I am not sure whether it is a new kernel bug so I post it in here. Recently I have compiled glibc-2.3.2 with NTPL support based on linux-headers-2.6.5 and also I have compiled kernel-2.6.7-rc2. Running 2.6.7-rc2 version I get a lot of seg. faults: cups daemon, evolution, mozilla, emerge, epiphany etc. Running kde apps like konqueror and konsole with no problems. Running gentoo-dev-sources-2.6.5 everything is O.K. Reproducible: Always Steps to Reproduce: //gdb 6.0 $ gdb evolution (no debugging symbols found)...Using host libthread_db library "/lib/libthread_db.so.1". (gdb) run Starting program: /usr/bin/evolution Program received signal SIGSEGV, Segmentation fault. 0x40010b28 in ?? () $ gdb /usr/lib/mozilla/mozilla-bin (no debugging symbols found)...Using host libthread_db library "/lib/libthread_db.so.1". (gdb) run Starting program: /usr/lib/mozilla/mozilla-bin (no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)... (no debugging symbols found)...(no debugging symbols found)... Program received signal SIGSEGV, Segmentation fault. 0x40010b28 in realloc () from /lib/ld-linux.so.2
Can you please try development-sources-2.6.6 so we can isolate what kernel version you got this problem in? Thanks!
I think this is NPTL troubles. All software needs to be recompiled to use nptl without worries.
It is still happening. I cannot get rid of it :( I disabled nptl, but it's not use. (gdb) run Starting program: /usr/bin/evolution warning: Unable to find dynamic linker breakpoint function. GDB will be unable to debug shared library initializers and track explicitly loaded dynamic code. Program received signal SIGSEGV, Segmentation fault. 0x4000ec58 in ?? () (gdb) where #0 0x4000ec58 in ?? () #1 0x40007217 in ?? () #2 0x64eacf19 in ?? () #3 0x4136a267 in ?? () #4 0x400124fc in ?? () #5 0x08052144 in ?? () #6 0x00000020 in ?? ()
I forgot to mention that: 1. If I compile __ANY__ new kernel and then run it - the same story with seg. faults happens. 2. Sometimes this all thing dissapears after some minutes on running machine.
Can you do a ``FEATURES="nostrip" emerge glibc'' [ Use NPTL if you wish ] so GDB backtraces work properly? Also; can you please merge 2.6.6 headers before merging glibc again? Thanks...
Ok I will do it. But I will not use NPTL and especialy 2.6.6 headers (I use 2.4.21). Is that ok?
Recompiled glibc just like you said - didn't work.
So you mean if you run a backtrace you still get the: "warning: Unable to find dynamic linker breakpoint function. GDB will be unable to debug shared library initializers and track explicitly loaded dynamic code." ... message, or do you mean that you still get the segmentation faults? This doesn't look like a kernel bug to me since I've been running 2.6.7-rcX with 2.6 headers and NPTL and I haven't had any issues. Reassigning to toolchain@gentoo.org. One suggestion is to reboot - that solved a few issues when I migrated over to NPTL and I was getting various Segmentation Faults before I rebooted.
The warning is gone (thanks for the tip) but seg faults - no. And yes - this is not kernel bug. I don't have any ideas what to do :( It seems I will have to rebuild whole system from scratch.
do you still get this problem? also, have you run memtest86 lately?
I reinstalled gentoo from scratch, sorry.
