is already in pages, so we should not convert from bytes to pages.
The result of this bug was bad scaling of the VHPT relative to the
available memory.
Submitted by: Arun Sharma <arun@sharma-home.net>
It's unnecessary for two reasons: (1) Giant is at present already held in
such cases and (2) our various implementations of pmap_growkernel() look to
be MP safe. (For example, for sparc64 the proof of (2) is trivial.)
freebsd4_sigaction() and osigaction() instead of around the whole
body of those functions. They now no longer hold Giant around calls
to copyin() and copyout(), and it is slightly more obvious what
Giant is protecting.
barrier between free'ing filedesc structures. Basically if you want to
access another process's filedesc, you want to hold this mutex over the
entire operation.
values for the initial inode generation numbers in newfs and for
newly allocated inode generation numbers in the kernel.
Submitted by: Theo de Raadt <deraadt@cvs.openbsd.org>
Sponsored by: DARPA & NAI Labs.
opposed to returning the top of the old chain when there was one and
the top of the newly allocated chain if there was no old chain.
Actually, it should be noted that prior to this fix, although the
comment above m_getm() advertised that m_getm() would return the
top of the old chain (if an old chain was being passed in) it
actually [wrongly] was returning the tail mbuf in the old chain
instead. This is a bug but since the one use of m_getm() in
the tree luckily did not depend on the behavior, it happened
to work out without notice.
Harti Brandt pointed out that the advertised behavior was actually
not the real behavior and so this change makes m_getm() ALWAYS
return the newly allocated chain (and fixes the comment). This
is less confusing and is the best course of action as then the
caller is always able to have both a reference to the top of
the original chain (because it's passing it in in the call) and
a reference to the newly attached chain. Although the API is
slightly modified, I don't think that any third-party code uses
m_getm() and if it does, it surely can't be working properly
because the old behavior was bogus.
API bug pointed out by: Harti Brandt <brandt@fokus.fraunhofer.de>
To fix scsi, don't wait for ithreads if we're dumping, it makes the
debugger sad.
To fix ata, use what appears to be a polling method if we're dumping,
I stole this from tmm but added code to ensure that this change is
only in effect while dumping.
Tested by: des
for the agp module, and add agp to the list of modules to compile for alpha.
Add an alpha_mb() to agp_flush_cache for alpha -- it's not correct but may
improve the situation, and it's what linux and NetBSD do.
we can have additional different types of bridges.
o remove now bogus comment.
o Don't clear CARD_OK when we can't attach a card.
o minor style nits
# this make kldload of cardbus drivers work for me when the card is
# present on boot.
o chip_name arrays ifdef'd out.
o use the OLDCARD-like get/put functions so we can support differnt types
of mappings.
o Write the beggings of is this a valid exca device and introduce more
chipset support.
# this is partially a wip, but also needed because some other cahnges I've
# made require some of these changes.
previous revision fixed the panic, I found the problem exits in
another part of the function by investigating the crom dump sent by him.
The search was started in the middle of bus info block and the
routine misunderstood the EUI64 as a crom entry. This problem is fixed.
PR: kern/48129
Fix incorrect type mask included in a logical unit number and check
the validity of the lun.
Introdice RTLD_SELF special handle and properly process it within
dlsym() and dlinfo() functions.
The intention is to improve our compatibility with Solaris and
to make a Java port easier.
Partially submitted by: phantom
- Drain fwohci TX queue first then drain xfer queue which has not started.
- Check validity of the received packet length.
- Don't allocate too large buffer for xfer receive buf.
sbp
- Fix panic for some CROM which doesn't have a text leaf.
This could fix the PR kern/48129 but no feedback has been gotten from
the originator yet.
- Put back some M_NOWAIT flags into malloc which could be called
in interrupt context for 4-stable.
warning which breaks builds.
cc1: warnings being treated as errors
src/sys/net/bridge.c: In function `bdg_forward':
sys/net/bridge.c:931: warning: suggest parentheses around assignment used as truth value
*** Error code 1
lower extremities.
Setting bit 4 in debugflags (sysctl kern.geom.debugflags=16) will
allow any open to succeed on rank#1 providers. This will generally
correspond to the physical disk devices: ad0, da0, md0 etc.
