NENormalizationLayer fails to run

[joey2014]

Test environment :
android M+ arm_compute v17.04
in some case, NENormalizationLayer fails to run. I provide two cases one is workable case and the other is failure case in attach code.
test.zip

build the test sample:
aarch64-linux-android-g++ norm_test1.cpp -std=c++11 -I. -pie -o test2 -Lbuild -larm_compute

terminate called after throwing an instance of 'std::runtime_error'
what(): in void arm_compute::Window::validate() const ./arm_compute/core/Window.inl:73: _dims[i].end() <= _dims[i].start()
Aborted

call stack:
04-26 04:36:37.522 309 309 F DEBUG : backtrace:
04-26 04:36:37.522 309 309 F DEBUG : #00 pc 000000000006a8e4 /system/lib64/libc.so (tgkill+8)
04-26 04:36:37.522 309 309 F DEBUG : #1 pc 0000000000068074 /system/lib64/libc.so (pthread_kill+68)
04-26 04:36:37.522 309 309 F DEBUG : #2 pc 0000000000020e08 /system/lib64/libc.so (raise+28)
04-26 04:36:37.522 309 309 F DEBUG : #3 pc 000000000001b5a8 /system/lib64/libc.so (abort+60)
04-26 04:36:37.522 309 309 F DEBUG : #4 pc 00000000001c1744 /data/local/tmp/libarm_compute.so (_ZN9__gnu_cxx27__verbose_terminate_handlerEv+348)
04-26 04:36:37.523 309 309 F DEBUG : #5 pc 000000000016fac8 /data/local/tmp/libarm_compute.so (_ZN10__cxxabiv111__terminateEPFvvE+8)
04-26 04:36:37.523 309 309 F DEBUG : #6 pc 000000000016fb34 /data/local/tmp/libarm_compute.so (_ZSt9terminatev+12)
04-26 04:36:37.523 309 309 F DEBUG : #7 pc 000000000016fc70 /data/local/tmp/libarm_compute.so (__cxa_throw+136)
04-26 04:36:37.523 309 309 F DEBUG : #8 pc 0000000000050340 /data/local/tmp/libarm_compute.so (_ZN11arm_compute5errorEPKcS1_iS1_z+320)
04-26 04:36:37.523 309 309 F DEBUG : #9 pc 000000000005158c /data/local/tmp/libarm_compute.so (_ZNK11arm_compute21AccessWindowRectangle23update_window_if_neededERNS_6WindowE+1068)
04-26 04:36:37.523 309 309 F DEBUG : #10 pc 00000000000662b8 /data/local/tmp/libarm_compute.so (ZN11arm_compute25update_window_and_paddingIJNS_22AccessWindowHorizontalES1_RS1_EEEbRNS_6WindowEDpOT+52)
04-26 04:36:37.523 309 309 F DEBUG : #11 pc 000000000012a7fc /data/local/tmp/libarm_compute.so (_ZN11arm_compute31NEPixelWiseMultiplicationKernel9configureEPKNS_7ITensorES3_PS1_fNS_13ConvertPolicyENS_14RoundingPolicyE+1220)
04-26 04:36:37.523 309 309 F DEBUG : #12 pc 000000000016935c /data/local/tmp/libarm_compute.so (_ZN11arm_compute20NENormalizationLayer9configureEPKNS_7ITensorEPS1_NS_22NormalizationLayerInfoE+204)
04-26 04:36:37.524 309 309 F DEBUG : #13 pc 0000000000002bc4 /data/local/tmp/test2 (main+768)
04-26 04:36:37.524 309 309 F DEBUG : #14 pc 00000000000196ac /system/lib64/libc.so (__libc_init+100)
04-26 04:36:37.524 309 309 F DEBUG : #15 pc 00000000000027cc /data/local/tmp/test2 (do_arm64_start+60)

Can you help check the issue?

[mpflanzer]

In the failure case you allocate before calling configure(). In that case configure() cannot increase the padding but instead shrinks the execution window. Since your tensor is only 13x13 and we process 16 elements per row the window will be reduced to zero.

We already have a patch to improve the error message in that case. It will be included in the next release.

[mpflanzer]

Please feel to re-open if your problem still remains after allocating after the configuration.