This is the mail archive of the libc-alpha@sourceware.org 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]
Other format: [Raw text]

Re: Patch for bz#11589 - jn calculation


From: Andreas Schwab <schwab@redhat.com>
Date: Thu, 29 Sep 2011 16:54:17 +0200

> David Miller <davem@davemloft.net> writes:
> 
>> From: Andreas Schwab <schwab@redhat.com>
>> Date: Wed, 28 Sep 2011 09:35:55 +0200
>>
>>> David Miller <davem@davemloft.net> writes:
>>> 
>>>> I've been getting getting testsuite failures on sparc in this area
>>>> ever since the powerpc changes went in.
> 
> How big are the deltas?

Here is, for example, math/test-float.out

testing float (without inline functions)
Failure: Test: jn (2, 2.4048255576957729) == 0.43175480701968038399746111312430703
Result:
 is:          4.31754827499389648438e-01   0x1.ba1df000000000000000p-2
 should be:   4.31754797697067260742e-01   0x1.ba1dee00000000000000p-2
 difference:  2.98023223876953125000e-08   0x1.00000000000000000000p-25
 ulp       :  1.0000
 max.ulp   :  0.0000
Failure: Test: jn (5, 2.4048255576957729) == 0.163892432048058525099230549946147698E-1
Result:
 is:          1.63892414420843124390e-02   0x1.0c857600000000000000p-6
 should be:   1.63892433047294616699e-02   0x1.0c857800000000000000p-6
 difference:  1.86264514923095703125e-09   0x1.00000000000000000000p-29
 ulp       :  1.0000
 max.ulp   :  0.0000
Failure: Test: jn (6, 2.4048255576957729) == 0.34048184720278336646673682895929161E-2
Result:
 is:          3.40481777675449848175e-03   0x1.be46ba00000000000000p-9
 should be:   3.40481847524642944336e-03   0x1.be46c000000000000000p-9
 difference:  6.98491930961608886719e-10   0x1.80000000000000000000p-31
 ulp       :  3.0000
 max.ulp   :  0.0000
Failure: Test: jn (7, 2.4048255576957729) == 0.60068836573295394221291569249883076E-3
Result:
 is:          6.00688101258128881454e-04   0x1.3aeefe00000000000000p-11
 should be:   6.00688392296433448792e-04   0x1.3aef0800000000000000p-11
 difference:  2.91038304567337036133e-10   0x1.40000000000000000000p-32
 ulp       :  5.0000
 max.ulp   :  0.0000
Failure: Test: jn (8, 2.4048255576957729) == 0.92165786705344923232879022467054148E-4
Result:
 is:          9.21657701837830245495e-05   0x1.82923e00000000000000p-14
 should be:   9.21657847356982529163e-05   0x1.82924200000000000000p-14
 difference:  1.45519152283668518066e-11   0x1.00000000000000000000p-36
 ulp       :  2.0000
 max.ulp   :  0.0000
Failure: Test: jn (9, 2.4048255576957729) == 0.12517270977961513005428966643852564E-4
Result:
 is:          1.25172691696207039058e-05   0x1.a4028600000000000000p-17
 should be:   1.25172709886101074517e-05   0x1.a4028a00000000000000p-17
 difference:  1.81898940354585647583e-12   0x1.00000000000000000000p-39
 ulp       :  2.0000
 max.ulp   :  0.0000
Maximal error of `jn'
 is      : 5 ulp
 accepted: 4 ulp

Test suite completed:
  3038 test cases plus 2651 tests for exception flags executed.
  7 errors occurred.


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