crypto_{new.free}session() to be called with the "crypto_mtx"
spinlock held.
This doesn't change much for now because these functions acquire
the said mutex first on entry now, but at least it keeps the nasty
locks local to the opencrypto core.
-RFC2104 says that the block size of the hash algorithm must be used
for key/ipad/opad calculations. While formerly all ciphers used a block
length of 64, SHA384 and SHA512 use 128 bytes. So we can't use the
HMAC_BLOCK_LEN constant anymore. Add a new field to "struct auth_hash"
for the per-cipher blocksize.
-Due to this, there can't be a single "CRYPTO_SHA2_HMAC" external name
anymore. Replace this by 3 for the 3 different keysizes.
This was done by Open/FreeBSD before.
-Also fix the number of authenticator bits used tor ESP and AH to
conform to RFC4868, and remove uses of AH_HMAC_HASHLEN which did
assume a fixed authenticator size of 12 bytes.
FAST_IPSEC will not interoperate with KAME IPSEC anymore if sha2 is used,
because the latter doesn't implement these standards. It should
interoperate with at least modern Free/OpenBSD now.
(I've only tested with NetBSD-current/FAST_IPSEC on both ends.)
decompression:
-seperate the IPCOMP specific rule that compression must not grow the
data from general compression semantics: Introduce a special name
CRYPTO_DEFLATE_COMP_NOGROW/comp_algo_deflate_nogrow to describe
the IPCOMP semantics and use it there. (being here, fix the check
so that equal size is considered failure as well as required by
RFC2393)
Customers of CRYPTO_DEFLATE_COMP/comp_algo_deflate now always get
deflated data back, even if they are not smaller than the original.
-allow to pass a "size hint" to the DEFLATE decompression function
which is used for the initial buffer allocation. Due to the changes
done there, additional allocations and extra copies are avoided if the
initial allocation is sufficient. Set the size hint to MCLBYTES (=2k)
in IPCOMP which should be good for many use cases.
into account that the extension header type is not in the extension
header itself but in the previous one -- this makes a difference
because (a) the length field is different for AH than for all others
and (b) the offset of the "next type" field isn't the same in primary
and extension headers.
(I didn't manage to trigger the bug in my tests, no extension headers
besides AH made it to that point. Didn't try hard enough -- the fix
is still valid.)
they are initialized -- during lifetime, no changes are expected
plus some constification of input to comparision functions etc
mostly required by the former
the refcount in the (global) policies gets decremented
(This apparently was missed when the policy cache code was copied
over from KAME IPSEC.)
From Wolfgang Stukenbrock per PR kern/44410, just fixed differently
to avoid unecessary differences to KAME.
Before, setting the IP_RAWOUTPUT flag did imply that the ip_id
(the fragmentation thing) was used as-is.
Now, a new ID is diced unless the new IP_NOIPNEWID flag is set.
The ip_id is part of the data which are used to calculate the hash
for AH, so set the IP_NOIPNEWID flag to make sure the IP header
is not modified behind AH's back. Otherwise, the recipient will detect
a checksum mismatch and discard the packet.
everywhere splsoftnet() was used before, to fix MP concurrency problems
-pull KERNEL_LOCK where ip(6)_output() is called, as this is what
the network stack (unfortunately) expects, in particular to avoid
races for packets in the interface send queues
From Wolfgang Stukenbrock per PR kern/44418, with the application
of KERNEL_LOCK to what I think are the essential points, tested
on a dual-core i386.
1) RFC2367 says in 2.3.3 Address Extension: "All non-address
information in the sockaddrs, such as sin_zero for AF_INET sockaddrs,
and sin6_flowinfo for AF_INET6 sockaddrs, MUST be zeroed out."
the IPSEC_NAT_T code was expecting the port information it needs
to be conveyed in the sockaddr instead of exclusively by
SADB_X_EXT_NAT_T_SPORT and SADB_X_EXT_NAT_T_DPORT,
and was not zeroing out the port information in the non-nat-traversal
case.
Since it was expecting the port information to reside in the sockaddr
it could get away with (re)setting the ports after starting to use them.
-> Set the natt ports before setting the SA mature.
2) RFC3947 has two Original Address fields, initiator and responder,
so we need SADB_X_EXT_NAT_T_OAI and SADB_X_EXT_NAT_T_OAR and not just
SADB_X_EXT_NAT_T_OA
The change has been created using vanhu's patch for FreeBSD as reference.
Note that establishing actual nat-t sessions has not yet been tested.
Likely fixes the following:
PR bin/41757
PR net/42592
PR net/42606
"initializing IPsec..."" done" is of somewhat limited value.
(I normally wouldn't care; but on my box the (root) uhub(4)s attach
between the first and last portion of the line.)
on the amount of physical memory and limited by NMBCLUSTERS if present.
Architectures without direct mapping also limit it based on the kmem_map
size, which is used as backing store. On i386 and ARM, the maximum KVA
used for mbuf clusters is limited to 64MB by default.
The old default limits and limits based on GATEWAY have been removed.
key_registered_sb_max is hard-wired to a value derived from 2048
clusters.
- Socket layer becomes MP safe.
- Unix protocols become MP safe.
- Allows protocol processing interrupts to safely block on locks.
- Fixes a number of race conditions.
With much feedback from matt@ and plunky@.
(actually splnet) and condvars instead of tsleep/wakeup. Fix a few
miscellaneous problems and add some debugging printfs while there.
Restore set of CRYPTO_F_DONE in crypto_done() which was lost at some
point after this code came from FreeBSD -- it made it impossible to wait
properly for a condition.
Add flags analogous to the "crp" flags to the key operation's krp struct.
Add a new flag, CRYPTO_F_ONRETQ which tells us a request finished before
the kthread had a chance to dequeue it and call its callback -- this was
letting requests stick on the queues before even though done and copied
out.
Callers of crypto_newsession() or crypto_freesession() must now take the
mutex. Change netipsec to do so. Dispatch takes the mutex itself as
needed.
This was tested fairly extensively with the cryptosoft backend and lightly
with a new hardware driver. It has not been tested with FAST_IPSEC; I am
unable to ascertain whether FAST_IPSEC currently works at all in our tree.
pjd@FreeBSD.ORG, ad@NetBSD.ORG, and darran@snark.us pointed me in the
right direction several times in the course of this. Remaining bugs
are mine alone.
Only record an IPSEC_OUT_DONE tag when we have finished the processing
In ip{,6}_output, check this tag to know if we have already processed this
packet.
Remove some dead code (IPSEC_PENDING_TDB is not used in NetBSD)
Fix pr/36870
in IPsec processing that is dependent on protocol and/or port can be bypassed.
Bug report, analysis and initial fix from Karl Knutsson.
Final patch and ok from degroote@
the seq used by the request. It will improve consistency with the answer of SADB_GET
request and helps some applications which relies both on seq and pid.
Reported by Karl Knutsson by pr/36119.
socket. If we don't make an exact match, we may use a cached rule which
has lower priority than a rule that would otherwise have matched the
packet.
Code submitted by Karl Knutsson in PR/36051
