bugGNU Octave - Bugs: bug #33736, ./configure on MinGW / WinXP...

 
 

bug #33736: ./configure on MinGW / WinXP introduces hard disk "errors"

Submitted by:  None
Submitted on:  Sat 09 Jul 2011 08:34:52 PM UTC  
 
Category: Configuration and Build SystemSeverity: 5 - Blocker
Priority: 5 - NormalItem Group: Other
Status: FixedAssigned to: None
Originator Name: Philip NienhuisOriginator Email: -unavailable-
Open/Closed: ClosedRelease: 3.4.2
Operating System: Microsoft Windows

Add a New Comment(Rich Markup)
   

You are not logged in

Please log in, so followups can be emailed to you.

 

Thu 24 Nov 2011 10:30:20 AM UTC, comment #3:

Indeed, this issue is gone with a new tip checkout & newest gnulib (as of today).
I'll close this bug and set its status to "fixed".

Philip Nienhuis <philipnienhuis>
Project Member
Thu 24 Nov 2011 08:20:27 AM UTC, comment #2:

Indeed, "blocker" may be too severe. After all, it doesn't affect building Octave itself; rather it is a nuisance (although a big one at that).

I'll investigate if it still exists & report back; may take some time.
If it doesn't pop up again, I think we can close this bug.

Philip Nienhuis <philipnienhuis>
Project Member
Thu 24 Nov 2011 06:13:47 AM UTC, comment #1:

We should not add more system specific knowledge to the configure script. It should test for features, not types of systems or version numbers.

The test that prints this message comes from the gnulib rename module, so I think the -unavailable- mailing list is the right place to report the problem.

John W. Eaton <jwe>
Project Administrator
Sat 09 Jul 2011 08:34:52 PM UTC, original submission:

The configure script, when run under MinGW/MSYS on Windows-XP, introduces a hard disk "error". While trying to "check if rename manages hard links correctly" a systray message pops up:

"conftest.exe - Corrupt file"

and on reboot a chkdsk is performed.
(The latter is nasty on 150 GB NTFS partitions with > 250,000 files on it. Takes ages to complete).

I think there are no real disk errors introduced, it's just that configure's rename/hard link test under MinGW/MSYS seems to confuse Windows' HD disk driver so that it sets the "dirty" flag.
Tatsuro reported that he didn't see this anymore on Windows 7.

Nevertheless, as it is known that hard links don't exist on Windows, this configure test had better be left out when ./configure detects it runs under Windows or MinGW (is that at all possible?)
Or could a configure option "--enable-MinGW" or so be added to avoid this test?

BTW I didn't hit this issue when building additional tools and dependencies (which also invoke configure).

Anonymous

 

(Note: upload size limit is set to 16384 kB, after insertion of the required escape characters.)

Attach File(s):
   
   
Comment:
   

No files currently attached

 

Depends on the following items: None found

Items that depend on this one: None found

 

Carbon-Copy List
  • -unavailable- added by philipnienhuis (Posted a comment)
  • -unavailable- added by jwe (Posted a comment)
  • -unavailable- added by rik5 (Updated the item)
  • -unavailable- added by tmacchant
  • -unavailable- added by None (Submitted the item)
  •  

    Do you think this task is very important?
    If so, you can click here to add your encouragement to it.
    This task has 0 encouragements so far.

    Only project members can vote.

     

    Please enter the title of George Orwell's famous dystopian book (it's a date):

     

     

    Follow 4 latest changes.

    Date Changed By Updated Field Previous Value => Replaced By
    Thu 24 Nov 2011 10:31:35 AM UTCphilipnienhuisOpen/ClosedOpen=>Closed
    Thu 24 Nov 2011 10:30:20 AM UTCphilipnienhuisStatusNone=>Fixed
    Thu 24 Nov 2011 04:04:27 AM UTCrik5Severity3 - Normal=>5 - Blocker
    Sat 09 Jul 2011 09:29:38 PM UTCtmacchantCarbon-Copy-=>Added -unavailable-

    Back to the top


    Powered by Savane 3.1-cleanup1