so_gencnt, numopensockets, and the per-socket field so_gencnt. Annotate
this this might be better done with atomic operations.
Annotate what accept_mtx protects.
to warn about attempts to sleep in the I/O path. This change pushes the
definition and use of 'mymutex' behind #ifdef WITNESS to avoid the cost
in non-debugging cases. This results in a clear .22% performance win for
512 byte and 1k I/O tests on my SMP test box. Not much, but every bit
counts.
associated with performing a wakeup on the socket buffer:
- When performing an sbappend*() followed by a so[rw]wakeup(), explicitly
acquire the socket buffer lock and use the _locked() variants of both
calls. Note that the _locked() sowakeup() versions unlock the mutex on
return. This is done in uipc_send(), divert_packet(), mroute
socket_send(), raw_append(), tcp_reass(), tcp_input(), and udp_append().
- When the socket buffer lock is dropped before a sowakeup(), remove the
explicit unlock and use the _locked() sowakeup() variant. This is done
in soisdisconnecting(), soisdisconnected() when setting the can't send/
receive flags and dropping data, and in uipc_rcvd() which adjusting
back-pressure on the sockets.
For UNIX domain sockets running mpsafe with a contention-intensive SMP
mysql benchmark, this results in a 1.6% query rate improvement due to
reduce mutex costs.
The overhead of unconditionally allocating TIDs (and likewise,
unconditionally deallocating them), is amortized across multiple
thread creations by the way UMA makes it possible to have type-stable
storage.
Previously the cost was kept down by having threads created as part
of a fork operation use the process' PID as the TID. While this had
some nice properties, it also introduced complexity in the way TIDs
were allocated. Most importantly, by using the type-stable storage
that UMA gives us this was also unnecessary.
This change affects how core dumps are created and in particular how
the PRSTATUS notes are dumped. Since we don't have a thread with a
TID equalling the PID, we now need a different way to preserve the
old and previous behavior. We do this by having the given thread (i.e.
the thread passed to the core dump code in td) dump it's state first
and fill in pr_pid with the actual PID. All other threads will have
pr_pid contain their TIDs. The upshot of all this is that the debugger
will now likely select the right LWP (=TID) as the initial thread.
Credits to: julian@ for spotting how we can utilize UMA.
Thanks to: all who provided julian@ with test results.
uhid.c (1.61), author: jdolecek
add support for USB_GET_DEVICEINFO and USB_GET_STRING_DESC ioctls,
with same meaning as for ugen(4)
usbdi_util.h (1.29), usb_quirks.c (1.50), uhid.c (1.62),
ugen.c (1.68), usb_subr.c (1.114) author: mycroft
Yes, some devices return incorrect lengths in their string
descriptors. Rather than losing, do what Windows does: just
request the maximum size, and allow a shorter response. Obsoletes
the need for UQ_NO_STRINGS, and therefore these "quirks" are removed.
usb_subr.c (1.116), author: mycroft
In the "seemed like a good idea until I found the fatal flaw"
department... Attempting to read a maximum-size string descriptor
causes my kue device to go completely apeshit. So, go back to the
original method, but allow the device to return a shorter string than
it claimed.
Obtained from: NetBSD
copies.
No current line disciplines have a dynamically changing hotchar, and
expecting to receive anything sensible during a change in ldisc is
insane so no locking of the hotchar field is necessary.
ohci.c (1.147), author: mycroft
Failure to properly mask off UE_DIR_IN from the endpoint address
was causing OHCI_ED_FORMAT_ISO and EHCI_QH_HRECL to get set
spuriously, causing rather interesting lossage.
Suddenly I get MUCH better performance with ehci...
ohci.c (1.148), author: mycroft
Adjust a couple of comments to make it clear WTF is going on.
Obtained from: NetBSD
we have to revert to TTYDISC which we know will successfully open
rather than try the previous ldisc which might also fail to open.
Do not let ldisc implementations muck about with ->t_line, and remove
code which checks for reopens, it should never happen.
Move ldisc->l_hotchar to tty->t_hotchar and have ldisc implementation
initialize it in their open routines. Reset to zero when we enter
TTYDISC. ("no" should really be -1 since zero could be a valid
hotchar for certain old european mainframe protocols.)
Now that the devs files are marked before-depend, we can remvoe them
from a few places they were explicitly mentioned (along with
BEFORE_DEPEND).
Noticed by: bde
Reduce the need for hard coded *devs in various makefiles by declaring
them before-depend.
Other bugs in the handling of *devs remain, but this is the start of
the cleanup. These will be address in future commits.
Cleanup Motivator: bde
ehci.c (1.55), ehcireg.h (1.16); author: mycroft
Set the data toggle correctly, and use EHCI_QTD_DTC. This fixes
problems with my ALi-based drive enclosure (it works now, rather
than failing to attach). Also seems to work with a GL811-based
enclosure and an ASUS enclosure with a CD-RW, on both Intel and
NEC controllers.
