mbuma is an Mbuf & Cluster allocator built on top of a number of
extensions to the UMA framework, all included herein.
Extensions to UMA worth noting:
- Better layering between slab <-> zone caches; introduce
Keg structure which splits off slab cache away from the
zone structure and allows multiple zones to be stacked
on top of a single Keg (single type of slab cache);
perhaps we should look into defining a subset API on
top of the Keg for special use by malloc(9),
for example.
- UMA_ZONE_REFCNT zones can now be added, and reference
counters automagically allocated for them within the end
of the associated slab structures. uma_find_refcnt()
does a kextract to fetch the slab struct reference from
the underlying page, and lookup the corresponding refcnt.
mbuma things worth noting:
- integrates mbuf & cluster allocations with extended UMA
and provides caches for commonly-allocated items; defines
several zones (two primary, one secondary) and two kegs.
- change up certain code paths that always used to do:
m_get() + m_clget() to instead just use m_getcl() and
try to take advantage of the newly defined secondary
Packet zone.
- netstat(1) and systat(1) quickly hacked up to do basic
stat reporting but additional stats work needs to be
done once some other details within UMA have been taken
care of and it becomes clearer to how stats will work
within the modified framework.
From the user perspective, one implication is that the
NMBCLUSTERS compile-time option is no longer used. The
maximum number of clusters is still capped off according
to maxusers, but it can be made unlimited by setting
the kern.ipc.nmbclusters boot-time tunable to zero.
Work should be done to write an appropriate sysctl
handler allowing dynamic tuning of kern.ipc.nmbclusters
at runtime.
Additional things worth noting/known issues (READ):
- One report of 'ips' (ServeRAID) driver acting really
slow in conjunction with mbuma. Need more data.
Latest report is that ips is equally sucking with
and without mbuma.
- Giant leak in NFS code sometimes occurs, can't
reproduce but currently analyzing; brueffer is
able to reproduce but THIS IS NOT an mbuma-specific
problem and currently occurs even WITHOUT mbuma.
- Issues in network locking: there is at least one
code path in the rip code where one or more locks
are acquired and we end up in m_prepend() with
M_WAITOK, which causes WITNESS to whine from within
UMA. Current temporary solution: force all UMA
allocations to be M_NOWAIT from within UMA for now
to avoid deadlocks unless WITNESS is defined and we
can determine with certainty that we're not holding
any locks when we're M_WAITOK.
- I've seen at least one weird socketbuffer empty-but-
mbuf-still-attached panic. I don't believe this
to be related to mbuma but please keep your eyes
open, turn on debugging, and capture crash dumps.
This change removes more code than it adds.
A paper is available detailing the change and considering
various performance issues, it was presented at BSDCan2004:
http://www.unixdaemons.com/~bmilekic/netbuf_bmilekic.pdf
Please read the paper for Future Work and implementation
details, as well as credits.
Testing and Debugging:
rwatson,
brueffer,
Ketrien I. Saihr-Kesenchedra,
...
Reviewed by: Lots of people (for different parts)
been developed for use with FreeBSD, version 4.8 and later.
Submitted by: Hema Joyce
Reviewed by: Prafulla Deuskar
Approved by: Prafulla Deuskar
MFC after: 1 week
Fixed profiling of trap, syscall and interrupt handlers and some
ordinary functions, essentially by backing out half of rev.1.106 of
i386/exception.s. The handlers must be between certain labels for
the purposes of profiling, and this was broken by scattering them in
separately compiled .s files, especially for ordinary functions that
ended up between the labels. Merge the files by #including them as
before, except with different pathnames and better comments and
organization. Changes to the scattered files are minimal -- just
move the labels to the file that does the #includes.
This also partly fixes profiling of IPIs -- all IPI handlers are now
correctly classified as interrupt handlers, but many are still missing
mcount calls.
vm86bios.s is included as before, but it is now between the labels for
interrupt handlers again, which seems to be wrong since half of it is
for a non-interrupt handler.
converts miidevs to a .h file, so rename to reflect that.
