ache [Tue, 28 Jan 2014 13:29:54 +0000 (13:29 +0000)]
MFC: r261146
Bad timespec_subtract() calculations produce negative tv_nsec on i386
which cause EINVAL returned from nanosleep() which cause loop in
cron_sleep() and making all cron jobs to start about 30 seconds earlier
(which cause f.e. logfiles rotation by newsyslog delayed by 1 hour).
Use simple and proved calculations from kernel's timespecsub() instead.
dim [Tue, 28 Jan 2014 08:07:19 +0000 (08:07 +0000)]
MFC r261162:
Pull in r195679 from upstream llvm trunk:
Don't use nopl in cpus that don't support it.
Patch by Mikulas Patocka. I added the test. I checked that for cpu names that
gas knows about, it also doesn't generate nopl.
The modified cpus:
i686 - there are i686-class CPUs that don't have nopl: Via c3, Transmeta
Crusoe, Microsoft VirtualBox - see
https://bbs.archlinux.org/viewtopic.php?pid=775414
k6, k6-2, k6-3, winchip-c6, winchip2 - these are 586-class CPUs
via c3 c3-2 - see https://bugs.archlinux.org/task/19733 as a proof that
Via c3 and c3-Nehemiah don't have nopl
pfg [Sun, 26 Jan 2014 19:49:54 +0000 (19:49 +0000)]
MFC r260311, r260831:
gcc: Add support for Apple's Block extension
Block objects [1] are a C-level syntactic and runtime feature. They
are similar to standard C functions, but in addition to executable
code they may also contain variable bindings to automatic (stack)
or managed (heap) memory. A block can therefore maintain a set of
state (data) that it can use to impact behavior when executed.
This port is based on Apple's GCC 5646 with some bugfixes from
Apple GCC 5666.3. It has some small differences with the support
in clang, which remains the recommended compiler.
Perhaps the most notable difference is that in GCC __block is not
actually a keyword, but a macro. There may be workaround for this
issue in the future. Other issues can be consulted in the clang
documentation [2]
For better compatiblity with Apple's GCC and llvm-gcc, some related
fixes and features from Apple have been included. Support for the
non-standard nested functions in GCC is now off by default.
No effort was made to update the ObjC support since FreeBSD doesn't
carry ObjC in the base system but some of the code crept in and
was more difficult to remove than to adjust.
mav [Sat, 25 Jan 2014 01:58:15 +0000 (01:58 +0000)]
MFC r260549:
Move xpt_run_devq() call before request completion callback where it was
originally.
I am not sure why exactly have I moved it during one of many refactorings
during camlock project, but obviously it opens race window that may cause
use after free panics during SIM (in reported cases umass(4)) detach.
bapt [Sat, 25 Jan 2014 01:08:35 +0000 (01:08 +0000)]
Implicit include of sys/queue.h instead of relying on gelf.h/libelf.h to bring it
Newer version of gelf.h and libelf.h does not include sys/queue.h anymore
hselasky [Fri, 24 Jan 2014 08:10:08 +0000 (08:10 +0000)]
MFC r260808 and r260814:
- Close a minor deadlock.
- Fix a possible memory use after free and leak situation associated
with USB device detach when using character device handles. This also
includes LibUSB. It turns out that "usb_close()" cannot always get a
reference to clean up its USB transfers and such, if called during the
kernel USB device detach.
hselasky [Fri, 24 Jan 2014 07:48:52 +0000 (07:48 +0000)]
MFC r260588 and r260589:
- Separate I/O errors from reception of STALL PID.
- Implement better error recovery for Transaction Translators, TTs,
found in High Speed USB HUBs which translate from High Speed USB into
FULL or LOW speed USB. In some rare cases SPLIT transactions might get
lost, which might leave the TT in an unknown state. Whenever we detect
such an error try to issue either a clear TT buffer request, or if
that is not possible reset the whole TT.
mav [Fri, 24 Jan 2014 00:41:02 +0000 (00:41 +0000)]
MFC r260521:
Remove not applicable PI_SDTR_ABLE and PI_WIDE_16 hba_inquiry flags to
make CAM to not try negotiate unsupported settings and suppress warnings.
