On Thu, Nov 25, 2010 at 5:55 PM, Talin <span dir="ltr"><<a href="mailto:viridia@gmail.com">viridia@gmail.com</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div><div></div><div class="h5">On Tue, Nov 9, 2010 at 9:04 AM, Devang Patel <span dir="ltr"><<a href="mailto:dpatel@apple.com" target="_blank">dpatel@apple.com</a>></span> wrote:<br></div></div><div class="gmail_quote">
<div><div></div><div class="h5"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div bgcolor="#FFFFFF"><div><div><br></div><div><br>On Nov 8, 2010, at 10:52 PM, Talin <<a href="mailto:viridia@gmail.com" target="_blank">viridia@gmail.com</a>> wrote:<br><br></div><div></div><blockquote type="cite">
<div>On Mon, Nov 8, 2010 at 9:56 AM, Devang Patel <span dir="ltr"><<a href="mailto:dpatel@apple.com" target="_blank"></a><a href="mailto:dpatel@apple.com" target="_blank">dpatel@apple.com</a>></span> wrote:<br><div class="gmail_quote">
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div style="word-wrap:break-word"><br><div><div><div>On Nov 6, 2010, at 7:35 PM, Talin wrote:</div><br><blockquote type="cite">After to speaking to Devang and a number of other people at the developer's conference, I was able to make some forward progress on getting debugging to work. I'm now able to actually single-step through my program and set breakpoints, and examine function parameters.<div>
<br></div><div>However, I'm also seeing a lot of new problems which weren't exposed before. After spending the better part of two days working on them, I'd like to describe them here and get some advice on what I might be doing wrong.</div>
<div><br></div><div>#1) Class sizes coming out as zero.</div><div><br></div><div>In my frontend, I call DebugFactory::CreateCompositeTypeEx as follows:</div><div><br></div><div><font face="'courier new', monospace"> DICompositeType di = dbgFactory_.CreateCompositeTypeEx(</font></div>
<div><font face="'courier new', monospace"> type->typeClass() == Type::Class ? dwarf::DW_TAG_class_type : dwarf::DW_TAG_structure_type,</font></div><div><font face="'courier new', monospace"> dbgCompileUnit_,</font></div>
<div><font face="'courier new', monospace"> type->typeDefn()->linkageName().c_str(),</font></div><div><font face="'courier new', monospace"> genDIFile(type->typeDefn()),</font></div>
<div><font face="'courier new', monospace"> getSourceLineNumber(type->typeDefn()->location()),</font></div><div><font face="'courier new', monospace"> getSizeOfInBits(type->irType()),</font></div>
<div><font face="'courier new', monospace"> getAlignOfInBits(type->irType()),</font></div><div><font face="'courier new', monospace"> getInt64Val(0), 0,</font></div>
<div><font face="'courier new', monospace"> DIType(),</font></div><div><font face="'courier new', monospace"> dbgFactory_.GetOrCreateArray(members.data(), members.size()));</font></div>
<div> <br clear="all">The 'getSizeOfInBits()' function and the others like it basically call llvm::ConstantExpr::getSizeOf() and then multiple the result by 8. Looking at the generated IR, I can see that the expressions look correct to me (I've added some line breaks for clarity):</div>
<div><br></div><div><div><font face="'courier new', monospace"> !15 = metadata !{i32 589826, metadata !2, metadata !"tart.core.Object", metadata !12, i32 9,</font></div><div><font face="'courier new', monospace"> i64 mul (i64 ptrtoint (%tart.core.Object* getelementptr (%tart.core.Object* null, i32 1) to i64), i64 8),</font></div>
</div></blockquote><div><br></div><div><br></div></div>This does not work. The DebugInfo accessors uses getSizeInBits() to get an integer for the size. You will have to modify it to get the value.</div></div></blockquote>
<div><br></div><div>I'm not sure I understand what you are saying - isn't llvm::ConstantExpr::getSizeOf() and llvm::ConstantExpr::getAlignOf() the proper way to compute the size and alignment of a field? If those functions can't be used for debug info, then that's really inconvenient - I don't want to have to write my own calculations for it, since (a) I'm likely to get it wrong in some subtle way, and (b) I'd be duplicating code, since the same calculations need to be done in a lot of other places.</div>
