with:
target: x86_64-unknown-linux-gnu
- name: Download Wasm LLVM
uses: ./.github/actions/get-llvm
with:
target: wasm32-unknown-emscripten
- name: Install emsdk
uses: ./.github/actions/get-emsdk
- name: Set LLVM Environment Variables
run: |
echo "LLVM_SYS_181_PREFIX=$(pwd)/llvm-x86_64-unknown-linux-gnu" >> $GITHUB_ENV
echo "REVIVE_LLVM_TARGET_PREFIX=$(pwd)/llvm-wasm32-unknown-emscripten" >> $GITHUB_ENV
- name: Build Revive
run: |
source emsdk/emsdk_env.sh
make install-wasm
- uses: actions/upload-artifact@v4
with:
name: revive-wasm
path: |
${{ env.REVIVE_WASM_INSTALL_DIR }}/resolc.js
${{ env.REVIVE_WASM_INSTALL_DIR }}/resolc.wasm
${{ env.REVIVE_WASM_INSTALL_DIR }}/resolc_web.js
retention-days: 1
test:
needs: build
strategy:
matrix:
os: ["ubuntu-24.04", "macos-14", "windows-2022"]
runs-on: ${{ matrix.os }}
steps:
- uses: actions/checkout@v4
- name: Create Target Directory
run: mkdir -p ${{ env.REVIVE_WASM_INSTALL_DIR }}
- name: Download Artifact
uses: actions/download-artifact@v4
with:
name: revive-wasm
path: ${{ env.REVIVE_WASM_INSTALL_DIR }}
- name: Set Up Node.js
uses: actions/setup-node@v3
with:
node-version: "20"
- name: Install Node Packages
run: npm install
- name: Run Playwright tests
run: |
cd js
npx playwright install --with-deps
npx playwright test
- name: Test revive
run: |
echo "Running tests for ${{ matrix.os }}"
npm run test:wasm
+57
View File
@@ -0,0 +1,57 @@
name: Test
on:
push:
branches: ["main"]
pull_request:
branches: ["main"]
types: [opened, synchronize]
concurrency:
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
cancel-in-progress: true
env:
CARGO_TERM_COLOR: always
jobs:
test:
runs-on: ubuntu-24.04
steps:
- uses: actions/checkout@v4
- uses: actions-rust-lang/setup-rust-toolchain@v1
with:
# without this it will override our rust flags
rustflags: ""
- name: Install Solc
uses: ./.github/actions/get-solc
- name: Download LLVM
uses: ./.github/actions/get-llvm
with:
target: x86_64-unknown-linux-gnu
- name: Set LLVM Environment Variables
run: |
echo "LLVM_SYS_181_PREFIX=$(pwd)/llvm-x86_64-unknown-linux-gnu" >> $GITHUB_ENV
- name: Install Geth
run: |
sudo add-apt-repository -y ppa:ethereum/ethereum
sudo apt update
sudo apt install -y ethereum
- name: Machete
uses: bnjbvr/cargo-machete@v0.7.1
- name: Format
run: make format
- name: Clippy
run: make clippy
- name: Test cargo workspace
run: make test-workspace
- name: Test CLI
run: make test-cli
+47 -6
View File
@@ -4,7 +4,48 @@
This is a development pre-release.
Supported `polkadot-sdk` rev: `274a781e8ca1a9432c7ec87593bd93214abbff50`
Supported `polkadot-sdk` rev:`c29e72a8628835e34deb6aa7db9a78a2e4eabcee`
### Added
### Changed
### Fixed
- Constructors avoid storing zero sized immutable data on exit.
## v0.1.0-dev.13
This is a development pre-release.
Supported `polkadot-sdk` rev:`c29e72a8628835e34deb6aa7db9a78a2e4eabcee`
### Added
- Support for solc v0.8.29
- Decouples the solc JSON-input-output type definitions from the Solidity fronted and expose them via a dedicated crate.
- `--supported-solc-versions` for `resolc` binary to return a `semver` range of supported `solc` versions.
- Support for passing LLVM command line options via the prcoess input or providing one or more `--llvm-arg='..'` resolc CLI flag. This allows more fine-grained control over the LLVM backend configuration.
### Changed
- Storage keys and values are big endian. This was a pre-mature optimization because for the contract itself it this is a no-op and thus not observable. However we should consider the storage layout as part of the contract ABI. The endianness of transient storage values are still kept as-is.
- Running `resolc` using webkit is no longer supported.
### Fixed
- A missing byte swap for the create2 salt value.
## v0.1.0-dev.12
This is a development pre-release.
Supported `polkadot-sdk` rev: `21f6f0705e53c15aa2b8a5706b208200447774a9`
### Added
- Per file output selection for `--standard-json` mode.
- The `ir` output selection option for `--standard-json` mode.
### Changed
- Improved code size: Large contracts compile to smaller code blobs when enabling aggressive size optimizations (`-Oz`).
### Fixed
## v0.1.0-dev.11
@@ -14,7 +55,7 @@ Supported `polkadot-sdk` rev: `274a781e8ca1a9432c7ec87593bd93214abbff50`
### Added
### Changed
### Changed
### Fixed
- A bug causing incorrect loads from the emulated EVM linear memory.
@@ -30,7 +71,7 @@ Supported `polkadot-sdk` rev: `274a781e8ca1a9432c7ec87593bd93214abbff50`
- Support for the `coinbase` opcode.
- The resolc web JS version.
### Changed
### Changed
- Missing the `--overwrite` flag emits an error instead of a warning.
- The `resolc` executable prints the help by default.
- Removed support for legacy EVM assembly (EVMLA) translation.
@@ -52,7 +93,7 @@ This is a development pre-release.
### Added
### Changed
### Changed
- Syscalls with more than 6 arguments now pack them into registers.
### Fixed
@@ -92,9 +133,9 @@ This is a development pre-release.
- Calls forward maximum weight limits instead of 0, anticipating a change in polkadot-sdk where weight limits of 0 no longer interprets as uncapped limit.
### Fixed
- A linker bug which was preventing certain contracts from linking with the PVM linker.
- A linker bug which was preventing certain contracts from linking with the PVM linker.
- JS: Fix encoding conversion from JS string (UTF-16) to UTF-8.
- The git commit hash slug is always displayed in the version string.
- The git commit hash slug is always displayed in the version string.
## v0.1.0-dev.6
Generated
+644 -722
View File
File diff suppressed because it is too large Load Diff
+16 -15
View File
@@ -3,7 +3,7 @@ resolver = "2"
members = ["crates/*"]
[workspace.package]
version = "0.1.0-dev.11"
version = "0.1.0-dev.13"
authors = [
"Cyrill Leutwiler <cyrill@parity.io>",
"Parity Technologies <admin@parity.io>",
@@ -14,19 +14,20 @@ repository = "https://github.com/paritytech/revive"
rust-version = "1.81.0"
[workspace.dependencies]
revive-benchmarks = { version = "0.1.0-dev.11", path = "crates/benchmarks" }
revive-builtins = { version = "0.1.0-dev.11", path = "crates/builtins" }
revive-common = { version = "0.1.0-dev.11", path = "crates/common" }
revive-differential = { version = "0.1.0-dev.11", path = "crates/differential" }
revive-integration = { version = "0.1.0-dev.11", path = "crates/integration" }
revive-linker = { version = "0.1.0-dev.11", path = "crates/linker" }
lld-sys = { version = "0.1.0-dev.11", path = "crates/lld-sys" }
revive-llvm-context = { version = "0.1.0-dev.11", path = "crates/llvm-context" }
revive-runtime-api = { version = "0.1.0-dev.11", path = "crates/runtime-api" }
revive-runner = { version = "0.1.0-dev.11", path = "crates/runner" }
revive-solidity = { version = "0.1.0-dev.11", path = "crates/solidity" }
revive-stdlib = { version = "0.1.0-dev.11", path = "crates/stdlib" }
revive-build-utils = { version = "0.1.0-dev.11", path = "crates/build-utils" }
revive-benchmarks = { version = "0.1.0-dev.13", path = "crates/benchmarks" }
revive-builtins = { version = "0.1.0-dev.13", path = "crates/builtins" }
revive-common = { version = "0.1.0-dev.13", path = "crates/common" }
revive-differential = { version = "0.1.0-dev.13", path = "crates/differential" }
revive-integration = { version = "0.1.0-dev.13", path = "crates/integration" }
revive-linker = { version = "0.1.0-dev.13", path = "crates/linker" }
lld-sys = { version = "0.1.0-dev.13", path = "crates/lld-sys" }
revive-llvm-context = { version = "0.1.0-dev.13", path = "crates/llvm-context" }
revive-runtime-api = { version = "0.1.0-dev.13", path = "crates/runtime-api" }
revive-runner = { version = "0.1.0-dev.13", path = "crates/runner" }
revive-solc-json-interface = { version = "0.1.0-dev.13", path = "crates/solc-json-interface" }
revive-solidity = { version = "0.1.0-dev.13", path = "crates/solidity" }
revive-stdlib = { version = "0.1.0-dev.13", path = "crates/stdlib" }
revive-build-utils = { version = "0.1.0-dev.13", path = "crates/build-utils" }
hex = "0.4.3"
cc = "1.2"
@@ -72,7 +73,7 @@ assert_fs = "1.1"
# polkadot-sdk and friends
codec = { version = "3.6.12", default-features = false, package = "parity-scale-codec" }
scale-info = { version = "2.11.6", default-features = false }
polkadot-sdk = { git = "https://github.com/paritytech/polkadot-sdk", rev = "274a781e8ca1a9432c7ec87593bd93214abbff50" }
polkadot-sdk = { git = "https://github.com/paritytech/polkadot-sdk", rev = "c29e72a8628835e34deb6aa7db9a78a2e4eabcee" }
# llvm
[workspace.dependencies.inkwell]
+1 -1
View File
@@ -11,7 +11,7 @@ RUN make install-llvm-builder
RUN revive-llvm --target-env musl clone
RUN revive-llvm --target-env musl build --llvm-projects lld --llvm-projects clang
FROM messense/rust-musl-cross:x86_64-musl AS resolc-builder
FROM messense/rust-musl-cross@sha256:68b86bc7cb2867259e6b233415a665ff4469c28b57763e78c3bfea1c68091561 AS resolc-builder
WORKDIR /opt/revive
RUN apt update && \
+1 -1
View File
@@ -23,7 +23,7 @@
install: install-bin install-npm
install-bin:
cargo install --path crates/solidity
cargo install --locked --path crates/solidity
install-npm:
npm install && npm fund
+1 -1
View File
@@ -1,4 +1,4 @@
![CI](https://github.com/paritytech/revive/actions/workflows/rust.yml/badge.svg)
![CI](https://github.com/paritytech/revive/actions/workflows/test.yml/badge.svg)
[![Docs](https://img.shields.io/badge/Docs-contracts.polkadot.io-brightgreen.svg)](https://contracts.polkadot.io/revive_compiler/)
# revive
+8 -4
View File
@@ -4,10 +4,14 @@ Prior to the first stable release we neither have formal release processes nor d
To create a new pre-release:
1. Merge a release PR which updates the `-dev.X` versions in the workspace `Cargo.toml` and updates the `CHANGELOG.md` accordingly. The release workflow will attempt to build and publish a new release whenever the latest git tag does not match the cargo package version.
2. Wait for the `Release` workflow to finish. If the workflow fails after the `build-linux-all` step, check if a tag has been created and delete it before restarting or pushing updates. Note: It's more convenient to debug the release workflow in a fork (the fork has to be under the `paritytech` org to access `parity-large` runners).
3. Check draft release on [Releases page](https://github.com/paritytech/revive/releases) and publish (should contain `resolc.js`, `resolc.wasm`, `resolc-web.js`, and `resolc-static-linux` release assets)
4. Update the [contract-docs](https://github.com/paritytech/contract-docs/) accordingly
1. Create a release PR which updates the `-dev.X` versions in the workspace `Cargo.toml` and updates the `CHANGELOG.md` accordingly.
2. If the CI passes, merge the release PR.
3. Push a tag that has the same `-dev.X` version as in `Cargo.toml`
4. The release workflow will attempt to build and publish a new pre-release if the latest tag does match the cargo package version.
5. Wait for the `Release` workflow to finish. It should create the pre-release with the same `-dev.X` name.
6. Check that pre-release was created on the [Releases page](https://github.com/paritytech/revive/releases) with all artifacts.
7. After the release is published, another workflow should start automatically and update json files in https://github.com/paritytech/resolc-bin. Check the changes.
8. Update the [contract-docs](https://github.com/paritytech/contract-docs/) accordingly
# LLVM release
+24 -24
View File
@@ -15,58 +15,58 @@
### Baseline
| | `EVM` | `PVMInterpreter` |
|:--------|:------------------------|:-------------------------------- |
| **`0`** | `3.36 us` (✅ **1.00x**) | `11.84 us` (*3.52x slower*) |
| | `EVM` | `PVMInterpreter` |
|:--------|:-------------------------|:-------------------------------- |
| **`0`** | `10.08 us` (✅ **1.00x**) | `10.32 us` (**1.02x slower**) |
### OddPorduct
| | `EVM` | `PVMInterpreter` |
|:-------------|:-------------------------|:-------------------------------- |
| **`10000`** | `3.11 ms` (✅ **1.00x**) | `1.53 ms` (🚀 **2.03x faster**) |
| **`100000`** | `30.70 ms` (✅ **1.00x**) | `15.54 ms` (🚀 **1.98x faster**) |
| **`300000`** | `92.68 ms` (✅ **1.00x**) | `45.47 ms` (🚀 **2.04x faster**) |
| | `EVM` | `PVMInterpreter` |
|:-------------|:--------------------------|:-------------------------------- |
| **`10000`** | `3.60 ms` (✅ **1.00x**) | `1.57 ms` (🚀 **2.28x faster**) |
| **`100000`** | `34.72 ms` (✅ **1.00x**) | `14.82 ms` (🚀 **2.34x faster**) |
| **`300000`** | `105.01 ms` (✅ **1.00x**) | `44.11 ms` (🚀 **2.38x faster**) |
### TriangleNumber
| | `EVM` | `PVMInterpreter` |
|:-------------|:-------------------------|:-------------------------------- |
| **`10000`** | `2.29 ms` (✅ **1.00x**) | `1.09 ms` (🚀 **2.11x faster**) |
| **`100000`** | `22.84 ms` (✅ **1.00x**) | `10.66 ms` (🚀 **2.14x faster**) |
| **`360000`** | `82.29 ms` (✅ **1.00x**) | `37.01 ms` (🚀 **2.22x faster**) |
| **`10000`** | `2.43 ms` (✅ **1.00x**) | `1.12 ms` (🚀 **2.17x faster**) |
| **`100000`** | `24.20 ms` (✅ **1.00x**) | `10.86 ms` (🚀 **2.23x faster**) |
| **`360000`** | `88.69 ms` (✅ **1.00x**) | `38.46 ms` (🚀 **2.31x faster**) |
### FibonacciRecursive
| | `EVM` | `PVMInterpreter` |
|:---------|:--------------------------|:--------------------------------- |
| **`12`** | `135.67 us` (✅ **1.00x**) | `125.02 us` (✅ **1.09x faster**) |
| **`16`** | `903.75 us` (✅ **1.00x**) | `762.79 us` (✅ **1.18x faster**) |
| **`20`** | `6.12 ms` (✅ **1.00x**) | `4.96 ms` (✅ **1.23x faster**) |
| **`24`** | `42.05 ms` (✅ **1.00x**) | `33.86 ms` (✅ **1.24x faster**) |
| **`12`** | `144.17 us` (✅ **1.00x**) | `150.85 us` (✅ **1.05x slower**) |
| **`16`** | `938.71 us` (✅ **1.00x**) | `922.11 us` (✅ **1.02x faster**) |
| **`20`** | `6.54 ms` (✅ **1.00x**) | `6.20 ms` (✅ **1.05x faster**) |
| **`24`** | `45.73 ms` (✅ **1.00x**) | `41.98 ms` (✅ **1.09x faster**) |
### FibonacciIterative
| | `EVM` | `PVMInterpreter` |
|:----------|:-------------------------|:-------------------------------- |
| **`64`** | `15.04 us` (✅ **1.00x**) | `29.45 us` (❌ *1.96x slower*) |
| **`128`** | `26.36 us` (✅ **1.00x**) | `42.19 us` (❌ *1.60x slower*) |
| **`256`** | `48.61 us` (✅ **1.00x**) | `65.71 us` (*1.35x slower*) |
| **`64`** | `23.00 us` (✅ **1.00x**) | `31.88 us` (❌ *1.39x slower*) |
| **`128`** | `35.28 us` (✅ **1.00x**) | `42.43 us` (❌ *1.20x slower*) |
| **`256`** | `60.12 us` (✅ **1.00x**) | `61.20 us` (**1.02x slower**) |
### FibonacciBinet
| | `EVM` | `PVMInterpreter` |
|:----------|:-------------------------|:-------------------------------- |
| **`64`** | `15.22 us` (✅ **1.00x**) | `41.46 us` (❌ *2.72x slower*) |
| **`128`** | `17.05 us` (✅ **1.00x**) | `42.84 us` (❌ *2.51x slower*) |
| **`256`** | `19.00 us` (✅ **1.00x**) | `44.36 us` (❌ *2.34x slower*) |
| **`64`** | `23.01 us` (✅ **1.00x**) | `47.74 us` (❌ *2.07x slower*) |
| **`128`** | `25.44 us` (✅ **1.00x**) | `49.67 us` (❌ *1.95x slower*) |
| **`256`** | `28.66 us` (✅ **1.00x**) | `53.01 us` (❌ *1.85x slower*) |
### SHA1
| | `EVM` | `PVMInterpreter` |
|:----------|:--------------------------|:--------------------------------- |
| **`1`** | `110.04 us` (✅ **1.00x**) | `216.11 us` (❌ *1.96x slower*) |
| **`64`** | `209.04 us` (✅ **1.00x**) | `309.48 us` (❌ *1.48x slower*) |
| **`512`** | `903.65 us` (✅ **1.00x**) | `980.49 us` (✅ **1.09x slower**) |
| **`1`** | `135.87 us` (✅ **1.00x**) | `243.75 us` (❌ *1.79x slower*) |
| **`64`** | `258.45 us` (✅ **1.00x**) | `355.70 us` (❌ *1.38x slower*) |
| **`512`** | `1.10 ms` (✅ **1.00x**) | `1.09 ms` (✅ **1.01x faster**) |
---
Made with [criterion-table](https://github.com/nu11ptr/criterion-table)
+1
View File
@@ -15,6 +15,7 @@ serde_json = { workspace = true }
revive-solidity = { workspace = true }
revive-runner = { workspace = true }
revive-llvm-context = { workspace = true }
[dev-dependencies]
sha1 = { workspace = true }
+8 -8
View File
@@ -1,10 +1,10 @@
{
"Baseline": 1110,
"Computation": 2389,
"DivisionArithmetics": 14822,
"ERC20": 23973,
"Events": 1605,
"FibonacciIterative": 2023,
"Flipper": 1989,
"SHA1": 17026
"Baseline": 950,
"Computation": 2262,
