git@vger.kernel.org mailing list mirror (one of many)
 help / color / mirror / code / Atom feed
From: Junio C Hamano <gitster@pobox.com>
To: "Adam Szkoda via GitGitGadget" <gitgitgadget@gmail.com>
Cc: git@vger.kernel.org, Phillip Wood <phillip.wood123@gmail.com>,
	Adam Szkoda <adaszko@gmail.com>, Fabian Stelzer <fs@gigacodes.de>
Subject: Re: [PATCH v2] ssh signing: better error message when key not in agent
Date: Tue, 24 Jan 2023 09:52:20 -0800	[thread overview]
Message-ID: <xmqq1qnjhlbf.fsf@gitster.g> (raw)
In-Reply-To: <pull.1270.v2.git.git.1674573972087.gitgitgadget@gmail.com> (Adam Szkoda via GitGitGadget's message of "Tue, 24 Jan 2023 15:26:11 +0000")

"Adam Szkoda via GitGitGadget" <gitgitgadget@gmail.com> writes:

> From: Adam Szkoda <adaszko@gmail.com>
>
> When signing a commit with a SSH key, with the private key missing from
> ssh-agent, a confusing error message is produced:
>
>     error: Load key
>     "/var/folders/t5/cscwwl_n3n1_8_5j_00x_3t40000gn/T//.git_signing_key_tmpkArSj7":
>     invalid format? fatal: failed to write commit object
>
> The temporary file .git_signing_key_tmpkArSj7 created by git contains a
> valid *public* key.  The error message comes from `ssh-keygen -Y sign' and
> is caused by a fallback mechanism in ssh-keygen whereby it tries to
> interpret .git_signing_key_tmpkArSj7 as a *private* key if it can't find in
> the agent [1].  A fix is scheduled to be released in OpenSSH 9.1. All that
> needs to be done is to pass an additional backward-compatible option -U to
> 'ssh-keygen -Y sign' call.  With '-U', ssh-keygen always interprets the file
> as public key and expects to find the private key in the agent.
>
> As a result, when the private key is missing from the agent, a more accurate
> error message gets produced:
>
>     error: Couldn't find key in agent
>
> [1] https://bugzilla.mindrot.org/show_bug.cgi?id=3429
>
> Signed-off-by: Adam Szkoda <adaszko@gmail.com>
> ---

Well explained.

> Fetch-It-Via: git fetch https://github.com/gitgitgadget/git pr-git-1270/radicle-dev/maint-v2
> Pull-Request: https://github.com/git/git/pull/1270
>
> Range-diff vs v1:
>
>  1:  0ce06076242 < -:  ----------- ssh signing: better error message when key not in agent
>  -:  ----------- > 1:  03dfca79387 ssh signing: better error message when key not in agent

This is a fairly useless range-diff.

Even when a range-diff shows the differences in the patches,
mechanically generated range-diff can only show _what_ changed.  It
is helpful to explain the changes in your own words to highlight
_why_ such changes are done, and this place after the "---" line
and the diffstat we see below is the place to do so.

Does GitGitGadget allow its users to describe the differences since
the previous iteration yourself?

>  gpg-interface.c | 15 ++++++++++-----
>  1 file changed, 10 insertions(+), 5 deletions(-)
>
> diff --git a/gpg-interface.c b/gpg-interface.c
> index f877a1ea564..33899a450eb 100644
> --- a/gpg-interface.c
> +++ b/gpg-interface.c
> @@ -998,6 +998,7 @@ static int sign_buffer_ssh(struct strbuf *buffer, struct strbuf *signature,
>  	char *ssh_signing_key_file = NULL;
>  	struct strbuf ssh_signature_filename = STRBUF_INIT;
>  	const char *literal_key = NULL;
> +	int literal_ssh_key = 0;
>  
>  	if (!signing_key || signing_key[0] == '\0')
>  		return error(
> @@ -1005,6 +1006,7 @@ static int sign_buffer_ssh(struct strbuf *buffer, struct strbuf *signature,
>  
>  	if (is_literal_ssh_key(signing_key, &literal_key)) {
>  		/* A literal ssh key */
> +		literal_ssh_key = 1;
>  		key_file = mks_tempfile_t(".git_signing_key_tmpXXXXXX");
>  		if (!key_file)
>  			return error_errno(
> @@ -1036,11 +1038,14 @@ static int sign_buffer_ssh(struct strbuf *buffer, struct strbuf *signature,
>  	}
>  
>  	strvec_pushl(&signer.args, use_format->program,
> -		     "-Y", "sign",
> -		     "-n", "git",
> -		     "-f", ssh_signing_key_file,
> -		     buffer_file->filename.buf,
> -		     NULL);
> +			"-Y", "sign",
> +			"-n", "git",
> +			"-f", ssh_signing_key_file,
> +			NULL);

Please avoid making a pointless indentation change like this.  We do
not pass filename yet with this pushl(), because ...

> +	if (literal_ssh_key) {
> +		strvec_push(&signer.args, "-U");
> +	}

... when we give a literal key, we want to insert "-U" in front, and then

> +	strvec_push(&signer.args, buffer_file->filename.buf);

... the filename.  Which makes sense.

The insertion of "-U" is a single statement as the body of a if()
statement.  We do not want {} around it, by the way.

Other than that, nicely done.  Thanks.

>  	sigchain_push(SIGPIPE, SIG_IGN);
>  	ret = pipe_command(&signer, NULL, 0, NULL, 0, &signer_stderr, 0);
>
> base-commit: 844ede312b4e988881b6e27e352f469d8ab80b2a

  reply	other threads:[~2023-01-24 17:52 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-01-18  8:17 [PATCH] ssh signing: better error message when key not in agent Adam Szkoda via GitGitGadget
2023-01-18 11:10 ` Phillip Wood
2023-01-18 14:34   ` Phillip Wood
2023-01-18 15:28     ` Adam Szkoda
2023-01-18 16:29       ` Phillip Wood
2023-01-20  9:03         ` Fabian Stelzer
2023-01-23  9:33           ` Phillip Wood
2023-01-23 10:02             ` Fabian Stelzer
2023-01-23 16:17               ` Adam Szkoda
2023-01-24 15:26 ` [PATCH v2] " Adam Szkoda via GitGitGadget
2023-01-24 17:52   ` Junio C Hamano [this message]
2023-01-25 12:46     ` Adam Szkoda
2023-01-25 17:04       ` Junio C Hamano
2023-01-25 17:17       ` Junio C Hamano
2023-01-25 21:42       ` Eric Sunshine
2023-01-25 22:22         ` Junio C Hamano
2023-02-15  1:22           ` Eric Sunshine
2023-01-25 12:40   ` [PATCH v3] " Adam Szkoda via GitGitGadget

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: http://vger.kernel.org/majordomo-info.html

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=xmqq1qnjhlbf.fsf@gitster.g \
    --to=gitster@pobox.com \
    --cc=adaszko@gmail.com \
    --cc=fs@gigacodes.de \
    --cc=git@vger.kernel.org \
    --cc=gitgitgadget@gmail.com \
    --cc=phillip.wood123@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://80x24.org/mirrors/git.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).