Re: mozilla-devel port on current
On Fri, Dec 06, 2002 at 10:26:24PM -0500, Alexander Kabaev [EMAIL PROTECTED] wrote: On Fri, 6 Dec 2002 23:26:52 +0200 Vallo Kallaste [EMAIL PROTECTED] wrote: I've built it yesterday together with a lots of other stuff. Using other -march values than i686 is unofficially claimed to be unsupported (kan@freebsd). As others I'll bet the -march=p4 is causing problems, i686 works for me. The only thing I claimed was that using -march=p4 and other options which imply SSE2 with gcc 3.2.x is asking for trouble. Judging by the number of bugfixes that were committed to GCC mainline for SSE2 related problems, I am not that far off base. AFAIK, none of those bugfixes made it back to GCC 3.2.x branch. Ok, this is a good claim and making it public will be even better. The p4 optimisation is known to be problematic and users should be notified before they'll shoot the foot off. IMO it will be a good idea to add some logic into appropriate global makefile, to notify users. It's long time problem now and comes up periodically in the lists. -- Vallo Kallaste [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: mozilla-devel port on current
On Sat, Dec 07, 2002 at 07:24:02PM +0200, Vallo Kallaste wrote: On Fri, Dec 06, 2002 at 10:26:24PM -0500, Alexander Kabaev [EMAIL PROTECTED] wrote: On Fri, 6 Dec 2002 23:26:52 +0200 Vallo Kallaste [EMAIL PROTECTED] wrote: I've built it yesterday together with a lots of other stuff. Using other -march values than i686 is unofficially claimed to be unsupported (kan@freebsd). As others I'll bet the -march=p4 is causing problems, i686 works for me. The only thing I claimed was that using -march=p4 and other options which imply SSE2 with gcc 3.2.x is asking for trouble. Judging by the number of bugfixes that were committed to GCC mainline for SSE2 related problems, I am not that far off base. AFAIK, none of those bugfixes made it back to GCC 3.2.x branch. Ok, this is a good claim and making it public will be even better. The p4 optimisation is known to be problematic and users should be notified before they'll shoot the foot off. IMO it will be a good idea to add some logic into appropriate global makefile, to notify users. It's long time problem now and comes up periodically in the lists. Perhaps it should be mentioned in the relnotes/errata. Kris msg48336/pgp0.pgp Description: PGP signature
Re: mozilla-devel port on current
mozilla-devel port fails to build on current. I would imagine this is already known, but I haven't seen any posts on the mailing list. ign -Wno-long-long -O -pipe -march=pentium4 -pipe -DNDEBUG -DTRIMMED I'd bet that -march=pentium4 is your problem. Try with -march=pentiumpro (the default). Marc Premature optimization is the root of all evil. -- Donald E. Knuth To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: mozilla-devel port on current
On Fri, 6 Dec 2002, Marc Recht wrote: mozilla-devel port fails to build on current. I would imagine this is already known, but I haven't seen any posts on the mailing list. ign -Wno-long-long -O -pipe -march=pentium4 -pipe -DNDEBUG -DTRIMMED I'd bet that -march=pentium4 is your problem. Try with -march=pentiumpro (the default). Marc Premature optimization is the root of all evil. -- Donald E. Knuth A very apt signature, in this case Premature enabling of optimization in a devel gcc is the root of all evil. :) -Nate To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: mozilla-devel port on current
On Fri, Dec 06, 2002 at 10:19:23AM -0800, James Satterfield [EMAIL PROTECTED] wrote: mozilla-devel port fails to build on current. I would imagine this is already known, but I haven't seen any posts on the mailing list. I've built it yesterday together with a lots of other stuff. Using other -march values than i686 is unofficially claimed to be unsupported (kan@freebsd). As others I'll bet the -march=p4 is causing problems, i686 works for me. -- Vallo Kallaste [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: mozilla-devel port on current
