sos [Mon, 5 Dec 2005 17:33:57 +0000 (17:33 +0000)]
Add support for writing Intel MatrixRAID arrays.
Do a little better on handling volumes as well, however we cant create
multiple volumes from FreeBSD yet.
rwatson [Mon, 5 Dec 2005 13:04:00 +0000 (13:04 +0000)]
Add a simple KVM tool to dump the kernel state of UMA, including walking
the keg/zone lists, summarizing cache state, and walking bucket lists in
each zone. I seem to get inconsistent results on SMP, possibly due to
local header problems, but it seems to work quite well on UP. This tool
requires sufficient privilege to read /dev/mem (or a core dump), and is
for debugging purposes rather than administrative monitoring purposes
(use vmstat instead).
marius [Mon, 5 Dec 2005 12:32:20 +0000 (12:32 +0000)]
- Rev. 1.175 fixed compilation on sparc64 but also backed out zeroing of
the eaddr array (introduced in rev. 1.174) prior to writing to it. As
dc_read_eeprom() is told to write only 3 16-bit words to eaddr but eaddr
in fact is somewhat larger removal of the zeroing defeated the check
whether the MAC address is all zero as there can be some random garbage
in eaddr past the 3 words written to it and the check verifys all bits
in eaddr. Solve this by changing the check to verify only the 3 words
(happenning to be ETHER_ADDR_LEN bytes) written to eaddr.
- While here change the notation of "FCode" in a nearby comment to the
official way.
dougb [Mon, 5 Dec 2005 07:04:15 +0000 (07:04 +0000)]
Change how *.sh scripts are handled. If the script is in /etc/rc.d,
source it into the shell. If not, handle it in a subshell the same
way that "real" rc.d-style scripts are handled. This will dramatically
ease the "process local scripts in the base rcorder" transition.
Add *.bak to the list of files in */rc.d that we ignore.
davidxu [Mon, 5 Dec 2005 03:23:27 +0000 (03:23 +0000)]
Handle SIGEV_NONE, if notification is SIGEV_NONE, error status and
return status will be set, but no notification will be registered.
Increase hard limit of maxmsg to 100, so posixtestsuite ports can run.
stefanf [Sun, 4 Dec 2005 18:44:21 +0000 (18:44 +0000)]
Add the times builtin. It reports the user and system time for the shell
itself and its children. Instead of calling times() (as implied by POSIX) this
implementation directly calls getrusage() to get the times because this is more
convenient.
bde [Sun, 4 Dec 2005 13:52:46 +0000 (13:52 +0000)]
Fixed the approximation to pio4. pio4_hi must be pio2_hi/2 since it
shares its low half with pio2_hi. pio2_hi is rounded down although
rounding to nearest would be a tiny bit better, so pio4_hi must be
rounded down too. It was rounded to nearest, which happens to be
different in float precision but the same in double precision.
This fixes about 13.5 million errors of more than 1 ulp in asinf().
The largest error was 2.81 ulps on amd64 and 2.57 ulps on i386 -O1.
Now the largest error is 0.93 ulps on amd65 and 0.67 ulps on i386 -O1.
bde [Sun, 4 Dec 2005 12:30:44 +0000 (12:30 +0000)]
For log1pf(), fixed the approximations to sqrt(2), sqrt(2)-1 and
sqrt(2)/2-1. For log1p(), fixed the approximation to sqrt(2)/2-1.
The end result is to fix an error of 1.293 ulps in
log1pf(0.41421395540 (hex 0x3ed413da))
and an error of 1.783 ulps in
log1p(-0.292893409729003961761) (hex 0x12bec4 00000001)).
The former was the only error of > 1 ulp for log1pf() and the latter
is the only such error that I know of for log1p().
The approximations don't need to be very accurate, but the last 2 need
to be related to the first and be rounded up a little (even more than
1 ulp for sqrt(2)/2-1) for the following implementation-detail reason:
when the arg (x) is not between (the approximations to) sqrt(2)/2-1
and sqrt(2)-1, we commit to using a correction term, but we only
actually use it if 1+x is between sqrt(2)/2 and sqrt(2) according to
the first approximation. Thus we must ensure that
!(sqrt(2)/2-1 < x < sqrt(2)-1) implies !(sqrt(2)/2 < x+1 < sqrt(2)),
where all the sqrt(2)'s are really slightly different approximations
to sqrt(2) and some of the "<"'s are really "<="'s. This was not done.
