Remove always from inline to allow compiler decide that to do

[xgreenx]

It's funny, but removing always with combination use-ink/cargo-contract#358, reduces the size of Erc20 contract in master from 31.4K to 27.4K

polkadot address: 1nNaTpU9GHFvF7ZrSMu2CudQjXftR8Aqx58oMDgcuoH8dKe

[HCastano]

This looks promising! We should wait until the ink-waterfall pipeline is fixed though, since there we can see how this PR changes the sizes of all the example contracts instead of just the ERC-20 one.

[codecov-commenter]

Codecov Report

Merging #1012 (843b06d) into master (c7cc828) will decrease coverage by 0.02%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #1012      +/-   ##
==========================================
- Coverage   78.86%   78.84%   -0.03%     
==========================================
  Files         245      245              
  Lines        9246     9246              
==========================================
- Hits         7292     7290       -2     
- Misses       1954     1956       +2     

Impacted Files	Coverage Δ
crates/env/src/chain_extension.rs	0.00% <ø> (ø)
...odegen/src/generator/as_dependency/call_builder.rs	100.00% <ø> (ø)
...ng/codegen/src/generator/trait_def/call_builder.rs	100.00% <ø> (ø)
crates/lang/src/codegen/dispatch/execution.rs	0.00% <ø> (ø)
crates/lang/src/result_info.rs	100.00% <ø> (ø)
crates/primitives/src/key.rs	93.42% <ø> (ø)
crates/storage/src/collections/stash/storage.rs	60.97% <ø> (ø)
crates/storage/src/memory.rs	100.00% <ø> (ø)
crates/storage/src/traits/impls/prims.rs	96.92% <ø> (ø)
...ates/storage/src/collections/hashmap/fuzz_tests.rs	97.87% <0.00%> (-2.13%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update c7cc828...843b06d. Read the comment docs.

[cmichi]

Very nice discovery!

[HCastano]

So here are the changes in size of our example contracts compared to master. Note that all sizes are in KB, and - means that it is smaller with this PR applied.

contract-terminate,0.0,0.0
rand-extension,-0.055000305,-0.12400007
incrementer,0.035999298,-0.06700015
erc1155,-63.791008,-62.012
trait-erc20,-2.6740036,-2.7460003
dns,-0.10300064,-0.22099876
delegator,0.38000107,0.15100002
multisig,-14.578995,-13.453999
subber,0.20700073,0.14900017
flipper,0.069000244,-0.03699994
erc20,-2.675003,-2.7460003
trait-flipper,0.024999619,-0.06700003
contract-transfer,0.0,0.0
trait-incrementer,-0.20399857,-0.32500005
erc721,-3.8440018,-3.9720001
adder,0.20700073,0.14900017
accumulator,0.20699883,0.14900017

Notice the dramatic size reduction to the ERC-1155 example 🤯. We should really do some digging and figure out why this "simple" change has produced such a great improvement there.

[cmichi]

@xgreenx We want to give you a tip for all the great stuff you have contributed now and in the past. We're happy to have you as part of our ecosystem!

Could you edit your description of this PR and add either one of those?

polkadot address: <SS58 Address>

or

kusama address: <SS58 Address>

[xgreenx]

@xgreenx We want to give you a tip for all the great stuff you have contributed now and in the past. We're happy to have you as part of our ecosystem!

Glad to hear it=) Thank you, it is part of our grant, so you don't need to worry about it(=

So here are the changes in size of our example contracts compared to master. Note that all sizes are in KB, and - means that it is smaller with this PR applied.

Hmm, contracts that use builder have a bad impact(delegator,0.38000107,0.15100002), maybe we should keep inlining for them. I will try it.

[xgreenx]

Hmm, contracts that use builder have a bad impact(delegator,0.38000107,0.15100002), maybe we should keep inlining for them. I will try it.

I tried to return #[inline(always)] back in some places to reduce the size of the delegator back. Seems that use-ink/cargo-contract#358 should cover that case. Because the size is the same with #[inline(always)] and #[inline] if I build with use-ink/cargo-contract#358. So I think we can merge it=)

[cmichi]

/tip medium

[substrate-tip-bot]

A medium tip was successfully submitted for xgreenx (1nNaTpU9GHFvF7ZrSMu2CudQjXftR8Aqx58oMDgcuoH8dKe on polkadot).

https://polkadot.js.org/apps/#/treasury/tips