date: 2006/05/27 19:32:28; author: krion; state: Exp; lines: +1 -2
Fix using "P" command on empty pattern space. If the "P" command is
used once on a non-empty pattern space and then again on an empty
pattern space, the second usage restores the pattern space length to
the length that it had when the first "P" was used.
PR: bin/96052
Submitted by: Andrey Zholos <aaz@althenia.net>
When allocating a bucket to hold a free'd item in UMA fails, don't
report this as an allocation failure for the item type. The failure
will be separately recorded with the bucket type. This may eliminate
high mbuf allocation failure counts under some circumstances, which
can be alarming in appearance, but not actually a problem in
practice.
Reported by: ps, Peter J. Blok <pblok at bsd4all dot org>,
OxY <oxy at field dot hu>,
Gabor MICSKO <gmicskoa at szintezis dot hu>
This is a RELENG_6_1 merge candidate.
Tested by: Peter J. Blok <pblok at bsd4all dot org>,
OxY <oxy at field dot hu>,
Michael Butler <imb at protected-networks dot net>
Use only lower 64bit of src/dest (and src/dest port) for hashing of IPv6
connections and get rid of the flow_id as it is not guaranteed to be stable
some (most?) current implementations seem to just zero it out.
PR: kern/88664
Move call to ignore SIGPIPE signals before calling fork(),
so that both parent and child processes ignore this signal.
PR: bin/97768
Submitted by: Gea-Suan Lin <gslin at csie dot nctu dot edu dot tw>
vmspace_exitfree() and vmspace_free() which could result in the same
vmspace being freed twice.
Factor out part of exit1() into new function vmspace_exit(). Attach
to vmspace0 to allow old vmspace to be freed earlier.
Add new function, vmspace_acquire_ref(), for obtaining a vmspace
reference for a vmspace belonging to another process. Avoid changing
vmspace refcount from 0 to 1 since that could also lead to the same
vmspace being freed twice.
Change vmtotal() and swapout_procs() to use vmspace_acquire_ref().
date: 2006/05/27 18:27:41; author: krion; state: Exp; lines: +35 -3
Add the capability for a trailing scale indicator to cause the
specified size to be read in the more familiar units of kilobytes,
megabytes, gigabytes, terabytes and petabytes.
PR: bin/50988
Submitted by: Matthew Seaman <m.seaman@infracaninophile.co.uk>
- When setting up a packet for transmit, if we the tx ring is over half
full, kick the binary blob to force it to complete any pending tx
completions.
- In the watchdog routine, poke the binary blob to force it to flush any
pending tx completions and only reset the chip if the blob doesn't
complete any of them.
regression, multi-CDROM package loading problems, mbufs denied
bug, i386/nextboot problem.
These items will be updated as necessary to reflect reality.
Edited alpha sysinstall(8) item for markup and punctuation nits.
Correct a bug in the handling of backslash characters in smbfs which can
allow an attacker to escape from a chroot(2). [2]
Security: FreeBSD-SA-06:15.ypserv [1]
Security: FreeBSD-SA-06:16.smbfs [2]
Commands like gmirror, graid3, ... and others which use dlopen() to load
classes from say, /lib/geom, cannot be statically linked completely.
Moreover, those shared objects may require other shared objects (i.e.
for geom, libraries like -lmd, -lcrypto).
The libs_so extension to crunchgen fixes this by allowing some libraries
to be linked in dynamically. This requires that a copy of rtld and the
shared libraries be made available to the crunched binary, and so is not
suitable for all environments. Crunchgen configurations which do not
use the 'libs_so' keyword are unaffected and produce identical binaries
with and without this commit.
Approved by: jhb
In collaboration with: Adrian Steinmann <ast at marabu dot ch>
mimicing the NFS reference implementation.
NFS over TCP does not need fast retransmit timeouts, since network loss
and congestion are managed by the transport (TCP), unlike with NFS over
UDP. A long timeout prevents the unnecessary retransmission of non-
idempotent NFS requests.
