[cfe-commits] r101155 - in /cfe/trunk:lib/CodeGen/CGExprConstant.cpp test/CodeGen/decl.c

[Nuno Lopes]

> On Apr 13, 2010, at 3:04 PM, Nuno Lopes wrote:
>
>> Thanks!
>> I assume it's now ok to turn padding into undef again?  (btw, do you want 
>> to
>> review the patch before commit?)
>
> Yes, iff it doesn't break anything.  Please verify that sqlite works in 
> the jit etc,

Ok, thank you; will do it in my next llvm hacking session :)
Nuno


>> ----- Original Message ----- 
>> From: "Chris Lattner" <sabre at nondot.org>
>> To: <cfe-commits at cs.uiuc.edu>
>> Sent: Tuesday, April 13, 2010 7:16 PM
>> Subject: [cfe-commits] r101155 - in
>> /cfe/trunk:lib/CodeGen/CGExprConstant.cpp test/CodeGen/decl.c
>>
>>
>>> Author: lattner
>>> Date: Tue Apr 13 13:16:19 2010
>>> New Revision: 101155
>>>
>>> URL: http://llvm.org/viewvc/llvm-project?rev=101155&view=rev
>>> Log:
>>> Rework the ConstStructBuilder code to emit missing initializer
>>> elements with explicit zero values instead of with tail padding.
>>> On an example like this:
>>>
>>> struct foo { int a; int b; };
>>>
>>> struct foo fooarray[] = {
>>>   {1, 2},
>>>   {4},
>>> };
>>>
>>> We now lay this out as:
>>>
>>> @fooarray = global [2 x %struct.foo] [%struct.foo { i32 1, i32 2 },
>>> %struct.foo { i32 4, i32 0 }]
>>>
>>> instead of as:
>>>
>>> @fooarray = global %0 <{ %struct.foo { i32 1, i32 2 }, %1 { i32 4, [4 x
>>> i8] zeroinitializer } }>
>>>
>>> Preserving both the struct type of the second element, but also the 
>>> array
>>> type of the entire thing.