new makefile scheme

George Boudreau georgeb at linuxfromscratch.org
Mon Aug 28 12:20:51 PDT 2006


M.Canales.es wrote:
> El Domingo, 27 de Agosto de 2006 17:17, George Boudreau escribió:
>> Manuel,
>>    I have almost completed the modifications to the makefile design (I
>> have a fresh build running in the background). The only outstanding
>> issue left is to add the size statistics back in.
>>
>>    Once this build finishes I will send you a copy of the LFS
>> master.sh/Makefile and you can make comments. If all looks ok I will
>> work on the other books.
> 
> What is the status of that new Makefile generation?
   I am doing a final pass for the LFS book with ICA enabled. I found 
some problems with the ICA/farce scripts that exist in both trunk and 
the experimental branch that I have not fixed yet.
   When we adopted the 'logs/$@' convention the impact on ICA/farce code 
was overlooked. Right now ICA/farce is broken in the trunk (and 1.0?). 
All instances of logs/$@ in chroot code must be replaced with 
logs/${this_script}. ( I never watched the development of the comparison 
code and did not realized how it handled multiple passes)
> 
> If it looks that work, maybe the new code should be commit to experimental to 
> can add the blfs-tool dependencies installation based on that new Makefile 
> code.
> 
> To avoid having a lot of new features mixed in experimental, the "use only 
> menuconfig" changes can be merged to trunk.
> 
> The current blfs-tool support code could be merged also, if needed to keep a 
> clean experimental tree, commenting-out the required lines in Config.in.
> 
> I'm working now in the build scripts for blfs-tool dependencies.
> 
   I think you should commit all your blfs tools before I add my latest 
batch of bug riddled code. The new makefile format does not add any new 
features and is only an internal style change. I am modifying/replacing 
every wrt_xxxx function and making many changes to each xLFS/master.sh.

  I have a free evening and will make the necessary changes to trunk to 
fix the ICA problem.

   G



More information about the alfs-discuss mailing list