home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!spool.mu.edu!agate!dog.ee.lbl.gov!horse.ee.lbl.gov!torek
- From: torek@horse.ee.lbl.gov (Chris Torek)
- Newsgroups: comp.lang.c
- Subject: Re: GCC libraries (was GCC and all that)
- Date: 6 Nov 1992 11:26:50 GMT
- Organization: Lawrence Berkeley Laboratory, Berkeley
- Lines: 36
- Distribution: world
- Message-ID: <27266@dog.ee.lbl.gov>
- References: <alien.00qf@acheron.amigans.gen.nz>
- Reply-To: torek@horse.ee.lbl.gov (Chris Torek)
- NNTP-Posting-Host: 128.3.112.15
-
- In article <alien.00qf@acheron.amigans.gen.nz> alien@acheron.amigans.gen.nz
- (Ross Smith) writes:
- >... This library code is, of course, covered by the FSF's licence
- >(the library licence is slightly different from the application licence,
- >but in this context the result is the same), and, therefore, so is the
- >resulting executable.
-
- There is some question as to whether part or all of the libraries
- included as an integral part of GCC `force' code redistribution a la
- the usual (GPL) license. Regardless, however, there is an alternative.
- The HP300 and SPARCstation ports of 4.4BSD both use GCC as the
- standard compiler; however, the entire C libraries of both ports
- are covered only by the usual Berkeley license, occasionally augmented
- with author names and/or LBL credit requirements.
-
- This includes my replacement for the `long long' library used to do
- 64-bit operations on 32-bit CPUs. Mine should be considerably more
- efficient than the one FSF version I saw (which may well have been
- replaced since then). We decided it was safest to rewrite, rather than
- depend on one particular interpretation of the `gnulib' license text,
- and I was appalled at the idea of adding 64-bit quantities 16 bits at
- a time. :-)
-
- Note that a few of the primitives are necessarily written in assembly,
- or make heavy use of asm() inlines (which amounts to the same thing),
- and are therefore not portable. The majority of the code is either
- fully portable or includes `nearly portable' C implementations (such
- as the C version of bcopy/memcpy/memmove).
-
- >In theory, I could get around this by writing my own linker libraries; in
- >practice, I'm not wizard enough for that to be a realistic alternative.
-
- So, get the net.2 and/or 4.4BSD distributions. :-)
- --
- In-Real-Life: Chris Torek, Lawrence Berkeley Lab CSE/EE (+1 510 486 5427)
- Berkeley, CA Domain: torek@ee.lbl.gov
-