Updated to alpha 51

3 files changed, 11 insertions, 300 deletions

diff --git a/.SRCINFO b/.SRCINFO
index c112f0569802..2bc075f927a2 100644
--- a/.SRCINFO
+++ b/.SRCINFO
@@ -1,7 +1,7 @@
 pkgbase = vmd-src
 	pkgdesc = Visual Molecular Dynamics
-	pkgver = 1.9.4a48
-	pkgrel = 2
+	pkgver = 1.9.4a51
+	pkgrel = 1
 	url = http://www.ks.uiuc.edu/Research/vmd/
 	arch = x86_64
 	license = custom
@@ -21,11 +21,10 @@ pkgbase = vmd-src
 	optdepends = cuda: NVIDIA CUDA GPU acceleration functions
 	conflicts = vmd
 	conflicts = vmd-bin
-	source = local://vmd-1.9.4a48.src.tar.gz
+	source = local://vmd-1.9.4a51.src.tar.gz
 	source = configure.patch
 	source = mpi.patch
-	source = wkfthreads_reversal.patch
-	sha256sums = d1ef1260be2bb43aed891832cc2ad15df1bb8402cb47d848073342c3df917ea4
+	sha256sums = b1c40b21111f5bab56d43d5e442c468d327159b07915af2ec175ba6b12842e5c
 	sha256sums = 1a36eeea92a362576fc1167f0bafb12ecf4c97562e0ae4b001743f5de6f0a48c
 	sha256sums = e281a57831b8ff60c5a644219f0b6289d32bee239978af676474941c7d8548c0
 	sha256sums = 1284e19124743fe85969ebbee67e4e4e5b6ec07e9c4c83f75f960db75b0587ad
diff --git a/PKGBUILD b/PKGBUILD
index 5d3e46436514..e9288ab1b249 100644
--- a/PKGBUILD
+++ b/PKGBUILD
@@ -5,8 +5,8 @@
 
 pkgname=vmd-src
 _pkgname=vmd
-pkgver=1.9.4a48
-pkgrel=2
+pkgver=1.9.4a51
+pkgrel=1
 pkgdesc="Visual Molecular Dynamics"
 url="http://www.ks.uiuc.edu/Research/vmd/"
 license=('custom')
@@ -23,9 +23,8 @@ conflicts=("$_pkgname" "$_pkgname-bin")
 # and put it in the PKGBUILD folder.
 source=("local://$_pkgname-${pkgver}.src.tar.gz"
         "configure.patch"
-        "mpi.patch"
-        "wkfthreads_reversal.patch")
-sha256sums=('d1ef1260be2bb43aed891832cc2ad15df1bb8402cb47d848073342c3df917ea4'
+        "mpi.patch")
+sha256sums=('b1c40b21111f5bab56d43d5e442c468d327159b07915af2ec175ba6b12842e5c'
             '1a36eeea92a362576fc1167f0bafb12ecf4c97562e0ae4b001743f5de6f0a48c'
             'e281a57831b8ff60c5a644219f0b6289d32bee239978af676474941c7d8548c0'
             '1284e19124743fe85969ebbee67e4e4e5b6ec07e9c4c83f75f960db75b0587ad')
@@ -37,9 +36,6 @@ prepare() {
   sed -i 's#:${LD_LIBRARY_PATH}/:${LD_LIBRARY_PATH}:#/opt/optix/lib64#g' bin/*
   patch -p0 < ../configure.patch
   
-  # Removing CPU intrinsics from WKFThreads.C as they break compilation
-  patch -p0 < ../wkfthreads_reversal.patch
-  
   # Assuming openmpi; if not the case edit mpi.patch
   patch -p0 < ../mpi.patch
   export TCLINC="-I/usr/include"
@@ -68,7 +64,8 @@ build() {
 
 package() {
   cd "$srcdir/$_pkgname-$pkgver"
-  install -D -m644 LICENSE "$pkgdir/usr/share/licenses/$pkgname/LICENSE"
-  cd src; make install
+  install -Dm644 LICENSE "$pkgdir/usr/share/licenses/$pkgname/LICENSE"
+  cd src
+  make install
   sed -i 's#set defaultvmddir=.*#set defaultvmddir=/usr/lib/vmd#' "$pkgdir/usr/bin/vmd"
