c++ - Get precise line/column debug info from LLVM IR -


i trying locate instructions in llvm pass line , column number (reported third-party tool) instrument them. achieve this, compiling source files clang -g -o0 -emit-llvm , looking information in metadata using code:

const debugloc &location = instruction->getdebugloc(); // location.getline() // location.getcol()

unfortunately, information absolutely imprecise. consider following implementation of fibonacci function: 

unsigned fib(unsigned n) {     if (n < 2)         return n;      unsigned f = fib(n - 1) + fib(n - 2);     return f; }

i locate single llvm instruction corresponding assignment unsigned f = ... in resulting llvm ir. not interested in calculations of right-hand side. generated llvm block including relevant debug metadata is: 

[...]  if.end:                                           ; preds = %entry   call void @llvm.dbg.declare(metadata !{i32* %f}, metadata !17), !dbg !18   %2 = load i32* %n.addr, align 4, !dbg !19   %sub = sub i32 %2, 1, !dbg !19   %call = call i32 @fib(i32 %sub), !dbg !19   %3 = load i32* %n.addr, align 4, !dbg !20   %sub1 = sub i32 %3, 2, !dbg !20   %call2 = call i32 @fib(i32 %sub1), !dbg !20   %add = add i32 %call, %call2, !dbg !20   store i32 %add, i32* %f, align 4, !dbg !20   %4 = load i32* %f, align 4, !dbg !21   store i32 %4, i32* %retval, !dbg !21   br label %return, !dbg !21  [...]  !17 = metadata !{i32 786688, metadata !4, metadata !"f", metadata !5, i32 5, metadata !8, i32 0, i32 0} ; [ dw_tag_auto_variable ] [f] [line 5] !18 = metadata !{i32 5, i32 11, metadata !4, null} !19 = metadata !{i32 5, i32 15, metadata !4, null} !20 = metadata !{i32 5, i32 28, metadata !4, null} !21 = metadata !{i32 6, i32 2, metadata !4, null} !22 = metadata !{i32 7, i32 1, metadata !4, null}

as can see, metadata !dbg !20 of store instruction points line 5 column 28, call fib(n - 2). worse, add operation , subtraction n - 2 both point function call, identified !dbg !20.

interestingly, clang ast emitted clang -xclang -ast-dump -fsyntax-only has information. thus, suspect somehow lost during code generation phase. seems during code generation clang reaches internal sequence point , associates following instructions position until next sequence point (e.g. function call) occurs. completeness, here declaration statement in ast: 

|-declstmt 0x7ffec3869f48 <line:5:2, col:38> | `-vardecl 0x7ffec382d680 <col:2, col:37> col:11 used f 'unsigned int' cinit |   `-binaryoperator 0x7ffec3869f20 <col:15, col:37> 'unsigned int' '+' |     |-callexpr 0x7ffec382d7e0 <col:15, col:24> 'unsigned int' |     | |-implicitcastexpr 0x7ffec382d7c8 <col:15> 'unsigned int (*)(unsigned int)' <functiontopointerdecay> |     | | `-declrefexpr 0x7ffec382d6d8 <col:15> 'unsigned int (unsigned int)' function 0x7ffec382d490 'fib' 'unsigned int (unsigned int)' |     | `-binaryoperator 0x7ffec382d778 <col:19, col:23> 'unsigned int' '-' |     |   |-implicitcastexpr 0x7ffec382d748 <col:19> 'unsigned int' <lvaluetorvalue> |     |   | `-declrefexpr 0x7ffec382d700 <col:19> 'unsigned int' lvalue parmvar 0x7ffec382d3d0 'n' 'unsigned int' |     |   `-implicitcastexpr 0x7ffec382d760 <col:23> 'unsigned int' <integralcast> |     |     `-integerliteral 0x7ffec382d728 <col:23> 'int' 1 |     `-callexpr 0x7ffec3869ef0 <col:28, col:37> 'unsigned int' |       |-implicitcastexpr 0x7ffec3869ed8 <col:28> 'unsigned int (*)(unsigned int)' <functiontopointerdecay> |       | `-declrefexpr 0x7ffec3869e10 <col:28> 'unsigned int (unsigned int)' function 0x7ffec382d490 'fib' 'unsigned int (unsigned int)' |       `-binaryoperator 0x7ffec3869eb0 <col:32, col:36> 'unsigned int' '-' |         |-implicitcastexpr 0x7ffec3869e80 <col:32> 'unsigned int' <lvaluetorvalue> |         | `-declrefexpr 0x7ffec3869e38 <col:32> 'unsigned int' lvalue parmvar 0x7ffec382d3d0 'n' 'unsigned int' |         `-implicitcastexpr 0x7ffec3869e98 <col:36> 'unsigned int' <integralcast> |           `-integerliteral 0x7ffec3869e60 <col:36> 'int' 2

is either possible improve accuracy of debug metadata, or resolve corresponding instruction in different way? ideally, leave clang untouched, i.e. not modify , recompile it.

