CODEWITHSUNDEEP
pythonnumpynumba
corvus
corvus

Reputation: 595

python numba.guvectorize fails: "LV: Can't vectorize due to memory conflicts"

I am trying to figure out how to use numba to generate numpy-style ufuncs for vectorized array operations. I noticed I am having very slow performance so I tried to debug by calling the following in my code as per the numba FAQ:

import llvmlite.binding as llvm
llvm.set_option('', '--debug-only=loop-vectorize')

Apparently my loop is not being vectorized 'due to memory conflicts.'

Again from the FAQ page, I see that this occurs 'when memory access pattern is non-trivial.' I'm not clear on what this means, but the code I am trying to vectorize seems pretty trivial to me:

@guvectorize(['void(f4[:,:], b1[:,:], f8, f4, f4[:,:])'],
             '(n,m), (n,m), (), () -> (n,m)', cache=True)
def enforce_cutoff(img, mask, max, nodata, out):
    for i in range(img.shape[0]):
        for j in range(img.shape[1]):
            if mask[i,j]:
                out[i,j] = nodata
            else:
                if img[i,j]<max:
                    out[i,j] = img[i,j]
                else:
                    out[i,j] = max-0.1

Any clues on why this cannot be vectorized and how I could get around it would be very much appreciated. I am pretty new to numba & have no familiarity at all with LLVM so I don't understand much of this well.

The full output from LLVM is here:

LV: Checking a loop in "_ZN7AtmCorr18enforce_cutoff$241E5ArrayIfLi2E1A7mutable7alignedE5ArrayIbLi2E1A7mutable7alignedEdf5ArrayIfLi2E1A7mutable7alignedE" from enforce_cutoff
LV: Loop hints: force=? width=0 unroll=0
LV: Found a loop: B40.us
LV: Found an induction variable.
LV: Found an induction variable.
LV: Can't vectorize due to memory conflicts
LV: Not vectorizing: Cannot prove legality.

LV: Checking a loop in "__gufunc__._ZN7AtmCorr18enforce_cutoff$241E5ArrayIfLi2E1A7mutable7alignedE5ArrayIbLi2E1A7mutable7alignedEdf5ArrayIfLi2E1A7mutable7alignedE" from <numba.npyufunc.wrappers._GufuncWrapper object at 0x0000020A848A6438>
LV: Loop hints: force=? width=0 unroll=0
LV: Not vectorizing: Cannot prove legality.

LV: Checking a loop in "_ZN7AtmCorr18enforce_cutoff$241E5ArrayIfLi2E1A7mutable7alignedE5ArrayIbLi2E1A7mutable7alignedEdf5ArrayIfLi2E1A7mutable7alignedE" from <numba.npyufunc.wrappers._GufuncWrapper object at 0x0000020A848A6438>
LV: Loop hints: force=? width=0 unroll=0
LV: Found a loop: B40.us
LV: Found an induction variable.
LV: Found an induction variable.
LV: Found an induction variable.
LV: Found an induction variable.
LV: Did not find one integer induction var.
LV: Can't vectorize due to memory conflicts
LV: Not vectorizing: Cannot prove legality.

LV: Checking a loop in "_ZN7AtmCorr18enforce_cutoff$241E5ArrayIfLi2E1A7mutable7alignedE5ArrayIbLi2E1A7mutable7alignedEdf5ArrayIfLi2E1A7mutable7alignedE" from <numba.npyufunc.wrappers._GufuncWrapper object at 0x0000020A848A6438>
LV: Loop hints: force=? width=0 unroll=0
LV: Found a loop: B20.us.us
LV: Found an induction variable.
LV: Can't vectorize due to memory conflicts
LV: Not vectorizing: Cannot prove legality.

Upvotes: 1

Views: 270

Answers (1)

max9111
max9111

Reputation: 6482

A possible workaround would be to ensure that the arrays are C-contiguous. If they are not c-contigous they will be copied.

Example

import numba as nb
import numpy as np
@nb.njit(cache=True,parallel=True)
def enforce_cutoff_2(img, mask, max, nodata, out):
    #create a contigous copy if array isn't c-contiguous
    img=np.ascontiguousarray(img)
    mask=np.ascontiguousarray(mask)

    for i in nb.prange(img.shape[0]):
        for j in range(img.shape[1]):
            if mask[i,j]:
                out[i,j] = nodata
            else:
                if img[i,j]<max:
                    out[i,j] = img[i,j]
                else:
                    out[i,j] = max-0.1

Timings

#contiguous arrays
img=np.random.rand(1000,1000).astype(np.float32)
mask=np.random.rand(1000,1000)>0.5
max=0.5
nodata=1.
out=np.empty((img.shape[0],img.shape[1]),dtype=np.float32)

%timeit enforce_cutoff_2(img, mask, max, nodata, out)
#single-thread
#678 µs ± 3.72 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)
#parallel
#143 µs ± 1.87 µs per loop (mean ± std. dev. of 7 runs, 10000 loops each)

#non contigous arrays
img=np.random.rand(2000,2000).astype(np.float32)
mask=np.random.rand(2000,2000)>0.5
img=img[0:-1:2,0:-1:2]
mask=mask[0:-1:2,0:-1:2]
max=0.5
nodata=1.
out=np.empty((img.shape[0],img.shape[1]),dtype=np.float32)

%timeit enforce_cutoff_2(img, mask, max, nodata, out)
#single threaded
#with contiguous copy
#1.78 ms ± 9.58 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)
#without contiguous copy
#5.76 ms ± 20.1 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)

#parallel
#with contiguous copy
#1.42 ms ± 7.03 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)
##without contiguous copy
#1.08 ms ± 75.9 µs per loop (mean ± std. dev. of 7 runs, 1 loop each)

Upvotes: 1

Related Questions