Bump subxt to 0.50 by RomarQ · Pull Request #12096 · paritytech/polkadot-sdk

RomarQ · 2026-05-18T15:14:26Z

Follow-up to #11422.

Bumps the workspace subxt, subxt-metadata, and subxt-signer dependencies from 0.44 to 0.50.

subxt 0.50 introduces a block-anchored client API and splits subxt::Error into per-operation sub-errors. Affected crates were updated to the new API; the JSON-RPC and CLI surfaces are unchanged.

- Migrate .number()/.hash() to .block_number()/.block_hash() on at-block handles - Update MockBlockInfo number field literals to u64 (SubstrateBlockNumber widened) - tx_client needs mut for create_signed (now takes &mut self) - node_client.tx() is async in 0.50, requires .await before chaining - find_first returns Option<Result<...>> not Result<Option<...>> - system::calls::Remark field renamed from value to remark

subxt 0.50.1's SubstrateConfigBuilder::set_genesis_hash stores the value but build() drops it, and the Config trait's genesis_hash() defaults to None. As a result, create_signable_offline fails at runtime with GenesisHashNotProvided. PolkadotConfig has the same issue since it delegates to SubstrateConfig. Wrap SubstrateConfig in a local OfflineConfig that holds the genesis hash and returns it from genesis_hash(). The remaining Config methods delegate to the inner SubstrateConfig.

RomarQ · 2026-05-18T20:51:44Z

Blocked by paritytech/subxt#2236. Will re-open the PR once a new version of subxt gets released.

acatangiu · 2026-06-22T11:53:29Z

Blocked by paritytech/subxt#2236. Will re-open the PR once a new version of subxt gets released.

@RomarQ Looks like the fix in subxt should now be available?

RomarQ · 2026-06-22T12:04:10Z

Blocked by paritytech/subxt#2236. Will re-open the PR once a new version of subxt gets released.

@RomarQ Looks like the fix in subxt should now be available?

Waiting for a new version of subxt to be released. But can fix the conflicts in the meanwhile.

…t-0.50 # Conflicts: # cumulus/zombienet/zombienet-sdk/Cargo.toml # substrate/client/statement-store/statement-latency-bench/Cargo.toml # substrate/frame/revive/rpc/src/client/runtime_api.rs # substrate/frame/revive/rpc/src/lib.rs

…t-0.50 # Conflicts: # Cargo.lock # substrate/frame/revive/rpc/src/client.rs # substrate/frame/revive/rpc/src/client/runtime_api.rs # substrate/frame/revive/rpc/src/lib.rs # substrate/frame/revive/rpc/src/receipt_provider.rs

socket-security · 2026-07-11T12:15:51Z

Review the following changes in direct dependencies. Learn more about Socket for GitHub.

Diff	Package	Supply Chain Security	Vulnerability	Quality	Maintenance	License
	cargo/subxt@0.50.0 ⏵ 0.50.2
	cargo/subxt-metadata@0.50.0 ⏵ 0.50.2
	cargo/subxt-signer@0.50.0 ⏵ 0.50.2
	cargo/async-channel@2.3.0 ⏵ 2.5.0

View full report

subxt 0.50 removed OnlineClient::blocks(); fetch the receipt's block via at_block(number) instead of blocks().at_latest(). Header access is now the async block_header(), and block extrinsics come from extrinsics().fetch(). iter() yields Results, so propagate the error before decoding into OpaqueExtrinsic.

The workspace subxt is now 0.50, but zombienet-sdk still depends on subxt 0.44 and its client API (node.wait_client, OnlineClient) is typed against that version. The zombienet test crates are coupled to that API, so they cannot use the 0.50 workspace subxt. Route them through zombienet_sdk::subxt / zombienet_sdk::subxt_signer (the 0.44 re-export) and drop their direct subxt deps, so no subxt 0.44 pin appears in any Cargo.toml; it stays only as a transitive dependency of zombienet-sdk. - substrate, polkadot: drop direct subxt/subxt-signer deps and use the zombienet_sdk re-exports. - polkadot weights.rs: point the subxt codegen macro at ::zombienet_sdk::subxt::ext::subxt_core. - cumulus runtime_upgrade.rs: build the big-message client via jsonrpsee (zombienet_sdk::subxt::ext::jsonrpsee) rather than the reconnecting client, which is not enabled in zombienet-sdk's default subxt features.

