Date: Tue, 1 Dec 1998 12:45:41 +1030 From: Greg Lehey <grog@lemis.com> To: Jerry Hicks <jhicks@glenatl.glenayre.com> Cc: Asif Haswarey <Asif.Haswarey@Eng.Sun.COM>, freebsd-sparc@FreeBSD.ORG, bright@hotjobs.com, paipai@tin.it Subject: Re: [Ultra] Compiler, again Message-ID: <19981201124541.S455@freebie.lemis.com> In-Reply-To: <199811301830.NAA17927@jhicks.glenatl.glenayre.com>; from Jerry Hicks on Mon, Nov 30, 1998 at 01:30:11PM -0500 References: <19981126110842.D67961@freebie.lemis.com> <199811301830.NAA17927@jhicks.glenatl.glenayre.com>
next in thread | previous in thread | raw e-mail | index | archive | help
[Format autorecovered at freebie.lemis.com] On Monday, 30 November 1998 at 13:30:11 -0500, Jerry Hicks wrote: >> [Format autorecovered at freebie.lemis.com] >> >> On Wednesday, 25 November 1998 at 14:11:42 -0500, Jerry Hicks wrote: >>> Chuck Youse - Icon CMT Network Operations Center wrote: >>>> Use GNU make. >>> >>> Definitely good advice! >> >> I prefer GNU make to BSD make, but it's still not good advice. >> FreeBSD uses BSD make, and if we start requiring two different makes >> to build the product, we're going to have both procedural trouble and >> lots of resistance from the core team. > > Hi Greg, > > Right, but for a first pass just trying to bootstrap GCC it *is* > good advice. Hmm, was that the context? I can't remember. I thought that getting gcc running on a Sparc was as simple as running configure. > The port to BSD make must happen, but until the regression tests can be run successfully we don't have a good starting point (do we?) > > Trying to build a cross-compiler from x86 FreeBSD to sparc64-unknown-elf64 it seems to me that both the FSF and Cygnus EGCS are broken. I'm currently trying to build from Solaris to target sparc64-unknown-elf64 to see if it is just a cross architecture problem. > > IMHO, we need to get this fixed first and will necessarily need to work with other GCC developers to do this. I wholeheartedly agree that the port to BSD make needs to be done... just not right now. > > I also agree that gcc28 is preferable over egcs, although we might need to keep an eye out for useful patches from egcs too. It would be a *lot* easier to read your mail if you made your lines about the normal length they are in books, like this: > The port to BSD make must happen, but until the regression tests can > be run successfully we don't have a good starting point (do we?) Sure, the i386 Makefiles. Once we've built the initial port, the next thing should be to squeeze it into the FreeBSD framework. The longer we go a different way, the more difficult it'll be to rejoin the main tree. > Trying to build a cross-compiler from x86 FreeBSD to > sparc64-unknown-elf64 it seems to me that both the FSF and Cygnus > EGCS are broken. I'm currently trying to build from Solaris to > target sparc64-unknown-elf64 to see if it is just a cross > architecture problem. I'd guess it's a configuration problem. > IMHO, we need to get this fixed first and will necessarily need to > work with other GCC developers to do this. I wholeheartedly agree > that the port to BSD make needs to be done... just not right now. I still think it should be done as quickly as possible. > I also agree that gcc28 is preferable over egcs, although we might > need to keep an eye out for useful patches from egcs too. I don't know the differences and the problems enough to judge this. If it's true that future development effort is going into egcs, I suppose we don't have much option than to follow, sooner or later. Given that this is a new port with little ballast, it seems to me that the sooner, the better. Greg -- See complete headers for address, home page and phone numbers finger grog@lemis.com for PGP public key To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-sparc" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19981201124541.S455>