On Sat, Dec 29, 2001 at 11:37:04PM +0000, Russell King - ARM Linux wrote:
> On Sat, Dec 29, 2001 at 02:24:04PM +0000, Nicholas Clark wrote:
> > Are both the a.out and ELF binaries your compiled against shared libc?
> 
> Yep.
> 
> > If so, does recompiling them with static linking change the performance
> > ratio?
> 
> Recompiling them as static ELF reveals:

Try this set instead; sorry I goofed 8/

So yes, static ELF looks a lot better for execl, but the process creation
is still down (lower than dynamic ELF results which is puzzling).

  BYTE UNIX Benchmarks (Version 3.11)
  System -- flint
  Start Benchmark Run: Sat Dec 29 23:46:54 GMT 2001
   9 interactive users.
Arithmetic Test (type = double)             445.1 lps   (10 secs, 6 samples)
Dhrystone 2 without register variables   311749.1 lps   (10 secs, 6 samples)
Execl Throughput Test                       378.2 lps   (9 secs, 6 samples)
Pipe-based Context Switching Test          2502.6 lps   (10 secs, 6 samples)
Shell scripts (1 concurrent)                385.2 lpm   (24 secs, 3 samples)
Shell scripts (2 concurrent)                200.5 lpm   (17 secs, 3 samples)
Shell scripts (4 concurrent)                101.6 lpm   (23 secs, 3 samples)
Shell scripts (8 concurrent)                 52.6 lpm   (24 secs, 3 samples)
Process Creation Test                       280.1 lps   (10 secs, 6 samples)


                     INDEX VALUES            
TEST                                        BASELINE     RESULT      INDEX

Arithmetic Test (type = double)               2541.7      445.1        0.2
Dhrystone 2 without register variables       22366.3   311749.1       13.9
Execl Throughput Test                           16.5      378.2       22.9
Pipe-based Context Switching Test             1318.5     2502.6        1.9
Shell scripts (8 concurrent)                     4.0       52.6       13.2
                                                                 =========
     SUM of  5 items                                                  52.1
     AVERAGE                                                          10.4

To follow up Phil's reply:

02000000-02001000 r-xp 00000000 09:07 528108     /mnt/src/byte-bench/pgms/spawn
02008000-02009000 rw-p 00000000 09:07 528108     /mnt/src/byte-bench/pgms/spawn
40000000-40016000 r-xp 00000000 09:01 320        /lib/ld-2.2.2.so
4001d000-4001e000 rw-p 00015000 09:01 320        /lib/ld-2.2.2.so
4001e000-4001f000 rwxp 00000000 00:00 0
40028000-40133000 r-xp 00000000 09:01 361        /lib/libc-2.2.2.so
40133000-40138000 ---p 0010b000 09:01 361        /lib/libc-2.2.2.so
40138000-40141000 rw-p 00108000 09:01 361        /lib/libc-2.2.2.so
40141000-40144000 rw-p 00000000 00:00 0
bfffe000-c0000000 rwxp fffff000 00:00 0


00008000-00010000 r-xp 00000000 00:0d 528109     /mnt/src/byte-bench/pgms/spawn
00010000-00018000 rwxp 00008000 00:0d 528109     /mnt/src/byte-bench/pgms/spawn
00bf8000-00c9f000 rwxp 00000000 03:03 3630       /lib/libc.so.4.6.27
00c9f000-00ccc000 rwxp 00000000 00:00 0
01799000-017a0000 rwxp 00019000 03:03 3694       /lib/ld.so.1.7.14
03fff000-04000000 rwxp 00000000 00:00 0


_______________________________________________
http://lists.arm.linux.org.uk/mailman/listinfo/linux-arm
Please visit the above address for information on this list.

Reply via email to