chore(test): Remove duplicate brillig tests

[aakoshh]

Description

Problem*

Resolves #6512

Summary*

Removed duplicate test_programs/execution_success/brillig_<test> from where an equivalent test_programs/execution_success/<test> exists, listed by:

for dir in (ls | grep -v brillig_); if [ -d "brillig_$dir" ]; echo duplicate $dir; end; end;

The brillig_ version of following has been removed from execution_success:

• blake3 -> duplicate; unconstrained main
• references -> duplicate; unconstrained main
• slices -> duplicate; unconstrained main
• assert -> ACIR calls unconstrained; moved conditional
• hash_to_field -> ACIR to unconstrained, otherwise identical
• loop -> ACIR to unconstrained, otherwise identical
• array_to_slice -> not a duplicate, but seems redundant
• ecdsa_secp256k1 -> ACIR to unconstrained, but very similar
• ecdsa_secp256r1 -> ACIR to unconstrained, but very similar
• sha256 -> ACIR to unconstrained, subset

Additional Context

I wasn't sure if the ones where ACIR calls unconstrained are considered duplicates, but hash_to_field was such an example and it's listed in the ticket, so I removed all of them.

Note that there is execution_failure/brillig_assert_msg_runtime which seems like it could be removed in favour of execution_failure/assert_msg_runtime, however currently these tests are only executed with the default settings which means the --force-brillig option isn't applied to them; if we want to test brillig behaviour we need to either run it both ways or have two tests.

Documentation*

Check one:

• No documentation needed.
• Documentation included in this PR.
• [For Experimental Features] Documentation to be submitted in a separate PR.

PR Checklist*

• I have tested the changes locally.
• I have formatted the changes with Prettier and/or cargo fmt on default settings.

[github-actions]

Changes to Brillig bytecode sizes

Generated at commit: 3e483953e9feb1233204fd5dca6d4257fb069928, compared to commit: 1cad7c887893ebfb5de57a71d7965c8b88158a14

🧾 Summary (10% most significant diffs)

Program	Brillig opcodes (+/-)	%
loop	+27 ❌	+36.00%
sha256	+608 ❌	+29.57%

---

Full diff report 👇

Program	Brillig opcodes (+/-)	%
loop	102 (+27)	+36.00%
sha256	2,664 (+608)	+29.57%

[github-actions]

Changes to circuit sizes

Generated at commit: 3e483953e9feb1233204fd5dca6d4257fb069928, compared to commit: 1cad7c887893ebfb5de57a71d7965c8b88158a14

🧾 Summary (10% most significant diffs)

Program	ACIR opcodes (+/-)	%	Circuit size (+/-)	%
sha256	+157 ❌	+9.98%	+4,025 ❌	+19.24%

---

Full diff report 👇

Program	ACIR opcodes (+/-)	%	Circuit size (+/-)	%
sha256	1,730 (+157)	+9.98%	24,946 (+4,025)	+19.24%

[github-actions]

Changes to number of Brillig opcodes executed

Generated at commit: 3e483953e9feb1233204fd5dca6d4257fb069928, compared to commit: 1cad7c887893ebfb5de57a71d7965c8b88158a14

🧾 Summary (10% most significant diffs)

Program	Brillig opcodes (+/-)	%
loop	+59 ❌	+43.07%
sha256	+3,534 ❌	+32.58%

---

Full diff report 👇

Program	Brillig opcodes (+/-)	%
loop	196 (+59)	+43.07%
sha256	14,381 (+3,534)	+32.58%

[jfecher]

LGTM