On Thu, Oct 11, 2007 at 10:58:36AM -0400, Kaloyan M. Penev wrote: > Hi, > > I recently installed debian stable. Immediately after that I downloaded > and compiled myself the latest kernel version: 2.6.22.8. Since my machine > is quad core I enabled SMP and other things that seemed appropriate. The > system boots up and works almost fine, however I am experiencing the > following weird problems: > > When I try to compile something with g++ every now and then I get the > following message:
compiling is a great way to find hardware problems. my first thought is bad memory. Use a systematic approach to test each stick in each slot. Lots of people recommend memtest, but I find it doesn't always find the errors. > > ./matrix.cpp:602: internal compiler error: Segmentation fault > Please submit a full bug report, > with preprocessed source if appropriate. > See <URL:http://gcc.gnu.org/bugs.html> for instructions. > For Debian GNU/Linux specific bug reporting instructions, > see <URL:file:///usr/share/doc/gcc-4.1/README.Bugs>. > The bug is not reproducible, so it is likely a hardware or OS problem. > > Also because I had to install the fglrx driver from the radeon web page > (v 8.41.7), which for some reason requires libstdc++5, and everything else > requires libstdc++6 when I compile opengl applications I get a warning > that there may be possible conflicts between libstdc++5 and libstdc++6 and > sometimes the compiled executable works and sometimes does not. > > Also every now and then everything just freezes and the only thing left > for me to do is press the reset button. Looking at kern.log and syslog > shows nothing suspicious around the time this happens. I did search > through kern.log and syslog for suspicios messages and I have attached all > I found below (perhaps there are more): > > kern.log: > potato kernel: No NUMA configuration found > potato kernel: Faking a node at 0000000000000000-000000007fef0000 > potato kernel: swsusp: Registered nosave memory region: 000000000009b000 - > 000000000009c000 > > potato kernel: usb 2-2: reset high speed USB device using ehci_hcd and > address 2 (this repeats many times while running, every few minutes if I > use the device) > > potato kernel: ClientGUI[7156] general protection rip:2b6259cf66ce > rsp:7fff534bf4c0 error:0 > potato kernel: ClientGUI[7296]: segfault at ffffffffffffffe8 rip > 00002b9d486c76d4 rsp 00007fff64aecaf0 error 4 > potato kernel: ClientGUI[8543] general protection rip:2b2c244416ce > rsp:7fff88d72d80 error:0 > potato kernel: ClientGUI[8706]: segfault at 0000000000000110 rip > 00002b363e36c6ce rsp 00007fff6ee47e50 error 4 > (where ClientGUI is an opengl application that I am writing, uses > wxwidgets, gl and glu) > > Oct 4 15:03:26 potato kernel: WARNING: at fs/buffer.c:570 > __remove_assoc_queue() > Oct 4 15:03:26 potato kernel: > Oct 4 15:03:26 potato kernel: Call Trace: > Oct 4 15:03:26 potato kernel: [<ffffffff802991b6>] > drop_buffers+0xa4/0xec Oct 4 15:03:26 potato kernel: > [<ffffffff80299255>] try_to_free_buffers+0x57/0x > 9b > Oct 4 15:03:26 potato kernel: [<ffffffff8026030f>] > shrink_inactive_list+0x4ed/ > 0x83a > Oct 4 15:03:26 potato kernel: [<ffffffff8025eef4>] > __pagevec_release+0x19/0x22 > Oct 4 15:03:26 potato kernel: [<ffffffff8025fc40>] > shrink_active_list+0x483/0x491 > Oct 4 15:03:26 potato kernel: [<ffffffff804a7c8c>] > thread_return+0x0/0xdb Oct 4 15:03:26 potato kernel: > [<ffffffff80260750>] shrink_zone+0xf4/0x11d > Oct 4 15:03:26 potato kernel: [<ffffffff8026116e>] kswapd+0x2da/0x489 > Oct 4 15:03:26 potato kernel: [<ffffffff8023e10e>] > autoremove_wake_function+0x > 0/0x2e > Oct 4 15:03:26 potato kernel: [<ffffffff80260e94>] kswapd+0x0/0x489 > Oct 4 15:03:26 potato kernel: [<ffffffff8023dfee>] kthread+0x47/0x75 > Oct 4 15:03:26 potato kernel: [<ffffffff8020a378>] child_rip+0xa/0x12 > Oct 4 15:03:26 potato kernel: [<ffffffff8023dfa7>] kthread+0x0/0x75 > Oct 4 15:03:26 potato kernel: [<ffffffff8020a36e>] child_rip+0x0/0x12 > > > The same messages repeat of course in syslog. > > The last simptom that I notice is that when I mount an external SATA over > USB disk it starts reading from the disk and it takes it about 30 sec to a > minute to stop (the same exact disk does not behave that way when I plug > it into another computer running debian). I don't think this is necessarily indicative of anything unless you know you've got the same packages installed on both machines, although that USB message above may point to a flaky controller or driver. Look at the package differences between the two systems first, as that's fairly easy to compare. Tail -f the logs while using that disk to see if those errors crop up during that spin-up time and try plugging into a different USB port to see if you can get on a different controller to start isolating that as potential problem (lsusb helps here). Finally, don't discount the possibility of a failing power supply. A failing power supply can wreak subtle havoc everywhere. A
signature.asc
Description: Digital signature
