similar to: [LLVMdev] Declare multiple data type for a register class in tblegen

Hi Matt, I did call addRegisterClass in TargetLowering for all the possible types in the register. And for typecasting instructions (i32 to i64), it works. Any other possiblilities? On Thu, Jul 2, 2015 at 6:12 PM Matt Arsenault <Matthew.Arsenault at amd.com> wrote: > On 07/02/2015 05:56 PM, Xiaochu Liu wrote: > > Hi everyone, > > > > I tried to declare multiple data

Oh, they have selection details in the end. Let me check that first... On Sat, Jul 4, 2015 at 4:05 PM Xiaochu Liu <xiaochu1122 at gmail.com> wrote: > Hi Matt, > > I tried debug-only=isel and have some more informations. > The steps before 'Legalized selection'( excluding it) all use v2i32 load. > At the step of 'Legalized selection', it replaced one v2i32

Thanks. I'm gonna try tomorrow and let you know. On Thu, Jul 2, 2015 at 6:51 PM Matt Arsenault <Matthew.Arsenault at amd.com> wrote: > On 07/02/2015 06:41 PM, Xiaochu Liu wrote: > > Hi Matt, > > > > I did call addRegisterClass in TargetLowering for all the possible > > types in the register. And for typecasting instructions (i32 to i64), > > it works.

Hi there, I'm trying to map extract_vector_elt use the following pattern in tbl file. Def : Pat <(i64 (extractelt v2i32:$src, 0)), (i64 (SRLIMM GPR:$src, 32))>; But the tblgen shows : Type inference contradiction found , forcing v2i32 to have a vector element of type i64 But the manual says this instruction allows return type to be larger than element type. Anyone can show me any

I'm having some troubles implementing vector support to our custom backend It seems that llvm cannot emulate shuffle with extracts, inserts and builds? I've enabled vector registers with addRegisterClass(MVT::v2i32, TCE::V2I32RegsRegisterClass); addRegisterClass(MVT::v2f32, TCE::V2F32RegsRegisterClass); and created patterns for most vector instructions, including insert, extract and

Hi Hal, Thanks for your reply and it is helpful! I have a quick question: When I use BuildMI to build instructions in this case, do I have to add all three of the register operands explicitly (operand 0 and 1 are the same)? Thanks, Xiaochu On Tue, Nov 24, 2015 at 3:14 PM, Hal Finkel <hfinkel at anl.gov> wrote: > ----- Original Message ----- >> From: "Xiaochu Liu via

> On Apr 2, 2015, at 2:07 PM, Tom Stellard <tom at stellard.net> wrote: > > On Thu, Apr 02, 2015 at 01:35:55PM -0700, Pete Cooper wrote: >> Hi James, Jim >> >> If you *really* want this to work in selection DAG then there is a solution, but its not pretty. >> >> First make i64 not be legal. Then, assuming the regclass you gave has some subregs, you

I'm not compiling it to x86. Should loop optimizer something independent of the target? If so, should the vectorized code on IR level? On Aug 12, 2016 11:39 AM, "Daniel Berlin" <dberlin at dberlin.org> wrote: > cat > test.c > > #define SIZE 128 > > void bar(int *restrict A, int* restrict B,int K) { > > #pragma clang loop vectorize(enable)

Hi there, I was trying to build llvm on CentOS. The problem is that library on CentOS is always so old (Gcc, python). I managed to install local gcc and let cmake know. But I couldn't do so with python. I end up comment out the check in CMakeList.txt which is not good. I was wondering if anyone has experience building llvm (using cmake) using a local python? Thanks, XIaochu --------------

Hi Marcello, Thanks for your reply. I will try to pass down the mask! I have one more question. In my backend I return CSR_RegMask in getCallPreservedMask and return CSR_SaveList in getCalleeSavedRegs. Is that a correct setup? I dumped the regmask and found that callee saved regs are marked 1 and non-callee saved regs are 0. Thanks, Xiaochu On Wed, Aug 26, 2015 at 5:58 PM Marcello Maggioni

