Re: How can i see just -stable logs
On Wed, 6 Dec 2000 12:12:44 -0500 "Ken Menzel" [EMAIL PROTECTED] wrote: Ahh- Ok subscribe to the commit mailing list and just filter what I don't want yes?! Thanks I don't know why I didn't think of that. I guess I was oriented towards getting in a web page or file, but that works. This came up a while back and someone set up some filtered lists. I don't recall who or where but a search in the -stable list archives should bring it up. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-stable" in the body of the message
Re: mysql: a lot of lost connections...
On Wed, 6 Dec 2000, Milan Salajka wrote: Milan, 4.1.1 Release has different thread bugs, at least for me MySQL crashed randomly about once per week. 4.2-STABLE doesn't work, latest release (4.2) doesn't work, previous release (4.1.1) has bugs too... uff, what is the recommended FREEBSD version for MYSQL ? LINUX ? many people use APACHE, PHP MYSQL trio on FreeBSD, but it seems, that RELENG_4 isn't good choice. my 'upgrades': 4.2-stable - 4.2-release - 4.1.1-release when it will end ? ;o))) Milan As has been mentioned on this list, you can use 4.2-STABLE, but you must change the compiler. 4.2-RELEASE has a known mutex bug that was discussed on this list and fixed about two days after the release. Since this was a well-known bug, why on earth did you even bother with 4.2-RELEASE? Tom Uniserve To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-stable" in the body of the message
Re: Mysql segfaults; is the culprit libstdc++, pthread, regex ...?
hi, there!
On Wed, 6 Dec 2000, David O'Brien wrote:
now I can confirm that on latest -stable without "linking shared objects
with libgcc_[r]_pic" changes ACE wrappers tests run correctly without
segfaults. test program ("bad" one) from PR/23252 also does not segfault
anymore.
sorry, no workaround so far except backing out that changes
(patch for testing attached) and have no time to find more correct
solution.
Lets get the patch right for people to test. Remove all the bogus
-kthread stuff. As that was only for Linux threads (or does ACE use
that)?? If so, then it isn't being linked the way Jason Evans says it
should.
ACE can use both linuxthreads or pthreads. we use it with pthreads. Sorry,
have no time to test patch without -kthread stuff (attached) or to find
out what is the deal with threads when we are linking shared libraries
against libgcc_[r_]pic.a.
I think that changes to compiler and binutils should be tested
in -current for much longer period of time before MFC and should be MFC'ed
long before release date.
If you understood what was broken by not doing this commit, you'd
understand why it was done. C++ exceptions (even w/o threads) was rather
broken. W/o this patch we were going to dropped as a GCC 3.0 release
test platform.
AFAIR C++ exceptions without this patch are not broken (I am running
latest -stable without them now). When we switched to gcc crtstuff the
only thing that was broken were C++ multithreaded programs because threads
were not initialized when crtstuff tried to use them to initialize eh
context information.
I think it does not matter why it was done -- we need
someone to find out why linking shared libraries against libgcc_[r_]pic.a
breaks multithreaded programs.
