r184438 - Fix CodeGenCXX/debug-info.cpp test on Windows

[David Blaikie]

On Thu, Jun 20, 2013 at 10:13 AM, Reid Kleckner <rnk at google.com> wrote:
> On Thu, Jun 20, 2013 at 1:01 PM, Reid Kleckner <rnk at google.com> wrote:
>>
>> On Thu, Jun 20, 2013 at 12:43 PM, David Blaikie <dblaikie at gmail.com>
>> wrote:
>>>
>>>
>>> On Jun 20, 2013 9:30 AM, "Reid Kleckner" <reid at kleckner.net> wrote:
>>> >
>>> > Author: rnk
>>> > Date: Thu Jun 20 11:28:24 2013
>>> > New Revision: 184438
>>> >
>>> > URL: http://llvm.org/viewvc/llvm-project?rev=184438&view=rev
>>> > Log:
>>> > Fix CodeGenCXX/debug-info.cpp test on Windows
>>> >
>>> > On Windows, it looks like FlagIndirectVariable is being set in Flags
>>> > for
>>> > DIBuilder::createLocalVariable(), giving us an i32 of 8192 instead of
>>> > 0,
>>> > as on Linux.
>>> >
>>> > Fixes breakage from r184367.
>>> >
>>> > Modified:
>>> >     cfe/trunk/test/CodeGenCXX/debug-info.cpp
>>> >
>>> > Modified: cfe/trunk/test/CodeGenCXX/debug-info.cpp
>>> > URL:
>>> > http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGenCXX/debug-info.cpp?rev=184438&r1=184437&r2=184438&view=diff
>>> >
>>> > ==============================================================================
>>> > --- cfe/trunk/test/CodeGenCXX/debug-info.cpp (original)
>>> > +++ cfe/trunk/test/CodeGenCXX/debug-info.cpp Thu Jun 20 11:28:24 2013
>>> > @@ -92,7 +92,7 @@ incomplete (*x)[3];
>>> >  }
>>> >
>>> >  // For some reason the argument for PR14763 ended up all the way down
>>> > here
>>> > -// CHECK: = metadata !{i32 {{[0-9]*}}, metadata [[FUNC]], {{.*}},
>>> > metadata [[FOO]], i32 8192, i32 0} ; [ DW_TAG_arg_variable ] [f]
>>> > +// CHECK: = metadata !{i32 {{[0-9]*}}, metadata [[FUNC]], {{.*}},
>>> > metadata [[FOO]], i32 {{[0-9]+}}, i32 0} ; [ DW_TAG_arg_variable ] [f]
>>>
>>> Except I need that test that the flag is indirect (8193) to ensure we
>>> emit the right location info for this parameter. So now the test is too
>>> permissive.
>>
>> I got my commit description backwards between looking at expected output
>> and real output.  We're tripping the byval check in there which turns *off*
>> that flag.
>>>
>>> Should we xfail this on windows &mention a PR for it instead?
>>
>> I'd rather pick a non-win32 triple for this test to keep coverage.

Done in r184442.

>> I'll file a PR for it.
>
>
> http://llvm.org/bugs/show_bug.cgi?id=16390

Thanks!

> See also http://llvm.org/bugs/show_bug.cgi?id=16226 for more fun.

Yep, remember you mentioning this craziness. Good luck with that one.