Steve Hay <[EMAIL PROTECTED]> writes: >Nick Ing-Simmons wrote: > >>The source for the native Win32 layer is in perforce. >>I think the layer even gets built one can do >> >>set PERLIO=:win32:crlf >>dmake test >> >I just tried this (using VC++ / nmake) with bleadperl (normally all >tests OK). Seemingly every test fails, many saying "FAILED before any >test output arrived" and many producing popup application error windows, >or hanging. I soon gave up and killed it :(
It used to do better than that! But I can't remember if I was building threaded perls then. > >Things seem better when not using test/harness, e.g.: > >C:\p5p\bleadperl\t>.\perl -I..\lib harness base\cond.t >base/cond....FAILED before any test output arrived >FAILED--1 test script could be run, alas--no output ever seen > >C:\p5p\bleadperl\t>.\perl -I..\lib base\cond.t >1..4 >ok 1 >ok 2 >ok 3 >ok 4 > >so maybe it's not as bad as it looks. Not sure what's wrong with t/harness. Hmm, that is a pipe interface and possibly fork. If I remember my history it is possible that :win32 layer's Dup() hook isn't up to final spec of that entry point. > >PERLIO=:stdio:crlf doesn't quite work properly either: > >Failed Test Stat Wstat Total Fail Failed List of Failed >------------------------------------------------------------------------------- >../ext/Socket/t/socketpair.t 1 256 45 80 177.78% 6-45 >../lib/FileHandle.t 1 256 12 4 33.33% 11-12 >../lib/IPC/Open2.t 7 2 28.57% 3 7 >../lib/IPC/Open3.t 1 256 22 23 104.55% 3-4 9-13 15-22 >io/fflush.t 7 2 28.57% 6-7 >io/openpid.t 10 3 30.00% 6 8-9 > >(Most of the failing tests there hung and needed killing for the test >suite to continue.) > >- Steve > > > >------------------------------------------------ >Radan Computational Ltd. > >The information contained in this message and any files transmitted with it are >confidential and intended for the addressee(s) only. If you have received this >message in error or there are any problems, please notify the sender immediately. >The unauthorized use, disclosure, copying or alteration of this message is strictly >forbidden. Note that any views or opinions presented in this email are solely those >of the author and do not necessarily represent those of Radan Computational Ltd. The >recipient(s) of this message should check it and any attached files for viruses: >Radan Computational will accept no liability for any damage caused by any virus >transmitted by this email.
