Re: [gentoo-user] another openoffice build failure

[Allan Gottlieb <gottlieb@nyu.edu>]

At Sat, 02 Sep 2006 19:14:56 -0400 Allan Gottlieb <gottlieb@nyu.edu> wrote:

> At Sat, 02 Sep 2006 14:10:10 -0700 Richard Fish <bigfish@asmallpond.org> wrote:
>
>> On 9/2/06, Allan Gottlieb <gottlieb@nyu.edu> wrote:
>>> Inconsistency detected by ld.so: dl-minimal.c: 137: realloc: Assertion `new == ptr' failed!
>>
>> Well this looks suspiciously like a hardware problem to me, but it
>> could also be simply running out of memory.  Can you try:
>>
>> MAKEOPTS=-j1 emerge --oneshot openoffice
>>
>> You might also want to add a fan to blow on your laptop while the
>> build is running... :-)
>
> I should have thought of -j1, thanks.  It is running now.  Will try
> the fan if this fails.  I will also check to see if it dies in the
> same place, which would argue against the hardware problem.

Failed at exactly the same place.  Since each log file is 30,544 lines
long, it doesn't seem likely to be heat related (i.e., I would suspect
that a heat related hardware failure would not be so deterministic as
to fail after so many compiles in the same place).

I am not sure, but it might be that the ebuild forces -j1 by itself

,----
| src_compile() {
| 
| 	unset LIBC
| 	addpredict "/bin"
| 	addpredict "/root/.gconfd"
| 	addpredict "/root/.gnome"
| 
| 	# Should the build use multiprocessing? Not enabled by default, as it tends to break
| 	export JOBS="1"
| 
| ...
| 
| 	./configure ${MYCONF} \
| 		--with-distro="${DISTRO}" \
| 		--with-arch="${ARCH}" \
| 		--with-srcdir="${DISTDIR}" \
| 		--with-lang="${LINGUAS_OOO}" \
| 		--with-num-cpus="${JOBS}" \
`----

I then though of possibly running out of disk space.  That seemed
unlikely since I had 7.5G and it failed comparatively early in the
run.  But to be sure I ran it again with

   while true; do df -h | grep /mnt/hdc10; sleep 10; done >> /tmp/time

also running and the available space never went below 5.9G.  Once
again it failed at the same place.

So I don't see it as a hardware failure, due to the repeatability.
It is not inadequate disk space.
I have 1GB of mem
    ajglap hdc10 # cat /proc/meminfo
    MemTotal:      1033932 kB
    MemFree:         72632 kB
    Buffers:         40180 kB
    Cached:         559036 kB
    SwapCached:          0 kB
    Active:         628840 kB
    Inactive:       253460 kB
    HighTotal:      130876 kB
    HighFree:          284 kB
    LowTotal:       903056 kB
    LowFree:         72348 kB
    SwapTotal:     2008084 kB
    SwapFree:      2007840 kB
    Dirty:              44 kB
    Writeback:           0 kB
    Mapped:         398488 kB
    Slab:            63836 kB
    CommitLimit:   2525048 kB
    Committed_AS:   923704 kB
    PageTables:       2320 kB
    VmallocTotal:   114680 kB
    VmallocUsed:     33344 kB
    VmallocChunk:    77812 kB

I am not sure what to try next.  I suppose I could try to revert back
to gcc3.4.6
    ajglap hdc10 # gcc-config --list-profiles
     [1] i686-pc-linux-gnu-3.4.6
     [2] i686-pc-linux-gnu-3.4.6-hardened
     [3] i686-pc-linux-gnu-3.4.6-hardenednopie
     [4] i686-pc-linux-gnu-3.4.6-hardenednopiessp
     [5] i686-pc-linux-gnu-3.4.6-hardenednossp
     [6] i686-pc-linux-gnu-4.1.1 *
    ajglap hdc10 #

Any ideas would be welcome

thanks,
allan
-- 
gentoo-user@gentoo.org mailing list