</div></div></blockquote><div><br></div><div><br></div><div><br></div></div>llvm::Constant* refers to llvm struct and fields. They do not necessarily map directly to source lang structs and fields. Debugging information cares about size, offset and alignment as per lang. That said, if anyone prepares appropriate DebugInfo and Dwarfwriter patches then I won't oppose.<div>
<br></div></div></blockquote></div></div><div>First, apologies for responding to this after such a long interval.</div><div><br></div><div>So I think I understand what you are saying here, although the way that I would say it is "Dwarfwriter doesn't understand constant expressions, only simple constants". At least, when I changed my code to use simple constants instead of ConstantExpr::getSizeOf(), it worked.</div>
<div><br></div><div>Also, in my original message in this thread, I listed 3 issues. The one we've been discussing is issue #1 (class sizes coming out as zero) which I will consider solved. However, I never did figure out the answers to the other two issues I mentioned: 2) <span style="font-family:arial, sans-serif;font-size:13px;border-collapse:collapse">llvm.dbg.declare() not working for local variables, and 3) </span><span style="font-family:arial, sans-serif;font-size:13px;border-collapse:collapse">dwarfdump printing lots of errors when trying to dump a module produced by my frontend - even though when I check the debug info manually it all appears to be correct.</span></div>
<div><div></div><div class="h5">
<div><br></div></div></div></div></blockquote><div>Also, there's another issue I wanted to mention: I have noticed that debugging and inlining don't play very well together. Even if other optimizations are disabled, if I turn on inlining, I get null pointer errors in this piece of code in DwarfDebug::constructInlinedScopeDIE():</div>
<div><meta http-equiv="content-type" content="text/html; charset=utf-8"><span class="Apple-style-span" style="font-family: monospace; font-size: medium; white-space: pre; "><pre> <b><font color="#A020F0">if</font></b> (!Scope->getScopeNode())
<b><font color="#A020F0">return</font></b> NULL;
DIScope DS(Scope->getScopeNode());
DIE *ScopeDIE = <b><font color="#A020F0">new</font></b> DIE(dwarf::DW_TAG_inlined_subroutine);
DISubprogram InlinedSP = getDISubprogram(DS);
CompileUnit *TheCU = getCompileUnit(InlinedSP);
DIE *OriginDIE = TheCU->getDIE(InlinedSP);
assert(OriginDIE && <b><font color="#BC8F8F">"Unable to find Origin DIE!"</font></b>);
addDIEEntry(ScopeDIE, dwarf::DW_AT_abstract_origin,
<b><font color="#5F9EA0">dwarf</font></b>::DW_FORM_ref4, OriginDIE);</pre></span></div><div>What's happening is that InlinedSP.DbgNode is NULL, which causes getCompileUnit() to assert.</div><div><br></div>
<div>Note that this occurs in both llc and my linker (tartln) because they both use the same set of passes to generate machine code. So basically whenever I want to debug, I have to disable inlining.</div><div><br></div>
<div>
<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div class="gmail_quote"><div><div class="h5"><div></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div bgcolor="#FFFFFF"><div></div><div><font color="#888888">Devang</font><div><div></div><div><br><div><blockquote type="cite"><div><div class="gmail_quote">
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div style="word-wrap:break-word"><div><div><div></div><div><font face="'courier new', monospace"><br>
</font><blockquote type="cite"><div>
<div><font face="'courier new', monospace"> i6 4 mul (i64 ptrtoint (%tart.core.Object* getelementptr ({ i1, %tart.core.Object }* null, i64 0, i32 1) to i64), i64 8),</font></div><div>
