running formatter (#126)

* running formatter

* consolidate CI runs

* use cluster to run formatter

* use cluster to run formatter

* Add a CI-Success step

6 files changed, 29 insertions, 16 deletions

diff --git a/src/cuda-sim/cuda-sim.cc b/src/cuda-sim/cuda-sim.cc
index 69d1eb7..14ae504 100644
--- a/src/cuda-sim/cuda-sim.cc
+++ b/src/cuda-sim/cuda-sim.cc
@@ -2782,7 +2782,9 @@ void print_ptxinfo() {
   }
 }
 
-struct gpgpu_ptx_sim_info get_ptxinfo() { return g_ptxinfo; }
+struct gpgpu_ptx_sim_info get_ptxinfo() {
+  return g_ptxinfo;
+}
 
 std::map<unsigned, const char *> get_duplicate() { return g_duplicate; }
 
diff --git a/src/cuda-sim/cuda-sim.h b/src/cuda-sim/cuda-sim.h
index b1caf0c..21e1ca0 100644
--- a/src/cuda-sim/cuda-sim.h
+++ b/src/cuda-sim/cuda-sim.h
@@ -101,8 +101,8 @@ class functionalCoreSim : public core_t {
   bool *m_warpAtBarrier;
 };
 
-#define RECONVERGE_RETURN_PC ((address_type) - 2)
-#define NO_BRANCH_DIVERGENCE ((address_type) - 1)
+#define RECONVERGE_RETURN_PC ((address_type)-2)
+#define NO_BRANCH_DIVERGENCE ((address_type)-1)
 address_type get_return_pc(void *thd);
 const char *get_ptxinfo_kname();
 void print_ptxinfo();
diff --git a/src/cuda-sim/cuda_device_printf.cc b/src/cuda-sim/cuda_device_printf.cc
index 69ebc63..ef780e9 100644
--- a/src/cuda-sim/cuda_device_printf.cc
+++ b/src/cuda-sim/cuda_device_printf.cc
@@ -65,7 +65,7 @@ void my_cuda_printf(const char *fmtstr, const char *arg_list) {
         fprintf(fp, buf, *((unsigned *)ptr));
         arg_offset += 4;
       } else if (c == 'f') {
-        if (acc%2) {
+        if (acc % 2) {
           arg_offset += 4;
           ptr = (void *)&arg_list[arg_offset];
         }
diff --git a/src/cuda-sim/half.h b/src/cuda-sim/half.h
index 67e607b..fab1a22 100644
--- a/src/cuda-sim/half.h
+++ b/src/cuda-sim/half.h
@@ -846,8 +846,10 @@ uint16 int2half_impl(T value) {
       bits |= 0x7BFF + (R != std::round_toward_zero);

   } else if (value) {

     unsigned int m = value, exp = 24;

-    for (; m < 0x400; m <<= 1, --exp);

-    for (; m > 0x7FF; m >>= 1, ++exp);

+    for (; m < 0x400; m <<= 1, --exp)

+      ;

+    for (; m > 0x7FF; m >>= 1, ++exp)

+      ;

     bits |= (exp << 10) + m;

     if (exp > 24) {

       if (R == std::round_to_nearest)

@@ -1272,7 +1274,8 @@ inline double half2float_impl(uint16 value, double, true_type) {
   int abs = value & 0x7FFF;

   if (abs) {

     hi |= 0x3F000000 << static_cast<unsigned>(abs >= 0x7C00);

-    for (; abs < 0x400; abs <<= 1, hi -= 0x100000);

+    for (; abs < 0x400; abs <<= 1, hi -= 0x100000)

+      ;

     hi += static_cast<uint32>(abs) << 10;

   }

   uint64 bits = static_cast<uint64>(hi) << 32;

@@ -2113,7 +2116,8 @@ struct functions {
   static half frexp(half arg, int *exp) {

     int m = arg.data_ & 0x7FFF, e = -14;

     if (m >= 0x7C00 || !m) return *exp = 0, arg;

-    for (; m < 0x400; m <<= 1, --e);

+    for (; m < 0x400; m <<= 1, --e)

+      ;

     return *exp = e + (m >> 10),

            half(binary, (arg.data_ & 0x8000) | 0x3800 | (m & 0x3FF));