- Use a `subxt` module alias (`use zombienet_sdk::subxt::{self, ...}`) in block_building.rs and beefy_and_mmr.rs so call sites keep the bare `subxt::` paths instead of repeating `zombienet_sdk::subxt::` at each use. - Drop the now-dead `reconnecting-rpc-client` feature from the cumulus subxt dependency; nothing uses the reconnecting client after runtime_upgrade.rs moved to a plain jsonrpsee ws client (`native` is still needed for the 0.50 statement-store client). - Apply the `crate = "::zombienet_sdk::subxt::ext::subxt_core"` codegen override to the two subxt macros in coretime_revenue.rs that were missed; they only compile under `--features zombie-metadata,zombie-ci` (what CI uses), so the earlier zombie-ci-only check did not catch them.

check-semver detected larger changes than the prdoc declared: - frame-benchmarking-cli is a major bump: DynamicRemarkBuilder lost its generic Config parameter (now uses an internal wrapper config). - polkadot-omni-node-lib and sc-statement-store are minor, not patch.

check-semver flags pallet-revive as changed-but-not-listed: it depends on subxt-signer, which this PR bumps 0.44 -> 0.50. The dependency is behind a non-default feature, so there is no public API change (detected as "Dependency changed", not a major/minor API change); declare a minor bump.

This reverts commit 7211885.

ggwpez

Some grumbles in the pre-existing code, feel free to ignore.

…string - extrinsic.rs: use `find_first().transpose()?` to match the previous `find_first()?` behaviour (propagate decode errors, ignore a missing event). - remark-extrinsic.rs: propagate decode errors instead of silently dropping them via `filter_map(Result::ok)`; the previous `unwrap` was intentional. - runtime_api.rs: document that the "is not found" match originates from sc-executor ("Exported method <name> is not found").

RomarQ · 2026-07-14T08:15:11Z

@bkchr @acatangiu @ggwpez

Can we put this in the merge queue?

marian-radu

Revive changes look good to me, thank you!

…t-0.50 # Conflicts: # Cargo.lock

Address review: match on decode_call_data_fields_as and warn when an extrinsic is a revive EthTransact call but its fields fail to decode, so the dropped transaction is visible in logs instead of vanishing.

…e/bump-subxt-0.50

… arm Address review: split the catch-all match arm so a failed ContractEmitted field decode logs the underlying error, and document that None is unreachable because is_event() already confirmed the variant.

Address review: inspect_err on at_block_hash_and_number so a failure to resolve the block for an eth_block_hash query is logged at debug before being swallowed by .ok()?.

RomarQ · 2026-07-15T09:47:59Z

Revive changes look good to me, thank you!

Addressed the nits 👍

Can you add it to the merge queue?

PR paritytech#12096 bumped the workspace subxt/subxt-signer pin to 0.50 and moved the statement-store and zombienet-sdk crates onto workspace = true, so the ignore list (which existed to let them pin 0.50 while the workspace was on 0.44) no longer suppresses anything. Two entries also referenced the removed statement-latency-bench crate. The inheritance check stays green without it.

RomarQ and others added 7 commits May 18, 2026 16:13

Bump subxt to 0.50

9b7aba3

Add prdoc for subxt 0.50 bump

6a4cbe6

Rename prdoc to pr_12096.prdoc

55f026f

Use SubstrateBlockNumber instead of u64 in fee_history

2d0a61d

Use SubstrateBlockNumber in fetch_eth_block_hash callbacks

938334b

Merge branch 'master' into chore/bump-subxt-0.50

b2c4594

cargo fmt

554e554

RomarQ marked this pull request as ready for review May 18, 2026 16:51

RomarQ requested a review from a team as a code owner May 18, 2026 16:51

RomarQ added 3 commits May 18, 2026 20:37

Log inner subxt error from create_signable_offline

0dc3747

RomarQ marked this pull request as draft May 18, 2026 20:35

RomarQ closed this May 18, 2026

acatangiu reopened this Jun 22, 2026

RomarQ added 2 commits June 22, 2026 13:25

RomarQ marked this pull request as ready for review July 11, 2026 12:29

RomarQ and others added 6 commits July 11, 2026 15:31

Merge branch 'master' into chore/bump-subxt-0.50

cc5e6e7

bkchr added T17-primitives Changes to primitives that are not covered by any other label. and removed T17-primitives Changes to primitives that are not covered by any other label. labels Jul 13, 2026

RomarQ added 2 commits July 13, 2026 15:22

Remove dead SubstrateBlockHeader alias and unused Config import

7211885

Revert "Remove dead SubstrateBlockHeader alias and unused Config import"

6913eb8

This reverts commit 7211885.