The usb and pccard versions have also been renamed and will be hooked
into the build system shortly (I've made the conversion in my p4
tree).
ordinary functions, essentially by backing out half of rev.1.115 of
amd64/exception.S. The handlers must be between certain labels for
the purposes of profiling, and this was broken by scattering them in
separately compiled .S files, especially for ordinary functions that
ended up between the labels. Merge the files by #including them as
before, except with different pathnames and better comments and
organization. Changes to the scattered files are minimal -- just
move the labels to the file that does the #includes.
This also partly fixes profiling of IPIs -- all IPI handlers are now
correctly classified as interrupt handlers, but many are still missing
mcount calls.
Kernel profiling for amd64's (normal and high resolution) should now
compile and work as (un)well as on i386's. It works better than user
profiling because:
- it uses _cyg_profile_func_*() instead of .mcount(), so it doesn't suffer
from gcc misspelling .mcount as mcount.
- it doesn't neglect saving %rax in .mcount().
The SMP case hasn't been tested. The high resolution subcase of this uses
the i8254, and as on i386's, the locking for this is deficient and the
i8254 is too inefficient. The acpi timer is also too inefficient.
- Connect geom_stripe and geom_nop modules to the build.
- Connect STRIPE and NOP classes to the LINT build.
- Disconnect gconcat(8) from the build.
Supported by: Wheel - Open Technologies - http://www.wheel.pl
repocopied. Soon there will be additional bus attachments and
specialization for isa, acpi and pccard (and maybe pc98's cbus).
This was approved by nate, joerg and myself. bde dissented on the new
location, but appeared to be OK after some discussion.
yet, but building kld's is OK now and they can be loaded by kldload(2).
(but the machine will likely crash soon afterwards, a "minor" problem :-)
Brought to you by: my injured knee (from moving)
an option. Note that this option doesn't follow the normal USB_ or
Uxxx_ convention. That's because it is this way in the upstream
provider and I didn't want to change that.
"options OFW_NEWPCI").
This is a bit overdue, the new sparc64 OFW PCI code which is
meant to replace the old one is in place for 10 months and
enabled by default in GENERIC for 8 months. FreeBSD 5.2 and
5.2.1 also shipped with the new code enabled by default.
- Some minor clean-up, e.g. remove functions that encapsulated
the #ifdefs for OFW_NEWPCI, remove unused resp. no longer
required includes, etc.
Approved by: tmm, no objections on freebsd-sparc64
2. Note that ct device uses ctau name as driver name (due to name conflict
with ct driver) and also mark it as a driver inside the CVS tree.
MFC after: 10 days
register controlled the trigger mode and polarity of EISA interrupts.
However, it appears that most (all?) PCI systems use the ELCR to manage
the trigger mode and polarity of ISA interrupts as well since ISA IRQs used
to route PCI interrupts need to be level triggered with active low
polarity. We check to see if the ELCR exists by sanity checking the value
we get back ensuring that IRQS 0 (8254), 1 (atkbd), 2 (the link from the
slave PIC), and 8 (RTC) are all clear indicating edge trigger and active
high polarity.
This mini-driver will be used by the atpic driver to manage the trigger and
polarity of ISA IRQs. Also, the mptable parsing code will use this mini
driver rather than examining the ELCR directly.
returns okay when HW probe fails. This happens when comconsole flag is
set but VGA console is used instead.
Back out requested by: bde (He will be looking at other solutions from scratch)
synchronization protecting against dynamic load and unload of MAC
policies, and instead simply blocks load and unload. In a static
configuration, this allows you to avoid the synchronization costs
associated with introducing dynamicism.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, McAfee Research
parameter).
Keep using it only in the i386 NOTES for now. It is fairly MI, but it
doesn't use bus-space and has a couple of i386 i/o instructions in pci
intitialization.
