search for: recordlayoutbuilder

So, when performing expression evaluation, lldb trips over an assert in clang/lib/AST/RecordLayoutBuilder because ExternalFieldOffsets doesn't contain a FieldDecl that updateExternalFieldOffset expected. I found that the assert occurs when both static and non-static member variables are present. For instance, with the following, the lldb command 'expr my_test.length()' does not assert unl...

...On Behalf Of Thirumurthi, Ashok Sent: Friday, February 08, 2013 5:59 PM To: llvmdev at cs.uiuc.edu Cc: Greg Clayton Subject: [LLVMdev] assert when mixing static and non-static members with an external AST source So, when performing expression evaluation, lldb trips over an assert in clang/lib/AST/RecordLayoutBuilder because ExternalFieldOffsets doesn't contain a FieldDecl that updateExternalFieldOffset expected. I found that the assert occurs when both static and non-static member variables are present. For instance, with the following, the lldb command 'expr my_test.length()' does not assert unl...

While investigating compile times, I noticed that there's no information in the LLVM coding standards about code in headers. Many LLVM/Clang headers have lots of complex code in headers - I'm specifically looking at the Static Analyzer projects (seemingly slowest compile times, biggest hinderance to my productivity) - that probably doesn't need to be there. Nobody likes slow compile

My target implementation has 2 byte (16 bit ints) For my target implementation I would want that Clang would always use 2 byte aligned padding for Structs, which would match the size of an int. The current situation is that for this struct: struct AA { char n; char m; char j; }; I get it aligned by 1: %a = alloca %struct.AA, align 1 I would want it to implicitly use 4 bytes

...ccio <http://about.me/ariccio> If left to my own devices, I will build more. ⁂ On Wed, Feb 10, 2016 at 7:13 PM, Reid Kleckner <rnk at google.com> wrote: > Some parts of Clang are very careful about keeping implementation details > out of headers, and others aren't. Consider RecordLayoutBuilder, for > example, whose class definition is in the .cpp file. > > In general I would say that we are in favor of hiding as many > implementation details as we can, it's just thankless work to refactor > headers to sink code code into .cpp files, prune transitive includes, and >...

...lways (at least) 2 byte aligned ? > > I don't think there's a feature you can toggle for this (except, > maybe, making the alignment of every basic type 2 bytes; but that > would obviously affect arrays and even normal variables too). So you > probably have to modify lib/AST/RecordLayoutBuilder.cpp directly. > > Might be worth asking this on the cfe-dev mailing list though. That's > where most of the Clang experts live. > > Cheers. > > Tim. -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-...