This change sweeps remaining lock decisions based on if locked or not by moving
utility functions of rtentry updates from rtsock.c and ensuring holding the
rt_lock. It also improves the atomicity of a update of a rtentry.
Thanks to removal of LIST_ENTRY of struct route, rtcaches are accessed only by
their users. And in existing usages a rtcache is guranteed to be not accessed
simultaneously. So the rtcache framework doesn't need any exclusion controls
in itself.
The change introduces a global generation counter that is incremented when any
routes have been added or deleted. When a rtcache caches a rtentry into itself,
it also stores a snapshot of the generation counter. If the snapshot equals to
the global counter, the cache is still valid, otherwise invalidated.
One drawback of the change is that all rtcaches of all protocol families are
invalidated when any routes of any protocol families are added or deleted.
If that matters, we should have separate generation counters based on
protocol families.
This change removes LIST_ENTRY from struct route, which fixes a part of
PR kern/52515.
This change needs a tweak in route_output_change to unbreak route
change commands (e.g., route change -inet6 default -reject).
PR kern/52077 (s-yamaguchi@IIJ and ozaki-r@)
See the following descriptions for details.
Proposed on tech-kern and tech-net
Overview
--------
We protect the routing table with a rwock and protect
rtcaches with another rwlock. Each rtentry is protected
from being freed or updated via reference counting and psref.
Global rwlocks
--------------
There are two rwlocks; one for the routing table (rt_lock) and
the other for rtcaches (rtcache_lock). rtcache_lock covers
all existing rtcaches; there may have room for optimizations
(future work).
The locking order is rtcache_lock first and rt_lock is next.
rtentry references
------------------
References to an rtentry is managed with reference counting
and psref. Either of the two mechanisms is used depending on
where a rtentry is obtained. Reference counting is used when
we obtain a rtentry from the routing table directly via
rtalloc1 and rtrequest{,1} while psref is used when we obtain
a rtentry from a rtcache via rtcache_* APIs. In both cases,
a caller can sleep/block with holding an obtained rtentry.
The reasons why we use two different mechanisms are (i) only
using reference counting hurts the performance due to atomic
instructions (rtcache case) (ii) ease of implementation;
applying psref to APIs such rtaloc1 and rtrequest{,1} requires
additional works (adding a local variable and an argument).
We will finally migrate to use only psref but we can do it
when we have a lockless routing table alternative.
Reference counting for rtentry
------------------------------
rt_refcnt now doesn't count permanent references such as for
rt_timers and rtcaches, instead it is used only for temporal
references when obtaining a rtentry via rtalloc1 and rtrequest{,1}.
We can do so because destroying a rtentry always involves
removing references of rt_timers and rtcaches to the rtentry
and we don't need to track such references. This also makes
it easy to wait for readers to release references on deleting
or updating a rtentry, i.e., we can simply wait until the
reference counter is 0 or 1. (If there are permanent references
the counter can be arbitrary.)
rt_ref increments a reference counter of a rtentry and rt_unref
decrements it. rt_ref is called inside APIs (rtalloc1 and
rtrequest{,1} so users don't need to care about it while
users must call rt_unref to an obtained rtentry after using it.
rtfree is removed and we use rt_unref and rt_free instead.
rt_unref now just decrements the counter of a given rtentry
and rt_free just tries to destroy a given rtentry.
See the next section for destructions of rtentries by rt_free.
Destructions of rtentries
-------------------------
We destroy a rtentry only when we call rtrequst{,1}(RTM_DELETE);
the original implementation can destroy in any rtfree where it's
the last reference. If we use reference counting or psref, it's
easy to understand if the place that a rtentry is destroyed is
fixed.
rt_free waits for references to a given rtentry to be released
before actually destroying the rtentry. rt_free uses a condition
variable (cv_wait) (and psref_target_destroy for psref) to wait.
