PATCH: PR math/12775: Typo in sysdeps/x86_64/fpu/e_powl.S

H.J. Lu hjl.tools@gmail.com
Wed May 18 18:56:00 GMT 2011


On Wed, May 18, 2011 at 11:35 AM, Jakub Jelinek <jakub@redhat.com> wrote:
> On Wed, May 18, 2011 at 09:30:03AM -0700, H.J. Lu wrote:
>> @@ -90,7 +90,8 @@ distribute += $(filter-out $(generated),$(long-m-yes:=.c) $(long-c-yes:=.c))
>>  # Rules for the test suite.
>>  tests = test-matherr test-fenv atest-exp atest-sincos atest-exp2 basic-test \
>>       test-misc test-fpucw tst-definitions test-tgmath test-tgmath-ret \
>> -     bug-nextafter bug-nexttoward bug-tgmath1 test-tgmath-int test-tgmath2
>> +     bug-nextafter bug-nexttoward bug-tgmath1 test-tgmath-int \
>> +     test-tgmath2 test-powl
>>  # We do the `long double' tests only if this data type is available and
>>  # distinct from `double'.
>>  test-longdouble-yes = test-ldouble test-ildoubl
>> @@ -131,12 +132,14 @@ CFLAGS-test-ldouble.c = -fno-inline -ffloat-store -fno-builtin
>>  CFLAGS-test-tgmath.c = -fno-builtin
>>  CFLAGS-test-tgmath2.c = -fno-builtin
>>  CFLAGS-test-tgmath-ret.c = -fno-builtin
>> +CFLAGS-test-powl.c = -O0
>
> Why not -fno-builtin instead of -O0?
>
>> +#ifndef NO_LONG_DOUBLE
>> +    {
>> +      long double x = 1e-20;
>> +      union ieee854_long_double u;
>> +      u.ieee.mantissa0 = 1;
>> +      u.ieee.mantissa1 = 1;
>> +      u.ieee.exponent = 0;
>> +      u.ieee.negative = 0;
>
> This won't work correctly for IEEE quad long double or IBM double double.
> So, either you should add the test for extended long double only, or
> add guards here for extended long double too (e.g. LDBL_MANT_DIG == 64).
>
>        Jakub
>

Here is the updated patch.

Thanks.

-- 
H.J.
----
2011-05-18  H.J. Lu  <hongjiu.lu@intel.com>

	PR math/12775
	* math/Makefile (tests): Add test-powl.
	(CFLAGS-test-powl.c): New.

	* math/test-powl.c: New.

	* sysdeps/x86_64/fpu/e_powl.S: Fix a typo.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: libc-pr12775-3.patch
Type: text/x-diff
Size: 3311 bytes
Desc: not available
URL: <http://sourceware.org/pipermail/libc-alpha/attachments/20110518/06f52eb2/attachment.bin>


More information about the Libc-alpha mailing list