"DivisionArithmetics": 8915,
"ERC20": 17233,
"Events": 1628,
"FibonacciIterative": 1485,
"Flipper": 2132,
"SHA1": 8381
}
@@ -0,0 +1,33 @@
// SPDX-License-Identifier: MIT
pragma solidity ^0.8.28;
contract Predicted {
uint public salt;
constructor(uint _salt) {
salt = _salt;
}
}
contract AddressPredictor {
constructor(uint _salt, bytes memory _bytecode) payable {
address deployed = address(new Predicted{salt: bytes32(_salt)}(_salt));
address predicted = predictAddress(_salt, _bytecode);
assert(deployed == predicted);
}
function predictAddress(
uint _foo,
bytes memory _bytecode
) public view returns (address predicted) {
bytes32 addr = keccak256(
abi.encodePacked(
bytes1(0xff),
address(this),
bytes32(_foo),
keccak256(abi.encodePacked(_bytecode, abi.encode(_foo)))
)
);
predicted = address(uint160(uint(addr)));
}
}
@@ -0,0 +1,38 @@
// SPDX-License-Identifier: MIT
pragma solidity ^0.8.28;
/* runner.json
{
"differential": true,
"actions": [
{
"Instantiate": {
"code": {
"Solidity": {
"contract": "DelegateCaller"
}
}
}
},
{
"Call": {
"dest": {
"Instantiated": 0
},
"data": "e466c6c9"
}
}
]
}
*/
contract DelegateCaller {
function delegateNoContract() external returns (bool) {
address testAddress = 0x0000000000000000000000000000000000000000;
(bool success, ) = testAddress.delegatecall(
abi.encodeWithSignature("test()")
);
return success;
}
}
@@ -0,0 +1,36 @@
// SPDX-License-Identifier: MIT
pragma solidity ^0.8.28;
/* runner.json
{
"differential": true,
"actions": [
{
"Instantiate": {
"code": {
"Solidity": {
"contract": "FunctionType"
}
}
}
},
{
"Call": {
"dest": {
"Instantiated": 0
},
"data": "b8c9d365"
}
}
]
}
*/
contract FunctionType {
uint public immutable x = 42;
function h() public view returns (function() external view returns (uint)) {
return this.x;
}
}
+52
View File
@@ -0,0 +1,52 @@
// SPDX-License-Identifier: MIT
pragma solidity ^0.8.29;
/* runner.json
{
"differential": true,
"actions": [
{
"Instantiate": {
"code": {
"Solidity": {
"contract": "LayoutAt"
}
}
}
},
{
"Call": {
"dest": {
"Instantiated": 0
},
"data": "a7a0d537"
}
},
{
"Call": {
"dest": {
"Instantiated": 0
},
"data": "15393349"
}
}
]
}
*/
contract LayoutAt layout at 0xDEADBEEF + 0xCAFEBABE {
uint[3] public something;
constructor() payable {
something[0] = 1337;
something[1] = 42;
something[2] = 69;
}
function slotOfSomething() public pure returns (uint ret) {
assembly {
ret := something.slot
}
}
}
@@ -0,0 +1,51 @@
// SPDX-License-Identifier: MIT
pragma solidity ^0.8;
/* runner.json
{
"differential": true,
"actions": [
{
"Instantiate": {
"code": {
"Solidity": {
"contract": "MCopyOverlap"
}
}
}
},
{
"Call": {
"dest": {
"Instantiated": 0
},
"data": "afdce848"
}
}
]
}
*/
function copy(
uint dstOffset,
uint srcOffset,
uint length
) pure returns (bytes memory out) {
out = hex"2222222222222222333333333333333344444444444444445555555555555555"
hex"6666666666666666777777777777777788888888888888889999999999999999"
hex"aaaaaaaaaaaaaaaabbbbbbbbbbbbbbbbccccccccccccccccdddddddddddddddd";
assembly {
mcopy(
add(add(out, 0x20), dstOffset),
add(add(out, 0x20), srcOffset),
length
)
}
}
contract MCopyOverlap {
function mcopy_to_right_overlap() public pure returns (bytes memory) {
return copy(0x20, 0x10, 0x30);
}
}
+1 -1
View File
@@ -15,7 +15,7 @@ pragma solidity ^0.8;
"Instantiated": 0
},
"key": "0000000000000000000000000000000000000000000000000000000000000000",
"expected": "0100000000000000000000000000000000000000000000000000000000000000"
"expected": "0000000000000000000000000000000000000000000000000000000000000001"
}
},
{
+63 -10
View File
@@ -1,6 +1,7 @@
use alloy_primitives::{Address, Bytes, I256, U256};
use alloy_sol_types::{sol, SolCall, SolConstructor};
use revive_llvm_context::OptimizerSettings;
use revive_solidity::test_utils::*;
#[derive(Clone)]
@@ -249,8 +250,19 @@ sol!(
);
case!("Storage.sol", Storage, transientCall, storage_transient, value: U256);
sol!(
contract Predicted {
constructor(uint _foo);
}
contract AddressPredictor {
constructor(uint _foo, bytes memory _bytecode) payable;
}
);
case!("AddressPredictor.sol", Predicted, constructorCall, predicted_constructor, salt: U256);
case!("AddressPredictor.sol", AddressPredictor, constructorCall, address_predictor_constructor, salt: U256, bytecode: Bytes);
impl Contract {
fn build(calldata: Vec<u8>, name: &'static str, code: &str) -> Self {
pub fn build(calldata: Vec<u8>, name: &'static str, code: &str) -> Self {
Self {
name,
evm_runtime: compile_evm_bin_runtime(name, code),
@@ -258,11 +270,19 @@ impl Contract {
calldata,
}
}
pub fn build_size_opt(calldata: Vec<u8>, name: &'static str, code: &str) -> Self {
Self {
name,
evm_runtime: compile_evm_bin_runtime(name, code),
pvm_runtime: compile_blob_with_options(name, code, true, OptimizerSettings::size()),
calldata,
}
}
}
#[cfg(test)]
mod tests {
use alloy_primitives::{Bytes, U256};
use rayon::iter::{IntoParallelIterator, ParallelIterator};
use serde::{de::Deserialize, Serialize};
use std::{collections::BTreeMap, fs::File};
@@ -302,14 +322,47 @@ mod tests {
};
[
Contract::baseline as fn() -> Contract,
Contract::flipper as fn() -> Contract,
(|| Contract::odd_product(0)) as fn() -> Contract,
(|| Contract::fib_iterative(U256::ZERO)) as fn() -> Contract,
Contract::erc20 as fn() -> Contract,
(|| Contract::sha1(Bytes::new())) as fn() -> Contract,
(|| Contract::division_arithmetics_div(U256::ZERO, U256::ZERO)) as fn() -> Contract,
(|| Contract::event(U256::ZERO)) as fn() -> Contract,
(|| {
Contract::build_size_opt(
vec![],
"Baseline",
include_str!("../contracts/Baseline.sol"),
)
}) as _,
(|| {
Contract::build_size_opt(
vec![],
"Flipper",
include_str!("../contracts/flipper.sol"),
)
}) as _,
(|| {
Contract::build_size_opt(
vec![],
"Computation",
include_str!("../contracts/Computation.sol"),
)
}) as _,
(|| {
Contract::build_size_opt(
vec![],
"FibonacciIterative",
include_str!("../contracts/Fibonacci.sol"),
)
}) as _,
(|| Contract::build_size_opt(vec![], "ERC20", include_str!("../contracts/ERC20.sol")))
as _,
(|| Contract::build_size_opt(vec![], "SHA1", include_str!("../contracts/SHA1.sol")))
as _,
(|| {
Contract::build_size_opt(
vec![],
"DivisionArithmetics",
include_str!("../contracts/DivisionArithmetics.sol"),
)
}) as _,
(|| Contract::build_size_opt(vec![], "Events", include_str!("../contracts/Events.sol")))
as _,
]
.into_par_iter()
.map(extract_code_size)
+32
View File
@@ -33,6 +33,7 @@ test_spec!(msize, "MSize", "MSize.sol");
test_spec!(sha1, "SHA1", "SHA1.sol");
test_spec!(block, "Block", "Block.sol");
test_spec!(mcopy, "MCopy", "MCopy.sol");
test_spec!(mcopy_overlap, "MCopyOverlap", "MCopyOverlap.sol");
test_spec!(events, "Events", "Events.sol");
test_spec!(storage, "Storage", "Storage.sol");
test_spec!(mstore8, "MStore8", "MStore8.sol");
@@ -56,6 +57,9 @@ test_spec!(transfer, "Transfer", "Transfer.sol");
test_spec!(send, "Send", "Send.sol");
test_spec!(function_pointer, "FunctionPointer", "FunctionPointer.sol");
test_spec!(mload, "MLoad", "MLoad.sol");
test_spec!(delegate_no_contract, "DelegateCaller", "DelegateCaller.sol");
test_spec!(function_type, "FunctionType", "FunctionType.sol");
test_spec!(layout_at, "LayoutAt", "LayoutAt.sol");
fn instantiate(path: &str, contract: &str) -> Vec<SpecsAction> {
vec![Instantiate {
@@ -482,3 +486,31 @@ fn transfer_denies_reentrancy() {
}
.run();
}
#[test]
fn create2_salt() {
let salt = U256::from(777);
let predicted = Contract::predicted_constructor(salt).pvm_runtime;
let predictor = Contract::address_predictor_constructor(salt, predicted.clone().into());
Specs {
actions: vec![
Upload {
origin: TestAddress::Alice,
code: Code::Bytes(predicted),
storage_deposit_limit: None,
},
Instantiate {
origin: TestAddress::Alice,
value: 0,
gas_limit: Some(GAS_LIMIT),
storage_deposit_limit: None,
code: Code::Bytes(predictor.pvm_runtime),
data: predictor.calldata,
salt: OptionalHex::default(),
},
],
differential: false,
..Default::default()
}
.run();
}
-1
View File
@@ -52,7 +52,6 @@ pub fn link<T: AsRef<[u8]>>(input: T) -> anyhow::Result<Vec<u8>> {
let ld_args = [
"ld.lld",
"--lto=full",
"--error-limit=0",
"--relocatable",
"--emit-relocs",
+1 -1
View File
@@ -96,7 +96,7 @@ fn main() {
revive_build_utils::llvm_cxx_flags()
.split_whitespace()
.fold(&mut cc::Build::new(), |builder, flag| builder.flag(flag))
.flag("-Wno-unused-parameter")
.warnings(false)
.cpp(true)
.file("src/linker.cpp")
.compile("liblinker.a");
+1
View File
@@ -32,6 +32,7 @@ tar = { workspace = true }
flate2 = { workspace = true }
env_logger = { workspace = true }
log = { workspace = true }
which = { workspace = true }
[dev-dependencies]
assert_cmd = { workspace = true }
+41 -9
View File
@@ -1,5 +1,8 @@
//! Utilities for compiling the LLVM compiler-rt builtins.
use crate::utils::path_windows_to_unix as to_unix;
use std::{env::consts::EXE_EXTENSION, process::Command};
/// Static CFLAGS variable passed to the compiler building the compiler-rt builtins.
const C_FLAGS: [&str; 6] = [
"--target=riscv64",
@@ -44,24 +47,31 @@ fn cmake_dynamic_args(
let mut clang_path = llvm_target_host.to_path_buf();
clang_path.push("bin/clang");
clang_path.set_extension(EXE_EXTENSION);
let mut clangxx_path = llvm_target_host.to_path_buf();
clangxx_path.push("bin/clang++");
clangxx_path.set_extension(EXE_EXTENSION);
let mut llvm_config_path = llvm_target_host.to_path_buf();
llvm_config_path.push("bin/llvm-config");
llvm_config_path.set_extension(EXE_EXTENSION);
let mut ar_path = llvm_target_host.to_path_buf();
ar_path.push("bin/llvm-ar");
ar_path.set_extension(EXE_EXTENSION);
let mut nm_path = llvm_target_host.to_path_buf();
nm_path.push("bin/llvm-nm");
nm_path.set_extension(EXE_EXTENSION);
let mut ranlib_path = llvm_target_host.to_path_buf();
ranlib_path.push("bin/llvm-ranlib");
ranlib_path.set_extension(EXE_EXTENSION);
let mut linker_path = llvm_target_host.to_path_buf();
linker_path.push("bin/ld.lld");
linker_path.set_extension(EXE_EXTENSION);
Ok([
format!(
@@ -76,12 +86,18 @@ fn cmake_dynamic_args(
format!("-DCMAKE_C_FLAGS='{}'", C_FLAGS.join(" ")),
format!("-DCMAKE_ASM_FLAGS='{}'", C_FLAGS.join(" ")),
format!("-DCMAKE_CXX_FLAGS='{}'", C_FLAGS.join(" ")),
format!("-DCMAKE_C_COMPILER='{}'", clang_path.to_string_lossy()),
format!("-DCMAKE_ASM_COMPILER='{}'", clang_path.to_string_lossy()),
format!("-DCMAKE_CXX_COMPILER='{}'", clangxx_path.to_string_lossy()),
format!("-DCMAKE_AR='{}'", ar_path.to_string_lossy()),
format!("-DCMAKE_NM='{}'", nm_path.to_string_lossy()),
format!("-DCMAKE_RANLIB='{}'", ranlib_path.to_string_lossy()),
format!(
"-DCMAKE_C_COMPILER='{}'",
to_unix(clang_path.clone())?.display()
),
format!("-DCMAKE_ASM_COMPILER='{}'", to_unix(clang_path)?.display()),
format!(
"-DCMAKE_CXX_COMPILER='{}'",
to_unix(clangxx_path)?.display()
),
format!("-DCMAKE_AR='{}'", to_unix(ar_path)?.display()),
format!("-DCMAKE_NM='{}'", to_unix(nm_path)?.display()),
format!("-DCMAKE_RANLIB='{}'", to_unix(ranlib_path)?.display()),
format!(
"-DLLVM_CONFIG_PATH='{}'",
llvm_config_path.to_string_lossy()
@@ -101,7 +117,13 @@ pub fn build(
log::info!("building compiler-rt for rv64emac");
crate::utils::check_presence("cmake")?;
crate::utils::check_presence("ninja")?;
let generator = if cfg!(target_os = "windows") {
"Visual Studio 17 2022"
} else {
crate::utils::check_presence("ninja")?;
"Ninja"
};
let llvm_module_compiler_rt = crate::LLVMPath::llvm_module_compiler_rt()?;
let llvm_compiler_rt_build = crate::LLVMPath::llvm_build_compiler_rt()?;
@@ -114,7 +136,7 @@ pub fn build(
"-B",
llvm_compiler_rt_build.to_string_lossy().as_ref(),
"-G",
"Ninja",
generator,
])
.args(CMAKE_STATIC_ARGS)
.args(cmake_dynamic_args(build_type, target_env)?)
@@ -131,7 +153,17 @@ pub fn build(
"LLVM compiler-rt building cmake",
)?;
crate::utils::ninja(&llvm_compiler_rt_build)?;
crate::utils::command(
Command::new("cmake").args([
"--build",
llvm_compiler_rt_build.to_string_lossy().as_ref(),
"--target",
"install",
"--config",
build_type.to_string().as_str(),
]),
"Building",
)?;
Ok(())
}
+1 -1
View File
@@ -214,7 +214,7 @@ pub fn build(
sanitizer,
)?;
} else if cfg!(target_os = "windows") {
platforms::x86_64_windows_gnu::build(
platforms::x86_64_windows_msvc::build(
build_type,
targets,
llvm_projects,
+1 -1
View File
@@ -8,7 +8,7 @@ pub mod wasm32_emscripten;
pub mod x86_64_linux_gnu;
pub mod x86_64_linux_musl;
pub mod x86_64_macos;
pub mod x86_64_windows_gnu;
pub mod x86_64_windows_msvc;
use std::str::FromStr;
+3 -1
View File
@@ -8,7 +8,7 @@ use std::path::Path;
use std::process::Command;
/// The build options shared by all platforms.
pub const SHARED_BUILD_OPTS: [&str; 19] = [
pub const SHARED_BUILD_OPTS: [&str; 21] = [
"-DPACKAGE_VENDOR='Parity Technologies'",
"-DCMAKE_BUILD_WITH_INSTALL_RPATH=1",
"-DLLVM_BUILD_DOCS='Off'",
@@ -28,6 +28,8 @@ pub const SHARED_BUILD_OPTS: [&str; 19] = [
"-DCMAKE_EXPORT_COMPILE_COMMANDS='On'",
"-DPython3_FIND_REGISTRY='LAST'", // Use Python version from $PATH, not from registry
"-DBUG_REPORT_URL='https://github.com/paritytech/contract-issues/issues/'",
"-DCLANG_ENABLE_ARCMT='Off'",
"-DCLANG_ENABLE_STATIC_ANALYZER='Off'",
];
/// The build options shared by all platforms except MUSL.
@@ -1,7 +1,6 @@
//! The revive LLVM amd64 `windows-gnu` builder.
use std::collections::HashSet;
use std::path::PathBuf;
use std::process::Command;
use crate::build_type::BuildType;
@@ -28,10 +27,6 @@ pub fn build(
sanitizer: Option<Sanitizer>,
) -> anyhow::Result<()> {
crate::utils::check_presence("cmake")?;
crate::utils::check_presence("clang")?;
crate::utils::check_presence("clang++")?;
crate::utils::check_presence("lld")?;
crate::utils::check_presence("ninja")?;
let llvm_module_llvm =
LLVMPath::llvm_module_llvm().and_then(crate::utils::path_windows_to_unix)?;
@@ -48,15 +43,12 @@ pub fn build(
"-B",
llvm_build_final.to_string_lossy().as_ref(),
"-G",
"Ninja",
"Visual Studio 17 2022",
format!(
"-DCMAKE_INSTALL_PREFIX='{}'",
llvm_target_final.to_string_lossy().as_ref(),
)
.as_str(),
format!("-DCMAKE_BUILD_TYPE='{build_type}'").as_str(),
"-DCMAKE_C_COMPILER='clang'",
"-DCMAKE_CXX_COMPILER='clang++'",
format!(
"-DLLVM_TARGETS_TO_BUILD='{}'",
targets
@@ -75,7 +67,7 @@ pub fn build(
.join(";")
)
.as_str(),
"-DLLVM_USE_LINKER='lld'",
"-DLLVM_BUILD_LLVM_C_DYLIB=Off",
])
.args(crate::platforms::shared::shared_build_opts_default_target(
default_target,
@@ -107,20 +99,16 @@ pub fn build(
"LLVM building cmake",
)?;
crate::utils::ninja(llvm_build_final.as_ref())?;
let libstdcpp_source_path = match std::env::var("LIBSTDCPP_SOURCE_PATH") {
Ok(libstdcpp_source_path) => PathBuf::from(libstdcpp_source_path),
Err(error) => anyhow::bail!(
"The `LIBSTDCPP_SOURCE_PATH` must be set to the path to the libstdc++.a static library: {}", error
),
};
let mut libstdcpp_destination_path = llvm_target_final;
libstdcpp_destination_path.push("./lib/libstdc++.a");
fs_extra::file::copy(
crate::utils::path_windows_to_unix(libstdcpp_source_path)?,
crate::utils::path_windows_to_unix(libstdcpp_destination_path)?,
&fs_extra::file::CopyOptions::default(),
crate::utils::command(
Command::new("cmake").args([
"--build",
llvm_build_final.to_string_lossy().as_ref(),
"--target",
"install",
"--config",
build_type.to_string().as_str(),
]),
"Building with msbuild",
)?;
Ok(())
+3 -5
View File
@@ -7,6 +7,7 @@ use std::process::Command;
use std::process::Stdio;
use std::time::Duration;
use anyhow::Context;
use path_slash::PathBufExt;
/// The LLVM host repository URL.
@@ -131,11 +132,8 @@ pub fn path_windows_to_unix<P: AsRef<Path> + PathBufExt>(path: P) -> anyhow::Res
/// Checks if the tool exists in the system.
pub fn check_presence(name: &str) -> anyhow::Result<()> {
let description = &format!("checking the `{name}` executable");
log::info!("{description}");
command(Command::new("which").arg(name), description)
.map_err(|_| anyhow::anyhow!("Tool `{}` is missing. Please install", name))
which::which(name).with_context(|| format!("Tool `{name}` is missing. Please install"))?;
Ok(())
}
/// Identify XCode version using `pkgutil`.
+44 -9
View File
@@ -53,6 +53,16 @@ fn clone_build_and_clean_musl() -> anyhow::Result<()> {
.assert()
.success();
Command::cargo_bin(common::REVIVE_LLVM)?
