Make this more lint-friendly. This file seems to be invoked in just
about any .c file that includes a .h, and lint produces copious
whining because of the asm ...; stuff.
Back out most of revision 1.28: lists of diagnostics must use -diag,
not -tag. Instead, put a period after the error messages to aide
those using dumb terminals not capable of properly displaying markup.
o Disallow two "allow this" exceptions in p_cansignal() restricting
the ability of unprivileged processes to deliver arbitrary signals
to daemons temporarily taking on unprivileged effective credentials
when P_SUGID is not set on the target process:
Removed:
(p1->p_cred->cr_ruid != ps->p_cred->cr_uid)
(p1->p_ucred->cr_uid != ps->p_cred->cr_uid)
o Replace two "allow this" exceptions in p_cansignal() restricting
the ability of unprivileged processes to deliver arbitrary signals
to daemons temporarily taking on unprivileged effective credentials
when P_SUGID is set on the target process:
Replaced:
(p1->p_cred->p_ruid != p2->p_ucred->cr_uid)
(p1->p_cred->cr_uid != p2->p_ucred->cr_uid)
With:
(p1->p_cred->p_ruid != p2->p_ucred->p_svuid)
(p1->p_ucred->cr_uid != p2->p_ucred->p_svuid)
o These changes have the effect of making the uid-based handling of
both P_SUGID and non-P_SUGID signal delivery consistent, following
these four general cases:
p1's ruid equals p2's ruid
p1's euid equals p2's ruid
p1's ruid equals p2's svuid
p1's euid equals p2's svuid
The P_SUGID and non-P_SUGID cases can now be largely collapsed,
and I'll commit this in a few days if no immediate problems are
encountered with this set of changes.
o These changes remove a number of warning cases identified by the
proc_to_proc inter-process authorization regression test.
o As these are new restrictions, we'll have to watch out carefully for
possible side effects on running code: they seem reasonable to me,
but it's possible this change might have to be backed out if problems
are experienced.
peter [Fri, 13 Apr 2001 01:45:22 +0000 (01:45 +0000)]
Remove the 'DO NOT EDIT THIS FILE' crud that we spit out with 'crontab -l'.
Otherwise, "crontab -l > file; vi file; crontab file" adds an extra set
of "DO NOT EDIT" markers each and every time which is a bit silly.
Match ip6fw's command line options to those of ipfw (specifically, added
the ability to use a preprocessor, use the -q (quiet) flag when reading
from a file). The source used is from ipfw.
Clean up exit codes while I am here.
KAME has been informed and plans on integrating these patches into their
own source as well.
o Disable two "allow this" exceptions in p_cansched()m retricting the
ability of unprivileged processes to modify the scheduling properties
of daemons temporarily taking on unprivileged effective credentials.
These cases (p1->p_cred->p_ruid == p2->p_ucred->cr_uid) and
(p1->p_ucred->cr_uid == p2->p_ucred->cr_uid), respectively permitting
a subject process to influence the scheduling of a daemon if the subject
process has the same real uid or effective uid as the daemon's effective
uid. This removes a number of the warning cases identified by the
proc_to_proc iner-process authorization regression test.
o As these are new restrictions, we'll have to watch out carefully for
possible side effects on running code: they seem reasonable to me,
but it's possible this change might have to be backed out if problems
are experienced.
o Make kqueue's filt_procattach() function use the error value returned
by p_can(...P_CAN_SEE), rather than returning EACCES directly. This
brings the error code used here into line with similar arrangements
elsewhere, and prevents the leakage of pid usage information.
o Reduce information leakage into jails by adding invocations of
p_can(...P_CAN_SEE...) to getpgid(), getsid(), and setpgid(),
blocking these operations on processes that should not be visible
by the requesting process. Required to reduce information leakage
in MAC environments.
o Replace p_cankill() with p_cansignal(), remove wrappage of p_can()
from signal authorization checking.
o p_cansignal() takes three arguments: subject process, object process,
and signal number, unlike p_cankill(), which only took into account
the processes and not the signal number, improving the abstraction
such that CANSIGNAL() from kern_sig.c can now also be eliminated;
previously CANSIGNAL() special-cased the handling of SIGCONT based
on process session. privused is now deprecated.
