FS#35159 - [qt4] segfaults on x86

Attached to Project: Arch Linux
Opened by Thomas Lübking (luebking) - Wednesday, 08 May 2013, 16:02 GMT
Last edited by Andrea Scarpino (BaSh) - Friday, 10 May 2013, 18:10 GMT
Task Type Bug Report
Category Packages: Extra
Status Closed
Assigned To Andrea Scarpino (BaSh)
Evangelos Foutras (foutrelis)
Architecture i686
Severity Medium
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 1
Private No

Details

Description:
See summary, this simple app reliably segfaults here:

----------------

#include <QCoreApplication>
#include <QDateTime>
#include <QDebug>

int main (int argc, char **argv)
{
QCoreApplication a(argc, argv);

QTime t;
qDebug() << t.addMSecs(-123456789).toString();
QDateTime().setMSecsSinceEpoch(5);

return a.exec();
}

---------
// g++ `pkg-config --libs --cflags QtCore` -o qtimetest qtimetest.cpp



It doesn't matter whether QDateTime is the default constructed, an explicit or ::currentateTime()
The segfault then occurs in QTime::addMSecs() which however works (regardless of the parameter being >=< 0), why i assume this is some binary offset issue.

Also reverting to qt4-4.8.4-16 *only* passes this test (but because of the broken libpng dep would fail in QtGui) and re-upgrading reliably fails it.

Marking as "critical" since if this is not because my HW has a very magic bug, this will have widespread impact on Qt applications "randomly" crashing (eg. Qupzilla when writing to history)

Additional info:
* package version(s)
qt4-4.8.4-17
This task depends upon

Closed by  Andrea Scarpino (BaSh)
Friday, 10 May 2013, 18:10 GMT
Reason for closing:  Fixed
Additional comments about closing:  qt4 4.8.4-18
Comment by Andrea Scarpino (BaSh) - Wednesday, 08 May 2013, 19:32 GMT
  • Field changed: Summary (qt4-4.8.4-17 / x86 segfaults in QDateTime::setMSecsSinceEpoch() - qt4-4.8.4-16 is ok. → [qt4] segfaults on x86)
  • Field changed: Details
  • Field changed: Status (Unconfirmed → Assigned)
  • Field changed: Severity (Critical → Medium)
  • Task assigned to Andrea Scarpino (BaSh)
(fixed the test code)
Comment by Thomas Lübking (luebking) - Friday, 10 May 2013, 17:13 GMT Comment by Evangelos Foutras (foutrelis) - Friday, 10 May 2013, 17:58 GMT
The last GCC snapshot (4.8-20130502 packaged as 4.8.0-4) contains a fix for segfaults happening on i686. (Firefox had a similar issue.)

A simple rebuild seems to fix this; expect new packages shortly.
Comment by Andrea Scarpino (BaSh) - Friday, 10 May 2013, 18:01 GMT

Loading...