turns out, has been fixed introduction of mdlocation in llvm release 3.6.0. @ time of writing, current clang compiler shipped xcode command line tools still generates former "buggy" location information, though it's version string says apple llvm version 6.1.0 (clang-602.0.49) (based on llvm 3.6.0svn). after downloading pre-built binary, generated llvm ir looks this:

[...]  ; <label>:7                                       ; preds = %0   call void @llvm.dbg.declare(metadata i32* %f, metadata !21, metadata !14), !dbg !22   %8 = load i32* %2, align 4, !dbg !23   %9 = sub i32 %8, 1, !dbg !23   %10 = call i32 @fib(i32 %9), !dbg !24   %11 = load i32* %2, align 4, !dbg !25   %12 = sub i32 %11, 2, !dbg !25   %13 = call i32 @fib(i32 %12), !dbg !26   %14 = add i32 %10, %13, !dbg !24   store i32 %14, i32* %f, align 4, !dbg !22   %15 = load i32* %f, align 4, !dbg !27   store i32 %15, i32* %1, !dbg !28   br label %16, !dbg !28   [...]  !22 = !mdlocation(line: 5, column: 14, scope: !4) !23 = !mdlocation(line: 5, column: 22, scope: !4) !24 = !mdlocation(line: 5, column: 18, scope: !4) !25 = !mdlocation(line: 5, column: 35, scope: !4) !26 = !mdlocation(line: 5, column: 31, scope: !4) !27 = !mdlocation(line: 6, column: 12, scope: !4) !28 = !mdlocation(line: 6, column: 5, scope: !4)

the location metadata points beginning of expression. assignment, instance, left hand side specifier f @ line 5 column 14. seen in !dbg !24, might still ambiguous, unfortunately.

there has been 1 more change: access getline() , getcolumn() fail if no debug metadata attached instruction. debugloc class offers convenient way check this:

const debugloc &location = instruction->getdebugloc(); if (location) {     // location.getline()     // location.getcol() } else {     // no location metadata available }

-

Popular posts from this blog

c# - ODP.NET Oracle.ManagedDataAccess causes ORA-12537 network session end of file -

overview i want replace oracle.dataaccess orcale. managed dataaccess, opening connection latter throws ora-12537 network session end of file exception. exception message / stack trace {oracleinternal.network.networkexception (0x000030f9): ora-12537 : netzwerksession: dateiende @ oracleinternal.network.readerstream.read(orabuf ob) @ oracleinternal.ttc.orabufreader.getdatafromnetwork() @ oracleinternal.ttc.orabufreader.read(boolean bignoredata) @ oracleinternal.ttc.marshallingengine.unmarshalub1(boolean bignoredata) @ oracleinternal.ttc.ttcprotocolnegotiation.readresponse()} i trying connect oracle 11g database , not have client installed on local machine. working test application (unmanaged) using oracle.dataaccess works fine . using system; using oracle.dataaccess.client; namespace app.odp.unmanaged { internal class program { private static void main(string[] args) { //dummy connection string...
Read more

matlab - Compression and Decompression of ECG Signal using HUFFMAN ALGORITHM -

i want compress , decompress ecg signal stored text message text file , find out file size, cr,prd ratio,and qs. here code compression , decompression of ecg signal using huffman algorithm. i'm trying find out file size, compression ratio,prd ratio,qs. but error i'm getting the ecg file in text format. %clearing variables , screen enter code here clear all; close all; clc; [file1]=uigetfile('*.txt'); %fprintf('enter ecg file name :%s',file1); p=file1; t=sprintf('%s',p); b1=dlmread([file1]); len=length(b1); lead=input('enter lead number :'); %lead=str2double(lead); col2=b1(1:end,lead); e=fix(sqrt(len)); m=1; i=1:e j=1:e g2(i,j)=fix(col2(m)*1000); %amplifying 1000 m=m+1; end end g3=g2; i=1:e j=1:e ...
Read more

utf 8 - split utf-8 string into bytes in python -

i trying split utf-8 string bytes in python 3. problem is, when use bytearray, byte, encode etc functions array size of element 14 bytes, not 1 byte expected. need split text file sequence of bytes , send them byte after byte using sockets. tried this: infile = open (file, "r") str = infile.read() byte_str = bytes(str, 'utf-8') print("size of byte_str",sys.getsizeof(byte_str[0])) print gives me 14, need 1... suggestion? quoting official documentation : sys.getsizeof(object[, default]) return size of object in bytes. object can type of object. built-in objects return correct results, not have hold true third-party extensions implementation specific. only memory consumption directly attributed object accounted for, not memory consumption of objects refers to. if given, default returned if object not provide means retrieve size. otherwise typeerror raised. getsizeof() calls object’s __sizeof__ method , adds ...
Read more