feat(wasm-builder): Set optimisation params

[ukint-vs]

Resolves # .

• update fungible-token changes from refactor: avoid unnecessary syscalls gear-foundation/dapps-fungible-token#52.
• set opt-level = z in toml
• set wasm-opt level to 4 and shrink 2
• set codegen-units = 1
• set lto = "fat", which is the same as true

These changes reduce the size of wasm artifacts and speed up programs.

stress_transfer benchmark (updated):

before: Gas burned for one transfer operation = 0.9770345439034853 * 10^9. Calculated as geometric mean from 100 transfer operations.
after:  Gas burned for one transfer operation = 0.9441607030742862 * 10^9. Calculated as geometric mean from 100 transfer operations.

Artifact size:

fungible-token upload fees

before:
4.254320105 unit
after:
3.609106373 unit

Before	After	program
54 Ki	46 Ki	demo_async.opt.wasm
35 Ki	32 Ki	demo_async_custom_entry.opt.wasm
53 Ki	44 Ki	demo_async_init.opt.wasm
50 Ki	42 Ki	demo_async_recursion.opt.wasm
51 Ki	42 Ki	demo_async_signal_entry.opt.wasm
55 Ki	45 Ki	demo_async_tester.opt.wasm
19 Ki	18 Ki	demo_btree.opt.wasm
18 Ki	14 Ki	demo_calc_hash_in_one_block.opt.wasm
37 Ki	21 Ki	demo_calc_hash_over_blocks.opt.wasm
51 Ki	43 Ki	demo_compose.opt.wasm
48 Ki	30 Ki	demo_constructor.opt.wasm
56 Ki	48 Ki	demo_distributor.opt.wasm
41 Ki	28 Ki	demo_fungible_token.opt.wasm
58 Ki	48 Ki	demo_futures_unordered.opt.wasm
51 Ki	42 Ki	demo_incomplete_async_payloads.opt.wasm
50 Ki	40 Ki	demo_init_fail_sender.opt.wasm
48 Ki	39 Ki	demo_init_wait_reply_exit.opt.wasm
51 Ki	43 Ki	demo_ncompose.opt.wasm
17 Ki	16 Ki	demo_node.opt.wasm
5.9 Ki	6.7 Ki	demo_piggy_bank.opt.wasm
5.9 Ki	5.9 Ki	demo_ping.opt.wasm
71 Ki	62 Ki	demo_rwlock.opt.wasm
13 Ki	14 Ki	demo_send_from_reservation.opt.wasm
27 Ki	27 Ki	demo_signal_entry.opt.wasm
27 Ki	13 Ki	demo_stack_allocations.opt.wasm
50 Ki	41 Ki	demo_sync_duplicate.opt.wasm
56 Ki	46 Ki	demo_wait_timeout.opt.wasm
72 Ki	60 Ki	demo_waiter.opt.wasm
49 Ki	40 Ki	demo_waiting_proxy.opt.wasm
61 Ki	58 Ki	test_syscalls.opt.wasm

@gear-tech/dev

[breathx]

👍🏻
It'd be good compare upload program cost for some of them

[clearloop]

what about going with O4 plus shrink-level-2?

for -Oz vs -O4, the difference is optimization level 2 ( + shrink-level-2) vs optimization level 4, see
https://docs.rs/wasm-opt/0.113.0/wasm_opt/struct.OptimizationOptions.html#method.new_optimize_for_size_aggressively for structured docs

---

btw, super curious why tests::locking_gas_for_waitlist works in this PR but failed in all of my PRs xd

[ukint-vs]

what about going with O4 plus shrink-level-2?

for -Oz vs -O4, the difference is optimization level 2 ( + shrink-level-2) vs optimization level 4, see docs.rs/wasm-opt/0.113.0/wasm_opt/struct.OptimizationOptions.html#method.new_optimize_for_size_aggressively for structured docs

in some cases it is faster, but the size is similar

[ukint-vs]

👍🏻 It'd be good compare upload program cost for some of them

Added to description