- Define option FORCECONSPEED to force the serial console to
be CONSPEED. I've run into a lot of boards in which
the detect for prior speed doesn't work and ends up with
broken console since it is at the wrong speed.
- If a serial port is marked as a console, but console=vidconsole
and if the serial ports doesn't exist it will be probed and
attached at a 8250 chip. Then writes to that will freeze the
system.
- Add an option flags 0x400000 to mark this as a potential
comconsole in-case the one flaged with 0x10 does not exist
in the system.
This makes it easier to deploy on systems with one or two serial ports.
Obtained from: IronPort
gadgets (hotkeys, lcd, ...) on Asus laptops. I aim to closely track the
acpi4asus project which implements these features in the Linux kernel.
If this breaks your laptop, please let me know how it does it :-)
Approved by: njl (mentor)
like "the foo(4) manual page" to "foo(4)". Uniformized the remaining
instances of "manual page" and "manpage" to "man page". Uniformized
some nearby sentence breaks. Reformatted the whole paragraph containing
these changes only for DUMMYNET.
The VIA Nehemias is so obviously specific to i386 that it should not
be compiled on non-i386 platforms. The obviousness is in the fact that
all functions in nehemias.c are purely i386 inline assembly, guarded
by #ifdef __i386__
can more easily be used INSTEAD OF the hard-working Yarrow.
The only hardware source used at this point is the one inside
the VIA C3 Nehemiah (Stepping 3 and above) CPU. More sources will
be added in due course. Contributions welcome!
to awaken all waiters when a contested mutex is released instead of just
the highest priority waiter. If the various threads are awakened in
sequence then each thread may acquire and release the lock in question
without contention resulting in fewer expensive unlock and lock
operations. This old behavior of waking just the highest priority is
still used if this option is specified. Making the algorithm conditional
on a kernel option will allows us to benchmark both cases later and
determine which one should be used by default.
Requested by: tanimura-san
stuff was here (NFS) was fixed by Alfred in November. The only remaining
consumer of the stub functions was umapfs, which is horribly horribly
broken. It has missed out on about the last 5 years worth of maintenence
that was done on nullfs (from which umapfs is derived). It needs major
work to bring it up to date with the vnode locking protocol. umapfs really
needs to find a caretaker to bring it into the 21st century.
Functions GC'ed:
vop_noislocked, vop_nolock, vop_nounlock, vop_sharedlock.
was not present in what I originally tested when checking to see if
the kernel built/ran with the -O2 change. Recent instability in
sparc64 kernel was tracked to this. A reproducible kernel stack
traceback followed by hard hang during the call to msleep() at the
point the kernel waits 15 seconds for the SCSI bus to settle crept in
to recent kernel builds and it seems to go away with this patch.
Noticed by: kris
Approved by: rwatson (mentor)
extra entry for if_ndis_pci.c that depends on cardbus, just to cover
all the bases. (I don't think you can have cardbus without PCI, but
just in case...)
implementation could be characterized as a hybrid of the amd64 and i386
implementations. Specifically, the direct virtual-to-physical mapping is
used if possible and sf_buf_alloc() is used if the direct map cannot.
to select a serial console and debug port (resp). On ia64 these replace
the use of hints completely and take precedence over hints on alpha,
amd64 and i386. On sparc64 these variables are not yet recognised.
The reasons for introducing these variables are:
1. Hints have side-effects. They reserve the unit number for use by
isa or acpi devices and therefore cannot be used to select a pci
device. Also, the use of a unit number to select a device prior
to bus enumeration is nonsense. The new variables have no side-
effects and are not based on unit numbers.
2. Hints don't have the expression power to allow the sysadmin to
select UARTs that are not legacy PC devices and need the support
of compile-time constants to give the sysadmin some level of
flexibility.
The hw.uart.console and hw.uart.dbgport variables specify a list of
attributes. An attribute is a tag-value pair, seperated by a colon.