While there, enable command queuing on pass-through devices, announced
in hba_inquiry, but disabled. Even though queue size is very small, It
seems working well enough.
jhb [Thu, 23 Jan 2014 20:35:32 +0000 (20:35 +0000)]
MFC 259826,259997,259998:
Support soft power-off via the ACPI S5 state for bhyve guests and wire
up a virtual power button to SIGTERM:
- Implement the PM1_EVT and PM1_CTL registers required by ACPI.
- Emulate the Reset Control register at I/O port 0xcf9.
- Advertise an _S5 package.
- Implement an SMI_CMD register with commands to enable and disable ACPI.
Currently the only change when ACPI is enabled is to enable the virtual
power button via SIGTERM.
- Implement a fixed-feature power button when ACPI is enabled by asserting
PWRBTN_STS in PM1_EVT when SIGTERM is received.
- Add support for EVFILT_SIGNAL events to mevent.
- Implement support for the ACPI system command interrupt (SCI) and assert
it when needed based on the values in PM1_EVT. Mark the SCI as active-low
and level triggered in the MADT and MP Table.
jhb [Thu, 23 Jan 2014 20:21:39 +0000 (20:21 +0000)]
MFC 257422,257661,258075,258476,258494,258579,258609,258699:
Several enhancements to the I/O APIC support in bhyve including:
- Move the I/O APIC device model from userspace into vmm.ko and add
ioctls to assert and deassert I/O APIC pins.
- Add HPET device emulation including a single timer block with 8 timers.
- Remove the 'vdev' abstraction.
loos [Thu, 23 Jan 2014 12:32:30 +0000 (12:32 +0000)]
MFC r256959:
Add the Raspberry Pi BSC (I2C compliant) controller driver.
Reviewed by: rpaulo
MFC r256961:
Enable the build of OFW I2C bus for FDT systems.
MFC r258045:
As all the IIC controllers on system uses the same 'iichb' prefix we cannot
rely only on checking the device unit to indentify the BSC unit we are
attaching to. Make use of the device base address to identify our BSC unit.
MFC r259127:
Bring the RPi I2C driver in line with ti_i2c. Make it treat any slave
address as a 7-bit address.
- Introduce additional hash to group requests by hash of sockref. This
allows to process TCP acknowledgements without looping though all the cache,
and as result allows to do it every time.
- Indroduce additional callbacks to notify application layer about sockets
disconnection. Without this last few requests processed just before socket
disconnection never processed their ACKs and stuck in cache for many hours.
- Implement transport-specific method for tracking reply acknowledgements.
New implementation does not cross multiple stack layers to get the data and
does not have race conditions that previously made some requests stuck
in cache. This could be done more efficiently at sockbuf layer, but that
would broke some KBIs, while I don't know other consumers for it aside NFS.
- Instead of traversing all DRC twice per request, run cleaning only once
per request, and except in some conditions traverse only single hash slot
at a time.
Together this limits NFS DRC growth only to situations of real connectivity
problems. If network is working well, and so all replies are acknowledged,
cache remains almost empty even after hours of heavy load. Without this
change on the same test cache was growing to many thousand requests even
with perfectly working local network.
As another result this reduces CPU time spent on the DRC handling during
SPEC NFS benchmark from about 10% to 0.5%.
mav [Wed, 22 Jan 2014 23:51:12 +0000 (23:51 +0000)]
MFC r260036:
Introduce xprt_inactive_self() -- variant for use when sure that port
is assigned to thread. For example, withing receive handlers. In that
case the function reduces to single assignment and can avoid locking.
mav [Wed, 22 Jan 2014 23:50:13 +0000 (23:50 +0000)]
MFC r260031:
In addition to r259632 completely block receive upcalls if we have more
data than we need. This reduces lock pressure from xprt_active() side.
mav [Wed, 22 Jan 2014 23:49:37 +0000 (23:49 +0000)]
MFC r259877:
Slightly simplify expiration logic introduced in r254337.
- Do not update the histogram for items we are any way deleting from cache.
- Do not update the histogram if nfsrc_tcphighwater is not set.
- Remove some extra math operations.
mav [Wed, 22 Jan 2014 23:48:15 +0000 (23:48 +0000)]
MFC r259765:
Fix RPC server threads file handle affinity to work better with ZFS.
Instead of taking 8 specific bytes of file handle to identify file during
RPC thread affitinity handling, use trivial hash of the full file handle.
