Mo/feat/v10/add shared kmp and jvm module by typfel · Pull Request #1795 · wireapp/core-crypto

typfel · 2026-01-28T12:38:52Z

What's new in this PR

PR Submission Checklist for internal contributors

The PR Title
- conforms to the style of semantic commits messages¹ supported in Wire's Github Workflow²
- contains a reference JIRA issue number like SQPIT-764
- answers the question: If merged, this PR will: ... ³

https://sparkbox.com/foundry/semantic_commit_messages
https://github.com/wireapp/.github#usage
E.g. feat(conversation-list): Sort conversations by most emojis in the title #SQPIT-764.

github-actions · 2026-01-28T12:53:33Z

Bencher Report

Branch	mo/feat/v10/add-shared-kmp-and-jvm-module
Testbed	ubuntu-latest

⚠️ WARNING: No Threshold found!
Without a Threshold, no Alerts will ever be generated.
Latency (nanoseconds (ns))
Click here to create a new Threshold
For more information, see the Threshold documentation.
To only post results if a Threshold exists, set the --ci-only-thresholds flag.

Click to view all benchmark results

Benchmark	Latency	microseconds (µs)
Commit add f(group size)/cs1/mem/1002	📈 view plot ⚠️ NO THRESHOLD	15,846.00 µs
Commit add f(group size)/cs1/mem/2	📈 view plot ⚠️ NO THRESHOLD	663.47 µs
Commit add f(group size)/cs1/mem/202	📈 view plot ⚠️ NO THRESHOLD	4,201.30 µs
Commit add f(group size)/cs1/mem/402	📈 view plot ⚠️ NO THRESHOLD	6,959.80 µs
Commit add f(group size)/cs1/mem/602	📈 view plot ⚠️ NO THRESHOLD	10,676.00 µs
Commit add f(group size)/cs1/mem/802	📈 view plot ⚠️ NO THRESHOLD	13,416.00 µs
Commit add f(number clients)/cs1/mem/1002	📈 view plot ⚠️ NO THRESHOLD	984,110.00 µs
Commit add f(number clients)/cs1/mem/2	📈 view plot ⚠️ NO THRESHOLD	661.87 µs
Commit add f(number clients)/cs1/mem/202	📈 view plot ⚠️ NO THRESHOLD	80,219.00 µs
Commit add f(number clients)/cs1/mem/402	📈 view plot ⚠️ NO THRESHOLD	215,460.00 µs
Commit add f(number clients)/cs1/mem/602	📈 view plot ⚠️ NO THRESHOLD	424,530.00 µs
Commit add f(number clients)/cs1/mem/802	📈 view plot ⚠️ NO THRESHOLD	671,360.00 µs
Commit pending proposals f(group size)/cs1/mem/1002	📈 view plot ⚠️ NO THRESHOLD	115,020.00 µs
Commit pending proposals f(group size)/cs1/mem/2	📈 view plot ⚠️ NO THRESHOLD	22,749.00 µs
Commit pending proposals f(group size)/cs1/mem/202	📈 view plot ⚠️ NO THRESHOLD	40,939.00 µs
Commit pending proposals f(group size)/cs1/mem/402	📈 view plot ⚠️ NO THRESHOLD	56,516.00 µs
Commit pending proposals f(group size)/cs1/mem/602	📈 view plot ⚠️ NO THRESHOLD	76,417.00 µs
Commit pending proposals f(group size)/cs1/mem/802	📈 view plot ⚠️ NO THRESHOLD	91,744.00 µs
Commit pending proposals f(pending size)/cs1/mem/1	📈 view plot ⚠️ NO THRESHOLD	15,503.00 µs
Commit pending proposals f(pending size)/cs1/mem/101	📈 view plot ⚠️ NO THRESHOLD	113,560.00 µs
Commit pending proposals f(pending size)/cs1/mem/21	📈 view plot ⚠️ NO THRESHOLD	32,601.00 µs
Commit pending proposals f(pending size)/cs1/mem/41	📈 view plot ⚠️ NO THRESHOLD	54,046.00 µs
Commit pending proposals f(pending size)/cs1/mem/61	📈 view plot ⚠️ NO THRESHOLD	72,952.00 µs
Commit pending proposals f(pending size)/cs1/mem/81	📈 view plot ⚠️ NO THRESHOLD	93,525.00 µs
Commit remove f(group size)/cs1/mem/1002	📈 view plot ⚠️ NO THRESHOLD	10,192.00 µs
Commit remove f(group size)/cs1/mem/2	📈 view plot ⚠️ NO THRESHOLD	497.10 µs
Commit remove f(group size)/cs1/mem/202	📈 view plot ⚠️ NO THRESHOLD	1,933.90 µs
Commit remove f(group size)/cs1/mem/402	📈 view plot ⚠️ NO THRESHOLD	3,832.50 µs
Commit remove f(group size)/cs1/mem/602	📈 view plot ⚠️ NO THRESHOLD	5,446.10 µs
Commit remove f(group size)/cs1/mem/802	📈 view plot ⚠️ NO THRESHOLD	7,204.60 µs
Commit remove f(number clients)/cs1/mem/1002	📈 view plot ⚠️ NO THRESHOLD	12,726.00 µs
Commit remove f(number clients)/cs1/mem/2	📈 view plot ⚠️ NO THRESHOLD	133,000.00 µs
Commit remove f(number clients)/cs1/mem/202	📈 view plot ⚠️ NO THRESHOLD	108,930.00 µs
Commit remove f(number clients)/cs1/mem/402	📈 view plot ⚠️ NO THRESHOLD	85,193.00 µs
Commit remove f(number clients)/cs1/mem/602	📈 view plot ⚠️ NO THRESHOLD	60,669.00 µs
Commit remove f(number clients)/cs1/mem/802	📈 view plot ⚠️ NO THRESHOLD	36,739.00 µs
Commit update f(group size)/cs1/mem/1002	📈 view plot ⚠️ NO THRESHOLD	132,590.00 µs
Commit update f(group size)/cs1/mem/2	📈 view plot ⚠️ NO THRESHOLD	666.34 µs
Commit update f(group size)/cs1/mem/202	📈 view plot ⚠️ NO THRESHOLD	27,422.00 µs
Commit update f(group size)/cs1/mem/402	📈 view plot ⚠️ NO THRESHOLD	53,788.00 µs
Commit update f(group size)/cs1/mem/602	📈 view plot ⚠️ NO THRESHOLD	81,148.00 µs
Commit update f(group size)/cs1/mem/802	📈 view plot ⚠️ NO THRESHOLD	106,850.00 µs