In log1pf(), the last 2 approximations were rounded up by about 6 ulps
more than needed relative to a good approximation to sqrt(2), but the
actual approximation to sqrt(2) was off by 3 ulps. The approximation
to sqrt(2)-1 ended up being 4 ulps too small, so the algoritm was
broken in 4 cases. The result happened to be broken in 1 case. This
is fixed by using a natural approximation to sqrt(2) and derived
approximations for the others.
In logf(), all the approximations made sense, but the approximation
to sqrt(2)/2-1 was 2 ulps too small (a tiny amount, since we compare
with a granularity of 2**32 ulps), so the algorithm was broken in 2
cases. The result was broken in 1 case. This is fixed by rounding
up the approximation to sqrt(2)/2-1 by 2**32 ulps, so 2**32 cases are
now handled a little differently (still correctly according to my
assertion that the approximations don't need to be very accurate, but
this has not been checked).
There no longer seems to be this bug in gcc(1). Remove the
badly implemented workaround that caused a workaround to be
applied to all architectures, not only amd64.
stefanf [Sun, 4 Dec 2005 09:34:56 +0000 (09:34 +0000)]
Merge NetBSD's revision 1.27. This bug can be observed eg. when browsing
through the history in sh.
| Refresh bug reported by Julien Torres:
|
| going from:
| activate -verbose
| to:
| reset -activation
| results in:
| reset -activationverbose"
| instead of:
| reset -activation
|
| This is because we choose to insert "reset -" before the current line,
| and the delete "e -" and insert "ion" in the appropriate place. The
| cleareol code did not handle this case properly; we now cleareol to
| the maximum number of characters of the first difference, the second
| difference and the difference in line length.
bde [Sun, 4 Dec 2005 08:57:54 +0000 (08:57 +0000)]
Use the usual volatile hack to trick gcc into clipping any extra precision
on assignment.
Extra precision on i386's broke hi+lo decomposition in the usual way.
It caused all except 1 of the 62343 errors of more than 1 ulp for
log1pf() on i386's with gcc -O [-fno-float-store].
sam [Sun, 4 Dec 2005 04:50:27 +0000 (04:50 +0000)]
Adhoc mode fixups:
o plug memory leak in adhoc mode: on rx the sender may be the
current master so simply checking against ic_bss is not enough
to identify if the packet comes from an unknown sender; must
also check the mac address
o split neighbor node creation into two routines and fillin state
of nodes faked up on xmit when a beacon or probe response frame
is later received; this ensures important state like the rate set
and advertised capabilities are correct
philip [Sun, 4 Dec 2005 00:28:40 +0000 (00:28 +0000)]
Add "yet another" moused(8) hack: suspend handling mouse events when SIGUSR1
is caught. Can be assigned to a window manager shortcut to prevent accidents
with touchpads.
jhb [Sat, 3 Dec 2005 21:17:17 +0000 (21:17 +0000)]
Force any hardwire-routed interrupts to level trigger and active low
polarity. Some machines route PCI IRQs to an ISA IRQ but fail to include
an interrupt override entry to set the polarity and trigger of the given
ISA IRQ in their MADT table.
PR: usb/74989
Reported by: Julien Gabel jpeg at thilelli dot net
MFC after: 1 week
marius [Sat, 3 Dec 2005 19:52:20 +0000 (19:52 +0000)]
- Move the declaration of struct upa_ranges and the UPA_RANGE_* macros
from sys/sparc64/include/ofw_upa.h to sys/sparc64/pci/ofw_pci.h and
rename them to struct ofw_pci_ranges and OFW_PCI_RANGE_* respectively.
This ranges struct only applies to host-PCI bridges but no to other
bridges found on UPA. At the same time it applies to all host-PCI
bridges regardless of whether the interconnection bus is Fireplane/
Safari, JBus or UPA.