Note: The ALi enclosure is currently very SLOW, due to some issue
with taking too long to notice that the QTD is complete. This
requires more investigation.
ehci.c (1.56); author: mycroft
Failure to properly mask off UE_DIR_IN from the endpoint address
was causing OHCI_ED_FORMAT_ISO and EHCI_QH_HRECL to get set
spuriously, causing rather interesting lossage.
Suddenly I get MUCH better performance with ehci...
ehci.c (1.58); author: mycroft
Fix a stupid bug in ehci_check_intr() that caused use to try to
complete a transaction that was still running. Now ehci can
handle multiple devices being active at once.
ehci.c (1.59); author: enami
As the ehci_idone() now uses the variable `epipe'
unconditionally, always declare it (in other words, make this
file compile w/o EHCI_DEBUG).
ehci.c (1.60); author: mycroft
Remove comment about the data toggle being borked.
ehci.c (1.61); author: mycroft
Update comment.
ehci.c (1.62); author: mycroft
Adjust a couple of comments to make it clear WTF is going on.
ehci.c (1.63); author: mycroft
Fix an error in a debug printf().
ehci.c (1.64), ehcireg.h (1.17); author: mycroft
Further cleanup of toggle handling. Now that we use EHCI_QH_DTC,
we don't need to fiddle with the TOGGLE bit in the overlay
descriptor, so minimize how much we fuss with it.
Obtained from: NetBSD
Thanks to Sam for importing tags in a way that allowed this to be done.
Submitted by: Gleb Smirnoff <glebius@cell.sick.ru>
Also allow the sr and ar drivers to create netgraph versions of their modules.
Document the change to the ksocket node.
when not propogated on fork (due to minherit(2)). Consistency checks
otherwise fail when the vm_map is freed and it appears to have not been
emptied completely, causing an INVARIANTS panic in vm_map_zdtor().
PR: kern/68017
Submitted by: Mark W. Krentel <krentel@dreamscape.com>
Reviewed by: alc
because UFS uses fixed-size directory blocks. When using this code with
other file systems, such as HFS+, the value of auio.uio_resid will need
to be taken into account.
local to a function. Remove a couple of blank lines in variable
declarations.
In one case, explicitly test against NULL rather than using a pointer
as a boolean directly.
older API to list attributes on a file (zero-length attribute name)
to function. extattr_list_*() are now the only available APIs to
use when listing attributes.
only allow this to be further processed when bridging is active on
that interface, but also if the current packet has a VLAN tag and
VLANs are active on our interface. This gives the VLAN layers a
chance to also consider the packet (and perhaps drop it instead of the
main dispatcher).
This fixes a situation where bridging was only active on VLAN
interfaces but ether_demux() called on behalf of the main interface
had already thrown the packet away.
MFC after: 4 weeks
are currently all bad BIOS revisions that will never be able to support
ACPI. They were derived by examining which BIOS's are blacklisted by other
operating systems. Other types of quirks will be possible here as well.
network interfaces. This global mutex will protect all ifnet labels.
Acquire the mutex across various MAC activities on interfaces, such
as security checks, propagating interface labels to mbufs generated
from the interface, retrieving and setting the interface label.
Introduce mpo_copy_ifnet_label MAC policy entry point to copy the
value of an interface label from one label to another. Use this
to avoid performing a label externalize while holding mac_ifnet_mtx;
copy the label to a temporary ifnet label and then externalize that.
Implement mpo_copy_ifnet_label for various MAC policies that
implement interface labeling using generic label copying routines.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, McAfee Research
locking in tcp_input() for TCP packets with urgent data pointers to
hold the socket buffer lock across testing and updating oobmark
from just protecting sb_state.
Update socket locking annotations
Ultra2 users may want to set OFWCONS_POLL_HZ to a value of '20'.
I have left default value at '4' as higher values can consume a more
than is acceptable amount of CPU, and we don't have a consensus yet
what is an optimal value.
Submitted by: Pyun YongHyeon <yongari@kt-is.co.kr>
not active GEOM providers, it will result in a kernel panic.
If the GEOM provider or disk goes away before the volume
configuration data gets written to the disk, it will result
in another kernel panic.
o Make sure that the drives specified for volume creation
are active GEOM providers.
o When writing out volume configuration data to associated drives,
make sure that the GEOM provider is active, otherwise continue
to the next drive in the volume.
Approved by: le, bmilekic (mentor)
Giant if debug.mpsafenet=0, as any points that require synchronization
in the SMPng world also required it in the Giant-world:
- inpcb locks (including IPv6)
- inpcbinfo locks (including IPv6)
- dummynet subsystem lock
- ipfw2 subsystem lock
- Assert the mutex in NG_IDHASH_FIND() since the mutex is required to
safely walk the node lists in the ng_ID_hash table.
