[PATCH] Fix crash in ARM backend inside of ARMConstantIslandPass

[David Majnemer]

Hi James,

I've attached a patch that address your concerns regarding the testcase.

-- 
David Majnemer



On Tue, Jun 4, 2013 at 3:32 AM, James Molloy <james at jamesmolloy.co.uk>wrote:

> Hi David,
>
> Thanks for working on this. I was initially concerned that the patch
> didn't address LDRH too, but I've noticed that uses a different addressing
> mode so isn't affected.
>
> Minor comments: your testcase can be reduced further. It's complicated by
> the fact that you took it directly from Clang's output - you can change the
> global constructor into a simple function and remove the TBAA data - that
> leaves you with a simpler testcase which is more obvious to a maintainer
> (we're not testing global constructors in this test!)
>
> Also, could you please rename your testcase to something more descriptive
> than a PR number? You can say inside the testcase that it addresses a PR,
> but the test name is what shows up when we break it so please make it more
> descriptive.
>
> Cheers,
>
> James
>
> On 4 June 2013 06:54, David Majnemer <david.majnemer at gmail.com> wrote:
>
>> Hi,
>>
>> The ARM backend did not expect LDRBi12 to hold a constant pool operand.
>>
>> The attached patch allows for LLVM to deal with the instruction similar
>> to how it deals with LDRi12.
>>
>> This addresses PR16215.
>>
>> --
>> David Majnemer
>>
>> _______________________________________________
>> llvm-commits mailing list
>> llvm-commits at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20130604/a5b6fa4e/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: PR16215-v2.diff
Type: application/octet-stream
Size: 1130 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20130604/a5b6fa4e/attachment.obj>