similar to: [LLVMdev] bb-vectorizer transforms only part of the block

Hi all! It takes the current SLP vectorizer too long to vectorize my scalar code. I am talking here about functions that have a single, huge basic block with O(10^6) instructions. Here's an example: %0 = getelementptr float* %arg1, i32 49 %1 = load float* %0 %2 = getelementptr float* %arg1, i32 4145 %3 = load float* %2 %4 = getelementptr float* %arg2, i32 49 %5 = load

I seem to have problem to get the SLP vectorizer to make use of the full 8 floats available in a SIMD vector on a Sandy Bridge CPU with AVX. The function is attached, the CPU flags are: flags : fpu vme de pse tsc msr pae mce cx8 apic mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good

I looked more into this. For the previously sent IR the vector width of 256 bit is found mistakenly (and reproducibly) on this hardware: model name : Intel(R) Xeon(R) CPU E5-2650 0 @ 2.00GHz For the same IR the loop vectorizer finds the correct vector width (128 bit) on: model name : Intel(R) Xeon(R) CPU E5630 @ 2.53GHz model name : Intel(R) Core(TM) i7 CPU M 640 @

Hi Frank, I'm not an Intel expert, but it seems that your Xeon E5 supports AVX, which does have 256-bit vectors. The other two only supports SSE instructions, which are only 128-bit long. cheers, --renato On 10 November 2013 06:05, Frank Winter <fwinter at jlab.org> wrote: > I looked more into this. For the previously sent IR the vector width of > 256 bit is found mistakenly

The loop vectorizer is doing an amazing job so far. Most of the time. I just came across one function which led to unexpected behavior: On this function the loop vectorizer finds a 256 bit vector as the wides vector type for the x86-64 architecture. (!) This is strange, as it was always finding the correct size of 128 bit as the widest type. I isolated the IR of the function to check if this is

Hey guys, in an HPC project I am working on I am given an LLVM program consisting of a linear sequence of repetitive junks of code with an uniform memory access pattern. Each code junk does the following: 1) loads some memory, 2) performs some arithmetic operations, 3) stores the result back to memory. The memory stride between consecutive junks is constant over the whole program, thus the

The loop vectorizer relies on cleanup passes to be run after it: from Transforms/IPO/PassManagerBuilder.cpp: // Add the various vectorization passes and relevant cleanup passes for // them since we are no longer in the middle of the main scalar pipeline. MPM.add(createLoopVectorizePass(DisableUnrollLoops)); MPM.add(createInstructionCombiningPass());

The instcombine pass cleans up a lot. Any idea why there are still shufflevector, insertelement, *and* bitcast (!!) etc. instructions left? The original loop is so clean, a textbook example I'd say. There is no need to shuffle anything.At least I don't see it. Frank vector.ph: ; preds = %L5 %broadcast.splatinsert1 = insertelement <4 x

The following IR implements the following nested loop: for (int i = start ; i < end ; ++i ) for (int p = 0 ; p < 4 ; ++p ) a[i*4+p] = b[i*4+p] + c[i*4+p]; define void @main(i64 %arg0, i64 %arg1, i1 %arg2, i64 %arg3, float* noalias %arg4, float* noalias %arg5, float* noalias %arg6) { entrypoint: br i1 %arg2, label %L0, label %L1 L0:

I am trying a setup where the one loop is rewritten as two loops. This avoids the 'rem' and 'div' instructions in the index calculation (which give the loop vectorizer a hard time). However, with this setup the loop vectorizer complains about a too small loop. LV: Checking a loop in "main" LV: Found a loop: L3 LV: Found a loop with a very small trip count. This loop

Hi Renato, you are right! There is 'avx' support: fpu vme de pse tsc msr pae mce cx8 apic mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good xtopology nonstop_tsc aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 cx16 xtpr pdcm dca sse4_1 sse4_2 x2apic popcnt aes xsave

In the case when coming from C it was probably the loop unroller and SLP vectorizer which vectorized the code. Potentially I could do the same in the IR. However, the loop body that is generated in the IR can get very large. Thus, the loop unroller will refuse to unroll the loop in a large number of (important) cases. Isn't there a way to convince the loop vectorizer that it should

Hi! When compiling the attached module with the JIT engine on an Intel KNL I see wrong code getting emitted. I attach a complete exploit program which shows the bug in LLVM 3.8. It loads and JIT compiles the module and prints the assembler. I stumbled on this since the result of an actual calculation was wrong. So, it's not only the text version of the assembler also the machine

Frank, It looks like the loop vectorizer is unable to tell that the two stores in your code never overlap. This is probably because of the sign-extend in your code. Can you extend the indices to 64bit ? Thanks, Nadav On Oct 28, 2013, at 1:38 PM, Frank Winter <fwinter at jlab.org> wrote: > Verifying function > running passes ... > LV: Checking a loop in "bar" > LV:

Verifying function running passes ... LV: Checking a loop in "bar" LV: Found a loop: L0 LV: Found an induction variable. LV: We need to do 0 pointer comparisons. LV: Checking memory dependencies LV: Bad stride - Not an AddRecExpr pointer %13 = getelementptr float* %arg2, i32 %1 SCEV: ((4 * (sext i32 {(256 + %arg0),+,1}<nw><%L0> to i64)) + %arg2) LV: Src Scev: {((4 * (sext

MCJIT when lowering to x86-64 generates a MOVAPS (Move Aligned Packed Single-Precision Floating-Point Values) on a non-aligned memory address: movaps 88(%rdx), %xmm0 where %rdx comes in as a function argument with only natural alignment (float*). This x86 instruction requires the memory address to be 16 byte aligned which 88 plus something aligned to 4 byte isn't. Here the

I am trying to change this define void @main(float* noalias %arg0, float* noalias %arg1, float* noalias %arg2) { entrypoint: %0 = bitcast float* %arg1 to <4 x float>* intothis define void @main(float* noalias %arg0, float* noalias %arg1, float* noalias %arg2) { entrypoint: %0 = getelementptr float* %arg1, i64 0 %1 = bitcast float* %0 to <4 x float>* I must be close but

Yes, you need the latest ToT version of llvm or you run -loop-vectorize -earlycse -instcombine -simplifycfg The bitcast essentially is a noop to satisfy the type system. This is how your example looks like for me: vector.body: ; preds = %vector.body, %vector.ph %index = phi i64 [ 0, %vector.ph ], [ %index.next, %vector.body ] %.lhs = shl i64 %6, 2

Hi Frank, What does --debug-only=vectorize says? You may try to get the datalayout and the triple on the IR header, just to make sure you got everything right. LLVM will honour those, and front-ends should create them correctly. --renato On 1 July 2015 at 19:06, Frank Winter <fwinter at jlab.org> wrote: > I realized that the function parameters had no alignment attributes on them.

Frank, It sounds like the SLP vectorizer thinks that it is more profitable to use 128bit wide operations (because 256bit operations are double pumped on Sandybridge). Did you see a different result on Haswell? Thanks, Nadav > On Jul 1, 2015, at 11:06 AM, Frank Winter <fwinter at jlab.org> wrote: > > I realized that the function parameters had no alignment attributes on them.