Date: Mon, 31 Dec 2007 13:25:48 +0100 From: "Julian H. Stacey" <jhs@berklix.org> To: freebsd-stable@freebsd.org Cc: mvoorhis@cs.wpi.edu, Nikos Ntarmos <ntarmos@ceid.upatras.gr>, mvoorhis@eressea.wpi.edu, Ivan Voras <ivoras@freebsd.org> Subject: Re: 7.0BETA4 ../cc_tools/insn-attrtab.c compiles forever Message-ID: <200712311226.lBVCPm7I099990@fire.js.berklix.net> In-Reply-To: <20071220113331.GA83912@ace.netcins.ceid.upatras.gr> References: <200712192036.lBJKaWYp008250@fire.js.berklix.net>
next in thread | previous in thread | raw e-mail | index | archive | help
Sorry for delay (travelling), Answers to 1 private mail & 2 list posted replies further below, first my original posting for reference: >> Anyone else noticed on 7.0BETA4 (& hence Ive cc'd re@) >> >> This runs forever: >> >> ===> cc_int (all) >> Warning: Object directory not changed from original /usr/src/gnu/usr.bin/cc/cc_int >> cc -O2 -fno-strict-aliasing -march=i586 -DIN_GCC -DHAVE_CONFIG_H -DPREFIX=\"/usr\" -I/usr/src/gnu/usr.bin/cc/cc_int/../cc_tools -I/usr/src/gnu/usr.bin/cc/cc_int/../cc_tools -I/usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcc -I/usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcc/config -I/usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcclibs/include -I/usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcclibs/libcpp/include -I/usr/src/gnu/usr.bin/cc/cc_int/../../../../contrib/gcclibs/libdecnumber -c ../cc_tools/insn-attrtab.c >> >> With or without -pipe & with or without /usr/obj >> All the rest of /usr/src compiles OK though. >> >> FreeBSD lapn.js.berklix.net 7.0-BETA4 FreeBSD 7.0-BETA4 #0: Sun Dec 16 09:18:21 CET 2007 jhs@lapn.js.berklix.net:/usr/src/sys/i386/compile/GENERIC i386 ---------- From: Michael Voorhis Wed, 19 Dec 2007 21:11:25 -0500 (Thu 03:11 CET): > What sort of machine do you have? I seem to recall that compile-line > taking an age, but not forever. It takes a while to finish on my > Thinkpad t42, which is a > > Pentium(R) M processor 1.70GHz (1694.51-MHz 686-class CPU) This an old laptop with small slow disc: http://www.berklix.com/~jhs/hardware/laptops/dell_latitude_xpi_p133st/ CPU: Pentium/P54C (133.64-MHz 586-class CPU) avail memory = 43507712 (41 MB) /boot/loader.conf: hw.ata.ata_dma=0 hw.ata.wc=0 swapinfo Device 1K-blocks Used Avail Capacity /dev/ad0s1b 102400 10156 92244 10% /dev/md0 100000 10144 89856 10% ---------- From: Ivan Voras <ivoras@freebsd.org> Thu, 20 Dec 2007 11:47:57 +0100 > I did see some weird behaviour while compiling gcc, though I don't > remember if it's the same file (looks like it). They usually disappeared > after removing CPUTYPE and -O2 from compiler flags. My /etc/make.conf just has CFLAGS += -march=i586 I assume -O2 is coming out of /usr/src/ gcc itself. ---------- From: Nikos Ntarmos <ntarmos@ceid.upatras.gr> Thu, 20 Dec 2007 13:33:31 +0200 > I'd bet you're seeing some serious thrashing there. I've come across at > least one other user having problems with that particular part of > buildworld and -O2/-Os optimization levels. Seems like gcc4 is quite the > memory hog so I don't know whether something can be done, other than > either adding more RAM, trying to tweak swappiness, or taking a break > while this thing compiles itself. Yes, on return from travels, I powered up, left it running for day[s?] looked again & finally it had finished. Now trying from BETA4 to 7-Stable. Seems your right about gcc4 & memory hog. (-pipe by default makes it even worse, subject of another thread). Thanks all. -- Julian Stacey. Munich Computer Consultant, BSD Unix C Linux. http://berklix.com Ihr Rauch = mein allergischer Kopfschmerz. Dump cigs 4 snuff.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200712311226.lBVCPm7I099990>