🐰 View full continuous benchmarking report in Bencher

…e to all targets

…e 8.14.3 interop was using gradle 9.x which caused incomaptibllity issues for kotlin multiplatform. Upgrading gradle is not an option until we are ready to also upgrade kotlin multiplatform to 2.3.0 or above.

istankovic · 2026-02-24T10:40:37Z

Why is commit 12fb0ac necessary?
The message mentions ANDROID_HOME but the changes check for ANDROID_NDK_HOME.

typfel · 2026-02-24T10:43:40Z

Why is commit 12fb0ac necessary? The message mentions ANDROID_HOME but the changes check for ANDROID_NDK_HOME.

Right commits message is misleading I'll update it, it was necessary to expose ANDROID_NDK_ROOT since that's used by the Gobley gradle plugin.

.gitignore

istankovic · 2026-02-24T11:04:37Z

What is the purpose of buildScriptOutputDirectories files?

kmp/cinterop/
├── cargoBuildAndroidArm64
│   ├── buildScriptOutputDirectories
│   └── nativeStaticLibsDef
├── cargoBuildAndroidArmV7
│   ├── buildScriptOutputDirectories
│   └── nativeStaticLibsDef
├── cargoBuildAndroidX86
│   ├── buildScriptOutputDirectories
│   └── nativeStaticLibsDef
├── cargoBuildIosArm64
│   ├── buildScriptOutputDirectories
│   └── nativeStaticLibsDef
├── cargoBuildIosSimulatorArm64
│   ├── buildScriptOutputDirectories
│   └── nativeStaticLibsDef
├── cargoBuildLinuxX64
│   ├── buildScriptOutputDirectories
│   └── nativeStaticLibsDef
└── cargoBuildMacOSArm64
    ├── buildScriptOutputDirectories
    └── nativeStaticLibsDef

istankovic · 2026-02-24T11:10:39Z

Running ktlint found these:

