".." vnode. This is cheaper storagewise than keeping it in the
namecache, and it makes more sense since it's a 1:1 mapping.
2. Also handle the case of "." more intelligently rather than stuff
the namecache with pointless entries.
3. Add two lists to the vnode and hang namecache entries which go from
or to this vnode. When cleaning a vnode, delete all namecache
entries it invalidates.
4. Never reuse namecache enties, malloc new ones when we need it, free
old ones when they die. No longer a hard limit on how many we can
have.
5. Remove the upper limit on namelength of namecache entries.
6. Make a global list for negative namecache entries, limit their number
to a sysctl'able (debug.ncnegfactor) fraction of the total namecache.
Currently the default fraction is 1/16th. (Suggestions for better
default wanted!)
7. Assign v_id correctly in the face of 32bit rollover.
8. Remove the LRU list for namecache entries, not needed. Remove the
#ifdef NCH_STATISTICS stuff, it's not needed either.
9. Use the vnode freelist as a true LRU list, also for namecache accesses.
10. Reuse vnodes more aggresively but also more selectively, if we can't
reuse, malloc a new one. There is no longer a hard limit on their
number, they grow to the point where we don't reuse potentially
usable vnodes. A vnode will not get recycled if still has pages in
core or if it is the source of namecache entries (Yes, this does
indeed work :-) "." and ".." are not namecache entries any longer...)
11. Do not overload the v_id field in namecache entries with whiteout
information, use a char sized flags field instead, so we can get
rid of the vpid and v_id fields from the namecache struct. Since
we're linked to the vnodes and purged when they're cleaned, we don't
have to check the v_id any more.
12. NFS knew about the limitation on name length in the namecache, it
shouldn't and doesn't now.
Bugs:
The namecache statistics no longer includes the hits for ".."
and "." hits.
Performance impact:
Generally in the +/- 0.5% for "normal" workstations, but
I hope this will allow the system to be selftuning over a
bigger range of "special" applications. The case where
RAM is available but unused for cache because we don't have
any vnodes should be gone.
Future work:
Straighten out the namecache statistics.
"desiredvnodes" is still used to (bogusly ?) size hash
tables in the filesystems.
I have still to find a way to safely free unused vnodes
back so their number can shrink when not needed.
There is a few uses of the v_id field left in the filesystems,
scheduled for demolition at a later time.
Maybe a one slot cache for unused namecache entries should
be implemented to decrease the malloc/free frequency.
- parse command options using getopt for consistancy
- sanitise the command parsing so that it's less like spaghetti
- implement a "-l" option (idea from NetBSD - just list names)
- attempt to clean up the sysctl parsing loop some more. It still needs
to be taken out the back and shot though.
- cut down on global usage, but there's a lot more scope for this.
- make usage string a bit closer to reality (it was missing lots of things)
Unfortunately, I did this for the second time but with the memory of
the NetBSD version still recently in my mind. It's hard to redo simple
changes or getopt stuff without making it look like what you've been
working with a few hours ago.
(good thing for apache f.e.) This change is backward compatible with old
newsyslog.conf files.
2) Do not compress log if SIGHUP sending failed for some reason
(f.e. pid file deleted). Newcoming messages will be lost
otherwise.
3) Misc cleanup while I am here.
found that my syslogd is now running them for several months...
Add an option to syslogd to restrict the IP addresses that are allowed
to log to this syslogd. It's too late to develop the inter-syslogd
communications protocol mentioned in the BUGS section, some 10 years
too late. Thus, restricting the IP address range is about the most
effective change we can do if we want to allow incoming syslog
messages at all.
IMHO, we should encourage the system administrators to use this option,
and thus provide a knob in /etc/rc.* for it, defaulting to -a 127.0.0.1/32
(just as a hint about the usage).
Please state opinions about whether to merge this change into 2.2 or
not (i've got it running on RELENG_2_2 anyway).
to 0775.
This does *not* instantly make any program which "ensures"
mail spool consistency by creating lock files safe in any way
since other tools, like mail.local, will be using flock() semantics
and any such lock file will simply be ignored. It does, however,
allow a lot of things which are currently suid root in order to create
such bogus lockfiles to, at least, be bogus at a much lower level of
privilege (and this is good). Ultimately, of course, everybody should
just use flock.
I am not going to commit anything to this area for a few days.
This is because
1) I want everyone to be DARN sure there is no export of crypto
that may get our USA friends it trouble.
2) I have been asked by the folk developing KTH-eBones to hold off
for their new release.
Worked with: rkw, jdp
CVS:
CVS:
so that all these makefiles can be used to build libc_r too.
Added .if ${LIB} == "c" tests to restrict man page builds to libc
to avoid needlessly building them with libc_r too.
Split libc Makefile into Makefile and Makefile.inc to allow the
libc_r Makefile to include Makefile.inc too.
