just to ensure 32-bit variables. Doing so broke and/or pessimized
i386's with 64-bit longs (unnecessary use of 64-bit variables
caused remarkably few problems in C code, but the inline asm here
tended to fail because there are no 64-bit registers). Since the
interfaces here are very machine-dependent and shouldn't be used
outside of the kernel, use a standard types of "known" width instead
of fixed-width types.
Changed all quad_t's to u_int64_t's. quad_t isn't standard, and
using signed types for 64-bit registers was bogus (but made no
difference).
to ensure 32-bit variables. Doing so broke i386's with 64-bit
longs. Use fixed-size integral types instead of plain ints, shorts,
chars and pointers since the bootinfo struct layout is a binary
interface. The boot blocks could reasonably be implemented using
16-bit code.
least unsuitable for holding an object pointer. This should have been
used to fix warnings about casts between pointers and ints on alphas.
Moved corresponding existing general typedef (fptrint_t) for function
pointers from the i386 <machine/profile.h> to a kernel-only typedef
in <machine/types.h>. Kludged libc/gmon/mcount.c so that it can
still see this typedef.
is the kernel part of my commits, the userlevel stuff will be done in
a separate commit. Add the ability to suspend as well as hibernate to
syscons. Create a new virtual key like hibernate for suspend. Update
apm_bios.h to define more apm bios goodies.
There is only cdevsw (which should be renamed in a later edit to deventry
or something). cdevsw contains the union of what were in both bdevsw an
cdevsw entries. The bdevsw[] table stiff exists and is a second pointer
to the cdevsw entry of the device. it's major is in d_bmaj rather than
d_maj. some cleanup still to happen (e.g. dsopen now gets two pointers
to the same cdevsw struct instead of one to a bdevsw and one to a cdevsw).
rawread()/rawwrite() went away as part of this though it's not strictly
the same patch, just that it involves all the same lines in the drivers.
cdroms no longer have write() entries (they did have rawwrite (?)).
tapes no longer have support for bdev operations.
Reviewed by: Eivind Eklund and Mike Smith
Changes suggested by eivind.
Kapok Computer Co. notebook with AMI 'WinBIOS' which seems to insist
on having a short jump and nop as the first instructions in the
boot sector code. The prevailing theory is that the BIOS is doing
some sort of boot sector virus detection and refusing to run any
boot block that doesn't start with the same instruction sequence as
MS-DOG boot sector code. If this is the case, it would be nice if it
actually printed an error message to this effect instead of just
saying 'FAILED.'
This workaround has no effect on the boot sector code other than to
increase its size by three bytes.
code. Apart from the first one, none really affect typical configurations
but are nevertheless unnecessary limitations. We use netbooted PCs as
student X-terminals and all of the below fixes have been useful. Apologies
for including them all in one PR, but some are just too silly or trivial
to send on their own!
a) Newer SMC cards have hardware addresses starting with 00:E0.
Netboot compares the MAC address with 00:00:C0 to determine
if it is a WD/SMC card, so it fails to detect these.
b) Netboot is unable to boot kzipped kernels, as it assumes that
the kernel load address is 0x100000.
c) Users can abort the booting process and enter arbitrary network
addresses, or boot from a floppy disk. This can be a problem when
netbooted machines are used in a student environment.
d) It is not possible to set all options via bootp. For example there
is no way to remotely force a client to boot from disk. With both
SECURE_BOOT(patch below) and NO_TFTP defined, short of unplugging
the eprom there is no way at all to get the client to boot locally.
A generic solution is to allow complete netboot commands to be sent
using bootp lines such as:
:T132="diskboot":
e) The last character of netboot command names is not checked. You
can type 'iz 10.0.0.1' and it will be interpreted as 'ip'. This
is only important if you try to add a new command which is the
same as an existing one except for the last character.
f) We have a configuration where multiple servers are willing to serve
a diskless client. The tftp config file, or the bootptab entry on
each server must specify the root and swap filesystems as 'ip:/fs'
even though 'ip' will usually be the responding server's IP address.