o The new p_cansignal() further limits the set of signals that may
be delivered to processes with P_SUGID set, and restructures the
access control check to allow it to be extended more easily.
o These changes take into account work done by the OpenBSD Project,
as well as by Robert Watson and Thomas Moestl on the TrustedBSD
Project.
o Enable -DSETSUGID_SUPPORTED in inter-process authorization regression
test by default, as setugid() is now part of the base kernel (assuming
(options REGRESSION) has been enabled for the running kernel).
o Introduce a new system call, __setsugid(), which allows a process to
toggle the P_SUGID bit explicitly, rather than relying on it being
set implicitly by other protection and credential logic. This feature
is introduced to support inter-process authorization regression testing
by simplifying userland credential management allowing the easy
isolation and reproduction of authorization events with specific
security contexts. This feature is enabled only by "options REGRESSION"
and is not intended to be used by applications. While the feature is
not known to introduce security vulnerabilities, it does allow
processes to enter previously inaccessible parts of the credential
state machine, and is therefore disabled by default. It may not
constitute a risk, and therefore in the future pending further analysis
(and appropriate need) may become a published interface.
o Introduce "options REGRESSION", a kernel option which enables
interfaces and functionality intended for use during correctness and
regression testing. Features enabled by "options REGRESSION" may
in and of themselves introduce security or correctness problems if
used improperly, and so are not intended for use in production
systems, only in testing environments.
o First pass at an inter-process authorization regression testing suite.
This test utility attempts to evaluate the current kernel policy
for authorization inter-process activities, currently ptrace(),
kill(, SIGHUP), getpriority(), and setpriority(). The utility creates
pairs of processes, initializes their credential sets to useful
cases, and reports on whether the results are in keeping with hard-coded
safety expectations.
o Currently, this utility relies on the availability of __setugid(),
an uncomitted system call used for managing the P_SUGID bit. Due to
continuing discussion of optional regression testing kernel components
("options REGRESSION") I'll hold off on committing that until the
discussion has reached its natural termination.
o A number of additional testing factors should be taken into account
in the testing, including tests for different classes of signals,
interactions with process session characteristics, I/O signalling,
broadcast activities such as broadcast signalling, mass priority
setting, and to take into group-related aspects of credentials.
Additional operations should also be taken into account, such as ktrace,
debugging attach using procfs, and so on.
o This testing suite is intended to prevent the introduction of bugs
in the upcoming sets of authorization changes associated with the
introduction of process capabilities and mandatory access control.
Parse the various argument registers in the printtrap() function so that
one doesn't have to go grab a reference manual to decode them by hand every
time the alpha kernel falls over.
peter [Wed, 11 Apr 2001 00:39:20 +0000 (00:39 +0000)]
Create debug.hashstat.[raw]nchash and debug.hashstat.[raw]nfsnode to
enable easy access to the hash chain stats. The raw prefixed versions
dump an integer array to userland with the chain lengths. This cheats
and calls it an array of 'struct int' rather than 'int' or sysctl -a
faithfully dumps out the 128K array on an average machine. The non-raw
versions return 4 integers: count, number of chains used, maximum chain
length, and percentage utilization (fixed point, multiplied by 100).
The raw forms are more useful for analyzing the hash distribution, while
the other form can be read easily by humans and stats loggers.
Fix a typo relating to the "-U" (force UDP for mount protocol)
option. When specified, make sure to use the correct netid for the
getnetconfigent() call, and also in error messages.
brian [Tue, 10 Apr 2001 23:16:55 +0000 (23:16 +0000)]
o The -s limit is ARG_MAX - 4K, not ARG_MAX - 2K.
o Mention that the current environment is part of the -s calculation.
o Add a BUGS section that warns against executing a program that increases
the size of the argument list or the size of the environment.
I have wondered for a while what the difference is between
get a big list | xargs sudo command
which fails and
get a big list | sudo xargs command
which succeeds. The answer is that in the first case, sudo expands
the environment and pushes the amount of data passed into execve over
the E2BIG threshold.
Remove the old APIC I/O higher level IPI API in favor of the newer MI
API for IPI's that isn't tied to the Intel APIC. MD code can still use
the apic_ipi() function or dink with the apic directly if needed to send
MD IPI's.
