| Age | Commit message (Collapse) | Author |
|---|
|
ever used to pass on uint32 (for ipsec). stop that madness and just pass
the uint32, 0 in all cases but the two that pass the ipsec flowinfo.
ok deraadt reyk guenther
|
|
ok mlarkin@
|
|
ok gcc & md5 (alas, no binary change)
|
|
enough.
remove a define in an #ifdef notdef /* obsolete */ - 14 years are enough
|
|
|
|
|
|
and set location on ports as best we can. Take the port mutex too.
|
|
|
|
|
|
ok henning@
|
|
ok claudio reyk
|
|
carp_rewrite_lladdr to overwrite the src lladdr, get the intended src
lladdr before assembling the ethernet header.
carp_rewrite_lladdr -> carp_get_srclladdr
ok reyk claudio
|
|
src lladdr just to copy it from the esrc buffer into the ethernet header
a few lines later, use an esrc pointer to figure out where to copy the
src lladdr from. ok claudio reyk
|
|
|
|
RB lookup tree because the rdomain id is part of the lookup key.
Without this the RB tree gets corrupted and in the worst case a use after
free can happen when the interface is destroyed.
Why the sadl addresses are added to the tree in the first place is something
to reconsider.
OK henning@, mpi@, sthen@
|
|
loop ports and other fabric ports that have logged in to us, and check
that we're still logged in to other fabric ports. Rearrange the update
processing loop so we attach and detach targets last, since we need to get
all the way through before we've identified what's gone missing. Handle
fabric port login errors a bit more usefully too.
|
|
nothing except in_proto_cksum_out() uses it any more, and that's a good
thing. was on tech for 3 months, discussed with many
|
|
don't we all love functions implemented in header files? was under #ifdef
_KERNEL at least.
incremental checksum updates don't really make sense any more, this is
incredibly hard to get right, and doesn't fit the way our kernel deals
with the checksums these days. consequently, nothing uses in_cksum_addword
any more.
was on tech for 3 months, tested by & discussed with many.
|
|
is to check xs->status on completion to make sure it worked.
|
|
|
|
it was RX only, nothing on the TX side
rather complex, might eat up (or more) offloading benefits in many cases
the hardware miscomputes the cksums sometimes, so we don't trust it
claiming a cksum is bad and re-do the cksum verification in sw then...
(but we trust it to not mark bad ones good? hmmmmmmmmm.)
diff was on tech for 3 months, nobody disagreed, everybody who spoke up
agrees or doesn't care.
pretty exactly a revert of the cksum offloading addition years ago, tested
by several, the only report i quickly find now is from David Higgs <higgsd
at gmail dot com>, thanks!
|
|
|
|
|
|
|
|
|
|
keep ifqueue and ifaltq in sync is gone and thus the comment obsolete,
and finally there is no more need to include if_altq.h either
|
|
|
|
|
|
|
|
while there, get rid of the altq ioctls and assciated now obsolete code
|
|
|
|
|
|
of tulip_ifstart if the sendqueue was empty, but only if altq wasn't
compiled in (i. e., that's a _compile time_ decision and not based on
wether altq was actually used).
just use tulip_ifstart all the time, as before in our regular kernels.
kill tulip_ifstart_one completely.
makes sense to sthen
|
|
convince ourselves that that was right to begin with. anyway, begone.
|
|
out the entire codepath is unreachable. glad I'm not our ppp maintainer, he
has work to do.
kill that unreachable code, with & ok claudio
|
|
big WTF regarding the fastq use there while verifying w/ claudio, but
that's for the ppp maintainer and unrelated
|
|
tested by & ok mlarkin@
|
|
file becoming readable increase.
|
|
* altq for loop is just for debugging.
* only used when called for loop interface (not for
* a simplex interface).
*/
bye bye!
|
|
ALTQ has served us well for years and was extremely important not just for
us, but for the entire bandwidth management arena. Back when we got altq,
the subject was not yet well researched and understood, which is why altq
is the framework with pluggable schedulers it is. Kenjiro Cho (kjc@) did an
amazing job there.
Now, more than 10 years later, we do have a good understanding and can use
a simpler framework with just one priority queueing and one bandwidth
shaping mechanism each - the new queueing subsystem. Last not least because
it is incredibly painful to maintain both in parallel, it is time for altq
to depart. Farewell, thanks for many years of good service. Everybody
using any form of "not just fifo" queueing owes Kenjiro a lot. At least
buy him a beer when you meet him.
And, allow me this personal note, thanks Kenjiro, working with you on the
topic has always been a great pleasure and I learned a lot from you. Thanks!
|
|
ok dlg@
|
|
tested by jmatthew@
|
|
|
|
add attributes to it later; as wd33c93 is not a valid device name.
ok miod@
|
|
the process, not just the thread.
ok kettenis@
|
|
Add a check for IPv4-mapped IPv6 destination addresses, like in the most
recent KAME code, for non-connected sockets. This prevents packets from
reaching the wire through the default route, if a reject route
for ::ffff:0.0.0.0/96 isn't present. ok claudio@
|
|
|
|
|
|
manually, ok naddy (in january)
|
|
ok benno lteo naddy (back in january)
|
