done in other places. Until we have no support for command queueing we have
no any benefit from FBS, while enabling it only here somehow leads to
"port not ready" errors on Intel 63XXESB2 controller.
Tested by: Larry Rosenman <ler AT lerctr.org>
pointers together, move padding to the bottom of the structure, and add
two new integer spares due to attrition over time. Remove unused spare
"flags" field, we can use one of the spare ints if we need it later.
This change requires a rebuild of device driver modules that depend on
the layout of ifnet for binary compatibility reasons.
Discussed with: kmacy
which are not in a module of their own like gif.
Single kernel compiles and universe will fail if the size of the struct
changes. Th expected values are given in sys/vimage.h.
See the comments where how to handle this.
Requested by: peter
architecture to implement size-guards on the vimage vnet_* structures.
As CTASSERT_EQUAL() needs special compile time options we back it
by CTASSERT() in the default case. Unfortunately CTASSERT() triggers
first, thus add an option to allow compilation with CTASSERT_EQUAL() only.
See the comments how to get new values if you trigger the assert
and what to do in that case.
Reviewed by: rwatson, zec (earlier versions)
It's better to just use internal language constructs, because it is
likely the compiler has a better opinion on whether to perform inlining,
which is very likely to happen to struct winsize.
Submitted by: Christoph Mallon <christoph mallon gmx de>
It takes a positive integer constant (the expected value) and
another positive integer, usually compile-time evaluated,
e.g. CTASSERT_EQUAL(FOO_EXPECTED_SIZE, sizeof (struct foo));
While the classic CTASSERT() gives:
error: size of array '__assert60' is negative
this gives you:
In function '__ctassert_equal_at_line_60':
warning: '__expected_42_but_got[464ul]' is used uninitialized in this function
and you can directly see the difference in the expected and the
real value.
CTASSERT_EQUAL() needs special compile time options to trigger
thus keep it locally to this header. If it proves to be of general
interest it can be moved to systm.h.
Submitted by: jmallett
Reviewed by: sam, warner, rwatson, jmallett (earlier versions)
* Add RB_FOREACH_FROM() which continues traversal *at*
the y-node provided. There is no pre-increment.
* Nuke RB_FOREACH_SAFE as it was buggy; it would omit the final node.
* Replace RB_FOREACH_SAFE() with a working implementation
derived from RB_FOREACH_FROM().
The key observation is that we now only check the loop-control
variable, but still cache the next member pointer.
* Add RB_FOREACH_REVERSE_FROM() which continues backwards
traversal *at* the y-node provided. There is no pre-increment.
Typically this is used to back out of allocations made
whilst walking an RB-tree.
* Add RB_FOREACH_REVERSE_SAFE() which performs insertion and
deletion safe backwards traversal.
and partially r188903. Revert breaks new drives detection on reinit to the
state as it was before me, but fixes series of new bugs reported by some
people.
Unconditional queueing of ata_completed() calls can lead to deadlock if
due to timeout ata_reinit() was called at the same thread by previous
ata_completed(). Calling of ata_identify() on ata_reinit() in current
implementation opens numerous races and deadlocks.
Problems I was touching here are still exist and should be addresed, but
probably in different way.
When copying big structures, LLVM generates calls to memmove(), because
it may not be able to figure out whether structures overlap. This caused
linker errors to occur. memmove() is now implemented using bcopy().
Ideally it would be the other way around, but that can be solved in the
future. On ARM we don't do add anything, because it already has
memmove().
Discussed on: arch@
Reviewed by: rdivacky
drivers' probe routines. It allows not to sleep and so not drop Giant inside
ata_identify() critical section and so avoid crash if it reentered on
request timeout. Reentering of probe call checked inside of it.
Give device own knowledge about it's type (ata/atapi/atapicam). It is not
a good idea to ask channel status for device type inside ata_getparam().
Add softc memory deallocation on device destruction.
On FreeBSD, this is the default behaviour. According to the spec, we may
give this flag a value of zero, but I'd rather not do this. If we define
it to a non-zero value, we can always change default behaviour without
changing the ABI. This is very unlikely to happen, though.
wcscasecmp(), and wcsncasecmp().
- Make some previously non-standard extensions visible
if POSIX_VISIBLE >= 200809.
- Use restrict qualifiers in stpcpy().
- Declare off_t and size_t in stdio.h.
- Bump __FreeBSD_version in case the new symbols (particularly
getline()) cause issues with ports.