Hi Tim, Thanks for your message! It turns out that the infrastructure (an outdated one) that I am working on is using gcc+dragonegg to generate llvm code: gcc -m32 -S -c -O0 -fplugin=$(DRAGONEGG_SO) -fplugin-arg-dragonegg-emit-ir $< -o $@.tmp It directly generates llvm code with fadd, etc. I'm not familiar with dragonegg plugin... Thanks, XIaochu On Wed, Mar 16, 2016 at 12:00 PM,

Hi Daniel, I increased the size of your test to be 128 but -stats still shows no loop optimized... Xiaochu On Aug 12, 2016 11:11 AM, "Daniel Berlin" <dberlin at dberlin.org> wrote: > It's not possible to know that A and B don't alias in this example. It's > almost certainly not profitable to add a runtime check given the size of > the loop. > > >

Hello everyone, I think I have found an gvn / alias analysis related bug, but before opening an issue on the tracker I wanted to see if I am missing something. I have the following testcase: define spir_kernel void @test(<2 x i32*> %in1, <2 x i32*> %in2, i32* %out) { > entry: > ; Just some temporary storage > %tmp.0 = alloca i32 > %tmp.1 = alloca i32 > %tmp.i =

Dear there, I'm developing an instruction layout like: opcode | rd| ts and its semantics is: rd= rd opcode rs But when I describe it in td file like this: class R<bits<5> Op, string OpcodeStr, list<dag> Pattern> : InstV<(outs GPR:$rd), (ins GPR:$rd, GPR:$rs), !strconcat(OpcodeStr, "\t$rd, $rs"), Pattern> { bits<5> rd; bits<6> rs; let

this is definitely a bug in AA. 225 for (auto I = CS2.arg_begin(), E = CS2.arg_end(); I != E; ++I) { 226 const Value *Arg = *I; 227 if (!Arg->getType()->isPointerTy()) -> 228 continue; 229 unsigned CS2ArgIdx = std::distance(CS2.arg_begin(), I); 230 auto CS2ArgLoc = MemoryLocation::getForArgument(CS2, CS2ArgIdx, TLI);

+ a few others. After following this rabbit hole a bit, there are a lot of mutually recursive calls, etc, that may or may not do the right thing with vectors of pointers. I can fix *this* particular bug with the attached patch. However, it's mostly papering over stuff. Nothing seems to know what to do with a memorylocation that is a vector of pointers. They all expect memorylocation to be a

Okay, so then it sounds like, for now, the right fix is to stop marking masked.gather and masked.scatter with intrarg* options. On Mon, Aug 29, 2016, 1:26 PM Philip Reames <listmail at philipreames.com> wrote: > We might have specification bug here, but we appear to implement what we > specified. argmemonly is specified as only considering pointer typed > arguments. It's

----- Original Message ----- > From: "Daniel Berlin" <dberlin at dberlin.org> > To: "Philip Reames" <listmail at philipreames.com>, "Davide Italiano" > <davide at freebsd.org>, "Chandler Carruth" <chandlerc at gmail.com> > Cc: "Chris Sakalis" <chrissakalis at gmail.com>, "David Majnemer" >

Thank you for the quick fix, I can no longer reproduce the issue. As far a releases go, I am guessing that this is going to be in 4.0? Best, Chris On Tue, Aug 30, 2016 at 9:26 PM, Daniel Berlin <dberlin at dberlin.org> wrote: > Yeah, i just hope it doesn't regress scatter/gather vector code badly. > But at least it's correct now? > > > On Tue, Aug 30, 2016 at 1:11

Great, thank you! On Wed, Aug 31, 2016 at 2:07 PM, Hal Finkel <hfinkel at anl.gov> wrote: > > ------------------------------ > > *From: *"Chris Sakalis" <chrissakalis at gmail.com> > *To: *"Daniel Berlin" <dberlin at dberlin.org> > *Cc: *"Hal Finkel" <hfinkel at anl.gov>, "David Majnemer" < > david.majnemer