wip: fix Err syscall wrongly loaded

[edg-l]

Still havent found the fix but looking into it

while doing lambdaclass/starknet_in_rust#1102 i found a bug, the felt error was not there and the deserialization crashed (its now fixed with a deserialization error, but it should find the felt because its passed)

When a syscall returns a Err (that always contains a vec of felts) it wont see the vec of felts but they are there
I think the problem here is how we handle the return syscall payload in the starknet libfuncs, when the return type is () we use a hack to get the enum type have 2 variants (but we cant make a concretetypeid from nowhere so we copy the variant from the other err branch)
We also build the ok and err payload always regardless of whether the enum tag is 0 or 1 for each variant, i think the problem lies there (edited)

[github-actions]

Benchmarking results

Benchmark for program factorial_1M

Open benchmarks

Command	Mean [s]	Min [s]	Max [s]	Relative
Cairo-vm (Rust, Cairo 1)	15.848 ± 0.115	15.721	16.046	59.08 ± 0.43
cairo-native (JIT MLIR ORC Engine)	1.215 ± 0.007	1.205	1.229	4.53 ± 0.03
cairo-native (AOT Native binary)	0.293 ± 0.002	0.292	0.298	1.09 ± 0.01
cairo-native (AOT Native binary with host CPU features, march=native)	0.268 ± 0.000	0.268	0.269	1.00

Benchmark for program fib_2M

Open benchmarks

Command	Mean [s]	Min [s]	Max [s]	Relative
Cairo-vm (Rust, Cairo 1)	28.203 ± 0.273	28.019	28.754	728.81 ± 7.22
cairo-native (JIT MLIR ORC Engine)	1.215 ± 0.007	1.201	1.225	31.39 ± 0.19
cairo-native (AOT Native binary)	0.039 ± 0.000	0.038	0.039	1.00
cairo-native (AOT Native binary with host CPU features, march=native)	0.039 ± 0.000	0.039	0.040	1.01 ± 0.00

Benchmark for program map

Open benchmarks

Command	Mean [s]	Min [s]	Max [s]	Relative
Cairo-vm (Rust, Cairo 1)	2.393 ± 0.009	2.378	2.405	28.20 ± 0.11
cairo-native (JIT MLIR ORC Engine)	1.228 ± 0.005	1.221	1.235	14.47 ± 0.06
cairo-native (AOT Native binary)	0.085 ± 0.000	0.085	0.085	1.00
cairo-native (AOT Native binary with host CPU features, march=native)	0.087 ± 0.000	0.087	0.087	1.02 ± 0.00