mirror of
https://github.com/strongswan/strongswan.git
synced 2025-10-03 00:00:24 -04:00
kernel-interface: Change reqid if seq. nos. are supported and narrowing occurred
With the sequence numbers we don't have to maintain the reqid to delete the temporary state. One exception is with labels. There we currently only install trap policies with the generic label. SAs created from those don't have policies installed, so we have to reuse the reqid of the trap even if narrowing occurs. And as before, we reuse the reqid without checking traffic selectors if sequence numbers are not supported. Note that if a CHILD_SA is manually initiated (i.e. has no sequence number assigned) right before an acquire is triggered, there are several possible outcomes depending on whether narrowing occurs. If there is no narrowing, the same reqid is assigned and the kernel will remove the temporary SA when the SA is installed (no seq => reqid match). Afterwards, the queued duplicate CHILD_SA is destroyed and the acquire state in the trap manager gets removed. If there is narrowing, a new reqid is allocated, so the installation of the SA will not remove the temporary state. However, due to the narrowing, the duplicate check fails and when the duplicate is installed (with sequence number), the temporary state is deleted (as is the state in the trap manager).
This commit is contained in:
parent
c5b2a8eaa3
commit
46c338a78f
@ -392,12 +392,19 @@ METHOD(kernel_interface_t, alloc_reqid, status_t,
|
|||||||
/* search by reqid if given */
|
/* search by reqid if given */
|
||||||
entry = this->reqids->get(this->reqids, tmpl);
|
entry = this->reqids->get(this->reqids, tmpl);
|
||||||
}
|
}
|
||||||
if (entry && entry_equals_selectors(entry, tmpl))
|
|
||||||
|
/* if the IPsec stack supports sequence numbers for acquires, we can
|
||||||
|
* allocate a new reqid if narrowing occurred (otherwise, we get the same
|
||||||
|
* reqid back anyway). if not, we want to reuse the existing reqid of
|
||||||
|
* the trap polices and explicitly don't want to match the traffic
|
||||||
|
* selectors. another case where we want to reuse an existing reqid is
|
||||||
|
* when labels are used. because we currently only install policies once
|
||||||
|
* with the generic label, the reqid has to stay the same even if narrowing
|
||||||
|
* occurs. however, in either case we don't want to reuse the reqid if the
|
||||||
|
* additional selectors (e.g. marks) are different */
|
||||||
|
if (entry && (label || !(get_features(this) & KERNEL_ACQUIRE_SEQ)) &&
|
||||||
|
entry_equals_selectors(entry, tmpl))
|
||||||
{
|
{
|
||||||
/* we don't require a traffic selector match for existing reqids,
|
|
||||||
* as we want to reuse a reqid for trap-triggered policies that
|
|
||||||
* got narrowed during negotiation, but we don't want to reuse the
|
|
||||||
* reqid if the additional selectors (e.g. marks) are different */
|
|
||||||
reqid_entry_destroy(tmpl);
|
reqid_entry_destroy(tmpl);
|
||||||
}
|
}
|
||||||
else
|
else
|
||||||
|
@ -81,6 +81,8 @@ enum kernel_feature_t {
|
|||||||
KERNEL_POLICY_SPI = (1<<4),
|
KERNEL_POLICY_SPI = (1<<4),
|
||||||
/** IPsec backend reports use time per SA via query_sa() */
|
/** IPsec backend reports use time per SA via query_sa() */
|
||||||
KERNEL_SA_USE_TIME = (1<<5),
|
KERNEL_SA_USE_TIME = (1<<5),
|
||||||
|
/** IPsec backend associates acquires and SAs with a sequence number */
|
||||||
|
KERNEL_ACQUIRE_SEQ = (1<<6),
|
||||||
};
|
};
|
||||||
|
|
||||||
/**
|
/**
|
||||||
|
@ -2163,13 +2163,14 @@ child_sa_t *child_sa_create(host_t *me, host_t *other, child_cfg_t *config,
|
|||||||
if (!this->reqid)
|
if (!this->reqid)
|
||||||
{
|
{
|
||||||
/* reuse old reqid if we are rekeying an existing CHILD_SA and when
|
/* reuse old reqid if we are rekeying an existing CHILD_SA and when
|
||||||
* initiating a trap policy. While the reqid cache would find the same
|
* initiating a trap policy. the reqid cache will generally find the
|
||||||
* reqid for our selectors, this does not work in a special case: If an
|
* same reqid for our selectors. but this does not work in a special
|
||||||
* SA is triggered by a trap policy, but the negotiated TS get
|
* case: if the IPsec stack does not use sequence numbers for acquires,
|
||||||
* narrowed, we still must reuse the same reqid to successfully
|
* an SA is triggered by a trap policy and the negotiated TS get
|
||||||
* replace the temporary SA on the kernel level. Rekeying such an SA
|
* narrowed, we still must reuse the same reqid to successfully replace
|
||||||
* requires an explicit reqid, as the cache currently knows the original
|
* the temporary SA on the kernel level. however, if sequence numbers
|
||||||
* selectors only for that reqid. */
|
* are supported, the reqid will later get updated in case of narrowing
|
||||||
|
* when alloc_reqid() is called */
|
||||||
if (data->reqid &&
|
if (data->reqid &&
|
||||||
charon->kernel->ref_reqid(charon->kernel, data->reqid) == SUCCESS)
|
charon->kernel->ref_reqid(charon->kernel, data->reqid) == SUCCESS)
|
||||||
{
|
{
|
||||||
|
Loading…
x
Reference in New Issue
Block a user