crypto-ffi/bindings/kmp/build.gradle.kts:1:1: Imports must be ordered in lexicographic order without any empty lines in-between with "java", "javax", "kotlin" and aliases in the end (standard:import-ordering)
crypto-ffi/bindings/kmp/build.gradle.kts:3:1: Wildcard import (standard:no-wildcard-imports)
crypto-ffi/bindings/kmp/build.gradle.kts:177:23: Line is exceeding max line length. Break line between assignment and expression (standard:binary-expression-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:68: Argument should be on a separate line (unless all arguments can fit a single line) (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:87: Argument should be on a separate line (unless all arguments can fit a single line) (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:111: Missing newline before ")" (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:113: Line is exceeding max line length. Break line before '?:' (standard:binary-expression-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:138: Argument should be on a separate line (unless all arguments can fit a single line) (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:141: Exceeded max line length (140) (standard:max-line-length)
crypto-ffi/bindings/kmp/build.gradle.kts:177:180: Missing newline before ")" (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:178:51: Redundant curly braces (standard:string-template)
crypto-ffi/bindings/kmp/build.gradle.kts:204:61: Missing spacing after "," (standard:comma-spacing)
crypto-ffi/bindings/kmp/build.gradle.kts:205:56: Unnecessary long whitespace (standard:no-multi-spaces)

Seems like our CI does not catch this due to the way kotlin-check Makefile rule works.

typfel · 2026-02-24T11:46:06Z

What is the purpose of buildScriptOutputDirectories files?

kmp/cinterop/
├── cargoBuildAndroidArm64
│   ├── buildScriptOutputDirectories
│   └── nativeStaticLibsDef
├── cargoBuildAndroidArmV7
│   ├── buildScriptOutputDirectories
│   └── nativeStaticLibsDef
├── cargoBuildAndroidX86
│   ├── buildScriptOutputDirectories
│   └── nativeStaticLibsDef
├── cargoBuildIosArm64
│   ├── buildScriptOutputDirectories
│   └── nativeStaticLibsDef
├── cargoBuildIosSimulatorArm64
│   ├── buildScriptOutputDirectories
│   └── nativeStaticLibsDef
├── cargoBuildLinuxX64
│   ├── buildScriptOutputDirectories
│   └── nativeStaticLibsDef
└── cargoBuildMacOSArm64
    ├── buildScriptOutputDirectories
    └── nativeStaticLibsDef

If you let the Gobley cargo plugin handle the cargo build it will contain paths to rust build output which it scans for dynamically linked libraries. We've disabled this part of the cargo plugin and use our externally build ffi libraries instead, but we still need provide it since the cinterop tasks in the plugin expects it to be there.

typfel · 2026-02-24T11:54:05Z

Running ktlint found these:

crypto-ffi/bindings/kmp/build.gradle.kts:1:1: Imports must be ordered in lexicographic order without any empty lines in-between with "java", "javax", "kotlin" and aliases in the end (standard:import-ordering)
crypto-ffi/bindings/kmp/build.gradle.kts:3:1: Wildcard import (standard:no-wildcard-imports)
crypto-ffi/bindings/kmp/build.gradle.kts:177:23: Line is exceeding max line length. Break line between assignment and expression (standard:binary-expression-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:68: Argument should be on a separate line (unless all arguments can fit a single line) (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:87: Argument should be on a separate line (unless all arguments can fit a single line) (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:111: Missing newline before ")" (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:113: Line is exceeding max line length. Break line before '?:' (standard:binary-expression-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:138: Argument should be on a separate line (unless all arguments can fit a single line) (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:141: Exceeded max line length (140) (standard:max-line-length)
crypto-ffi/bindings/kmp/build.gradle.kts:177:180: Missing newline before ")" (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:178:51: Redundant curly braces (standard:string-template)
crypto-ffi/bindings/kmp/build.gradle.kts:204:61: Missing spacing after "," (standard:comma-spacing)
crypto-ffi/bindings/kmp/build.gradle.kts:205:56: Unnecessary long whitespace (standard:no-multi-spaces)

Seems like our CI does not catch this due to the way kotlin-check Makefile rule works.

But was the linter every applied to any gradle files? Looking at the Makefile it only includes the source files.

istankovic · 2026-02-24T12:09:08Z

Running ktlint found these:

crypto-ffi/bindings/kmp/build.gradle.kts:1:1: Imports must be ordered in lexicographic order without any empty lines in-between with "java", "javax", "kotlin" and aliases in the end (standard:import-ordering)
crypto-ffi/bindings/kmp/build.gradle.kts:3:1: Wildcard import (standard:no-wildcard-imports)
crypto-ffi/bindings/kmp/build.gradle.kts:177:23: Line is exceeding max line length. Break line between assignment and expression (standard:binary-expression-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:68: Argument should be on a separate line (unless all arguments can fit a single line) (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:87: Argument should be on a separate line (unless all arguments can fit a single line) (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:111: Missing newline before ")" (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:113: Line is exceeding max line length. Break line before '?:' (standard:binary-expression-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:138: Argument should be on a separate line (unless all arguments can fit a single line) (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:177:141: Exceeded max line length (140) (standard:max-line-length)
crypto-ffi/bindings/kmp/build.gradle.kts:177:180: Missing newline before ")" (standard:argument-list-wrapping)
crypto-ffi/bindings/kmp/build.gradle.kts:178:51: Redundant curly braces (standard:string-template)
crypto-ffi/bindings/kmp/build.gradle.kts:204:61: Missing spacing after "," (standard:comma-spacing)
crypto-ffi/bindings/kmp/build.gradle.kts:205:56: Unnecessary long whitespace (standard:no-multi-spaces)

