This is the mail archive of the gcc-patches@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: speedup collect2 (by not using it)


Mike Stump <mrs@apple.com> writes:

> On Thursday, May 22, 2003, at 01:02 PM, Zack Weinberg wrote:
>> I think this is a fine idea in principle.  We need to nail down
>> precisely what collect2 is or is not used for, of course, and make
>> sure it's invoked when necessary.
>
> I've done that as much as I can.  The HP and AIX folks kicked in
> exactly what I hoped to hear.

Good.

>> Better check unified tree builds and the like, though.
>
> It should not impact them any.   What concerns do you have?

gcc/configure.in, when it detects a unified tree build, symlinks
../ld/ld-new to collect-ld.  And only collect-ld.  If collect2.c
looks for the linker as "collect-ld", but gcc.c doesn't, it won't
work.

After installation it should all be fine, but I think it may blow up
building the target libraries.

Please check whether or not this is a problem, and if it is, fix it.

> + #else
> + #undef USE_COLLECT2
> + #define USE_COLLECT2 1
> + #endif

Ick.  Just make configure define USE_COLLECT2 to 1 instead of the
empty string.  (It'll be easiest to do this by making USE_COLLECT2 be
a real AC_DEFINE rather than something stuffed into xm_defines.)

> ! 	  /* On a system that does USE_COLLECT2, we must always run
> ! 	     it, otherwise, we don't run collect2 for any reason other
> ! 	     than -frepo.  */

Awkward wording; I would say

  /* If the host configuration has set USE_COLLECT2, we must always
     run collect2.  Otherwise we need only run it if -frepo is in use.  */

zw


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