Repeated error message caused by omitting "--via-ir"

[haoyang9804]

Description

The reproducible code (I guess the error is not related to the code but only the compilation flags):

contract contract0 {
  event event1(int32 var2);

  struct struct3 {
    string var4;
  }

  struct struct5 {
    struct3[] array6;
    int32 var8;
  }

  struct3 internal struct_instance9;
  string internal var10;

  constructor(string[2] memory array11) {
    emit contract0.event1(int32(-273486664));
  }

  function func13() internal pure returns (int32 var14, int32 var15) {
    struct5 memory struct_instance22;
    struct_instance22 = struct_instance22;
    struct5 memory struct_instance23;
    struct_instance23 = struct_instance23;
    (struct_instance23.var8) ^ struct_instance22.var8;
    struct5 memory struct_instance24;
    struct_instance24 = struct_instance24;
    struct5 memory struct_instance25;
    struct_instance25 = struct_instance25;
    return (struct_instance24.var8, ~struct_instance25.var8);
  }

  function func16(mapping(int32 => int32) storage mapping17) internal returns (string memory var20, struct5 memory struct_instance21) {
    var20 = var20;
    struct_instance21 = struct_instance21;
    bool var26;
    while ((!var26)) {}
  }
}

contract contract27 {
  error error30(uint128[] array31);

  event event28(contract0.struct5 struct_instance29);

  struct struct33 {
    contract0.struct3 struct_instance34;
    mapping(uint128 => string) mapping35;
  }

  struct struct38 {
    string var39;
    mapping(int32 => struct33) mapping40;
  }

  mapping(int32 => struct33) internal mapping47;
  bool internal var43;

  constructor() {
    contract27.struct38 storage struct_instance46;
    struct_instance46 = struct_instance46;
    var43 ? (mapping47) : struct_instance46.mapping40;
  }

  function func44() internal returns (string memory var45) {
    var45 = var45;
    contract0.struct5 memory struct_instance50;
    struct_instance50 = struct_instance50;
    (struct_instance50.var8)++;
  }
}

The command:

../AFLs/new-solidity/build/solc/solc generated_programs/program_2024-12-16_11:25:29_43.sol --bin-runtime --yul-optimizations amexguhEiFCUrOdI --optimize-yul --optimize --model-checker-ext-calls trusted --model-checker-targets balance --model-checker-invariants reentrancy --model-checker-div-mod-no-slacks --model-checker-bmc-loop-iterations 10 --model-checker-show-proved-safe --model-checker-show-unproved --model-checker-engine bmc

Error message

Error: Stack too deep. Try compiling with `--via-ir` (cli) or the equivalent `viaIR: true` (standard JSON) while enabling the optimizer. Otherwise, try removing local variables. When compiling inline assembly: Variable dataEnd is 1 slot(s) too deep inside the stack. Stack too deep. Try compiling with `--via-ir` (cli) or the equivalent `viaIR: true` (standard JSON) while enabling the optimizer. Otherwise, try removing local variables.

The Try compiling with --via-ir message is thrown by twice.

Environment

• Compiler version:0.8.29-develop.2024.11.30+commit.b4ecc58b.Linux.g++

[cameel]

The Try compiling with --via-ir message is thrown by twice.

Regarding this bit, this is caused by AsmCodeGen always prepending a canned message when reporting StackTooDeep errors from the lower layer, regardless of what that intercepted message says:

solidity/libyul/backends/evm/AsmCodeGen.cpp

Lines 62 to 68 in 10fbfca

	if (!transform.stackErrors().empty())
	assertThrow(
	false,
	langutil::StackTooDeepError,
	util::stackTooDeepString + " When compiling inline assembly" +
	(transform.stackErrors().front().comment() ? ": " + *transform.stackErrors().front().comment() : ".")
	);

We should perhaps remove that bit from that lower layer message, after making sure that it always gets intercepted.