.current_dir(test_dir.path())
.arg("build")
.arg("--llvm-projects")
.arg("clang")
.arg("--llvm-projects")
.arg("lld")
.assert()
.success();
Command::cargo_bin(common::REVIVE_LLVM)?
.arg("--target-env")
.arg("musl")
@@ -65,12 +75,6 @@ fn clone_build_and_clean_musl() -> anyhow::Result<()> {
.assert()
.success();
Command::cargo_bin(common::REVIVE_LLVM)?
.current_dir(test_dir.path())
.arg("builtins")
.assert()
.success();
Command::cargo_bin(common::REVIVE_LLVM)?
.current_dir(test_dir.path())
.arg("clean")
@@ -83,6 +87,7 @@ fn clone_build_and_clean_musl() -> anyhow::Result<()> {
/// This test verifies that the LLVM repository can be successfully cloned and built in debug mode
/// with tests and coverage enabled.
#[test]
#[cfg(target_os = "linux")]
fn debug_build_with_tests_coverage() -> anyhow::Result<()> {
let test_dir = common::TestDir::with_lockfile(None)?;
@@ -99,6 +104,10 @@ fn debug_build_with_tests_coverage() -> anyhow::Result<()> {
.arg("--enable-tests")
.arg("--build-type")
.arg("Debug")
.arg("--llvm-projects")
.arg("clang")
.arg("--llvm-projects")
.arg("lld")
.assert()
.success();
@@ -107,6 +116,7 @@ fn debug_build_with_tests_coverage() -> anyhow::Result<()> {
/// This test verifies that the LLVM repository can be successfully built with address sanitizer.
#[test]
#[cfg(target_os = "linux")]
fn build_with_sanitizers() -> anyhow::Result<()> {
let test_dir = common::TestDir::with_lockfile(None)?;
@@ -121,6 +131,10 @@ fn build_with_sanitizers() -> anyhow::Result<()> {
.arg("build")
.arg("--sanitizer")
.arg("Address")
.arg("--llvm-projects")
.arg("lld")
.arg("--llvm-projects")
.arg("clang")
.assert()
.success();
@@ -129,15 +143,36 @@ fn build_with_sanitizers() -> anyhow::Result<()> {
/// Tests the clone, build, and clean process of the LLVM repository for the emscripten target.
#[test]
#[cfg(any(target_os = "linux", target_os = "macos"))]
#[cfg(target_os = "linux")]
fn clone_build_and_clean_emscripten() -> anyhow::Result<()> {
let test_dir = common::TestDir::with_lockfile(None)?;
let command = Command::cargo_bin(common::REVIVE_LLVM)?;
let program = command.get_program().to_string_lossy();
Command::cargo_bin(common::REVIVE_LLVM)?
.current_dir(test_dir.path())
.arg("clone")
.assert()
.success();
Command::cargo_bin(common::REVIVE_LLVM)?
.current_dir(test_dir.path())
.arg("build")
.arg("--llvm-projects")
.arg("lld")
.arg("--llvm-projects")
.arg("clang")
.assert()
.success();
// Two little shell-dependent things here:
// Doing `. ./emsdk_env.sh` instead of `source`, as `source` might be missing in some shells
// `cd {} && . ./emsdk_env.sh && cd ..` helps the script to locate `emsdk.py`
// @see https://github.com/emscripten-core/emsdk/blob/9dbdc4b3437750b85d16931c7c801bb71a782122/emsdk_env.sh#L61-L69
let emsdk_wrapped_build_command = format!(
"{program} --target-env emscripten clone && \
source {}emsdk_env.sh && \
{program} --target-env emscripten build --llvm-projects clang --llvm-projects lld",
cd {} && . ./emsdk_env.sh && cd .. && \
{program} --target-env emscripten build --llvm-projects lld",
revive_llvm_builder::LLVMPath::DIRECTORY_EMSDK_SOURCE,
);
+2
View File
@@ -22,6 +22,7 @@ num = { workspace = true }
hex = { workspace = true }
sha3 = { workspace = true }
inkwell = { workspace = true }
libc = { workspace = true }
polkavm-disassembler = { workspace = true }
polkavm-common = { workspace = true }
@@ -29,3 +30,4 @@ revive-common = { workspace = true }
revive-runtime-api = { workspace = true }
revive-linker = { workspace = true }
revive-stdlib = { workspace = true }
revive-solc-json-interface = { workspace = true }
+53 -10
View File
@@ -1,9 +1,7 @@
//! The LLVM context library.
pub(crate) mod debug_config;
pub(crate) mod optimizer;
pub(crate) mod polkavm;
pub(crate) mod target_machine;
use std::ffi::CString;
use std::sync::OnceLock;
pub use self::debug_config::ir_type::IRType as DebugConfigIR;
pub use self::debug_config::DebugConfig;
@@ -21,17 +19,27 @@ pub use self::polkavm::context::function::declaration::Declaration as PolkaVMFun
pub use self::polkavm::context::function::intrinsics::Intrinsics as PolkaVMIntrinsicFunction;
pub use self::polkavm::context::function::llvm_runtime::LLVMRuntime as PolkaVMLLVMRuntime;
pub use self::polkavm::context::function::r#return::Return as PolkaVMFunctionReturn;
pub use self::polkavm::context::function::runtime::arithmetics::Division as PolkaVMDivisionFunction;
pub use self::polkavm::context::function::runtime::arithmetics::Remainder as PolkaVMRemainderFunction;
pub use self::polkavm::context::function::runtime::arithmetics::SignedDivision as PolkaVMSignedDivisionFunction;
pub use self::polkavm::context::function::runtime::arithmetics::SignedRemainder as PolkaVMSignedRemainderFunction;
pub use self::polkavm::context::function::runtime::deploy_code::DeployCode as PolkaVMDeployCodeFunction;
pub use self::polkavm::context::function::runtime::entry::Entry as PolkaVMEntryFunction;
pub use self::polkavm::context::function::runtime::immutable_data_load::ImmutableDataLoad as PolkaVMImmutableDataLoadFunction;
pub use self::polkavm::context::function::runtime::revive::Exit as PolkaVMExitFunction;
pub use self::polkavm::context::function::runtime::revive::WordToPointer as PolkaVMWordToPointerFunction;
pub use self::polkavm::context::function::runtime::runtime_code::RuntimeCode as PolkaVMRuntimeCodeFunction;
pub use self::polkavm::context::function::runtime::FUNCTION_DEPLOY_CODE as PolkaVMFunctionDeployCode;
pub use self::polkavm::context::function::runtime::FUNCTION_ENTRY as PolkaVMFunctionEntry;
pub use self::polkavm::context::function::runtime::FUNCTION_LOAD_IMMUTABLE_DATA as PolkaVMFunctionImmutableDataLoad;
pub use self::polkavm::context::function::runtime::FUNCTION_RUNTIME_CODE as PolkaVMFunctionRuntimeCode;
pub use self::polkavm::context::function::yul_data::YulData as PolkaVMFunctionYulData;
pub use self::polkavm::context::function::Function as PolkaVMFunction;
pub use self::polkavm::context::global::Global as PolkaVMGlobal;
pub use self::polkavm::context::pointer::heap::LoadWord as PolkaVMLoadHeapWordFunction;
pub use self::polkavm::context::pointer::heap::StoreWord as PolkaVMStoreHeapWordFunction;
pub use self::polkavm::context::pointer::storage::LoadTransientWord as PolkaVMLoadTransientStorageWordFunction;
pub use self::polkavm::context::pointer::storage::LoadWord as PolkaVMLoadStorageWordFunction;
pub use self::polkavm::context::pointer::storage::StoreTransientWord as PolkaVMStoreTransientStorageWordFunction;
pub use self::polkavm::context::pointer::storage::StoreWord as PolkaVMStoreStorageWordFunction;
pub use self::polkavm::context::pointer::Pointer as PolkaVMPointer;
pub use self::polkavm::context::r#loop::Loop as PolkaVMLoop;
pub use self::polkavm::context::solidity_data::SolidityData as PolkaVMContextSolidityData;
@@ -47,14 +55,17 @@ pub use self::polkavm::evm::create as polkavm_evm_create;
pub use self::polkavm::evm::crypto as polkavm_evm_crypto;
pub use self::polkavm::evm::ether_gas as polkavm_evm_ether_gas;
pub use self::polkavm::evm::event as polkavm_evm_event;
pub use self::polkavm::evm::event::EventLog as PolkaVMEventLogFunction;
pub use self::polkavm::evm::ext_code as polkavm_evm_ext_code;
pub use self::polkavm::evm::immutable as polkavm_evm_immutable;
pub use self::polkavm::evm::immutable::Load as PolkaVMLoadImmutableDataFunction;
pub use self::polkavm::evm::immutable::Store as PolkaVMStoreImmutableDataFunction;
pub use self::polkavm::evm::math as polkavm_evm_math;
pub use self::polkavm::evm::memory as polkavm_evm_memory;
pub use self::polkavm::evm::r#return as polkavm_evm_return;
pub use self::polkavm::evm::return_data as polkavm_evm_return_data;
pub use self::polkavm::evm::storage as polkavm_evm_storage;
pub use self::polkavm::metadata_hash::MetadataHash as PolkaVMMetadataHash;
pub use self::polkavm::r#const as polkavm_const;
pub use self::polkavm::Dependency as PolkaVMDependency;
pub use self::polkavm::DummyDependency as PolkaVMDummyDependency;
@@ -63,9 +74,41 @@ pub use self::polkavm::WriteLLVM as PolkaVMWriteLLVM;
pub use self::target_machine::target::Target;
pub use self::target_machine::TargetMachine;
/// Initializes the target machine.
pub fn initialize_target(target: Target) {
pub(crate) mod debug_config;
pub(crate) mod optimizer;
pub(crate) mod polkavm;
pub(crate) mod target_machine;
static DID_INITIALIZE: OnceLock<()> = OnceLock::new();
/// Initializes the LLVM compiler backend.
///
/// This is a no-op if called subsequentially.
///
/// `llvm_arguments` are passed as-is to the LLVM CL options parser.
pub fn initialize_llvm(target: Target, name: &str, llvm_arguments: &[String]) {
let Ok(_) = DID_INITIALIZE.set(()) else {
return; // Tests don't go through a recursive process
};
let argv = [name.to_string()]
.iter()
.chain(llvm_arguments)
.map(|arg| CString::new(arg.as_bytes()).unwrap())
.collect::<Vec<_>>();
let argv: Vec<*const libc::c_char> = argv.iter().map(|arg| arg.as_ptr()).collect();
let overview = CString::new("").unwrap();
unsafe {
inkwell::llvm_sys::support::LLVMParseCommandLineOptions(
argv.len() as i32,
argv.as_ptr(),
overview.as_ptr(),
);
}
inkwell::support::enable_llvm_pretty_stack_trace();
match target {
Target::PVM => self::polkavm::initialize_target(),
Target::PVM => inkwell::targets::Target::initialize_riscv(&Default::default()),
}
}
@@ -2,6 +2,7 @@
pub mod size_level;
use revive_solc_json_interface::SolcStandardJsonInputSettingsOptimizer;
use serde::Deserialize;
use serde::Serialize;
@@ -226,3 +227,18 @@ impl std::fmt::Display for Settings {
)
}
}
impl TryFrom<&SolcStandardJsonInputSettingsOptimizer> for Settings {
type Error = anyhow::Error;
fn try_from(value: &SolcStandardJsonInputSettingsOptimizer) -> Result<Self, Self::Error> {
let mut result = match value.mode {
Some(mode) => Self::try_from_cli(mode)?,
None => Self::cycles(),
};
if value.fallback_to_optimizing_for_size.unwrap_or_default() {
result.enable_fallback_to_size();
}
Ok(result)
}
}
@@ -9,6 +9,9 @@ pub static XLEN: usize = revive_common::BIT_LENGTH_X32;
/// The calldata size global variable name.
pub static GLOBAL_CALLDATA_SIZE: &str = "calldatasize";
/// The spill buffer global variable name.
pub static GLOBAL_ADDRESS_SPILL_BUFFER: &str = "address_spill_buffer";
/// The deployer call header size that consists of:
/// - bytecode hash (32 bytes)
pub const DEPLOYER_CALL_HEADER_SIZE: usize = revive_common::BYTE_LENGTH_WORD;
@@ -8,10 +8,6 @@ pub enum AddressSpace {
Stack,
/// The heap memory.
Heap,
/// The generic memory page.
Storage,
/// The transient storage.
TransientStorage,
}
impl From<AddressSpace> for inkwell::AddressSpace {
@@ -19,8 +15,6 @@ impl From<AddressSpace> for inkwell::AddressSpace {
match value {
AddressSpace::Stack => Self::from(0),
AddressSpace::Heap => Self::from(1),
AddressSpace::Storage => Self::from(5),
AddressSpace::TransientStorage => Self::from(6),
}
}
}
@@ -4,61 +4,98 @@
#[derive(Debug, Clone)]
pub struct Argument<'ctx> {
/// The actual LLVM operand.
pub value: inkwell::values::BasicValueEnum<'ctx>,
pub value: Value<'ctx>,
/// The original AST value. Used mostly for string literals.
pub original: Option<String>,
/// The preserved constant value, if available.
pub constant: Option<num::BigUint>,
}
/// The function argument can be either a pointer or a integer value.
/// This disambiguation allows for lazy loading of variables.
#[derive(Clone, Debug)]
pub enum Value<'ctx> {
Register(inkwell::values::BasicValueEnum<'ctx>),
Pointer {
pointer: crate::polkavm::context::Pointer<'ctx>,
id: String,
},
}
impl<'ctx> Argument<'ctx> {
/// The calldata offset argument index.
pub const ARGUMENT_INDEX_CALLDATA_OFFSET: usize = 0;
/// The calldata length argument index.
pub const ARGUMENT_INDEX_CALLDATA_LENGTH: usize = 1;
/// A shortcut constructor.
pub fn new(value: inkwell::values::BasicValueEnum<'ctx>) -> Self {
/// A shortcut constructor for register arguments.
pub fn value(value: inkwell::values::BasicValueEnum<'ctx>) -> Self {
Self {
value,
value: Value::Register(value),
original: None,
constant: None,
}
}
/// A shortcut constructor.
pub fn new_with_original(
value: inkwell::values::BasicValueEnum<'ctx>,
original: String,
) -> Self {
/// A shortcut constructor for stack arguments.
pub fn pointer(pointer: crate::polkavm::context::Pointer<'ctx>, id: String) -> Self {
Self {
value,
original: Some(original),
value: Value::Pointer { pointer, id },
original: None,
constant: None,
}
}
/// A shortcut constructor.
pub fn new_with_constant(
value: inkwell::values::BasicValueEnum<'ctx>,
constant: num::BigUint,
) -> Self {
Self {
value,
original: None,
constant: Some(constant),
}
/// Set the original decleratation value.
pub fn with_original(mut self, original: String) -> Self {
self.original = Some(original);
self
}
/// Set the constant value.
pub fn with_constant(mut self, constant: num::BigUint) -> Self {
self.constant = Some(constant);
self
}
/// Returns the inner LLVM value.
pub fn to_llvm(&self) -> inkwell::values::BasicValueEnum<'ctx> {
self.value
///
/// Panics if `self` is a pointer argument.
pub fn _to_llvm_value(&self) -> inkwell::values::BasicValueEnum<'ctx> {
match &self.value {
Value::Register(value) => *value,
Value::Pointer { .. } => unreachable!("invalid register value access"),
}
}
/// Access the underlying value.
///
/// Will emit a stack load if `self` is a pointer argument.
pub fn access<D: crate::polkavm::Dependency + Clone>(
&self,
context: &crate::polkavm::context::Context<'ctx, D>,
) -> anyhow::Result<inkwell::values::BasicValueEnum<'ctx>> {
match &self.value {
Value::Register(value) => Ok(*value),
Value::Pointer { pointer, id } => context.build_load(*pointer, id),
}
}
/// Access the underlying value.
///
/// Will emit a stack load if `self` is a pointer argument.
pub fn as_pointer<D: crate::polkavm::Dependency + Clone>(
&self,
context: &crate::polkavm::context::Context<'ctx, D>,
) -> anyhow::Result<crate::polkavm::context::Pointer<'ctx>> {
match &self.value {
Value::Register(value) => {
let pointer = context.build_alloca_at_entry(context.word_type(), "pvm_arg");
context.build_store(pointer, *value)?;
Ok(pointer)
}
Value::Pointer { pointer, .. } => Ok(*pointer),
}
}
}
impl<'ctx> From<inkwell::values::BasicValueEnum<'ctx>> for Argument<'ctx> {
fn from(value: inkwell::values::BasicValueEnum<'ctx>) -> Self {
Self::new(value)
Self::value(value)
}
}
@@ -1,9 +1,6 @@
//! The LLVM runtime functions.
use inkwell::types::BasicType;
use crate::optimizer::Optimizer;
use crate::polkavm::context::address_space::AddressSpace;
use crate::polkavm::context::function::declaration::Declaration as FunctionDeclaration;
use crate::polkavm::context::function::Function;
@@ -19,9 +16,6 @@ pub struct LLVMRuntime<'ctx> {
pub exp: FunctionDeclaration<'ctx>,
/// The corresponding LLVM runtime function.
pub sign_extend: FunctionDeclaration<'ctx>,
/// The corresponding LLVM runtime function.
pub sha3: FunctionDeclaration<'ctx>,
}
impl<'ctx> LLVMRuntime<'ctx> {
@@ -37,9 +31,6 @@ impl<'ctx> LLVMRuntime<'ctx> {
/// The corresponding runtime function name.
pub const FUNCTION_SIGNEXTEND: &'static str = "__signextend";
/// The corresponding runtime function name.
pub const FUNCTION_SHA3: &'static str = "__sha3";
/// A shortcut constructor.
pub fn new(
llvm: &'ctx inkwell::context::Context,
@@ -65,43 +56,11 @@ impl<'ctx> LLVMRuntime<'ctx> {
Function::set_default_attributes(llvm, sign_extend, optimizer);
Function::set_pure_function_attributes(llvm, sign_extend);
let sha3 = Self::declare(
module,
Self::FUNCTION_SHA3,
llvm.custom_width_int_type(revive_common::BIT_LENGTH_WORD as u32)
.fn_type(
vec![
llvm.ptr_type(AddressSpace::Heap.into())
.as_basic_type_enum()
.into(),
llvm.custom_width_int_type(revive_common::BIT_LENGTH_WORD as u32)
.as_basic_type_enum()
.into(),
llvm.custom_width_int_type(revive_common::BIT_LENGTH_BOOLEAN as u32)
.as_basic_type_enum()
.into(),
]
.as_slice(),
false,
),
Some(inkwell::module::Linkage::External),
);
Function::set_default_attributes(llvm, sha3, optimizer);
Function::set_attributes(
llvm,
sha3,
//vec![Attribute::ArgMemOnly, Attribute::ReadOnly],
vec![],
false,
);
Self {
add_mod,
mul_mod,
exp,
sign_extend,
sha3,
}
}
@@ -81,8 +81,7 @@ impl<'ctx> Function<'ctx> {
|| (name.starts_with("__")
&& name != self::runtime::FUNCTION_ENTRY
&& name != self::runtime::FUNCTION_DEPLOY_CODE
&& name != self::runtime::FUNCTION_RUNTIME_CODE
&& name != self::runtime::FUNCTION_LOAD_IMMUTABLE_DATA)
&& name != self::runtime::FUNCTION_RUNTIME_CODE)
}
/// Returns the LLVM function declaration.