- While here rename the PCI_CS_* macros in sys/sparc64/pci/ofw_pci.h
to OFW_PCI_CS_* in order to be consistent and change this header to
use uintXX_t instead of u_intXX_t.
philip [Sat, 3 Dec 2005 17:32:39 +0000 (17:32 +0000)]
Add [-J jid_file] option to write out a JidFile, similar to a PidFile,
containing the jailid, path, hostname, ip and the command used to start
the jail.
PR: misc/89883
Submitted by: L. Jason Godsey <lannygodsey -at- yahoo.com>
Reviewed by: phk
MFC after: 1 week
marius [Sat, 3 Dec 2005 16:36:54 +0000 (16:36 +0000)]
- Register the PCI bus error interrupt handler according to which half of
the bridge (PCI bus A or B) we are attaching to rather than registering
both handlers at once when attaching to the first half we encounter.
This is a bit cleaner as it corresponds to which PCI bus error interrupt
actually is assigned to the respective half by the OFW and allows to
collapse both PCI bus error interrupt handlers into one function easily.
- Use the actual RID of the respective interrupt resource as index into
sc_irq_res and also use it when allocating the resource. For now this
is a bit cleaner and will be mandatory later on.
- According to OpenSolaris the spare hardware interrupt is used as the
over-temperature interrupt in systems with Psycho bridges. Unlike as
with the SBus-based workstations I didn't manage to trigger it when
covering the fan outlets of an U60 but better be safe than sorry and
register a handler anyway.
marius [Sat, 3 Dec 2005 13:08:05 +0000 (13:08 +0000)]
- Improve the comment regarding the workaround for the E250 interrupt map
bug by explaining what the problem is and how the workaround works.
- Fix some cosmetics nits, mainly properly terminate sentences in comments,
which I missed when backporting the style changes to psycho(4) in psycho.c
rev. 1.54 due to lack of corresponding code.
- The "USIIe version of the Sabre bridge" actually is termed "Hummingbird";
name it as such in comments and messages.
bde [Sat, 3 Dec 2005 11:57:19 +0000 (11:57 +0000)]
Fixed fdlibm[+cygnus] logbf() and logb() on denormals. Adjustment
according to the highest nonzero bit in a denormal was missing.
fdlibm ilogbf() and ilogb() have always had the adjustment, but only
use a small part of their method for handling denormals; use the
normalization method in log[f]() for the main part.
bde [Sat, 3 Dec 2005 09:00:29 +0000 (09:00 +0000)]
Restored removal of the special handling needed for a result of +-0.
It was lost in rev.1.9. The log message for rev.1.9 says that the
special case of +-0 is handled twice, but it was only handled once,
so it became unhandled, and this happened to break half of the cases
that return +-0:
- round-towards-minus-infinity: 0 < x < 1: result was -0 not 0
- round-to-nearest: -0.5 <= x < 0: result was 0 not -0
- round-towards-plus-infinity: -1 < x < 0: result was 0 not -0
- round-towards-zero: -1 < x < 0: result was 0 not -0
bde [Sat, 3 Dec 2005 07:38:35 +0000 (07:38 +0000)]
Simplified the fix in rev.1.3. Instead of using long double for
TWO52[sx] to trick gcc into correctly converting TWO52[sx]+x to double
on assignment to "double w", force a correct assignment by assigning
to *(double *)&w. This is cleaner and avoids the double rounding
problem on machines that evaluate double expressions in double
precision. It is not necessary to convert w-TWO52[sx] to double
precision on return as implied in the comment in rev.1.3, since
the difference is exact.
bde [Sat, 3 Dec 2005 07:23:30 +0000 (07:23 +0000)]
Fixed rint(x) in the following cases:
(1) In round-to-nearest mode, on all machines, fdlibm rint() never
worked for |x| = n+0.75 where n is an even integer between 262144
and 524286 inclusive (2*131072 cases). To avoid double rounding
on some machines, we begin by adjusting x to a value with the 0.25
bit not set, essentially by moving the 0.25 bit to a lower bit
where it works well enough as a guard, but we botched the adjustment
when log2(|x|) == 18 (2*2**52 cases) and ended up just clearing
the 0.25 bit then. Most subcases still worked accidentally since
another lower bit serves as a guard. The case of odd n worked
accidentally because the rounding goes the right way then. However,
for even n, after mangling n+0.75 to 0.5, rounding gives n but the
correct result is n+1.