 }
diff --git a/wkfthreads_reversal.patch b/wkfthreads_reversal.patch
deleted file mode 100644
index 9850f31212b1..000000000000
--- a/wkfthreads_reversal.patch
+++ /dev/null
@@ -1,285 +0,0 @@
---- src/WKFThreads.C	2020-10-16 19:12:35.700720013 +0300
-+++ src/WKFThreads.C	2020-10-20 23:43:13.354885730 +0300
-@@ -3,7 +3,7 @@
-  *
-  *      $RCSfile: WKFThreads.C,v $
-  *      $Author: johns $        $Locker:  $             $State: Exp $
-- *      $Revision: 1.25 $       $Date: 2020/07/20 04:17:49 $
-+ *      $Revision: 1.20 $       $Date: 2020/02/26 15:43:31 $
-  *
-  ***************************************************************************/
- /**
-@@ -22,7 +22,7 @@
-  *       running the sequence of 'sed' commands shown at the top of the
-  *       "threads.c" file within the Tachyon source distribution.
-  *
-- * $Revision: 1.25 $       $Date: 2020/07/20 04:17:49 $
-+ * $Revision: 1.20 $       $Date: 2020/02/26 15:43:31 $
-  *
-  * \author John E. Stone - john.stone@gmail.com
-  * \copyright
-@@ -233,211 +233,12 @@
- }
- 
- 
--/*
-- * Functions supporting processor-specific runtime dispatch for hand-written
-- * kernels using SIMD vector intrinsics or other highly specialized routines.
-- */
--#define WKF_USEINTCPUID 1
--#if defined(WKF_USEINTCPUID) && (defined(__GNUC__) || defined(__INTEL_COMPILER)) && (defined(__i386__) || defined(__x86_64__))
--#if 1
--//static void wkf_cpuid(uint32_t eax, uint32_t ecx, uint32_t* abcd) {
--static void wkf_cpuid(unsigned int eax, unsigned int ecx, unsigned int* abcd) {
--#if defined(_MSC_VER)
--  __cpuidex((int*)abcd, eax, ecx);
--#else
--//  uint32_t ebx, edx;
--  unsigned int ebx=0, edx=0;
--#if defined(__i386__) && defined (__PIC__)
--  /* in case of PIC under 32-bit EBX cannot be clobbered */
--  __asm__("movl %%ebx, %%edi \n\t cpuid \n\t xchgl %%ebx, %%edi" : "=D" (ebx),
--#else
--  __asm__("cpuid" : "+b" (ebx),
--#endif
--          "+a" (eax), "+c" (ecx), "=d" (edx));
--          abcd[0] = eax; abcd[1] = ebx; abcd[2] = ecx; abcd[3] = edx;
--#endif
--  }
--#else
--static void wkf_cpuid(unsigned int eax, unsigned int ecx, unsigned int *info) {
--  __asm__ __volatile__(
--    "xchg %%ebx, %%edi;"
--    "cpuid;"
--    "xchg %%ebx, %%edi;"
--    :"=a" (info[0]), "=D" (info[1]), "=c" (info[2]), "=d" (info[3])
--    :"0" (eax)
--  );
--}
--#endif
--
--static unsigned long long wkf_xgetbv(unsigned int index) {
--  unsigned int eax=0, edx=0;
--  __asm__ __volatile__(
--    "xgetbv;"
--    : "=a" (eax), "=d"(edx)
--    : "c" (index)
--  );
--  return ((unsigned long long) edx << 32) | eax;
--}
--#endif
--
--
--static int wkf_x86cpu_vector_reg_state(wkf_cpu_caps_t *cpucaps) {
--
--#if defined(WKF_USEINTCPUID) && (defined(__GNUC__) || defined(__INTEL_COMPILER)) && (defined(__i386__) || defined(__x86_64__))