@@ -110,30 +109,21 @@ impl<'ctx> Function<'ctx> {
pub fn set_attributes(
llvm: &'ctx inkwell::context::Context,
declaration: Declaration<'ctx>,
attributes: Vec<Attribute>,
attributes: &[Attribute],
force: bool,
) {
for attribute_kind in attributes.into_iter() {
for attribute_kind in attributes {
match attribute_kind {
Attribute::Memory => unimplemented!("`memory` attributes are not implemented"),
attribute_kind @ Attribute::AlwaysInline if force => {
let is_optimize_none_set = declaration
.value
.get_enum_attribute(
inkwell::attributes::AttributeLoc::Function,
Attribute::OptimizeNone as u32,
)
.is_some();
if !is_optimize_none_set {
declaration.value.remove_enum_attribute(
inkwell::attributes::AttributeLoc::Function,
Attribute::NoInline as u32,
);
declaration.value.add_attribute(
inkwell::attributes::AttributeLoc::Function,
llvm.create_enum_attribute(attribute_kind as u32, 0),
);
}
declaration.value.remove_enum_attribute(
inkwell::attributes::AttributeLoc::Function,
Attribute::NoInline as u32,
);
declaration.value.add_attribute(
inkwell::attributes::AttributeLoc::Function,
llvm.create_enum_attribute(*attribute_kind as u32, 0),
);
}
attribute_kind @ Attribute::NoInline if force => {
declaration.value.remove_enum_attribute(
@@ -142,12 +132,12 @@ impl<'ctx> Function<'ctx> {
);
declaration.value.add_attribute(
inkwell::attributes::AttributeLoc::Function,
llvm.create_enum_attribute(attribute_kind as u32, 0),
llvm.create_enum_attribute(*attribute_kind as u32, 0),
);
}
attribute_kind => declaration.value.add_attribute(
inkwell::attributes::AttributeLoc::Function,
llvm.create_enum_attribute(attribute_kind as u32, 0),
llvm.create_enum_attribute(*attribute_kind as u32, 0),
),
}
}
@@ -178,27 +168,16 @@ impl<'ctx> Function<'ctx> {
declaration: Declaration<'ctx>,
optimizer: &Optimizer,
) {
if optimizer.settings().level_middle_end == inkwell::OptimizationLevel::None {
Self::remove_attributes(
declaration,
&[Attribute::OptimizeForSize, Attribute::AlwaysInline],
);
if optimizer.settings().level_middle_end_size == SizeLevel::Z {
Self::set_attributes(
llvm,
declaration,
vec![Attribute::OptimizeNone, Attribute::NoInline],
false,
);
} else if optimizer.settings().level_middle_end_size == SizeLevel::Z {
Self::set_attributes(
llvm,
declaration,
vec![Attribute::OptimizeForSize, Attribute::MinSize],
&[Attribute::OptimizeForSize, Attribute::MinSize],
false,
);
}
Self::set_attributes(llvm, declaration, vec![Attribute::NoFree], false);
Self::set_attributes(llvm, declaration, &[Attribute::NoFree], false);
}
/// Sets the front-end runtime attributes.
@@ -208,7 +187,7 @@ impl<'ctx> Function<'ctx> {
optimizer: &Optimizer,
) {
if optimizer.settings().level_middle_end_size == SizeLevel::Z {
Self::set_attributes(llvm, declaration, vec![Attribute::NoInline], false);
Self::set_attributes(llvm, declaration, &[Attribute::NoInline], false);
}
}
@@ -220,7 +199,7 @@ impl<'ctx> Function<'ctx> {
Self::set_attributes(
llvm,
declaration,
vec![
&[
Attribute::MustProgress,
Attribute::NoUnwind,
Attribute::WillReturn,
@@ -0,0 +1,269 @@
//! Translates the arithmetic operations.
use inkwell::values::BasicValue;
use crate::polkavm::context::runtime::RuntimeFunction;
use crate::polkavm::context::Context;
use crate::polkavm::Dependency;
use crate::polkavm::WriteLLVM;
/// Implements the division operator according to the EVM specification.
pub struct Division;
impl<D> RuntimeFunction<D> for Division
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_division";
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context.word_type().fn_type(
&[context.word_type().into(), context.word_type().into()],
false,
)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<inkwell::values::BasicValueEnum<'ctx>>> {
let operand_1 = Self::paramater(context, 0).into_int_value();
let operand_2 = Self::paramater(context, 1).into_int_value();
wrapped_division(context, operand_2, || {
Ok(context
.builder()
.build_int_unsigned_div(operand_1, operand_2, "DIV")?)
})
.map(Into::into)
}
}
impl<D> WriteLLVM<D> for Division
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
/// Implements the signed division operator according to the EVM specification.
pub struct SignedDivision;
impl<D> RuntimeFunction<D> for SignedDivision
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_signed_division";
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context.word_type().fn_type(
&[context.word_type().into(), context.word_type().into()],
false,
)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<inkwell::values::BasicValueEnum<'ctx>>> {
let operand_1 = Self::paramater(context, 0).into_int_value();
let operand_2 = Self::paramater(context, 1).into_int_value();
let block_calculate = context.append_basic_block("calculate");
let block_overflow = context.append_basic_block("overflow");
let block_select = context.append_basic_block("select_result");
let block_origin = context.basic_block();
context.builder().build_switch(
operand_2,
block_calculate,
&[
(context.word_type().const_zero(), block_select),
(context.word_type().const_all_ones(), block_overflow),
],
)?;
context.set_basic_block(block_calculate);
let quotient = context
.builder()
.build_int_signed_div(operand_1, operand_2, "SDIV")?;
context.build_unconditional_branch(block_select);
context.set_basic_block(block_overflow);
let max_uint = context.builder().build_int_z_extend(
context
.integer_type(revive_common::BIT_LENGTH_WORD - 1)
.const_all_ones(),
context.word_type(),
"max_uint",
)?;
let is_operand_1_overflow = context.builder().build_int_compare(
inkwell::IntPredicate::EQ,
operand_1,
context.builder().build_int_neg(max_uint, "min_uint")?,
"is_operand_1_overflow",
)?;
context.build_conditional_branch(is_operand_1_overflow, block_select, block_calculate)?;
context.set_basic_block(block_select);
let result = context.builder().build_phi(context.word_type(), "result")?;
result.add_incoming(&[
(&operand_1, block_overflow),
(&context.word_const(0), block_origin),
(&quotient.as_basic_value_enum(), block_calculate),
]);
Ok(Some(result.as_basic_value()))
}
}
impl<D> WriteLLVM<D> for SignedDivision
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
/// Implements the remainder operator according to the EVM specification.
pub struct Remainder;
impl<D> RuntimeFunction<D> for Remainder
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_remainder";
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context.word_type().fn_type(
&[context.word_type().into(), context.word_type().into()],
false,
)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<inkwell::values::BasicValueEnum<'ctx>>> {
let operand_1 = Self::paramater(context, 0).into_int_value();
let operand_2 = Self::paramater(context, 1).into_int_value();
wrapped_division(context, operand_2, || {
Ok(context
.builder()
.build_int_unsigned_rem(operand_1, operand_2, "MOD")?)
})
.map(Into::into)
}
}
impl<D> WriteLLVM<D> for Remainder
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
/// Implements the signed remainder operator according to the EVM specification.
pub struct SignedRemainder;
impl<D> RuntimeFunction<D> for SignedRemainder
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_signed_remainder";
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context.word_type().fn_type(
&[context.word_type().into(), context.word_type().into()],
false,
)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<inkwell::values::BasicValueEnum<'ctx>>> {
let operand_1 = Self::paramater(context, 0).into_int_value();
let operand_2 = Self::paramater(context, 1).into_int_value();
wrapped_division(context, operand_2, || {
Ok(context
.builder()
.build_int_signed_rem(operand_1, operand_2, "SMOD")?)
})
.map(Into::into)
}
}
impl<D> WriteLLVM<D> for SignedRemainder
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
/// Wrap division operations so that zero will be returned if the
/// denominator is zero (see also Ethereum YP Appendix H.2).
///
/// The closure is expected to calculate and return the quotient.
///
/// The result is either the calculated quotient or zero,
/// selected at runtime.
fn wrapped_division<'ctx, D, F, T>(
context: &Context<'ctx, D>,
denominator: inkwell::values::IntValue<'ctx>,
f: F,
) -> anyhow::Result<inkwell::values::BasicValueEnum<'ctx>>
where
D: Dependency + Clone,
F: FnOnce() -> anyhow::Result<T>,
T: inkwell::values::IntMathValue<'ctx>,
{
assert_eq!(
denominator.get_type().get_bit_width(),
revive_common::BIT_LENGTH_WORD as u32
);
let block_calculate = context.append_basic_block("calculate");
let block_select = context.append_basic_block("select");
let block_origin = context.basic_block();
context.builder().build_switch(
denominator,
block_calculate,
&[(context.word_const(0), block_select)],
)?;
context.set_basic_block(block_calculate);
let calculated_value = f()?.as_basic_value_enum();
context.build_unconditional_branch(block_select);
context.set_basic_block(block_select);
let result = context.builder().build_phi(context.word_type(), "result")?;
result.add_incoming(&[
(&context.word_const(0), block_origin),
(&calculated_value, block_calculate),
]);
Ok(result.as_basic_value())
}
@@ -31,6 +31,14 @@ impl Entry {
context.xlen_type().get_undef(),
);
let address_type = context.integer_type(revive_common::BIT_LENGTH_ETH_ADDRESS);
context.set_global(
crate::polkavm::GLOBAL_ADDRESS_SPILL_BUFFER,
address_type,
AddressSpace::Stack,
address_type.const_zero(),
);
Ok(())
}
@@ -145,7 +153,7 @@ where
crate::PolkaVMFunction::set_attributes(
context.llvm(),
entry,
vec![crate::PolkaVMAttribute::NoReturn],
&[crate::PolkaVMAttribute::NoReturn],
true,
);
@@ -1,118 +0,0 @@
//! The immutable data runtime function.
use crate::polkavm::context::address_space::AddressSpace;
use crate::polkavm::context::function::runtime;
use crate::polkavm::context::pointer::Pointer;
use crate::polkavm::context::Context;
use crate::polkavm::Dependency;
use crate::polkavm::WriteLLVM;
/// A function for requesting the immutable data from the runtime.
/// This is a special function that is only used by the front-end generated code.
///
/// The runtime API is called lazily and subsequent calls are no-ops.
///
/// The bytes written is asserted to match the expected length.
/// This should never fail; the length is known.
/// However, this is a one time assertion, hence worth it.
#[derive(Debug)]
pub struct ImmutableDataLoad;
impl<D> WriteLLVM<D> for ImmutableDataLoad
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
context.add_function(
runtime::FUNCTION_LOAD_IMMUTABLE_DATA,
context.void_type().fn_type(Default::default(), false),
0,
Some(inkwell::module::Linkage::External),
)?;
Ok(())
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
context.set_current_function(runtime::FUNCTION_LOAD_IMMUTABLE_DATA, None)?;
context.set_basic_block(context.current_function().borrow().entry_block());
let immutable_data_size_pointer = context
.get_global(revive_runtime_api::immutable_data::GLOBAL_IMMUTABLE_DATA_SIZE)?
.value
.as_pointer_value();
let immutable_data_size = context.build_load(
Pointer::new(
context.xlen_type(),
AddressSpace::Stack,
immutable_data_size_pointer,
),
"immutable_data_size_load",
)?;
let load_immutable_data_block = context.append_basic_block("load_immutables_block");
let return_block = context.current_function().borrow().return_block();
let immutable_data_size_is_zero = context.builder().build_int_compare(
inkwell::IntPredicate::EQ,
context.xlen_type().const_zero(),
immutable_data_size.into_int_value(),
"immutable_data_size_is_zero",
)?;
context.build_conditional_branch(
immutable_data_size_is_zero,
return_block,
load_immutable_data_block,
)?;
context.set_basic_block(load_immutable_data_block);
let output_pointer = context
.get_global(revive_runtime_api::immutable_data::GLOBAL_IMMUTABLE_DATA_POINTER)?
.value
.as_pointer_value();
context.build_runtime_call(
revive_runtime_api::polkavm_imports::GET_IMMUTABLE_DATA,
&[
context
.builder()
.build_ptr_to_int(output_pointer, context.xlen_type(), "ptr_to_xlen")?
.into(),
context
.builder()
.build_ptr_to_int(
immutable_data_size_pointer,
context.xlen_type(),
"ptr_to_xlen",
)?
.into(),
],
);
let bytes_written = context.builder().build_load(
context.xlen_type(),
immutable_data_size_pointer,
"bytes_written",
)?;
context.builder().build_store(
immutable_data_size_pointer,
context.xlen_type().const_zero(),
)?;
let overflow_block = context.append_basic_block("immutable_data_overflow");
let is_overflow = context.builder().build_int_compare(
inkwell::IntPredicate::UGT,
immutable_data_size.into_int_value(),
bytes_written.into_int_value(),
"is_overflow",
)?;
context.build_conditional_branch(is_overflow, overflow_block, return_block)?;
context.set_basic_block(overflow_block);
context.build_call(context.intrinsics().trap, &[], "invalid_trap");
context.build_unreachable();
context.set_basic_block(return_block);
context.build_return(None);
context.pop_debug_scope();
Ok(())
}
}
@@ -1,8 +1,9 @@
//! The front-end runtime functions.
pub mod arithmetics;
pub mod deploy_code;
pub mod entry;
pub mod immutable_data_load;
pub mod revive;
pub mod runtime_code;
/// The main entry function name.
@@ -13,6 +14,3 @@ pub const FUNCTION_DEPLOY_CODE: &str = "__deploy";
/// The runtime code function name.
pub const FUNCTION_RUNTIME_CODE: &str = "__runtime";
/// The immutable data load function name.
pub const FUNCTION_LOAD_IMMUTABLE_DATA: &str = "__immutable_data_load";
@@ -0,0 +1,147 @@
//! The revive compiler runtime functions.
use inkwell::values::BasicValue;
use crate::polkavm::context::function::Attribute;
use crate::polkavm::context::runtime::RuntimeFunction;
use crate::polkavm::context::Context;
use crate::polkavm::Dependency;
use crate::polkavm::WriteLLVM;
/// Pointers are represented as opaque 256 bit integer values in EVM.
/// In practice, they should never exceed a register sized bit value.
/// However, we still protect against this possibility here: Heap index
/// offsets are generally untrusted and potentially represent valid
/// (but wrong) pointers when truncated.
pub struct WordToPointer;
impl<D> RuntimeFunction<D> for WordToPointer
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_int_truncate";
const ATTRIBUTES: &'static [Attribute] = &[
Attribute::WillReturn,
Attribute::NoFree,
Attribute::AlwaysInline,
];
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context
.xlen_type()
.fn_type(&[context.word_type().into()], false)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<inkwell::values::BasicValueEnum<'ctx>>> {
let value = Self::paramater(context, 0).into_int_value();
let truncated =
context
.builder()
.build_int_truncate(value, context.xlen_type(), "offset_truncated")?;
let extended = context.builder().build_int_z_extend(
truncated,
context.word_type(),
"offset_extended",
)?;
let is_overflow = context.builder().build_int_compare(
inkwell::IntPredicate::NE,
value,
extended,
"compare_truncated_extended",
)?;
let block_continue = context.append_basic_block("offset_pointer_ok");
let block_trap = context.append_basic_block("offset_pointer_overflow");
context.build_conditional_branch(is_overflow, block_trap, block_continue)?;
context.set_basic_block(block_trap);
context.build_call(context.intrinsics().trap, &[], "invalid_trap");
context.build_unreachable();
context.set_basic_block(block_continue);
Ok(Some(truncated.as_basic_value_enum()))
}
}
impl<D> WriteLLVM<D> for WordToPointer
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
/// The revive runtime exit function.
pub struct Exit;
impl<D> RuntimeFunction<D> for Exit
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_exit";
const ATTRIBUTES: &'static [Attribute] = &[
Attribute::NoReturn,
Attribute::NoFree,
Attribute::AlwaysInline,
];
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context.void_type().fn_type(
&[
context.xlen_type().into(),
context.word_type().into(),
context.word_type().into(),
],
false,
)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<inkwell::values::BasicValueEnum<'ctx>>> {
let flags = Self::paramater(context, 0).into_int_value();
let offset = Self::paramater(context, 1).into_int_value();
let length = Self::paramater(context, 2).into_int_value();
let offset_truncated = context.safe_truncate_int_to_xlen(offset)?;
let length_truncated = context.safe_truncate_int_to_xlen(length)?;
let heap_pointer = context.build_heap_gep(offset_truncated, length_truncated)?;
let offset_pointer = context.builder().build_ptr_to_int(
heap_pointer.value,
context.xlen_type(),
"return_data_ptr_to_int",
)?;
context.build_runtime_call(
revive_runtime_api::polkavm_imports::RETURN,
&[flags.into(), offset_pointer.into(), length_truncated.into()],
);
context.build_unreachable();
Ok(None)
}
}
impl<D> WriteLLVM<D> for Exit
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
+95 -213
View File
@@ -10,6 +10,7 @@ pub mod function;
pub mod global;
pub mod r#loop;
pub mod pointer;
pub mod runtime;
pub mod solidity_data;
pub mod yul_data;
@@ -31,6 +32,8 @@ use crate::polkavm::DebugConfig;
use crate::polkavm::Dependency;
use crate::target_machine::target::Target;
use crate::target_machine::TargetMachine;
use crate::PolkaVMLoadHeapWordFunction;
use crate::PolkaVMStoreHeapWordFunction;
use self::address_space::AddressSpace;
use self::attribute::Attribute;
@@ -41,10 +44,13 @@ use self::function::declaration::Declaration as FunctionDeclaration;
use self::function::intrinsics::Intrinsics;
use self::function::llvm_runtime::LLVMRuntime;
use self::function::r#return::Return as FunctionReturn;
use self::function::runtime::revive::Exit;
use self::function::runtime::revive::WordToPointer;
use self::function::Function;
use self::global::Global;
use self::pointer::Pointer;
use self::r#loop::Loop;
use self::runtime::RuntimeFunction;
use self::solidity_data::SolidityData;
use self::yul_data::YulData;
@@ -77,6 +83,8 @@ where
current_function: Option<Rc<RefCell<Function<'ctx>>>>,
/// The loop context stack.
loop_stack: Vec<Loop<'ctx>>,
/// The extra LLVM arguments that were used during target initialization.
llvm_arguments: &'ctx [String],
/// The project dependency manager. It can be any entity implementing the trait.
/// The manager is used to get information about contracts and their dependencies during
@@ -217,6 +225,7 @@ where
dependency_manager: Option<D>,
include_metadata_hash: bool,
debug_config: DebugConfig,
llvm_arguments: &'ctx [String],
) -> Self {
Self::set_data_layout(llvm, &module);
Self::link_stdlib_module(llvm, &module);
@@ -244,6 +253,7 @@ where
functions: HashMap::with_capacity(Self::FUNCTIONS_HASHMAP_INITIAL_CAPACITY),
current_function: None,
loop_stack: Vec::with_capacity(Self::LOOP_STACK_INITIAL_CAPACITY),
llvm_arguments,
dependency_manager,
include_metadata_hash,
@@ -633,6 +643,7 @@ where
self.optimizer.settings().to_owned(),
self.include_metadata_hash,
self.debug_config.clone(),
self.llvm_arguments,
)
})
}
@@ -721,6 +732,7 @@ where
name: &str,
) -> Pointer<'ctx> {
let pointer = self.builder.build_alloca(r#type, name).unwrap();
pointer
.as_instruction()
.unwrap()
@@ -738,7 +750,9 @@ where
address: inkwell::values::IntValue<'ctx>,
) -> anyhow::Result<Pointer<'ctx>> {
let address_type = self.integer_type(revive_common::BIT_LENGTH_ETH_ADDRESS);
let address_pointer = self.build_alloca_at_entry(address_type, "address_pointer");
let address_pointer = self
.get_global(crate::polkavm::GLOBAL_ADDRESS_SPILL_BUFFER)?
.into();
let address_truncated =
self.builder()
.build_int_truncate(address, address_type, "address_truncated")?;
@@ -759,69 +773,49 @@ where
.into())
}
/// Builds a stack load instruction.
/// Builds a stack load instruction with a direct assignment.