It would be nice if netboot could automatically prepend the server's
IP address to an entry specified as just '/fs', so that multiple
servers can use the same tftp or bootp configuration files. Admittedly
this is hardly a major problem!
PR: 7098
Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
NFS_MINATTRTIMO VREG attrib cache timeout in sec
NFS_MAXATTRTIMO
NFS_MINDIRATTRTIMO VDIR attrib cache timeout in sec
NFS_MAXDIRATTRTIMO
NFS_GATHERDELAY Default write gather delay (msec)
NFS_UIDHASHSIZ Tune the size of nfssvc_sock with this
NFS_WDELAYHASHSIZ and with this
NFS_MUIDHASHSIZ Tune the size of nfsmount with this
NFS_NOSERVER (already documented in LINT)
NFS_DEBUG turn on NFS debugging
also, because NFS_ROOT is used by very different files, it has been
renamed to opt_nfsroot.h instead of the old opt_nfs.h....
end of the main options section.
Turned on documented option OVERRIDE_TUNER. LINT is primarily
for turning on options, not for documenting them.
Don't list IPFILTER twice (once as broken).
mmioctl() to fix hundreds of style bugs and a few error handling bugs
(don't check for superuser privilege for inappropriate ioctls, don't
check the input arg for the output-only MEM_RETURNIRQ ioctl, and don't
return EPERM for null changes).
`void *' arg. Fixed or hid most of the resulting type mismatches.
Handlers can now be updated locally (except for reworking their
global declarations in isa_device.h).
Don't generate declarations for isa interrupt handlers at all.
Isa interrupt handlers are now declared in <i386/isa/isa_device.h>
but should be converted take a `void *' arg and staticized as
soon as possible.
Updated CONFIGVERS. New configs are very incompatible with
previous versions.
and don't depend on them being declared there. This will cause lots of
warnings for a few minutes until config is updated. Interrupt handlers
should never have been configured by config, and the machine generated
declarations get in the way of changing the arg type from int to void *.
over from the probe are now expected for incompatible UARTs that
deliver IRQs as a strobe (low) instead of a level (high).
Discard events on going-away devices too. Endless loops may have
been possible when an active pccard was removed.
Submitted by: Randall Hopper <rhh@ct.picker.com>
The patch supports using the X10 Mouse Remote in both stand-alone and
pass-through configurations, so you can plug your mouse and remote into the
same serial port, use the mouse for X, and use the remote for other apps
like Fxtv. For instance, we can now control fxtv via the remote control
just like a TV : change channels, mute, increase volume, zoom video,
freeze frame 8)
The mouse events are channeled through the syscons/sysmouse I/F like
normal, and the remote buttons are "syphoned off" to a UNIX-domain stream
socket (defined as _PATH_MOUSEREMOTE in <machine/mouse.h>) for a
remote-aware app to grab and use.
For further info on the X10 Mouse Remote see:
http://www.x10.com/products/x10_mk19a.htm
Define an unsigned version of off_t (uoff_t) if KERNEL is defined.
This is useful for micro-optimizing expressions involving off_t's
where you know that the values are non-negative but the compiler
doesn't know it.
Fixed a gratuitous ANSIism.
explicit use of `[unsigned] long long' and `[u]quad_t'.
Fixed disorder of and lack of comments for the types added in the previous
two commits. Improved some comments. The comments should be elswhere
(not duplicated in every MD file).
machine-independent code and try mounting the devices in the
lists instead of guessing alternative root devices in a machine-
dependent way.
autoconf.c:
Reject preposterous slice numbers instead of silently converting
them to COMPATIBILITY_SLICE.
Don't forget to force slice = COMPATIBILITY_SLICE in the floppy
device name.
Eliminated most magic numbers and magic device names in setroot().
