Before the h2 frame dispatch we only need to check that coming from a
HEADERS frame the very next frame is a CONTINUATION, when the HPACK
block didn't fit in the former.

The CONTINUATION dispatch will then make the stream consistency check.

When a stream times out waiting for window credits, and all the other
streams are broke, declare the whole connection bankrupt.

It replaces idle_send_timeout when the stream is waiting for window
credits before sending more DATA frames. The idle_send_timeout will
still apply to individual writes to the socket, but triggering it is
considered a failure condition (this was already the case).

The two loops are merged into a single one to better deal with the
lack of ordering guarantees of request vs connection stream control
flow window crediting.

Refs #2980

This parameter needs to be somewhat high because web browsers may for
example only credit streams for image resources just enough to parse
metadata like width and height of a picture in order to perform layout
computations and fetch resources more sensitive to latency before the
effective image payloads.

The new meaning for h2->sess->t_idle is now the time since the last
complete frame, except before the first frame where it corresponds to
the creation of the session.

And after issuing or receiving a goaway, stop processing frames when
there are no streams left.

Initially submitted as part of listen sockets management.

Refs #3959

Reverting "retrieving" to "processing" is motivated by the fact that we
process ESI on the client side and we don't know whether we are in fact
retrieving an object from a backend, or just reusing one from the cache.

Even if the word "processing" appears again in the sentence, I think it
is more accurate here.

Better diff with the --word-diff option.

- before coffee
+ after coffee

Spotted by Nils.

All paths to EXP_Remove() raise the OC_F_DYING flag to dismiss the
objcore in HSH_Lookup(). A regular purge used to EXP_Rearm() an oc
for immediate (yet delayed) expiry, leaving it to the expiry thread
to apply this flag and call EXP_Remove().

When the regular purge went from EXP_Rearm() to EXP_Remove() in order
to avoid counting purged objects as expired, the OC_F_DYING flag was
overlooked.

Refs 709f71ab

Ubuntu noble tries to use `fchmodat2` (new syscall) and gets permission denied instead of ENOSYS.

This is a small security risk but it's running inside of circleci containers anyway so i think its acceptable.

See previous commit.

But spread the actual changes in their respective sections.

Quick attempt to address 3f69805f.

The remaining TODO in the changes needs to be addressed, and upgrade
notes are empty.

weird behavior was observed while working on a new feature: This test
would terminate s1 after the delay before varnish could finish reading
all of the chunked body, which would leave an unfinished backend request
in varnishd when it terminated.

Signed-off-by: Asad Sajjad Ahmed <asadsa@varnish-software.com>

Mention the old/new cli_limit values

Polish wordning of bits parameter change

Add param.show example in place of what looks like an unmotivated
param.set -j example