Description of problem: /usr/bin/vncviewer from tigervnc segfaults on x86_64 and i386 Fedora 12 systems. Version-Release number of selected component (if applicable): tigervnc-1.0.0-2.fc11.x86_64 How reproducible: Use vncviewer for a period of time. Eventually, it crashes. There doesn't appear to be a reliable set of circumstances under which it crashes, but I suspect that it's related to automatic colour-depth (or possibly encoding) changes. Additional info: gdb crash info and backtrace: Program received signal SIGSEGV, Segmentation fault. rfb::TransImageGetter::setColourMapEntries (this=0x0, firstCol=0, nCols=0, writer=0x0) at TransImageGetter.cxx:244 244 nCols = (1 << pb->getPF().depth) - firstCol; (gdb) bt #0 rfb::TransImageGetter::setColourMapEntries (this=0x0, firstCol=0, nCols=0, writer=0x0) at TransImageGetter.cxx:244 #1 0x0000000000408384 in DesktopWindow::handleTimeout (this=0x17f96f0, timer=0x0) at DesktopWindow.cxx:323 #2 0x000000000042e666 in rfb::Timer::checkTimeouts () at Timer.cxx:106 #3 0x000000000040aa51 in CConn::blockCallback (this=0x7fffd87fc820) at CConn.cxx:178 #4 0x0000000000462599 in rdr::FdInStream::readWithTimeoutOrCallback ( this=0x17f01d0, buf=<value optimized out>, len=<value optimized out>, wait=true) at FdInStream.cxx:239 #5 0x0000000000462cf2 in rdr::FdInStream::readBytes (this=0x17f01d0, data=<value optimized out>, length=14173) at FdInStream.cxx:122 #6 0x000000000042fb0f in rfb::DecompressJpegRect(const rfb::Rect &, struct rdr::InStream *, rdr::U32 *, struct rfb::CMsgHandler *) (r=@0x7fffd87fc230, is=<value optimized out>, buf=<value optimized out>, handler= 0x7fffd87fc820) at ../rfb/tightDecode.h:249 #7 0x0000000000430e01 in rfb::tightDecode32 (r=@0x7fffd87fc230, is=0x17f01d0, zis=<value optimized out>, buf=<value optimized out>, handler=<value optimized out>) at ../rfb/tightDecode.h:100 #8 0x00000000004506cc in rfb::CMsgReader::readRect (this=0x17f8eb0, r=@0x7fffd87fc230, encoding=7) at CMsgReader.cxx:115 #9 0x000000000041df62 in rfb::CMsgReaderV3::readMsg (this=0x17f8eb0) at CMsgReaderV3.cxx:94 #10 0x00000000004160bb in main (argc=<value optimized out>, argv=0x7fffd87fff28) at vncviewer.cxx:407 (gdb) print nCols $1 = 0 (gdb) print pb->getFP().depth Cannot access memory at address 0x10 (gdb) print firstCol $2 = 0
Another curiosity: abrt isn't catching these segfaults (on multiple boxes).
*** This bug has been marked as a duplicate of bug 551262 ***