gcc 3.4.3 make bootstrap error chapter 5

Henry Grefrath henry at grefrath.net
Wed Oct 26 02:51:10 PDT 2005

Ok - it seems that I will have to accept that :-(.

So the approach will be to use the hint for small footprint system 
and compile it on another machine manually forcing the target to be 586?


At 25.10.2005 08:07 +0200, you wrote:

>Matthew Burgess wrote:
>>Henry Grefrath wrote:
>>>Hi Tony,
>>>I was hoping I could build a small system on this laptop. It has 
>>>32MB of RAM and I setup a 75MB swap partition. The rest of the 
>>>hard disk 1.2GB is mounted as /mnt/lfs.
>>I think you might struggle with that, I'm afraid.  Building just 
>>chapter 5 on an Athlon XP 2400+ with 512MB RAM and 512MB swap took 
>>just over 3.5 hours over the weekend (admittedly that included all 
>>the toolchain testsuites as well).  A better approach, if possible, 
>>would be to utilise a much faster machine than your laptop, then 
>>copy the resulting system across, minus any binaries and libraries 
>>you don't require.
>Have to agree, and maybe even stretch it to; it won't work.. :(
>I compiled LFS 6.0 on a P-200MMX a few months ago. 96MB RAM and 7GB 
>HD. First followed my standard procedure and set up 190MB (ca 2xRAM) 
>swap, and started building. Don't remember exactly where the problem 
>started, but think it was the bootstrap phase in ch5. Strange 
>errors.. Tried again, similar, but not the same.. Ok, finally I set 
>up 1GB swap, and managed to compile the whole system. Took me about 
>a week, wouldn't recommend it. :)
>The case is, with low RAM, you gonna need alot of swap, and then 
>there's not enough space left on your 1.2GB HD for building.. :(
>--Tor Olav
>FAQ: http://www.linuxfromscratch.org/lfs/faq.html
>Unsubscribe: See the above information page

More information about the lfs-support mailing list