From owner-fam@oss.sgi.com Mon Jan 7 14:21:13 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g07MLDQ27631 for fam-outgoing; Mon, 7 Jan 2002 14:21:13 -0800 Received: from moutng0.schlund.de (moutng0.kundenserver.de [212.227.126.170]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g07ML4g27625 for ; Mon, 7 Jan 2002 14:21:04 -0800 Received: from [195.20.224.219] (helo=mrvdom03.kundenserver.de) by moutng0.schlund.de with esmtp (Exim 3.22 #2) id 16NhCM-000811-00; Mon, 07 Jan 2002 22:20:50 +0100 Received: from [217.5.91.210] (helo=shrek.lisa.de) by mrvdom03.kundenserver.de with esmtp (Exim 2.12 #2) id 16NhCL-0000Vc-00; Mon, 7 Jan 2002 22:20:50 +0100 Received: from there (elfe.lisa.de [172.16.24.108]) by shrek.lisa.de (Postfix/Linux) with SMTP id 37EA613D8; Mon, 7 Jan 2002 22:20:49 +0100 (CET) Content-Type: text/plain; charset="iso-8859-15" From: Hans-Peter Jansen To: wardle@adacel.com.au Subject: [fam] imon with current linux kernels Date: Mon, 7 Jan 2002 22:20:47 +0100 X-Mailer: KMail [version 1.3.2] Organization: TreeWater Society Berlin Cc: fam@oss.sgi.com MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Message-Id: <20020107212049.37EA613D8@shrek.lisa.de> Sender: owner-fam@oss.sgi.com Precedence: bulk Hi Michael et al., just tried to patch a current kernel (2.4.18-pre1:ia32), but doesn't looked very promising. May someone dived into this before? I'm running home rolled KDE 2.2.2 diskless on linux nfs-v3 servers, based on SuSE 7.3. They distribute fam-2.6.4 with imon disabled. I noticed, there are issues with fam and successive logins via kdm (killing fam). I will try 2.6.7 now, and would like to know, if it is worth the hassle of going through this: (and other known issues of fam/imon in my nfs world) hp@elfe:/usr/src/linux> zcat ~/Downloads/lib/imon-0.0.2-2.4.0.gz | drypatch -p1 patching file Documentation/Configure.help Hunk #1 succeeded at 15551 with fuzz 2 (offset 3439 lines). patching file Documentation/imon.txt patching file fs/Config.in Hunk #1 succeeded at 161 (offset 36 lines). patching file fs/Makefile Hunk #1 FAILED at 22. 1 out of 1 hunk FAILED -- saving rejects to file fs/Makefile.rej patching file fs/attr.c Hunk #1 succeeded at 9 with fuzz 1. Hunk #2 succeeded at 147 (offset 13 lines). patching file fs/exec.c Hunk #1 succeeded at 37 with fuzz 1 (offset 1 line). Hunk #2 succeeded at 860 (offset 18 lines). Hunk #3 succeeded at 890 (offset 18 lines). Hunk #4 succeeded at 911 (offset 18 lines). patching file fs/filesystems.c Hunk #1 FAILED at 25. Hunk #2 FAILED at 40. 2 out of 2 hunks FAILED -- saving rejects to file fs/filesystems.c.rej patching file fs/imon/Config.in patching file fs/imon/Makefile patching file fs/imon/README patching file fs/imon/imon_dynamic.c patching file fs/imon/imon_static.c patching file fs/namei.c Hunk #1 FAILED at 21. Hunk #2 succeeded at 949 (offset 33 lines). Hunk #3 succeeded at 1271 (offset 34 lines). Hunk #4 succeeded at 1300 (offset 34 lines). Hunk #5 succeeded at 1391 (offset 35 lines). Hunk #6 succeeded at 1464 (offset 35 lines). Hunk #7 succeeded at 1537 (offset 35 lines). Hunk #8 succeeded at 1612 (offset 35 lines). Hunk #9 succeeded at 1776 (offset 39 lines). Hunk #10 succeeded at 1822 (offset 39 lines). 1 out of 10 hunks FAILED -- saving rejects to file fs/namei.c.rej patching file fs/read_write.c Hunk #2 succeeded at 186 (offset 29 lines). Hunk #3 succeeded at 335 (offset 29 lines). patching file include/linux/fs.h Hunk #1 succeeded at 435 (offset 42 lines). patching file include/linux/imon.h patching file include/linux/sched.h Hunk #1 succeeded at 389 (offset 13 lines). Hunk #2 succeeded at 497 (offset 26 lines). patching file kernel/exit.c Hunk #1 succeeded at 15 (offset 3 lines). Hunk #2 succeeded at 440 (offset 11 lines). patching file kernel/fork.c Hunk #1 succeeded at 603 (offset 12 lines). patching file kernel/ksyms.c Hunk #1 succeeded at 39 (offset -2 lines). Hunk #2 succeeded at 270 (offset 16 lines). Thanks for your patience. Cheers, Hans-Peter -- NT, now approaching 23x6 availability -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Mon Jan 7 21:00:02 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g08502Q20407 for fam-outgoing; Mon, 7 Jan 2002 21:00:02 -0800 Received: from nexus.adacel.com (shelob.adacel.com.au [203.36.26.146] (may be forged)) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g084xug20394 for ; Mon, 7 Jan 2002 20:59:56 -0800 Received: (qmail 13602 invoked from network); 8 Jan 2002 03:51:45 -0000 Received: from unknown (HELO adacel.com) (192.168.75.20) by nexus.adacel.com with SMTP; 8 Jan 2002 03:51:45 -0000 Message-ID: <3C3A6F31.8080202@adacel.com> Date: Tue, 08 Jan 2002 15:01:53 +1100 From: Michael Wardle Organization: Adacel Technologies User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:0.9.4) Gecko/20011019 Netscape6/6.2 X-Accept-Language: en-au, en-us, en-gb, en, eo, de-8j÷` MIME-Version: 1.0 To: Hans-Peter Jansen CC: fam@oss.sgi.com Subject: Re: [fam] imon with current linux kernels References: <20020107212049.37EA613D8@shrek.lisa.de> Content-Type: text/plain; charset=ISO-8859-15; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-fam@oss.sgi.com Precedence: bulk Hi Hans-Peter. Hans-Peter Jansen wrote: > just tried to patch a current kernel (2.4.18-pre1:ia32), but doesn't looked > very promising. May someone dived into this before? There have been some patches contributed for more recent 2.4 kernel sources. They're currently attached to some bugs on http://oss.sgi.com/bugzilla, but I'll find some place for them under the downloads area. You should try the one with the latest kernel version. It should be pointed out that more Linux users and vendors are moving towards using DNotify rather than IMon, as it's available in the standard kernel sources. Alex Larsson has made a patch to add DNotify support. You can get it from the Red Hat FAM SRPMs, or wait until FAM 2.7.0, as I plan to integrate it. > I'm running home rolled KDE 2.2.2 diskless on linux nfs-v3 servers, based on > SuSE 7.3. They distribute fam-2.6.4 with imon disabled. If FAM comes with IMon disabled, are they using another kernel file monitor or are they polling? > I noticed, there are issues with fam and successive logins via kdm (killing > fam). I will try 2.6.7 now, and would like to know, if it is worth the hassle > of going through this: (and other known issues of fam/imon in my nfs world) There was a problem in FAM where it was trying to chown from one unprivileged user to another, and this was causing problems with multiple users using FAM. This might be related to the problem you're seeing, and was fixed in FAM 2.6.6. Perhaps try 2.6.7 or post the error messages you're seeing. Thanks for your message. Let us know how you go. -- MICHAEL WARDLE | WORK +61-2-6024-2699 SGI Desktop & Admin Software | MOBILE +61-415-439-838 Adacel Technologies Limited | WEB http://www.adacel.com/ -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Mon Jan 7 21:39:29 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g085dTL21203 for fam-outgoing; Mon, 7 Jan 2002 21:39:29 -0800 Received: from nexus.adacel.com (shelob.adacel.com.au [203.36.26.146] (may be forged)) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g085dOg21200 for ; Mon, 7 Jan 2002 21:39:25 -0800 Received: (qmail 15055 invoked from network); 8 Jan 2002 04:31:15 -0000 Received: from unknown (HELO adacel.com) (192.168.75.20) by nexus.adacel.com with SMTP; 8 Jan 2002 04:31:15 -0000 Message-ID: <3C3A7873.7000202@adacel.com> Date: Tue, 08 Jan 2002 15:41:23 +1100 From: Michael Wardle Organization: Adacel Technologies User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:0.9.4) Gecko/20011019 Netscape6/6.2 X-Accept-Language: en-au, en-us, en-gb, en, eo, de-8j÷` MIME-Version: 1.0 To: Michael Wardle CC: Hans-Peter Jansen , fam@oss.sgi.com Subject: Re: [fam] imon with current linux kernels References: <20020107212049.37EA613D8@shrek.lisa.de> <3C3A6F31.8080202@adacel.com> Content-Type: text/plain; charset=ISO-8859-15; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-fam@oss.sgi.com Precedence: bulk Michael Wardle wrote: >> just tried to patch a current kernel (2.4.18-pre1:ia32), but doesn't >> looked very promising. May someone dived into this before? > > There have been some patches contributed for more recent 2.4 kernel > sources. They're currently attached to some bugs on > http://oss.sgi.com/bugzilla, but I'll find some place for them under the > downloads area. You should try the one with the latest kernel version. I've placed them in: ftp://oss.sgi.com/projects/fam/download/imon You'll notice the latest version is for 2.4.12. You should have more luck with this one, but there will probably still be changes between 2.4.12 and 2.4.18 that make this patch fail (or at least use some fuzz). Enjoy! -- MICHAEL WARDLE | WORK +61-2-6024-2699 SGI Desktop & Admin Software | MOBILE +61-415-439-838 Adacel Technologies Limited | WEB http://www.adacel.com/ -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Tue Jan 8 18:38:02 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g092c2H12130 for fam-outgoing; Tue, 8 Jan 2002 18:38:02 -0800 Received: from a.mx.canon.com.au (bergeron.research.canon.com.au [203.12.172.124]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g092bRg12126 for ; Tue, 8 Jan 2002 18:37:28 -0800 Received: from bellmann.research.canon.com.au (kwanon.research.canon.com.au [203.12.172.254]) by a.mx.canon.com.au (Postfix) with ESMTP id BB2E233937 for ; Wed, 9 Jan 2002 01:37:22 +0000 (GMT) Received: from bax.research.canon.com.au (bax.research.canon.com.au [10.2.1.246]) by bellmann.research.canon.com.au (Postfix) with ESMTP id ABCD18B3B for ; Wed, 9 Jan 2002 12:16:46 +1100 (EST) Received: from research.canon.com.au (localhost.localdomain [127.0.0.1]) by bax.research.canon.com.au (Postfix) with ESMTP id 898AC34479 for ; Wed, 9 Jan 2002 12:37:21 +1100 (EST) Message-ID: <3C3B9ED0.5000005@research.canon.com.au> Date: Wed, 09 Jan 2002 12:37:20 +1100 From: "kennys@research.canon.com.au" User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.7) Gecko/20011221 X-Accept-Language: en-us MIME-Version: 1.0 To: fam@oss.sgi.com Subject: [fam] client keeps on dying Content-Type: multipart/mixed; boundary="------------050603050602070402000700" Sender: owner-fam@oss.sgi.com Precedence: bulk This is a multi-part message in MIME format. --------------050603050602070402000700 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit I am fiddling around with the test program that comes with the download, but at some point in time, the client stops for no obvious reasons. So, I need to understand this little piece of code. for ( ; FAMPending(&fc); ) { if( FAMNextEvent(&fc, &fe) < 0 ) { printf("Gahh, FAMNextEvent() returned < 0 ! \n"); exit(1); } processDirEvents(&fe); fflush(stdout); } okay what i need is to know 1. why the purpose of the if statement, 2. why it comes true as in what happens.. 3. is there anyway to make sure that if statement never comes true any help is much appreciated attached is the complete source . --------------050603050602070402000700 Content-Type: text/plain; name="test.c++" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="test.c++" #include #include #include #include #include #include #include //#include #include "fam.h" #include //#include "Boolean.h" /* FILE test.c - simple fam test program Usage: test [-r] [-f ] [-d ] ex. test -f /etc/hosts -d /tmp -r includes request IDs in output (if you use this, put it first) -f [file] monitors the given file -d [dir] monitors the given directory */ void processDirEvents(FAMEvent* fe); bool suspend = false; bool cont = false; bool intr = false; bool usr1 = false; bool usr2 = false; bool showReqIDs = false; //int lockfd = -1; //FILE *fout; void handle_stop(int) { printf("Suspended!\n"); suspend = true; } void handle_cont(int) { printf("Resumed!\n"); signal(SIGCONT, handle_cont); cont = true; } void handle_int(int) { printf("Interupted!\n"); signal(SIGINT, handle_int); intr = true; } void handle_usr1(int) { printf("Got USR1!\n"); signal(SIGUSR1, handle_usr1); usr1 = true; } void handle_usr2(int) { printf("Got USR2!\n"); signal(SIGUSR2, handle_usr2); usr2 = true; } struct TestRequest { FAMRequest fr; char path[PATH_MAX]; char userData[PATH_MAX]; bool isDir; }; void sendRequests(FAMConnection &fc, TestRequest tr[], int trlen) { for (int ii = 0; ii < trlen; ++ii) { if (tr[ii].isDir) { printf("FAMMonitorDirectory(\"%s\")\n", tr[ii].path); if (FAMMonitorDirectory(&fc, tr[ii].path, &(tr[ii].fr), tr[ii].userData) < 0) { fprintf(stderr, "FAMMonitorDirectory error\n"); exit(1); } if (showReqIDs) printf("req %d: ", tr[ii].fr.reqnum); printf("FAMMonitorDirectory(\"%s\")\n", tr[ii].path); } else { printf("FAMMonitorFile(\"%s\")\n", tr[ii].path); if (FAMMonitorFile(&fc, tr[ii].path, &(tr[ii].fr), tr[ii].userData) < 0) { fprintf(stderr, "FAMMonitorFile error\n"); exit(1); } if (showReqIDs) printf("req %d: ", tr[ii].fr.reqnum); printf("FAMMonitorFile(\"%s\")\n", tr[ii].path); } } } void main(int argc, char **argv) { /* Create fam test */ FAMConnection fc; TestRequest tr[100]; int requests = 0; FAMEvent fe; fd_set readfds; fd_set readfdscpy; setbuf(stdout, NULL); if (argc < 2) { printf("usage: %s [-r] [-f ] [-d ]\n", argv[0]); exit(1); } int arg; extern char *optarg; extern int optind; if(FAMOpen(&fc) < 0) { fprintf(stderr, "FAMOpen failed!\n"); exit(1); } FD_ZERO(&readfds); FD_SET(fc.fd, &readfds); while ((arg = getopt(argc, argv, "f:d:r")) != -1) { switch (arg) { case 'd': case 'f': snprintf(tr[requests].userData, PATH_MAX, "%s %s: ", (arg == 'd') ? "DIR blablabla" : "FILE", optarg); snprintf(tr[requests].path, PATH_MAX, "%s", optarg); tr[requests].isDir = (arg == 'd'); ++requests; // just don't use more than 100, OK? break; case 'r': showReqIDs = true; break; } } signal(SIGTSTP, handle_stop); signal(SIGCONT, handle_cont); signal(SIGINT, handle_int); signal(SIGUSR1, handle_usr1); signal(SIGUSR2, handle_usr2); sendRequests(fc, tr, requests); while (1) { if (suspend) { for (int ii = 0; ii < requests; ii++) { FAMSuspendMonitor(&fc, &(tr[ii].fr)); printf("Suspended monitoring of request %i\n", ii); } fflush(stdout); suspend = false; kill(getpid(),SIGTSTP); signal(SIGTSTP, handle_stop); } if (cont) { for (int ii = 0; ii < requests; ii++) { FAMResumeMonitor(&fc, &(tr[ii].fr)); printf("Resumed monitoring of request %i\n", ii); } fflush(stdout); cont = false; } if (intr) { // Cancel monitoring of every other request. This makes // sure fam can handle the case where the connection goes // down with requests outstanding. for (int ii = 0; ii < requests ; ii++) { if (ii % 2 == 0) continue; FAMCancelMonitor(&fc, &(tr[ii].fr)); printf("Canceled monitoring of request %i\n", ii); } FAMClose(&fc); exit(0); } if (usr1) { // Cancel all requests, close the connection, and reopen it. // This makes sure long-lived clients can connect, monitor, and // disconnect repeatedly without leaking. usr1 = false; int sleeptime = 1; for (int ii = 0; ii < requests ; ii++) { FAMCancelMonitor(&fc, &(tr[ii].fr)); printf("Canceled monitoring of request %i\n", ii); } FAMClose(&fc); printf("Closed connection, sleeping %d...