llvm dev - Aug 2018 - Passing arguments to var args function

Hey Tim, thanks for the quick answer,
this looks exactly like what I was trying to achieve, I was hoping there
was something more clean,
I was wondering how the community would accept a change in the TD,
if the code at CallinvConvLower.cpp:138 would've passed the IsFixed flag to
the AssignFn then all those workarounds wouldn't be needed,
this example is only one of a few I can think of for expanding the
CallingConvention tablegen capabilities regarding this issue.

Thanks for the answer.

On Wed, Aug 8, 2018 at 4:21 PM, Tim Northover <t.p.northover at gmail.com>
wrote:
> Hi Liad,
>
> On Wed, 8 Aug 2018 at 13:28, Liad Mordekoviz via llvm-dev
> <llvm-dev at lists.llvm.org> wrote:
> > Did anyone else come by such requirement? is there any other way
anyone
> could recommend handling such requirement?
>
> Apple's calling convention on iOS does this. We have two separate
> CC_Whatever definitions and then AArch64ISelLowering.cpp chooses which
> one to use based on whether the argument is fixed or not (see line
> 3446 onwards).
>
> Cheers.
>
> Tim.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20180808/11c30251/attachment.html>

On Wed, 8 Aug 2018 at 14:42, Liad Mordekoviz
<liad.mordekoviz at nextsilicon.com> wrote:
> I was wondering how the community would accept a change in the TD,
> if the code at CallinvConvLower.cpp:138 would've passed the IsFixed
flag to the AssignFn then all those workarounds wouldn't be needed,
I think that would be a pretty good idea. I suspect the main reason it
wasn't done for AArch64 is that the backend was secret at the time and
Jim wanted to minimize merge conflicts.

Cheers.

Tim.

On 8 August 2018 at 14:49, Tim Northover via llvm-dev
<llvm-dev at lists.llvm.org> wrote:
> On Wed, 8 Aug 2018 at 14:42, Liad Mordekoviz
> <liad.mordekoviz at nextsilicon.com> wrote:
>> I was wondering how the community would accept a change in the TD,
>> if the code at CallinvConvLower.cpp:138 would've passed the IsFixed
flag to the AssignFn then all those workarounds wouldn't be needed,
>
> I think that would be a pretty good idea. I suspect the main reason it
> wasn't done for AArch64 is that the backend was secret at the time and
> Jim wanted to minimize merge conflicts.
Anyone looking at this issue may want to take a look at my patch here
https://reviews.llvm.org/D42374 which adds an IsFixed field to
ISD::ArgFlagsTy.

Best,

Alex