[Bug c++/21314] New: C++ performance regression with -Os

["deemkay at gmail dot com" <gcc-bugzilla at gcc dot gnu dot org>]

the following code takes more than twice as long to run when compiled with -Os
using gcc-4.0.0 than it does with 3.3.5 or 3.4.3.


#include <stdio.h>
#include <vector>

class mytype_A {
 public:
  int id;
  mytype_A():id(0) {}
};

class mytype_B {
 public:
  mytype_A* A;
  mytype_B(mytype_A* p):A(p) {}
};

class mytype_C {
 public:
  mytype_B* B;
  mytype_C(mytype_B* p):B(p) {}
};


class mytype_D {   
 public:
  // mytype_C* C[2];          // less performance difference if we use simple arrays
  std::vector<mytype_C*> C;   
  int junk[3];                // affects performance

 public:
  mytype_D(mytype_A* a0, mytype_A* a1) {
    //    C[0] = new mytype_C(new mytype_B(a0));
    //    C[1] = new mytype_C(new mytype_B(a0));
    C.push_back(new mytype_C(new mytype_B(a0)));
    C.push_back(new mytype_C(new mytype_B(a0)));
  }
};

int main() {
  int k = 5000;                    // run-time not linear in k
  mytype_A* A[k];
  mytype_D* D[k];
  for (int i=0;i<=k;i++)
    A[i] = new mytype_A();
  for (int i=0;i<k;i++)
    D[i] = new mytype_D(A[i],A[k-i]);    // intentionally make some pointers
farther apart

  clock_t before = clock();

  int k0 = 0;
  for (int i=0;i<k;i++) {
    k0 = 0;
    for (int j=0;j<k;j++) {         // run through list of D's, and reference
pointers
      mytype_D* d = D[j];
      if (d->C[0]->B->A->id)         k0++;
      if (d->C[1]->B->A->id)         k0++;
    }
  }
  printf("%d\n",k0);                // don't allow compiler to optimize away k0

  printf("time: %f\n",(double)(clock()-before)/CLOCKS_PER_SEC);

  return 0;
}


dirtyepic ~/work $ g++ -Os -Wall -march=pentium3 cpptest.cpp

i686-pc-linux-gnu-3.3.5-20050130 - time: 6.530000
i686-pc-linux-gnu-3.4.3-20050110 - time: 6.920000
i686-pc-linux-gnu-4.0.0(release) - time: 16.790000
i686-pc-linux-gnu-4.0.0-20050430 - time: 16.700000


dirtyepic ~/work $ g++ -v -save-temps -Os -Wall -march=pentium3 cpptest.cpp -o
gcc401
Using built-in specs.
Target: i686-pc-linux-gnu
Configured with:
/var/tmp/portage/gcc-4.0.1_alpha20050430/work/gcc-4.0-20050430/configure
--enable-version-specific-runtime-libs --prefix=/usr
--bindir=/usr/i686-pc-linux-gnu/gcc-bin/4.0.1-alpha20050430
--includedir=/usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/include
--datadir=/usr/share/gcc-data/i686-pc-linux-gnu/4.0.1-alpha20050430
--mandir=/usr/share/gcc-data/i686-pc-linux-gnu/4.0.1-alpha20050430/man
--infodir=/usr/share/gcc-data/i686-pc-linux-gnu/4.0.1-alpha20050430/info
--with-gxx-include-dir=/usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/include/g++-v4
--host=i686-pc-linux-gnu --build=i686-pc-linux-gnu --disable-altivec
--disable-nls --with-system-zlib --disable-checking --disable-werror
--disable-libunwind-exceptions --disable-multilib --disable-libgcj
--enable-languages=c,c++ --enable-shared --enable-threads=posix
--enable-__cxa_atexit --enable-clocale=gnu
Thread model: posix
gcc version 4.0.1-alpha20050430 (Gentoo Linux 4.0.1_alpha20050430)
 /usr/libexec/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/cc1plus -E -quiet -v
-D_GNU_SOURCE cpptest.cpp -march=pentium3 -Wall -Os -fpch-preprocess -o cpptest.ii
ignoring nonexistent directory "/usr/local/include"
#include "..." search starts here:
#include <...> search starts here:
 /usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/include/g++-v4
 /usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/include/g++-v4/i686-pc-linux-gnu
 /usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/include/g++-v4/backward
 /usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/include
 /usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/../../../../i686-pc-linux-gnu/include
 /usr/include
End of search list.
 /usr/libexec/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/cc1plus -fpreprocessed
cpptest.ii -quiet -dumpbase cpptest.cpp -march=pentium3 -auxbase cpptest -Os
-Wall -version -o cpptest.s
GNU C++ version 4.0.1-alpha20050430 (Gentoo Linux 4.0.1_alpha20050430)
(i686-pc-linux-gnu)
	compiled by GNU C version 4.0.1-alpha20050430 (Gentoo Linux 4.0.1_alpha20050430).
GGC heuristics: --param ggc-min-expand=55 --param ggc-min-heapsize=48339
 /usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/../../../../i686-pc-linux-gnu/bin/as
-V -Qy -o cpptest.o cpptest.s
GNU assembler version 2.16.90.0.2 (i686-pc-linux-gnu) using BFD version
2.16.90.0.2 20050429
 /usr/libexec/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/collect2 --eh-frame-hdr
-m elf_i386 -dynamic-linker /lib/ld-linux.so.2 -o gcc401
/usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/../../../crt1.o
/usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/../../../crti.o
/usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/crtbegin.o
-L/usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430
-L/usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430
-L/usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/../../../../i686-pc-linux-gnu/lib
-L/usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/../../.. cpptest.o -lstdc++
-lm -lgcc_s -lgcc -lc -lgcc_s -lgcc
/usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/crtend.o
/usr/lib/gcc/i686-pc-linux-gnu/4.0.1-alpha20050430/../../../crtn.o

-- 
           Summary: C++ performance regression with -Os
           Product: gcc
           Version: 4.0.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: c++
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: deemkay at gmail dot com
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: i686-pc-linux-gnu
  GCC host triplet: i686-pc-linux-gnu
GCC target triplet: i686-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21314