/fjoe
--- gnu/usr.bin/cc/cc_tools/freebsd-native.h.orig Sun Nov 12 03:52:40 2000
+++ gnu/usr.bin/cc/cc_tools/freebsd-native.hWed Dec 6 14:54:29 2000
@@ -1,4 +1,4 @@
-/* $FreeBSD: src/gnu/usr.bin/cc/cc_tools/freebsd-native.h,v 1.5.2.1 2000/11/11
21:52:40 obrien Exp $ */
+/* $FreeBSD: src/gnu/usr.bin/cc/cc_tools/freebsd-native.h,v 1.4 1999/12/22 05:00:27
+obrien Exp $ */
/* FREEBSD_NATIVE is defined when gcc is integrated into the FreeBSD
source tree so it can be configured appropriately without using
@@ -50,9 +50,4 @@
/* Tell gcc to locate libgcc.a for us according to the -m rules. */
#undef LIBGCC_SPEC
#define LIBGCC_SPEC \
- "%{!shared: \
-%{!pthread:libgcc.a%s} \
-%{pthread:libgcc_r.a%s}} \
- %{shared: \
-%{!pthread:libgcc_pic.a%s} \
-%{pthread:libgcc_r_pic.a%s}}"
+ "%{!shared:%{!pthread:libgcc.a%s}%{pthread:libgcc_r.a%s}}"
--- Makefile.inc1.orig Sat Dec 2 03:58:09 2000
+++ Makefile.inc1 Wed Dec 6 15:01:03 2000
@@ -682,7 +682,7 @@
#
# The following dependencies exist between the libraries:
#
-# lib*: csu libgcc_pic
+# lib*: csu
# libatm: libmd
# libcrypt: libmd
# libdialog: libncurses
@@ -690,7 +690,7 @@
# libg++: libm
# libkrb: libcrypt
# libopie: libmd
-# libpam: libcom_err libcrypt libcrypto libkrb libopie libradius \
+# libpam: libcom_err libcrypt libcrypto libgcc_pic libkrb libopie libradius \
#libskey libtacplus libutil libz libssh
# libradius: libmd
# libreadline: libncurses
@@ -703,7 +703,7 @@
#
# gnu/lib: lib/libm lib/libncurses
# kerberosIV/lib kerberos5/lib: lib/libcrypt
-# lib/libpam: secure/lib/libcrypto kerberosIV/lib/libkrb \
+# lib/libpam: secure/lib/libcrypto kerberosIV/lib/libkrb gnu/lib/libgcc \
# secure/lib/libssh lib/libz
# secure/lib: lib/libmd
#
@@ -742,9 +742,8 @@
.endif
libraries:
-.for _lib in ${_csu} gnu/lib/csu gnu/lib/libgcc lib/libmd lib/libcrypt \
-${_secure_lib} ${_kerberosIV_lib} \
-${_kerberos5_lib} lib/libcom_err ${_libm} lib/libncurses \
+.for _lib in ${_csu} gnu/lib/csu lib/libmd lib/libcrypt ${_secure_lib}
+${_kerberosIV_lib} \
+${_kerberos5_lib} gnu/lib/libgcc lib/libcom_err ${_libm} lib/libncurses \
lib/libopie lib/libradius lib/libskey lib/libtacplus lib/libutil \
lib/libz lib gnu/lib \
${_libperl} usr.bin/lex/lib ${_libkeycap}
Re: How can i see just -stable logs
* Steve O'Hara-Smith [EMAIL PROTECTED] [001206 11:49]: On Wed, 6 Dec 2000 12:12:44 -0500 "Ken Menzel" [EMAIL PROTECTED] wrote: Ahh- Ok subscribe to the commit mailing list and just filter what I don't want yes?! Thanks I don't know why I didn't think of that. I guess I was oriented towards getting in a web page or file, but that works. This came up a while back and someone set up some filtered lists. I don't recall who or where but a search in the -stable list archives should bring it up. mail to [EMAIL PROTECTED] subscribe freebsd-stable-4 in the body. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-stable" in the body of the message -- Larry Rosenman http://www.lerctr.org/~ler Phone: +1 972-414-9812 E-Mail: [EMAIL PROTECTED] US Mail: 1905 Steamboat Springs Drive, Garland, TX 75044-6749 To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-stable" in the body of the message
Re: console (XF86 3.3.6) dumps core with des root password
Several weeks ago, Patrick Hausen reported a problem with "xconsole" core-dumping under "wdm" if he used a DES-encrypted root password. I ran into the same problem with "xconsole", but I'm using "xdm" and an MD5-encrypted root password (4.1-RELEASE and 4.2-RELEASE). After fiddling around with the XFree86 code a bit, I managed to narrow the problem down to XFree86's _XGetpwuid() wrapper macro, and I also came up with a workaround. I added the following lines to my /usr/X11R6/lib/X11/xdm/Xsetup_0 startup script, just before the "xconsole" invocation: export USER=root export HOME=/root Actually, I suspect it's probably enough just to set HOME. The relevant portion of "xconsole" is trying to get the home directory path name -- if possible from $HOME, or else from $USER (followed by an OS-specific-wrapped getpwnam() call), or via getuid() and a wrapped getpwuid() call if neither of these environment variables is set. Apparently, there's something wrong with the way _XGetpwuid() is being expanded for the FreeBSD environment. I've reported this problem to the XFree86 people. I also submitted a FreeBSD PR (misc/23168), though this report got shot down on the grounds that X isn't part of the FreeBSD base system. (I did post a followup to my PR, with the above workaround, just in case anyone else with the problem goes looking for help in the GNATS database.) Rich Wales [EMAIL PROTECTED] http://www.webcom.com/richw/ To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-stable" in the body of the message
