process: Don't close all FDs on new spawn

Closes #40990

See merge request !848

Fixes #40990

Signed-off-by: David Goulet <dgoulet@torproject.org>

This reverts commit 4f3aafa1.

Once merged upstream, the CI failed here with:

Warning: Failed to confirm worker threads' start up after timeout.
Warning: Bug:     /builds/tpo/core/tor/build/src/app/tor(_start+0x2a) [0x56404d21ddaa] (on Tor 0.4.9.0-alpha-dev 4f3aafa1)
Warning: Bug:     /builds/tpo/core/tor/build/src/app/tor(cpuworker_init+0x53) [0x56404d373d53] (on Tor 0.4.9.0-alpha-dev 4f3aafa1)
Warning: Bug:     /builds/tpo/core/tor/build/src/app/tor(log_backtrace_impl+0x57) [0x56404d29e1f7] (on Tor 0.4.9.0-alpha-dev 4f3aafa1)
Warning: Bug:     /builds/tpo/core/tor/build/src/app/tor(main+0x19) [0x56404d21dd59] (on Tor 0.4.9.0-alpha-dev 4f3aafa1)
Warning: Bug:     /builds/tpo/core/tor/build/src/app/tor(run_tor_main_loop+0xce) [0x56404d22188e] (on Tor 0.4.9.0-alpha-dev 4f3aafa1)
Warning: Bug:     /builds/tpo/core/tor/build/src/app/tor(threadpool_new+0x295) [0x56404d3e28f5] (on Tor 0.4.9.0-alpha-dev 4f3aafa1)
Warning: Bug:     /builds/tpo/core/tor/build/src/app/tor(tor_assertion_failed_+0x148) [0x56404d2a9248] (on Tor 0.4.9.0-alpha-dev 4f3aafa1)
Warning: Bug:     /builds/tpo/core/tor/build/src/app/tor(tor_main+0x49) [0x56404d21e179] (on Tor 0.4.9.0-alpha-dev 4f3aafa1)
Warning: Bug:     /builds/tpo/core/tor/build/src/app/tor(tor_run_main+0x1e5) [0x56404d221db5] (on Tor 0.4.9.0-alpha-dev 4f3aafa1)
Warning: Bug:     /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xea) [0x7f6aa3b1cd7a] (on Tor 0.4.9.0-alpha-dev 4f3aafa1)
Warning: Bug: Tor 0.4.9.0-alpha-dev (git-4f3aafa1): Assertion 0 failed in threadpool_new at /builds/tpo/core/tor/src/lib/evloop/workqueue.c:641: . Stack trace: (on Tor 0.4.9.0-alpha-dev 4f3aafa1)

We are figuring it out but revert this so we can release an alpha.

excurso authored 4 months ago and David Goulet committed 3 months ago

See issue #40991

This should be a pure refactor to simplify the logic a bit now that we
no longer probe for IPv6 support.

Previously we would incorrectly detect that ipv6 isn't supported if the
ping binary isn't present (as it may not be in a relatively stripped
down container image), or if ICMP packets aren't permitted (as they
often aren't in containers).

We don't really have a need to run these network tests on non-IPv6
systems, so it makes more sense to just run them unconditionally.

Jim Newsome authored 3 months ago and David Goulet committed 3 months ago

This is to work around
https://github.com/docker-library/official-images/issues/17896

Jim Newsome authored 3 months ago and David Goulet committed 3 months ago

* buster is now very old
* all the other jobs run bullseye
* This will facilitate switching to a TPA-maintained image (there is no
  buster image).

Update images

See merge request !846

This is to work around
https://github.com/docker-library/official-images/issues/17896

* buster is now very old
* all the other jobs run bullseye
* This will facilitate switching to a TPA-maintained image (there is no
  buster image).

excurso authored 4 months ago and David Goulet committed 4 months ago

See issue #40989

CI: use a fixed version of chutney

See merge request !843

While chutney currently runs tor's chutney test in its own CI,
it's difficult to guarantee the two won't accidentally diverge.
Probably best to use a fixed version here so that we can control
chutney version bumps and avoid surprise breakage in tor's CI.

This will also free us to intentionally make breaking changes in
chutney (though I don't have any immediate plans for any).

whitespace fix while debugging #40981

See merge request !841

We're about to add a few more steps, which would make the single-line
format a bit unwieldy.

Closes #40984

This reverts commit 3d2f7c34.

Create socket with correct family as given by sockshost

See merge request !836

Conflux circuit building was ignoring the "predicted ports" feature,
which aims to make Tor stop building circuits if there have been
no user requests lately. This bug led to every idle Tor on the
network building and discarding circuits every 30 seconds, which
added overall load to the network, used bandwidth and battery from
clients that weren't actively using their Tor, and kept sockets open
on guards which added connection padding essentially forever.

Bug went in on commit 39c2927d when we added preemptive conflux circuit
pools.

Fixes bug 40981; bugfix on 0.4.8.1-alpha.