(2) In round-towards-minus-infinity mode, on all machines, fdlibm rint()
never for x = n+0.25 where n is any integer between -524287 and
-262144 inclusive (262144 cases). In these cases, after mangling
n+0.25 to n, rounding gives n but the correct result is n-1.
(3) In round-towards-plus-infinity mode, on all machines, fdlibm rint()
never for x = n+0.25 where n is any integer between 262144 and
524287 inclusive (262144 cases). In these cases, after mangling
n+0.25 to n, rounding gives n but the correct result is n+1.
A variant of this bug was fixed for the float case in rev.1.9 of s_rintf.c,
but the analysis there is incomplete (it only mentions (1)) and the fix
is buggy.
Example of the problem with double rounding: rint(1.375) on a machine
which evaluates double expressions with just 1 bit of extra precision
and is in round-to-nearest mode. We evaluate the result using
(double)(2**52 + 1.375) - 2**52. Evaluating 2**52 + 1.375 in (53+1) bit
prcision gives 2**52 + 1.5 (first rounding). (Second) rounding of this
to double gives 2**52 + 2.0. Subtracting 2**52 from this gives 2.0 but
we want 1.0. Evaluating 2**52 + 1.375 in double precision would have
given the desired intermediate result of 2**52 + 1.0.
The double rounding problem is relatively rare, so the botched adjustment
can be fixed for most machines by removing the entire adjustment. This
would be a wrong fix (using it is 1 of the bugs in rev.1.9 of s_rintf.c)
since fdlibm is supposed to be generic, but it works in the following cases:
- on all machines that evaluate double expressions in double precision,
provided either long double has the same precision as double (alpha,
and i386's with precision forced to double) or my earlier fix to use
a long double 2**52 is modified to avoid using long double precision.
- on all machines that evaluate double expressions in many more than 11
bits of extra precision. The 1 bit of extra precision in the example
is the worst case. With N bits of extra precision, it sufices to
adjust the bit N bits below the 0.5 bit. For N >= about 52 there is
no such bit so the adjustment is both impossible and unnecessary. The
fix in rev.1.9 of s_rintf.c apparently depends on corresponding magic
in float precision: on all supported machines N is either 0 or >= 24,
so double rounding doesn't occur in practice.
- on all machines that don't use fdlibm rint*() (i386's).
So under FreeBSD, the double rounding problem only affects amd64 now, but
should only affect i386 in future (when double expressions are evaluated
in long double precision).
ambrisko [Sat, 3 Dec 2005 05:11:07 +0000 (05:11 +0000)]
Switch BUILD_ARCH in Makefile to use uname -p suggested by ru.
Switch strncpy to strlcpy suggested by gad and issue found by pjd.
Add to uname(3) man page describing:
UNAME_s
UNAME_r
UNAME_v
UNAME_m
Add to getosreldate(3) man page describing:
OSVERSION
anholt [Fri, 2 Dec 2005 23:51:36 +0000 (23:51 +0000)]
Add support for i915 GMCH AGP. This diff is a combination of work by myself
and some fixes from Motomichi Matsuzaki. Testing involved many people, but the
final, successful testing was from rwatson who endured several rounds of "it
crashes at XYZ stage" "oh, please correct this typo and try again." The Linux
driver, and to a small extent the limited specs, were both used as a reference
for how to program the chipset.
grehan [Fri, 2 Dec 2005 22:36:14 +0000 (22:36 +0000)]
Register definitions for the ancient via6522. This 20+ year-old chip
still exists as a cell in the Macio asic on Apples, and is used to communicate
through the shift register with the external PMU microcontroller.
"-o rw" is invalid and undocumented mount option that
is only present for fstab(5) compatibility, and is
otherwise ignored by mount(8) (not passed to mount_*
programs, and not passed to nmount(2)).
