This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Potential gcc-3.1 install problem


>> Thank you for the detailed report.  We do not recommend non-developers
>> (of gcc) rerunning make within a bootstrap tree (fully made or partial).

> Sure, and I was not trying to do that, I was only running make
> install.

Sure, how about this instead:

``Thank you for the detailed report.  We do not recommend
  non-developers (of gcc) rerunning make [with any particular
  arguments] within a bootstrap tree (fully made or partial) [after
  source changes have been made].''

> What happened to me could happen to anyone who runs into
> some trouble during make install and tries to troubleshoot editing the
> Makefile.  I did not expect that editing the Makefile causes
> recompilation of some objects, and the result is a compiler that seems
> to work correctly, except it handles errors and warnings incorrectly.
> When it happened to me, I thought gcc was broken, and I was searching
> the mail list if anyone else had this problem, and I did not find
> anything, so I figured I must be doing something wrong, and I have to
> debug this myself.

How true.  What would you have the Makefile do differently?  BTW, the
rules that force recompilation when you touch the generated Makefile
are themselves within the Makefile...

I think what you are failing to recognize is that running quickstrap
is the only (one-step) documented way to update a completely bootstrap
tree after making the sort of change you made.  Some developers know
when it is safe to do otherwise, but this information is not well
documented.

Regards,
Loren


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]