Attributes are seperated by a comma. Where possible, tags are the
same as those in /etc/remote (only br and pa in practice). Details
can be found in the manpage (not part of this commit).
Not tested on: amd64, pc98
from ddp_usrreq.c. Functions moved are:
at_pcballoc()
at_pcbconnect()
at_pcbdetach()
at_pcbdisconnect()
at_pcbsetaddr()
at_sockaddr()
Also moved are ddp_ports and ddpcb, global variables associated with DDP
pcbs. This makes PCB implementation more parallel to inet, inet6, and
ipx.
I've added -fno-strict-aliasing for now so we can ease into this.
I wanted to shoot for -O3, but the inlining caused problems due to GCC's
size heuristics; so also add -frename-registers, which is one of the things
-O3 would have given us.
mini-layer. I don't have time to bing it forward into the GEOM world, and
no one else has stepped forward to claim it. It'll be in the Attic for safe
keeping for now.
Only cy, bs and wd in the tree still use it. I have a replacement for
cy that I need to test on ISA and PCI cards. bs and wd are pc98 only
drivers that appear to no longer be necessary. I'll be removing them
when I hear back from the pc98 people.
this driver is being retired. Remove it from the tree. If someone
wants to update it to the latest APIs and can test the hardware, it
can return to the tree.
COMPAT_PCI api. This API is going away, so this driver is going away
also.
If users are interested in updating this, please contact the author
since he has some preliminary work to move this to newer APIs.
driver uses COMPAT_ISA shims, and those shims are going away.
It can be brought back if someone updates it to the latest APIs, and
moves it to the appropriate place in the tree.
Fix 'broken' ifdefs.
icc does not support profiling yet so remove unfinished code which was
supposed to help.
Submitted by: netchild (original version)
Reviewed by: ru
Intel C/C++ compiler (lang/icc) to build the kernel.
The icc CPUTYPE CFLAGS use icc v7 syntax, icc v8 moans about them, but
doesn't abort. They also produce CPU specific code (new instructions
of the CPU, not only CPU specific scheduling), so if you get coredumps
with signal 4 (SIGILL, illegal instruction) you've used the wrong
CPUTYPE.
Incarnations of this patch survive gcc compiles and my make universe.
I use it on my desktop.
To use it update share/mk, add
/usr/local/intel/compiler70/ia32/bin (icc v7, works)
or
/usr/local/intel_cc_80/bin (icc v8, doesn't work)
to your PATH, make sure you have a new kernel compile directory
(e.g. MYKERNEL_icc) and run
CFLAGS="-O2 -ip" CC=icc make depend
CFLAGS="-O2 -ip" CC=icc make
in it.
Don't compile with -ipo, the build infrastructure uses ld directly to
link the kernel and the modules, but -ipo needs the link step to be
performed with Intel's linker.
Problems with icc v8:
- panic: npx0 cannot be emulated on an SMP system
- UP: first start of /bin/sh results in a FP exception
Parts of this commit contains suggestions or submissions from
Marius Strobl <marius@alchemy.franken.de>.
Reviewed by: silence on -arch
Submitted by: netchild
This adds the former ports registered groups: proxy and authpf as well as
the proxy user. Make sure to run mergemaster -p in oder to complete make
installworld without errors.
This also provides the passive OS fingerprints from OpenBSD (pf.os) and an
example pf.conf.
For those who want to go without pf; it provides a NO_PF knob to make.conf.
__FreeBSD_version will be bumped soon to reflect this and to be able to
change ports accordingly.
Approved by: bms(mentor)
ethernet (tested) and FDDI (not tested). The main use for this is on ADSL (or
other ATM) connections where bridged ethernet is used, PPPoE being a prime
example.
There is no manual page as yet, I will write one shortly.
Reviewed by: harti
but a bit more reamins to be done. For now, it is usable.
PR:
Submitted by: Taku YAMAMOTO <taku@cent.saitama-u.ac.jp>
Reviewed by:
Approved by:
Obtained from:
MFC after:
generic watchdoc(9) interface.