/// Sets the alignment to 256 bits for the stack and 1 bit for the heap, parent, and child.
pub fn build_load(
pub fn build_load_assign(
&self,
pointer: Pointer<'ctx>,
name: &str,
) -> anyhow::Result<inkwell::values::BasicValueEnum<'ctx>> {
assignment_pointer: &mut Option<inkwell::values::PointerValue<'ctx>>,
) -> anyhow::Result<Option<inkwell::values::BasicValueEnum<'ctx>>> {
match pointer.address_space {
AddressSpace::Heap => {
let heap_pointer = self.build_heap_gep(
self.builder().build_ptr_to_int(
pointer.value,
self.xlen_type(),
"offset_ptrtoint",
)?,
pointer
.r#type
.size_of()
.expect("should be IntValue")
.const_truncate(self.xlen_type()),
)?;
let value = self
.builder()
.build_load(pointer.r#type, heap_pointer.value, name)?;
self.basic_block()
.get_last_instruction()
.expect("Always exists")
.set_alignment(revive_common::BYTE_LENGTH_BYTE as u32)
.expect("Alignment is valid");
self.build_byte_swap(value)
}
AddressSpace::Storage | AddressSpace::TransientStorage => {
let storage_value_pointer =
self.build_alloca(self.word_type(), "storage_value_pointer");
self.build_store(storage_value_pointer, self.word_const(0))?;
let storage_value_length_pointer =
self.build_alloca(self.xlen_type(), "storage_value_length_pointer");
self.build_store(
storage_value_length_pointer,
self.word_const(revive_common::BIT_LENGTH_WORD as u64),
)?;
let transient = pointer.address_space == AddressSpace::TransientStorage;
self.build_runtime_call(
revive_runtime_api::polkavm_imports::GET_STORAGE,
&[
self.xlen_type().const_int(transient as u64, false).into(),
pointer.to_int(self).into(),
self.xlen_type().const_all_ones().into(),
storage_value_pointer.to_int(self).into(),
storage_value_length_pointer.to_int(self).into(),
],
);
// We do not to check the return value.
// Solidity assumes infallible SLOAD.
// If a key doesn't exist the "zero" value is returned.
self.build_load(storage_value_pointer, "storage_value_load")
let declaration =
<PolkaVMLoadHeapWordFunction as RuntimeFunction<D>>::declaration(self);
match assignment_pointer.take() {
Some(assignment_pointer) => {
let arguments = [
self.builder()
.build_ptr_to_int(
pointer.value,
self.xlen_type(),
"offset_ptrtoint",
)?
.as_basic_value_enum(),
assignment_pointer.into(),
];
self.build_call(declaration, &arguments, "heap_load");
Ok(None)
}
None => {
let pointer = self.build_alloca_at_entry(self.word_type(), "pointer");
let arguments = [
self.builder()
.build_ptr_to_int(
pointer.value,
self.xlen_type(),
"offset_ptrtoint",
)?
.as_basic_value_enum(),
pointer.value.into(),
];
self.build_call(declaration, &arguments, "heap_load");
Ok(Some(self.build_load(pointer, "storage_value")?))
}
}
}
AddressSpace::Stack => {
let value = self
@@ -834,11 +828,23 @@ where
.set_alignment(revive_common::BYTE_LENGTH_STACK_ALIGN as u32)
.expect("Alignment is valid");
Ok(value)
Ok(Some(value))
}
}
}
/// Builds a stack load instruction.
/// Sets the alignment to 256 bits for the stack and 1 bit for the heap, parent, and child.
pub fn build_load(
&self,
pointer: Pointer<'ctx>,
name: &str,
) -> anyhow::Result<inkwell::values::BasicValueEnum<'ctx>> {
Ok(self
.build_load_assign(pointer, name, &mut None)?
.expect("without an assignment pointer loads return a value"))
}
/// Builds a stack store instruction.
/// Sets the alignment to 256 bits for the stack and 1 bit for the heap, parent, and child.
pub fn build_store<V>(&self, pointer: Pointer<'ctx>, value: V) -> anyhow::Result<()>
@@ -847,60 +853,13 @@ where
{
match pointer.address_space {
AddressSpace::Heap => {
let heap_pointer = self.build_heap_gep(
self.builder().build_ptr_to_int(
pointer.value,
self.xlen_type(),
"offset_ptrtoint",
)?,
value
.as_basic_value_enum()
.get_type()
.size_of()
.expect("should be IntValue")
.const_truncate(self.xlen_type()),
)?;
let value = value.as_basic_value_enum();
let value = match value.get_type().into_int_type().get_bit_width() as usize {
revive_common::BIT_LENGTH_WORD => self.build_byte_swap(value)?,
revive_common::BIT_LENGTH_BYTE => value,
_ => unreachable!("Only word and byte sized values can be stored on EVM heap"),
};
self.builder
.build_store(heap_pointer.value, value)?
.set_alignment(revive_common::BYTE_LENGTH_BYTE as u32)
.expect("Alignment is valid");
}
AddressSpace::Storage | AddressSpace::TransientStorage => {
assert_eq!(
value.as_basic_value_enum().get_type(),
self.word_type().as_basic_type_enum()
);
let storage_value_pointer = self.build_alloca(self.word_type(), "storage_value");
let storage_value_pointer_casted = self.builder().build_ptr_to_int(
storage_value_pointer.value,
self.xlen_type(),
"storage_value_pointer_casted",
)?;
self.builder()
.build_store(storage_value_pointer.value, value)?;
let transient = pointer.address_space == AddressSpace::TransientStorage;
self.build_runtime_call(
revive_runtime_api::polkavm_imports::SET_STORAGE,
&[
self.xlen_type().const_int(transient as u64, false).into(),
pointer.to_int(self).into(),
self.xlen_type().const_all_ones().into(),
storage_value_pointer_casted.into(),
self.integer_const(crate::polkavm::XLEN, 32).into(),
],
);
let declaration =
<PolkaVMStoreHeapWordFunction as RuntimeFunction<D>>::declaration(self);
let arguments = [
pointer.to_int(self).as_basic_value_enum(),
value.as_basic_value_enum(),
];
self.build_call(declaration, &arguments, "heap_store");
}
AddressSpace::Stack => {
let instruction = self.builder.build_store(pointer.value, value).unwrap();
@@ -946,9 +905,6 @@ where
where
T: BasicType<'ctx>,
{
assert_ne!(pointer.address_space, AddressSpace::Storage);
assert_ne!(pointer.address_space, AddressSpace::TransientStorage);
let value = unsafe {
self.builder
.build_gep(pointer.r#type, pointer.value, indexes, name)
@@ -1115,38 +1071,16 @@ where
offset: inkwell::values::IntValue<'ctx>,
length: inkwell::values::IntValue<'ctx>,
) -> anyhow::Result<()> {
let offset_truncated = self.safe_truncate_int_to_xlen(offset)?;
let length_truncated = self.safe_truncate_int_to_xlen(length)?;
let offset_into_heap = self.build_heap_gep(offset_truncated, length_truncated)?;
let length_pointer = self.safe_truncate_int_to_xlen(length)?;
let offset_pointer = self.builder().build_ptr_to_int(
offset_into_heap.value,
self.xlen_type(),
"return_data_ptr_to_int",
)?;
self.build_runtime_call(
revive_runtime_api::polkavm_imports::RETURN,
&[flags.into(), offset_pointer.into(), length_pointer.into()],
self.build_call(
<Exit as RuntimeFunction<D>>::declaration(self),
&[flags.into(), offset.into(), length.into()],
"exit",
);
self.build_unreachable();
Ok(())
}
/// Truncate a memory offset to register size, trapping if it doesn't fit.
/// Pointers are represented as opaque 256 bit integer values in EVM.
/// In practice, they should never exceed a register sized bit value.
/// However, we still protect against this possibility here. Heap index
/// offsets are generally untrusted and potentially represent valid
/// (but wrong) pointers when truncated.
///
/// TODO: Splitting up into a dedicated function
/// could potentially decrease code sizes (LLVM can still decide to inline).
/// However, passing i256 parameters is counter productive and
/// I've found that splitting it up actualy increases code size.
/// Should be reviewed after 64bit support.
pub fn safe_truncate_int_to_xlen(
&self,
value: inkwell::values::IntValue<'ctx>,
@@ -1160,29 +1094,19 @@ where
"expected XLEN or WORD sized int type for memory offset",
);
let truncated =
self.builder()
.build_int_truncate(value, self.xlen_type(), "offset_truncated")?;
let extended =
self.builder()
.build_int_z_extend(truncated, self.word_type(), "offset_extended")?;
let is_overflow = self.builder().build_int_compare(
inkwell::IntPredicate::NE,
value,
extended,
"compare_truncated_extended",
)?;
let block_continue = self.append_basic_block("offset_pointer_ok");
let block_trap = self.append_basic_block("offset_pointer_overflow");
self.build_conditional_branch(is_overflow, block_trap, block_continue)?;
self.set_basic_block(block_trap);
self.build_call(self.intrinsics().trap, &[], "invalid_trap");
self.build_unreachable();
self.set_basic_block(block_continue);
Ok(truncated)
Ok(self
.build_call(
<WordToPointer as RuntimeFunction<D>>::declaration(self),
&[value.into()],
"word_to_pointer",
)
.unwrap_or_else(|| {
panic!(
"revive runtime function {} should return a value",
<WordToPointer as RuntimeFunction<D>>::NAME,
)
})
.into_int_value())
}
/// Build a call to PolkaVM `sbrk` for extending the heap from offset by `size`.
@@ -1222,34 +1146,6 @@ where
Ok(memory_size_value.into_int_value())
}
/// Call PolkaVM `sbrk` for extending the heap by `offset` + `size`,
/// trapping the contract if the call failed.
pub fn build_heap_alloc(
&self,
offset: inkwell::values::IntValue<'ctx>,
size: inkwell::values::IntValue<'ctx>,
) -> anyhow::Result<()> {
let end_of_memory = self.build_sbrk(offset, size)?;
let return_is_nil = self.builder().build_int_compare(
inkwell::IntPredicate::EQ,
end_of_memory,
self.llvm().ptr_type(Default::default()).const_null(),
"compare_end_of_memory_nil",
)?;
let continue_block = self.append_basic_block("sbrk_not_nil");
let trap_block = self.append_basic_block("sbrk_nil");
self.build_conditional_branch(return_is_nil, trap_block, continue_block)?;
self.set_basic_block(trap_block);
self.build_call(self.intrinsics().trap, &[], "invalid_trap");
self.build_unreachable();
self.set_basic_block(continue_block);
Ok(())
}
/// Returns a pointer to `offset` into the heap, allocating
/// enough memory if `offset + length` would be out of bounds.
/// # Panics
@@ -1262,19 +1158,8 @@ where
assert_eq!(offset.get_type(), self.xlen_type());
assert_eq!(length.get_type(), self.xlen_type());
self.build_heap_alloc(offset, length)?;
let heap_start = self
.module()
.get_global(revive_runtime_api::polkavm_imports::MEMORY)
.expect("the memory symbol should have been declared")
.as_pointer_value();
Ok(self.build_gep(
Pointer::new(self.byte_type(), AddressSpace::Stack, heap_start),
&[offset],
self.byte_type(),
"heap_offset_via_gep",
))
let pointer = self.build_sbrk(offset, length)?;
Ok(Pointer::new(self.byte_type(), AddressSpace::Stack, pointer))
}
/// Returns a boolean type constant.
@@ -1445,13 +1330,6 @@ where
inkwell::attributes::AttributeLoc::Param(index as u32),
self.llvm.create_enum_attribute(Attribute::NoFree as u32, 0),
);
if function == self.llvm_runtime().sha3 {
call_site_value.add_attribute(
inkwell::attributes::AttributeLoc::Param(index as u32),
self.llvm
.create_enum_attribute(Attribute::ReadOnly as u32, 0),
);
}
if Some(argument.get_type()) == function.r#type.get_return_type() {
if function
.r#type
@@ -1580,4 +1458,8 @@ where
anyhow::bail!("The immutable size data is not available");
}
}
pub fn optimizer_settings(&self) -> &OptimizerSettings {
self.optimizer.settings()
}
}
@@ -0,0 +1,121 @@
//! The revive simulated EVM linear memory pointer functions.
use inkwell::values::BasicValueEnum;
use crate::polkavm::context::address_space::AddressSpace;
use crate::polkavm::context::pointer::Pointer;
use crate::polkavm::context::runtime::RuntimeFunction;
use crate::polkavm::context::Context;
use crate::polkavm::Dependency;
use crate::polkavm::WriteLLVM;
/// Load a word size value from a heap pointer.
pub struct LoadWord;
impl<D> RuntimeFunction<D> for LoadWord
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_load_heap_word";
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context.void_type().fn_type(
&[
context.xlen_type().into(),
context.llvm().ptr_type(Default::default()).into(),
],
false,
)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<BasicValueEnum<'ctx>>> {
let offset = Self::paramater(context, 0).into_int_value();
let assignment_pointer = Self::paramater(context, 1).into_pointer_value();
let length = context
.xlen_type()
.const_int(revive_common::BYTE_LENGTH_WORD as u64, false);
let pointer = context.build_heap_gep(offset, length)?;
let value = context
.builder()
.build_load(context.word_type(), pointer.value, "value")?;
context
.basic_block()
.get_last_instruction()
.expect("Always exists")
.set_alignment(revive_common::BYTE_LENGTH_BYTE as u32)
.expect("Alignment is valid");
let swapped_value = context.build_byte_swap(value)?;
context.build_store(
Pointer::new(context.word_type(), AddressSpace::Stack, assignment_pointer),
swapped_value,
)?;
Ok(None)
}
}
impl<D> WriteLLVM<D> for LoadWord
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
/// Store a word size value through a heap pointer.
pub struct StoreWord;
impl<D> RuntimeFunction<D> for StoreWord
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_store_heap_word";
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context.void_type().fn_type(
&[context.xlen_type().into(), context.word_type().into()],
false,
)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<BasicValueEnum<'ctx>>> {
let offset = Self::paramater(context, 0).into_int_value();
let length = context
.xlen_type()
.const_int(revive_common::BYTE_LENGTH_WORD as u64, false);
let pointer = context.build_heap_gep(offset, length)?;
let value = context.build_byte_swap(Self::paramater(context, 1))?;
context
.builder()
.build_store(pointer.value, value)?
.set_alignment(revive_common::BYTE_LENGTH_BYTE as u32)
.expect("Alignment is valid");
Ok(None)
}
}
impl<D> WriteLLVM<D> for StoreWord
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
@@ -7,6 +7,9 @@ use crate::polkavm::context::global::Global;
use crate::polkavm::context::Context;
use crate::polkavm::Dependency;
pub mod heap;
pub mod storage;
/// The LLVM pointer.
#[derive(Debug, Clone, Copy)]
pub struct Pointer<'ctx> {
@@ -0,0 +1,286 @@
//! The revive storage pointer functions.
use inkwell::values::BasicValueEnum;
use crate::polkavm::context::address_space::AddressSpace;
use crate::polkavm::context::pointer::Pointer;
use crate::polkavm::context::runtime::RuntimeFunction;
use crate::polkavm::context::Context;
use crate::polkavm::Dependency;
use crate::polkavm::WriteLLVM;
/// Load a word size value from a storage pointer.
pub struct LoadWord;
impl<D> RuntimeFunction<D> for LoadWord
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_load_storage_word";
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context.void_type().fn_type(
&[
context.llvm().ptr_type(Default::default()).into(),
context.llvm().ptr_type(Default::default()).into(),
],
false,
)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<BasicValueEnum<'ctx>>> {
let key = Self::paramater(context, 0);
let assignment_pointer = Self::paramater(context, 1).into_pointer_value();
let value = emit_load(context, key, false)?;
context.build_store(
Pointer::new(context.word_type(), AddressSpace::Stack, assignment_pointer),
value,
)?;
Ok(None)
}
}
impl<D> WriteLLVM<D> for LoadWord
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
/// Load a word size value from a transient storage pointer.
pub struct LoadTransientWord;
impl<D> RuntimeFunction<D> for LoadTransientWord
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_load_transient_storage_word";
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context
.word_type()
.fn_type(&[context.llvm().ptr_type(Default::default()).into()], false)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<BasicValueEnum<'ctx>>> {
Ok(Some(emit_load(context, Self::paramater(context, 0), true)?))
}
}
impl<D> WriteLLVM<D> for LoadTransientWord
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
/// Store a word size value through a storage pointer.
pub struct StoreWord;
impl<D> RuntimeFunction<D> for StoreWord
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_store_storage_word";
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context.void_type().fn_type(
&[
context.llvm().ptr_type(Default::default()).into(),
context.llvm().ptr_type(Default::default()).into(),
],
false,
)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<BasicValueEnum<'ctx>>> {
emit_store(
context,
Self::paramater(context, 0),
Self::paramater(context, 1),
false,
)?;
Ok(None)
}
}
impl<D> WriteLLVM<D> for StoreWord
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
/// Store a word size value through a transient storage pointer.
pub struct StoreTransientWord;
impl<D> RuntimeFunction<D> for StoreTransientWord
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_store_transient_storage_word";
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context.void_type().fn_type(
&[
context.llvm().ptr_type(Default::default()).into(),
context.llvm().ptr_type(Default::default()).into(),
],
false,
)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<BasicValueEnum<'ctx>>> {
emit_store(
context,
Self::paramater(context, 0),
Self::paramater(context, 1),
true,
)?;
Ok(None)
}
}
impl<D> WriteLLVM<D> for StoreTransientWord
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
fn emit_load<'ctx, D: Dependency + Clone>(
context: &mut Context<'ctx, D>,
key: BasicValueEnum<'ctx>,
transient: bool,
) -> anyhow::Result<BasicValueEnum<'ctx>> {
let mut key = context.build_load(
super::Pointer::new(
context.word_type(),
Default::default(),
key.into_pointer_value(),
),
"key",
)?;
if !transient {
key = context.build_byte_swap(key)?;
}
let key_pointer = context.build_alloca_at_entry(context.word_type(), "key_pointer");
let value_pointer = context.build_alloca_at_entry(context.word_type(), "value_pointer");
let length_pointer = context.build_alloca_at_entry(context.xlen_type(), "length_pointer");
context.builder().build_store(key_pointer.value, key)?;
context.build_store(value_pointer, context.word_const(0))?;
context.build_store(
length_pointer,
context
.xlen_type()
.const_int(revive_common::BYTE_LENGTH_WORD as u64, false),
)?;
let is_transient = context.xlen_type().const_int(transient as u64, false);
let arguments = [
is_transient.into(),
key_pointer.to_int(context).into(),
context.xlen_type().const_all_ones().into(),
value_pointer.to_int(context).into(),
length_pointer.to_int(context).into(),
];
context.build_runtime_call(revive_runtime_api::polkavm_imports::GET_STORAGE, &arguments);
// We do not to check the return value: Solidity assumes infallible loads.
// If a key doesn't exist the "zero" value is returned (ensured by above write).
let value = context.build_load(value_pointer, "storage_value")?;
Ok(if transient {
value
} else {
context.build_byte_swap(value)?
})
}
fn emit_store<'ctx, D: Dependency + Clone>(
context: &mut Context<'ctx, D>,
key: BasicValueEnum<'ctx>,
value: BasicValueEnum<'ctx>,
transient: bool,
) -> anyhow::Result<()> {
let mut key = context.build_load(
super::Pointer::new(
context.word_type(),
Default::default(),
key.into_pointer_value(),
),
"key",
)?;
let mut value = context.build_load(
super::Pointer::new(
context.word_type(),
Default::default(),
value.into_pointer_value(),
),
"key",
)?;
if !transient {
key = context.build_byte_swap(key)?;
value = context.build_byte_swap(value)?;
}
let key_pointer = context.build_alloca_at_entry(context.word_type(), "key_pointer");
let value_pointer = context.build_alloca_at_entry(context.word_type(), "value_pointer");
context.build_store(key_pointer, key)?;
context.build_store(value_pointer, value)?;
let is_transient = context.xlen_type().const_int(transient as u64, false);
let arguments = [
is_transient.into(),
key_pointer.to_int(context).into(),
context.xlen_type().const_all_ones().into(),
value_pointer.to_int(context).into(),
context.integer_const(crate::polkavm::XLEN, 32).into(),
];
context.build_runtime_call(revive_runtime_api::polkavm_imports::SET_STORAGE, &arguments);
Ok(())
}
@@ -0,0 +1,113 @@
//! The revive compiler runtime function interface definition.