Reviewed by: standards@
at irq install/uninstall time, but when we vt switch, we uninstall the
irq handler. When the irq handler is reinstalled, the modeset ioctl
happens first. The modeset ioctl is supposed to tell us that we can
disable vblank interrupts if there are no active consumers. This will
fail after a vt switch until another modeset ioctl is called via dpms
or xrandr. Leading to cases where either interrupts are on and can't
be disabled, or worse, no interrupts at all.
MFC after: 2 weeks
usb stack rather than with the rest of the processor support code.
Not sure that's a good idea, as we were moving away from it, but this
fixes the build in the mean time so we can have that discussion.
- Fix the copy, we can't do a blind copy but must transfer
the data from the old to the new.
- Fix the ACK processing so we properly stop retransmitting
the thing.
- Fix it so if we get a retran we will properly reply with
the saved response without doing anything.
MFC after: 1 month
the devfs clone handler to open the (invisible) devices on the fly.
The /dev entries are layed out as follows,
/dev/usbctl = master device
/dev/usb/0.1.0.5 = usb device, (<bus>.<dev>.<iface>.<endpoint>)
/dev/ugen0.1 -> usb/0.1.0.0 = ugen link to ctrl endpoint
This also removes the custom permissions model from USB. Bump
__FreeBSD_version to 800066.
Submitted by: rink (earlier version)
net/route.h.
Remove the hidden include of opt_route.h and net/route.h from net/vnet.h.
We need to make sure that both opt_route.h and net/route.h are included
before net/vnet.h because of the way MRT figures out the number of FIBs
from the kernel option. If we do not, we end up with the default number
of 1 when including net/vnet.h and array sizes are wrong.
This does not change the list of files which depend on opt_route.h
but we can identify them now more easily.
printf() and vprintf() are exactly the same, except the way arguments
are passed. Just like we see in other pieces of code (i.e. libc's
printf()), implement printf() using vprintf().
Submitted by: Christoph Mallon <christoph mallon gmx de>
As mentioned by bz and bde, the change I made wasn't the proper way to
fix. Inspired by bde's patch, perform some small cleanups to uprintf().
Reviewed by: bz
Previously, DBCR0 flags were set "globally", but this leads to problems
because Book-E fine grained debug settings work only in conjuction with the
debug master enable bit in MSR: in scenarios when the DBCR0 was set with
intention to debug one process, but another one with MSR[DE] set got
scheduled, the latter would immediately cause debug exceptions to occur upon
execution of its own code instructions (and not the one intended for
debugging).
To avoid such problems and properly handle debugging context, DBCR0 state
should be managed individually per process.
Submitted by: Grzegorz Bernacki gjb ! semihalf dot com
Reviewed by: marcel
The function pow() in libmp(3) clashes with pow(3) in libm. We could
rename this single function, but we can just take the same approach as
the Solaris folks did, which is to prefix all function names with mp_.
libmp(3) isn't really popular nowadays. I suspect not a single
application in ports depends on it. There's still a chance, so I've
increased the SHLIB_MAJOR and __FreeBSD_version.
Reviewed by: deischen, rdivacky
kernel dumping case.
ata_completed() may initiate ata_reinit() on error, that may lead to drives
attach or detach. Attach and detach are sending requests to drives and sleep
waiting for results. But ata_finish() can be called directly from
interrupt handler where sleeping is prohibited, so we must break this chain
somewhere. This place seems to fit best.
go before the standard -I$S to have the search happen before /sys/.
Make this conditional on 'makeoptions WITH_LEGACY=1' in the kernel config.
Prodded by: sam
Disable MSI for nVidia MCP51 controller. Enabling MSI there leads to
unexpected errors and timeouts, that should not happen even if interrupts
are not working completely.
Currently bread()-ing through device vnode with
(1) VMIO enabled,
(2) bo_bsize != DEV_BSIZE
(3) more than 1 block
results in data being incorrectly cached.
So instead a more common approach of using a vnode belonging to fs is now
employed.
Also, prevent attempt to bread more than MAXBSIZE bytes because of
adjustments made to account for offset that doesn't start on block
boundary.
Add expanded comments to explain the calculations.
Also drop unused inline function while here.
PR: kern/120967
PR: kern/129084
Reviewed by: scottl, kib
Approved by: jhb (mentor)
Inside do_execve(), we have a pointer `ndp', which always points to
`&nd'. I can imagine a primitive (non-optimizing) compiler to really
reserve space for such a pointer, so just remove the variable and use
`&nd' directly.
kern_time.c:
- Unused variable `p'.
kern_thr.c:
- Variable `error' is always caught immediately, so no reason to
initialize it. There is no way that error != 0 at the end of
create_thread().
kern_sig.c:
- Unused variable `code'.
kern_synch.c:
- `rval' is always assigned in all different cases.
kern_rwlock.c:
- `v' is always overwritten with RW_UNLOCKED further on.
kern_malloc.c:
- `size' is always initialized with the proper value before being used.
kern_exit.c:
- `error' is always caught and returned immediately. abort2() never
returns a non-zero value.
kern_exec.c:
- `len' is always assigned inside the if-statement right below it.
tty_info.c:
- `td' is always overwritten by FOREACH_THREAD_IN_PROC().