- Acquire the ng_nodelist_mtx when walking ng_allnodes or ng_allhooks
to generate state dump output from the netgraph sysctls.
the socket buffer having its limits adjusted. sbreserve() now acquires
the lock before calling sbreserve_locked(). In soreserve(), acquire
socket buffer locks across read-modify-writes of socket buffer fields,
and calls into sbreserve/sbrelease; make sure to acquire in keeping
with the socket buffer lock order. In tcp_mss(), acquire the socket
buffer lock in the calling context so that we have atomic read-modify
-write on buffer sizes.
smp_rendezvous() to ensure we run on the BSP. This reverts rev 1.128.
Add a comment indicating that MI code should be the one that runs all
shutdown functions on the BSP with the APs halted. This should work
around problems in power off while waiting for the MI code to be improved.
waiting for the socket to connect and use msleep() on the socket
mute rather than tsleep(). Acquire socket buffer mutexes around
read-modify-write of socket buffer flags.
The general UMA lock is a recursion-allowed lock because
there is a code path where, while we're still configured
to use startup_alloc() for backend page allocations, we
may end up in uma_reclaim() which calls zone_foreach(zone_drain),
which grabs uma_mtx, only to later call into startup_alloc()
because while freeing we needed to allocate a bucket. Since
startup_alloc() also takes uma_mtx, we need to be able to
recurse on it.
This exact explanation also added as comment above mtx_init().
Trace showing recursion reported by: Peter Holm <peter-at-holm.cc>
originated on RELENG_4 and was ported to -CURRENT.
The scoreboarding code was obtained from OpenBSD, and many
of the remaining changes were inspired by OpenBSD, but not
taken directly from there.
You can enable/disable sack using net.inet.tcp.do_sack. You can
also limit the number of sack holes that all senders can have in
the scoreboard with net.inet.tcp.sackhole_limit.
Reviewed by: gnn
Obtained from: Yahoo! (Mohan Srinivasan, Jayanth Vijayaraghavan)
This should fix problems with older SMP systems that only have ISA/EISA
IRQs when routing virgin PCI interrupts as well as on other boxes whose
MADT does not have any interrupt override entries for ISA IRQs that are
used to route PCI interrupts even in APIC mode.
actually used. For most ACPI devices this means deferring the call
until bus_alloc_resource().
- Add a function acpi_config_intr() to call BUS_CONFIG_INTR() for an
ACPI IRQ resource using the trigger mode and polarity information
stored in the ACPI resource object.
- Add a function acpi_lookup_irq_resource() to lookup the ACPI IRQ
resource that corresponds to a specified rid and new-bus resource.
- Have the ACPI PCI bridge driver call BUS_CONFIG_INTR() on interrupts
that it routes through link devices.
- Remove needactivate variable from acpi_alloc_resource() by changing the
function not modify the flags variable but just mask off RF_ACTIVE when
calling rman_reserve_resource().
Reviewed by: njl (1, an earlier version)
- Allow ioapic_set_{nmi,smi,extint}() to be called multiple times on the
same pin so long as the pin's mode is the same as the mode being
requested.
- Add a notion of bus type for the interrupt associated with interrupt pin.
This is needed so that we can force all EISA interrupts to be active high
in the forthcoming ioapic_config_intr().
- Fix a bug for EISA systems that didn't remap IRQs. This would have broken
EISA systems that tried to disable mixed mode for IRQ 0.
case of NFS mounted swap, so do not try to dereference it.
While we're here, brucify the printf() call which happens when we
time out on acquisition of vm_page_queue_mtx.
PR: kern/67898
Submitted by: bde (style)
device associated with any PCI devices that are enumerated in the ACPI
tree when adding children to an ACPI PCI bus and remove the duplicate
ACPI-only device_t and replace the device_t associated with the handle with
the ACPI and PCI aware device_t.
Several changes:
* Implement read for ulpt.
* If the device is not opened for reading, occasionally drain any
data the printer might have (but don't hammer the printer with reads).
* Lower the buffer size to one page.
The driver seems to work with more printers now.
Obtained from: NetBSD
depending on namespace pollution in <sys/vnode.h> for the definition
of mutex interfaces used in SOCKBUF_*LOCK().
Sorted includes.
Removed unused includes.
the SS_NBIO flag from the parent socket to the child socket during an
accept() operation.
The file descriptor O_NONBLOCK flag would have been propagated already
by the fflag assignment, and therefore would have been inconsistent
with the underlying socket's so_state member.
This makes accept() more closely adhere to the API contract we effectively
outline in the manual page. Note also that Linux continues to differ here;
O_NONBLOCK is not propagated. The other BSDs do propagate the flag, as
does Solaris. The Single UNIX Specification does not offer specific
advice on this issue.
PR: kern/45733
Requested by: Jayanth Vijayaraghavan
Reviewed by: rwatson
%di will already point to the character after the nul char when the
'repnz scasb' terminates.
Submitted by: Tom Cosgrove tom dot cosgrove at arches-consulting dot com
- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.
Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