//!
//! Common routines should not be inlined but extracted into smaller functions.
//! This benefits contract code size.
use crate::optimizer::settings::size_level::SizeLevel;
use crate::polkavm::context::function::declaration::Declaration;
use crate::polkavm::context::function::Function;
use crate::polkavm::context::Attribute;
use crate::polkavm::context::Context;
use crate::polkavm::Dependency;
/// The revive runtime function interface simplifies declaring runtime functions
/// and code emitting by providing helpful default implementations.
pub trait RuntimeFunction<D>
where
D: Dependency + Clone,
{
/// The function name.
const NAME: &'static str;
const ATTRIBUTES: &'static [Attribute] = &[
Attribute::NoFree,
Attribute::NoRecurse,
Attribute::WillReturn,
];
/// The function type.
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx>;
/// Declare the function.
fn declare(&self, context: &mut Context<D>) -> anyhow::Result<()> {
let function = context.add_function(
Self::NAME,
Self::r#type(context),
0,
Some(inkwell::module::Linkage::External),
)?;
let mut attributes = Self::ATTRIBUTES.to_vec();
attributes.extend_from_slice(match context.optimizer_settings().level_middle_end_size {
SizeLevel::Zero => &[],
_ => &[Attribute::OptimizeForSize, Attribute::MinSize],
});
Function::set_attributes(
context.llvm(),
function.borrow().declaration(),
&attributes,
true,
);
Ok(())
}
/// Get the function declaration.
fn declaration<'ctx>(context: &Context<'ctx, D>) -> Declaration<'ctx> {
context
.get_function(Self::NAME)
.unwrap_or_else(|| panic!("runtime function {} should be declared", Self::NAME))
.borrow()
.declaration()
}
/// Emit the function.
fn emit(&self, context: &mut Context<D>) -> anyhow::Result<()> {
context.set_current_function(Self::NAME, None)?;
context.set_basic_block(context.current_function().borrow().entry_block());
let return_value = self.emit_body(context)?;
self.emit_epilogue(context, return_value);
context.pop_debug_scope();
Ok(())
}
/// Emit the function body.
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<inkwell::values::BasicValueEnum<'ctx>>>;
/// Emit the function return instructions.
fn emit_epilogue<'ctx>(
&self,
context: &mut Context<'ctx, D>,
return_value: Option<inkwell::values::BasicValueEnum<'ctx>>,
) {
let return_block = context.current_function().borrow().return_block();
context.build_unconditional_branch(return_block);
context.set_basic_block(return_block);
match return_value {
Some(value) => context.build_return(Some(&value)),
None => context.build_return(None),
}
}
/// Get the nth function paramater.
fn paramater<'ctx>(
context: &Context<'ctx, D>,
index: usize,
) -> inkwell::values::BasicValueEnum<'ctx> {
let name = Self::NAME;
context
.get_function(name)
.unwrap_or_else(|| panic!("runtime function {name} should be declared"))
.borrow()
.declaration()
.function_value()
.get_nth_param(index as u32)
.unwrap_or_else(|| panic!("runtime function {name} should have parameter #{index}"))
}
}
@@ -10,12 +10,20 @@ pub fn create_context(
llvm: &inkwell::context::Context,
optimizer_settings: OptimizerSettings,
) -> Context<DummyDependency> {
crate::polkavm::initialize_target();
crate::initialize_llvm(crate::Target::PVM, "resolc", Default::default());
let module = llvm.create_module("test");
let optimizer = Optimizer::new(optimizer_settings);
Context::<DummyDependency>::new(llvm, module, optimizer, None, true, Default::default())
Context::<DummyDependency>::new(
llvm,
module,
optimizer,
None,
true,
Default::default(),
Default::default(),
)
}
#[test]
+25 -107
View File
@@ -2,8 +2,13 @@
use inkwell::values::BasicValue;
use crate::polkavm::context::runtime::RuntimeFunction;
use crate::polkavm::context::Context;
use crate::polkavm::Dependency;
use crate::PolkaVMDivisionFunction;
use crate::PolkaVMRemainderFunction;
use crate::PolkaVMSignedDivisionFunction;
use crate::PolkaVMSignedRemainderFunction;
/// Translates the arithmetic addition.
pub fn addition<'ctx, D>(
@@ -59,11 +64,11 @@ pub fn division<'ctx, D>(
where
D: Dependency + Clone,
{
wrapped_division(context, operand_2, || {
Ok(context
.builder()
.build_int_unsigned_div(operand_1, operand_2, "DIV")?)
})
let name = <PolkaVMDivisionFunction as RuntimeFunction<D>>::NAME;
let declaration = <PolkaVMDivisionFunction as RuntimeFunction<D>>::declaration(context);
Ok(context
.build_call(declaration, &[operand_1.into(), operand_2.into()], "div")
.unwrap_or_else(|| panic!("revive runtime function {name} should return a value",)))
}
/// Translates the arithmetic remainder.
@@ -75,11 +80,11 @@ pub fn remainder<'ctx, D>(
where
D: Dependency + Clone,
{
wrapped_division(context, operand_2, || {
Ok(context
.builder()
.build_int_unsigned_rem(operand_1, operand_2, "MOD")?)
})
let name = <PolkaVMRemainderFunction as RuntimeFunction<D>>::NAME;
let declaration = <PolkaVMRemainderFunction as RuntimeFunction<D>>::declaration(context);
Ok(context
.build_call(declaration, &[operand_1.into(), operand_2.into()], "rem")
.unwrap_or_else(|| panic!("revive runtime function {name} should return a value",)))
}
/// Translates the signed arithmetic division.
@@ -94,54 +99,11 @@ pub fn division_signed<'ctx, D>(
where
D: Dependency + Clone,
{
assert_eq!(
operand_2.get_type().get_bit_width(),
revive_common::BIT_LENGTH_WORD as u32
);
let block_calculate = context.append_basic_block("calculate");
let block_overflow = context.append_basic_block("overflow");
let block_select = context.append_basic_block("select_result");
let block_origin = context.basic_block();
context.builder().build_switch(
operand_2,
block_calculate,
&[
(context.word_type().const_zero(), block_select),
(context.word_type().const_all_ones(), block_overflow),
],
)?;
context.set_basic_block(block_calculate);
let quotient = context
.builder()
.build_int_signed_div(operand_1, operand_2, "SDIV")?;
context.build_unconditional_branch(block_select);
context.set_basic_block(block_overflow);
let max_uint = context.builder().build_int_z_extend(
context
.integer_type(revive_common::BIT_LENGTH_WORD - 1)
.const_all_ones(),
context.word_type(),
"max_uint",
)?;
let is_operand_1_overflow = context.builder().build_int_compare(
inkwell::IntPredicate::EQ,
operand_1,
context.builder().build_int_neg(max_uint, "min_uint")?,
"is_operand_1_overflow",
)?;
context.build_conditional_branch(is_operand_1_overflow, block_select, block_calculate)?;
context.set_basic_block(block_select);
let result = context.builder().build_phi(context.word_type(), "result")?;
result.add_incoming(&[
(&operand_1, block_overflow),
(&context.word_const(0), block_origin),
(&quotient.as_basic_value_enum(), block_calculate),
]);
Ok(result.as_basic_value())
let name = <PolkaVMSignedDivisionFunction as RuntimeFunction<D>>::NAME;
let declaration = <PolkaVMSignedDivisionFunction as RuntimeFunction<D>>::declaration(context);
Ok(context
.build_call(declaration, &[operand_1.into(), operand_2.into()], "sdiv")
.unwrap_or_else(|| panic!("revive runtime function {name} should return a value",)))
}
/// Translates the signed arithmetic remainder.
@@ -153,53 +115,9 @@ pub fn remainder_signed<'ctx, D>(
where
D: Dependency + Clone,
{
wrapped_division(context, operand_2, || {
Ok(context
.builder()
.build_int_signed_rem(operand_1, operand_2, "SMOD")?)
})
}
/// Wrap division operations so that zero will be returned if the
/// denominator is zero (see also Ethereum YP Appendix H.2).
///
/// The closure is expected to calculate and return the quotient.
///
/// The result is either the calculated quotient or zero,
/// selected at runtime.
fn wrapped_division<'ctx, D, F, T>(
context: &Context<'ctx, D>,
denominator: inkwell::values::IntValue<'ctx>,
f: F,
) -> anyhow::Result<inkwell::values::BasicValueEnum<'ctx>>
where
D: Dependency + Clone,
F: FnOnce() -> anyhow::Result<T>,
T: inkwell::values::IntMathValue<'ctx>,
{
assert_eq!(
denominator.get_type().get_bit_width(),
revive_common::BIT_LENGTH_WORD as u32
);
let block_calculate = context.append_basic_block("calculate");
let block_select = context.append_basic_block("select");
let block_origin = context.basic_block();
context.builder().build_switch(
denominator,
block_calculate,
&[(context.word_const(0), block_select)],
)?;
context.set_basic_block(block_calculate);
let calculated_value = f()?.as_basic_value_enum();
context.build_unconditional_branch(block_select);
context.set_basic_block(block_select);
let result = context.builder().build_phi(context.word_type(), "result")?;
result.add_incoming(&[
(&context.word_const(0), block_origin),
(&calculated_value, block_calculate),
]);
Ok(result.as_basic_value())
let name = <PolkaVMSignedRemainderFunction as RuntimeFunction<D>>::NAME;
let declaration = <PolkaVMSignedRemainderFunction as RuntimeFunction<D>>::declaration(context);
Ok(context
.build_call(declaration, &[operand_1.into(), operand_2.into()], "srem")
.unwrap_or_else(|| panic!("revive runtime function {name} should return a value",)))
}
+28 -14
View File
@@ -63,7 +63,6 @@ where
let non_overflow_block = context.append_basic_block("shift_left_non_overflow");
let join_block = context.append_basic_block("shift_left_join");
let result_pointer = context.build_alloca(context.word_type(), "shift_left_result_pointer");
let condition_is_overflow = context.builder().build_int_compare(
inkwell::IntPredicate::UGT,
shift,
@@ -73,7 +72,6 @@ where
context.build_conditional_branch(condition_is_overflow, overflow_block, non_overflow_block)?;
context.set_basic_block(overflow_block);
context.build_store(result_pointer, context.word_const(0))?;
context.build_unconditional_branch(join_block);
context.set_basic_block(non_overflow_block);
@@ -81,11 +79,17 @@ where
context
.builder()
.build_left_shift(value, shift, "shift_left_non_overflow_result")?;
context.build_store(result_pointer, value)?;
context.build_unconditional_branch(join_block);
context.set_basic_block(join_block);
context.build_load(result_pointer, "shift_left_result")
let result = context
.builder()
.build_phi(context.word_type(), "shift_left_value")?;
result.add_incoming(&[
(&value, non_overflow_block),
(&context.word_const(0), overflow_block),
]);
Ok(result.as_basic_value())
}
/// Translates the bitwise shift right.
@@ -101,7 +105,6 @@ where
let non_overflow_block = context.append_basic_block("shift_right_non_overflow");
let join_block = context.append_basic_block("shift_right_join");
let result_pointer = context.build_alloca(context.word_type(), "shift_right_result_pointer");
let condition_is_overflow = context.builder().build_int_compare(
inkwell::IntPredicate::UGT,
shift,
@@ -111,7 +114,6 @@ where
context.build_conditional_branch(condition_is_overflow, overflow_block, non_overflow_block)?;
context.set_basic_block(overflow_block);
context.build_store(result_pointer, context.word_const(0))?;
context.build_unconditional_branch(join_block);
context.set_basic_block(non_overflow_block);
@@ -121,11 +123,17 @@ where
false,
"shift_right_non_overflow_result",
)?;
context.build_store(result_pointer, value)?;
context.build_unconditional_branch(join_block);
context.set_basic_block(join_block);
context.build_load(result_pointer, "shift_right_result")
let result = context
.builder()
.build_phi(context.word_type(), "shift_right_value")?;
result.add_incoming(&[
(&value, non_overflow_block),
(&context.word_const(0), overflow_block),
]);
Ok(result.as_basic_value())
}
/// Translates the arithmetic bitwise shift right.
@@ -145,8 +153,6 @@ where
let non_overflow_block = context.append_basic_block("shift_right_arithmetic_non_overflow");
let join_block = context.append_basic_block("shift_right_arithmetic_join");
let result_pointer =
context.build_alloca(context.word_type(), "shift_right_arithmetic_result_pointer");
let condition_is_overflow = context.builder().build_int_compare(
inkwell::IntPredicate::UGT,
shift,
@@ -174,11 +180,9 @@ where
)?;
context.set_basic_block(overflow_positive_block);
context.build_store(result_pointer, context.word_const(0))?;
context.build_unconditional_branch(join_block);
context.set_basic_block(overflow_negative_block);
context.build_store(result_pointer, context.word_type().const_all_ones())?;
context.build_unconditional_branch(join_block);
context.set_basic_block(non_overflow_block);
@@ -188,11 +192,21 @@ where
true,
"shift_right_arithmetic_non_overflow_result",
)?;
context.build_store(result_pointer, value)?;
context.build_unconditional_branch(join_block);
context.set_basic_block(join_block);
context.build_load(result_pointer, "shift_right_arithmetic_result")
let result = context
.builder()
.build_phi(context.word_type(), "shift_arithmetic_right_value")?;
result.add_incoming(&[
(&value, non_overflow_block),
(
&context.word_type().const_all_ones(),
overflow_negative_block,
),
(&context.word_const(0), overflow_block),
]);
Ok(result.as_basic_value())
}
/// Translates the `byte` instruction, extracting the byte of `operand_2`
+16 -16
View File
@@ -2,6 +2,7 @@
use inkwell::values::BasicValue;
use crate::polkavm::context::pointer::Pointer;
use crate::polkavm::context::Context;
use crate::polkavm::Dependency;
@@ -49,7 +50,9 @@ where
D: Dependency + Clone,
{
let address_type = context.integer_type(revive_common::BIT_LENGTH_ETH_ADDRESS);
let address_pointer = context.build_alloca_at_entry(address_type, "origin_address");
let address_pointer: Pointer<'_> = context
.get_global(crate::polkavm::GLOBAL_ADDRESS_SPILL_BUFFER)?
.into();
context.build_store(address_pointer, address_type.const_zero())?;
context.build_runtime_call(
revive_runtime_api::polkavm_imports::ORIGIN,
@@ -97,13 +100,13 @@ where
D: Dependency + Clone,
{
let output_pointer = context.build_alloca_at_entry(context.word_type(), "blockhash_out_ptr");
let index_ptr = context.build_alloca_at_entry(context.word_type(), "blockhash_index_ptr");
context.build_store(index_ptr, index)?;
let index_pointer = context.build_alloca_at_entry(context.word_type(), "blockhash_index_ptr");
context.build_store(index_pointer, index)?;
context.build_runtime_call(
revive_runtime_api::polkavm_imports::BLOCK_HASH,
&[
index_ptr.to_int(context).into(),
index_pointer.to_int(context).into(),
output_pointer.to_int(context).into(),
],
);
@@ -127,10 +130,9 @@ pub fn coinbase<'ctx, D>(
where
D: Dependency + Clone,
{
let pointer = context.build_alloca_at_entry(
context.integer_type(revive_common::BIT_LENGTH_ETH_ADDRESS),
"coinbase_output",
);
let pointer: Pointer<'_> = context
.get_global(crate::polkavm::GLOBAL_ADDRESS_SPILL_BUFFER)?
.into();
context.build_runtime_call(
revive_runtime_api::polkavm_imports::BLOCK_AUTHOR,
&[pointer.to_int(context).into()],
@@ -155,10 +157,9 @@ pub fn address<'ctx, D>(
where
D: Dependency + Clone,
{
let pointer = context.build_alloca_at_entry(
context.integer_type(revive_common::BIT_LENGTH_ETH_ADDRESS),
"address_output",
);
let pointer: Pointer<'_> = context
.get_global(crate::polkavm::GLOBAL_ADDRESS_SPILL_BUFFER)?
.into();
context.build_runtime_call(
revive_runtime_api::polkavm_imports::ADDRESS,
&[pointer.to_int(context).into()],
@@ -173,10 +174,9 @@ pub fn caller<'ctx, D>(
where
D: Dependency + Clone,
{
let pointer = context.build_alloca_at_entry(
context.integer_type(revive_common::BIT_LENGTH_ETH_ADDRESS),
"address_output",
);
let pointer: Pointer<'_> = context
.get_global(crate::polkavm::GLOBAL_ADDRESS_SPILL_BUFFER)?
.into();
context.build_runtime_call(
revive_runtime_api::polkavm_imports::CALLER,
&[pointer.to_int(context).into()],
+7 -10
View File
@@ -32,6 +32,7 @@ where
let salt_pointer = match salt {
Some(salt) => {
let salt_pointer = context.build_alloca_at_entry(context.word_type(), "salt_pointer");
let salt = context.build_byte_swap(salt.into())?;
context.build_store(salt_pointer, salt)?;
salt_pointer
}
@@ -118,10 +119,8 @@ where
_ => error,
})?;
if contract_path.as_str() == parent {
return Ok(Argument::new_with_constant(
context.word_const(0).as_basic_value_enum(),
num::BigUint::zero(),
));
return Ok(Argument::value(context.word_const(0).as_basic_value_enum())
.with_constant(num::BigUint::zero()));
} else if identifier.ends_with("_deployed") && code_type == CodeType::Runtime {
anyhow::bail!("type({}).runtimeCode is not supported", identifier);
}
@@ -130,7 +129,7 @@ where
let hash_value = context
.word_const_str_hex(hash_string.as_str())
.as_basic_value_enum();
Ok(Argument::new_with_original(hash_value, hash_string))
Ok(Argument::value(hash_value).with_original(hash_string))
}
/// Translates the deploy call header size instruction. the header consists of
@@ -159,10 +158,8 @@ where
_ => error,
})?;
if contract_path.as_str() == parent {
return Ok(Argument::new_with_constant(
context.word_const(0).as_basic_value_enum(),
num::BigUint::zero(),
));
return Ok(Argument::value(context.word_const(0).as_basic_value_enum())
.with_constant(num::BigUint::zero()));
} else if identifier.ends_with("_deployed") && code_type == CodeType::Runtime {
anyhow::bail!("type({}).runtimeCode is not supported", identifier);
}
@@ -171,5 +168,5 @@ where
let size_value = context
.word_const(crate::polkavm::DEPLOYER_CALL_HEADER_SIZE as u64)
.as_basic_value_enum();
Ok(Argument::new_with_constant(size_value, size_bigint))
Ok(Argument::value(size_value).with_constant(size_bigint))
}
@@ -28,7 +28,7 @@ pub fn value<'ctx, D>(
where
D: Dependency + Clone,
{
let output_pointer = context.build_alloca(context.value_type(), "value_transferred");
let output_pointer = context.build_alloca_at_entry(context.value_type(), "value_transferred");
context.build_store(output_pointer, context.word_const(0))?;
context.build_runtime_call(
revive_runtime_api::polkavm_imports::VALUE_TRANSFERRED,
@@ -46,8 +46,7 @@ where
D: Dependency + Clone,
{
let address_pointer = context.build_address_argument_store(address)?;
let balance_pointer = context.build_alloca(context.word_type(), "balance_pointer");
let balance_pointer = context.build_alloca_at_entry(context.word_type(), "balance_pointer");
let balance = context.builder().build_ptr_to_int(
balance_pointer.value,
context.xlen_type(),
@@ -69,7 +68,7 @@ pub fn self_balance<'ctx, D>(
where
D: Dependency + Clone,
{
let balance_pointer = context.build_alloca(context.word_type(), "balance_pointer");
let balance_pointer = context.build_alloca_at_entry(context.word_type(), "balance_pointer");
let balance = context.builder().build_ptr_to_int(
balance_pointer.value,
context.xlen_type(),
+167 -68
View File
@@ -2,84 +2,183 @@
use inkwell::values::BasicValue;
use crate::polkavm::context::runtime::RuntimeFunction;
use crate::polkavm::context::Context;
use crate::polkavm::Dependency;
use crate::polkavm::WriteLLVM;
/// A function for emitting EVM event logs from contract code.
pub struct EventLog<const N: usize>;
impl<D, const N: usize> RuntimeFunction<D> for EventLog<N>
where
D: Dependency + Clone,
{
const NAME: &'static str = match N {
0 => "__revive_log_0",
1 => "__revive_log_1",
2 => "__revive_log_2",
3 => "__revive_log_3",
4 => "__revive_log_4",
_ => unreachable!(),
};
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
let mut parameter_types = vec![context.xlen_type().into(), context.xlen_type().into()];
parameter_types.extend_from_slice(&[context.word_type().into(); N]);
context.void_type().fn_type(&parameter_types, false)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<inkwell::values::BasicValueEnum<'ctx>>> {
let input_offset = Self::paramater(context, 0).into_int_value();
let input_length = Self::paramater(context, 1).into_int_value();
let input_pointer = context.builder().build_ptr_to_int(
context.build_heap_gep(input_offset, input_length)?.value,
context.xlen_type(),
"event_input_offset",
)?;
let arguments = if N == 0 {
[
context.xlen_type().const_zero().as_basic_value_enum(),
context.xlen_type().const_zero().as_basic_value_enum(),
input_pointer.as_basic_value_enum(),
input_length.as_basic_value_enum(),
]
} else {
let topics_buffer_size = N * revive_common::BYTE_LENGTH_WORD;
let topics_buffer_pointer = context.build_alloca_at_entry(
context.byte_type().array_type(topics_buffer_size as u32),
"topics_buffer",
);
for n in 0..N {
let topic = Self::paramater(context, n + 2);
let topic_buffer_offset = context
.xlen_type()
.const_int((n * revive_common::BYTE_LENGTH_WORD) as u64, false);
context.build_store(
context.build_gep(
topics_buffer_pointer,
&[context.xlen_type().const_zero(), topic_buffer_offset],
context.byte_type(),
&format!("topic_buffer_{N}_gep"),
),
context.build_byte_swap(topic.as_basic_value_enum())?,
)?;
}
[
context
.builder()
.build_ptr_to_int(
topics_buffer_pointer.value,
context.xlen_type(),
"event_topics_offset",
)?
