conncache: More efficient implementation of cpool_remove_bundle

[mkauf]

No description provided.

[dfandrich]

Analysis of PR #15292 at 5b234286:

Test http/test_17_ssl_use.py::TestSSLUse::test_17_09_ssl_min_max[TLSv1-0--2] failed, which has NOT been flaky recently, so there could be a real issue in this PR.

Test http/test_17_ssl_use.py::TestSSLUse::test_17_09_ssl_min_max[TLSv1-0--1] failed, which has NOT been flaky recently, so there could be a real issue in this PR.

Test http/test_17_ssl_use.py::TestSSLUse::test_17_09_ssl_min_max[TLSv1-0-0] failed, which has NOT been flaky recently, so there could be a real issue in this PR.

Test http/test_17_ssl_use.py::TestSSLUse::test_17_09_ssl_min_max[TLSv1-1--2] failed, which has NOT been flaky recently, so there could be a real issue in this PR.

Test http/test_17_ssl_use.py::TestSSLUse::test_17_09_ssl_min_max[TLSv1-1--1] failed, which has NOT been flaky recently, so there could be a real issue in this PR.

Test http/test_17_ssl_use.py::TestSSLUse::test_17_09_ssl_min_max[TLSv1-1-0] failed, which has NOT been flaky recently, so there could be a real issue in this PR.

Test http/test_17_ssl_use.py::TestSSLUse::test_17_09_ssl_min_max[TLSv1-2--2] failed, which has NOT been flaky recently, so there could be a real issue in this PR.

There are more failures, but that's enough from Gha.

Generated by Testclutch

[mkauf]

Test http/test_17_ssl_use.py is also failing in other recent pull requests -> ignore