On Fri, Dec 06, 2002 at 07:29:55PM +0100, Marc Recht wrote: mozilla-devel port fails to build on current. I would imagine this is already known, but I haven't seen any posts on the mailing list. ign -Wno-long-long -O -pipe -march=pentium4 -pipe -DNDEBUG -DTRIMMED I'd bet that -march=pentium4 is your problem. Try with -march=pentiumpro (the default). -march=pentiumpro is not the default, -mcpu=pentiumpro is. Anyway, I doubt this is the cause of the problem. It's more likely to be a stale dependency. Kris msg48279/pgp0.pgp Description: PGP signature
Re: mozilla-devel port on current
-march=pentiumpro is not the default, -mcpu=pentiumpro is. Anyway, I Nope, -march=pentiumpro is added, if the CPUTYPE is i686 (the default). At least if I had read examples/etc/make.conf and bsd.cpu.mk correctly. doubt this is the cause of the problem. It's more likely to be a stale dependency. Hmm, IMO it's the optimizer: jsdtoa.c:1874: this is the insn: (insn 1797 1796 1798 (set (reg:CCNO 17 flags) [...] Marc Premature optimization is the root of all evil. -- Donald E. Knuth To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: mozilla-devel port on current
On Sat, Dec 07, 2002 at 12:09:04AM +0100, Marc Recht wrote: -march=pentiumpro is not the default, -mcpu=pentiumpro is. Anyway, I Nope, -march=pentiumpro is added, if the CPUTYPE is i686 (the default). At least if I had read examples/etc/make.conf and bsd.cpu.mk correctly. The CPUTYPE is not i686 by default. Kris msg48283/pgp0.pgp Description: PGP signature
Re: mozilla-devel port on current
Marc Recht wrote: -march=pentiumpro is not the default, -mcpu=pentiumpro is. Anyway, I Nope, -march=pentiumpro is added, if the CPUTYPE is i686 (the default). At least if I had read examples/etc/make.conf and bsd.cpu.mk correctly. doubt this is the cause of the problem. It's more likely to be a stale dependency. Hmm, IMO it's the optimizer: jsdtoa.c:1874: this is the insn: (insn 1797 1796 1798 (set (reg:CCNO 17 flags) [...] Marc Premature optimization is the root of all evil. -- Donald E. Knuth Marc was correct. Everything built fine with CPUTYPE=i686 and no -O. Still having trouble with lcms tho. James. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: mozilla-devel port on current
The CPUTYPE is not i686 by default. Stupid me... Of course, you're right. Marc Premature optimization is the root of all evil. -- Donald E. Knuth To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: mozilla-devel port on current
There is no default CPUTYPE. If there is no CPUTYPE, and you are on i386, -mcpu=pentiumpro gets added (top of bsd.cpu.mk). Oops! Yes, you're right. I think I got confused by the #CPUTYPE?=i686 in the example make.conf. Marc Premature optimization is the root of all evil. -- Donald E. Knuth To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: mozilla-devel port on current
In the last episode (Dec 07), Marc Recht said: -march=pentiumpro is not the default, -mcpu=pentiumpro is. Anyway, I Nope, -march=pentiumpro is added, if the CPUTYPE is i686 (the default). At least if I had read examples/etc/make.conf and bsd.cpu.mk correctly. There is no default CPUTYPE. If there is no CPUTYPE, and you are on i386, -mcpu=pentiumpro gets added (top of bsd.cpu.mk). -- Dan Nelson [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: mozilla-devel port on current
On Fri, 6 Dec 2002 23:26:52 +0200 Vallo Kallaste [EMAIL PROTECTED] wrote: I've built it yesterday together with a lots of other stuff. Using other -march values than i686 is unofficially claimed to be unsupported (kan@freebsd). As others I'll bet the -march=p4 is causing problems, i686 works for me. The only thing I claimed was that using -march=p4 and other options which imply SSE2 with gcc 3.2.x is asking for trouble. Judging by the number of bugfixes that were committed to GCC mainline for SSE2 related problems, I am not that far off base. AFAIK, none of those bugfixes made it back to GCC 3.2.x branch. -- Alexander Kabaev To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
