Martin Sebor wrote:
Thanks for fixing it, Travis. I can't help but be a little worried
about the other "little problem on Windows" that Farid said he'd
found in STDCXX-426 and fixed by changing the "r" to "rb" I guess
we'll see tomorrow...

Windows is not looking good. With Farid's patch reverted locales
seem to be processed correctly but all examples and tests are
failing with an ERROR. Andrew, I'd appreciate if you'd look into
it ASAP and either fix the problem or revert your patch.

Martin


Martin

Travis Vitek wrote:
Andrew Black wrote:

Greetings Martin.

I haven't looked with a debugger, but my suspicion is that the new difs
are caused by http://svn.apache.org/viewvc?rev=644314&view=rev .  I
retrieved the results from a random build (34606522 - icl 10, Windows
XP, 12d, 4/1, r643226), and the failures are not observed.  This
particular build occurred after
http://svn.apache.org/viewvc?rev=643120&view=rev was checked in (my
resolution for STDCXX-426), but before Farid's change (
http://svn.apache.org/viewvc?rev=644314&view=rev ) was checked in.

--Andrew Black

Thanks Andrew. That looks like the problem. After spending over an hour
wading through the windows build infrastructure, I finally got enough of
an idea of what was going on so that I could test the change...

The following command line

    ..\bin\exec -d c:\build\stdcxx\trunk\examples spell.exe radix.exe
tele.exe

Produces this output with Farid's change,

    NAME         STATUS WARN ASSERTS FAILED PERCNT    USER     SYS
REAL
    spell.exe      DIFF    0                         0.015   0.000
0.000
    radix.exe      DIFF    0                         0.015   0.015
0.000
    tele.exe       DIFF    0                         0.015   0.000
0.000

And produces this after removing it and rebuilding the exec utility.

    NAME         STATUS WARN ASSERTS FAILED PERCNT    USER     SYS
REAL
    spell.exe         0    0                         0.015   0.000
0.000
    radix.exe         0    0                         0.015   0.000
0.000
    tele.exe          0    0                         0.015   0.000
0.000

I'll be committing a fix in a few seconds.

Travis

Martin Sebor wrote:
Looks like all examples are being reported in the DIFF status
in all recent Windows builds, including a few from this morning.
I suspect it might have something to do with Andrew's recent
patch to the exec utility. Andrew, or someone with access to
a Windows machine, can you please look into this? It would
be good to get this resolved before the weekend.

  http://people.apache.org/~sebor/stdcxx/results/

Thanks
Martin


Reply via email to