--  unsigned int cpuinfo[4] = { 0 };
--  int havexmmymm = 0;
--  int havezmmmask = 0;
--  int haveosxsave = 0;
--  unsigned long long xcrFeatureMask = 0;
--  unsigned long long xcrXinuse = 0;
--
--  wkf_cpuid(1, 0, cpuinfo);     /* get various SIMD extension flags */
--  haveosxsave = (cpuinfo[2] & (1 << 27)) != 0; /* OS save/restore xmm regs */
--
--  /* if we have AVX, we need to call xgetbv too */
--  if ((cpuinfo[2] & (1 << 28)) != 0) {
--    xcrFeatureMask = wkf_xgetbv(0);
--    havexmmymm  = (xcrFeatureMask & 0x06) == 0x06;
--    havezmmmask = (xcrFeatureMask & 0xE6) == 0xE6;
--    xcrXinuse = wkf_xgetbv(1);
--  }
--
--  printf("wkf_x86cpu_vector_reg_state():\n");
--  printf("  XINUSE %16lx\n", xcrXinuse);
--  printf("    ");
--  if (xcrXinuse & 0x01)
--    printf("x87  ");
--  if (xcrXinuse & 0x02)
--    printf("SSE  ");
--  if (xcrXinuse & 0x04)
--    printf("YMM_Hi128  ");
--  if (xcrXinuse & 0x20)
--    printf("OpMask  ");
--  if (xcrXinuse & 0x40)
--    printf("ZMM_Hi256  ");
--  if (xcrXinuse & 0x80)
--    printf("Hi16_ZMM  ");
--  printf("\n");
--
--#endif
--
--  return 0;   
--}
--
--
- int wkf_cpu_capability_flags(wkf_cpu_caps_t *cpucaps) {
-   int flags=CPU_UNKNOWN;
--  int smtdepth = CPU_SMTDEPTH_UNKNOWN;
--
--#if defined(WKF_USEINTCPUID) && (defined(__GNUC__) || defined(__INTEL_COMPILER)) && (defined(__i386__) || defined(__x86_64__))
--  // https://software.intel.com/content/www/us/en/develop/articles/how-to-detect-new-instruction-support-in-the-4th-generation-intel-core-processor-family.html
--  // https://stackoverflow.com/questions/6121792/how-to-check-if-a-cpu-supports-the-sse3-instruction-set
--  // https://gist.github.com/hi2p-perim/7855506
--  // http://www.hugi.scene.org/online/coding/hugi%2016%20-%20corawhd4.htm
--  // http://www.geoffchappell.com/studies/windows/km/cpu/precpuid.htm
--  // http://www.geoffchappell.com/studies/windows/km/cpu/cpuid/index.htm
--  // https://www.sandpile.org/x86/cpuid.htm
--  unsigned int vendcpuinfo[4] = { 0 };
--  unsigned int cpuinfo[4] = { 0 };
--  unsigned long long xcrFeatureMask = 0;
--  int havexmmymm = 0;
--  int havezmmmask = 0;
--  int haveosxsave = 0;
--
--  wkf_cpuid(0, 0, vendcpuinfo); /* get vendor string, highest function code */
--  if (vendcpuinfo[0] == 0)
--    goto nocpuinfo; /* bail on very primitive CPU type, max fctn code==0 */
--
--  wkf_cpuid(1, 0, cpuinfo);     /* get various SIMD extension flags */
--  haveosxsave = (cpuinfo[2] & (1 << 27)) != 0; /* OS save/restore xmm regs */
--
--  flags = 0;
--  flags |= ((cpuinfo[2] & (1 << 19)) != 0)           * CPU_SSE4_1;
--  flags |= ((cpuinfo[2] & (1 << 29)) != 0)           * CPU_F16C;
--  flags |= ((cpuinfo[2] & (1 << 31)) != 0)           * CPU_HYPERVISOR;