Seems like our CI does not catch this due to the way kotlin-check Makefile rule works.

But was the linter every applied to any gradle files? Looking at the Makefile it only includes the source files.

Currently it does not include the gradle files, but Makefile has been through many changes so it could be that we dropped it at some point.

Running ktlint inside bindings only complains about the new kmp file, not the other ones, so my guess is we had that linting before.

istankovic · 2026-02-24T12:10:35Z

If you let the Gobley cargo plugin handle the cargo build it will contain paths to rust build output which it scans for dynamically linked libraries. We've disabled this part of the cargo plugin and use our externally build ffi libraries instead, but we still need provide it since the cinterop tasks in the plugin expects it to be there.

Right, but note that buildScriptOutputDirectories are all empty and I am wondering what purpose that serves.

istankovic · 2026-02-24T12:20:38Z

Why is commit 12fb0ac necessary? The message mentions ANDROID_HOME but the changes check for ANDROID_NDK_HOME.

Right commits message is misleading I'll update it, it was necessary to expose ANDROID_NDK_ROOT since that's used by the Gobley gradle plugin.

In that case, I would just copy the android env setup from .github/workflows/build-android.yml:

      - name: setup android sdk
        uses: android-actions/setup-android@v3
        with:
          # Don't download unnecessary packages.
          # We only need sdkmanager for the next step.
          packages: ''

      - name: install ndk and set up environment
        run: |
          ndk_version=$(cat crypto-ffi/bindings/android/ndk.version)
          echo "y" | sdkmanager --install "ndk;$ndk_version"
          echo ANDROID_NDK_HOME=$ANDROID_HOME/ndk/$ndk_version >> $GITHUB_ENV

and adjust it accordingly. This would make it more consistent with other workflows and also avoid downloading unnecessary packages.

typfel · 2026-02-24T12:26:29Z

If you let the Gobley cargo plugin handle the cargo build it will contain paths to rust build output which it scans for dynamically linked libraries. We've disabled this part of the cargo plugin and use our externally build ffi libraries instead, but we still need provide it since the cinterop tasks in the plugin expects it to be there.

Right, but note that buildScriptOutputDirectories are all empty and I am wondering what purpose that serves.

They need to be there even if they are empty otherwise the gobley cargo plugin will throw an error.

istankovic · 2026-02-24T15:14:56Z

Right, but note that buildScriptOutputDirectories are all empty and I am wondering what purpose that serves.

They need to be there even if they are empty otherwise the gobley cargo plugin will throw an error.

Oh well. 🥴

crypto-ffi/bindings/kmp/build.gradle.kts

…e 2.0.21 downgrade kotlinx-serialization-json to 1.8.0 for kotlin 2.0.21 compatibility

…go plugin. also improve NDK setup by validating ANDROID_HOME and adding installation step

typfel · 2026-02-25T08:43:56Z

Right, but note that buildScriptOutputDirectories are all empty and I am wondering what purpose that serves.

They need to be there even if they are empty otherwise the gobley cargo plugin will throw an error.

Oh well. 🥴

Took another pass on this changed it so that we generate these empty files on the fly. Also removed the folders so that only thing remaining in the source tree is the actual def files.

istankovic

Looks good, thanks!

typfel requested a review from a team January 28, 2026 12:38

typfel force-pushed the mo/feat/v10/add-shared-kmp-and-jvm-module branch from 8c69730 to e40745b Compare January 28, 2026 12:49

typfel force-pushed the mo/feat/v10/add-shared-kmp-and-jvm-module branch 2 times, most recently from 3a3b05f to 646e1d6 Compare February 4, 2026 12:51