This fundamentally violates the mechanics of GEOMs autoconfiguration,
and is only provided as a debugging facility, so obviously error
reports on GEOM where this bit is or has been set will not be
accepted.
Kill the slightly bogus #define for DECODE_PROTOTYPE
Be less verbose. Hide most (all I hope) of the CIS
parsing behind cardbus_debug_cis (which is set with
hw.cardbus.debug_cis=1).
This doesn't fix problems with parsing, but should make cardbus
less chatty. There appears to be some issues still with the
parsing of the CIS, but this won't fix them.
Prompted by: scottl
Second part of the kldload patches for cardbus. This makes
kldload of a driver for a device that's inserted now appears
to work. To make it work, we only do a power cycle of the card
if there's no children drivers attached.
This likely is papering over bogosities in the power system. The
power sequence needs to be re-written, so I'll not worry about
the papering over until the re-write.
disk I/O processing.
The intent is that the disk driver in its hardware interrupt
routine will simply schedule the bio on the task queue with
a routine to finish off whatever needs done.
The g_up thread will then schedule this routine, the likely
outcome of which is a biodone() which queues the bio on
g_up's regular queue where it will be picked up and processed.
Compared to the using the regular taskqueue, this saves one
contextswitch.
Change our scheduling of the g_up and g_down queues to be water-tight,
at the cost of breaking the userland regression test-shims.
Input and ideas from: scottl
Cut up requests into smaller bits if they are longer than the drivers
disk->d_maxsize or dev->si_iosize_max.
Properly handle the race condition when using g_clone_bio() is used
without having the single-threadedness of g_down/g_up secure locking.
and d_stripesisze;
Introduce si_stripesize and si_stripeoffset in struct cdev so we
can make the visible to clustering code.
Add stripesize and stripeoffset to providers.
DTRT with stripesize and stripeoffset in various places in GEOM.
The locking here needs to be revisited, but this ought to get rid of the
LOR messages that people are complaining about for now. I imagine either
I or someone else interested with smp will eventually clear this up.
unconditionally. kldloading a cardbus driver was shooting down other
attached devices because most drivers assume that one cannot
power-cycle cards w/o the driver knowning about it.
Submitted by: simokawa-san
- Use the ratio of kg_runtime / kg_slptime to determine our dynamic priority.
- Scale kg_runtime and kg_slptime back when the sum of the two exceeds
SCHED_SLP_RUN_MAX. This allows us to slowly forget old behavior.
- Scale back the runtime and slptime in fork so that the new process has the
same ratio but much less accumulated time. This causes new behavior to be
noticed more quickly.
blocks now, which should eliminate problems with the driver failing to
attach due to insufficient contiguous RAM. Allow the FIB pool to grow
from the default of 128 to the max of 512 as demand grows. Also pad the
adapter init struct to work around the 2120/2200 DMA bug now that there
is no longer a FIB slab.
idle time.
Statistics now default to "on" and can be turned off with
sysctl kern.geom.collectstats=0
Performance impact of statistics collection is on the order of
800 nsec per consumer/provider set on a 700MHz Athlon.
that is protected by the vnode lock.
- Move B_SCANNED into b_vflags and call it BV_SCANNED.
- Create a vop_stdfsync() modeled after spec's sync.
- Replace spec_fsync, msdos_fsync, and hpfs_fsync with the stdfsync and some
fs specific processing. This gives all of these filesystems proper
behavior wrt MNT_WAIT/NOWAIT and the use of the B_SCANNED flag.
- Annotate the locking in buf.h
buf lists, synchronization variables, and atomic ops for the counters.
This change does not remove giant from any code although some pushdown
may be possible.
- In vfs_bio_awrite() don't access buf fields without the buf lock.
* implement watchdog timer.
* check all standing transactions in firewire_xfer_timeout().
- Add firewire_xferq_drain() for fw_busreset().
- Add/improve some debug messages.