\n", sleeptime); sleep(sleeptime); // Now reconnect and resend the requests. if(FAMOpen(&fc) < 0) { fprintf(stderr, "FAMOpen failed!\n"); exit(1); } FD_ZERO(&readfds); FD_SET(fc.fd, &readfds); sendRequests(fc, tr, requests); } if (usr2) { // Clean things up like a well-behaved client and exit. for (int ii = 0; ii < requests ; ii++) { FAMCancelMonitor(&fc, &(tr[ii].fr)); printf("Canceled monitoring of request %i\n", ii); } FAMClose(&fc); printf("Closed connection\n"); exit(0); } readfdscpy = readfds; if (select(FD_SETSIZE, &readfdscpy, NULL, NULL, NULL) < 0) { if (errno == EINTR) { continue; } break; } for (; FAMPending(&fc); ) { if (FAMNextEvent(&fc, &fe) < 0) { printf("Gahh, FAMNextEvent() returned < 0!\n"); exit(1); } processDirEvents(&fe); fflush(stdout); } } } void processDirEvents(FAMEvent* fe) { if (fe->userdata) printf("%s ", (char *)fe->userdata); if (showReqIDs) printf("req %d: ", fe->fr.reqnum); switch (fe->code) { case FAMExists: printf("%s Exists\n", fe->filename); break; case FAMEndExist: printf("%s EndExist\n", fe->filename); break; case FAMChanged: printf("%s Changed\n", fe->filename); break; case FAMDeleted: printf("%s was deleted\n", fe->filename); break; case FAMStartExecuting: printf("%s started executing\n", fe->filename); break; case FAMStopExecuting: printf("%s stopped executing\n", fe->filename); break; case FAMCreated: printf("'%s' was created\n", fe->filename); break; case FAMMoved: printf("%s was moved\n", fe->filename); break; default: printf("(unknown event %d on %s)\n", fe->code, fe->filename); } } --------------050603050602070402000700-- -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Tue Jan 8 20:11:02 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g094B2M16374 for fam-outgoing; Tue, 8 Jan 2002 20:11:02 -0800 Received: from nexus.adacel.com (shelob.adacel.com.au [203.36.26.146] (may be forged)) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g094Asg16368 for ; Tue, 8 Jan 2002 20:10:54 -0800 Received: (qmail 14246 invoked from network); 9 Jan 2002 03:02:39 -0000 Received: from unknown (HELO adacel.com) (192.168.75.20) by nexus.adacel.com with SMTP; 9 Jan 2002 03:02:39 -0000 Message-ID: <3C3BB532.4040603@adacel.com> Date: Wed, 09 Jan 2002 14:12:50 +1100 From: Michael Wardle Organization: Adacel Technologies User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:0.9.4) Gecko/20011019 Netscape6/6.2 X-Accept-Language: en-au, en-us, en-gb, en, eo, de-8j÷` MIME-Version: 1.0 To: "kennys@research.canon.com.au" CC: fam@oss.sgi.com Subject: Re: [fam] client keeps on dying References: <3C3B9ED0.5000005@research.canon.com.au> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-fam@oss.sgi.com Precedence: bulk kennys@research.canon.com.au wrote: > I am fiddling around with the test program that comes with the download, > but at some point in time, the client stops for no obvious reasons. > > So, I need to understand this little piece of code. > > for ( ; FAMPending(&fc); ) { > if( FAMNextEvent(&fc, &fe) < 0 ) { > printf("Gahh, FAMNextEvent() returned < 0 ! \n"); > exit(1); > } > processDirEvents(&fe); > fflush(stdout); > } > > > okay what i need is to know > 1. why the purpose of the if statement, > 2. why it comes true as in what happens.. > 3. is there anyway to make sure that if statement never comes true Okay, I've just had a quick look at the source. FAMPending and FAMNextEvent both come from libfam/fam.c++. These are the steps that we go thru: 1. Do FAMPending() to see if there's an event waiting on our connection. 2. If there is, do FAMNextEvent() to try to process it. 3. If the event couldn't be processed, print an error message and exit. So the answers to your questions: 1. To see whether the incoming event can be processed. 2. Perhaps the FAM daemon died. 3. Stop the FAM daemon from dying. ;-) These are some of the things I would think about: * architecture * operating system * compiler * FAM version * if there's anything weird about your system (such as non-standard partitioning, nearly full disks, ...) * what kernel monitor you're using (imon/dnotify/other/none) * how FAM was executed (shell/inetd/xinetd/other) * how long FAM has been running for * how long it takes for "test" to stop * if you have any other FAM client programs running * what command line you ran "test" with * what file or directory you are monitoring * what user you ran "test" as * any error messages (screen/system log/other) * whether it's reproducible The more information about these points you can post here, the easier it will be to help me find the problem. Regards. -- MICHAEL WARDLE | WORK +61-2-6024-2699 SGI Desktop & Admin Software | MOBILE +61-415-439-838 Adacel Technologies Limited | WEB http://www.adacel.com/ -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Tue Jan 15 10:01:22 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0FI1MP27710 for fam-outgoing; Tue, 15 Jan 2002 10:01:22 -0800 Received: from devserv.devel.redhat.com (nat-pool-meridian.redhat.com [12.107.208.200]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0FI11P27698 for ; Tue, 15 Jan 2002 10:01:02 -0800 Received: from localhost (alexl@localhost) by devserv.devel.redhat.com (8.11.0/8.11.0) with ESMTP id g0FH0xC05899 for ; Tue, 15 Jan 2002 12:00:59 -0500 X-Authentication-Warning: devserv.devel.redhat.com: alexl owned process doing -bs Date: Tue, 15 Jan 2002 12:00:59 -0500 (EST) From: Alex Larsson X-X-Sender: To: Subject: [fam] [Patch] Fam leaving .fam files in /tmp Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-fam@oss.sgi.com Precedence: bulk I've gotten bugreports (and noticed myself) that fam leaves lots of .famXXX type files in /tmp. These should have been removed by the LocalClient destructor, but perhaps they didn't due to fam abruptly quiting or something. Anyway, here is a patch that should fix this problem. We don't really need the name of the socket after the client has connected. I applied it to the current Red Hat RPM. / Alex --- fam-2.6.7/fam/Listener.c++.cleanup Mon Dec 17 20:03:37 2001 +++ fam-2.6.7/fam/Listener.c++ Tue Jan 15 11:44:45 2002 @@ -287,10 +287,18 @@ // Keep the scheduler from helpfully cleaning this up. Scheduler::remove_onetime_task(cleanup_negotiation, nc); + // Remove the temp file + uid_t preveuid = geteuid(); + if (preveuid) seteuid(0); + seteuid(nc->uid); + unlink(nc->sun.sun_path); + if (nc->uid) seteuid(0); + seteuid(preveuid); + Log::debug("client fd %d is local/trusted (socket %s, uid %d).", client_fd, nc->sun.sun_path, nc->uid); Cred cred(nc->uid, client_fd); - new LocalClient(client_fd, &(nc->sun), cred); + new LocalClient(client_fd, cred); // We don't need a reference to this object. The constructor // takes care of registering it with the Scheduler. --- fam-2.6.7/fam/LocalClient.c++.cleanup Mon Dec 17 20:03:37 2001 +++ fam-2.6.7/fam/LocalClient.c++ Tue Jan 15 11:43:04 2002 @@ -29,17 +29,12 @@ #include "Log.h" #include "Cred.h" -LocalClient::LocalClient(int fd, const struct sockaddr_un *addr, Cred &cred) +LocalClient::LocalClient(int fd, Cred &cred) : TCP_Client(LOCALHOST(), fd, cred) { assert(cred.is_valid()); - sun.sun_family = AF_UNIX; - strncpy(sun.sun_path, addr->sun_path, sizeof(sun.sun_path)); - sun.sun_path[sizeof(sun.sun_path) - 1] = '\0'; } LocalClient::~LocalClient() { - if (geteuid() != cred.uid()) cred.become_user(); - unlink(sun.sun_path); } --- fam-2.6.7/fam/LocalClient.h.cleanup Mon Dec 17 20:03:37 2001 +++ fam-2.6.7/fam/LocalClient.h Tue Jan 15 11:43:04 2002 @@ -33,13 +33,8 @@ public: - LocalClient(int fd, const struct sockaddr_un *addr, Cred &cred); + LocalClient(int fd, Cred &cred); ~LocalClient(); - -private: - - struct sockaddr_un sun; - }; #endif /* !RemoteClient_included */ -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Wed Jan 16 13:44:19 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0GLiJM23893 for fam-outgoing; Wed, 16 Jan 2002 13:44:19 -0800 Received: from smtp017.mail.yahoo.com (smtp017.mail.yahoo.com [216.136.174.114]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0GLiCP23889 for ; Wed, 16 Jan 2002 13:44:12 -0800 Received: from 24-28-199-157.gf.cox.rr.com (HELO viper.ascellatech.com) (24.28.199.157) by smtp.mail.vip.sc5.yahoo.com with SMTP; 16 Jan 2002 20:44:09 -0000 Subject: [fam] fam and nautilus From: Amith Varghese To: fam@oss.sgi.com Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Evolution/1.0 (Preview Release) Date: 16 Jan 2002 15:44:08 -0500 Message-Id: <1011213849.15824.25.camel@viper> Mime-Version: 1.0 Sender: owner-fam@oss.sgi.com Precedence: bulk Hi... I'm trying to compile nautilus with fam support but i am running into some problems. I can run and install fam alright, but when I run the nautilus configure script it says it can't find the FAM library so it won't compile support in for it. However when I look at the config.log I see the following configure:6661: checking for FAMOpen in -lfam configure:6680: gcc -o conftest -g -O2 conftest.c -lfam 1>&5 /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `__rtti_user' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `__rethrow' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `__builtin_vec_new' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `__builtin_vec_delete' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `__builtin_delete' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `__builtin_new' collect2: ld returned 1 exit status configure: failed program was: #line 6669 "configure" #include "confdefs.h" /* Override any gcc2 internal prototype to avoid an error. */ /* We use char because int might match the return type of a gcc2 builtin and then its argument prototype would still apply. */ char FAMOpen(); int main() { FAMOpen() ; return 0; } It looks like there are some problems when compiling the test program so that's why it failed. I don't understand why there are undefined symbols when I compiled FAM itself. I'm running redhat with a 2.4.13 kernel and glibc 2.2.4. If anyone has any suggestions I would appreciate it. Thanks Amith _________________________________________________________ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Wed Jan 16 13:51:52 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0GLpqP24146 for fam-outgoing; Wed, 16 Jan 2002 13:51:52 -0800 Received: from rcsnet.net (APuteaux-102-2-1-4.abo.wanadoo.fr [193.251.91.4]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0GLphP24133 for ; Wed, 16 Jan 2002 13:51:44 -0800 Received: from xfiles.rcsnet.net (IDENT:rcoscali@localhost.localdomain [127.0.0.1]) by rcsnet.net (8.12.2.Beta3/8.12.1) with ESMTP id g0GKlOAZ025135; Wed, 16 Jan 2002 21:47:24 +0100 Received: (from rcoscali@localhost) by xfiles.rcsnet.net (8.12.2.Beta3/8.12.1/Submit) id g0GKlOhB025134; Wed, 16 Jan 2002 21:47:24 +0100 X-Authentication-Warning: xfiles.rcsnet.net: rcoscali set sender to Remi@Cohen-Scali.com using -f Subject: Re: [fam] fam and nautilus From: =?ISO-8859-1?Q?R=E9mi?= Cohen-Scali To: Amith Varghese Cc: FAM mailling list In-Reply-To: <1011213849.15824.25.camel@viper> References: <1011213849.15824.25.camel@viper> Content-Type: text/plain; charset=ISO-8859-1 X-Mailer: Evolution/1.0.1 Date: 16 Jan 2002 21:47:23 +0100 Message-Id: <1011214044.23273.0.camel@xfiles> Mime-Version: 1.0 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by oss.sgi.com id g0GLpjP24137 Sender: owner-fam@oss.sgi.com Precedence: bulk Fam is a c++ module and need g++ to link le mer 16-01-2002 à 21:44, Amith Varghese a écrit : > Hi... I'm trying to compile nautilus with fam support but i am running > into some problems. I can run and install fam alright, but when I run > the nautilus configure script it says it can't find the FAM library so > it won't compile support in for it. However when I look at the > config.log I see the following > > > configure:6661: checking for FAMOpen in -lfam > configure:6680: gcc -o conftest -g -O2 conftest.c -lfam 1>&5 > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__rtti_user' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__rethrow' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__builtin_vec_new' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__builtin_vec_delete' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__builtin_delete' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__builtin_new' > collect2: ld returned 1 exit status > configure: failed program was: > #line 6669 "configure" > #include "confdefs.h" > /* Override any gcc2 internal prototype to avoid an error. */ > /* We use char because int might match the return type of a gcc2 > builtin and then its argument prototype would still apply. */ > char FAMOpen(); > > int main() { > FAMOpen() > ; return 0; } > > It looks like there are some problems when compiling the test program so > that's why it failed. I don't understand why there are undefined > symbols when I compiled FAM itself. I'm running redhat with a 2.4.13 > kernel and glibc 2.2.4. If anyone has any suggestions I would > appreciate it. > > Thanks > Amith > > > > > > > _________________________________________________________ > Do You Yahoo!? > Get your free @yahoo.com address at http://mail.yahoo.com > > > -- > Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ > To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com -- -o) Remi Cohen-Scali /\\ _\_v ---- -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Wed Jan 16 14:03:02 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0GM32324633 for fam-outgoing; Wed, 16 Jan 2002 14:03:02 -0800 Received: from smtp018.mail.yahoo.com (smtp018.mail.yahoo.com [216.136.174.115]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0GM2pP24628 for ; Wed, 16 Jan 2002 14:02:51 -0800 Received: from 24-28-199-157.gf.cox.rr.com (HELO viper.ascellatech.com) (24.28.199.157) by smtp.mail.vip.sc5.yahoo.com with SMTP; 16 Jan 2002 21:02:44 -0000 Subject: Re: [fam] fam and nautilus From: Amith Varghese To: =?ISO-8859-1?Q?R=E9mi?