View previous topic :: View next topic |
Author |
Message |
childc2 n00b
Joined: 17 Nov 2010 Posts: 33
|
Posted: Wed Nov 17, 2010 5:20 pm Post subject: [SOLVED] compiler troubles |
|
|
Hey everyone,
I'm working on installing Gentoo amd64 on desktop of mine and it's having a bit of trouble compiling the kernel.
Usually it seg faults on some random .c file halfway though. Right now it gets stuck here:
Code: | jail livecd linux # make && make modules_install
CHK include/linux/version.h
CHK include/generated/utsrelease.h
CALL scripts/checksyscalls.sh
CHK include/generated/compile.h
LD vmlinux.o
ld: kernel/built-in.o symbol number 4933 references nonexistent SHT_SYMTAB_SHNDX section
kernel/built-in.o: could not read symbols: File format not recognized
make: *** [vmlinux.o] Error 1 |
I've also tried the 32 bit version. I can usually get the kernel to compile then, but anything I emerge fails to compile.
I know random seg faults in gcc tend to be hardware problems so I tried the CPU test here and I ran memtest a few times. My CPU is only 34C after quite a few attempts at compiling 7 or 8 kernels at once and memtest passed a few times so they seem okay.
Anyone know what might be causing this?
edit: added solved to subject - childc2 12/29/2010
Last edited by childc2 on Wed Dec 29, 2010 9:51 pm; edited 1 time in total |
|
Back to top |
|
|
Sadako Advocate
Joined: 05 Aug 2004 Posts: 3792 Location: sleeping in the bathtub
|
Posted: Wed Nov 17, 2010 7:26 pm Post subject: |
|
|
If subsequent "make"s of the kernel all fail with the same error, it suggests to me that some object was compiled "wrong", but you're not actually seeing any error until the kernel tries to build the kernel binary itself from the pre-compiled binaries.
Run `make clean` to clear out all compiled objects and run make again to compile everything from scratch, but this is just as likely to error out on "some random .c" as it was before...
I'd install sci-mathematics/gimps and play around with mprime, which will give your cpu a proper stress test.
Seriously, it sounds like a hardware issue to me too, and mprime will push the cpu a lot more than just a few parallel compiles, it even uncovered a ram issue for me under stress which memtester failed to catch. _________________ "You have to invite me in" |
|
Back to top |
|
|
childc2 n00b
Joined: 17 Nov 2010 Posts: 33
|
Posted: Wed Dec 29, 2010 9:50 pm Post subject: |
|
|
Looks like you were right. I put in a new CPU and memory and it compiled like normal.
Thanks! |
|
Back to top |
|
|
ThinkTankTed n00b
Joined: 05 Nov 2010 Posts: 4
|
Posted: Mon Jan 10, 2011 5:53 pm Post subject: Doh! Only change one thing at a time when troubleshooting! |
|
|
childc2 wrote: | Looks like you were right. I put in a new CPU and memory and it compiled like normal.
Thanks! |
Now you don't know if it was the memory, CPU or both that were causing your issue...moot point if both exchanged at no cost for new components..
Ted |
|
Back to top |
|
|
childc2 n00b
Joined: 17 Nov 2010 Posts: 33
|
Posted: Tue Jan 11, 2011 6:55 pm Post subject: |
|
|
I haven't looked back over the old hardware yet to find out what broke, but I'll let ya know if I find something.
Edit: Okay turns out it was one of the RAM slots on the motherboard. Skipped it and now the machine works nicely. - childc2 02/01/2011 |
|
Back to top |
|
|
tophyr n00b
Joined: 02 Dec 2012 Posts: 1 Location: Seattle
|
Posted: Sun Dec 02, 2012 5:30 pm Post subject: |
|
|
Sorry to ressurect an old thread - but I am getting the same error on an Android build in a virtual machine. How did you go about determining that the problem was a bad RAM slot? I'm currently trying to figure out if there is something wrong with my hardware, or if VMware is somehow introducing this error virtually. |
|
Back to top |
|
|
childc2 n00b
Joined: 17 Nov 2010 Posts: 33
|
Posted: Mon Dec 03, 2012 3:43 pm Post subject: |
|
|
It's been a while since I've dealt with this, but if I remember correctly it was a bad RAM slot. I could use each stick of RAM alone without any errors, but whenever there was a stick in the 2nd or 3rd slot (I don't remember which) the kernel wouldn't compile.
Hope that helps. |
|
Back to top |
|
|
|