Unfortunately rtrequst{,1}(RTM_DELETE) can be called in softint
that we cannot use cv_wait. In that case, we have to defer the
destruction to a workqueue.
rtentry#rt_cv, rtentry#rt_psref and global variables
(see rt_free_global) are added to conduct the procedure.
Updates of rtentries
--------------------
One difficulty to use refcnt/psref instead of rwlock for rtentry
is updates of rtentries. We need an additional mechanism to
prevent readers from seeing inconsistency of a rtentry being
updated.
We introduce RTF_UPDATING flag to rtentries that are updating.
While the flag is set to a rtentry, users cannot acquire the
rtentry. By doing so, we avoid users to see inconsistent
rtentries.
There are two options when a user tries to acquire a rtentry
with the RTF_UPDATING flag; if a user runs in softint context
the user fails to acquire a rtentry (NULL is returned).
Otherwise a user waits until the update completes by waiting
on cv.
The procedure of a updater is simpler to destruction of
a rtentry. Wait on cv (and psref) and after all readers left,
proceed with the update.
Global variables (see rt_update_global) are added to conduct
the procedure.
Currently we apply the mechanism to only RTM_CHANGE in
rtsock.c. We would have to apply other codes. See
"Known issues" section.
psref for rtentry
-----------------
When we obtain a rtentry from a rtcache via rtcache_* APIs,
psref is used to reference to the rtentry.
rtcache_ref acquires a reference to a rtentry with psref
and rtcache_unref releases the reference after using it.
rtcache_ref is called inside rtcache_* APIs and users don't
need to take care of it while users must call rtcache_unref
to release the reference.
struct psref and int bound that is needed for psref is
embedded into struct route. By doing so we don't need to
add local variables and additional argument to APIs.
However this adds another constraint to psref other than
reference counting one's; holding a reference of an rtentry
via a rtcache is allowed by just one caller at the same time.
So we must not acquire a rtentry via a rtcache twice and
avoid a recursive use of a rtcache. And also a rtcache must
be arranged to be used by a LWP/softint at the same time
somehow. For IP forwarding case, we have per-CPU rtcaches
used in softint so the constraint is guaranteed. For a h
rtcache of a PCB case, the constraint is guaranteed by the
solock of each PCB. Any other cases (pf, ipf, stf and ipsec)
are currently guaranteed by only the existence of the global
locks (softnet_lock and/or KERNEL_LOCK). If we've found the
cases that we cannot guarantee the constraint, we would need
to introduce other rtcache APIs that use simple reference
counting.
psref of rtcache is created with IPL_SOFTNET and so rtcache
shouldn't used at an IPL higher than IPL_SOFTNET.
Note that rtcache_free is used to invalidate a given rtcache.
We don't need another care by my change; just keep them as
they are.
Performance impact
------------------
When NET_MPSAFE is disabled the performance drop is 3% while
when it's enabled the drop is increased to 11%. The difference
comes from that currently we don't take any global locks and
don't use psref if NET_MPSAFE is disabled.
We can optimize the performance of the case of NET_MPSAFE
on by reducing lookups of rtcache that uses psref;
currently we do two lookups but we should be able to trim
one of two. This is a future work.
Known issues
------------
There are two known issues to be solved; one is that
a caller of rtrequest(RTM_ADD) may change rtentry (see rtinit).
We need to prevent new references during the update. Or
we may be able to remove the code (perhaps, need more
investigations).
The other is rtredirect that updates a rtentry. We need
to apply our update mechanism, however it's not easy because
rtredirect is called in softint and we cannot apply our
mechanism simply. One solution is to defer rtredirect to
a workqueue but it requires some code restructuring.
Some functions use rt_walktree to scan the routing table and delete
matched routes. However, we shouldn't use rt_walktree to delete
routes because rt_walktree is recursive to the routing table (radix
tree) and isn't friendly to MP-ification. rt_walktree allows a caller
to pass a callback function to delete an matched entry. The callback
function is called from an API of the radix tree (rn_walktree) but
also calls an API of the radix tree to delete an entry.
