Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux - #162063
Merged
Conversation
For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected And adds a dist-aarch64-linux-quick: * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). Once we decide where to enable this (e.g., in try builds by default) we can consider aligning with other tasks and saving $/build if we're not able to benefit from increased speed (e.g., because perf won't run until the try build as a whole finishes). For dist-x86_64-linux-full we have this breakdown: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore.
Collaborator
|
|
Member
Author
|
@bors try |
This comment has been minimized.
This comment has been minimized.
rust-bors Bot
pushed a commit
that referenced
this pull request
Aug 31, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux
Mark-Simulacrum
marked this pull request as ready for review
August 31, 2026 13:11
Collaborator
|
Warning If you are changing how CI LLVM is built or linked, make sure to bump cc @jieyouxu |
Contributor
Member
|
Looks good, thanks! |
Member
|
Forgot to approve it, lol. @bors r+ |
Contributor
JonathanBrouwer
added a commit
to JonathanBrouwer/rust
that referenced
this pull request
Aug 31, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
JonathanBrouwer
added a commit
to JonathanBrouwer/rust
that referenced
this pull request
Aug 31, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
JonathanBrouwer
added a commit
to JonathanBrouwer/rust
that referenced
this pull request
Aug 31, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
GuillaumeGomez
added a commit
to GuillaumeGomez/rust
that referenced
this pull request
Aug 31, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
JonathanBrouwer
added a commit
to JonathanBrouwer/rust
that referenced
this pull request
Aug 31, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
rust-bors Bot
pushed a commit
that referenced
this pull request
Aug 31, 2026
…uwer Rollup of 12 pull requests Successful merges: - #162045 (`rust-analyzer` subtree update) - #162077 (compiler-builtins subtree update) - #137720 (support `#[target_feature(enable = ...)]` on `#[naked]` functions) - #160534 (stabilize smart pointer map functions) - #160551 (mir_build: Clearly distinguish or/refutable/irrefutable patterns during match lowering) - #161929 (explicitly track inherent const generic args kind) - #162063 (Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux) - #161937 (A series of Polonius Alpha refactors) - #162014 (Move more `rustdoc-html` tests using `--test` into the right folder) - #162051 (`rustc_feature` cleanups) - #162055 (remove `_{style}` recovery for diagnostic structs) - #162075 (Move track_caller on closures gating to attribute parsing)
jhpratt
added a commit
to jhpratt/rust
that referenced
this pull request
Sep 1, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
jhpratt
added a commit
to jhpratt/rust
that referenced
this pull request
Sep 1, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
JonathanBrouwer
added a commit
to JonathanBrouwer/rust
that referenced
this pull request
Sep 1, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
JonathanBrouwer
added a commit
to JonathanBrouwer/rust
that referenced
this pull request
Sep 1, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
JonathanBrouwer
added a commit
to JonathanBrouwer/rust
that referenced
this pull request
Sep 1, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
JonathanBrouwer
added a commit
to JonathanBrouwer/rust
that referenced
this pull request
Sep 1, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
JonathanBrouwer
added a commit
to JonathanBrouwer/rust
that referenced
this pull request
Sep 1, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
JonathanBrouwer
added a commit
to JonathanBrouwer/rust
that referenced
this pull request
Sep 1, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
JonathanBrouwer
added a commit
to JonathanBrouwer/rust