--  flags |= ((cpuinfo[3] & (1 << 26)) != 0)           * CPU_SSE2;
--  flags |= ((cpuinfo[3] & (1 << 28)) != 0)           * CPU_HT;
--
--  /* if we have AVX, we need to call xgetbv too */
--  if ((cpuinfo[2] & (1 << 28)) != 0) {
--    xcrFeatureMask = wkf_xgetbv(0);
--    havexmmymm  = (xcrFeatureMask & 0x06) == 0x06;
--    havezmmmask = (xcrFeatureMask & 0xE6) == 0xE6;
--  }
--
--  flags |= (((cpuinfo[2] & (1 << 28)) != 0) &&
--            havexmmymm && haveosxsave)               * CPU_AVX;
--
--  /* check that we can call CPUID function 7 */
--  if (cpuinfo[0] >= 0x7) {
--    unsigned int extcpuinfo[4] = { 0 };
--    wkf_cpuid(7, 0, extcpuinfo);
--
--    flags |= (((extcpuinfo[1] & (1 << 5)) != 0) &&
--              havexmmymm && haveosxsave)               * CPU_AVX2;
--
--    flags |= (((extcpuinfo[1] & (1 << 16)) != 0) &&
--              havezmmmask && haveosxsave)              * CPU_AVX512F;
--    flags |= (((extcpuinfo[1] & (1 << 26)) != 0) &&
--              havezmmmask && haveosxsave)              * CPU_AVX512PF;
--    flags |= (((extcpuinfo[1] & (1 << 27)) != 0) &&
--              havezmmmask && haveosxsave)              * CPU_AVX512ER;
--    flags |= (((extcpuinfo[1] & (1 << 28)) != 0) &&
--              havezmmmask && haveosxsave)              * CPU_AVX512CD;
--  }
- 
--  smtdepth = 1;
--  if (flags & CPU_HT) {
--#if 1
--    /* XXX correct this for Phi, OS/BIOS settings */
--    smtdepth = 2;
--
--    /* XXX Hack to detect Xeon Phi CPUs since no other CPUs */
--    /* support AVX-512ER or AVX-512PF (yet...)              */        
--    if ((flags & CPU_AVX512ER) && (flags & CPU_AVX512PF)) {
--      smtdepth = 4;
--    }
--#else
--    int logicalcores = (cpuinfo[1] >> 16) && 0xFF;
--    int physicalcores = logicalcores;
--    char vendor[16] = { 0 };
--    ((unsigned *)vendor)[0] = vendcpuinfo[1]; 
--    ((unsigned *)vendor)[1] = vendcpuinfo[3]; 
--    ((unsigned *)vendor)[2] = vendcpuinfo[2]; 
--
--    /* hmm, not quite right yet */
--    if (!strcmp(vendor, "GenuineIntel")) {
--      unsigned int corecpuinfo[4] = { 0 };
--      wkf_cpuid(4, 0, corecpuinfo);
--      physicalcores = ((corecpuinfo[0] >> 26) & 0x3f) + 1;
--    } else if (!strcmp(vendor, "AuthenticAMD")) {
--      unsigned int corecpuinfo[4] = { 0 };
--      wkf_cpuid(0x80000008, 0, corecpuinfo);
--      physicalcores = (corecpuinfo[2] & 0xFF) + 1;        
--    }
--
--printf("cpuinfo: %d / %d  vend: %s\n", logicalcores, physicalcores, vendor);
--
--    smtdepth = logicalcores / physicalcores; 
--#endif
--  }  
--
--#if 1
--  __m256 qq = _mm256_set1_ps(3.14159f);
--  __m512 qqqq = _mm512_set1_ps(3.141f); 
--  float q1 = ((float *) &qq)[0];
--  float q2 = ((float *) &qqqq)[0];
--//  _mm256_zeroupper();
--  printf("  q1: %f  q2: %f\n\n", q1, q2);
