#39097 [Opn->Csd]: Decimal order of magnitude of precision lost between 5.2.0RC4 and RC5

ID: 39097 Updated by: bjori@php.net Reported By: phpbugrep-20061009 at pgregg dot com -Status: Open +Status: Closed Bug Type: Unknown/Other Function Operating System: FreeBSD PHP Version: 5.2.0RC5 New Comment: This bug has been fixed in CVS. Snapshots of the sources are packaged every three hours; this change will be in the next snapshot. You can grab the snapshot at http://snaps.php.net/. Thank you for the report, and for helping us make PHP better. Thanks. Previous Comments: ------------------------------------------------------------------------ [2006-10-09 16:51:24] phpbugrep-20061009 at pgregg dot com Fix for the test: *** ext/standard/tests/math/constants.phpt.bak Mon Oct 9 17:47:37 2006 --- ext/standard/tests/math/constants.phpt Mon Oct 9 17:49:02 2006 *************** *** 25,46 **** printf("%-10s: %.14s\n", $constant, constant($constant)); } ?> ! --EXPECT-- ! M_E : 2.718281828459 ! M_LOG2E : 1.442695040889 ! M_LOG10E : 0.434294481903 ! M_LN2 : 0.693147180559 ! M_LN10 : 2.302585092994 ! M_PI : 3.141592653589 ! M_PI_2 : 1.570796326794 ! M_PI_4 : 0.785398163397 ! M_1_PI : 0.318309886183 ! M_2_PI : 0.636619772367 ! M_SQRTPI : 1.772453850905 ! M_2_SQRTPI: 1.128379167095 ! M_LNPI : 1.144729885849 ! M_EULER : 0.577215664901 ! M_SQRT2 : 1.414213562373 ! M_SQRT1_2 : 0.707106781186 ! M_SQRT3 : 1.732050807568 --- 25,46 ---- printf("%-10s: %.14s\n", $constant, constant($constant)); } ?> ! --EXPECTREGEX-- ! M_E : 2.718281[0-9]* ! M_LOG2E : 1.442695[0-9]* ! M_LOG10E : 0.434294[0-9]* ! M_LN2 : 0.693147[0-9]* ! M_LN10 : 2.302585[0-9]* ! M_PI : 3.141592[0-9]* ! M_PI_2 : 1.570796[0-9]* ! M_PI_4 : 0.785398[0-9]* ! M_1_PI : 0.318309[0-9]* ! M_2_PI : 0.636619[0-9]* ! M_SQRTPI : 1.772453[0-9]* ! M_2_SQRTPI: 1.128379[0-9]* ! M_LNPI : 1.144729[0-9]* ! M_EULER : 0.577215[0-9]* ! M_SQRT2 : 1.414213[0-9]* ! M_SQRT1_2 : 0.707106[0-9]* ! M_SQRT3 : 1.732050[0-9]* ------------------------------------------------------------------------ [2006-10-09 16:40:33] phpbugrep-20061009 at pgregg dot com If that is the case, then the report becomes one of a broken test. How can it be bogus when a standard "make test" fails with ext/standard/tests/math/constants.phpt The "goal of the test is to make sure they are defined, not their value." <--- this sure is not what the test is coded to do. ------------------------------------------------------------------------ [2006-10-09 16:20:17] bjori@php.net The constants are system dependent. If your system doesn't define them, PHP should do so. The goal of the test is to make sure they are defined, not their value. ------------------------------------------------------------------------ [2006-10-09 15:45:21] phpbugrep-20061009 at pgregg dot com Sorry "Description" should say: Running make test on 5.2.0RC5 fails with ------------------------------------------------------------------------ [2006-10-09 15:43:56] phpbugrep-20061009 at pgregg dot com Description: ------------ Running make test on 5.2.0RC4 fails with: FAIL Math constants [ext/standard/tests/math/constants.phpt] Reproduce code: --------------- Add this additional debug to run-tests.php at line 1494: $output_pgarr = explode("\n", $output); $wanted_pgarr = explode("\n", $wanted); foreach($wanted_pgarr as $line) { list($pgleft, $pgright) = explode(":", $line, 2); $pgleft = trim($pgleft); $pgright = trim($pgright); $pgwanted[$pgleft] = "$pgright"; } foreach($output_pgarr as $line) { list($pgleft, $pgright) = explode(":", $line, 2); $pgleft = trim($pgleft); $pgright = trim($pgright); if (strcmp($pgright, $pgwanted[$pgleft]) !== 0) printf("%-10s: Expected: %-16s Got: %-16s Diff: %s\n", $pgleft, $pgwanted[$pgleft], $pgright, bcsub($pgwant ed[$pgleft], $pgright,12)); } % echo "ext/standard/tests/math/constants.phpt" > pgtest.txt % php run-tests.php -l pgtest.txt Expected result: ---------------- Expected test to PASS Actual result: -------------- Added more code to run-tests.php to report on what was going on and got: M_E : Expected: 2.718281828459 Got: 2.71828182846 Diff: -0.000000000001 M_LOG2E : Expected: 1.442695040889 Got: 1.44269504089 Diff: -0.000000000001 M_LN2 : Expected: 0.693147180559 Got: 0.69314718056 Diff: -0.000000000001 M_LN10 : Expected: 2.302585092994 Got: 2.30258509299 Diff: 0.000000000004 M_PI : Expected: 3.141592653589 Got: 3.14159265359 Diff: -0.000000000001 M_PI_2 : Expected: 1.570796326794 Got: 1.57079632679 Diff: 0.000000000004 M_1_PI : Expected: 0.318309886183 Got: 0.318309886184 Diff: -0.000000000001 M_2_PI : Expected: 0.636619772367 Got: 0.636619772368 Diff: -0.000000000001 M_SQRTPI : Expected: 1.772453850905 Got: 1.77245385091 Diff: -0.000000000005 M_2_SQRTPI: Expected: 1.128379167095 Got: 1.1283791671 Diff: -0.000000000005 M_LNPI : Expected: 1.144729885849 Got: 1.14472988585 Diff: -0.000000000001 M_EULER : Expected: 0.577215664901 Got: 0.577215664902 Diff: -0.000000000001 M_SQRT2 : Expected: 1.414213562373 Got: 1.41421356237 Diff: 0.000000000003 M_SQRT1_2 : Expected: 0.707106781186 Got: 0.707106781187 Diff: -0.000000000001 M_SQRT3 : Expected: 1.732050807568 Got: 1.73205080757 Diff: -0.000000000002 FAIL Math constants [ext/standard/tests/math/constants.phpt] As you can see, we have lost a decimal place of precision. plop:root/p2-/usr/local/src/Web/Apache2_PHP5/php-5.2.0RC5-862#->uname -a 4:38PM FreeBSD plop.pgregg.com 5.5-PRERELEASE FreeBSD 5.5-PRERELEASE #4: Tue Mar 7 13:49:31 GMT 2006 pgregg@plop.pgregg.com:/usr/obj/usr/src/sys/PGREGG_PLOP i386 This test passed ok in 5.2.0RC4 ------------------------------------------------------------------------ -- Edit this bug report at http://bugs.php.net/?id=39097&edit=1