= Cohen-Scali Cc: FAM mailling list In-Reply-To: <1011214044.23273.0.camel@xfiles> References: <1011213849.15824.25.camel@viper> <1011214044.23273.0.camel@xfiles> Content-Type: text/plain; charset=ISO-8859-1 X-Mailer: Evolution/1.0 (Preview Release) Date: 16 Jan 2002 16:02:43 -0500 Message-Id: <1011214964.15824.27.camel@viper> Mime-Version: 1.0 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by oss.sgi.com id g0GM2pP24629 Sender: owner-fam@oss.sgi.com Precedence: bulk I don't think because it says gcc instead of g++ it is causing the problem. To prove the point I pasted the test program into a file and did the following: g++ -o conftest -g -O2 famtest.c -lfam and i got: /tmp/ccVa4c0d.o: In function `main': /tmp/famtest.c:8: undefined reference to `FAMOpen() ' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `__rtti_user' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `__rethrow' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `__builtin_vec_new' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `__builtin_vec_delete' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `__builtin_delete' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `__builtin_new' collect2: ld returned 1 exit status If you notice i still get the same problem. Any other ideas? Thanks Amith On Wed, 2002-01-16 at 15:47, Rémi Cohen-Scali wrote: > Fam is a c++ module and need g++ to link > > le mer 16-01-2002 à 21:44, Amith Varghese a écrit : > > Hi... I'm trying to compile nautilus with fam support but i am running > > into some problems. I can run and install fam alright, but when I run > > the nautilus configure script it says it can't find the FAM library so > > it won't compile support in for it. However when I look at the > > config.log I see the following > > > > > > configure:6661: checking for FAMOpen in -lfam > > configure:6680: gcc -o conftest -g -O2 conftest.c -lfam 1>&5 > > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > > undefined reference to `__rtti_user' > > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > > undefined reference to `__rethrow' > > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > > undefined reference to `__builtin_vec_new' > > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > > undefined reference to `__builtin_vec_delete' > > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > > undefined reference to `__builtin_delete' > > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > > undefined reference to `__builtin_new' > > collect2: ld returned 1 exit status > > configure: failed program was: > > #line 6669 "configure" > > #include "confdefs.h" > > /* Override any gcc2 internal prototype to avoid an error. */ > > /* We use char because int might match the return type of a gcc2 > > builtin and then its argument prototype would still apply. */ > > char FAMOpen(); > > > > int main() { > > FAMOpen() > > ; return 0; } > > > > It looks like there are some problems when compiling the test program so > > that's why it failed. I don't understand why there are undefined > > symbols when I compiled FAM itself. I'm running redhat with a 2.4.13 > > kernel and glibc 2.2.4. If anyone has any suggestions I would > > appreciate it. > > > > Thanks > > Amith > > > > > > > > > > > > > > _________________________________________________________ > > Do You Yahoo!? > > Get your free @yahoo.com address at http://mail.yahoo.com > > > > > > -- > > Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ > > To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com > -- > -o) > Remi Cohen-Scali /\\ > _\_v > ---- _________________________________________________________ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Wed Jan 16 14:12:16 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0GMCG525194 for fam-outgoing; Wed, 16 Jan 2002 14:12:16 -0800 Received: from ns1.epinetworx.com (ns1.epinetworx.com [213.41.61.153]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0GMC3P25191 for ; Wed, 16 Jan 2002 14:12:04 -0800 Received: from [172.16.3.24] (helo=pierre) by ns1.epinetworx.com with esmtp (Exim 3.33 #1 ) id 16QxP2-000E7k-00 for ; Wed, 16 Jan 2002 22:15:24 +0100 Date: Wed, 16 Jan 2002 22:11:59 +0100 From: Pierre To: FAM mailling list Subject: Re: [fam] fam and nautilus Message-ID: <20020116221159.G6484@pierre.epinetworx.com> Mime-Version: 1.0 Content-Type: text/plain; format=flowed; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit X-Mailer: Balsa 1.2.1 Lines: 87 Sender: owner-fam@oss.sgi.com Precedence: bulk On 2002.01.16 21:44 Amith Varghese wrote: > Hi... I'm trying to compile nautilus with fam support but i am running > into some problems. I can run and install fam alright, but when I run > the nautilus configure script it says it can't find the FAM library so > it won't compile support in for it. However when I look at the > config.log I see the following > > > configure:6661: checking for FAMOpen in -lfam > configure:6680: gcc -o conftest -g -O2 conftest.c -lfam 1>&5 > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__rtti_user' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__rethrow' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__builtin_vec_new' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__builtin_vec_delete' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__builtin_delete' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__builtin_new' > collect2: ld returned 1 exit status > configure: failed program was: > #line 6669 "configure" > #include "confdefs.h" > /* Override any gcc2 internal prototype to avoid an error. */ > /* We use char because int might match the return type of a gcc2 > builtin and then its argument prototype would still apply. */ > char FAMOpen(); > > int main() { > FAMOpen() > ; return 0; } > > It looks like there are some problems when compiling the test program > so > that's why it failed. I don't understand why there are undefined > symbols when I compiled FAM itself. I'm running redhat with a 2.4.13 > kernel and glibc 2.2.4. If anyone has any suggestions I would > appreciate it. > > Thanks > Amith > I also have problem configuring nautilus with fam... Here is what i got : configure:6661: checking for FAMOpen in -lfam configure:6680: gcc -o conftest -g -O2 conftest.c -lfam 1>&5 /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `operator new[](unsigned)' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `operator delete(void*)' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `__gxx_personality_v0' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `std::ios_base::Init::Init()' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `vtable