Found by: LLVM's scan-build
... as opposed to files with data in extents.
Some UDF authoring tools produce this type of file for sufficiently small
data files.
Approved by: jhb (mentor)
Use bit-shift instead of division/multiplication.
Act on error as soon as it is detected.
Report attempt to read data embedded in file entry via regular way.
While there, fix lblktosize macro and make use of it.
No functionality should change as a result.
Approved by: jhb (mentor)
`p' is already initialized with `td->td_proc'. Because td is always
curthread, it is safe to initialize it without any locks.
Found by: LLVM's scan-build
priv:kernel:priv_check:priv_ok fires for granted privileges
priv:kernel:priv_check:priv_errr fires for denied privileges
The first argument is the requested privilege number. The naming
convention is a little different from the OpenSolaris equivilent
because we can't have '-' in probefunc names, and our privilege
namespace is different.
MFC after: 1 week
fault. In r188331 this update was relocated because of synchronization
changes to a place where it would occur on both hard and soft faults. This
change again restricts the update to hard faults.
required to make 3CR990 familiy controllers run on NV flash
firmware version 03.001.008.
The latest firmware added HMAC digest information so teach txp(4)
to pass them to sleep image before downloading is started.
While I'm here restore previous IMR/IER register if firmware
downloading have failed.
PR: kern/89876, kern/132047
The old BTX passed the general purpose registers from the 32-bit client to
the routines called via virtual 86 mode. The new BTX did the same thing.
However, it turns out that some instructions behave differently in virtual 86
mode and real mode (even though this is under-documented). For example, the
LEAVE instruction will cause an exception in real mode if any of the upper
16-bits of %ebp are non-zero after it executes. In virtual 8086 mode the
upper 16-bits are simply ignored. This could cause faults in hardware
interrupt handlers that inherited an %ebp larger than 0xffff from the 32-bit
client (loader, boot2, etc.) while running in real mode.
To fix, when executing hardware interrupt handlers provide an explicit clean
state where all the general purpose and segment registers are zero upon
entry to the interrupt handler. While here, I attempted to simplify the
control flow in the 'intusr' code that sets up the various stack frames
and exits protected mode to invoke the requested routine via real mode.
A huge thanks to Tor Egge (tegge@) for debugging this issue.
Submitted by: tegge
Reviewed by: tegge
Tested by: bz
MFC after: 1 week
function, done in r188334. Instead, collect the entries that shall be
freed, in the deferred_freelist member of the map. Automatically purge
the deferred freelist when map is unlocked.
Tested by: pho
Reviewed by: alc
an NFS file. Now the priming is conditional on a new
vfs.nfs.prime_access_cache sysctl. For now I've left the default setting
to disabling the priming.
Requested by: scottl
checks for the tcpcb, previously used to detect complete disconnection,
with INP_DROPPED checks. Correct that, preventing shutdown() from
improperly generating a TCP segment with destination IP and port of
0.0.0.0:0.
PR: kern/132050
Reported by: david gueluy <david.gueluy at netasq.com>
MFC after: 3 weeks
- We don't need to exit the Giant mutex when sleeping. This is done
automatically. Replace Giant by NULL mutex for all control requests in the
enumeration path.
- Optimise away duplicate alternate interface selection requests in USB Host
mode.
Submitted by: Hans Petter Selasky
Improvements to "usb2_transfer_setup()" and "usb2_transfer_unsetup()". Set
"ppxfer[n]" when the transfer setup is complete to prevent races. Remove
redundant NULL-checks from "usb2_transfer_unsetup()".
Submitted by: Hans Petter Selasky
- The software computed HID size is not always correct, because the algoritm
does not handle unsorted HID descriptors.
- Change the way we obtain the report ID.
- Use the X/Y/Z+button locations instead for report ID source for ums.
- Add more range checks.
- Remove Microsoft Mouse quirks. If the positions are moduloed the report
length multiplied by 8, the values seem correct.
- Some minor style changes.