MohamadJaara force-pushed the mo/feat/v10/add-shared-kmp-and-jvm-module branch from b7767b8 to 8d6886a Compare February 5, 2026 13:24

typfel force-pushed the mo/feat/v10/add-shared-kmp-and-jvm-module branch 3 times, most recently from c168e0d to e72fb55 Compare February 11, 2026 07:59

typfel force-pushed the mo/feat/v10/add-shared-kmp-and-jvm-module branch 5 times, most recently from eecd970 to 0996e82 Compare February 24, 2026 09:57

MohamadJaara and others added 6 commits February 24, 2026 11:03

feat: Add shared KMP and JVM module

62fa358

feat: Add KMP module

f76bd51

test: move kotlin tests to the shared module

32e33ca

ci: run kmp jvm tests on CI

74de06e

ci: move kotlin file definitions to the top so that they are availabl…

ecc8e27

…e to all targets

fix(interop): align gradle version in interop and bindings to both us…

87352c5

…e 8.14.3 interop was using gradle 9.x which caused incomaptibllity issues for kotlin multiplatform. Upgrading gradle is not an option until we are ready to also upgrade kotlin multiplatform to 2.3.0 or above.

typfel force-pushed the mo/feat/v10/add-shared-kmp-and-jvm-module branch from 0996e82 to 6fc4c7f Compare February 24, 2026 10:03

istankovic reviewed Feb 24, 2026

View reviewed changes

.gitignore Show resolved Hide resolved

typfel force-pushed the mo/feat/v10/add-shared-kmp-and-jvm-module branch from d82cef2 to 464609f Compare February 24, 2026 11:51

typfel requested a review from istankovic February 24, 2026 13:24

istankovic reviewed Feb 24, 2026

View reviewed changes

crypto-ffi/bindings/kmp/build.gradle.kts Outdated Show resolved Hide resolved

istankovic reviewed Feb 24, 2026

View reviewed changes

crypto-ffi/bindings/kmp/build.gradle.kts Show resolved Hide resolved

istankovic reviewed Feb 24, 2026

View reviewed changes

crypto-ffi/bindings/kmp/build.gradle.kts Show resolved Hide resolved

typfel and others added 4 commits February 25, 2026 09:14

fix(interop): align kotlin version in interop and bindings to both us…

78c55bb

…e 2.0.21 downgrade kotlinx-serialization-json to 1.8.0 for kotlin 2.0.21 compatibility

chore: add gradle build directories to gitignore

9d26b8d

chore(interop): stop using compose activity

3158e71

fix(interop): export ANDROID_NDK_ROOT since it's read by Gobley's car…

ca3aa3a

…go plugin. also improve NDK setup by validating ANDROID_HOME and adding installation step

typfel requested a review from istankovic February 25, 2026 08:44

typfel force-pushed the mo/feat/v10/add-shared-kmp-and-jvm-module branch from d6c3eda to b98140a Compare February 25, 2026 08:51

istankovic approved these changes Feb 25, 2026

View reviewed changes

typfel added 3 commits February 25, 2026 10:08

ci: re-use already existing ffi artifacts when building kmp

d0fa68f

ci: include jvm and ffi lib as dependencies for kmp jvm tests

09ca136

ci: include gradle files when running the kotlin linter

fdbcab1

typfel force-pushed the mo/feat/v10/add-shared-kmp-and-jvm-module branch from b98140a to fdbcab1 Compare February 25, 2026 09:09

typfel merged commit fdbcab1 into main Feb 25, 2026
50 checks passed

typfel deleted the mo/feat/v10/add-shared-kmp-and-jvm-module branch February 25, 2026 09:27

typfel mentioned this pull request Feb 25, 2026

feat: add kmp bindings #1705

Closed

3 tasks

Conversation

typfel commented Jan 28, 2026

What's new in this PR

PR Submission Checklist for internal contributors

Uh oh!

github-actions bot commented Jan 28, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Bencher Report

⚠️ WARNING: No Threshold found!

Uh oh!

istankovic commented Feb 24, 2026

Uh oh!

typfel commented Feb 24, 2026

Uh oh!

Uh oh!

istankovic commented Feb 24, 2026

Uh oh!

istankovic commented Feb 24, 2026

Uh oh!

typfel commented Feb 24, 2026

Uh oh!

typfel commented Feb 24, 2026

Uh oh!

istankovic commented Feb 24, 2026

Uh oh!

istankovic commented Feb 24, 2026

Uh oh!

istankovic commented Feb 24, 2026

Uh oh!

typfel commented Feb 24, 2026

Uh oh!

istankovic commented Feb 24, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

typfel commented Feb 25, 2026

Uh oh!

istankovic left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

github-actions bot commented Jan 28, 2026 •

edited

Loading