ubuntu-sdk-bugs team mailing list archive

[Nathan Osman <1441961@xxxxxxxxxxxxxxxxxx>]

Public bug reported:

This is probably the strangest bug I have ever observed, but it is
completely reproducible. When running the attached Qt application, I
receive the following error:

    QObject::connect: signal not found in Test

This error is 100% reproducible on my Raspberry Pi 2 running Ubuntu
14.04.2 (Trusty). This device has an ARMv7 CPU (Broadcom BCM2836). I
cannot reproduce this error on any other hardware (x86 and amd64 have
been tested). This does not appear to be a compiler bug since both g++
and Clang produce an executable that prints the above error.

Here's the thing that really confuses me: the error disappears (and
everything works correctly) if I comment out the "dataChanged(...)" line
in test.cpp. This is completely bizarre because the line is never
executed by any code.

I've tried digging through Qt's headers and buried myself neck-deep in
macros, templates, and MOC-generated files. I can't make any sense out
of this. Therefore I am reporting it as a bug. To compile the example,
enter the source directory, run "qmake", and then run "make" to build
the executable. Remember, the bug only manifests itself on an ARM CPU.

** Affects: qtbase-opensource-src (Ubuntu)
     Importance: Undecided
         Status: New

** Attachment added: "Source code for the example"
   https://bugs.launchpad.net/bugs/1441961/+attachment/4370042/+files/test-src.tar.gz

-- 
You received this bug notification because you are a member of Ubuntu
SDK bug tracking, which is subscribed to qtbase-opensource-src in
Ubuntu.
https://bugs.launchpad.net/bugs/1441961

Title:
  dataChanged signal cannot be used on ARM under certain conditions

Status in qtbase-opensource-src package in Ubuntu:
  New

Bug description:
  This is probably the strangest bug I have ever observed, but it is
  completely reproducible. When running the attached Qt application, I
  receive the following error:

      QObject::connect: signal not found in Test

  This error is 100% reproducible on my Raspberry Pi 2 running Ubuntu
  14.04.2 (Trusty). This device has an ARMv7 CPU (Broadcom BCM2836). I
  cannot reproduce this error on any other hardware (x86 and amd64 have
  been tested). This does not appear to be a compiler bug since both g++
  and Clang produce an executable that prints the above error.

  Here's the thing that really confuses me: the error disappears (and
  everything works correctly) if I comment out the "dataChanged(...)"
  line in test.cpp. This is completely bizarre because the line is never
  executed by any code.

  I've tried digging through Qt's headers and buried myself neck-deep in
  macros, templates, and MOC-generated files. I can't make any sense out
  of this. Therefore I am reporting it as a bug. To compile the example,
  enter the source directory, run "qmake", and then run "make" to build
  the executable. Remember, the bug only manifests itself on an ARM CPU.

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/qtbase-opensource-src/+bug/1441961/+subscriptions