apr-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Garrett Rooney" <roo...@electricjellyfish.net>
Subject Re: testing on Windows
Date Fri, 14 Apr 2006 20:49:38 GMT
On 4/13/06, John Vandenberg <jayvdb@gmail.com> wrote:
> Hi,
> I am setting up a Windows XP build environment using Microsoft Visual
> C++ 2005 Express and Windows(r) Server 2003 SP1 Platform SDK, and have
> a few questions about the results because I can't compare them with an
> older VC++ setup.
> After MS VC++ upgraded the project files, I found that I needed to
> manually add advapi32.lib to the libapr project properties under
> Linker : Input : Additional Dependencies.

Yeah, I've encountered that problem as well.  If anyone can find a way
to tweak the dsp files such that this does not occur I'd much
appreciate it...

> The only annoyance during the build was this warning occurring for all
> apr-util and apr-iconv commands:
> cl : Command line warning D9002 : ignoring unknown option '/YXiconv.h'
> Replacing /YX with /Yc & Yu fixes the warning, and based on MSDN, that
> syntax is supported by VC5.0
> While figuring out how to test the results, I opened
> apr/test/aprtest.dsp, and found that it depends on awk (contrary to
> the note on compiling_win32.html) and MakeWin32Make.awk, which
> disappeared somewhere between 1.2.1 and 1.2.2.  Is aprtest.dsp useful?
> Building testall reports the following warning.
> 1. cl : Command line warning D9035 : option 'GX' has been deprecated ...
>    cl : Command line warning D9036 : use 'EHsc' instead of 'GX'
> Following those instructions doesnt affect the results.
> Running apt/test/testall reports:
> Failed Tests            Total   Fail    Failed %
> ===================================================
> testpipe                    9      2     22.22%
> testsock                    7      1     14.29%
> Do these failures indicate there is a problem with my setup, or are
> these results to be expected on the Windows platform?

Those are known problems.  It would be nice if our tests had a way to
indicate that certain tests are known to fail on certain platforms...

Look on the bright side, until fairly recently those tests would hang,
not fail ;-)


View raw message