Submitted by: Hans Petter Selasky
o add ah_configPCIE and ah_disablePCIE for drivers to configure PCIE
power save operation (modeled after ath9k, may need changes)
o add private state flag to indicate if device is PCIE (replaces private
hack in 5212 code)
o add serdes programming ini bits for 5416 and later parts and setup
for each part (5416 and 9160 logic hand-crafted from existing routines);
5212 remains open-coded but is now hooked in via ah_configPCIE
o add PCIE workaround gunk
o add ar5416AttachPCIE for iodomatic code used by 5416 and later parts
o add output mux support
o gpio pin count is chip-dependent
o 9280 and 9285 do input handling different
o hookup gpio interrupts
o no need to save/restore soft led state around reset
1 will trigger a pass through the VM's low-memory handlers, such as
protocol and UMA drain routines. This makes it easier to exercise
these otherwise rarely-invoked code paths.
MFC after: 3 days
to in_rtrequest(); the radix head lock is already acquired before
rnh_walktree is called in in_rtqtimo_one(). This avoids a recursive
acquisition that is no longer permitted in 8.x due to use of an rwlock
for the radix head lock.
Reported by: dikshie <dikshie at gmail.com>
MFC after: 3 days
msdosfs_unmount() and ffs_unmount() exit early after getting ENXIO.
However, dounmount() treats ENXIO as a success and proceeds with
unmounting. In effect, the filesystem gets unmounted without closing
GEOM provider etc.
Reviewed by: kib
Approved by: rwatson (mentor)
Tested by: dho
Sponsored by: FreeBSD Foundation
register instead of AAC_RX_FWSTATUS, as that is the way it's done in
Adaptec's vendor driver and in the Linux drivers. (The same applies
to aac_rkt_get_fwstatus as well.)
However, a concern has been raised about the compatibility of this
change and old hardware / firmware versions. In the absense of
specific information, revert to the original behaviour if the firmware
does not support the "New comm." interface. Users of old cards or
firmware haven't reported the problems that are potentially solved by
switching to OMR0.
different cpu is still assigned to that vector by never clearing idt
entries. This was only provided as a debugging feature and the bugs
are caught by other means.
- Drop the sched lock when rebinding to reassign an interrupt vector
to a new cpu so that pending interrupts have a chance to be delivered
before removing the old vector.
Discussed with: tegge, jhb
- protect againtst recursions,
- add new devices detection using ata_identify().
Improve ata_identify():
- do not add duplicate device if device already exist.
Rework SATA hot-plug events handling. Instead of unsafe duplicate
implementation use common ata_reinit() to handle all state changes.
All together this gives quite stable and robust cold- and hot-plug operation,
invariant to false, lost and duplicate events.
properly. Otherwise the minimum of 1 is used and you can
only insert a single partition/slice and only at sector
0 (index 1).
o When adding a partition/slice, recalculate the index after
the start and size of the partition/slice are adjusted to
make them a multiple of the track size. Since the precheck
method sets the index based on the start of the partition
as provided by the user, we know that we're off by at most
1 and adjusting the index is safe.
[1] Add the support for the NARK controller which seems a variant of
the i960Rx.
[2] Split up memory regions and other resources in 2 different parts
as long as NARK uses them separately (it is not clear to me
why though as long as there are no more informations available
on this controller). Please note that in all the other cases,
the regions overlaps leaving the default behaviour for all the
other controllers.
[3] Implement a clock daemon responsible for maintain updated the
wall clock time of the controller (run any 30 minutes)*.
Submitted by: Adaptec (driver build 15317 [1, 2] and 15727 [3])
Reviewed by: emaste
Tested by: emaste
Sponsored by: Sandvine Incorporated
* Please note that originally, in the Adaptec driver, the clock daemon
is not implemented with callouts as in our in-tree driver.
predefined set of methods, which are set in osd_register() and called
via osd_call(). Currently, no methods are defined, though prison
objects will have some in the future.
Expand the locking from a single per-type mutex to three different kinds
of locks (four if you include the requirement that the container
(e.g. prison) be locked when getting/setting data). This clears up one
existing issue, as well as others added by the method support.
Approved by: bz (mentor)
ATA specification declares minimal reset time of 5us. SATA keeps it, but
requires devices to handle commands transmitted even one by one without
any gap.
The existing code calls kern_open() to resolve the vnode of a pathname
right after a stat(). This is not correct, because it causes random
character devices to be opened in /dev. This means ls'ing a tape
streamer will cause it to rewind, for example. Changes I have made:
- Add kern_statat_vnhook() to allow binary emulators to `post-process'
struct stat, using the proper vnode.
- Remove unneeded printf's from stat() and statfs().
- Make the Linuxolator use kern_statat_vnhook(), replacing
translate_path_major_minor_at().
