similar to: [LLVMdev] [PATCH][RFC] Add extra arguments to TargetLowering::LowerCall() so targets have more context in which to construct call chains

TargetLowering::LowerCall is already a mess, I would really prefer not to extend it any further. It's especially difficult to justify extending it without a use in the open source tree. We really should think hard about how to improve the API in two ways. Perhaps we should wrap the arguments in some struct rather than as individual ones. We should also make it easier to extend it in the

From: Evan Cheng [mailto:evan.cheng at apple.com] Sent: Thursday, April 19, 2012 10:47 AM To: Justin Holewinski Cc: llvmdev at cs.uiuc.edu; llvm-commits at cs.uiuc.edu; Vinod Grover Subject: Re: [llvm-commits] [PATCH][RFC] Add extra arguments to TargetLowering::LowerCall() so targets have more context in which to construct call chains TargetLowering::LowerCall is already a mess, I would really

On Apr 19, 2012, at 11:15 AM, Justin Holewinski wrote: > > From: Evan Cheng [mailto:evan.cheng at apple.com] > Sent: Thursday, April 19, 2012 10:47 AM > To: Justin Holewinski > Cc: llvmdev at cs.uiuc.edu; llvm-commits at cs.uiuc.edu; Vinod Grover > Subject: Re: [llvm-commits] [PATCH][RFC] Add extra arguments to TargetLowering::LowerCall() so targets have more context in which

In response to this discussion, I've prepared the attached patch as an initial prototype for the LowerCall/LowerCallTo change. All of the information currently needed by the back-ends, and the extra information needed by the NVPTX back-end, is now wrapped in a CallLoweringInfo struct. The various SelectionDAG classes have been modified so any calls to TargetLowering::LowerCallTo use this

> -----Original Message----- > From: Evan Cheng [mailto:evan.cheng at apple.com] > Sent: Wednesday, May 16, 2012 5:30 PM > To: Justin Holewinski > Cc: llvmdev at cs.uiuc.edu; llvm-commits at cs.uiuc.edu > Subject: Re: [llvm-commits] [PATCH][RFC] Add extra arguments to > TargetLowering::LowerCall() so targets have more context in which to > construct call chains > >

Thanks. This is going in the right direction IMHO. Obviously, please make sure you add comments to the data structure and convert all the targets first. Also, please don't forget to send an email to llvmdev to warn owners of all the out-of-tree targets about the ABI change. Evan On May 15, 2012, at 6:56 AM, Justin Holewinski <jholewinski at nvidia.com> wrote: > In response to this

On Apr 19, 2012, at 12:46 PM, Evan Cheng <evan.cheng at apple.com> wrote: > > On Apr 19, 2012, at 11:15 AM, Justin Holewinski wrote: > >> >> From: Evan Cheng [mailto:evan.cheng at apple.com] >> Sent: Thursday, April 19, 2012 10:47 AM >> To: Justin Holewinski >> Cc: llvmdev at cs.uiuc.edu; llvm-commits at cs.uiuc.edu; Vinod Grover >> Subject:

A change has just been committed that changes the interface for TargetLowering::LowerCallTo and TargetLowering::LowerCall. Instead of passing a mess of parameters to these functions, a new structure has been created that encapsulates these parameters. The motivation is that additional fields can be added to this structure (with appropriate changes to SelectionDAGBuilder for populating the new

On 24 Sep 2007, at 09:18, Evan Cheng wrote: > +; RUN: llvm-as < %s | llc -march=x86 -mattr=+sse2 -stats -info- > output-file - | grep asm-printer | grep 9 > +; change preceeding line form ... | grep 8 to ..| grep 9 since > +; with new fastcc has std call semantics causing a stack adjustment > +; after the function call > > Not sure if I understand this. Can you illustrate

On Sep 24, 2007, at 2:25 AM, Arnold Schwaighofer wrote: > > On 24 Sep 2007, at 09:18, Evan Cheng wrote: >> +; RUN: llvm-as < %s | llc -march=x86 -mattr=+sse2 -stats -info- >> output-file - | grep asm-printer | grep 9 >> +; change preceeding line form ... | grep 8 to ..| grep 9 since >> +; with new fastcc has std call semantics causing a stack adjustment >>

> > FastCC use to be caller pops arguments so there was no stack > > adjustment after the > > call to qux. Now FastCC has callee pops arguments on return semantics > > so the > > x86 backend inserts a stack adjustment after the call. > > > > _array: > > subl $12, %esp > > movss LCPI1_0, %xmm0 > > mulss

On Sep 16, 2009, at 5:58 AM, Kenneth Uildriks wrote: >> I recently made a major reorganization of the calling-convention >> lowering code which cleared away one of the major obstacles to >> doing this within codegen. >> >> Dan > > So what was the obstacle, and how was it cleared? The biggest obstacle is that there used to be two different methods for lowering

Hi Arnold, This is a very good first step! Thanks! Comments below. Evan Index: test/CodeGen/X86/constant-pool-remat-0.ll =================================================================== --- test/CodeGen/X86/constant-pool-remat-0.ll (revision 42247) +++ test/CodeGen/X86/constant-pool-remat-0.ll (working copy) @@ -1,8 +1,10 @@ ; RUN: llvm-as < %s | llc -march=x86-64 | grep LCPI | count 3 ;

Philip Reames wrote: > Any change outside of statepoint lowering is highly suspect. Notice that SelectionDAGBuilder::LowerCallTo (the one I'm modifying) has exactly one other caller: visitCall, which doesn't match vararg functions. Every other codepath directly calls TargetLowering::LowerCallTo, supplying CallLoweringInfo information explicity (it's a structure with a vararg

> what does "refuses to compile" mean? I.e. what error do you get? > Specifically I get this message when compiling with the default -mattr: Call result #2 has unhandled type i32 UNREACHABLE executed at CallingConvLower.cpp:162! 0 llc 0x0000000100a1e115 PrintStackTrace(void*) + 38 1 llc 0x0000000100a1e6d0 SignalHandler(int) + 254 2

The patch is against revision 42247. -------------- next part -------------- A non-text attachment was scrubbed... Name: tailcall-src.patch Type: application/octet-stream Size: 62639 bytes Desc: not available URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20070923/4770302f/attachment.obj>

Hi Nickhil, The Legalizer expects lower'd call is the node that produce the same number of values as the non-lowered node. That's what the assertion is checking. Take a look at the LowerCall routine for any other targets. You will see that in the non-void function case, it returns a MERGE_VALUES, i.e. all the results merged along with the chain. Cheers, Evan On Nov 15, 2006, at

Hi, I am trying to write a LowerCALL() function for my (custom) target ISA. All I need to do is map a CALL instruction directly onto an SDNode that takes an equal number of arguments (very much alike intrinsics, except that these are custom to my target.) I do not need to implement any call sequences, stack frames etc. I get the following assertion failure: llc: LegalizeDAG.cpp:834:

Hello, I am working on a backend for a custom ISA that is somewhat similar to the MicroBlaze ISA so I've decided to use that as a starting point. I am trying to compile a custom ray tracer (lots of floating point) and the llvm-g++ frontend generates an fneg instruction which is not supported by the MBlaze backend in the 2.8 release. I added code to emit an fneg assembly instruction and now

> I recently made a major reorganization of the calling-convention > lowering code which cleared away one of the major obstacles to > doing this within codegen. > > Dan So what was the obstacle, and how was it cleared? And how do you see the large struct return working in codegen? Anything you care to tell me would be welcome. I will be starting on this today or tomorrow.