"Illegal instruction" when running simple array additions #10330
Description
Numpy crashes with an "illegal instruction" when running a simple array addition (past a certain array size) when running inside a VM under Xen. It is probably somehow related to #9534. I narrowed it down the commit 37740eb, which introduced AVX2 simd code:
Python 3.5.2 (default, Nov 23 2017, 16:37:01)
[GCC 5.4.0 20160609] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy
>>> numpy.__version__
'1.12.0.dev0+37740eb'
>>> a = numpy.arange(15)
>>> a[1:] + a[:-1]
Illegal instruction (core dumped)I also tried the current master, which shows the same error.
It is working fine up to the preceding commit eeb4e17:
Python 3.5.2 (default, Nov 23 2017, 16:37:01)
[GCC 5.4.0 20160609] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy
>>> numpy.__version__
'1.12.0.dev0+eeb4e17'
>>> a = numpy.arange(15)
>>> a[1:] + a[:-1]
array([ 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27])`Note that /proc/cpuinfo doesn't list AVX2:
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 63
model name : Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
stepping : 2
microcode : 0x31
cpu MHz : 2494.282
cache size : 30720 KB
physical id : 0
siblings : 4
core id : 0
cpu cores : 4
apicid : 0
initial apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 13
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx rdtscp lm constant_tsc rep_good nopl pni pclmulqdq ssse3 cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes rdrand hypervisor lahf_lm abm fsgsbase tsc_adjust bmi1 smep bmi2 erms invpcid
bugs :
bogomips : 4988.56
clflush size : 64
cache_alignment : 64
address sizes : 46 bits physical, 48 bits virtual
power management:
processor : 1
vendor_id : GenuineIntel
cpu family : 6
model : 63
model name : Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
stepping : 2
microcode : 0x31
cpu MHz : 2494.282
cache size : 30720 KB
physical id : 0
siblings : 4
core id : 2
cpu cores : 4
apicid : 2
initial apicid : 2
fpu : yes
fpu_exception : yes
cpuid level : 13
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx rdtscp lm constant_tsc rep_good nopl pni pclmulqdq ssse3 cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes rdrand hypervisor lahf_lm abm fsgsbase tsc_adjust bmi1 smep bmi2 erms invpcid
bugs :
bogomips : 4988.56
clflush size : 64
cache_alignment : 64
address sizes : 46 bits physical, 48 bits virtual
power management:
processor : 2
vendor_id : GenuineIntel
cpu family : 6
model : 63
model name : Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
stepping : 2
microcode : 0x31
cpu MHz : 2494.282
cache size : 30720 KB
physical id : 0
siblings : 4
core id : 4
cpu cores : 4
apicid : 4
initial apicid : 4
fpu : yes
fpu_exception : yes
cpuid level : 13
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx rdtscp lm constant_tsc rep_good nopl pni pclmulqdq ssse3 cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes rdrand hypervisor lahf_lm abm fsgsbase tsc_adjust bmi1 smep bmi2 erms invpcid
bugs :
bogomips : 4988.56
clflush size : 64
cache_alignment : 64
address sizes : 46 bits physical, 48 bits virtual
power management:
processor : 3
vendor_id : GenuineIntel
cpu family : 6
model : 63
model name : Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
stepping : 2
microcode : 0x31
cpu MHz : 2494.282
cache size : 30720 KB
physical id : 0
siblings : 4
core id : 6
cpu cores : 4
apicid : 6
initial apicid : 6
fpu : yes
fpu_exception : yes
cpuid level : 13
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx rdtscp lm constant_tsc rep_good nopl pni pclmulqdq ssse3 cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes rdrand hypervisor lahf_lm abm fsgsbase tsc_adjust bmi1 smep bmi2 erms invpcid
bugs :
bogomips : 4988.56
clflush size : 64
cache_alignment : 64
address sizes : 46 bits physical, 48 bits virtual
power management:
I tried to understand how numpy decides when to use the avx2 simd codepaths but failed miserably.