- Call fw_xfer_done() if retry handler is NULL.
o Add a MD header private to libc called _fpmath.h; this header
contains bitfield layouts of MD floating-point types.
o Add a MI header private to libc called fpmath.h; this header
contains bitfield layouts of MI floating-point types.
o Add private libc variables to lib/libc/$arch/gen/infinity.c for
storing NaN values.
o Add __double_t and __float_t to <machine/_types.h>, and provide
double_t and float_t typedefs in <math.h>.
o Add some C99 manifest constants (FP_ILOGB0, FP_ILOGBNAN, HUGE_VALF,
HUGE_VALL, INFINITY, NAN, and return values for fpclassify()) to
<math.h> and others (FLT_EVAL_METHOD, DECIMAL_DIG) to <float.h> via
<machine/float.h>.
o Add C99 macro fpclassify() which calls __fpclassify{d,f,l}() based
on the size of its argument. __fpclassifyl() is never called on
alpha because (sizeof(long double) == sizeof(double)), which is good
since __fpclassifyl() can't deal with such a small `long double'.
This was developed by David Schultz and myself with input from bde and
fenner.
PR: 23103
Submitted by: David Schultz <dschultz@uclink.Berkeley.EDU>
(significant portions)
Reviewed by: bde, fenner (earlier versions)
Insted of embedding a struct g_stat in consumers and providers, merely
include a pointer.
Remove a couple of <sys/time.h> includes now unneeded.
Add a special allocator for struct g_stat. This allocator will allocate
entire pages and hand out g_stat functions from there. The "id" field
indicates free/used status.
Add "/dev/geom.stats" device driver whic exports the pages from the
allocator to userland with mmap(2) in read-only mode.
This mmap(2) interface should be considered a non-public interface and
the functions in libgeom (not yet committed) should be used to access
the statistics data.
so be more careful about calling stackgap_init.
Tested by: Fred Souza <fred@storming.org>
2) Linux_sendmsg was forgetting to fill out the bsd_args struct.
Reviewed by: ume
3) The args to linux_connect have differently named types on alpha and
i386, so add a cast to stop gcc complaining.
Spotted by: peter
- Cache temp. keys so they are preserved across suspend/resume (MPI-350)
- Reads and writes are real fast to the MPI-350 causing early timeouts so
wait do some DELAYs to slow things down in the spin loops.
- Stream line setting RIDs when they are better to be set via another
function
- Add better support for setting home key via "ifconfig an0 wepkey 9:<key>"
Tested by: Peter Radcliffe <pir@pir.net> (in -stable)
myself in -current & -stable
MFC in: 3 days
Add debug.sizeof.g_stat sysctl.
Set the id field of the g_stat when we create consumers and providers.
Remove biocount from consumer, we will use the counters in the g_stat
structure instead. Replace one field which will need to be atomically
manipulated with two fields which will not (stat.nop and stat.nend).
Change add companion field to bio_children: bio_inbed for the exact
same reason.
Don't output the biocount in the confdot output.
Fix KASSERT in g_io_request().
Add sysctl kern.geom.collectstats defaulting to off.
Collect the following raw statistics conditioned on this sysctl:
for each consumer and provider {
total number of operations started.
total number of operations completed.
time last operation completed.
sum of idle-time.
for each of BIO_READ, BIO_WRITE and BIO_DELETE {
number of operations completed.
number of bytes completed.
number of ENOMEM errors.
number of other errors.
sum of transaction time.
}
}
API for getting hold of these statistics data not included yet.
handy if the machine is on another floor. A minor issue with this is that
these functions are also used by the debugger, so its possible to break into
the debugger from the debugger.
PR: sparc64/47143
Reviewed by: benno
Approved by: jake (mentor)
correctly tell CAM to requeue the command and then freeze it's queue. The
problem was that when resources became available again, it wouldn't tell
CAM to unfreeze it's queue, so no more commands would ever be delivered.
This is simialr to the bug that was fixed in the cciss driver last year.
This is a bug in 4-STABLE also, but is probably masked by the OS being
fast enough to drain the completion queue before it fills up.
Also add some diagnostics avaialble when compiled with MLY_DEBUG.
Thanks very much to LSI Corp for donating equipment to track this down,
and Vaidus Damosevicius for pestering me long enough to get it fixed.
We may actually be increasing Giant contention doing so because the
actual stuff we do is very cheap.
Also I am not convinced there is not a tiny window for a race here.
Sync with userland test framework which now deals better with pcm feeder kobj
emulation.