for __cxxabiv1::__class_type_info' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `operator delete[](void*)' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `std::ios_base::Init::~Init()' /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: undefined reference to `operator new(unsigned)' collect2: ld returned 1 exit status configure: failed program was: #line 6669 "configure" #include "confdefs.h" /* Override any gcc2 internal prototype to avoid an error. */ /* We use char because int might match the return type of a gcc2 builtin and then its argument prototype would still apply. */ char FAMOpen(); int main() { FAMOpen() ; return 0; } Pierre. -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Wed Jan 16 14:13:58 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0GMDw825250 for fam-outgoing; Wed, 16 Jan 2002 14:13:58 -0800 Received: from rcsnet.net (APuteaux-102-2-1-4.abo.wanadoo.fr [193.251.91.4]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0GMDjP25246 for ; Wed, 16 Jan 2002 14:13:45 -0800 Received: from xfiles.rcsnet.net (IDENT:rcoscali@localhost.localdomain [127.0.0.1]) by rcsnet.net (8.12.2.Beta3/8.12.1) with ESMTP id g0GL9bAZ025572; Wed, 16 Jan 2002 22:09:37 +0100 Received: (from rcoscali@localhost) by xfiles.rcsnet.net (8.12.2.Beta3/8.12.1/Submit) id g0GL9bC7025570; Wed, 16 Jan 2002 22:09:37 +0100 X-Authentication-Warning: xfiles.rcsnet.net: rcoscali set sender to Remi@Cohen-Scali.com using -f Subject: Re: [fam] fam and nautilus From: =?ISO-8859-1?Q?R=E9mi?= Cohen-Scali To: Amith Varghese Cc: FAM mailling list In-Reply-To: <1011214964.15824.27.camel@viper> References: <1011213849.15824.25.camel@viper> <1011214044.23273.0.camel@xfiles> <1011214964.15824.27.camel@viper> Content-Type: text/plain; charset=ISO-8859-1 X-Mailer: Evolution/1.0.1 Date: 16 Jan 2002 22:09:37 +0100 Message-Id: <1011215377.23273.4.camel@xfiles> Mime-Version: 1.0 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by oss.sgi.com id g0GMDkP25247 Sender: owner-fam@oss.sgi.com Precedence: bulk I test it by myself with gcc and it build without any problem. perhaps your library is mis-builded ... ? le mer 16-01-2002 à 22:02, Amith Varghese a écrit : > I don't think because it says gcc instead of g++ it is causing the > problem. To prove the point I pasted the test program into a file and > did the following: > > g++ -o conftest -g -O2 famtest.c -lfam > > and i got: > > > /tmp/ccVa4c0d.o: In function `main': > /tmp/famtest.c:8: undefined reference to `FAMOpen() ' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__rtti_user' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__rethrow' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__builtin_vec_new' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__builtin_vec_delete' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__builtin_delete' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__builtin_new' > collect2: ld returned 1 exit status > > If you notice i still get the same problem. Any other ideas? > > Thanks > Amith > > On Wed, 2002-01-16 at 15:47, Rémi Cohen-Scali wrote: > > Fam is a c++ module and need g++ to link > > > > le mer 16-01-2002 à 21:44, Amith Varghese a écrit : > > > Hi... I'm trying to compile nautilus with fam support but i am running > > > into some problems. I can run and install fam alright, but when I run > > > the nautilus configure script it says it can't find the FAM library so > > > it won't compile support in for it. However when I look at the > > > config.log I see the following > > > > > > > > > configure:6661: checking for FAMOpen in -lfam > > > configure:6680: gcc -o conftest -g -O2 conftest.c -lfam 1>&5 > > > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > > > undefined reference to `__rtti_user' > > > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > > > undefined reference to `__rethrow' > > > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > > > undefined reference to `__builtin_vec_new' > > > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > > > undefined reference to `__builtin_vec_delete' > > > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > > > undefined reference to `__builtin_delete' > > > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > > > undefined reference to `__builtin_new' > > > collect2: ld returned 1 exit status > > > configure: failed program was: > > > #line 6669 "configure" > > > #include "confdefs.h" > > > /* Override any gcc2 internal prototype to avoid an error. */ > > > /* We use char because int might match the return type of a gcc2 > > > builtin and then its argument prototype would still apply. */ > > > char FAMOpen(); > > > > > > int main() { > > > FAMOpen() > > > ; return 0; } > > > > > > It looks like there are some problems when compiling the test program so > > > that's why it failed. I don't understand why there are undefined > > > symbols when I compiled FAM itself. I'm running redhat with a 2.4.13 > > > kernel and glibc 2.2.4. If anyone has any suggestions I would > > > appreciate it. > > > > > > Thanks > > > Amith > > > > > > > > > > > > > > > > > > > > > _________________________________________________________ > > > Do You Yahoo!? > > > Get your free @yahoo.com address at http://mail.yahoo.com > > > > > > > > > -- > > > Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ > > > To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com > > -- > > -o) > > Remi Cohen-Scali /\\ > > _\_v > > ---- > > > > _________________________________________________________ > Do You Yahoo!? > Get your free @yahoo.com address at http://mail.yahoo.com > -- -o) Remi Cohen-Scali /\\ _\_v ---- -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Wed Jan 16 15:36:11 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0GNaB526603 for fam-outgoing; Wed, 16 Jan 2002 15:36:11 -0800 Received: from mailrelay1.lrz-muenchen.de (mailrelay1.lrz-muenchen.de [129.187.254.101]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0GNa6P26593 for ; Wed, 16 Jan 2002 15:36:06 -0800 Received: from [212.224.54.173] by mailout.lrz-muenchen.de with ESMTP for fam@oss.sgi.com; Wed, 16 Jan 2002 23:35:58 +0100 Message-Id: <3C45FFD0.79B4E732@informatik.tu-muenchen.de> Date: Wed, 16 Jan 2002 23:33:52 +0100 From: Christian Kreibich X-Mailer: Mozilla 4.77 [en] (X11; U; Linux 2.4.3 i686) X-Accept-Language: en MIME-Version: 1.0 To: FAM mailling list Subject: Re: [fam] fam and nautilus References: <1011213849.15824.25.camel@viper> <1011214044.23273.0.camel@xfiles> <1011214964.15824.27.camel@viper> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-fam@oss.sgi.com Precedence: bulk Amith Varghese wrote: > > I don't think because it says gcc instead of g++ it is causing the > problem. To prove the point I pasted the test program into a file and > did the following: > > g++ -o conftest -g -O2 famtest.c -lfam > > and i got: > > /tmp/ccVa4c0d.o: In function `main': > /tmp/famtest.c:8: undefined reference to `FAMOpen() ' > /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.0.3/../../../libfam.so: > undefined reference to `__rtti_user' Try adding -lstdc++. I've received reports