Cleanup - test_interpreter_and_jit!() macros
#13
Conversation
Lichtso
force-pushed
the
refactor/test_utils_macros
branch
from
2a85b2b to
3665510
Compare
| @@ -3535,46 +3535,6 @@ fn test_invalid_call_imm() { | |||
| ); | |||
| } | |||
|
|
|||
| #[test] | |||
| #[should_panic(expected = "Invalid syscall should have been detected in the verifier.")] | |||
| fn test_invalid_exit_or_return() { | |||
There was a problem hiding this comment.
The idea of this test was to ensure correct behavior if execution reaches an invalid sys call not detected in the verifier.
If you think that will never happen, we can simply live without this test.
There was a problem hiding this comment.
Yes, I moved it to the verifier tests as these are execution only tests. And things which are supposed to fail verification have no defined execution behavior.
| @@ -3501,40 +3501,6 @@ fn test_total_chaos() { | |||
| } | |||
| } | |||
|
|
|||
| #[test] | |||
| fn test_invalid_call_imm() { | |||
There was a problem hiding this comment.
This test was here to make sure one previous bug I introduced in the codebase doesn't exist anymore. The bug was dispatching syscalls with call imm.
There was a problem hiding this comment.
I added test_call_imm_does_not_dispatch_syscalls()
test_utils/src/lib.rs
Outdated
| if !$override_budget { | ||
| const INSTRUCTION_METER_BUDGET: u64 = 1024; | ||
| context_object.remaining = INSTRUCTION_METER_BUDGET; | ||
| } |
There was a problem hiding this comment.
override_budget == true will NOT override the budget.
override_budget == false WILL override the budget.
There was a problem hiding this comment.
Ah yes, we should invert the flag and its use sites.
Lichtso
force-pushed
the
refactor/test_utils_macros
branch
from
3665510 to
0fe538f
Compare
tests/verifier.rs
Outdated
| @@ -374,10 +374,27 @@ fn test_verifier_err_unknown_opcode() { | |||
| } | |||
|
|
|||
| #[test] | |||
| #[should_panic(expected = "InvalidFunction(1811268607)")] | |||
| #[should_panic(expected = "UnknownOpCode(6, 0)")] | |||
There was a problem hiding this comment.
Are we testing the invalid exit? Shouldn't we be throwing an error for 0x95?
There was a problem hiding this comment.
test_verifier_err_invalid_return appears to have the same issue.
There was a problem hiding this comment.
You are right, will rewrite these two tests.
tests/verifier.rs
Outdated
| #[should_panic(expected = "InvalidFunction(3)")] | ||
| fn test_verifier_unknown_sycall() { | ||
| let prog = &[ | ||
| 0x85, 0x00, 0x00, 0x00, 0xfe, 0xc3, 0xf5, 0x6b, // call 0x6bf5c3fe | ||
| 0x85, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, // call_imm 2 |
There was a problem hiding this comment.
0x85 is not a sys call in V3.
Lichtso
force-pushed
the
refactor/test_utils_macros
branch
from
0fe538f to
fc1e828
Compare
tests/verifier.rs
Outdated
| fn test_verifier_unknown_sycall() { | ||
| let prog = &[ | ||
| 0x85, 0x00, 0x00, 0x00, 0xfe, 0xc3, 0xf5, 0x6b, // call 0x6bf5c3fe | ||
| 0x95, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, // call_imm 2 |
There was a problem hiding this comment.
| 0x95, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, // call_imm 2 | |
| 0x95, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, // syscall 2 |
…terpreter_and_jit_elf and test_syscall_asm into test_utils crate.
… with and without override_budget.
Lichtso
force-pushed
the
refactor/test_utils_macros
branch
from
fc1e828 to
f5e5e30
Compare
No description provided.