From Bugzilla Helper:
User-Agent: Mozilla/4.72 [en] (X11; U; Linux 2.2.14-5.0 i586)

Description of problem:
Compiles with g++ take significantly (4 times or more) longer than under RH
6.2.  The current gcc related patches in thesupport section of the RH web
site do not fix this bug.
Our experience seems to match those discussed in: 
http://gcc.gnu.org/ml/gcc-bugs/2000-08/msg00688.html
I see what may be a patch posted at:
http://www.geocrawler.com/mail/msg.php3?msg_id=4291993&list=356
Will RedHat be posting a patch for this bug soon?

How reproducible:
Always

Steps to Reproduce:
1.  On RH 6.2, g++ -c bug2.cpp.  Returns almost immediately (under 2
seconds).
2. On RH 7.1 (with faster cpu), g++ -c bug2.cpp.  Takes about 22 seconds.
3

Actual Results:  On RH 6.2, the compile finishes in under 2 seconds on an
AMD K6 400MHz system.
On RH 7.1, the compiles takes about 22 seconds on an AMD K6 400MHz system.
We can repeat a major compile time difference on systems with faster
processors also.
No compiler error is generated.

Expected Results:  We expected no difference in compile times.

Additional info:

Our "bug2" program is:
int main()
{
int var1, var2, var3, var4, var5, var6, var7, var8, var9, var10, var11;

var1 =
        ( (1 * (var2 == 1)) |
        (( (1 * (var3 == 4)) |
        (2 * (var3 == 5)) |
        (4 * (var3 == 1)) |
        (4 * (var3 == 3)) |
        (4 * (var3 == 2))) * (var2 == 2 )) |
        (1 * (var4 == 3)) |
        (1 * (var4 == 1)) |
        (1 * (var4 == 2)) |
        (1 * (var4 == 3)) |
        (3 * (var5 == 3 )) |
        (3 * (var6 == 1 )) |
        (3 * (var6 == 2 )) |
        (3 * (var6 == 3 )));

var7 =
        ( (1 * (var8 == 1 )) |
        (4 * (var8 == 4 )) |
        (( (1 * (var9 == 4)) |
        (2 * (var10 == 1)) |
        (3 * (var10 == 2)) |
        (4 * (var10 == 3))) * (var11 == 4)));
}