[cfe-dev] plans about kernel compilation?

[Chris Lattner]

On Oct 28, 2007, at 1:14 PM, Török Edvin wrote:
> On 10/28/07, Chris Lattner <clattner at apple.com> wrote:
>> On the other hand, if you're asking whether "llvm" can compile the
>> kernel, llvm-gcc is probably either capable of doing it now, or very
>> close to being able to.  I'd give it a try and file bugs if something
>> doesn't work.
>
> I am compiling a 2.6.24-rc1 kernel for x86-64.
> I filed 2 bugs for llvm-gcc and clang.

I fixed the llvm-gcc bugs.

I closed one clang bug as a dup of the other.  I fixed many minor  
clang bugs in various subsystems exposed by this, thanks!

Unfortunately, I don't have time at the moment to fix the rest of the  
issues.  Here is what is left:

1) Sema::MergeFunctionDecl needs to allow a prototype after a  
function body.
http://llvm.org/bugs/show_bug.cgi?id=1750#c5

2) GCC anonymous unions parse bug don't allow field access when you  
use P->b.
http://llvm.org/bugs/show_bug.cgi?id=1750#c6

3) We have a minor sema bug related to struct assignment and cv  
qualifiers, I asked Steve to investigate:
http://llvm.org/bugs/show_bug.cgi?id=1750#c8

If anyone is interested in tackling #1 or #2, they are both self- 
contained and would be a great way to dive into clang.

Thanks Edvin!

-Chris