hugin-devs team mailing list archive

[Matt Craig <814280@xxxxxxxxxxxxxxxxxx>]

Did some more testing tonight.

Much more to report...enblend does not *always* fail, even when run from
the gui. Command line does not always work contrary to Mikko Kuutti's
report above.

Details:

GUI or command line: 
Even starting from the same set of images the crash does not always occur at the same time (i.e. in my set of 17 images it sometimes crashed after 2, sometiems 4, sometimes 8, etc).

GUI:
I took a set of images and opened them in hugin. Creating the panorama crashed while running enblend with the abort trap: 6 error. Rerunning "Create panorama" will eventually generate the panaorama, though, if you try enough times. Sometimes it took only two, sometimes as many as 6 or 7.

In at least a couple cases, when the PTBatcherGUI had a few bad builds
in it, hugin successfully created the panorama on the first try. By
"couple cases" I mean I made several duplicates of a folder containing a
set of images, opened hugin and loaded the images from one of the
folders, tried stitching until it worked, then closed hugin and tried
again in a different folder, and sometimes the stitching worked right on
the first try.

There didn't seem to be much consistency: some times the second try
worked, oftentimes not, but eventually it always stitched. Once it
stitched once, it was likely to stitch again in the same hugin session
(i.e. clicking "Create panorama...: again after a success was often, but
not always, successful").

Command line: I believe that Mikko Kuutti was seeing the "eventualy
enblend will succeed" behavior. When I copied all of the tif files
generated by nona to a clean directory and ran the enblend command from
the PTBatchGUI log from the command line it failed. Running it again and
again eventually succeeded. And, on occasion, pasting the command into a
terminal window right after the GUI failed did build the panorama as
Mikko Kuutti reported.

I've attached the output to the terminal from repeated runs of the
enblend command; enblend does not always generate the same error, and
one of them is perhaps more useful than the others:

enblend(3820,0x7fff7afd1960) malloc: *** error for object 0x10d800148: incorrect checksum for freed object - object was probably modified after being freed.
*** set a breakpoint in malloc_error_break to debug
Abort trap: 6

In one of the runs it threw 
Segmentation fault: 11
instead of the abort trap error.

Just to be clear, the attached file was run in a directory that already
contained the tifs generated by nona, and copies of the original jpg
images.

** Attachment added: "command-line-output.txt"
   https://bugs.launchpad.net/hugin/+bug/814280/+attachment/2284953/+files/command-line-output.txt

-- 
You received this bug notification because you are a member of Hugin
Developers, which is subscribed to Hugin.
https://bugs.launchpad.net/bugs/814280

Title:
  OSX 10.7 Hugin 2010.2 and later. gnumake Abort trap: 6 on execution of
  enblend.

Status in Hugin - Panorama Tools GUI:
  Confirmed

Bug description:
  In Mac OS 10.7 Lion, the stitching process runs smoothly for a while,
  and then it stops abruptly with an error.

  The error occurs regardless of the number of images in the project.

  Error log attached.

  Hugin Version: 2011.0.1 (Error also occurs in version 2010.2.0)

  Hardware Info:

    General:

    Model Name:	iMac
    Model Identifier:	iMac5,1
    Processor Name:	Intel Core 2 Duo
    Processor Speed:	2 GHz
    Number of Processors:	1
    Total Number of Cores:	2
    L2 Cache:	4 MB
    Memory:	3 GB
    Bus Speed:	667 MHz
    Boot ROM Version:	IM51.0090.B09
    SMC Version (system):	1.8f2

    Graphics:

    Chipset Model:	ATY,RadeonX1600
    Type:	GPU
    Bus:	PCIe
    PCIe Lane Width:	x16
    VRAM (Total):	128 MB
    Vendor:	ATI (0x1002)
    Device ID:	0x71c5
    Revision ID:	0x0000
    ROM Revision:	113-xxxxxx-139
    EFI Driver Version:	01.00.139

To manage notifications about this bug go to:
https://bugs.launchpad.net/hugin/+bug/814280/+subscriptions