.as_basic_value_enum(),
context
.xlen_type()
.const_int(N as u64, false)
.as_basic_value_enum(),
input_pointer.as_basic_value_enum(),
input_length.as_basic_value_enum(),
]
};
context.build_runtime_call(
revive_runtime_api::polkavm_imports::DEPOSIT_EVENT,
&arguments,
);
Ok(None)
}
}
impl<D> WriteLLVM<D> for EventLog<0>
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
impl<D> WriteLLVM<D> for EventLog<1>
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
impl<D> WriteLLVM<D> for EventLog<2>
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
impl<D> WriteLLVM<D> for EventLog<3>
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
impl<D> WriteLLVM<D> for EventLog<4>
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
/// Translates a log or event call.
///
/// TODO: Splitting up into dedicated functions (log0..log4)
/// could potentially decrease code sizes (LLVM can still decide to inline).
/// However, passing i256 parameters is counter productive and
/// I've found that splitting it up actualy increases code size.
/// Should be reviewed after 64bit support.
pub fn log<'ctx, D>(
pub fn log<'ctx, D, const N: usize>(
context: &mut Context<'ctx, D>,
input_offset: inkwell::values::IntValue<'ctx>,
input_length: inkwell::values::IntValue<'ctx>,
topics: Vec<inkwell::values::IntValue<'ctx>>,
topics: [inkwell::values::BasicValueEnum<'ctx>; N],
) -> anyhow::Result<()>
where
D: Dependency + Clone,
{
let input_offset = context.safe_truncate_int_to_xlen(input_offset)?;
let input_length = context.safe_truncate_int_to_xlen(input_length)?;
let input_pointer = context.builder().build_ptr_to_int(
context.build_heap_gep(input_offset, input_length)?.value,
context.xlen_type(),
"event_input_offset",
)?;
let arguments = if topics.is_empty() {
[
context.xlen_type().const_zero().as_basic_value_enum(),
context.xlen_type().const_zero().as_basic_value_enum(),
input_pointer.as_basic_value_enum(),
input_length.as_basic_value_enum(),
]
} else {
let topics_buffer_size = topics.len() * revive_common::BYTE_LENGTH_WORD;
let topics_buffer_pointer = context.build_alloca(
context.byte_type().array_type(topics_buffer_size as u32),
"topics_buffer",
);
for (n, topic) in topics.iter().enumerate() {
let topic_buffer_offset = context
.xlen_type()
.const_int((n * revive_common::BYTE_LENGTH_WORD) as u64, false);
context.build_store(
context.build_gep(
topics_buffer_pointer,
&[context.xlen_type().const_zero(), topic_buffer_offset],
context.byte_type(),
"topic_buffer_gep",
),
context.build_byte_swap(topic.as_basic_value_enum())?,
)?;
}
[
context
.builder()
.build_ptr_to_int(
topics_buffer_pointer.value,
context.xlen_type(),
"event_topics_offset",
)?
.as_basic_value_enum(),
context
.xlen_type()
.const_int(topics.len() as u64, false)
.as_basic_value_enum(),
input_pointer.as_basic_value_enum(),
input_length.as_basic_value_enum(),
]
};
let _ = context.build_runtime_call(
revive_runtime_api::polkavm_imports::DEPOSIT_EVENT,
&arguments,
);
let declaration = <EventLog<N> as RuntimeFunction<D>>::declaration(context);
let mut arguments = vec![
context.safe_truncate_int_to_xlen(input_offset)?.into(),
context.safe_truncate_int_to_xlen(input_length)?.into(),
];
arguments.extend_from_slice(&topics);
context.build_call(declaration, &arguments, "log");
Ok(())
}
@@ -4,10 +4,206 @@ use inkwell::types::BasicType;
use crate::polkavm::context::address_space::AddressSpace;
use crate::polkavm::context::code_type::CodeType;
use crate::polkavm::context::function::runtime;
use crate::polkavm::context::pointer::Pointer;
use crate::polkavm::context::runtime::RuntimeFunction;
use crate::polkavm::context::Context;
use crate::polkavm::Dependency;
use crate::polkavm::WriteLLVM;
/// A function for requesting the immutable data from the runtime.
/// This is a special function that is only used by the front-end generated code.
///
/// The runtime API is called lazily and subsequent calls are no-ops.
///
/// The bytes written is asserted to match the expected length.
/// This should never fail; the length is known.
/// However, this is a one time assertion, hence worth it.
pub struct Load;
impl<D> RuntimeFunction<D> for Load
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_load_immutable_data";
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context.void_type().fn_type(Default::default(), false)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<inkwell::values::BasicValueEnum<'ctx>>> {
let immutable_data_size_pointer = context
.get_global(revive_runtime_api::immutable_data::GLOBAL_IMMUTABLE_DATA_SIZE)?
.value
.as_pointer_value();
let immutable_data_size = context.build_load(
Pointer::new(
context.xlen_type(),
AddressSpace::Stack,
immutable_data_size_pointer,
),
"immutable_data_size_load",
)?;
let load_immutable_data_block = context.append_basic_block("load_immutables_block");
let return_block = context.current_function().borrow().return_block();
let immutable_data_size_is_zero = context.builder().build_int_compare(
inkwell::IntPredicate::EQ,
context.xlen_type().const_zero(),
immutable_data_size.into_int_value(),
"immutable_data_size_is_zero",
)?;
context.build_conditional_branch(
immutable_data_size_is_zero,
return_block,
load_immutable_data_block,
)?;
context.set_basic_block(load_immutable_data_block);
let output_pointer = context
.get_global(revive_runtime_api::immutable_data::GLOBAL_IMMUTABLE_DATA_POINTER)?
.value
.as_pointer_value();
context.build_runtime_call(
revive_runtime_api::polkavm_imports::GET_IMMUTABLE_DATA,
&[
context
.builder()
.build_ptr_to_int(output_pointer, context.xlen_type(), "ptr_to_xlen")?
.into(),
context
.builder()
.build_ptr_to_int(
immutable_data_size_pointer,
context.xlen_type(),
"ptr_to_xlen",
)?
.into(),
],
);
let bytes_written = context.builder().build_load(
context.xlen_type(),
immutable_data_size_pointer,
"bytes_written",
)?;
context.builder().build_store(
immutable_data_size_pointer,
context.xlen_type().const_zero(),
)?;
let overflow_block = context.append_basic_block("immutable_data_overflow");
let is_overflow = context.builder().build_int_compare(
inkwell::IntPredicate::UGT,
immutable_data_size.into_int_value(),
bytes_written.into_int_value(),
"is_overflow",
)?;
context.build_conditional_branch(is_overflow, overflow_block, return_block)?;
context.set_basic_block(overflow_block);
context.build_call(context.intrinsics().trap, &[], "invalid_trap");
context.build_unreachable();
Ok(None)
}
}
impl<D> WriteLLVM<D> for Load
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
/// Store the immutable data from the constructor code.
pub struct Store;
impl<D> RuntimeFunction<D> for Store
where
D: Dependency + Clone,
{
const NAME: &'static str = "__revive_store_immutable_data";
fn r#type<'ctx>(context: &Context<'ctx, D>) -> inkwell::types::FunctionType<'ctx> {
context.void_type().fn_type(Default::default(), false)
}
fn emit_body<'ctx>(
&self,
context: &mut Context<'ctx, D>,
) -> anyhow::Result<Option<inkwell::values::BasicValueEnum<'ctx>>> {
let immutable_data_size_pointer = context
.get_global(revive_runtime_api::immutable_data::GLOBAL_IMMUTABLE_DATA_SIZE)?
.value
.as_pointer_value();
let immutable_data_size = context.build_load(
Pointer::new(
context.xlen_type(),
AddressSpace::Stack,
immutable_data_size_pointer,
),
"immutable_data_size_load",
)?;
let write_immutable_data_block = context.append_basic_block("write_immutables_block");
let join_return_block = context.append_basic_block("join_return_block");
let immutable_data_size_is_zero = context.builder().build_int_compare(
inkwell::IntPredicate::EQ,
context.xlen_type().const_zero(),
immutable_data_size.into_int_value(),
"immutable_data_size_is_zero",
)?;
context.build_conditional_branch(
immutable_data_size_is_zero,
join_return_block,
write_immutable_data_block,
)?;
context.set_basic_block(write_immutable_data_block);
let immutable_data_pointer = context
.get_global(revive_runtime_api::immutable_data::GLOBAL_IMMUTABLE_DATA_POINTER)?
.value
.as_pointer_value();
context.build_runtime_call(
revive_runtime_api::polkavm_imports::SET_IMMUTABLE_DATA,
&[
context
.builder()
.build_ptr_to_int(
immutable_data_pointer,
context.xlen_type(),
"immutable_data_pointer_to_xlen",
)?
.into(),
immutable_data_size,
],
);
context.build_unconditional_branch(join_return_block);
context.set_basic_block(join_return_block);
Ok(None)
}
}
impl<D> WriteLLVM<D> for Store
where
D: Dependency + Clone,
{
fn declare(&mut self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::declare(self, context)
}
fn into_llvm(self, context: &mut Context<D>) -> anyhow::Result<()> {
<Self as RuntimeFunction<_>>::emit(&self, context)
}
}
/// Translates the contract immutable load.
///
@@ -27,14 +223,15 @@ where
}
Some(CodeType::Deploy) => load_from_memory(context, index),
Some(CodeType::Runtime) => {
let name = <Load as RuntimeFunction<D>>::NAME;
context.build_call(
context
.get_function(runtime::FUNCTION_LOAD_IMMUTABLE_DATA)
.get_function(name)
.expect("is always declared for runtime code")
.borrow()
.declaration(),
&[],
runtime::FUNCTION_LOAD_IMMUTABLE_DATA,
name,
);
load_from_memory(context, index)
}
+13 -3
View File
@@ -29,7 +29,8 @@ where
pub fn load<'ctx, D>(
context: &mut Context<'ctx, D>,
offset: inkwell::values::IntValue<'ctx>,
) -> anyhow::Result<inkwell::values::BasicValueEnum<'ctx>>
assignment_pointer: &mut Option<inkwell::values::PointerValue<'ctx>>,
) -> anyhow::Result<Option<inkwell::values::BasicValueEnum<'ctx>>>
where
D: Dependency + Clone,
{
@@ -40,7 +41,7 @@ where
offset,
"memory_load_pointer",
);
context.build_load(pointer, "memory_load_result")
context.build_load_assign(pointer, "memory_load_result", assignment_pointer)
}
/// Translates the `mstore` instruction.
@@ -85,5 +86,14 @@ where
offset,
"mstore8_destination",
);
context.build_store(pointer, value)
let pointer = context.build_sbrk(
pointer.to_int(context),
context.xlen_type().const_int(1, false),
)?;
context
.builder()
.build_store(pointer, value)?
.set_alignment(revive_common::BYTE_LENGTH_BYTE as u32)
.expect("Alignment is valid");
Ok(())
}
+7 -55
View File
@@ -1,9 +1,9 @@
//! Translates the transaction return operations.
use crate::polkavm::context::address_space::AddressSpace;
use crate::polkavm::context::code_type::CodeType;
use crate::polkavm::context::pointer::Pointer;
use crate::polkavm::context::runtime::RuntimeFunction;
use crate::polkavm::context::Context;
use crate::polkavm::evm::immutable::Store;
use crate::polkavm::Dependency;
/// Translates the `return` instruction.
@@ -18,55 +18,11 @@ where
match context.code_type() {
None => anyhow::bail!("Return is not available if the contract part is undefined"),
Some(CodeType::Deploy) => {
let immutable_data_size_pointer = context
.get_global(revive_runtime_api::immutable_data::GLOBAL_IMMUTABLE_DATA_SIZE)?
.value
.as_pointer_value();
let immutable_data_size = context.build_load(
Pointer::new(
context.xlen_type(),
AddressSpace::Stack,
immutable_data_size_pointer,
),
"immutable_data_size_load",
)?;
let write_immutable_data_block = context.append_basic_block("write_immutables_block");
let join_return_block = context.append_basic_block("join_return_block");
let immutable_data_size_is_zero = context.builder().build_int_compare(
inkwell::IntPredicate::EQ,
context.xlen_type().const_zero(),
immutable_data_size.into_int_value(),
"immutable_data_size_is_zero",
)?;
context.build_conditional_branch(
immutable_data_size_is_zero,
join_return_block,
write_immutable_data_block,
)?;
context.set_basic_block(write_immutable_data_block);
let immutable_data_pointer = context
.get_global(revive_runtime_api::immutable_data::GLOBAL_IMMUTABLE_DATA_POINTER)?
.value
.as_pointer_value();
context.build_runtime_call(
revive_runtime_api::polkavm_imports::SET_IMMUTABLE_DATA,
&[
context
.builder()
.build_ptr_to_int(
immutable_data_pointer,
context.xlen_type(),
"immutable_data_pointer_to_xlen",
)?
.into(),
immutable_data_size,
],
context.build_call(
<Store as RuntimeFunction<D>>::declaration(context),
Default::default(),
"store_immutable_data",
);
context.build_unconditional_branch(join_return_block);
context.set_basic_block(join_return_block);
}
Some(CodeType::Runtime) => {}
}
@@ -100,11 +56,7 @@ pub fn stop<D>(context: &mut Context<D>) -> anyhow::Result<()>
where
D: Dependency + Clone,
{
r#return(
context,
context.integer_const(crate::polkavm::XLEN, 0),
context.integer_const(crate::polkavm::XLEN, 0),
)
r#return(context, context.word_const(0), context.word_const(0))
}
/// Translates the `invalid` instruction.
+55 -24
View File
@@ -1,65 +1,96 @@
//! Translates the storage operations.
use crate::polkavm::context::address_space::AddressSpace;
use crate::polkavm::context::runtime::RuntimeFunction;
use crate::polkavm::context::Context;
use crate::polkavm::Dependency;
use crate::PolkaVMArgument;
use crate::PolkaVMLoadStorageWordFunction;
use crate::PolkaVMLoadTransientStorageWordFunction;
use crate::PolkaVMStoreStorageWordFunction;
use crate::PolkaVMStoreTransientStorageWordFunction;
/// Translates the storage load.
pub fn load<'ctx, D>(
context: &mut Context<'ctx, D>,
position: inkwell::values::IntValue<'ctx>,
) -> anyhow::Result<inkwell::values::BasicValueEnum<'ctx>>
position: &PolkaVMArgument<'ctx>,
assignment_pointer: &mut Option<inkwell::values::PointerValue<'ctx>>,
) -> anyhow::Result<Option<inkwell::values::BasicValueEnum<'ctx>>>
where
D: Dependency + Clone,
{
let mut slot_ptr = context.build_alloca_at_entry(context.word_type(), "slot_pointer");
slot_ptr.address_space = AddressSpace::Storage;
context.builder().build_store(slot_ptr.value, position)?;
context.build_load(slot_ptr, "storage_load_value")
let _name = <PolkaVMLoadStorageWordFunction as RuntimeFunction<D>>::NAME;
let declaration = <PolkaVMLoadStorageWordFunction as RuntimeFunction<D>>::declaration(context);
match assignment_pointer.take() {
Some(assignment_pointer) => {
let arguments = [
position.as_pointer(context)?.value.into(),
assignment_pointer.into(),
];
context.build_call(declaration, &arguments, "storage_load");
Ok(None)
}
None => {
let pointer = context.build_alloca_at_entry(context.word_type(), "pointer");
let arguments = [
position.as_pointer(context)?.value.into(),
pointer.value.into(),
];
context.build_call(declaration, &arguments, "storage_load");
Ok(Some(context.build_load(pointer, "storage_value")?))