Reduce max rate from 96kHz to 48kHz as userland tests found a few bad
points about 90kHz and we don't care about operating up there for now.
DDB when the interrupt button (aka the "programmer's switch") is pressed.
This isn't unfortunately an NMI, but it's a handy way to get into DDB
quickly if needed.
- Move the command timeout check from a separate repeating timeout to the
kthread since the kthread is already running periodically.
- Move printing the hardware print buffer to the kthread.
- Properly shut down the kernel thread on detach.
- Detach the child array devices on detach.
- Don't issue a controller halt command on detach. Doing so requires a PCI
reset to wake the controller back up. The driver can now be unloaded as
long as CAM support is not enabled.
similar patch has been in 4.x for a while, but is more hacky there.)
For this to work, vinum has to be loaded early (e. g. from
boot/loader), for obvious reasons. If the kernel env variable
(aka. loader variable) "vinum.autostart" is set, vinum then asks the
sysctl kern.disks for all available disks in the system, and scans
them for possible vinum headers.
For statically compiled kernels, this behaviour can be obtained even
without boot/loader by using "options VINUM_AUTOSTART" (though this is
not the recommended way).
Alternatively, the 4.x way to specify "vinum.drives" is also supported.
No further hacks (like the 4.x "vinum.root" variable) are needed,
since in 5.x, mountroot() asks back at the drivers to have them
resolve the name of the root FS into a dev_t (using the dev_clone
eventhandler).
(The MFC reminder below is for a partial MFC for vinum.autostart, the
rest is already there in 4.x.)
Timed out on: grog
MFC after: 2 weeks
we have the rc4 code already in the kernel (via wlan stuff or awi).
Add a dependency on the rc4 module so if it doesn't exist then load it.
Reviewed by: archie
1) It is already called in init_main.c:proc0_post()
2) It is called each time read_random_phony() called, because "initialized"
variable is never set to 1.
Approved by: markm
- Add hid.h
Obtained from: NetBSD
NOTE: This undoes some changes I'd made to prefix the processor name defines
with PVR_. This was due to my original decision to use MPC750 as a cpu name.
With this changed, the PVR_ change is no longer required.
Remove all the stuff that does not relate to the TSC.
Change the calibration to use DELAY(1000000) rather than trying to check
it against the CMOS RTC, this drastically increases precision:
Using 25 samples on a Athlon 700MHz UP machine I find:
stddev min max average
CMOS 22200 Hz -74980 Hz 34301 Hz 704928721 Hz
DELAY 1805 Hz -1984 Hz 2678 Hz 704937583 Hz
(The difference between the two averages is not statistically significant.)
expressed in PPM of the frequency:
stddev min max
CMOS 31.49 PPM -106.37 PPM 48.66 PPM
DELAY 2.56 PPM 2.81 PPM 3.80 PPM
This code will not be used until a followup commit to sys/isa/clock.c
and sys/pc98/pc98/clock.c which will only happen after some field testing.
MAC policy. To load a KLD, require that the subject hold Biba privilege,
and the the kernel module be marked as high integrity. To unload a
KLD, require that the subject hold Biba privilege.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
and MLS policies: as we support both an effective (single) element and
range (available) elements, require that the single be in the range if
both the single and range are defined in the update. Remove comments
suggesting that such a check might be a good idea.
Don't introduce a similar check for network interfaces; due to different
interpretations of the single and range elements, it's not clear that
it's useful to do so.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
labeling for Biba.
Rename the variable 'level' to 'type' in interface parsing and
labeling for MLS.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, Network Associates Laboratories
uio segment is empty. In this case no dma segment is create by
bus_dmamap_load_buffer, but the calling routine clears the first flag.
Under certain combinations of addresses of the first and second mbuf/uio
buffer this leads to corrupted DMA segment descriptors. This was already
fixed by tmm in sparc64/sparc64/iommu.c.
PR: kern/47733
Reviewed by: sam
Approved by: jake (mentor)
some USB devices. (Make sure to set xfer data length when we force
a short inquiry.)
Obtained from: NetBSD(sys/dev/usb/umass_scsipi.c rev.1.8)
Original idea from: Shingo WATANABE <nabe@nabechan.org>
