Re: [PATCH 1/2] t5516/t5601: avoid using `localhost` for failing HTTPS requests

From: "Ævar Arnfjörð Bjarmason" <avarab@gmail.com>
To: Johannes Schindelin via GitGitGadget <gitgitgadget@gmail.com>
Cc: git@vger.kernel.org, Derrick Stolee <derrickstolee@github.com>,
	Johannes Schindelin <johannes.schindelin@gmx.de>
Subject: Re: [PATCH 1/2] t5516/t5601: avoid using `localhost` for failing HTTPS requests
Date: Mon, 31 Oct 2022 21:49:55 +0100	[thread overview]
Message-ID: <221031.86v8nz67jn.gmgdl@evledraar.gmail.com> (raw)
In-Reply-To: <25cc0f6d91a9d23eb1b755e1463d672e4958a4e9.1667245639.git.gitgitgadget@gmail.com>

On Mon, Oct 31 2022, Johannes Schindelin via GitGitGadget wrote:

> From: Johannes Schindelin <johannes.schindelin@gmx.de>
>
> In 6dcbdc0d6616 (remote: create fetch.credentialsInUrl config,
> 2022-06-06), we added four test cases that validate various behavior
> around passing credentials as part of the URL (which is considered
> unsafe in general).
>
> These tests do not _actually_ try to connect anywhere, but have to use
> the https:// protocol in order to validate the intended code paths.
>
> However, using `localhost` for such a connection causes several
> problems:
>
> - There might be a web server running on localhost, and we do not
>   actually want to connect to that.
>
> - The DNS resolver, or the local firewall, might take a substantial
>   amount of time (or forever, whichever comes first) to fail to connect,
>   slowing down the test cases unnecessarily.
>
> Let's instead use an IPv4 address that is guaranteed never to offer a
> web server: 224.0.0.1 (which is part of the IP multicast range).
>
> Incidentally, this seems to fix an issue where the tests fail in the
> Windows jobs of Git's CI builds.
> [...]
> diff --git a/t/t5601-clone.sh b/t/t5601-clone.sh
> index 45f0803ed4d..0b386c74818 100755
> --- a/t/t5601-clone.sh
> +++ b/t/t5601-clone.sh
> @@ -72,19 +72,19 @@ test_expect_success 'clone respects GIT_WORK_TREE' '
>  '
>  
>  test_expect_success LIBCURL 'clone warns or fails when using username:password' '
> -	message="URL '\''https://username:<redacted>@localhost/'\'' uses plaintext credentials" &&
> -	test_must_fail git -c transfer.credentialsInUrl=allow clone https://username:password@localhost attempt1 2>err &&
> +	message="URL '\''https://username:<redacted>@224.0.0.1/'\'' uses plaintext credentials" &&
> +	test_must_fail git -c transfer.credentialsInUrl=allow clone https://username:password@224.0.0.1 attempt1 2>err &&
>  	! grep "$message" err &&
>  
> -	test_must_fail git -c transfer.credentialsInUrl=warn clone https://username:password@localhost attempt2 2>err &&
> +	test_must_fail git -c transfer.credentialsInUrl=warn clone https://username:password@224.0.0.1 attempt2 2>err &&
>  	grep "warning: $message" err >warnings &&
>  	test_line_count = 2 warnings &&
>  
> -	test_must_fail git -c transfer.credentialsInUrl=die clone https://username:password@localhost attempt3 2>err &&
> +	test_must_fail git -c transfer.credentialsInUrl=die clone https://username:password@224.0.0.1 attempt3 2>err &&
>  	grep "fatal: $message" err >warnings &&
>  	test_line_count = 1 warnings &&
>  
> -	test_must_fail git -c transfer.credentialsInUrl=die clone https://username:@localhost attempt3 2>err &&
> +	test_must_fail git -c transfer.credentialsInUrl=die clone https://username:@224.0.0.1 attempt3 2>err &&
>  	grep "fatal: $message" err >warnings &&
>  	test_line_count = 1 warnings
>  '

For this one one at least, it eventually gets around to setting up an
actual httpd server with cloning etc. from $HTTPD_URL.

Can't we just do that for both of these tests rather than the the
224.0.0.0 hack? I.e. the root cause is that we're cleverly faking a
not-a-server here, and now we're going to add another somewhat clever
hack on top.

but since the test coverage is for https:// anyway, and we have other
https tests against an actual server...