Catch up to the dirpref changes by copying new fields in the alternate
superblock from the original superblock so that differences in those new
fields are ignored.
Split out all the RPC code into a separate function and address a
number of issues:
- Fix background mounts; these were broken in revision 1.40.
- Don't give up before trying all addresses returned by getaddrinfo().
- Use protocol-independent routines where possible.
- Improve error reporting for RPC errors.
- In non-background mode, give up after trying all protocols once.
- Use daemon(3) instead of rolling our own version.
- Never go ahead with the mount() syscall until we have received
a reply from the remote nfsd; this is especially important with
non-interruptible mounts, as otherwise a mistyped command might
require a reboot to correct.
Add an MI API for sending IPI's. I used the same API present on the alpha
because:
- it used a better namespace (smp_ipi_* rather than *_ipi),
- it used better constant names for the IPI's (IPI_* rather than
X*_OFFSET), and
- this API also somewhat exists for both alpha and ia64 already.
Add another card to the list of Neomagic 256AV's which don't have AC97
codecs. Also, add some additional code to check for future cards without
this feature - attempting to initialise them as AC97 cards will hang the
machine.
Add information about the new options to newfs and tunefs which set the
expected average file size and number of files per directory. Could do
with some fleshing out.
Directory layout preference improvements from Grigoriy Orlov <gluk@ptci.ru>.
His description of the problem and solution follow. My own tests show
speedups on typical filesystem intensive workloads of 5% to 12% which
is very impressive considering the small amount of code change involved.
------
One day I noticed that some file operations run much faster on
small file systems then on big ones. I've looked at the ffs
algorithms, thought about them, and redesigned the dirpref algorithm.
First I want to describe the results of my tests. These results are old
and I have improved the algorithm after these tests were done. Nevertheless
they show how big the perfomance speedup may be. I have done two file/directory
intensive tests on a two OpenBSD systems with old and new dirpref algorithm.
The first test is "tar -xzf ports.tar.gz", the second is "rm -rf ports".
The ports.tar.gz file is the ports collection from the OpenBSD 2.8 release.
It contains 6596 directories and 13868 files. The test systems are:
1. Celeron-450, 128Mb, two IDE drives, the system at wd0, file system for
test is at wd1. Size of test file system is 8 Gb, number of cg=991,
size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current
from Dec 2000 with BUFCACHEPERCENT=35
2. PIII-600, 128Mb, two IBM DTLA-307045 IDE drives at i815e, the system
at wd0, file system for test is at wd1. Size of test file system is 40 Gb,
number of cg=5324, size of cg is 8m, block size = 8k, fragment size = 1k
OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=50
You can get more info about the test systems and methods at:
http://www.ptci.ru/gluk/dirpref/old/dirpref.html
Test Results
tar -xzf ports.tar.gz rm -rf ports
mode old dirpref new dirpref speedup old dirprefnew dirpref speedup
First system
normal 667 472 1.41 477 331 1.44
async 285 144 1.98 130 14 9.29
sync 768 616 1.25 477 334 1.43
softdep 413 252 1.64 241 38 6.34
Second system
normal 329 81 4.06 263.5 93.5 2.81
async 302 25.7 11.75 112 2.26 49.56
sync 281 57.0 4.93 263 90.5 2.9
softdep 341 40.6 8.4 284 4.76 59.66
"old dirpref" and "new dirpref" columns give a test time in seconds.
speedup - speed increasement in times, ie. old dirpref / new dirpref.
------
Algorithm description
The old dirpref algorithm is described in comments:
/*
* Find a cylinder to place a directory.
*
* The policy implemented by this algorithm is to select from
* among those cylinder groups with above the average number of
* free inodes, the one with the smallest number of directories.
*/
A new directory is allocated in a different cylinder groups than its
parent directory resulting in a directory tree that is spreaded across
all the cylinder groups. This spreading out results in a non-optimal
access to the directories and files. When we have a small filesystem
it is not a problem but when the filesystem is big then perfomance
degradation becomes very apparent.
What I mean by a big file system ?
1. A big filesystem is a filesystem which occupy 20-30 or more percent
of total drive space, i.e. first and last cylinder are physically
located relatively far from each other.
2. It has a relatively large number of cylinder groups, for example
more cylinder groups than 50% of the buffers in the buffer cache.