ggwpez approved these changes Jul 13, 2026

View reviewed changes

Comment thread substrate/frame/revive/rpc/examples/extrinsic.rs Outdated

Comment thread substrate/frame/revive/rpc/examples/remark-extrinsic.rs Outdated

Comment thread substrate/frame/revive/rpc/src/client/runtime_api.rs

Comment thread substrate/frame/revive/rpc/src/client/runtime_api.rs

RomarQ and others added 3 commits July 13, 2026 17:29

Merge branch 'master' into chore/bump-subxt-0.50

b977336

Merge branch 'master' into chore/bump-subxt-0.50

d2882a3

bkchr approved these changes Jul 13, 2026

View reviewed changes

Merge branch 'master' into chore/bump-subxt-0.50

f784a99

bkchr enabled auto-merge July 14, 2026 15:56

marian-radu reviewed Jul 15, 2026

View reviewed changes

Comment thread substrate/frame/revive/rpc/src/receipt_extractor.rs Outdated

marian-radu reviewed Jul 15, 2026

View reviewed changes

Comment thread substrate/frame/revive/rpc/src/receipt_extractor.rs Outdated

marian-radu reviewed Jul 15, 2026

View reviewed changes

Comment thread substrate/frame/revive/rpc/src/receipt_extractor.rs Outdated

marian-radu approved these changes Jul 15, 2026

View reviewed changes

RomarQ added 4 commits July 15, 2026 10:36

Merge remote-tracking branch 'paritytech/master' into chore/bump-subx…

8a36c11

…t-0.50 # Conflicts: # Cargo.lock

Log decode failures instead of silently dropping EthTransact extrinsics

48f6d4c

Address review: match on decode_call_data_fields_as and warn when an extrinsic is a revive EthTransact call but its fields fail to decode, so the dropped transaction is visible in logs instead of vanishing.

Merge remote-tracking branch 'origin/chore/bump-subxt-0.50' into chor…

8b15337

…e/bump-subxt-0.50

Surface decode error and document unreachable None in ContractEmitted…

88555ae

… arm Address review: split the catch-all match arm so a failed ContractEmitted field decode logs the underlying error, and document that None is unreachable because is_event() already confirmed the variant.

auto-merge was automatically disabled July 15, 2026 09:44
Head branch was pushed to by a user without write access

Log block-resolution failures in eth_block_hash fetch

a647003

Address review: inspect_err on at_block_hash_and_number so a failure to resolve the block for an eth_block_hash query is logged at debug before being swallowed by .ok()?.

marian-radu enabled auto-merge July 15, 2026 09:55

marian-radu added this pull request to the merge queue Jul 15, 2026

Merged via the queue into paritytech:master with commit be832f6 Jul 15, 2026
404 of 412 checks passed

bkontur mentioned this pull request Jul 15, 2026

Bump subxt 0.44.3 -> 0.50.2 across all chain-facing crates paritytech/web3-storage#288

Merged

RomarQ mentioned this pull request Jul 15, 2026

CI: check workspace dependency inheritance #11422

Merged

ilchu mentioned this pull request Jul 15, 2026

Make provider coordinators event-driven off a single chain connection paritytech/web3-storage#291

Open

RomarQ deleted the chore/bump-subxt-0.50 branch July 16, 2026 16:50

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Bump subxt to 0.50#12096

Bump subxt to 0.50#12096
marian-radu merged 30 commits into
paritytech:masterfrom
RomarQ:chore/bump-subxt-0.50

RomarQ commented May 18, 2026 •

edited

Loading

Uh oh!

RomarQ commented May 18, 2026

Uh oh!

acatangiu commented Jun 22, 2026 •

edited

Loading

Uh oh!

RomarQ commented Jun 22, 2026

Uh oh!

socket-security Bot commented Jul 11, 2026 •

edited

Loading

Uh oh!

ggwpez left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

RomarQ commented Jul 14, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

marian-radu left a comment

Uh oh!

RomarQ commented Jul 15, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

Uh oh!

Conversation

RomarQ commented May 18, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

RomarQ commented May 18, 2026

Uh oh!

acatangiu commented Jun 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

RomarQ commented Jun 22, 2026

Uh oh!

socket-security Bot commented Jul 11, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ggwpez left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

RomarQ commented Jul 14, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

marian-radu left a comment

Choose a reason for hiding this comment

Uh oh!

RomarQ commented Jul 15, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

RomarQ commented May 18, 2026 •

edited

Loading

acatangiu commented Jun 22, 2026 •

edited

Loading

socket-security Bot commented Jul 11, 2026 •

edited

Loading