several critical bugs, including race conditions and lock order issues:
Replace the single rwlock, ifnet_lock, with two locks, an rwlock and an
sxlock. Either can be held to stablize the lists and indexes, but both
are required to write. This allows the list to be held stable in both
network interrupt contexts and sleepable user threads across sleeping
memory allocations or device driver interactions. As before, writes to
the interface list must occur from sleepable contexts.
Reviewed by: bz, julian
MFC after: 3 days
vnet.h, we now use jails (rather than vimages) as the abstraction
for virtualization management, and what remained was specific to
virtual network stacks. Minor cleanups are done in the process,
and comments updated to reflect these changes.
Reviewed by: bz
Approved by: re (vimage blanket)
network stacks, VNET_SYSINIT:
- Add VNET_SYSINIT and VNET_SYSUNINIT macros to declare events that will
occur each time a network stack is instantiated and destroyed. In the
!VIMAGE case, these are simply mapped into regular SYSINIT/SYSUNINIT.
For the VIMAGE case, we instead use SYSINIT's to track their order and
properties on registration, using them for each vnet when created/
destroyed, or immediately on module load for already-started vnets.
- Remove vnet_modinfo mechanism that existed to serve this purpose
previously, as well as its dependency scheme: we now just use the
SYSINIT ordering scheme.
- Implement VNET_DOMAIN_SET() to allow protocol domains to declare that
they want init functions to be called for each virtual network stack
rather than just once at boot, compiling down to DOMAIN_SET() in the
non-VIMAGE case.
- Walk all virtualized kernel subsystems and make use of these instead
of modinfo or DOMAIN_SET() for init/uninit events. In some cases,
convert modular components from using modevent to using sysinit (where
appropriate). In some cases, do minor rejuggling of SYSINIT ordering
to make room for or better manage events.
Portions submitted by: jhb (VNET_SYSINIT), bz (cleanup)
Discussed with: jhb, bz, julian, zec
Reviewed by: bz
Approved by: re (VIMAGE blanket)
unused custom mutex/condvar-based sleep locks with two locks: an
rwlock (for non-sleeping use) and sxlock (for sleeping use). Either
acquired for read is sufficient to stabilize the vnet list, but both
must be acquired for write to modify the list.
Replace previous no-op read locking macros, used in various places
in the stack, with actual locking to prevent race conditions. Callers
must declare when they may perform unbounded sleeps or not when
selecting how to lock.
Refactor vnet sysinits so that the vnet list and locks are initialized
before kernel modules are linked, as the kernel linker will use them
for modules loaded by the boot loader.
Update various consumers of these KPIs based on whether they may sleep
or not.
Reviewed by: bz
Approved by: re (kib)
(DPCPU), as suggested by Peter Wemm, and implement a new per-virtual
network stack memory allocator. Modify vnet to use the allocator
instead of monolithic global container structures (vinet, ...). This
change solves many binary compatibility problems associated with
VIMAGE, and restores ELF symbols for virtualized global variables.
Each virtualized global variable exists as a "reference copy", and also
once per virtual network stack. Virtualized global variables are
tagged at compile-time, placing the in a special linker set, which is
loaded into a contiguous region of kernel memory. Virtualized global
variables in the base kernel are linked as normal, but those in modules
are copied and relocated to a reserved portion of the kernel's vnet
region with the help of a the kernel linker.
Virtualized global variables exist in per-vnet memory set up when the
network stack instance is created, and are initialized statically from
the reference copy. Run-time access occurs via an accessor macro, which
converts from the current vnet and requested symbol to a per-vnet
address. When "options VIMAGE" is not compiled into the kernel, normal
global ELF symbols will be used instead and indirection is avoided.
This change restores static initialization for network stack global
variables, restores support for non-global symbols and types, eliminates
the need for many subsystem constructors, eliminates large per-subsystem
structures that caused many binary compatibility issues both for
monitoring applications (netstat) and kernel modules, removes the
per-function INIT_VNET_*() macros throughout the stack, eliminates the
need for vnet_symmap ksym(2) munging, and eliminates duplicate
definitions of virtualized globals under VIMAGE_GLOBALS.