Re: Dell C600/FreeBSD4.2-rel
In message [EMAIL PROTECTED] Danny Braniss writes: : which is the 'prefered' solution? : changing the configuration : device pcic0 at isa? irq 0 ... : to : device pcic0 at isa? irq 10 ... : (are there cards that behave correctly with irq 0?) : or the : /etc/pccard.conf : anyways, thanks, i now have the ethernet working - now to get the video ... Those are two different things. The first one is the IRQ for management events for the bridge (card insertion, removal, bad batterym etc). The second is the irq for the card itself. Warner To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-stable" in the body of the message
Re: WaveLAN PCI problem
In message [EMAIL PROTECTED] shizuka00 writes: : Should I assume that -current code should work with this PCI adapter? No. I've not been able to get these cards to work with -current yet, nor have I had the time to find why it hangs. : 1. Using GENERIC kernel (cvsup 30 Nov), machine hang after pccardd recognize the :card and load the driver. It also hang if I take the WaveLan card Same thing here, if it doesn't hang while it is waiting for my scsi bus to settle. : 2. Using NEWCARD kernel, machine hang while setting up pccard1 (?). Here're what I :wrote down NEWCARD doesn't work for 16bit cards at the present moment. : One more question, the PCMCIA-PCI adapter has only one slot, is there any reason why :pccbb card detect pccard0 pccard1? Yes. There really are 2 slots on these cards, as far as the hardware is concerned. Your card might have only one of them attached to the board. I have cards that have 1 or 2 attached. It could also be a bug in our slot detection code. My old Libretto detected two slots as well, even though it had only one... until you connected the docing port... Warner To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-stable" in the body of the message
Re: How can i see just -stable logs
On Wed, Dec 06, 2000, a militant turtle forced Ken Menzel to say: Hi Guy's, I know about http://docs.freebsd.org/mail/current/cvs-all.html which allows me to see what is committed, but there is too much! Is Here's a snippet from the FreeBSD-stable Conspectus: -snip- On May 30, 2000, [David Miller] made this proclamation: I've setup freebsd-stable-3 and freebsd-stable-4 majordomo lists at sparks.net. These use procmail to filter the RELENG_[3|4] messages out of cvs-all, so one can easily tell which commits affect them. Anyone could use procmail to filter the list himself, but I thought this was more convenient, especially for those not set up with procmail. To subscribe, send an email to freebsd-stable-[3|4][EMAIL PROTECTED] Digest versions are setup as well. -snip- -- Scott McDermott Unix Dude Election fault (Gore dumped) To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-stable" in the body of the message
Re: make installworld from nfs directory
Hurf Sheldon: We've built oa 4.2-STABLE /usr/src (loaded with cvsup) tree by running "make buildworld", then a "make installworld" on machine "A" (originally a 4.0-RELEASE), then built a new kernel, which worked fine. Now we NFS mount the machine "A":/usr/src to machine "B":/usr/src (also 4.0-RELEASE) and run "make installworld" which dies with the message: "install: /usr/src/include/osreldate.h: No such file or directory" The build logon machine "A" says this file was created on the fly during the "make buildworld" by 'newconf.sh' . Do we have to do a "make buildworld" on every system? You need to mount A:/usr/obj to /usr/obj on B as well. Helge To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-stable" in the body of the message
Re: VMware hanging
Ok (this to the -stable mailing list). Barry and I have figurd out
why VMWare was locking up on him.
