[llvm-commits] [llvm] r91184 - in /llvm/trunk: lib/Transforms/Scalar/ScalarReplAggregates.cpp test/Transforms/ScalarRepl/2009-12-11-NeonTypes.ll

[Eric Christopher]

On Dec 19, 2009, at 12:35 AM, Eli Friedman wrote:

> On Sat, Dec 19, 2009 at 12:19 AM, Eric Christopher <echristo at apple.com> wrote:
>> 
>> On Dec 18, 2009, at 11:10 PM, Bob Wilson wrote:
>> 
>>>> The returned Idx should be a uint64_t for large indexes into large arrays.
>>> 
>>> OK.  I had considered that earlier but since SROA only handles small arrays I decided it didn't matter.  I guess you could have a nested aggregate where the inner array was too big to address with 32 bits.  I'm still not sure if it wouldn't have been better to check for large arrays and exclude them earlier, but it's not a big deal either way.
>> 
>> Been meaning to ask, for things like this would size_t be better? Things like vector, list, etc return size_t for size().
> 
> size_t is fine for the size of an array, but not for the size of an
> LLVM array type.  Remember that we could be compiling 64-bit code on a
> 32-bit system.

Ah yes. Just had to fix a few warnings, it's late at night and I've been meaning to ask. :)

Thanks!

-eric