jhalfs experimental BLFS

George Boudreau georgeb at linuxfromscratch.org
Mon Mar 20 12:39:50 PST 2006



M.Canales.es wrote:
> El Lunes, 20 de Marzo de 2006 20:47, George Boudreau escribió:
>> Manuel,
>>    I see you have been very active with the hlfs code.

> I will ask in the hlfs-dev list if a default config file can be created, like 
> with the kernel.

   The uClibc config patch is actually the config file and does/did 
produce a bootable partition in the past..

  >>    2. The function 'build_makefile' will have to be reworked to handle
>> both a 'chroot jail' build as well as a 'boot minimal' style.  The
>> method I used in the makefile is miserable and relies on a failure to
>> exit the makefile script prematurely.. (bad.. very bad)
> 
> Maybe we will need to create a diferent set of Makefile's targets based on if 
> using the "chroot" or "boot" method. I have not run yet the ./clfs script, 
> then don't know how that is currently handled.
  yes.. this is what I see also. At the moment, for a chroot build, we 
are doing a 'make all'  For a boot build we chuld loose the 'all' and 
only have 'make tools' and after a reboot a 'make system'

>>    3. The package downloading functions pulls down all patches for a
>> specific version. In LFS,BLFS,HLFS this is not an issue but with clfs it
>> will pull patches for all the architectures.. i.e. x86 build ends up
>> with space/mips..etc patches.. They do not affect the build but the do
>> pollute the the SRC_ARCHIVE.. (more a pain in the buttocks than a problem)
> 
> For HLFS and the CLFS patches look easy. For CLFS I can't figure yet how to 
> solve it, but must be solved ;-)
>  
   Has someone broken the clfs x86 book. I am no longer able to run the 
bash without errors. I had no problem earlier in the day???

   grb



More information about the alfs-discuss mailing list