We tracked it down to excessive dirty filesystem buffers, created
through this codepath:
Debugger(c02ca083) at Debugger+0x35
panic(c02cd684,ca77cbe8,c018054a,c4f9cdb8,c4f9cdb8) at panic+0x70
bdirty(c4f9cdb8,c4f9cdb8,cb719000,0,c1218800) at bdirty+0x5d
bdwrite(c4f9cdb8,cb719000,0,0,2000) at bdwrite+0x52
cluster_write(c4f9cdb8,676a000,0,0) at cluster_write+0x313
ffs_write(ca77ccf8) at ffs_write+0x48b
vnode_pager_generic_putpages(cb719000,ca77ce08,1,0,ca77cd9c) at
vnode_pager_generic_putpages+0x181
ffs_putpages(ca77cd60) at ffs_putpages+0x1f
vnode_pager_putpages(cb72dae0,ca77ce08,10,0,ca77cd9c) at vnode_pager_putpages+0x
6a
vm_pageout_flush(ca77ce08,10,0,0,c0ad7a7c) at vm_pageout_flush+0xb1
vm_object_page_clean(cb72dae0,0,0,4,0) at vm_object_page_clean+0x36a
vfs_msync(c11e5200,2,ca768600,0,ca76e780) at vfs_msync+0xc5
sync_fsync(ca77cf7c) at sync_fsync+0x4b
sched_sync(0) at sched_sync+0xf3
fork_trampoline() at fork_trampoline+0x8
The problem occurs when msync() tries to flush non-contiguous dirty
pages. There is a hole in cluster code that does not check for
excessive dirty buffers before issuing a bdwrite(), resulting in
excessive dirty buffers. Since the cluster code is rather sensitive,
my fix occurs at a higher level (in the UFS code).
The patch below appears to solve the problem for Barry. I will commit
it to -current now, and to -stable in 2 days. I've also committed a
patch to writev() to -current and will backport it to -stable in 2 days
as well.
-Matt
Index: ufs/ufs/ufs_readwrite.c
===
RCS file: /home/ncvs/src/sys/ufs/ufs/ufs_readwrite.c,v
retrieving revision 1.65.2.3
diff -u -r1.65.2.3 ufs_readwrite.c
--- ufs/ufs/ufs_readwrite.c 2000/11/26 02:55:13 1.65.2.3
+++ ufs/ufs/ufs_readwrite.c 2000/12/06 20:03:59
@@ -495,6 +495,9 @@
if (ioflag IO_SYNC) {
(void)bwrite(bp);
+ } else if (vm_page_count_severe() || buf_dirty_count_severe()) {
+ bp-b_flags |= B_CLUSTEROK;
+ bawrite(bp);
} else if (xfersize + blkoffset == fs-fs_bsize) {
if ((vp-v_mount-mnt_flag MNT_NOCLUSTERW) == 0) {
bp-b_flags |= B_CLUSTEROK;
@@ -502,9 +505,6 @@
} else {
bawrite(bp);
}
- } else if (vm_page_count_severe() || buf_dirty_count_severe()) {
- bp-b_flags |= B_CLUSTEROK;
- bawrite(bp);
} else {
bp-b_flags |= B_CLUSTEROK;
bdwrite(bp);
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-stable" in the body of the message
sendmail, rbl orbs
Hi Can anyone send me two(or four:) simple lines to add to freebsd stock sendmail.cf that block mail from sites that are listed in RBL and ORBS databases ? TIA L.