The first results in long access times, while the second results in
many buffers being used by metadata operations. Such operations use
cylinder group blocks and on-disk inode blocks. The cylinder group
block (fs->fs_cblkno) contains struct cg, inode and block bit maps.
It is 2k in size for the default filesystem parameters. If new and
parent directories are located in different cylinder groups then the
system performs more input/output operations and uses more buffers.
On filesystems with many cylinder groups, lots of cache buffers are
used for metadata operations.
My solution for this problem is very simple. I allocate many directories
in one cylinder group. I also do some things, so that the new allocation
method does not cause excessive fragmentation and all directory inodes
will not be located at a location far from its file's inodes and data.
The algorithm is:
/*
* Find a cylinder group to place a directory.
*
* The policy implemented by this algorithm is to allocate a
* directory inode in the same cylinder group as its parent
* directory, but also to reserve space for its files inodes
* and data. Restrict the number of directories which may be
* allocated one after another in the same cylinder group
* without intervening allocation of files.
*
* If we allocate a first level directory then force allocation
* in another cylinder group.
*/
My early versions of dirpref give me a good results for a wide range of
file operations and different filesystem capacities except one case:
those applications that create their entire directory structure first
and only later fill this structure with files.
My solution for such and similar cases is to limit a number of
directories which may be created one after another in the same cylinder
group without intervening file creations. For this purpose, I allocate
an array of counters at mount time. This array is linked to the superblock
fs->fs_contigdirs[cg]. Each time a directory is created the counter
increases and each time a file is created the counter decreases. A 60Gb
filesystem with 8mb/cg requires 10kb of memory for the counters array.
The maxcontigdirs is a maximum number of directories which may be created
without an intervening file creation. I found in my tests that the best
performance occurs when I restrict the number of directories in one cylinder
group such that all its files may be located in the same cylinder group.
There may be some deterioration in performance if all the file inodes
are in the same cylinder group as its containing directory, but their
data partially resides in a different cylinder group. The maxcontigdirs
value is calculated to try to prevent this condition. Since there is
no way to know how many files and directories will be allocated later
I added two optimization parameters in superblock/tunefs. They are:
int32_t fs_avgfilesize; /* expected average file size */
int32_t fs_avgfpdir; /* expected # of files per directory */
These parameters have reasonable defaults but may be tweeked for special
uses of a filesystem. They are only necessary in rare cases like better
tuning a filesystem being used to store a squid cache.
I have been using this algorithm for about 3 months. I have done
a lot of testing on filesystems with different capacities, average
filesize, average number of files per directory, and so on. I think
this algorithm has no negative impact on filesystem perfomance. It
works better than the default one in all cases. The new dirpref
will greatly improve untarring/removing/coping of big directories,
decrease load on cvs servers and much more. The new dirpref doesn't
speedup a compilation process, but also doesn't slow it down.
alfred [Tue, 10 Apr 2001 05:29:26 +0000 (05:29 +0000)]
Add more diagnostic output for failure.
s/1518/ETHER_MAX_LEN
Some style changes, add some braces, mostly residual from having
a lot of debug hooks added while working on this driver.
Bring in a plethora of changes from NetBSD:
revision 1.58
date: 2001/03/08 11:07:08; author: ichiro; state: Exp; lines: +17 -1
it wait until busy flag disappears.
it was able to prevent some cards with late initializing faling in wi_reset().
revision 1.41
date: 2000/10/13 19:15:08; author: jonathan; state: Exp; lines: +4 -2
Fix wi_intr() to avoid touching card registers during insert/remove events,
when sharing an interrupt with other devices:
check sc->sc_enabled, and drop the interrupt if its' off.
revision 1.30
date: 2000/08/18 04:11:48; author: jhawk; state: Exp; lines: +4 -4
Copy wi_{dst,src}_addr from struct wi_frame into faked-up ether_header
instead of addr1 and addr2. THis means that tcpdump -e will show the
correct MAC address for communications with access points instead of showing
the BSSID.
In the future there should be 802.11 support for bpf/libpcap/tcpdump,
but that is aways down the road.
Clean up a bit. Use the correct TAILQ link when walking the thread
lists to free thread resources after a fork (in the child). Also
remember to free the dead thread list.