Fixed dozens of style bugs.
vfs_conf.c:
Put the actual root device name instead of "root_device" in the
mount struct if the actual name is available. This is useful after
booting with -s. If it were set in all cases then it could be used
to do mount(8)'s ROOTSLICE_HUNT and fsck(8)'s hotroot guess better.
or unsigned int (this doesn't change the struct layout, size or
alignment in any of the files changed in this commit, at least for
gcc on i386's. Using bitfields of type u_char may affect size and
alignment but not packing)).
FreeBSD/alpha. The most significant item is to change the command
argument to ioctl functions from int to u_long. This change brings us
inline with various other BSD versions. Driver writers may like to
use (__FreeBSD_version == 300003) to detect this change.
The prototype FreeBSD/alpha machdep will follow in a couple of days
time.
so that the new behaviour is now default.
Solves the "infinite loop in diversion" problem when more than one diversion
is active.
Man page changes follow.
The new code is in -stable as the NON default option.
miscconfigured case) if the port is the console. This fixes several
bugs:
- if all sioprobe()s failed, then the console driver followed null
pointers in cdevsw[].
- if the sioprobe() for the console failed but another sioprobe()
succeeded, then init hung early when the console couldn't be
opened.
- it was silly for the console to not be there after printing boot
messages on it.
Bugs introduced by this are hopefully no worse than old ones caused
by forcing the success of the `cn' level probe.
Only complain about an irq mismatch in the probe if the configured
irq doesn't become active, and then print the bitmap of irqs that
became active (including clock irqs) instead of just the first
(not including clock irqs).
Bugs reported by: msmith
probe and intialisation. This will ultimately remove the grubby (but
functional) hack that copies a real-mode function into low memory
early in locore.s.
layer does not like the null shmid_ds buffer pointer. The emulation layer
returned an error without ever calling FreeBSD's shmctl, so the segments
were not being deleted when the reference count went to zero."
Submitted by: Kevin Street <street@iname.com>
a test of the irq number, and made failure of this test non-fatal.
Removed related unused complications for the APIC_IO case. Removed the
no-test3 flag.
Deverbosified the failure messages for the other tests. Removed the
per-port verbose flag - just use the general verbose flag.
Clean up (or if antipodic: down) some of the msgbuf stuff.
Use an inline function rather than a macro for timecounter delta.
Maintain process "on-cpu" time as 64 bits of microseconds to avoid
needless second rollover overhead.
Avoid calling microuptime the second time in mi_switch() if we do
not pass through _idle in cpu_switch()
This should reduce our context-switch overhead a bit, in particular
on pre-P5 and SMP systems.
WARNING: Programs which muck about with struct proc in userland
will have to be fixed.
Reviewed, but found imperfect by: bde
i contains the contents of the EP_W0_CONFIG_CTRL register.
i was being used as the array index into an array on the stack.
j is initialized to 0 as it should be.
PR: kern/6757
Reviewed by: jmb
Submitted by: Stephane E. Potvin <sepotvin@videotron.ca>
unexpectedly do not complete writes even with sync I/O requests.
This should help the behavior of mmaped files when using
softupdates (and perhaps in other circumstances also.)
Althought the comments say the datasheet doesn't list the device ID
registers on the M2/MX, they seem to be there and quite alive.
(It's interesting to note that the M2/MX calls itself a 686 class cpu but
is missing a heck of a lot of features, including VME, PGE, PSE, etc)
available. The per-cpu variable ss_tpr has been replaced by ss_eflags.
This reduced the number of interrupts sent to the wrong CPU, due to
the cpu having the global lock being inside a critical region.
Remove some unneeded manipulation of tpr register in mplock.s.
Adjust code in mplock.s to be aware of variables on the stack being
destroyed by MPgetlock if GRAB_LOPRIO is defined.
update of cpu usage as shown by top when one process is cpu bound
(no system calls) while the system is otherwise idle (except for top).