Reviewed by: mohans, silby, rees?
Sponsored by: Network Appliance, Incorporated
the estimator to be more easily tuned and maintained.
There should be no functional change except there is now a lower limit
on the retransmit timeout to prevent the client from retransmitting
faster than the server's disks can fill requests, and an upper limit
to prevent the estimator from taking too long to retransmit during a
server outage.
Reviewed by: mohan, kris, silby
Sponsored by: Network Appliance, Incorporated
for nfsclient and nfs4client in order to prevent local root users
from panicing the system.
MFC of nfs4client/nfs4_dev.c revision 1.9.
PR: kern/77463
Submitted by: Wojciech A. Koszek
Reviewed by: cel, rees
Security: Local root users can panic the system at will
Sponsored by: Network Appliance, Incorporated
Remove hardcoded dependencies on ufs, cd9660, msdosfs, and ntfs.
MFC: mountd.c 1.84, 1.85, 1.86
- exclude changes to pass in errmsg mount option, because errmsg
is not in RELENG_6's vfs_mount.c as a default mount option.
Makefile 1.17
PR: bin/97642
In devfs_first(), set mp->mnt_opt to a valid empty list of mount options
instead of leaving it NULL. This eliminates a kernel panic
when trying to do a mount -o update of /dev.
Noticed by: cjsp
Reviewed by: phk
I encountered repeatable Unaligned access panics similar to:
fatal kernel trap:
trap entry = 0x4 (unaligned access fault)
cpuid = 0
faulting va = 0xfffffc0000fae1c2
opcode = 0x28
register = 0x1
pc = 0xfffffc00004d07dc
ra = 0xfffffc00004d0a1c
sp = 0xfffffe0031875bc0
curthread = 0xfffffc007da237a0
pid = 2, comm = g_event
These occur when a floppy or a CD (the 6.1R install CD will do in most
cases) is in the drive during boot. A floppy or CD inserted *after*
the system was booted did not give me this problem. Interestingly
enough booting from the 6.1-R install CD sometimes caused the crash,
sometimes it did not. When it did not panic an installation of 6.1-R
was possible without any incidents. A subsequent boot from HD with
the CD still present sometimes(...) worked fine, sometimes paniced.
Taking GEOM_GPT out of the kernel config file allows booting without
incidents. We do not need GPT on Alpha, so this is the easy way out.
Discussed with: phk
Add a /media to FreeBSD. /media is a directory designed to contain
subdirectories that will be used for removable media mount points (i.e.
mount points for CDs, floppy disks, USB drives, etc.). While the primary
purpose of /media is to provide a location for HAL
(http://www.freedesktop.org/wiki/Software_2fhal) to mount volumes, it could
be used by any application that needs to manage removable media volumes.
If the user asks for "kernel sources" to be installed, extract the
SRC_BASE package (src/[A-Z]*) as well as SRC_SYS (src/sys/*). This
allows users who only install the kernel source code to use the
modern "make buildkernel" approach.
usr.sbin/portsnap/portsnap/portsnap.8 revs 1.8-1.9,
usr.sbin/portsnap/portsnap/portsnap.sh revs 1.16-1.24:
* Allow "HTTP_PROXY" to be (mis)spelled "http_proxy".
* Add "-l descfile" option to portsnap, to merge local describes
output when building INDEX files.
* If a port directory is a symlink, remove the symlink when
extracting a new version of the port.
* Make portsnap's automatic mirror selection interoperate better
with HTTP proxies.
* Switch mirrors if the first mirror(s) selected fails.
* During mirror selection, parse the output of host(1) from
BIND 8 as well as BIND 9.
* NO_BIND -> WITHOUT_BIND in a comment.
PR: ports/93901, bin/96288
Sponsored by: FreeBSD security development fundraiser