about issues with gcc3 versions and compiling fam, as libstdc++ is apparently under certain circumstances not automatically linked in. Try an older gcc if available. Christian. -- ________________________________________________________________________ http://www.whoop.org -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Fri Jan 18 13:12:06 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0ILC6e09477 for fam-outgoing; Fri, 18 Jan 2002 13:12:06 -0800 Received: from monkeyiq.dnsalias.org (CPE-203-45-214-174.qld.bigpond.net.au [203.45.214.174]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0ILC1P09470 for ; Fri, 18 Jan 2002 13:12:01 -0800 Received: by monkeyiq.dnsalias.org id g0IKER806110 ; Sat, 19 Jan 2002 06:14:27 +1000 Date: Sat, 19 Jan 2002 06:14:27 +1000 Message-Id: <200201182014.g0IKER806110@monkeyiq.dnsalias.org> To: E Develop cc: fam-list Subject: [fam] Re: [E-devel] imon installation with fam? From: monkeyiq MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-fam@oss.sgi.com Precedence: bulk Christian Kreibich writes: > > I still haven't tried the new versions of fam that use dnotify, so you > don't need imon any more. If they work well, it's definitely the way to > go. I have the fam rpms from rawhide, fam-devel-2.6.7-4 fam-2.6.7-4 and there is still an outstanding bug (from what I can tell) using dnotify where it doesn't send a FAMEndExist for reading an empty dir. so apps waiting for this FAMEndExist event seem to "hang" but are just idle waiting for an event that will never come. IIRC a FAMEndExist is fired when using imon/polling to read an empty dir. > > Christian. > -- > ________________________________________________________________________ > http://www.whoop.org > > > _______________________________________________ > enlightenment-devel mailing list > enlightenment-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel > -- ----------------------------------------------------- http://witme.sourceforge.net/libferris.web/index.html -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Fri Jan 18 13:28:12 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0ILSCX10436 for fam-outgoing; Fri, 18 Jan 2002 13:28:12 -0800 Received: from devserv.devel.redhat.com (nat-pool-meridian.redhat.com [12.107.208.200]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0ILS8P10424 for ; Fri, 18 Jan 2002 13:28:08 -0800 Received: from localhost (alexl@localhost) by devserv.devel.redhat.com (8.11.0/8.11.0) with ESMTP id g0IKS3112551; Fri, 18 Jan 2002 15:28:03 -0500 X-Authentication-Warning: devserv.devel.redhat.com: alexl owned process doing -bs Date: Fri, 18 Jan 2002 15:28:03 -0500 (EST) From: Alex Larsson X-X-Sender: To: monkeyiq cc: E Develop , fam-list Subject: Re: [fam] Re: [E-devel] imon installation with fam? In-Reply-To: <200201182014.g0IKER806110@monkeyiq.dnsalias.org> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-fam@oss.sgi.com Precedence: bulk On Sat, 19 Jan 2002, monkeyiq wrote: > Christian Kreibich writes: > > > > > I still haven't tried the new versions of fam that use dnotify, so you > > don't need imon any more. If they work well, it's definitely the way to > > go. > > I have the fam rpms from rawhide, > fam-devel-2.6.7-4 > fam-2.6.7-4 > and there is still an outstanding bug (from what I can tell) using > dnotify where it doesn't send a FAMEndExist for reading an empty dir. > > so apps waiting for this FAMEndExist event seem to "hang" but are just > idle waiting for an event that will never come. IIRC a FAMEndExist > is fired when using imon/polling to read an empty dir. Are you sure? I don't see that: bash-2.05$ mkdir /tmp/foo bash-2.05$ ~/bin/fam_test -d /tmp/foo FAMMonitorDirectory("/tmp/foo") FAMMonitorDirectory("/tmp/foo") DIR /tmp/foo: /tmp/foo Exists DIR /tmp/foo: /tmp/foo EndExist / Alex -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Fri Jan 18 13:48:27 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0ILmRE11091 for fam-outgoing; Fri, 18 Jan 2002 13:48:27 -0800 Received: from monkeyiq.dnsalias.org (CPE-203-45-214-174.qld.bigpond.net.au [203.45.214.174]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0ILmJP11083 for ; Fri, 18 Jan 2002 13:48:20 -0800 Received: by monkeyiq.dnsalias.org id g0IKopK27249 ; Sat, 19 Jan 2002 06:50:51 +1000 Date: Sat, 19 Jan 2002 06:50:51 +1000 Message-Id: <200201182050.g0IKopK27249@monkeyiq.dnsalias.org> To: Alex Larsson Cc: monkeyiq , E Develop , fam-list Subject: Re: [fam] Re: [E-devel] imon installation with fam? From: monkeyiq MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-fam@oss.sgi.com Precedence: bulk Alex Larsson writes: > On Sat, 19 Jan 2002, monkeyiq wrote: > > > Christian Kreibich writes: > > > > > > > > I still haven't tried the new versions of fam that use dnotify, so you > > > don't need imon any more. If they work well, it's definitely the way to > > > go. > > > > I have the fam rpms from rawhide, > > fam-devel-2.6.7-4 > > fam-2.6.7-4 > > and there is still an outstanding bug (from what I can tell) using > > dnotify where it doesn't send a FAMEndExist for reading an empty dir. > > > > so apps waiting for this FAMEndExist event seem to "hang" but are just > > idle waiting for an event that will never come. IIRC a FAMEndExist > > is fired when using imon/polling to read an empty dir. > > Are you sure? I don't see that: > bash-2.05$ mkdir /tmp/foo > bash-2.05$ ~/bin/fam_test -d /tmp/foo > FAMMonitorDirectory("/tmp/foo") > FAMMonitorDirectory("/tmp/foo") > DIR /tmp/foo: /tmp/foo Exists > DIR /tmp/foo: /tmp/foo EndExist hmm, /me hides in shame. well, the fam_test works ok just as you have shown above. So its all good, except for my previous email, which should me moved to /dev/null. Thanks, I'll check out what could be making that event fail at a event loop level... but that is none of fam's fault. Sorry about that. > > / Alex > > -- ----------------------------------------------------- http://witme.sourceforge.net/libferris.web/index.html -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Fri Jan 18 15:25:58 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0INPwm13581 for fam-outgoing; Fri, 18 Jan 2002 15:25:58 -0800 Received: from andromeda ([65.200.107.1]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0INPtP13577 for ; Fri, 18 Jan 2002 15:25:55 -0800 Received: from nwaddell by andromeda with local (Exim 3.33 #1 (Debian)) id 16RhSA-0006OG-00 for ; Fri, 18 Jan 2002 16:25:42 -0600 Content-Type: text/plain; charset="iso-8859-1" From: Nathan Waddell Reply-To: ravenhall@kracked.com To: fam@oss.sgi.com Subject: [fam] 2.5.1 kernel patch?? Date: Fri, 18 Jan 2002 16:25:37 -0600 X-Mailer: KMail [version 1.3.2] MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Message-Id: Sender: owner-fam@oss.sgi.com Precedence: bulk -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello. Is the fam/imon project still being worked on?? I am a developer for E17 (http://www.enlightenment.org) and we are using fam as part of the basis for EFSD, the E filesystem daemon. I have recently upgraded to kernel 2.5.1, and I see that there has been no new news on the sgi project website since 10/12/2000. Is this project