Don't attempt to switch to the BSP in boot(). If the system was idle when
an interrupt caused a panic, this won't work. Instead, switch to the BSP
in cpu_reset.
Remove some spurious forward_statclock/forward_hardclock warnings.
Windows 95 after rebooting FreeBSD without power off. In PC-98
system, reboot mode is set via I/O port 0x37 in cpu_reset(), and
accessing of this port is the reason of the problem. To avnoid the
fault, current status of reboot mode should be checked before
accessing the I/O port.
that local APIC should be disabled in UP system. However, some of old
BIOS does not disable local APIC, and virtual wire mode through local
APIC may cause int 15.
submitted ioctl to clear the video buffer
prior to starting video capture
Amancio : clean up yuv12 so that it does not
affect rgb capture. Basically, fxtv after
capturing in yuv12 mode , switching to rgb
would cause the video capture to be too bright.
1.32 disable inverse gamma function for rgb and yuv
capture. fixed meteor brightness ioctl it now
converts the brightness value from unsigned to
signed.
1.33 added sysctl: hw.bt848.tuner, hw.bt848.reverse_mute,
hw.bt848.card
card takes a value from 0 to bt848_max_card
tuner takes a value from 0 to bt848_max_tuner
reverse_mute : 0 no effect, 1 reverse tuner
mute function some tuners are wired reversed :(
but doesn't do much of anything with it. I added it to siopnp_ids[]
and it was found and recognized as a serial port.
PR: 6605
Reviewed by: phk
Submitted by: Dave Marquardt <marquard@zilker.net>
This is a result of discussions on the mailing lists. Kudos to those who
have found the issue and created work-arounds. I have chosen Tor's fix
for now, before we can all work the issue more completely.
Submitted by: Tor Egge
possibly non-open devices, and we don't want to restrict dumping
to swap devices anwyay. It is especially invalid to call d_ioctl()
in non-process context for panics. d_psize() can be called on
non-open devices, at least on non-SLICED ones that support d_dump(),
and setdumpdev() has depended on this for a long time although it
is probably wrong, but even d_psize() can't be called in non-process
context - that's why dumpsys() depends on previously computed values
although these values may be stale. The historical restriction to
devices with dkpart(dev) == SWAP_PART should go away.
Support >8G drives in CHS mode. This is done by guesstimating the
cylinder count from the LBA size reported. It works on my shiny
new Maxtor 11.5G drive, YMMV.
Reports from users of other big drives (read Quantum bigfoot's)
are welcome...
Linux emulation. This make Allegro Common Lisp 4.3 work under
FreeBSD!
Submitted by: Fred Gilham <gilham@csl.sri.com>
Commented on by: bde, dg, msmith, tg
Hoping he got everything right: eivind
This code will be turned on with the TWO options
DEVFS and SLICE. (see LINT)
Two labels PRE_DEVFS_SLICE and POST_DEVFS_SLICE will deliniate these changes.
/dev will be automatically mounted by init (thanks phk)
on bootup. See /sys/dev/slice/slice.4 for more info.
All code should act the same without these options enabled.
Mike Smith, Poul Henning Kamp, Soeren, and a few dozen others
This code does not support the following:
bad144 handling.
Persistance. (My head is still hurting from the last time we discussed this)
ATAPI flopies are not handled by the SLICE code yet.
When this code is running, all major numbers are arbitrary and COULD
be dynamically assigned. (this is not done, for POLA only)
Minor numbers for disk slices ARE arbitray and dynamically assigned.
is that the previous commit spammed a hacked 2.2-stable onto -current,
deleting the DMA support etc. (I guess that's one way of minimizing diffs
between -current and -stable.. :-] )
Submitted by: Roger Hardiman <roger@cs.strath.ac.uk>
Roger Hardiman <roger@cs.strath.ac.uk> :
Revised autodetection code to correctly handle both
old and new VideoLogic Captivator PCI cards.
Added tsleep of 2 seconds to initialistion code for PAL users.