}
}
}
/// Translates the storage store.
pub fn store<'ctx, D>(
context: &mut Context<'ctx, D>,
position: inkwell::values::IntValue<'ctx>,
value: inkwell::values::IntValue<'ctx>,
position: &PolkaVMArgument<'ctx>,
value: &PolkaVMArgument<'ctx>,
) -> anyhow::Result<()>
where
D: Dependency + Clone,
{
let mut slot_ptr = context.build_alloca_at_entry(context.word_type(), "slot_pointer");
slot_ptr.address_space = AddressSpace::Storage;
context.builder().build_store(slot_ptr.value, position)?;
context.build_store(slot_ptr, value)?;
let declaration = <PolkaVMStoreStorageWordFunction as RuntimeFunction<D>>::declaration(context);
let arguments = [
position.as_pointer(context)?.value.into(),
value.as_pointer(context)?.value.into(),
];
context.build_call(declaration, &arguments, "storage_store");
Ok(())
}
/// Translates the transient storage load.
pub fn transient_load<'ctx, D>(
context: &mut Context<'ctx, D>,
position: inkwell::values::IntValue<'ctx>,
position: &PolkaVMArgument<'ctx>,
) -> anyhow::Result<inkwell::values::BasicValueEnum<'ctx>>
where
D: Dependency + Clone,
{
let mut slot_ptr = context.build_alloca_at_entry(context.word_type(), "slot_pointer");
slot_ptr.address_space = AddressSpace::TransientStorage;
context.builder().build_store(slot_ptr.value, position)?;
context.build_load(slot_ptr, "transient_storage_load_value")
let name = <PolkaVMLoadTransientStorageWordFunction as RuntimeFunction<D>>::NAME;
let arguments = [position.as_pointer(context)?.value.into()];
let declaration =
<PolkaVMLoadTransientStorageWordFunction as RuntimeFunction<D>>::declaration(context);
Ok(context
.build_call(declaration, &arguments, "transient_storage_load")
.unwrap_or_else(|| panic!("runtime function {name} should return a value")))
}
/// Translates the transient storage store.
pub fn transient_store<'ctx, D>(
context: &mut Context<'ctx, D>,
position: inkwell::values::IntValue<'ctx>,
value: inkwell::values::IntValue<'ctx>,
position: &PolkaVMArgument<'ctx>,
value: &PolkaVMArgument<'ctx>,
) -> anyhow::Result<()>
where
D: Dependency + Clone,
{
let mut slot_ptr = context.build_alloca_at_entry(context.word_type(), "slot_pointer");
slot_ptr.address_space = AddressSpace::TransientStorage;
context.builder().build_store(slot_ptr.value, position)?;
context.build_store(slot_ptr, value)?;
let declaration =
<PolkaVMStoreTransientStorageWordFunction as RuntimeFunction<D>>::declaration(context);
let arguments = [
position.as_pointer(context)?.value.into(),
value.as_pointer(context)?.value.into(),
];
context.build_call(declaration, &arguments, "transient_storage_store");
Ok(())
}
+2 -6
View File
@@ -3,7 +3,6 @@
pub mod r#const;
pub mod context;
pub mod evm;
pub mod metadata_hash;
pub use self::r#const::*;
@@ -18,11 +17,6 @@ use sha3::Digest;
use self::context::build::Build;
use self::context::Context;
/// Initializes the PolkaVM target machine.
pub fn initialize_target() {
inkwell::targets::Target::initialize_riscv(&Default::default());
}
/// Builds PolkaVM assembly text.
pub fn build_assembly_text(
contract_path: &str,
@@ -95,6 +89,7 @@ pub trait Dependency {
optimizer_settings: OptimizerSettings,
include_metadata_hash: bool,
debug_config: DebugConfig,
llvm_arguments: &[String],
) -> anyhow::Result<String>;
/// Resolves a full contract path.
@@ -115,6 +110,7 @@ impl Dependency for DummyDependency {
_optimizer_settings: OptimizerSettings,
_include_metadata_hash: bool,
_debug_config: DebugConfig,
_llvm_arguments: &[String],
) -> anyhow::Result<String> {
Ok(String::new())
}
+1
View File
@@ -34,3 +34,4 @@ polkadot-sdk.features = [
revive-solidity = { workspace = true, optional = true }
revive-differential = { workspace = true, optional = true }
revive-llvm-context = { workspace = true }
+1
View File
@@ -274,6 +274,7 @@ impl From<Code> for pallet_revive::Code {
&contract,
&source_code,
solc_optimizer.unwrap_or(true),
revive_llvm_context::OptimizerSettings::cycles(),
))
}
Code::Path(path) => pallet_revive::Code::Upload(std::fs::read(path).unwrap()),
+2 -6
View File
@@ -106,14 +106,10 @@ impl SpecsAction {
};
for (key, expected) in storage {
let mut key = **key;
let mut expected = **expected;
key.reverse();
expected.reverse();
actions.push(Self::VerifyStorage {
contract: account_pvm.clone(),
key,
expected,
key: **key,
expected: **expected,
});
}
+3 -3
View File
@@ -13,18 +13,18 @@ uint32_t __memory_size = 0;
void * __sbrk_internal(uint32_t offset, uint32_t size) {
if (offset >= MAX_MEMORY_SIZE || size > MAX_MEMORY_SIZE) {
return NULL;
POLKAVM_TRAP();
}
uint32_t new_size = ALIGN(offset + size);
if (new_size > MAX_MEMORY_SIZE) {
return NULL;
POLKAVM_TRAP();
}
if (new_size > __memory_size) {
__memory_size = new_size;
}
return (void *)&__memory[__memory_size];
return (void *)&__memory[offset];
}
void * memset(void *b, int c, size_t len) {
+23
View File
@@ -0,0 +1,23 @@
[package]
name = "revive-solc-json-interface"
version.workspace = true
authors.workspace = true
license.workspace = true
edition.workspace = true
repository.workspace = true
rust-version.workspace = true
description = "Rust bindings for the solc standard JSON and combined JSON interface"
[features]
default = ["parallel"]
parallel = ["rayon"]
resolc = [] # The resolc binary adds a bunch of custom fields to the format
[dependencies]
revive-common = { workspace = true }
anyhow = { workspace = true }
rayon = { workspace = true, optional = true }
semver = { workspace = true }
serde = { workspace = true }
serde_json = { workspace = true }
+27
View File
@@ -0,0 +1,27 @@
//! This crates provides (de)serializable Rust types for interacting
//! `solc` via the [JSON-input-output][0] interface.
//!
//! [0]: https://docs.soliditylang.org/en/latest/using-the-compiler.html#compiler-input-and-output-json-description
pub use self::combined_json::contract::Contract as CombinedJsonContract;
pub use self::standard_json::input::language::Language as SolcStandardJsonInputLanguage;
pub use self::standard_json::input::settings::metadata::Metadata as SolcStandardJsonInputSettingsMetadata;
pub use self::standard_json::input::settings::metadata_hash::MetadataHash as SolcStandardJsonInputSettingsMetadataHash;
pub use self::standard_json::input::settings::optimizer::Optimizer as SolcStandardJsonInputSettingsOptimizer;
pub use self::standard_json::input::settings::selection::file::flag::Flag as SolcStandardJsonInputSettingsSelectionFileFlag;
pub use self::standard_json::input::settings::selection::file::File as SolcStandardJsonInputSettingsSelectionFile;
pub use self::standard_json::input::settings::selection::Selection as SolcStandardJsonInputSettingsSelection;
pub use self::standard_json::input::settings::Settings as SolcStandardJsonInputSettings;
pub use self::standard_json::input::source::Source as SolcStandardJsonInputSource;
pub use self::standard_json::input::Input as SolcStandardJsonInput;
pub use self::standard_json::output::contract::evm::bytecode::Bytecode as SolcStandardJsonOutputContractEVMBytecode;
pub use self::standard_json::output::contract::evm::EVM as SolcStandardJsonOutputContractEVM;
pub use self::standard_json::output::contract::Contract as SolcStandardJsonOutputContract;
pub use self::standard_json::output::Output as SolcStandardJsonOutput;
#[cfg(feature = "resolc")]
pub use self::warning::Warning as ResolcWarning;
pub mod combined_json;
pub mod standard_json;
#[cfg(feature = "resolc")]
pub mod warning;
@@ -8,14 +8,15 @@ use std::collections::BTreeMap;
use std::collections::BTreeSet;
use std::path::PathBuf;
#[cfg(feature = "parallel")]
#[cfg(all(feature = "parallel", feature = "resolc"))]
use rayon::iter::{IntoParallelIterator, ParallelIterator};
use serde::Deserialize;
use serde::Serialize;
use crate::solc::standard_json::input::settings::metadata::Metadata as SolcStandardJsonInputSettingsMetadata;
use crate::solc::standard_json::input::settings::optimizer::Optimizer as SolcStandardJsonInputSettingsOptimizer;
use crate::solc::standard_json::input::settings::selection::Selection as SolcStandardJsonInputSettingsSelection;
use crate::standard_json::input::settings::metadata::Metadata as SolcStandardJsonInputSettingsMetadata;
use crate::standard_json::input::settings::optimizer::Optimizer as SolcStandardJsonInputSettingsOptimizer;
use crate::standard_json::input::settings::selection::Selection as SolcStandardJsonInputSettingsSelection;
#[cfg(feature = "resolc")]
use crate::warning::Warning;
use self::language::Language;
@@ -33,6 +34,7 @@ pub struct Input {
/// The compiler settings.
pub settings: Settings,
/// The suppressed warnings.
#[cfg(feature = "resolc")]
#[serde(skip_serializing)]
pub suppressed_warnings: Option<Vec<Warning>>,
}
@@ -60,7 +62,7 @@ impl Input {
output_selection: SolcStandardJsonInputSettingsSelection,
optimizer: SolcStandardJsonInputSettingsOptimizer,
metadata: Option<SolcStandardJsonInputSettingsMetadata>,
suppressed_warnings: Option<Vec<Warning>>,
#[cfg(feature = "resolc")] suppressed_warnings: Option<Vec<Warning>>,
) -> anyhow::Result<Self> {
let mut paths: BTreeSet<PathBuf> = paths.iter().cloned().collect();
let libraries = Settings::parse_libraries(library_map)?;
@@ -89,12 +91,14 @@ impl Input {
optimizer,
metadata,
),
#[cfg(feature = "resolc")]
suppressed_warnings,
})
}
/// A shortcut constructor from source code.
/// Only for the integration test purposes.
#[cfg(feature = "resolc")]
#[allow(clippy::too_many_arguments)]
pub fn try_from_sources(
evm_version: Option<revive_common::EVMVersion>,
@@ -3,18 +3,20 @@
use serde::Deserialize;
use serde::Serialize;
use crate::standard_json::input::settings::metadata_hash::MetadataHash;
/// The `solc --standard-json` input settings metadata.
#[derive(Debug, Default, Serialize, Deserialize)]
#[serde(rename_all = "camelCase")]
pub struct Metadata {
/// The bytecode hash mode.
#[serde(skip_serializing_if = "Option::is_none")]
pub bytecode_hash: Option<revive_llvm_context::PolkaVMMetadataHash>,
pub bytecode_hash: Option<MetadataHash>,
}
impl Metadata {
/// A shortcut constructor.
pub fn new(bytecode_hash: revive_llvm_context::PolkaVMMetadataHash) -> Self {
pub fn new(bytecode_hash: MetadataHash) -> Self {
Self {
bytecode_hash: Some(bytecode_hash),
}
@@ -1,6 +1,7 @@
//! The `solc --standard-json` input settings.
pub mod metadata;
pub mod metadata_hash;
pub mod optimizer;
pub mod selection;
@@ -49,18 +49,3 @@ impl Optimizer {
};
}
}
impl TryFrom<&Optimizer> for revive_llvm_context::OptimizerSettings {
type Error = anyhow::Error;
fn try_from(value: &Optimizer) -> Result<Self, Self::Error> {
let mut result = match value.mode {
Some(mode) => Self::try_from_cli(mode)?,
None => Self::cycles(),
};
if value.fallback_to_optimizing_for_size.unwrap_or_default() {
result.enable_fallback_to_size();
}
Ok(result)
}
}
@@ -40,6 +40,9 @@ pub enum Flag {
/// The assembly code
#[serde(rename = "evm.assembly")]
Assembly,
/// The Ir
#[serde(rename = "ir")]
Ir,
}
impl std::fmt::Display for Flag {
@@ -57,6 +60,7 @@ impl std::fmt::Display for Flag {
Self::EVMBC => write!(f, "evm.bytecode"),
Self::EVMDBC => write!(f, "evm.deployedBytecode"),
Self::Assembly => write!(f, "evm.assembly"),
Self::Ir => write!(f, "ir"),
}
}
}
@@ -10,7 +10,7 @@ use serde::Serialize;
use self::flag::Flag as SelectionFlag;
/// The `solc --standard-json` output file selection.
#[derive(Debug, Default, Serialize, Deserialize)]
#[derive(Debug, Default, Serialize, Deserialize, PartialEq)]
pub struct File {
/// The per-file output selections.
#[serde(rename = "", skip_serializing_if = "Option::is_none")]
@@ -45,6 +45,7 @@ impl File {
self.per_contract
.get_or_insert_with(HashSet::default)
.extend(required.per_contract.unwrap_or_default());
self
}
}
@@ -0,0 +1,99 @@
//! The `solc --standard-json` output selection.
pub mod file;
use std::collections::BTreeMap;
use serde::Deserialize;
use serde::Serialize;
use self::file::File as FileSelection;
/// The `solc --standard-json` output selection.
#[derive(Debug, Serialize, Deserialize, Default, PartialEq)]
pub struct Selection {
/// Only the 'all' wildcard is available for robustness reasons.
#[serde(rename = "*", skip_serializing_if = "Option::is_none")]
all: Option<FileSelection>,
#[serde(skip_serializing_if = "BTreeMap::is_empty", flatten)]
pub files: BTreeMap<String, FileSelection>,
}
impl Selection {
/// Creates the selection required by our compilation process.
pub fn new_required() -> Self {
Self {
all: Some(FileSelection::new_required()),
files: BTreeMap::new(),
}
}
/// Extends the user's output selection with flag required by our compilation process.
pub fn extend_with_required(&mut self) -> &mut Self {
self.all
.get_or_insert_with(FileSelection::new_required)
.extend_with_required();
for (_, v) in self.files.iter_mut() {
v.extend_with_required();
}
self
}
}
#[cfg(test)]
mod test {
use std::collections::BTreeMap;
use crate::SolcStandardJsonInputSettingsSelectionFile;
use super::Selection;
#[test]
fn per_file() {
let init = Selection {
all: None,
files: BTreeMap::from([(
"Test".to_owned(),
SolcStandardJsonInputSettingsSelectionFile::new_required(),
)]),
};
let deser = serde_json::to_string(&init)
.and_then(|string| serde_json::from_str(&string))
.unwrap();
assert_eq!(init, deser)
}
#[test]
fn all() {
let init = Selection {
all: Some(SolcStandardJsonInputSettingsSelectionFile::new_required()),
files: BTreeMap::new(),
};
let deser = serde_json::to_string(&init)
.and_then(|string| serde_json::from_str(&string))
.unwrap();
assert_eq!(init, deser)
}
#[test]
fn all_and_override() {
let init = Selection {
all: Some(SolcStandardJsonInputSettingsSelectionFile::new_required()),
files: BTreeMap::from([(
"Test".to_owned(),
SolcStandardJsonInputSettingsSelectionFile::new_required(),
)]),
};
let deser = serde_json::to_string(&init)
.and_then(|string| serde_json::from_str(&string))
.unwrap();
assert_eq!(init, deser)
}
}
@@ -0,0 +1,4 @@
//! The `solc <input>.sol --standard-json` interface input and output.
pub mod input;
pub mod output;
@@ -32,6 +32,9 @@ pub struct Contract {
/// Contract's bytecode and related objects
#[serde(default, skip_serializing_if = "Option::is_none")]
pub evm: Option<EVM>,
/// The contract IR code.
#[serde(default, skip_serializing_if = "Option::is_none")]
pub ir: Option<String>,
/// The contract optimized IR code.
#[serde(default, skip_serializing_if = "Option::is_none")]
pub ir_optimized: Option<String>,
@@ -0,0 +1,71 @@
//! The `solc --standard-json` output.
pub mod contract;
pub mod error;
pub mod source;
use std::collections::BTreeMap;
use serde::Deserialize;
use serde::Serialize;
#[cfg(feature = "resolc")]
use crate::warning::Warning;
use self::contract::Contract;
use self::error::Error as SolcStandardJsonOutputError;
use self::source::Source;
/// The `solc --standard-json` output.
#[derive(Debug, Serialize, Deserialize, Clone, Default)]
pub struct Output {
/// The file-contract hashmap.
#[serde(default, skip_serializing_if = "Option::is_none")]
pub contracts: Option<BTreeMap<String, BTreeMap<String, Contract>>>,
/// The source code mapping data.
#[serde(default, skip_serializing_if = "Option::is_none")]
pub sources: Option<BTreeMap<String, Source>>,
/// The compilation errors and warnings.
#[serde(default, skip_serializing_if = "Option::is_none")]
pub errors: Option<Vec<SolcStandardJsonOutputError>>,
/// The `solc` compiler version.
#[serde(skip_serializing_if = "Option::is_none")]
pub version: Option<String>,
/// The `solc` compiler long version.
#[serde(skip_serializing_if = "Option::is_none")]
pub long_version: Option<String>,
/// The `resolc` compiler version.
#[serde(skip_serializing_if = "Option::is_none")]
pub revive_version: Option<String>,
}
impl Output {
/// Traverses the AST and returns the list of additional errors and warnings.
#[cfg(feature = "resolc")]
pub fn preprocess_ast(&mut self, suppressed_warnings: &[Warning]) -> anyhow::Result<()> {
let sources = match self.sources.as_ref() {
Some(sources) => sources,
None => return Ok(()),
};
let mut messages = Vec::new();
for (path, source) in sources.iter() {
if let Some(ast) = source.ast.as_ref() {
let mut polkavm_messages = Source::get_messages(ast, suppressed_warnings);
for message in polkavm_messages.iter_mut() {
message.push_contract_path(path.as_str());
}
messages.extend(polkavm_messages);
}
}
self.errors = match self.errors.take() {
Some(mut errors) => {
errors.extend(messages);
Some(errors)
}
None => Some(messages),
};
Ok(())
}
}
@@ -3,7 +3,8 @@
use serde::Deserialize;
use serde::Serialize;
use crate::solc::standard_json::output::error::Error as SolcStandardJsonOutputError;
use crate::standard_json::output::error::Error as SolcStandardJsonOutputError;
#[cfg(feature = "resolc")]
use crate::warning::Warning;
/// The `solc --standard-json` output source.
@@ -131,6 +132,7 @@ impl Source {
}
/// Returns the list of messages for some specific parts of the AST.
#[cfg(feature = "resolc")]
pub fn get_messages(
ast: &serde_json::Value,
suppressed_warnings: &[Warning],
@@ -194,7 +196,7 @@ impl Source {
_ => None,
},
)
.last()
.next_back()
.ok_or_else(|| anyhow::anyhow!("The last contract not found in the AST"))
}
}
@@ -1,26 +1,22 @@
//! The compiler warning.
//! `resolc` custom compiler warnings.
//!
//! The revive compiler adds warnings only applicable when compilng
//! to the revive stack on Polkadot to the output.
use std::str::FromStr;
use serde::Deserialize;
use serde::Serialize;
/// The compiler warning.
// The `resolc` custom compiler warning.
#[derive(Debug, Serialize, Deserialize, Clone, Copy, PartialEq, Eq, Hash)]
pub enum Warning {
/// The warning for eponymous feature.
EcRecover,
/// The warning for eponymous feature.
SendTransfer,
/// The warning for eponymous feature.
ExtCodeSize,
/// The warning for eponymous feature.
TxOrigin,
/// The warning for eponymous feature.
BlockTimestamp,
/// The warning for eponymous feature.
BlockNumber,
/// The warning for eponymous feature.
BlockHash,
}
+1
View File
@@ -37,6 +37,7 @@ inkwell = { workspace = true }
revive-common = { workspace = true }
revive-llvm-context = { workspace = true }
revive-solc-json-interface = { workspace = true, features = ["resolc"] }
[target.'cfg(target_env = "musl")'.dependencies]
mimalloc = { version = "*", default-features = false }
+3 -4
View File
@@ -5,12 +5,11 @@ use std::fs::File;
use std::io::Write;
use std::path::Path;
use revive_solc_json_interface::CombinedJsonContract;
use revive_solc_json_interface::SolcStandardJsonOutputContract;
use serde::Deserialize;
use serde::Serialize;
use crate::solc::combined_json::contract::Contract as CombinedJsonContract;
use crate::solc::standard_json::output::contract::Contract as StandardJsonOutputContract;
/// The Solidity contract build.
#[derive(Debug, Serialize, Deserialize)]
pub struct Contract {
@@ -131,7 +130,7 @@ impl Contract {
/// Writes the contract text assembly and bytecode to the standard JSON.
pub fn write_to_standard_json(
self,
standard_json_contract: &mut StandardJsonOutputContract,
standard_json_contract: &mut SolcStandardJsonOutputContract,
) -> anyhow::Result<()> {
standard_json_contract.metadata = Some(self.metadata_json);

Some files were not shown because too many files have changed in this diff Show More