Microsoft MVP성태의 닷넷 이야기
글쓴 사람
정성태 (techsharer at outlook.com)
홈페이지
첨부 파일
 
(연관된 글이 1개 있습니다.)

ilasm.exe 컴파일 오류 - error : syntax error at token '-' in ... -inf

간단하게 다음의 예제 코드를,

Console.WriteLine(GetResult());

double GetResult()
{
    return double.NegativeInfinity;
}

빌드해 생성한 DLL을 ildasm.exe로 IL 코드로 변경합니다.

C:\temp\ConsoleApp1\ConsoleApp1\bin\Debug\net7.0> ildasm ConsoleApp1.dll /out=test.il

그리고 이렇게 생성한 IL 코드를 다시 컴파일하면,

C:\temp\ConsoleApp1\ConsoleApp1\bin\Debug\net7.0> ilasm test.il /out=test.dll

Microsoft (R) .NET Framework IL Assembler.  Version 4.8.9105.0
Copyright (c) Microsoft Corporation.  All rights reserved.
Assembling 'test.il'  to EXE --> 'test.dll'
Source file is ANSI

Assembled method Microsoft.CodeAnalysis.EmbeddedAttribute::.ctor
Assembled method System.Runtime.CompilerServices.RefSafetyRulesAttribute::.ctor
Assembled method Program::<Main>$
Assembled method Program::.ctor
test.il(131) : error : syntax error at token '-' in:     IL_0001:  ldc.r8     -inf

***** FAILURE *****

오류를 재현할 수 있습니다. 오류의 원인은, GetResult 메서드에서 사용한 double.NegativeInfinity 상수가 IL 코드로는 -inf로 번역되었기 때문인데요,

.method assembly hidebysig static float64 
        '<<Main>$>g__GetResult|0_0'() cil managed
{
    .custom instance void [System.Runtime]System.Runtime.CompilerServices.CompilerGeneratedAttribute::.ctor() = ( 01 00 00 00 ) 
    // Code size       15 (0xf)
    .maxstack  1
    .locals init (float64 V_0)
    IL_0000:  nop
    IL_0001:  ldc.r8     -inf
    IL_000a:  stloc.0
    IL_000b:  br.s       IL_000d

    IL_000d:  ldloc.0
    IL_000e:  ret
} // end of method Program::'<<Main>$>g__GetResult|0_0'

어째서 ildasm.exe가 그렇게 번역한 것인지는 모르겠지만 어쨌든 ilasm.exe에는 그것이 아직 반영되지 않아 -inf 토큰을 해석할 수 없었던 것입니다. 이에 대해 검색해 보면,

error : syntax error at token '-' in: IL_002d: ldc.r8 -nan(ind) with "return Double.NaN
; https://github.com/3F/DllExport/issues/158

ILDasm.exe regression with infinum floating point values in .NET Framework 4.8 SDK
; https://developercommunity.visualstudio.com/t/ildasmexe-regression-with-infinum-floating-point-v/545431

답이 나오는데요, 각각 다음의 표현으로 수정하는 것으로 어셈블이 가능해집니다.

ldc.r4 inf
 ==> ldc.r4 (00 00 80 7F)

ldc.r4 -inf
 ==> ldc.r4 (00 00 80 FF)

ldc.r8 inf
 ==> ldc.r8 (00 00 00 00 00 00 F0 7F)

ldc.r8 -inf
 ==> ldc.r8 (00 00 00 00 00 00 F0 FF)

그러니까, 제가 사용한 코드의 경우 이렇게 바꿔주면 되는 것입니다.

.method assembly hidebysig static float64 
        '<<Main>$>g__GetResult|0_0'() cil managed
{
    .custom instance void [System.Runtime]System.Runtime.CompilerServices.CompilerGeneratedAttribute::.ctor() = ( 01 00 00 00 ) 
    // Code size       15 (0xf)
    .maxstack  1
    .locals init (float64 V_0)
    IL_0000:  nop
    //     IL_0001:  ldc.r8     -inf
    IL_0001:  ldc.r8 (00 00 00 00 00 00 F0 FF)
    IL_000a:  stloc.0
    IL_000b:  br.s       IL_000d

    IL_000d:  ldloc.0
    IL_000e:  ret
} // end of method Program::'<<Main>$>g__GetResult|0_0'

이후 ilasm.exe로 컴파일하고, 그 결과를 dnSpy 등의 도구로 보면 double.NegativeInfinity로 나옵니다.

[CompilerGenerated]
internal static double <<Main>$>g__GetResult|0_0()
{
    return double.NegativeInfinity;
}

재미있는 건, (ildasm.exe와는 달리) 위의 코드를 dnSpy의 "IL" 보기로 전환하면 "IL_0001: ldc.r8 (00 00 00 00 00 00 F0 FF)" 코드가 나옵니다. (또한, 최신 버전의 ildasm을 사용해도 바이트 배열 코드가 나옵니다.)