Corrected clock selection code on format change.
--- Amancio
device. But with devfs, currently, /dev/psm0 is the blocking device
and /dev/npsm0 is the non-blocking one.
DEVFS must stay consistent with the older behaviour.
PR: 6260
Reviewed by: phk
Submitted by: Kapil Chowksey <kchowksey@hss.hns.com>
is a tremendous perf decrease due to the disabling of advanced
features such as DMA, Ultra DMA, and 32bit mode. This patch
might have been reported by someone else (I seem to remember
it.)
Use config flags 0x1000 to enable LBA mode. It should be enabled in
the BIOS too to avoid geometry confusion.
One catch though, I'm not sure all BIOS's uses the 64head/63secs
translation, all mine does but....
ints. Remove some no longer needed casts. Initialize the per-cpu
global data area using the structs rather than knowing too much about
layout, alignment, etc.
- Set the correct value scp->font_size in init_scp().
- Set scp->font_size to FONT_NONE for VGA_MODEX.
Interim fix for a font problem:
- A kludge to display the correct font on some video cards.
We should be able to load multiple fonts to the VGA plane #2 and switch
between fonts by setting the font select register in the VGA sequencer.
It appears that the current code isn't functioning as expected on
some VGA cards (I have reports on Millenium and Mach64 cards). This is
either a bug in syscons or a hardware compatibility problem ;-<
This kludge will always load only one font set at a time and always use
the font page #0 on the plane #2. It is an interim kludge until
we find the exact cause and solution.
Small adjustment for mouse cursor handling:
- Turn off the mouse cursor early when changing video modes.
Video mode switch fixes:
- Stop the screen saver when changing video modes.
- Enclose the critical section with a pair of spltty()/splx().
- A kludge to prevent scrn_update() from accessing video memory in less-
critical sections in video mode change; artificially turn on the
UNKNOWN_MODE flag.
PR: bin/5899, bin/5907
Tested by: ache and a couple of users
OKed by: sos
* Figure out UTC relative to boottime. Four new functions provide
time relative to boottime.
* move "runtime" into struct proc. This helps fix the calcru()
problem in SMP.
* kill mono_time.
* add timespec{add|sub|cmp} macros to time.h. (XXX: These may change!)
* nanosleep, select & poll takes long sleeps one day at a time
Reviewed by: bde
Tested by: ache and others
- Attempt to handle PCI devices where the interrupt is
an ISA/EISA interrupt according to the mp table.
- Attempt to handle multiple IO APIC pins connected to
the same PCI or ISA/EISA interrupt source. Print a
warning if this happens, since performance is suboptimal.
This workaround is only used for PCI devices.
With these two workarounds, the -SMP kernel is capable of running on
my Asus P/I-P65UP5 motherboard when version 1.4 of the MP table is disabled.
"time" wasn't a atomic variable, so splfoo() protection were needed
around any access to it, unless you just wanted the seconds part.
Most uses of time.tv_sec now uses the new variable time_second instead.
gettime() changed to getmicrotime(0.
Remove a couple of unneeded splfoo() protections, the new getmicrotime()
is atomic, (until Bruce sets a breakpoint in it).
A couple of places needed random data, so use read_random() instead
of mucking about with time which isn't random.
Add a new nfs_curusec() function.
Mark a couple of bogosities involving the now disappeard time variable.
Update ffs_update() to avoid the weird "== &time" checks, by fixing the
one remaining call that passwd &time as args.
Change profiling in ncr.c to use ticks instead of time. Resolution is
the same.
Add new function "tvtohz()" to avoid the bogus "splfoo(), add time, call
hzto() which subtracts time" sequences.
Reviewed by: bde
_KPOSIX_PRIORITY_SCHEDULING options to work. Changes:
Change all "posix4" to "p1003_1b". Misnamed files are left
as "posix4" until I'm told if I can simply delete them and add
new ones;
Add _POSIX_PRIORITY_SCHEDULING system calls for FreeBSD and Linux;
Add man pages for _POSIX_PRIORITY_SCHEDULING system calls;
Add options to LINT;
Minor fixes to P1003_1B code during testing.