"-u -o rw" worked with an old mount(8) with mount_ufs.c
because "-o rw" was stripped and simple "-u" caused an
update of UFS from read-only to read-write, due to
inability of mount(2) to track changes in options
(MNT_RDONLY is either set or not).
"-u" no longer causes the transition from RO to RW,
now that mount(8) was converted to use nmount(2), so
an explicit change to RW is required. Keep up with
this change, and use "-uw" to mount root read-write.
dougb [Fri, 2 Dec 2005 20:06:07 +0000 (20:06 +0000)]
Introduce startup scripts from the local_startup directories to
the base rcorder. This is accomplished by running rcorder twice,
first to get all the disks mounted (through mountcritremote),
then again to include the local_startup directories.
This dramatically changes the behavior of rc.d/localpkg, as
all "local" scripts that have the new rc.d semantics are now
run in the base rcorder, so only scripts that have not been
converted yet will run in rc.d/localpkg.
Make a similar change in rc.shutdown, and add some functions in
rc.subr to support these changes.
alc [Fri, 2 Dec 2005 18:02:54 +0000 (18:02 +0000)]
Change pmap_enter_quick() to use the vm_prot_t parameter introduced in
revision 1.179 to correctly set/clear execute permission on the mapping
it creates. Thus, mmap(2)ing a memory resident file will not result in
the file being mapped with execute permission when execute permission was
not requested.
Eliminate an unneeded Instruction Memory Barrier (IMB) in
pmap_enter_quick(). Since there was no previous (instruction) mapping
for the given virtual address prior to pmap_enter_quick(), there can be
no instructions from the given virtual address in the pipeline that need
flushing.
bde [Fri, 2 Dec 2005 13:45:06 +0000 (13:45 +0000)]
Fixed roundf(). The following cases never worked in FreeBSD:
- in round-towards-minus-infinity mode, on all machines, roundf(x) never
worked for 0 < |x| < 0.5 (2*0x3effffff cases in all, or almost half of
float space). It was -0 for 0 < x < 0.5 and 0 for -0.5 < x < 0, but
should be 0 and -0, respectively. This is because t = ceilf(|x|) = 1
for these args, and when we adjust t from 1 to 0 by subtracting 1, we
get -0 in this rounding mode, but we want and expected to get 0.
- in round-towards-minus-infinity, round towards zero and round-to-nearest
modes, on machines that evaluate float expressions in float precision
(most machines except i386's), roundf(x) never worked for |x| =
<float value immediately below 0.5> (2 cases in all). It was +-1 but
should have been +-0. This is because t = ceilf(|x|) = 1 for these
args, and when we try to classify |x| by subtracting it from 1 we
get an unexpected rounding error -- the result is 0.5 after rounding
to float in all 3 rounding modes, so we we have forgotten the
difference between |x| and 0.5 and end up returning the same value
as for +-0.5.
The fix is to use floorf() instead of ceilf() and to add 1 instead of
-1 in the adjustment. With floorf() all the expressions used are
always evaluated exactly so there are no rounding problems, and with
adjustments of +1 we don't go near -0 when adjusting.
Attempted to fix round() and roundl() by cloning the fix for roundf().
This has only been tested for round(), only on args representable as
floats. Double expressions are evaluated in double precision even on
i386's, so round(0.5-epsilon) was broken even on i386's. roundl()
must be completely broken on i386's since long double precision is not
really supported. There seem to be no other dependencies on the
precision.
sos [Fri, 2 Dec 2005 10:13:53 +0000 (10:13 +0000)]
Update the ICH7 support so it deals better with chips without AHCI.
Update Intel MatrixRAID support to be able to pick up RAID0+1 (RAID10)
and RAID5 arrays without panic'ing.
This has the side effect of now also supporting multiple volumes on
MatrixRAID's now I have the metadata better understood..
davidxu [Fri, 2 Dec 2005 08:23:49 +0000 (08:23 +0000)]
1. Check if message priority is less than MQ_PRIO_MAX.
2. Use getnanotime instead of getnanouptime.
3. Don't free message in _mqueue_send, mqueue_send will free it.