   }

@@ -2131,7 +2135,8 @@ struct functions {
     unsigned int mask = (1 << (25 - e)) - 1, m = arg.data_ & mask;

     iptr->data_ = arg.data_ & ~mask;

     if (!m) return half(binary, arg.data_ & 0x8000);

-    for (; m < 0x400; m <<= 1, --e);

+    for (; m < 0x400; m <<= 1, --e)

+      ;

     return half(binary, static_cast<uint16>((arg.data_ & 0x8000) | (e << 10) |

                                             (m & 0x3FF)));

   }

@@ -2143,7 +2148,8 @@ struct functions {
   static half scalbln(half arg, long exp) {

     unsigned int m = arg.data_ & 0x7FFF;

     if (m >= 0x7C00 || !m) return arg;

-    for (; m < 0x400; m <<= 1, --exp);

+    for (; m < 0x400; m <<= 1, --exp)

+      ;

     exp += m >> 10;

     uint16 value = arg.data_ & 0x8000;

     if (exp > 30) {

@@ -2185,7 +2191,8 @@ struct functions {
     if (abs < 0x7C00) {

       int exp = (abs >> 10) - 15;

       if (abs < 0x400)

-        for (; abs < 0x200; abs <<= 1, --exp);

+        for (; abs < 0x200; abs <<= 1, --exp)

+          ;

       return exp;

     }

     if (abs > 0x7C00) return FP_ILOGBNAN;

@@ -2201,11 +2208,13 @@ struct functions {
     if (abs < 0x7C00) {

       int exp = (abs >> 10) - 15;

       if (abs < 0x400)

-        for (; abs < 0x200; abs <<= 1, --exp);

+        for (; abs < 0x200; abs <<= 1, --exp)

+          ;

       uint16 bits = (exp < 0) << 15;

       if (exp) {

         unsigned int m = std::abs(exp) << 6, e = 18;

-        for (; m < 0x400; m <<= 1, --e);

+        for (; m < 0x400; m <<= 1, --e)

+          ;

         bits |= (e << 10) + m;

       }

       return half(binary, bits);

diff --git a/src/cuda-sim/ptx_sim.cc b/src/cuda-sim/ptx_sim.cc
index 9a22d02..d0de1ca 100644
--- a/src/cuda-sim/ptx_sim.cc
+++ b/src/cuda-sim/ptx_sim.cc
@@ -203,8 +203,8 @@ unsigned ptx_thread_info::get_builtin(int builtin_id, unsigned dim_mod) {
       return (unsigned)(m_gpu->gpu_sim_cycle + m_gpu->gpu_tot_sim_cycle);
     case CLOCK64_REG:
       // Change return value to unsigned long long?
-      // Currently returns 32-bit unsigned, which may cause truncation for large values.
-      // GPGPUSim clock is 4 times slower - multiply by 4
+      // Currently returns 32-bit unsigned, which may cause truncation for large
+      // values. GPGPUSim clock is 4 times slower - multiply by 4
       return (m_gpu->gpu_sim_cycle + m_gpu->gpu_tot_sim_cycle) * 4;
     case HALFCLOCK_ID:
       // GPGPUSim clock is 4 times slower - multiply by 4
diff --git a/src/cuda-sim/ptx_sim.h b/src/cuda-sim/ptx_sim.h
index 7128f8e..8eec922 100644
--- a/src/cuda-sim/ptx_sim.h
+++ b/src/cuda-sim/ptx_sim.h
@@ -340,7 +340,9 @@ class ptx_thread_info {
   dim3 get_ctaid() const { return m_ctaid; }
   dim3 get_tid() const { return m_tid; }
   dim3 get_ntid() const { return m_ntid; }
-  class gpgpu_sim *get_gpu() { return (gpgpu_sim *)m_gpu; }
+  class gpgpu_sim *get_gpu() {
+    return (gpgpu_sim *)m_gpu;
+  }
   unsigned get_hw_tid() const { return m_hw_tid; }
   unsigned get_hw_ctaid() const { return m_hw_ctaid; }
   unsigned get_hw_wid() const { return m_hw_wid; }