They are atomic, but return in essence what is in the "time" variable.
gettime() is now a macro front for getmicrotime().
Various patches to use the two new functions instead of the various
hacks used in their absence.
Some puntuation and grammer patches from Bruce.
A couple of XXX comments.
everything is contained inside #ifdef VM86, so this option must be
present in the config file to use this functionality.
Thanks to Tor Egge, these changes should work on SMP machines. However,
it may not be throughly SMP-safe.
Currently, the only BIOS calls made are memory-sizing routines at bootup,
these replace reading the RTC values.
- Implement proper EISA probing.
- Better support for the new transputer based host cards.
- use standard termios settings, one can use the intial/lock devices.
- use a simple bcopy since some cards/systems apparently don't support
32 bit accesses.
- hard reset and halt host card CPU prior to download in case of a soft
restart.
- recognize new remote module types (ASIC vs. CD1400 based)
- a number of cosmetic changes (my fault, not Nick's)
Submitted by: Nick Sayer <nsayer@quack.kfu.com>
Don't try to flush buffers if the drive says it has none.
More error checking and reporting.
Hack: if drive hangs, it can be reset by issuing a mt -f device offline.
I've been able to make several 4G backups. However there is still problems
with some configurations. It is not clear if it is hardware or driver
problems yet.
problems. Tor Egge and others have helped with various VM bugs
lately, but don't blame him -- blame me!!!
pmap.c:
1) Create an object for kernel page table allocations. This
fixes a bogus allocation method previously used for such, by
grabbing pages from the kernel object, using bogus pindexes.
(This was a code cleanup, and perhaps a minor system stability
issue.)
pmap.c:
2) Pre-set the modify and accessed bits when prudent. This will
decrease bus traffic under certain circumstances.
vfs_bio.c, vfs_cluster.c:
3) Rather than calculating the beginning virtual byte offset
multiple times, stick the offset into the buffer header, so
that the calculated offset can be reused. (Long long multiplies
are often expensive, and this is a probably unmeasurable performance
improvement, and code cleanup.)
vfs_bio.c:
4) Handle write recursion more intelligently (but not perfectly) so
that it is less likely to cause a system panic, and is also
much more robust.
vfs_bio.c:
5) getblk incorrectly wrote out blocks that are incorrectly sized.
The problem is fixed, and writes blocks out ONLY when B_DELWRI
is true.
vfs_bio.c:
6) Check that already constituted buffers have fully valid pages. If
not, then make sure that the B_CACHE bit is not set. (This was
a major source of Sig-11 type problems.)
vfs_bio.c:
7) Fix a potential system deadlock due to an incorrectly specified
sleep priority while waiting for a buffer write operation. The
change that I made opens the system up to serious problems, and
we need to examine the issue of process sleep priorities.
vfs_cluster.c, vfs_bio.c:
8) Make clustered reads work more correctly (and more completely)
when buffers are already constituted, but not fully valid.
(This was another system reliability issue.)
vfs_subr.c, ffs_inode.c:
9) Create a vtruncbuf function, which is used by filesystems that
can truncate files. The vinvalbuf forced a file sync type operation,
while vtruncbuf only invalidates the buffers past the new end of file,
and also invalidates the appropriate pages. (This was a system reliabiliy
and performance issue.)
10) Modify FFS to use vtruncbuf.
vm_object.c:
11) Make the object rundown mechanism for OBJT_VNODE type objects work
more correctly. Included in that fix, create pager entries for
the OBJT_DEAD pager type, so that paging requests that might slip
in during race conditions are properly handled. (This was a system
reliability issue.)
vm_page.c:
12) Make some of the page validation routines be a little less picky
about arguments passed to them. Also, support page invalidation
change the object generation count so that we handle generation
counts a little more robustly.
vm_pageout.c:
13) Further reduce pageout daemon activity when the system doesn't
need help from it. There should be no additional performance
decrease even when the pageout daemon is running. (This was
a significant performance issue.)
vnode_pager.c:
14) Teach the vnode pager to handle race conditions during vnode
deallocations.
on the IOAPIC being connected to the 8254 timer interrupt.
