Date: Sun, 29 Apr 2001 16:11:52 -0700 (PDT) From: David Wolfskill <david@catwhisker.org> To: current@FreeBSD.ORG Subject: Re: dirpref: a new form of rm -rf * (was: World is broken...) Message-ID: <200104292311.f3TNBqp26845@bunrab.catwhisker.org> In-Reply-To: <200104292023.f3TKNqK39533@vashon.polstra.com>
next in thread | previous in thread | raw e-mail | index | archive | help
>Date: Sun, 29 Apr 2001 13:23:52 -0700 (PDT) >From: John Polstra <jdp@polstra.com> >In article <200104291911.f3TJBw801041@mass.dis.org>, >Mike Smith <msmith@FreeBSD.ORG> wrote: >> > #define isalnum(c) (isalpha(c) || isdigit(c)) Actually, following the lead of the definition of isalpha(), the version I used had the last term of the disjunction wrapped in parens. I suppose there may be some pathological cases where that could be significant. :-} But inserting that definition of isalnum() into /usr/src/lib/libstand/stand.h allowed the build of today's -CURRENT to succeed for me. >> Just FWIW, you can't use <ctype.h> in the loader; it's too incestuously >> entangled with the rest of libc. I suspect that Daniel forgot to update >> libstand. Thanks for the insight as to why that is.... >Thanks. I wish it had just stayed broken. Running this morning's >kernel has immediately trashed my entire root filesystem. Given the fore-warning you provided, I was a little more cautious than usual: after mergemaster completed, I first booted single-user, did a manual "fsck -p". No weirdnesses found. (More on config stuff below.) Then I did a (manual) "mount -a"; again, nothing strange. So I rebooted, allowing the system to come up multi-user as usual. No problems observed. >I updated the kernel and modules and rebooted. Double-checked >UPDATING, which assured me that the "old fsck and new kernel >interactions appear to have been fixed." Started a make buildworld. >It hung almost immediately, in the "rm -rf /usr/obj/local0/src/i386" >step of "Rebuilding the temporary build tree". No response to the >keyboard, no response to pings. It wouldn't escape into DDB, and >there were no messages of any kind. Blecch. :-( >Upon rebooting, fsck complained about many unexpected soft-updates >inconsistencies and other problems. By the time it had finished >having its way with me, I was left with nothing but "/lost+found". Urggh. Sympathy. :-( I haven't tried a "make buildworld" yet -- I'd normally do that after updating my sources, and after the problem with isalnum(), I had re-CVSupped: CVSup started from cvsup14.freebsd.org at Sat Apr 28 03:47:00 PDT 2001 CVSup ended from cvsup14.freebsd.org at Sat Apr 28 03:54:18 PDT 2001 CVSup started from cvsup14.freebsd.org at Sun Apr 29 03:47:00 PDT 2001 CVSup ended from cvsup14.freebsd.org at Sun Apr 29 03:53:14 PDT 2001 CVSup started from cvsup14.freebsd.org at Sun Apr 29 10:00:22 PDT 2001 CVSup ended from cvsup14.freebsd.org at Sun Apr 29 10:06:31 PDT 2001 and did the "cvs update" in the /usr/src I use for -CURRENT. >It's a single-CPU system, SCSI disks, nothing special. Obviously I >can't give you the kernel config file or dmesg output at this point. On my side, I'm using my laptop -- single-CPU, single IDE drive; everything mounted with soft updates turned on. It's set up to boot any of 3 (FreeBSD) environments (so an FS that's root for one would be mounted elsewhere for others, which is one of the reasons I have soft updates on everywhere). Hope your system is recovered soon, david -- David H. Wolfskill david@catwhisker.org As a computing professional, I believe it would be unethical for me to advise, recommend, or support the use (save possibly for personal amusement) of any product that is or depends on any Microsoft product. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200104292311.f3TNBqp26845>