dead?? Please reply to me, I hate sending mail and never knowing if it was received. Ravenhall -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.6 (GNU/Linux) Comment: For info see http://www.gnupg.org iD8DBQE8SKDk/B8ikrtb+kwRAv6WAJ91rqpzy1eGQPChJtE5CazdC7hRHgCggd/L OMGdej1eMxSx/23ZTjA0yFw= =2L4G -----END PGP SIGNATURE----- -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Sun Jan 20 15:32:12 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0KNWCk32247 for fam-outgoing; Sun, 20 Jan 2002 15:32:12 -0800 Received: from nexus.adacel.com (shelob.adacel.com.au [203.36.26.146] (may be forged)) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0KNW5P32244 for ; Sun, 20 Jan 2002 15:32:06 -0800 Received: (qmail 5492 invoked from network); 20 Jan 2002 22:22:33 -0000 Received: from unknown (HELO adacel.com) (192.168.75.20) by nexus.adacel.com with SMTP; 20 Jan 2002 22:22:33 -0000 Message-ID: <3C4B45EF.3040809@adacel.com> Date: Mon, 21 Jan 2002 09:34:23 +1100 From: Michael Wardle Organization: Adacel Technologies User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:0.9.4) Gecko/20011019 Netscape6/6.2 X-Accept-Language: en-au, en-us, en-gb, en, eo, de-8j÷` MIME-Version: 1.0 To: ravenhall@kracked.com CC: fam@oss.sgi.com Subject: Re: [fam] 2.5.1 kernel patch?? References: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-fam@oss.sgi.com Precedence: bulk Hi Nathan. Nathan Waddell wrote: > Is the fam/imon project still being worked on?? Yes. FAM is in active development (FAM 2.6.7 was released a month ago). IMon is still being used, but most persons seem to be using DNotify instead, as it is available in standard Linux kernels. I'm working on adding DNotify support to our FAM tree, but until then, you can get a patch from Alex Larsson at . There is no IMon patch for 2.5.1 as yet, and I'm probably not likely to have one ready soon. If anyone would like to make one, feel free! I'm sorry that the FAM pages haven't been updated for a while now. I'm supposed to maintain them, but I've been quite busy. I'm looking forward to a stable release of E17! :-) Let me know if there's anything I can do to help. Regards -- MICHAEL WARDLE | WORK +61-2-6024-2699 SGI Desktop & Admin Software | MOBILE +61-415-439-838 Adacel Technologies Limited | WEB http://www.adacel.com/ -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Sun Jan 20 17:26:33 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0L1QXv01566 for fam-outgoing; Sun, 20 Jan 2002 17:26:33 -0800 Received: from devserv.devel.redhat.com (nat-pool-meridian.redhat.com [12.107.208.200]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0L1QUP01563 for ; Sun, 20 Jan 2002 17:26:30 -0800 Received: from localhost (alexl@localhost) by devserv.devel.redhat.com (8.11.0/8.11.0) with ESMTP id g0L0PdD06747; Sun, 20 Jan 2002 19:25:39 -0500 X-Authentication-Warning: devserv.devel.redhat.com: alexl owned process doing -bs Date: Sun, 20 Jan 2002 19:25:39 -0500 (EST) From: Alex Larsson X-X-Sender: To: Michael Wardle cc: , Subject: Re: [fam] 2.5.1 kernel patch?? In-Reply-To: <3C4B45EF.3040809@adacel.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-fam@oss.sgi.com Precedence: bulk On Mon, 21 Jan 2002, Michael Wardle wrote: > Hi Nathan. > > Nathan Waddell wrote: > > > Is the fam/imon project still being worked on?? > > > Yes. FAM is in active development (FAM 2.6.7 was released a month ago). > IMon is still being used, but most persons seem to be using DNotify > instead, as it is available in standard Linux kernels. I'm working on > adding DNotify support to our FAM tree, but until then, you can get a > patch from Alex Larsson at > . Newer patches are in the Red Hat fam package in Rawhide. / Alex -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Mon Jan 21 04:33:35 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0LCXZM21045 for fam-outgoing; Mon, 21 Jan 2002 04:33:35 -0800 Received: from mailout09.sul.t-online.com (mailout09.sul.t-online.com [194.25.134.84]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0LCXWP21041 for ; Mon, 21 Jan 2002 04:33:32 -0800 Received: from fwd03.sul.t-online.de by mailout09.sul.t-online.com with smtp id 16Scha-0003TM-02; Mon, 21 Jan 2002 12:33:26 +0100 Received: from b4 (520080802291-0001@[217.225.70.7]) by fmrl03.sul.t-online.com with esmtp id 16SchU-02CfJ2C; Mon, 21 Jan 2002 12:33:20 +0100 Received: from orbit.b4.mad (orbit.b4.mad [10.129.1.9]) by b4 (Postfix) with ESMTP id 6B283CBDC0 for ; Mon, 21 Jan 2002 12:33:19 +0100 (CET) Received: by orbit.b4.mad (Postfix on SuSE Linux 7.2 (i386), from userid 500) id 59AD3C09; Mon, 21 Jan 2002 12:33:19 +0100 (CET) Date: Mon, 21 Jan 2002 12:33:19 +0100 From: Stephan Duehr To: fam@oss.sgi.com Subject: [fam] imon hash table overflow Message-ID: <20020121123319.A1801@b4mad.net> Reply-To: duehr@b4mad.net Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.3.16i X-Sender: 520080802291-0001@t-dialin.net Sender: owner-fam@oss.sgi.com Precedence: bulk Hi, when using fam and imon on directories with about 30000 files, the kernel says: kernel: /dev/imon: hash table overflow and fam polls the directory (strace on the fam process shows many lstat64 calls). I tried to increase hash table size by modprobe imon imon_hashhisize=96 The hash table overflows remained. I increased the value until the kernel crashed (at about 128). Is there a way to enable the use of hash table for large directories? Or does it make no sense to do that for some reason? Regards Stephan -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com From owner-fam@oss.sgi.com Mon Jan 21 17:18:28 2002 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id g0M1ISN20058 for fam-outgoing; Mon, 21 Jan 2002 17:18:28 -0800 Received: from nexus.adacel.com (shelob.adacel.com.au [203.36.26.146] (may be forged)) by oss.sgi.com (8.11.2/8.11.3) with SMTP id g0M1IOP20052 for ; Mon, 21 Jan 2002 17:18:24 -0800 Received: (qmail 7349 invoked from network); 22 Jan 2002 00:08:43 -0000 Received: from unknown (HELO adacel.com) (192.168.75.20) by nexus.adacel.com with SMTP; 22 Jan 2002 00:08:43 -0000 Message-ID: <3C4CB058.40107@adacel.com> Date: Tue, 22 Jan 2002 11:20:40 +1100 From: Michael Wardle Organization: Adacel Technologies User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:0.9.4) Gecko/20011019 Netscape6/6.2 X-Accept-Language: en-au, en-us, en-gb, en, eo, de-8j÷` MIME-Version: 1.0 To: fam@oss.sgi.com Subject: [fam] updating FAM page content Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-fam@oss.sgi.com Precedence: bulk Hi. Just letting you know that I'm in the process of updating the FAM pages. This wouldn't normally be news, but it's been a long time since they've been updated. Some pages may still be in progress. The old site can be found here: . Your feedback is welcome, but I don't plan to spend too long on the web site at the moment. Regards -- MICHAEL WARDLE | WORK +61-2-6024-2699 SGI Desktop & Admin Software | MOBILE +61-415-439-838 Adacel Technologies Limited | WEB http://www.adacel.com/ -- Source code, list archive, and docs: http://oss.sgi.com/projects/fam/ To unsubscribe: echo unsubscribe fam | mail majordomo@oss.sgi.com