Date:	Sun, 25 Feb 2001 18:11:17 +0800
From:	Lin Zhemin <ljm AT ljm DOT qqjane DOT net>
To:	pgcc AT delorie DOT com
Subject:	pgcc-2.95.2.1 vs. glibc-2.2.2 test-suit failed.
Message-ID:	<20010225181117.A18753@ljm.qqjane.net>
Mime-Version:	1.0
User-Agent:	Mutt/1.2.5i
X-MIME-Autoconverted:	from 8bit to quoted-printable by ljm.qqjane.net id f1PABHO18758
X-MIME-Autoconverted:	from quoted-printable to 8bit by delorie.com id FAA18448
Reply-To:	pgcc AT delorie DOT com

Dear Sir,

I'm working on a 2 x Pentium-III 600MHz Linux kernel version 2.2.18
and tried to compile glibc-2.2.2 with pgcc-2.95.2.1, with
CFLAGS="-O -Os -march=pentiumpro". I've compiled pgcc-2.95.2.1
for two bootstraps in order to get a pure pgcc generated code.

Most glibc-2.2.2 test-suites passed but one had not:

-------------------------------------------------------------------------
GCONV_PATH=/src/tmp/gb/iconvdata LC_ALL=C /src/tmp/gb/elf/ld-linux.so.2 \
           --library-path /src/tmp/gb:/src/tmp/gb/math:/src/tmp/gb/elf: \
           /src/tmp/gb/dlfcn:/src/tmp/gb/nss:/src/tmp/gb/nis: \
           /src/tmp/gb/rt:/src/tmp/gb/resolv:/src/tmp/gb/crypt: \
           /src/tmp/gb/linuxthreads /src/tmp/gb/math/test-float
           > /src/tmp/gb/math/test-float.out
make[2]: *** [/src/tmp/gb/math/test-float.out] Erreur 1
make[2]: Quitte le répertoire /src/tmp/glibc-2.2.2/math'
make[1]: *** [math/tests] Erreur 2
make[1]: Quitte le répertoire /src/tmp/glibc-2.2.2'
make: *** [check] Erreur 2
-------------------------------------------------------------------------


And the content of test-float.out is:

-------------------------------------------------------------------------
testing float (without inline functions)
Failure: Test: tanh (0.7) == 0.60436777711716349631
Result:
 is:          6.04367733001708984375e-01   0x1.356fb000000000000000p-1
 should be:   6.04367792606353759766e-01   0x1.356fb200000000000000p-1
 difference:  5.96046447753906250000e-08   0x1.00000000000000000000p-24
 ulp       :  1.0000
 max.ulp   :  0.0000
Maximal error of `tanh'
 is      :  1.0000 ulp
 accepted:  0.0000 ulp
Failure: Test: Real part of: cacosh (-2 - 3 i) == -1.9833870299165354323 + 2.1414491111159960199 i
Result:
 is:         -1.98338806629180908203e+00  -0x1.fbbf5200000000000000p+0
 should be:  -1.98338699340820312500e+00  -0x1.fbbf4000000000000000p+0
 difference:  1.07288360595703125000e-06   0x1.20000000000000000000p-20
 ulp       :  9.0000
 max.ulp   :  4.0000
Maximal error of real part of: cacosh
 is      :  9.0000 ulp
 accepted:  4.0000 ulp
Maximal error of imaginary part of: cacosh
 is      :  0.0000 ulp
 accepted:  4.0000 ulp
Failure: Test: Real part of: csqrt (-2 - 3 i) == 0.89597747612983812471573375529004348 - 1.6741492280355400404480393008490519 i
Result:
 is:          8.95977437496185302734e-01   0x1.cabd8e00000000000000p-1
 should be:   8.95977497100830078125e-01   0x1.cabd9000000000000000p-1
 difference:  5.96046447753906250000e-08   0x1.00000000000000000000p-24
 ulp       :  1.0000
 max.ulp   :  0.0000
Failure: Test: Real part of: csqrt (-2 + 3 i) == 0.89597747612983812471573375529004348 + 1.6741492280355400404480393008490519 i
Result:
 is:          8.95977437496185302734e-01   0x1.cabd8e00000000000000p-1
 should be:   8.95977497100830078125e-01   0x1.cabd9000000000000000p-1
 difference:  5.96046447753906250000e-08   0x1.00000000000000000000p-24
 ulp       :  1.0000
 max.ulp   :  0.0000
Maximal error of real part of: csqrt
 is      :  1.0000 ulp
 accepted:  0.0000 ulp
Maximal error of imaginary part of: csqrt
 is      :  0.0000 ulp
 accepted:  0.0000 ulp
Failure: Test: Real part of: ctanh (-2 - 3 i) == -0.9653858790221331242 + 0.0098843750383224937 i
Result:
 is:         -9.65385794639587402344e-01  -0x1.ee470c00000000000000p-1
 should be:  -9.65385854244232177734e-01  -0x1.ee470e00000000000000p-1
 difference:  5.96046447753906250000e-08   0x1.00000000000000000000p-24
 ulp       :  1.0000
 max.ulp   :  0.0000
Failure: Test: yn (3, 0.1) == -5099.3323786129048894
Result:
 is:         -5.09933154296875000000e+03  -0x1.3eb54e00000000000000p+12
 should be:  -5.09933251953125000000e+03  -0x1.3eb55200000000000000p+12
 difference:  9.76562500000000000000e-04   0x1.00000000000000000000p-10
 ulp       :  2.0000
 max.ulp   :  1.0000
Failure: Test: yn (3, 0.7) == -15.819479052819633505
Result:
 is:         -1.58194799423217773438e+01  -0x1.fa392e00000000000000p+3
 should be:  -1.58194789886474609375e+01  -0x1.fa392c00000000000000p+3
 difference:  9.53674316406250000000e-07   0x1.00000000000000000000p-20
 ulp       :  1.0000
 max.ulp   :  0.0000
Failure: Test: yn (10, 0.1) == -0.11831335132045197885e19
Result:
 is:         -1.18313319792010854400e+18  -0x1.06b55600000000000000p+60
 should be:  -1.18313347279801548800e+18  -0x1.06b55a00000000000000p+60
 difference:  2.74877906944000000000e+11   0x1.00000000000000000000p+38
 ulp       :  2.0000
 max.ulp   :  1.0000
Failure: Test: yn (10, 1.0) == -0.12161801427868918929e9
Result:
 is:         -1.21618032000000000000e+08  -0x1.cfef9c00000000000000p+26
 should be:  -1.21618016000000000000e+08  -0x1.cfef9800000000000000p+26
 difference:  1.60000000000000000000e+01   0x1.00000000000000000000p+4
 ulp       :  2.0000
 max.ulp   :  1.0000

Test suite completed:
  2468 test cases plus 2266 tests for exception flags executed.
  12 errors occurred.
-------------------------------------------------------------------------

I don't know what's wrong here. 
I've compiled the entire glibc-2.2.2 again with CFLAGS='-O' but things
were the same. Thanks for helping.


Best regards,
Lin Zhemin.


-- 
Liebe mich mit treuem Sinn! Sieh auch meine Treue; mit ganzem
Herzen und mit ganzem Sinn bin ich bei dir, auch in der Ferne.
Fingerprint20 = CE32 D237 02C0 FE31 FEA9  B858 DE8F AE2D D810 F2D9

- Raw text -