This is the mail archive of the libc-hacker@cygnus.com mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

glibc-2.0.112 - "make check" failures on fdim et al., and minor errors


glibc-2.0.112
egcs-1.1.1
binutils-2.9.1.0.19a
Linux kernel headers 2.2.1
Intel Pentium 75 MHz (stepping 5)

(1)
I get errors
> Testing: fdim (9, 0) = 9
> Pass: Exception "Divide by zero" not set
> Pass: Exception "Invalid operation" not set
> Fail:
> Result:
> is:         -9.00000000000000000000e+00  -0x1.20000000000000000000p+3
> should be:   9.00000000000000000000e+00   0x1.20000000000000000000p+3
> difference:  1.80000000000000000000e+01   0x1.20000000000000000000p+4

[...]

> Testing: fdim (0, -9) = 9
> Pass: Exception "Divide by zero" not set
> Pass: Exception "Invalid operation" not set
> Fail:
> Result:
> is:         -9.00000000000000000000e+00  -0x1.20000000000000000000p+3
> should be:   9.00000000000000000000e+00   0x1.20000000000000000000p+3
>  difference:  1.80000000000000000000e+01   0x1.20000000000000000000p+4
>   
> Testing: fdim (+inf, 9) = +inf
> Pass: Exception "Divide by zero" not set
> Pass: Exception "Invalid operation" not set
> Fail:
> Value: -inf  -inf
>    
> Testing: fdim (+inf, -9) = +inf
> Pass: Exception "Divide by zero" not set
> Pass: Exception "Invalid operation" not set
> Fail:
> Value: -inf  -inf

[...]

> Testing: fdim (+9, -inf) = +inf
> Pass: Exception "Divide by zero" not set
> Pass: Exception "Invalid operation" not set
> Fail:
> Value: -inf  -inf
> 
> Testing: fdim (-9, -inf) = +inf
> Pass: Exception "Divide by zero" not set
> Pass: Exception "Invalid operation" not set
> Fail:
> Value: -inf  -inf
  
in test-float.out, test-double.out, test-ldouble.out, test-ildoubl.out,
test-ifloat.out, and test-idouble.out.

(2)
Also I see 

> __erfl not implemented
> __erfcl not implemented
> __ieee754_lgammal_r not implemented
> __ieee754_lgammal_r not implemented
> __ieee754_lgammal_r not implemented

but I am assuming that this is okay.

(3)
I fixed an obvious(?) typo, appended at the end.

Thanks.

	-Paul <kimoto@spacenet.tn.cornell.edu>


--- glibc-2.0.112/math/libm-test.c~     Sun Dec 13 17:59:28 1998
+++ glibc-2.0.112/math/libm-test.c      Sun Jan 31 12:22:48 1999
@@ -2178,7 +2178,7 @@ fdim_test (void)
   check ("fdim (+0, +0) = +0", FUNC(fdim) (0, 0), 0);
   check ("fdim (9, 0) = 9", FUNC(fdim) (9, 0), 9);
   check ("fdim (0, 9) = 0", FUNC(fdim) (0, 9), 0);
-  check ("fdim (-9, 0) = 9", FUNC(fdim) (-9, 0), 0);
+  check ("fdim (-9, 0) = 0", FUNC(fdim) (-9, 0), 0);
   check ("fdim (0, -9) = 9", FUNC(fdim) (0, -9), 9);
	     
   check_isinfp ("fdim (+inf, 9) = +inf", FUNC(fdim) (plus_infty, 9));


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]