Verify that timer interrupts are delivered. If they aren't, attempt
a fallback to mixed mode (i.e. routing the timer interrupt via the 8259 PIC).
Gee, I wish there was a better way to run makesyscalls.sh so that
a make world finds missing things like this. Running it manually
sucks.
Pointed out by: Peter Dufault
previous commit. Opportunities to clean pages were often missed,
and leaving of the idle state was sometimes delayed until the next
interrupt (after any that occurred while cleaning).
Fixed an unstaticization, a syntax error and a style bug in the
previous commit.
slice number passed in by the bootblocks. This means the kernel will
not use the compatability slice to obtain the root filesystem when
booting from a sliced disk.
Use the extraction macros from reboot.h rather than stating them in full
again.
has been some bitrot and incorrect assumptions in the vfs_bio code. These
problems have manifest themselves worse on NFS type filesystems, but can
still affect local filesystems under certain circumstances. Most of
the problems have involved mmap consistancy, and as a side-effect broke
the vfs.ioopt code. This code might have been committed seperately, but
almost everything is interrelated.
1) Allow (pmap_object_init_pt) prefaulting of buffer-busy pages that
are fully valid.
2) Rather than deactivating erroneously read initial (header) pages in
kern_exec, we now free them.
3) Fix the rundown of non-VMIO buffers that are in an inconsistent
(missing vp) state.
4) Fix the disassociation of pages from buffers in brelse. The previous
code had rotted and was faulty in a couple of important circumstances.
5) Remove a gratuitious buffer wakeup in vfs_vmio_release.
6) Remove a crufty and currently unused cluster mechanism for VBLK
files in vfs_bio_awrite. When the code is functional, I'll add back
a cleaner version.
7) The page busy count wakeups assocated with the buffer cache usage were
incorrectly cleaned up in a previous commit by me. Revert to the
original, correct version, but with a cleaner implementation.
8) The cluster read code now tries to keep data associated with buffers
more aggressively (without breaking the heuristics) when it is presumed
that the read data (buffers) will be soon needed.
9) Change to filesystem lockmgr locks so that they use LK_NOPAUSE. The
delay loop waiting is not useful for filesystem locks, due to the
length of the time intervals.
10) Correct and clean-up spec_getpages.
11) Implement a fully functional nfs_getpages, nfs_putpages.
12) Fix nfs_write so that modifications are coherent with the NFS data on
the server disk (at least as well as NFS seems to allow.)
13) Properly support MS_INVALIDATE on NFS.
14) Properly pass down MS_INVALIDATE to lower levels of the VM code from
vm_map_clean.
15) Better support the notion of pages being busy but valid, so that
fewer in-transit waits occur. (use p->busy more for pageouts instead
of PG_BUSY.) Since the page is fully valid, it is still usable for
reads.
16) It is possible (in error) for cached pages to be busy. Make the
page allocation code handle that case correctly. (It should probably
be a printf or panic, but I want the system to handle coding errors
robustly. I'll probably add a printf.)
17) Correct the design and usage of vm_page_sleep. It didn't handle
consistancy problems very well, so make the design a little less
lofty. After vm_page_sleep, if it ever blocked, it is still important
to relookup the page (if the object generation count changed), and
verify it's status (always.)
18) In vm_pageout.c, vm_pageout_clean had rotted, so clean that up.
19) Push the page busy for writes and VM_PROT_READ into vm_pageout_flush.
20) Fix vm_pager_put_pages and it's descendents to support an int flag
instead of a boolean, so that we can pass down the invalidate bit.
f00f_hack has run.
