Re: 答复: How to run gcc test suite in pure mingw32 environment?

[Tim Prince-4]

徐持恒 wrote:

> But can you tell me why there are no testresult of MinGW or Cygwin on
> gcc-testresults mailinglist ?
>

These topics hardly seems appropriate for gcc development; diverting to
gcc-help.
In your original message, you seemed to be aware that cygwin would be a
more appropriate framework for running testsuite, but you wanted a test
harness independent of cygwin.  Then you confused me totally by
suggesting that the mingw compiler should be built on newlib (a port of
newlib to mingw?  I'm not aware of any interest in that).  I didn't see
how that connected with your other topics.  I thought the non-use of
newlib and cygwin run-time was the central goal of mingw.
  I've been posting cygwin results on gcc-testresults regularly, but
gcc-4.5 hasn't been building on cygwin the last 7 weeks.
In the absence of support for 64-bit native compilation, usefulness of
and interest in cygwin compiler continues to decline.  I suppose it may
drop off the secondary platform list, if it continues not to build.
I didn't catch when or why make -j 4 test became broken on cygwin; it
used to work well.

[Dave Korn-6]

Tim Prince wrote:
> I've been posting cygwin
> results on gcc-testresults regularly, but gcc-4.5 hasn't been building on
> cygwin the last 7 weeks.

  WJFFM.  HEAD built just fine for me last night, but no test results yet.
Last posted results:

http://gcc.gnu.org/ml/gcc-testresults/2009-10/msg02682.html

> decline.  I suppose it may drop off the secondary platform list, if it
> continues not to build.

  How would anyone even know, since you never filed a PR?

> I didn't catch when or why make -j 4 test became
> broken on cygwin; it used to work well.

  WJFFM.  Well, "fine" apart from being slow compared to -j8, which is what I
use all the time.

    cheers,
      DaveK

[Tim Prince-4]

Dave Korn wrote:

> Tim Prince wrote:
>> I've been posting cygwin
>> results on gcc-testresults regularly, but gcc-4.5 hasn't been building on
>> cygwin the last 7 weeks.
>
>   WJFFM.  HEAD built just fine for me last night, but no test results yet.
> Last posted results:
>
> http://gcc.gnu.org/ml/gcc-testresults/2009-10/msg02682.html
>
>> decline.  I suppose it may drop off the secondary platform list, if it
>> continues not to build.
>
>   How would anyone even know, since you never filed a PR?
>

I didn't see the point in filing a PR for such a mysterious failure
(bootstrap compiler can't run the test to confirm .o suffix, test not
visible in log file).  I have several multi-year PRs posted, so that
isn't a short term solution.

[Dave Korn-6]

[ CCs snipped and subject adjusted since we're changing topic here. ]

Tim Prince wrote:

> Dave Korn wrote:
>> Tim Prince wrote:
>>> I've been posting cygwin
>>> results on gcc-testresults regularly, but gcc-4.5 hasn't been building on
>>> cygwin the last 7 weeks.
>>   WJFFM.  HEAD built just fine for me last night, but no test results yet.
>> Last posted results:
>>
>> http://gcc.gnu.org/ml/gcc-testresults/2009-10/msg02682.html
>>
>>> decline.  I suppose it may drop off the secondary platform list, if it
>>> continues not to build.
>>   How would anyone even know, since you never filed a PR?
>>
> I didn't see the point in filing a PR for such a mysterious failure
> (bootstrap compiler can't run the test to confirm .o suffix, test not
> visible in log file).  

  If
    - it doesn't happen to anyone else
and
    - you don't tell anyone about it
then
    - how do you expect it to lead to Cygwin falling off the secondary
platform list?

> I have several multi-year PRs posted, so that isn't a short term solution.

  I checked BZ(*).  ISTM that with the exception of some fairly specialised
problems in fortran vectorisation, every other PR you filed has been resolved.
 Your implicit assumption here that because some PRs have not been resolved
quickly, therefore all PRs will not be resolved quickly is transparently not
the case, so there's really no need to go cutting off your nose to spite your
face: if you have a reproducible problem, the odds are (based on past history)
that it will be worth reporting.

  I don't know why you have all this trouble and others don't.  I think I may
vaguely remember asking you once before if it had always been on the same
computer and you saying not so it can't be a hardware problem, but it's really
hard to imagine anything wrong that you could be doing in your procedure that
would have these kinds of effects.  What's the electricity supply quality like
where you live?  I'm starting to get to the point of having considered
everything possible and wondering about the impossible things.  Do you live
next to a nuclear reactor or in the beam of a particle accelerator by any
chance?  Are you orbiting in an unshielded spacecraft?  Have you painted your
house using a tin of old radium-based glow-in-the-dark paint you found in the
attic?  Finally, is your PC perhaps built on top of an old indian burial
ground, and/or have you disturbed any crypts, pyramids, or other tombs and/or
been placed under any ancient curses that you are aware of?

    cheers,
      DaveK
--
(*) -
http://gcc.gnu.org/bugzilla/buglist.cgi?query_format=advanced&short_desc_type=allwordssubstr&short_desc=&known_to_fail_type=allwordssubstr&known_to_work_type=allwordssubstr&long_desc_type=substring&long_desc=&bug_file_loc_type=allwordssubstr&bug_file_loc=&gcchost_type=allwordssubstr&gcchost=&gcctarget_type=allwordssubstr&gcctarget=&gccbuild_type=allwordssubstr&gccbuild=&keywords_type=allwords&keywords=&emailassigned_to1=1&emailreporter1=1&emailtype1=substring&email1=tprince&emailassigned_to2=1&emailreporter2=1&emailcc2=1&emailtype2=substring&email2=&bugidtype=include&bug_id=&votes=&chfieldfrom=&chfieldto=Now&chfieldvalue=&cmdtype=doit&order=Reuse+same+sort+as+last+time&field0-0-0=noop&type0-0-0=noop&value0-0-0=