Make watchdogd(8) perform as watchdog(8) as well, and make it
possible to specify a check command to run, timeout and sleep
periods.
Update watchdog(4) to talk about the generic interface and add
new watchdog(8) page.
sleep queue interface:
- Sleep queues attempt to merge some of the benefits of both sleep queues
and condition variables. Having sleep qeueus in a hash table avoids
having to allocate a queue head for each wait channel. Thus, struct cv
has shrunk down to just a single char * pointer now. However, the
hash table does not hold threads directly, but queue heads. This means
that once you have located a queue in the hash bucket, you no longer have
to walk the rest of the hash chain looking for threads. Instead, you have
a list of all the threads sleeping on that wait channel.
- Outside of the sleepq code and the sleep/cv code the kernel no longer
differentiates between cv's and sleep/wakeup. For example, calls to
abortsleep() and cv_abort() are replaced with a call to sleepq_abort().
Thus, the TDF_CVWAITQ flag is removed. Also, calls to unsleep() and
cv_waitq_remove() have been replaced with calls to sleepq_remove().
- The sched_sleep() function no longer accepts a priority argument as
sleep's no longer inherently bump the priority. Instead, this is soley
a propery of msleep() which explicitly calls sched_prio() before
blocking.
- The TDF_ONSLEEPQ flag has been dropped as it was never used. The
associated TDF_SET_ONSLEEPQ and TDF_CLR_ON_SLEEPQ macros have also been
dropped and replaced with a single explicit clearing of td_wchan.
TD_SET_ONSLEEPQ() would really have only made sense if it had taken
the wait channel and message as arguments anyway. Now that that only
happens in one place, a macro would be overkill.
pf/pflog/pfsync as modules. Do not list them in NOTES or modules/Makefile
(i.e. do not connect it to any (automatic) builds - yet).
Approved by: bms(mentor)
to a new mac_inet.c. This code is now conditionally compiled based
on inet support being compiled into the kernel.
Move socket related MAC Framework entry points from mac_net.c to a new
mac_socket.c.
To do this, some additional _enforce MIB variables are now non-static.
In addition, mbuf_to_label() is now mac_mbuf_to_label() and non-static.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, McAfee Research
also prints the actual numerical value of the symbol in question.
Users of addr2line(1) will be less proficient in hex arithmetic as a
consequence.
This amongst other things means that traceback lines change from:
siointr1(c4016800,c073bda0,0,c06b699c,69f) at siointr1+0xc5
to
siointr1(c4016800,c073bda0,0,c06b699c,69f) at 0xc062b0bd = siointr1+0xc5
I made this an option to avoid bikesheds.
~
~
~
The nonstandard formatting made my mega-patch scripts miss it.
Retire the static major number while we're here anyway.
Reported by: Niels Chr. Bank-Pedersen <ncbp@bank-pedersen.dk>
This commit adds a couple of functions for pseudodrivers to use for
implementing cloning in a manner we will be able to lock down (shortly).
Basically what happens is that pseudo drivers get a way to ask for
"give me the dev_t with this unit number" or alternatively "give
me a dev_t with the lowest guaranteed free unit number" (there is
unfortunately a lot of non-POLA in the exact numeric value of this
number, just live with it for now)
Managing the unit number space this way removes the need to use
rman(9) to do so in the drivers this greatly simplifies the code in
the drivers because even using rman(9) they still needed to manage
their dev_t's anyway.
I have taken the if_tun, if_tap, snp and nmdm drivers through the
mill, partly because they (ab)used makedev(), but mostly because
together they represent three different problems for device-cloning:
if_tun and snp is the plain case: just give me a device.
if_tap has two kinds of devices, with a flag for device type.
nmdm has paired devices (ala pty) can you can clone either of them.
Free approx 86 major numbers with a mostly automatically generated patch.
A number of strategic drivers have been left behind by caution, and a few
because they still (ab)use their major number.