RE: Huge executable file with gcc version 3.1.1

["D.Venkatasubramanian, Noida" <dvenkat at noida dot hcltech dot com>]

Hi,

I am using newlib and my target is h8300-elf,
so it always produce a statically linked executable.
The problem is, I stripped the file and still the size is considerable.
If you see the size of the .text section, it is around 253 KB.

There is no --disable-shared also in the -v output.

Isn't this considerable for such a small test case?

Thanks and Regards,

Venky

[venkat@peacock Bugs]$ h8300-elf-g++ -v
Reading specs from
/home/venkat/gcc_release/h8300-elf/tools/bin/../lib/gcc-lib/h8300-elf/3.1.1/
specs
Configured with: /export/niting/WorkArea/gcc-release/src/configure
--prefix=/home/venkat/gcc_release/h8300-elf/tools/ --target=h8300-elf
--with-newlib --enable-languages=c,c++ --with-gnu-ld --with-gnu-as
--with-ld=/home/venkat/gcc_release/h8300-elf/tools/h8300-elf/bin/ld
--with-as=/home/venkat/gcc_release/h8300-elf/tools/h8300-elf/bin/as
--with-headers=/export/niting/WorkArea/gcc-release/src/newlib/libc/include/
Thread model: single
gcc version 3.1.1
[venkat@peacock BugC37]$ readelf -S a.out
There are 15 section headers, starting at offset 0x4189c:

Section Headers:
  [Nr] Name              Type            Addr     Off    Size   ES Flg Lk
Inf Al
  [ 0]                   NULL            00000000 000000 000000 00      0
0  0
  [ 1] .init             PROGBITS        00000100 000094 00000a 00  AX  0
0  1
  [ 2] .text             PROGBITS        0000010c 00009e 03e026 00  AX  0
0  4
  [ 3] .fini             PROGBITS        0003e132 03e0c4 000006 00  AX  0
0  1
  [ 4] .rodata           PROGBITS        0003e138 03e0ca 0018b4 00   A  0
0  4
  [ 5] .data             PROGBITS        0003f9ec 03f97e 001510 00  WA  0
0  4
  [ 6] .eh_frame         PROGBITS        00040efc 040e8e 000004 00  WA  0
0  4
  [ 7] .gcc_except_table PROGBITS        00040f00 040e92 000964 00  WA  0
0  4
  [ 8] .ctors            PROGBITS        00041864 0417f6 00001c 00  WA  0
0  4
  [ 9] .dtors            PROGBITS        00041880 041812 000014 00  WA  0
0  4
  [10] .jcr              PROGBITS        00041894 041826 000004 00  WA  0
0  4
  [11] .sbss             PROGBITS        00041898 04182a 000000 00   W  0
0  1
  [12] .bss              NOBITS          00041898 04182a 000ee0 00  WA  0
0  4
  [13] .stack            PROGBITS        0002fefc 04182a 000004 00      0
0  1
  [14] .shstrtab         STRTAB          00000000 04182e 00006c 00      0
0  1
Key to Flags:
  W (write), A (alloc), X (execute), M (merge), S (strings)
  I (info), L (link order), G (group), x (unknown)
  O (extra OS processing required) o (OS specific), p (processor specific)
[venkat@peacock BugC37]$ du -sk -h a.out
268k    a.out
[venkat@peacock BugC37]$ file a.out
a.out: ELF 32-bit MSB executable, Hitachi H8/300H, version 1 (SYSV),
statically linked, stripped

-----Original Message-----
From: Douglas Richardson [mailto:douglas.richardson@gaussvip.com]
Sent: Tuesday, September 24, 2002 7:56 PM
To: D.Venkatasubramanian, Noida; Kazu Hirata
Cc: gcc@gcc.gnu.org; gcc-bugs@gcc.gnu.org
Subject: RE: Huge executable file with gcc version 3.1.1


You are probably statically linking with the runtime libraries. Look at the
output of gcc -v. It will tell you how your gcc was configured. Do you see a
--disable-shared anywhere in the output? If so, you are statically linking
with the runtime libraries.

Doug.

-----Original Message-----
From: D.Venkatasubramanian, Noida [mailto:dvenkat@noida.hcltech.com]
Sent: Tuesday, September 24, 2002 06:15
To: Kazu Hirata
Cc: gcc@gcc.gnu.org; gcc-bugs@gcc.gnu.org
Subject: Huge executable file with gcc version 3.1.1


Hi,

I found that even very small test cases like an Hello World Program in C++, 
creating huge executable files for h8300-elf architecture.

If you see the output from the shell given below, 
you find that for a simple program, the executable was 2.9 MB.

I noticed this happening only when some header file was included.
This causes the simulator to crash, and when I compiled with -g option,
and tried to load it on the debugger, I found the debugger crashing when 
I tried to load the executable on the simulator.Seems the debugger crash 
must be due to this large size.

I noticed this when I found many test cases failing when I ran make check.

Any help would be appreciated.

Thanks and Regards,

Venky



[venkat@peacock BugC37]$ cat test.C
#include<iostream.h>
int main()
{
        std::cout << "Hello World" << endl;
}

[venkat@peacock BugC37]$ h8300-elf-g++ -v
Reading specs from
/home/venkat/gcc_release/h8300-elf/tools/bin/../lib/gcc-lib/h8300-elf/3.1.1/
specs
Configured with: /export/niting/WorkArea/gcc-release/src/configure
--prefix=/home/venkat/gcc_release/h8300-elf/tools/ --target=h8300-elf
--with-newlib --enable-languages=c,c++ --with-gnu-ld --with-gnu-as
--with-ld=/home/venkat/gcc_release/h8300-elf/tools/h8300-elf/bin/ld
--with-as=/home/venkat/gcc_release/h8300-elf/tools/h8300-elf/bin/as
--with-headers=/export/niting/WorkArea/gcc-release/src/newlib/libc/include/
Thread model: single
gcc version 3.1.1
[venkat@peacock BugC37]$ h8300-elf-g++ -mh -mint32 test.C
In file included from
/home/venkat/gcc_release/h8300-elf/tools/include/c++/3.1.1/backward/iostream
.h:31,
                 from test.C:1:
/home/venkat/gcc_release/h8300-elf/tools/include/c++/3.1.1/backward/backward
_warning.h:32:2: warning: #warning This file includes at least one
deprecated or antiquated header. Please consider using one of the 32 headers
found in section 17.4.1.2 of the C++ standard. Examples include substituting
the <X> header for the <X.h> header for C++ includes, or <sstream> instead
of the deprecated header <strstream.h>. To disable this warning use
-Wno-deprecated.
[venkat@peacock BugC37]$ du -sk -h a.out
2.9M    a.out
[venkat@peacock BugC37]$ file a.out
a.out: ELF 32-bit MSB executable, Hitachi H8/300H, version 1 (SYSV),
statically linked, not stripped
[venkat@peacock BugC37]$ ls -al
total 2908
drwxrwxr-x    2 venkat   venkat       4096 Sep 24 18:44 .
drwxrwxr-x   65 venkat   venkat       4096 Sep 24 17:37 ..
-rwxrwxr-x    1 venkat   venkat    2958101 Sep 24 18:44 a.out
-rw-rw-r--    1 venkat   venkat         74 Sep 24 18:42 test.C
[venkat@peacock BugC37]$


_____________________________________________________

"To be or not to be?" is not the question, but "What do you
want to be?" is.  
   --- Anon