Arch Linux User Repository

Thank you @bulletmark, that does seem to work. Successfully ran

./stockfish bench 16 1 13 default depth mixed

@setone, yes that check is primitive and my fix here is primitive as well but should work. Can you please try it: http://ix.io/3tdC

Hello, I have an older AMD processor:

Architecture:            x86_64
  CPU op-mode(s):        32-bit, 64-bit
  Address sizes:         40 bits physical, 48 bits virtual
  Byte Order:            Little Endian
CPU(s):                  4
  On-line CPU(s) list:   0-3
Vendor ID:               AuthenticAMD
  Model name:            AMD A8-3820 APU with Radeon(tm) HD Graphics
    CPU family:          18
    Model:               1
    Thread(s) per core:  1
    Core(s) per socket:  4
    Socket(s):           1
    Stepping:            0
    Frequency boost:     enabled
    CPU max MHz:         2500.0000
    CPU min MHz:         800.0000
    BogoMIPS:            5002.60
    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 mmxext fxsr_opt pdpe1gb rdtscp lm 3dnowext 3dnow constant_tsc rep_good nopl nonstop_tsc cpuid extd_api
                         cid pni cx16 popcnt lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt cpb hw_pstate vmmcall arat npt lbrv svm_lock nrip_save pausefilter
Virtualization features: 
  Virtualization:        AMD-V
Caches (sum of all):     
  L1d:                   256 KiB (4 instances)
  L1i:                   256 KiB (4 instances)
  L2:                    4 MiB (4 instances)
NUMA:                    
  NUMA node(s):          1
  NUMA node0 CPU(s):     0-3
Vulnerabilities:         
  Itlb multihit:         Not affected
  L1tf:                  Not affected
  Mds:                   Not affected
  Meltdown:              Not affected
  Spec store bypass:     Not affected
  Spectre v1:            Mitigation; usercopy/swapgs barriers and __user pointer sanitization
  Spectre v2:            Mitigation; Full AMD retpoline, STIBP disabled, RSB filling
  Srbds:                 Not affected
  Tsx async abort:       Not affected

When I build using makepkg, PKGBUILD incorrectly identifies my architecture as x86-64-modern. This causes stockfish to crash in smmintrin.h:

281│ extern __inline __m128i __attribute__((__gnu_inline__, __always_inline__, __artificial__))
282│ _mm_max_epi8 (__m128i __X, __m128i __Y)
283│ {
284├─> return (__m128i) __builtin_ia32_pmaxsb128 ((__v16qi)__X, (__v16qi)__Y);
285│ }

I can work around the problem by commenting out 2 lines in PKGBUILD:

 30 #  elif grep -q popcnt /proc/cpuinfo; then
 31 #    _arch=x86-64-modern

which causes my architecture to be correctly identified as x86-64. So there seems to be an issue with the test.

Ah, my mistake. I wasn't sure. Thanks for the heads up.

@tjackson, note (e.g. for next time) there is no need or point bumping the pkgrel just because you add a new arch.

Hi, could add armv7h as well? I'm building a raspberry pi 4 to be my chesscomputer with SCID. (I simply added 'armv7h' to the PKGBUILD and it builts, install, and works just fine. Thank you.

The build also supports Clang if someone wants/needs it. Just change COMP=gcc to COMP=clang, I believe.

I fear I've done this horribly wrong but I added armv7 and bmi2 instructions. Feedback greatly welcome.

Could you also enable BMI2 instructions for those processors that support it? This provides a couple % speedup (versus x86-64-modern).

    elif grep bmi2 /proc/cpuinfo 2>&1; then
        _arch=x86-64-bmi2