>>>>> "James" == James Amundson <amundson@fnal.gov> writes:
James> Previously, I could do this:
James> (C1) %j[1](sqrt(x));
James> (D1) %J (SQRT(x))
James> 1
James> Now, I get this:
James> (C1) %j[1](sqrt(x));
James> Is x positive, negative, or zero?
James> This bug has to be fixed. I see you updated to tests to work around the
James> bug. Please roll back those changes. I haven't found the source of the
James> bug yet, but the above information should be a good lead.
Fixed. I didn't know that $realpart and/or $imagpart would cause such
a question to be asked. I rearranged the code, and it doesn't do this
anymore.
I've also undone the change to rtest14.mac. The tests pass fine.
I'm really sorry for this screwup. I'll remember to run the checks
more often as a good sanity check.
The only errors I get are the 2 known bugs in test15.
FWIW, on my 300 Mhz Ultra sparc machine, I now get:
Clisp:
Real time: 37.164703f0 sec.
Run time: 29.45f0 sec.
Space: 25961176 Bytes
GC: 29, GC time: 2.35f0 sec.
CMUCL:
; Evaluation took:
; 18.11f0 seconds of real time
; 10.22f0 seconds of user run time
; 1.09f0 seconds of system run time
; [Run times include 0.58f0 seconds GC run time]
; 0 page faults and
; 34256416 bytes consed.
On my 866 Mhz P-III from yesterday, I get:
Clisp:
Real time: 6.965162f0 sec.
Run time: 6.75f0 sec.
Space: 25421416 Bytes
GC: 31, GC time: 0.89f0 sec.
;; Loading of file tests.lisp is finished.
T
CMUCL:
Evaluation took:
2.83f0 seconds of real time
2.5f0 seconds of user run time
0.23f0 seconds of system run time
[Run times include 0.16f0 seconds GC run time]
5200 page faults and
34114120 bytes consed.
GCL:
real time : 7.560 secs
run time : 7.060 secs
T
Too bad GCL doesn't print out the GC time or the bytes allocated.
I'm surprised that CMUCL is now the fastest by far of the three
systems, and Clisp is comparable or slightly faster than GCL. (But
maybe I didn't compile GCL with the right bignum package.)
Ray