APPENDIX B: PRETEST REPORT FORM

ADA COMPILER EVALUATION SYSTEM Version 2.0

PRETEST REPORT

Evaluator's Name:____________________________________Date:____________

Compiler:____________________________________________Version:_________

Operating

System:______________________________________________Version:_________

Host

Hardware:____________________________________________Model:___________

Disk

Devices:_____________________________________________Memory:__________

Disk Size Required

for Compilation System:_______________________________________________

Information regarding

Installation of Compiler:_____________________________________________

_____________________________________________

Y/N/Answer

(N/A for not applicable)

1. Access the compilation system -

Is compiler able to output strings containing numerics? _________

2. Check label'ADDRESS capability -

a. Does the compilation system support label'ADDRESS

and is it accurate? _________

b. If label'ADDRESS is not supported or not accurate

was the assembly routine zg_getad adapted? _________

3. Compile the system clock/calendar tests -

a. Are CPU time results desired? _________

b. If yes to #3a, was "zg_cpu.dec" adapted and renamed

"zg_cpu<ada>"? _________

c. Did all necessary zp_tcal* files compile and

link successfully? _________

4. Execute the system clock/calendar tests -

Elapsed time - zp_tcal1 & zp_tcal2

a. Is the system clock accurate for elapsed time

measurements? _________

CPU time - zp_tcal3

b. Record percentage difference for run WITHOUT

contention. _________

c. Record percentage difference for run WITH

contention. _________

CPU time - zp_tcal4

d. Does the CPU function return Program or

Process time? _________

5. Compile the test suite baseline files -

a. Which zg_glob3 package was used?

(.ELG .ELL .CPG or .CPL)? _________

b. Which random number generator was used?

(zm_ran16 or zm_ran32)? _________

c. Which zm_math package was used?(vendor NUMWG, vendor

unique, ACES portable, or ACES tailored)? _________

d. Which zg_init program for initialization was used?

(.ELG .ELL .CPG or .CPL)? _________

MUST be the same as the above zg_glob3.

e. Which zg_incld program for text inclusion was used?

(.ELG .ELL .CPG or .CPL)? _________

MUST be the same as the above zg_glob3 & zg_init

f. Which time function (.EL or .CP) was used for the

creation of the compile & link time stamp programs?_________

6. Test the inner timing loop iteration count -

a. Which zp_basic program was used? (

.ELG .ELL .CPG or .CPL)? _________

MUST be the same as the above files.

b. Suggested value for basic_iteration_count from

zp_basic is _________

c. Actual value chosen was _________

7. Test the portable MATH package adaptation -

Was zp_dptst (DEPTEST) executed successfully? _________

8. Test ZM_MATH Package -

a. How many bits of mantissa in float6 type? _ _______bits

b. How many bits of precision in integer16 type? _______bits

c. Are all maximum error ranges within tolerance

for all float6 functions? _________

d. Do the values for selected inputs follow NUMWG

recommendations for float6 functions? _________

9. Test ZM_DBLMA Package -

a. How many bits of mantissa in float9 type? _______bits

b. How many bits of precision in integer32 type? _______bits

c. Are all maximum error ranges within tolerance

for all float9 functions? _________

d. Do the values for selected inputs follow NUMWG

recommendations for float9 functions? _________

10. Test the zg_incld (INCLUDE) preprocessor -

Is zg_incld working correctly? _________

11. Execute the test suite command files/Harness tool -

Were all test suite command files adapted? _________

Was a Harness executable created successfully? _________

Did the command file created by the Harness

execute correctly? _________

Were compilation/link time measurements output

to a file? _________

Were execution time measurements output to a file? _________

12. Did the following analysis tools compile?

Menu? _________

Condense? _________

Comparative Analysis? _________

Single System Analysis _________

Did the tools link together or did they have to be

linked separately? _________

13. Did the Condense tool format the output correctly? _________

Do the database files created by Condense contain

entries for each compile/link time? _________

Do the database files created by Condense contain

entries for each execution time? _________

14. Did each of the tools produce expected results

using the sample input values?

Menu? _________

Condense? _________

Comparative Analysis? _________

Were execution time results measured and reported

using a sample performance test? _________

15. Did Single System Analysis produce expected results? _________

______________________________________________________________________

GENERAL COMMENTS