Use the global r_idt descriptor in f00f_hack when in SMP mode,
so the APs find the relocated interrupt descriptor table.
Submitted by: Partially from David A Adkins <adkin003@tc.umn.edu>
interrupts are masked, and EOI is sent iff the corresponding ISR bit
is set in the local apic. If the CPU cannot obtain the interrupt
service lock (currently the global kernel lock) the interrupt is
forwarded to the CPU holding that lock.
Clock interrupts now have higher priority than other slow interrupts.
the signal handling latency for cpu-bound processes that performs very
few system calls.
The IPI for forcing an additional software trap is no longer dependent upon
BETTER_CLOCK being defined.
than rolling it's own. This means that it now uses the "safe"
exec_map_first_page() to get the ld.so headers rather than risking a panic
on a page fault failure (eg: NFS server goes down).
Since all the ELF tools go to a lot of trouble to make sure everything
lives in the first page for executables, this is a win. I have not seen
any ELF executable on any system where all the headers didn't fit in the
first page with lots of room to spare.
I have been running variations of this code for some time on my pure ELF
systems.
It does endeed work, but there is still some problems to solve.
I get a "nonrecovered data error" from time to time, but besides
this it has backed up several Gigs allready.
Please report any success/failure directly to me.
Thanks to Warner Losh for providing a drive to use in writing
this driver!
2) Do not unnecessarily force page blocking when paging
pages out.
3) Further improve swap pager performance and correctness,
including fixing the paging in progress deadlock (except
in severe I/O error conditions.)
4) Enable vfs_ioopt=1 as a default.
5) Fix and enable the page prezeroing in SMP mode.
All in all, SMP systems especially should show a significant
improvement in "snappyness."
it runs at a constant frequency. This was less of an issue before,
because the TSC only interpolated in the HZ intervals, but now where
the timecounter is used all the way, this becomes much more visible.
Nit: Fix a printf which triggered the bde-filter.
in a way identically as before.) I had problems with the system properly
handling the number of vnodes when there is alot of system memory, and the
default VM_KMEM_SIZE. Two new options "VM_KMEM_SIZE_SCALE" and
"VM_KMEM_SIZE_MAX" have been added to support better auto-sizing for systems
with greater than 128MB.
Add some accouting for vm_zone memory allocations, and provide properly
for vm_zone allocations out of the kmem_map. Also move the vm_zone
allocation stats to the VM OID tree from the KERN OID tree.
signals delivered to a process writing to the audio device the
system: if you try
cat /dev/zero > /dev/dsp (or cat /dev/zero > /dev/pcaudio)
and press Ctrl-C : for a second or two the system appears to freeze
(e.g. the cursor will disappear if you move the mouse, xclock
blocks, etc.). I think that interrupts etc. still run so the problem
is not too terrible, but still annoying
[ The problems appears to also be in isa/pcaudio.c, though that is ignored ]
Submitted by: Luigi Rizzo <luigi@labinfo.iet.unipi.it>
Highlights:
* Simple model for underlying hardware.
* Hardware basis for timekeeping can be changed on the fly.
* Only one hardware clock responsible for TOD keeping.
* Provides a real nanotime() function.
* Time granularity: .232E-18 seconds.
* Frequency granularity: .238E-12 s/s
* Frequency adjustment is continuous in time.
* Less overhead for frequency adjustment.
* Improves xntpd performance.
Reviewed by: bde, bde, bde
Uncommented css0. It compiles OK.
Moved trix0 so that it compiles OK when uncommented. Uncommented
it. Drivers with the same interrupt handler must be together in
config files so that config(8)'s simple avoidance of redundant
declarations of interrupt handlers works (config emits a declaration
unless it would duplicate the previous one).
Commented out NO_LKM. Negative options should not be configured
in LINT. There should be no negative options for subsystems.
LKMs should never have been standard or the default.
