Re: [Ltib] Binary toolchain distributed

[Stuart Hughes]

Hi Svein,

I just wanted to add to Steve's reply to say that from my point of view 
the reason for using binary toolchains is that they have been tested.

What we're trying to avoid is letting a user build a toolchain on an 
unknown platform and then find out later that this has injected some 
subtle bug rendering everything built with it invalid.

Of course people can build the toolchains (we upload srpms for each 
rpm), but we don't recommend it unless you really need to for some reason.

BTW: for some platforms LTIB will build gcc for the target (Canadian 
cross).  The implication of course is you have to test these yourself.

Another thing to bear in mind is we don't normally build glibc/uClibc 
with LTIB, instead we copy the headers/libraries from the cross 
toolchain being used.  This is done to guarantee compatibility between 
the toolchain and the target C library, and also to allow building for 
uClibc or glibc.

Regards, Stuart


Svein Seldal wrote:
> More questions if you don't mind:
>
>
> Why is the toolchain (like tc-fsl-x86lnx-armeabi-nptl-4.1.2-3.i386.rpm) 
> distributed in binary form? I see a note in ltib stating that it one 
> package we cannot build ourselves. Is there a particular reason why this 
> cannot be built during hostcf (except from time and cpu consumption)?
>
>
>
> - Svein