# New Ticket Created by Alex Gough # Please include the string: [perl #18762] # in the subject line of all future correspondence about this issue. # <URL: http://rt.perl.org/rt2/Ticket/Display.html?id=18762 >
t/pmc/nci.t doesn't like me or, at least, it doesn't like the way I don't have a lrint in my <math.h>. My K&R claims that everything in math.h ought to return a double (well, anything you can validly expect to find in every math.h), so using math.so is probably not the way to test the nci, instead a set of dummy functions ought to be provided by a little library of our own, and built as part of the build process. Oh, and the test dumps core when it doesn't find the symbol, this is probably not a good thing either. [12:30AM]percy/parrot% perl -Ilib t/pmc/nci.t 1..2 ok 1 - nic_d_d not ok 2 - nic_i_d # Failed test (t/pmc/nci.t at line 28) # got: 'Undefined symbol "lrint" # ' # expected: 'loaded # dlfunced # ok 1 # ' # Looks like you failed 1 tests of 2. ((( output_is(<<'CODE', <<'OUTPUT', "nic_i_d"); loadlib P1, "libm.so" print "loaded\n" dlfunc P0, P1, "lrint", "id" print "dlfunced\n" set N5, 4.1 invoke ne I5, 4, nok_1 print "ok 1\n" end nok_1: print "nok 1\n" print I5 print "\n" end CODE loaded dlfunced ok 1 OUTPUT ))) Alex Gough -- Why on earth was William of Orange? (Seriously, though.)