<font face="'courier new', monospace"> i64 0, i32 0, null, metadata !16, i32 0, null} ; [ DW_TAG_class_type ]</font></div></div><div><br></div><div>However, when I print out the debugging information via dwarfdump, you can see that the size of the object is zero:</div>
<div><br></div><div><div><font face="'courier new', monospace"><1>< 2233> DW_TAG_class_type</font></div><div><font face="'courier new', monospace"> DW_AT_sibling <2295></font></div>
<div><font face="'courier new', monospace"> DW_AT_name tart.core.Object</font></div><div><font face="'courier new', monospace"><b> DW_AT_byte_size 0</b></font></div>
<div><font face="'courier new', monospace"> DW_AT_decl_file 65</font></div><div><font face="'courier new', monospace"> DW_AT_decl_line 9</font></div>
<div><font face="'courier new', monospace"><2>< 2258> DW_TAG_member</font></div><div><font face="'courier new', monospace"> DW_AT_name __tib</font></div>
<div><font face="'courier new', monospace"> DW_AT_type <2228></font></div><div><font face="'courier new', monospace"> DW_AT_decl_file 65</font></div>
<div><font face="'courier new', monospace"> DW_AT_decl_line 10</font></div><div><font face="'courier new', monospace"> DW_AT_data_member_location DW_OP_plus_uconst 0</font></div>
<div><font face="'courier new', monospace"><2>< 2274> DW_TAG_member</font></div><div><font face="'courier new', monospace"> DW_AT_name __gcstate</font></div>
<div><font face="'courier new', monospace"> DW_AT_type <156></font></div><div><font face="'courier new', monospace"> DW_AT_decl_file 65</font></div>
<div><font face="'courier new', monospace"> DW_AT_decl_line 11</font></div><div><font face="'courier new', monospace"> DW_AT_data_member_location DW_OP_plus_uconst 0</font></div>
</div><div><br></div><div>In addition, it appears that all of the field offsets are zero as well - see the DIEs immediately following the class definition.</div><div><br></div><div>#2) I can't seem to get llvm.dbg.declare() to work for local variables, although it appears to work fine for function parameters. For example, look at the following snippet of IR:</div>
<div><br></div><blockquote style="margin:0 0 0 40px;border:none;padding:0px"><div><div><font face="'courier new', monospace">define {} @ReflectionTest.testModuleReflection(%ReflectionTest* %self) gc "tart-gc" {</font></div>
</div><div><div><font face="'courier new', monospace">prologue:</font></div></div><div><div><font face="'courier new', monospace"> %m = alloca %tart.reflect.Module*</font></div>
</div><div><div><font face="'courier new', monospace"> call void @llvm.dbg.value(metadata !{%ReflectionTest* %self}, i64 0, metadata !1176)</font></div></div><div><div><font face="'courier new', monospace"> call void @llvm.dbg.declare(metadata !{%tart.reflect.Module** %m}, metadata !1177)</font></div>
</div><div><div><font face="'courier new', monospace"> %0 = bitcast %tart.reflect.Module** %m to i8**, !dbg !1178</font></div></div><div><div><font face="'courier new', monospace"> call void @llvm.gcroot(i8** %0, i8* null), !dbg !1178</font></div>
</div></blockquote><div><br></div><div>In the debugger, I can examine the value of 'self', however when I attempt to print the value of 'm', gdb reports 'No symbol "m" in current context.'</div>
</blockquote><div><br></div></div></div><div>Are using -O0? You do not have line number info with dbg.declare intrinsic. FWIW, following works ...</div><div><br></div><div><div>define void @foo() nounwind ssp {</div><div>
entry:</div><div> %i = alloca i32, align 4</div><div> call void @llvm.dbg.declare(metadata !{i32* %i}, metadata !5), !dbg !8</div><div> store i32 2, i32* %i, align 4, !dbg !9</div><div> ret void, !dbg !10</div><div>}</div>
</div><div><br></div><div>-</div><div>Devang</div><br></div><br></div></blockquote></div><br><br clear="all"><br>-- <br>-- Talin<br>
</div></blockquote></div></div></div></div></div>
</blockquote></div></div></div><br><br clear="all"><br>-- <br><font color="#888888">-- Talin<br>
</font></blockquote></div><br><br clear="all"><br>-- <br>-- Talin<br>