This is the mail archive of the
libc-alpha@sources.redhat.com
mailing list for the glibc project.
K6-II+, 2.2.3pre3 test results
- To: Byeong-ryeol Kim <jinbo21 at hananet dot net>
- Subject: K6-II+, 2.2.3pre3 test results
- From: Stephen L Moshier <moshier at mediaone dot net>
- Date: Tue, 17 Apr 2001 07:33:08 -0400 (EDT)
- cc: libc-alpha at sourceware dot cygnus dot com
- Reply-To: moshier at moshier dot ne dot mediaone dot net
> I run K6-II+ 500 MHZ model 13 stepping 4 at 400 MHZ
Using a similar processor I get the similar but not identical
error report below. The same compiler and binutils gave no error
report on an Intel P133.
According to my calculation,
Jn(3,10) = 5.83793793051868123429E-2
which means the "should be" value printed in the test report is wrong
at the 8th figure. When I use the jn function in a test program,
it gives the correct answer. Also Drepper previously said the i387 asinl
function was very inaccurate but when I tested the routine it did
not seem to have much wrong with it. So I suspect the functions may be
ok but I wonder about what is going on in the test program.
$ gcc -v
Reading specs from /usr/lib/gcc-lib/i486-linux/2.95.3/specs
gcc version 2.95.3 20010312 (prerelease)
$ as --version
GNU assembler 2.10.1
$ cat /proc/cpuinfo
processor : 0
vendor_id : AuthenticAMD
cpu family : 5
model : 8
model name : AMD-K6(tm) 3D processor
stepping : 12
cpu MHz : 350.803
cache size : 64 KB
fdiv_bug : no
hlt_bug : no
sep_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr mce cx8 sep mtrr pge mmx
3dnow
bogomips : 699.60
$ cat test-ifloat.out
testing float (inline functions)
Failure: Test: jn (3, 10.0) == 0.058379379305186812343
Result:
is: 5.80999180674552917480e-02 0x1.dbf45c00000000000000p-5
should be: 5.83793781697750091553e-02 0x1.de3e6e00000000000000p-5
difference: 2.79460102319717407227e-04 0x1.25090000000000000000p-12
ulp : 75017.0000
max.ulp : 1.0000
Maximal error of `jn'
is : 75017.0000 ulp
accepted: 2.0000 ulp
Test suite completed:
2406 test cases plus 2204 tests for exception flags executed.
2 errors occurred.