that referenced
this pull request
Sep 1, 2026
Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
rust-bors Bot
pushed a commit
that referenced
this pull request
Sep 1, 2026
…uwer Rollup of 16 pull requests Successful merges: - #137720 (support `#[target_feature(enable = ...)]` on `#[naked]` functions) - #160534 (stabilize smart pointer map functions) - #160551 (mir_build: Clearly distinguish or/refutable/irrefutable patterns during match lowering) - #160989 (Make sin, cos, exp, exp2, log, log2, log10 generic) - #161861 (mir_build: Clarify parts of if-condition lowering) - #161929 (explicitly track inherent const generic args kind) - #162040 (bootstrap: stage0 to cbae9b4) - #162063 (Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux) - #161353 (Add test for parallel compiler reproducible build) - #161937 (A series of Polonius Alpha refactors) - #162051 (`rustc_feature` cleanups) - #162055 (remove `_{style}` recovery for diagnostic structs) - #162075 (Move track_caller on closures gating to attribute parsing) - #162079 (std: implement `File::fsync` for Hermit) - #162097 (Deduplicate `InstrumentFnAttr`) - #162115 (fix typo in feature documentation)
rust-bors Bot
pushed a commit
that referenced
this pull request
Sep 1, 2026
Rollup merge of #162063 - Mark-Simulacrum:aarch64-ec2, r=Kobzol Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux For dist-aarch64-linux (full): * GHA 8c takes 2h25m ($2.03/build) - current * c8g.8xl takes 1h20m ($1.69/build) * c9g.8xl takes 1h ($1.38/build) * c9g.4xl takes 1h10m ($0.81/build) * m9g.2xl takes 1h30m ($0.59/build) - selected Adds a dist-aarch64-linux-quick (default enabled in try builds): * c8g.8xl takes 50m ($1.059/build) * c9g.8xl takes 40m ($0.924/build) * c9g.4xl takes 47m ($0.543/build) - selected * m9g.2xl takes 64m ($0.417/build) For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down. For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient: * c8a.8xl takes 1h34m ($2.64/build) - current * c8a.4xl takes 1h45m ($1.51/build) - selected * m8a.2xl takes 2h10m ($1.05/build) I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore. Testing: * Regular try (no jobs selected) - https://github.com/rust-lang/rust/actions/runs/33395505441 * Runs both x86 and aarch64, both in quick variants * Try with jobs=dist-x86_64-linux*,dist-aarch64-linux* - https://github.com/rust-lang/rust/actions/runs/33394676009 * Runs both x86 and aarch64, quick and non-quick * This fixes a bug where dist-x86_64-linux-quick via explicit request used to not be a quick job AFAICT; nothing set DIST_TRY_BUILD on it. r? @Kobzol
Collaborator
|
Finished benchmarking commit (1e686b4): comparison URL. Overall result: no relevant changes - no action neededBenchmarking means the PR may be perf-sensitive. Consider adding rollup=never if this change is not fit for rolling up. @rustbot label: -S-waiting-on-perf -perf-regression Instruction countThis perf run didn't have relevant results for this metric. Max RSS (memory usage)This perf run didn't have relevant results for this metric. CyclesThis perf run didn't have relevant results for this metric. Binary sizeThis perf run didn't have relevant results for this metric. Bootstrap: missing data |
pull Bot
pushed a commit
to LeeeeeeM/miri
that referenced
this pull request
Sep 2, 2026
…uwer Rollup of 16 pull requests Successful merges: - rust-lang/rust#137720 (support `#[target_feature(enable = ...)]` on `#[naked]` functions) - rust-lang/rust#160534 (stabilize smart pointer map functions) - rust-lang/rust#160551 (mir_build: Clearly distinguish or/refutable/irrefutable patterns during match lowering) - rust-lang/rust#160989 (Make sin, cos, exp, exp2, log, log2, log10 generic) - rust-lang/rust#161861 (mir_build: Clarify parts of if-condition lowering) - rust-lang/rust#161929 (explicitly track inherent const generic args kind) - rust-lang/rust#162040 (bootstrap: stage0 to cbae9b4cae2b108f6a3d18cfe6075714bb739463) - rust-lang/rust#162063 (Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux) - rust-lang/rust#161353 (Add test for parallel compiler reproducible build) - rust-lang/rust#161937 (A series of Polonius Alpha refactors) - rust-lang/rust#162051 (`rustc_feature` cleanups) - rust-lang/rust#162055 (remove `_{style}` recovery for diagnostic structs) - rust-lang/rust#162075 (Move track_caller on closures gating to attribute parsing) - rust-lang/rust#162079 (std: implement `File::fsync` for Hermit) - rust-lang/rust#162097 (Deduplicate `InstrumentFnAttr`) - rust-lang/rust#162115 (fix typo in feature documentation)
