similar to: [LLVMdev] Promoting i1,i8,i16

Hrmm.... PromoteVectorOp doesn't seem to follow this at all. http://llvm.org/svn/llvm-project/llvm/trunk/lib/CodeGen/SelectionDAG/LegalizeVectorOps.cpp SDValue VectorLegalizer::PromoteVectorOp(SDValue Op) { // Vector "promotion" is basically just bitcasting and doing the operation // in a different type. For example, x86 promotes ISD::AND on v2i32 to // v1i64. EVT VT =

Notice that PromoteVectorOp is called after the type legalization legalized all of the types in the program. It legalizes the *operations*, not the types. So, you should only see legal types (Legal types are types that fit into your registers). So, if your target has v2i32, I suspect that v4i8 is an illegal because it has a different size. -----Original Message----- From: Villmow, Micah

I don't know how your target architecture looks like, but I suspect that <4 x i8> should not be legalized to <1 x i32>. I think that what you are seeing is that <4 x i8> is first split into <2 x i8>, and later promoted to <2 x i32>. At the moment different targets can only affect type-legalization by declaring different legal types. A number of us discussed the

v4i8 itself is a legal type, just not on the 'AND' operation. So there seems to be multiple problems here. 1) PromoteVectorOp doesn't handle the case where the types are not the same size, this occurs because #2 2) getTypeToPromoteTo doesn't actual check to see if the type it should promote to makes any sense. 3) PromoteVectorOp also doesn't handle the case where

If v4i8 is a legal type then getTypeToPromoteTo should return the pair v4i8 and 'legal'. This looks like the root of the problem. -----Original Message----- From: Villmow, Micah [mailto:Micah.Villmow at amd.com] Sent: Monday, July 30, 2012 22:10 To: Rotem, Nadav; Developers Mailing List Subject: RE: Vector promotion broken for <2 x [i8|i16]> v4i8 itself is a legal type, just not

Sorry, <4 x i8> should convert to a <1 x i32>. What currently is happening is that it is returning a <2 x i32> because <1 x i32> does not exist. Micah > -----Original Message----- > From: Rotem, Nadav [mailto:nadav.rotem at intel.com] > Sent: Monday, July 30, 2012 10:51 AM > To: Villmow, Micah; Developers Mailing List > Subject: RE: Vector promotion broken

On Tue, Feb 3, 2009 at 10:23 AM, Villmow, Micah <Micah.Villmow at amd.com> wrote: > Is there a way to force llvm to promote all smaller types to i32 instead of > i16? It should just work if i16 is also set to promote... what are you trying to do? -Eli

I want to promote i1 to i32 and not i16 as i32 is my native type and i16 is emulated, but I need to handle i16 as a special case so I don't want to promote it. I will see if what Bill pointed out is what I need. Micah -----Original Message----- From: llvmdev-bounces at cs.uiuc.edu [mailto:llvmdev-bounces at cs.uiuc.edu] On Behalf Of Eli Friedman Sent: Tuesday, February 03, 2009 11:08 AM To:

Dan Gohman wrote: > There certainly are wishlist items for TableGen and TableGen-based > instruction descriptions, though I don't know of an official list. > Offhand, > a few things that come to mind are the ability to handle nodes with > multiple results, Is there an official workaround, BTW? - Volodya

I've patched valuetypes.td/h to add data types that my backend needs to support. There seems to be a lot of assumptions made in other spots of the code that limit the number of data types to 32. I need to add a few more types, but once I go over this limit llvm starts acting wonky. I found all the items that are hard coded to 32 and a section that isn't, but I cannot figure out how to

On my target store/load of f32 or i32 are equivalents. Previously I had duplicate instructions def in .td to map f32 and i32 to the same opcode. I deleted all that and I instead tried a new approach (to simplify things) : setOperationAction(ISD::STORE, MVT::f32, Promote); AddPromotedToType(ISD::STORE, MVT::f32, MVT::i32); setOperationAction(ISD::LOAD, MVT::f32, Promote);

"Villmow, Micah" <Micah.Villmow at amd.com> writes: > Sorry, <4 x i8> should convert to a <1 x i32>. Why? I'm really confused. Shouldn't this converts to a <4 x i32>? -Dave

The comments in the code state it should do bitcast, op, then bitcast, not extend, op and truncate. "SDValue VectorLegalizer::PromoteVectorOp(SDValue Op) { // Vector "promotion" is basically just bitcasting and doing the operation // in a different type. For example, x86 promotes ISD::AND on v2i32 to // v1i64." So following the same logic <4 x i8> bitcasts into a

Micah, I think that your patch is missing the necessary modifications in lib/VMCore/ValueTypes.cpp to EVT::getEVTString() and EVT::getTypeForEVT. -Hal On Fri, 27 Jul 2012 22:54:24 +0000 "Villmow, Micah" <Micah.Villmow at amd.com> wrote: > Vector promotion which is new in LLVM 3.1 is broken for sub32 bit > types. The problem is in the VectorLegalizer::PromoteVectorOp. The

I think that you attached the wrong patch. The attached patch is the one which adds the new MVT types. From: llvmdev-bounces at cs.uiuc.edu [mailto:llvmdev-bounces at cs.uiuc.edu] On Behalf Of Villmow, Micah Sent: Saturday, July 28, 2012 01:54 To: Developers Mailing List Subject: [LLVMdev] Vector promotion broken for <2 x [i8|i16]> Vector promotion which is new in LLVM 3.1 is broken for

On Mon, Aug 05, 2013 at 02:09:58PM -0400, Francois Pichet wrote: > On my target store/load of f32 or i32 are equivalents. > Previously I had duplicate instructions def in .td to map f32 and i32 to > the same opcode. > > I deleted all that and I instead tried a new approach (to simplify things) : > > setOperationAction(ISD::STORE, MVT::f32, Promote); >

On Mon, Aug 5, 2013 at 2:25 PM, Tom Stellard <tom at stellard.net> wrote: > On Mon, Aug 05, 2013 at 02:09:58PM -0400, Francois Pichet wrote: > > On my target store/load of f32 or i32 are equivalents. > > Previously I had duplicate instructions def in .td to map f32 and i32 to > > the same opcode. > > > > I deleted all that and I instead tried a new approach

Micah, One more thing ;) -- llvm::getEnumName in utils/TableGen/CodeGenTarget.cpp -Hal On Tue, 31 Jul 2012 21:02:02 +0000 "Villmow, Micah" <Micah.Villmow at amd.com> wrote: > Ahh yep, thanks for catching that, new patch attached. > > > -----Original Message----- > > From: Hal Finkel [mailto:hfinkel at anl.gov] > > Sent: Tuesday, July 31, 2012 1:40 PM

>Though I am having trouble in understanding how  x86 supports the <1 x i64> type. Based on looking at the code, it should fail because v1i64 is not supported on the x86 platform as >far as I can tell. The Type-Legalizer can handle vector types in the following ways: 1. Split - this splits vectors into two halves. For example on SSE4, <4 x i64> is split to <2 x i64> 2.

The PowerPC backend on PPC64 for non-Darwin (SVR4 ABI) systems currently has a problem handling integer types smaller than 64 bits. This is because the ABI specifies that these types are zero-extended to 64 bits on the stack and the default logic provided in LegalizeDAG does not use that convention. Specifically, for these targets we have: setOperationAction(ISD::VAARG, MVT::Other, Expand);