- Let translate_fd_major_minor() use vp->v_rdev instead of
vp->v_un.vu_cdev.
Result:
crw-rw-rw- 1 root root 0, 14 Feb 20 13:54 /dev/ptmx
crw--w---- 1 root adm 136, 0 Feb 20 14:03 /dev/pts/0
crw--w---- 1 root adm 136, 1 Feb 20 14:02 /dev/pts/1
crw--w---- 1 ed tty 136, 2 Feb 20 14:03 /dev/pts/2
Before this commit, ptmx also had a major number of 136, because it
silently allocated and deallocated a pseudo-terminal. Device nodes that
cannot be opened now have proper major/minor-numbers.
Reviewed by: kib, netchild, rdivacky (thanks!)
as ATA RAID, but generic ATAPCI driver unable to detect drives there. AHCI
driver reported to handle them fine. Linux does the same.
Submitted by: Andrey V. Elsukov on stable@
This fixes the low "max device openings" count that has lead to poor
performance in FreeBSD 7.0 and 7.1.
Extra thanks goes to Mike Tancsa at Sentex for providing a debug system for
this.
1. Extend geom_dev by having it create the symlink (i.e. call
make_dev_alias) based on the DIOCGPROVIDERALIAS ioctl.
In this way the functionaility is generic and thus usable
by any geom/provider.
2. Have g_part handle said ioctl through the devalias method,
so that it's under control of the scheme itself. By design
the alias will not be created for newly added partitions.
stale entries, we save a copy of the directory's modification time when
the first negative cache entry was added in the directory's NFS node.
When a negative cache entry is hit during a pathname lookup, the parent
directory's modification time is checked. If it has changed, all of the
negative cache entries for that parent are purged and the lookup falls
back to using the RPC. This required adding a new cache_purge_negative()
method to the name cache to purge only negative cache entries for a given
directory.
Submitted by: mohans, Rick Macklem, Ricardo Labiaga @ NetApp
Reviewed by: mohans
opportunistic ACCESS RPC to populate both the access and attribute caches
of the file and instead always use a GETATTR RPC. On many modern NFS
servers, an ACCESS RPC is much more expensive to service than a GETATTR
RPC.
Submitted by: mohans
open() of the same file will load fresh attributes, so they do not need to
be explicitly flushed in close() to guarantee close to open consistency.
However, other file desciptors may still reference this file and clearing
the attributes in close() forces those other file descriptors to fetch
fresh attributes the next time they need them.
Reviewed by: mohans
MFC after: 1 week
- Don't return a negative errno when using an unknown ioctl() on a
pseudo-terminal master device. Be sure to convert ENOIOCTL to ENOTTY,
just like the TTY layer does.
- Even though we should return st_rdev of the master device node when
emulating pty(4) devices, FIODGNAME should still return the name of
the slave device. Otherwise ptsname(3) and ttyname(3) return an
invalid device name.
Not enough space in user-land buffer is not an error, userland
will read further until eof is reached. So instead of propagating
-1 to caller we convert it to zero/success.
cd9660 code works exactly the same way.
PR: kern/78987
Reviewed by: jhb (mentor)
Approved by: jhb (mentor)
This is triggered only if BIOS configures ACPI_BITREG_BUS_MASTER_RLD
aka BRLD_EN_BM to 1.
Rationale:
1. we do not support C3 on PIIX4E
2. bus master activity need not break out of C2 state
3. because of CPU_QUIRK_NO_BM_CTRL quirk we may reset bus master
status which would result in immediate break out from C2
So if you have seen
cpu0: too many short sleeps, backing off to C1
with this chipset before you may want to try cx_lowest of C2 again.
Reviewed by: rpaulo (mentor), njl
Approved by: rpaulo (mentor)
ata_detach() to implement IOCATAATTACH/IOCATADETACH ioctls.
This will permit channel drivers to properly shutdown port hardware on channel
detach and init it on attach.
and xmit parameters. This makes it possible to use tdma on fractional
channels.
o add IEEE80211_MODE_HALF and IEEE80211_MODE_QUARTER; note these are
band-agnostic (may need revisiting)
o setup all default rates in ic_sup_rates instead of doing it only
for active modes; we need these to calculate the default tx parameters
which are not recalculated after a regulatory update (can't just
recalculate after installing a new channel list because we might
clobber user settings)
o remove special case code in ieee80211_get_suprates; this is now
a candidate for an inline or removal
o add various entries for new modes (roaming+tx params, wme, rate
mapping, scan set setup, country ie construction, tdma, basic rates)
Note these modes are intentionally not visible through if_media.