Bump __FreeBSD_version and update UPDATING.
Portions submitted by: bz
Reviewed by: bz, zec
Discussed with: gnn, jamie, jeff, jhb, julian, sam
Suggested by: peter
Approved by: re (kensmith)
Introduce the new flag KNF_NOKQLOCK to allow event callers to be called
without KQ_LOCK mtx held.
- Modify VFS knote calls to always use KNF_NOKQLOCK flag. This is required
for ZFS as its getattr implementation may sleep.
Approved by: re (rwatson)
Reviewed by: kib
MFC after: 2 weeks
IF_ADDR_UNLOCK() across network device drivers when accessing the
per-interface multicast address list, if_multiaddrs. This will
allow us to change the locking strategy without affecting our driver
programming interface or binary interface.
For two wireless drivers, remove unnecessary locking, since they
don't actually access the multicast address list.
Approved by: re (kib)
MFC after: 6 weeks
network stack when reentering the inbound path from netgraph, and
force queueing of mbufs at the outbound netgraph node.
The mechanism relies on two components. First, in netgraph nodes
where outbound path of the network stack calls into netgraph, the
current thread has to be appropriately marked using the new
NG_OUTBOUND_THREAD_REF() macro before proceeding to call further
into the netgraph topology, and unmarked using the
NG_OUTBOUND_THREAD_UNREF() macro before returning to the caller.
Second, netgraph nodes which can potentially reenter the network
stack in the inbound path have to mark their inbound hooks using
NG_HOOK_SET_TO_INBOUND() macro. The netgraph framework will then
detect when there is a danger of a call graph looping back from
outbound to inbound path via netgraph, and defer handing off the
mbufs to the "inbound" node to a worker thread with a clean stack.
In this first pass only the most obvious netgraph nodes have been
updated to ensure no outbound to inbound calls can occur. Nodes
such as ng_ipfw, ng_gif etc. should be further examined whether a
potential for outbound to inbound call looping exists.
This commit changes the layout of struct thread, but due to
__FreeBSD_version number shortage a version bump has been omitted
at this time, nevertheless kernel and modules have to be rebuilt.
Reviewed by: julian, rwatson, bz
Approved by: julian (mentor)
If packet leaves ipfw to other kernel subsystem (dummynet, netgraph, etc)
it carries pointer to matching ipfw rule. If this packet then reinjected back
to ipfw, ruleset processing starts from that rule. If rule was deleted
meanwhile, due to existed race condition panic was possible (as well as
other odd effects like parsing rules in 'reap list').
P.S. this commit changes ABI so userland ipfw related binaries should be
recompiled.
MFC after: 1 month
Tested by: Mikolaj Golub
the ROUTETABLES kernel option thus there is no need to include opt_route.h
anymore in all consumers of vnet.h and no longer depend on it for module
builds.
Remove the hidden include in flowtable.h as well and leave the two
explicit #includes in ip_input.c and ip_output.c.
Vnet modules and protocol domains may now register destructor
functions to clean up and release per-module state. The destructor
mechanisms can be triggered by invoking "vimage -d", or a future
equivalent command which will be provided via the new jail framework.
While this patch introduces numerous placeholder destructor functions,
many of those are currently incomplete, thus leaking memory or (even
worse) failing to stop all running timers. Many of such issues are
already known and will be incrementaly fixed over the next weeks in
smaller incremental commits.
Apart from introducing new fields in structs ifnet, domain, protosw
and vnet_net, which requires the kernel and modules to be rebuilt, this
change should have no impact on nooptions VIMAGE builds, since vnet
destructors can only be called in VIMAGE kernels. Moreover,
destructor functions should be in general compiled in only in
options VIMAGE builds, except for kernel modules which can be safely
kldunloaded at run time.
Bump __FreeBSD_version to 800097.