This change adds a new API of the radix tree, rn_search_matched,
which returns a matched entry that is selected by a callback
function passed by a caller and the caller itself deletes the
entry. By using the API, we can avoid the recursive form.
We want to ensure that a rtentry is referenced by nobody after
RTM_DELETE (except for the caller). However, rt_timer could
have a reference to the rtentry after that.
If NET_MPSAFE is enabled, don't hold KERNEL_LOCK and softnet_lock in
part of the network stack such as IP forwarding paths. The aim of the
change is to make it easy to test the network stack without the locks
and reduce our local diffs.
By default (i.e., if NET_MPSAFE isn't enabled), the locks are held
as they used to be.
Reviewed by knakahara@
This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.
Timers (such as nd6_timer) typically free/destroy some data in callout
(softint). If we apply psz/psref for such data, we cannot do free/destroy
process in there because synchronization of psz/psref cannot be used in
softint. So run timer callbacks in workqueue works (normal LWP context).
Doing workqueue_enqueue a work twice (i.e., call workqueue_enqueue before
a previous task is scheduled) isn't allowed. For nd6_timer and
rt_timer_timer, this doesn't happen because callout_reset is called only
from workqueue's work. OTOH, ip{,6}flow_slowtimo's callout can be called
before its work starts and completes because the callout is periodically
called regardless of completion of the work. To avoid such a situation,
add a flag for each protocol; the flag is set true when a work is
enqueued and set false after the work finished. workqueue_enqueue is
called only if the flag is false.
Proposed on tech-net and tech-kern.
Addresses of an interface (struct ifaddr) have a (reverse) pointer of an
interface object (ifa->ifa_ifp). If the addresses are surely freed when
their interface is destroyed, the pointer is always valid and we don't
need a tweak of replacing the pointer to if_index like mbuf.
In order to make sure the assumption, the following changes are required:
- Deactivate the interface at the firstish of if_detach. This prevents
in6_unlink_ifa from saving multicast addresses (wrongly)
- Invalidate rtcache(s) and clear a rtentry referencing an address on
RTM_DELETE. rtcache(s) may delay freeing an address
- Replace callout_stop with callout_halt of DAD timers to ensure stopping
such timers in if_detach
We no longer need to change rtentry below if_output.
The change makes it clear where rtentries are changed (or not)
and helps forthcoming locking (os psrefing) rtentries.
rt_gwroute of rtentry is a reference to a rtentry of the gateway
for a rtentry with RTF_GATEWAY. That was used by L2 (arp and ndp)
to look up L2 addresses. By separating L2 nexthop caches, we don't
need a route for the purpose and we can stop using rt_gwroute.
By doing so, we can reduce referencing and modifying rtentries,
which makes it easy to apply a lock (and/or psref) to the
routing table and rtentries.
One issue to do this is to keep RTF_REJECT behavior. It seems it
was broken when we moved rtalloc1 things from L2 output routines
(e.g., ether_output) to ip_hresolv_output, but (fortunately?)
it works unexpectedly. What we mistook are:
- RTF_REJECT was checked for any routes in L2 output routines,
but in ip_hresolv_output it is checked only when the route
is RTF_GATEWAY
- The RTF_REJECT check wasn't copied to IPv6 (nd6_output)
It seems that rt_gwroute checks hid the mistakes and it looked
work (unexpectedly) and removing rt_gwroute checks unveil the
issue. So we need to fix RTF_REJECT checks in ip_hresolv_output
and also add them to nd6_output.
One more point we have to care is returning an errno; we need
to mimic looutput behavior. Originally RTF_REJECT check was
done either in L2 output routines or in looutput. The latter is
applied when a reject route directs to a loopback interface.
However, now RTF_REJECT check is done before looutput so to keep
the original behavior we need to return an errno which looutput
chooses. Added rt_check_reject_route does such tweaks.
show arptab command of ddb is now inappropriate because it actually dumps
routes but arp entries aren't routes anymore. So rename it to show routes
and move the code from if_arp.c to route.c.
ok christos@