ZFS's struct zfid_short does not have padding field after the length field,
as result, originally picked 8 bytes are loosing lower 16 bits of object ID,
causing many false matches and unneeded requests affinity to same thread.
This fix substantially improves NFS server latency and scalability in SPEC
NFS benchmark by more flexible use of multiple NFS threads.
mav [Wed, 22 Jan 2014 23:47:29 +0000 (23:47 +0000)]
MFC r259659, r259662:
Remove several linear list traversals per request from RPC server code.
Do not insert active ports into pool->sp_active list if they are success-
fully assigned to some thread. This makes that list include only ports that
really require attention, and so traversal can be reduced to simple taking
the first one.
Remove idle thread from pool->sp_idlethreads list when assigning some
work (port of requests) to it. That again makes possible to replace list
traversals with simple taking the first element.
mav [Wed, 22 Jan 2014 23:46:19 +0000 (23:46 +0000)]
MFC r259632:
Rework flow control for connection-oriented (TCP) RPC server.
When processing receive buffer, write the amount of data, expected
in present request record, into socket's so_rcv.sb_lowat to make stack
aware about our needs. When processing following upcalls, ignore them
until socket collect enough data to be read and processed in one turn.
This change reduces number of context switches and other operations
in RPC stack during large NFS writes (especially via non-Jumbo networks)
by order of magnitude.
After precessing current packet, take another look into the pending
buffer to find out whether the next packet had been already received.
If not, deactivate this port right there without making RPC code to
push this port to another thread just to find that there is nothing.
If the next packet is received partially, also deactivate the port, but
also update socket's so_rcv.sb_lowat to not be woken up prematurely.
This change additionally reduces number of context switches per NFS
request about in half.
mav [Wed, 22 Jan 2014 23:42:30 +0000 (23:42 +0000)]
MFC r258132:
Some minor tuning to rpc/svc.c:
- close cosmetic race in svc_exit();
- do not set wait timeout for idle threads if we have no use for wakeups;
- create new requested thread sooner, not only after some another thread
wakeup, that may happen later under constant load.
The first two were always GPL2. The last two were
added after the GPL3 transition, but were written
by aaw@google.com and Rafael EspÃndola got permission
to relicense them under the GPL2 for inclusion in
llvm-gcc.
pfg [Mon, 20 Jan 2014 19:37:38 +0000 (19:37 +0000)]
MFC r260014, r260099:
gcc: Add support for label attributes and "unavailable" attribute.
Apple GCC has extensions to support for both label attributes and
an "unavailable" attribute. These are critical for objc but are
also useful in regular C/C++.
Improve error message shown to the user when trying to load a module that is
already loaded or compiled withing the kernel
Point the user to dmesg(1) to get informations about why loading a module did fail
instead of printing the cryptic "Exec format error"
Update the BUGS section of kld(4) according the recent changes in kldload(8)
hrs [Mon, 20 Jan 2014 05:44:45 +0000 (05:44 +0000)]
- Fix a bxe(4) entry. This issue is not related to NFSv4, and poor
performance is caused by disabling TSO, not the issue itself.
- s/&os;-STABLE/&os; &release.current;/
gjb [Thu, 16 Jan 2014 16:44:23 +0000 (16:44 +0000)]
MFC r260772:
Update the pkg-stage target to be more compatible with pkg-1.2:
- Add a release-dvd.conf pkg(8) configuration file to override
the default FreeBSD.conf configuration.
- Remove architecture-specific pkg-stage.conf files, consolidate,
and move their contents to scripts/pkg-stage.sh.
- Use 'pkg -vv' to determine the ABI, which is used as the
cache directory.
Prior to these changes, it would be possible for pkg-stage to fetch
conflicting binary packages from multiple repositories.
dteske [Wed, 15 Jan 2014 08:17:31 +0000 (08:17 +0000)]
MFC r260261 (gavin):
Lower the priority of the "connect to any open wireless network" network
description, so guaranteeing that any other defined network will be tried
first.
MFC r260262 (gavin):
Allow bsdinstall to use WPA-Enterprise networks when installing. This
only allows basic username/password config, and does not provide the
ability to set any of the other WPA options. Regardless, this is
generally sufficient to associate.
Perhaps in the future this could allow full configuring (e.g. being able
to set "anonymous identity", and perhaps some of the more obscure WPA
options), though perhaps that will better belong in bsdconfig when that
grows wlan config ability.