Reviewed by: bz, julian
Approved by: rwatson, kib (re), julian (mentor)
- Each socket upcall is now invoked with the appropriate socket buffer
locked. It is not permissible to call soisconnected() with this lock
held; however, so socket upcalls now return an integer value. The two
possible values are SU_OK and SU_ISCONNECTED. If an upcall returns
SU_ISCONNECTED, then the soisconnected() will be invoked on the
socket after the socket buffer lock is dropped.
- A new API is provided for setting and clearing socket upcalls. The
API consists of soupcall_set() and soupcall_clear().
- To simplify locking, each socket buffer now has a separate upcall.
- When a socket upcall returns SU_ISCONNECTED, the upcall is cleared from
the receive socket buffer automatically. Note that a SO_SND upcall
should never return SU_ISCONNECTED.
- All this means that accept filters should now return SU_ISCONNECTED
instead of calling soisconnected() directly. They also no longer need
to explicitly clear the upcall on the new socket.
- The HTTP accept filter still uses soupcall_set() to manage its internal
state machine, but other accept filters no longer have any explicit
knowlege of socket upcall internals aside from their return value.
- The various RPC client upcalls currently drop the socket buffer lock
while invoking soreceive() as a temporary band-aid. The plan for
the future is to add a new flag to allow soreceive() to be called with
the socket buffer locked.
- The AIO callback for socket I/O is now also invoked with the socket
buffer locked. Previously sowakeup() would drop the socket buffer
lock only to call aio_swake() which immediately re-acquired the socket
buffer lock for the duration of the function call.
Discussed with: rwatson, rmacklem
previously always pointing to the default vnet context, to a
dynamically changing thread-local one. The currvnet context
should be set on entry to networking code via CURVNET_SET() macros,
and reverted to previous state via CURVNET_RESTORE(). Recursions
on curvnet are permitted, though strongly discuouraged.
This change should have no functional impact on nooptions VIMAGE
kernel builds, where CURVNET_* macros expand to whitespace.
The curthread->td_vnet (aka curvnet) variable's purpose is to be an
indicator of the vnet context in which the current network-related
operation takes place, in case we cannot deduce the current vnet
context from any other source, such as by looking at mbuf's
m->m_pkthdr.rcvif->if_vnet, sockets's so->so_vnet etc. Moreover, so
far curvnet has turned out to be an invaluable consistency checking
aid: it helps to catch cases when sockets, ifnets or any other
vnet-aware structures may have leaked from one vnet to another.
The exact placement of the CURVNET_SET() / CURVNET_RESTORE() macros
was a result of an empirical iterative process, whith an aim to
reduce recursions on CURVNET_SET() to a minimum, while still reducing
the scope of CURVNET_SET() to networking only operations - the
alternative would be calling CURVNET_SET() on each system call entry.
In general, curvnet has to be set in three typicall cases: when
processing socket-related requests from userspace or from within the
kernel; when processing inbound traffic flowing from device drivers
to upper layers of the networking stack, and when executing
timer-driven networking functions.
This change also introduces a DDB subcommand to show the list of all
vnet instances.
Approved by: julian (mentor)
initialize / release netgraph related state in iattach() / idetach()
functions called via the vnet module registration / initialization
framework, instead of initialization / cleanups being done in
mod_event handlers.
While here, introduce a crude hack aimed at preventing ng_ether to
autoattach to ng_eiface ifnets, which are also netgraph nodes already.
Reviewed by: bz
Approved by: julian (mentor)
Clang disallows structs with variable length arrays to be nested inside
other structs, because this is in violation with ISO C99. Even though we
can keep bugging the LLVM folks about this issue, we'd better just fix
our code to not do this. This code seems to be the only code in the
entire source tree that does this.
I haven't tested this patch by using the kernel modules in question, but
Diane Bruce and I have compared disassembled versions of these kernel
modules. We would have expected them to be exactly the same, but due to
randomness in the register allocator and reordering of instructions,
there were some minor differences.
Approved by: julian
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
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.