asukaminato0721
pushed a commit
to asukaminato0721/rust-analyzer
that referenced
this pull request
Sep 2, 2026
…uwer Rollup of 16 pull requests Successful merges: - rust-lang/rust#137720 (support `#[target_feature(enable = ...)]` on `#[naked]` functions) - rust-lang/rust#160534 (stabilize smart pointer map functions) - rust-lang/rust#160551 (mir_build: Clearly distinguish or/refutable/irrefutable patterns during match lowering) - rust-lang/rust#160989 (Make sin, cos, exp, exp2, log, log2, log10 generic) - rust-lang/rust#161861 (mir_build: Clarify parts of if-condition lowering) - rust-lang/rust#161929 (explicitly track inherent const generic args kind) - rust-lang/rust#162040 (bootstrap: stage0 to cbae9b4cae2b108f6a3d18cfe6075714bb739463) - rust-lang/rust#162063 (Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux) - rust-lang/rust#161353 (Add test for parallel compiler reproducible build) - rust-lang/rust#161937 (A series of Polonius Alpha refactors) - rust-lang/rust#162051 (`rustc_feature` cleanups) - rust-lang/rust#162055 (remove `_{style}` recovery for diagnostic structs) - rust-lang/rust#162075 (Move track_caller on closures gating to attribute parsing) - rust-lang/rust#162079 (std: implement `File::fsync` for Hermit) - rust-lang/rust#162097 (Deduplicate `InstrumentFnAttr`) - rust-lang/rust#162115 (fix typo in feature documentation)
github-actions Bot
pushed a commit
to rust-lang/rustc-dev-guide
that referenced
this pull request
Sep 4, 2026
…uwer Rollup of 16 pull requests Successful merges: - rust-lang/rust#137720 (support `#[target_feature(enable = ...)]` on `#[naked]` functions) - rust-lang/rust#160534 (stabilize smart pointer map functions) - rust-lang/rust#160551 (mir_build: Clearly distinguish or/refutable/irrefutable patterns during match lowering) - rust-lang/rust#160989 (Make sin, cos, exp, exp2, log, log2, log10 generic) - rust-lang/rust#161861 (mir_build: Clarify parts of if-condition lowering) - rust-lang/rust#161929 (explicitly track inherent const generic args kind) - rust-lang/rust#162040 (bootstrap: stage0 to cbae9b4cae2b108f6a3d18cfe6075714bb739463) - rust-lang/rust#162063 (Switch dist-aarch64-linux to EC2 and update dist-x86_64-linux) - rust-lang/rust#161353 (Add test for parallel compiler reproducible build) - rust-lang/rust#161937 (A series of Polonius Alpha refactors) - rust-lang/rust#162051 (`rustc_feature` cleanups) - rust-lang/rust#162055 (remove `_{style}` recovery for diagnostic structs) - rust-lang/rust#162075 (Move track_caller on closures gating to attribute parsing) - rust-lang/rust#162079 (std: implement `File::fsync` for Hermit) - rust-lang/rust#162097 (Deduplicate `InstrumentFnAttr`) - rust-lang/rust#162115 (fix typo in feature documentation)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
For dist-aarch64-linux (full):
Adds a dist-aarch64-linux-quick (default enabled in try builds):
For now I've chosen a balance between cost and speed (c9g.4xl). This is still faster than the x86 build so shouldn't slow us down.
For dist-x86_64-linux-full this reduces core count to be a bit more cost efficient:
I'll re-benchmark dist-x86_64-linux-quick in a future PR, for now it will stay on c8a.8xl. This PR also drops codebuild configuration (but not yet cleaning up various related pieces that are more tied into our CI) since it doesn't seem relevant anymore.
Testing:
r? @Kobzol