FreeBSD/Alpha Alignment Issues

Hey - I'm testing PHP 4.3.0-pre2 on my FreeBSD/Alpha box. (FreeBSD 4.7-RELEASE-p1, just built this weekend.) When I run "make test", I have a bunch of errors like this: pid 15878 (php): unaligned access: va=0x120227204 pc=0x12017fcd4 ra=0x12017fcc8 op=stq pid 15878 (php): unaligned access: va=0x120227124 pc=0x12017fcd4 ra=0x12017fcc8 op=stq Come flying to my screen. The tests then go on to run, and returns: TEST RESULT SUMMARY ---------------------------------------------------------------------

Exts skipped    :   80
Exts tested     :   10

--------------------------------------------------------------------- Number of tests : 334 Tests skipped : 119 (35.6%)

Tests failed    :   13 ( 3.9%)
Tests passed    :  202 (60.5%)

---------------------------------------------------------------------

Time taken      :  184 seconds

===================================================================== ===================================================================== FAILED TEST SUMMARY --------------------------------------------------------------------- mb_strtoupper() / mb_strtolower() [ext/mbstring/tests/casefold.phpt] HTML input/output [ext/mbstring/tests/htmlent.phpt] mb_split() [ext/mbstring/tests/mb_split.phpt] serializing references test case using globals [ext/session/tests/019.phpt] Test array_merge and array_walk [ext/standard/tests/array/001.phpt] Test arsort, asort, krsort, ksort, rsort, and sort [ext/standard/tests/array/002.phpt] Test usort, uksort and uasort [ext/standard/tests/array/003.phpt] var_dump float test [ext/standard/tests/general_functions/008.phpt] Simple math tests [ext/standard/tests/math/abs.phpt] overflow check for _php_math_basetozval [ext/standard/tests/math/hexdec.phpt] Various pow() tests [ext/standard/tests/math/pow.phpt] Simple math tests [ext/standard/tests/math/round.phpt] crc32() function [ext/standard/tests/strings/crc32.phpt] ===================================================================== I'd say there are some 64-bit issues in the current test version. Also: gcc reports that "-O2" produces buggy code on this platform. In order to build at all, I have to change "-O2" to "-O" in the Makefile after running configure. (This is no big deal, but it would be nice if the configure process did this for me.) I'm willing to do whatever I can to help debug this issue. - Matt