--#endif
--wkf_x86cpu_vector_reg_state(cpucaps);
--
--#elif defined(__INTEL_COMPILER) && (__INTEL_COMPILER >= 1300)
--
--  // https://software.intel.com/content/www/us/en/develop/documentation/cpp-compiler-developer-guide-and-reference/top/compiler-reference/intrinsics/intrinsics-for-all-intel-architectures/may-i-use-cpu-feature.html
-+#if defined(__INTEL_COMPILER) && (__INTEL_COMPILER >= 1300)
-   flags = 0;
-   flags |= _may_i_use_cpu_feature(_FEATURE_SSE2)     * CPU_SSE2;
--  flags |= _may_i_use_cpu_feature(_FEATURE_SSE4_1)   * CPU_SSE4_1;
-   flags |= _may_i_use_cpu_feature(_FEATURE_AVX)      * CPU_AVX;
-   flags |= _may_i_use_cpu_feature(_FEATURE_AVX2)     * CPU_AVX2;
-   flags |= _may_i_use_cpu_feature(_FEATURE_FMA)      * CPU_FMA;
-@@ -445,27 +246,9 @@
-   flags |= _may_i_use_cpu_feature(_FEATURE_AVX512CD) * CPU_AVX512CD;
-   flags |= _may_i_use_cpu_feature(_FEATURE_AVX512ER) * CPU_AVX512ER;
-   flags |= _may_i_use_cpu_feature(_FEATURE_AVX512PF) * CPU_AVX512PF;
--
--#elif defined(__GNUC__) && (defined(__i386__) || defined(__x86_64__))
--
--  // https://gcc.gnu.org/onlinedocs/gcc/x86-Built-in-Functions.html
--  flags = 0;
--  __builtin_cpu_init();
--  flags |= (__builtin_cpu_supports("sse2")!=0)       * CPU_SSE2;
--  flags |= (__builtin_cpu_supports("sse4.1")!=0)     * CPU_SSE4_1;
--  flags |= (__builtin_cpu_supports("avx")!=0)        * CPU_AVX;
--  flags |= (__builtin_cpu_supports("avx2")!=0)       * CPU_AVX2;
--  flags |= (__builtin_cpu_supports("fma")!=0)        * CPU_FMA;
--  flags |= (__builtin_cpu_supports("avx512f")!=0)    * CPU_AVX512F;
--  flags |= (__builtin_cpu_supports("avx512cd")!=0)   * CPU_AVX512CD;
--  flags |= (__builtin_cpu_supports("avx512er")!=0)   * CPU_AVX512ER;
--  flags |= (__builtin_cpu_supports("avx512pf")!=0)   * CPU_AVX512PF;
--
- #endif
- 
--nocpuinfo:
-   cpucaps->flags = flags;
--  cpucaps->smtdepth = smtdepth;
- 
-   if (flags == CPU_UNKNOWN)
-     return 1;
-@@ -474,23 +257,6 @@
- }
- 
- 
--int wkf_cpu_smt_depth(void) {
--  int smtdepth = CPU_SMTDEPTH_UNKNOWN;
--
--#if defined(WKF_USEINTCPUID) && (defined(__GNUC__) || defined(__INTEL_COMPILER)) && (defined(__i386__) || defined(__x86_64__))
--  // x86 examples:
--  //  https://software.intel.com/en-us/articles/methods-to-utilize-intels-hyper-threading-technology-with-linux
--  // https://stackoverflow.com/questions/2901694/how-to-detect-the-number-of-physical-processors-cores-on-windows-mac-and-linu
--  wkf_cpu_caps_t cpucaps;
--  if (!wkf_cpu_capability_flags(&cpucaps)) {
--    smtdepth = cpucaps.smtdepth;
--  }
--#endif
--
--  return smtdepth;
--}
--
--
- int * wkf_cpu_affinitylist(int *cpuaffinitycount) {
-   int *affinitylist = NULL;
-   *cpuaffinitycount = -1; /* return count -1 if unimplemented or err occurs */