한 예로, 근래의 "Oracle.ManagedDataAccess.dll"을 (pdb 디버깅을 위해) 역어셈블하면 -inf 예외가 발생하는데요, 이 문제를 위와 같은 방법으로 수정하거나 아니면 .NET Core/5+에 포함된 ildasm.exe를 써도 됩니다.




[이 글에 대해서 여러분들과 의견을 공유하고 싶습니다. 틀리거나 미흡한 부분 또는 의문 사항이 있으시면 언제든 댓글 남겨주십시오.]

[연관 글]






[최초 등록일: ]
[최종 수정일: 12/6/2023]

Creative Commons License
이 저작물은 크리에이티브 커먼즈 코리아 저작자표시-비영리-변경금지 2.0 대한민국 라이센스에 따라 이용하실 수 있습니다.
by SeongTae Jeong, mailto:techsharer at outlook.com

비밀번호

댓글 작성자
 




1  2  3  4  [5]  6  7  8  9  10  11  12  13  14  15  ...
NoWriterDateCnt.TitleFile(s)
13818정성태11/15/20245287Windows: 272. Windows 11 24H2 - sudo 추가
13817정성태11/14/20244929Linux: 106. eBPF / bpf2go - (BPF_MAP_TYPE_HASH) Map을 이용한 전역 변수 구현
13816정성태11/14/20245386닷넷: 2312. C#, C++ - Windows / Linux 환경의 Thread Name 설정파일 다운로드1
13815정성태11/13/20244809Linux: 105. eBPF - bpf2go에서 전역 변수 설정 방법
13814정성태11/13/20245283닷넷: 2311. C# - Windows / Linux 환경에서 Native Thread ID 가져오기파일 다운로드1
13813정성태11/12/20245047닷넷: 2310. .NET의 Rune 타입과 emoji 표현파일 다운로드1
13812정성태11/11/20245266오류 유형: 933. Active Directory - The forest functional level is not supported.
13811정성태11/11/20244848Linux: 104. Linux - COLUMNS 환경변수가 언제나 80으로 설정되는 환경
13810정성태11/10/20245389Linux: 103. eBPF (bpf2go) - Tracepoint를 이용한 트레이스 (BPF_PROG_TYPE_TRACEPOINT)
13809정성태11/10/20245261Windows: 271. 윈도우 서버 2025 마이그레이션
13808정성태11/9/20245257오류 유형: 932. Linux - 커널 업그레이드 후 "error: bad shim signature" 오류 발생
13807정성태11/9/20244994Linux: 102. Linux - 커널 이미지 파일 서명 (Ubuntu 환경)
13806정성태11/8/20244902Windows: 270. 어댑터 상세 정보(Network Connection Details) 창의 내용이 비어 있는 경우
13805정성태11/8/20244736오류 유형: 931. Active Directory의 adprep 또는 복제가 안 되는 경우
13804정성태11/7/20245366Linux: 101. eBPF 함수의 인자를 다루는 방법
13803정성태11/7/20245322닷넷: 2309. C# - .NET Core에서 바뀐 DateTime.Ticks의 정밀도
13802정성태11/6/20245704Windows: 269. GetSystemTimeAsFileTime과 GetSystemTimePreciseAsFileTime의 차이점파일 다운로드1
13801정성태11/5/20245492Linux: 100. eBPF의 2가지 방식 - libbcc와 libbpf(CO-RE)
13800정성태11/3/20246323닷넷: 2308. C# - ICU 라이브러리를 활용한 문자열의 대소문자 변환 [2]파일 다운로드1
13799정성태11/2/20244907개발 환경 구성: 732. 모바일 웹 브라우저에서 유니코드 문자가 표시되지 않는 경우
13798정성태11/2/20245521개발 환경 구성: 731. 유니코드 - 출력 예시 및 폰트 찾기
13797정성태11/1/20245508C/C++: 185. C++ - 문자열의 대소문자를 변환하는 transform + std::tolower/toupper 방식의 문제점파일 다운로드1
13796정성태10/31/20245382C/C++: 184. C++ - ICU dll을 이용하는 예제 코드 (Windows)파일 다운로드1
13795정성태10/31/20245178Windows: 268. Windows - 리눅스 환경처럼 공백으로 끝나는 프롬프트 만들기
13794정성태10/30/20245262닷넷: 2307. C# - 윈도우에서 한글(및 유니코드)을 포함한 콘솔 프로그램을 컴파일 및 실행하는 방법
13793정성태10/28/20245135C/C++: 183. C++ - 윈도우에서 한글(및 유니코드)을 포함한 콘솔 프로그램을 컴파일 및 실행하는 방법
1  2  3  4  [5]  6  7  8  9  10  11  12  13  14  15  ...