Git Mailing List Archive on lore.kernel.org
 help / color / Atom feed
From: Junio C Hamano <gitster@pobox.com>
To: "András Kucsma via GitGitGadget" <gitgitgadget@gmail.com>
Cc: git@vger.kernel.org, "András Kucsma" <r0maikx02b@gmail.com>
Subject: Re: [PATCH] Fix dir sep handling of GIT_ASKPASS on Windows
Date: Tue, 24 Mar 2020 13:51:17 -0700
Message-ID: <xmqqlfnp1np6.fsf@gitster.c.googlers.com> (raw)
In-Reply-To: <pull.587.git.1584997990694.gitgitgadget@gmail.com> (=?utf-8?Q?=22Andr=C3=A1s?= Kucsma via GitGitGadget"'s message of "Mon, 23 Mar 2020 21:13:10 +0000")

"András Kucsma via GitGitGadget"  <gitgitgadget@gmail.com> writes:

> From: Andras Kucsma <r0maikx02b@gmail.com>
>
> On Windows with git installed through cygwin, GIT_ASKPASS failed to run
> for relative and absolute paths containing only backslashes as directory
> separators.
>
> The reason was that git assumed that if there are no forward slashes in
> the executable path, it has to search for the executable on the PATH.

Also if I were reading the discussion correctly, there was a doubt
about locate_in_PATH() that may not work on Windows for at least two
reasons.  Is it OK to ignore these issues, and if so why?

I know if you have a full path, a broken locate_in_PATH() would be
skipped and won't cause an immediate issue, but this change to make
the code realize that "a\\b" is not asking to search in %PATH% feels
just a beginning of a fix, not the whole fix, at least to me.

> The fix is to look for OS specific directory separators, not just
> forward slashes.

Yes, but it is quite unfortunate that you would use a function that
has to scan the string to the end because it asks for the last one.

Perhaps introduce 

--------------------------------------------------
#ifndef has_dir_sep
static inline int git_has_dir_sep(const char *path)
{
	return !!strchr(path, '/');
}
#define has_dir_sep(path) git_has_dir_sep(path)
#endif
--------------------------------------------------

in <git-compat-util.h>, with a replacement definition in
<compat/win32/path-utils.h> that may read

--------------------------------------------------
#define has_dir_sep(path) win32_has_dir_sep(path)
static inline int has_dir_sep(const char *path)
{
        /* 
         * See how long the non-separator part of the given path is, and
         * if and only if it covers the whole path (i.e. path[len] is NUL),
         * there is no separator in the path---otherwise there is a separaptor.
         */
        size_t len = strcspn(path, "/\\");
        return !!path[len];
}
--------------------------------------------------

and use that instead?


> diff --git a/run-command.c b/run-command.c
> index f5e1149f9b3..9fcc12ebf9c 100644
> --- a/run-command.c
> +++ b/run-command.c
> @@ -421,12 +421,12 @@ static int prepare_cmd(struct argv_array *out, const struct child_process *cmd)
>  	}
>  
>  	/*
> -	 * If there are no '/' characters in the command then perform a path
> -	 * lookup and use the resolved path as the command to exec.  If there
> -	 * are '/' characters, we have exec attempt to invoke the command
> -	 * directly.
> +	 * If there are no dir separator characters in the command then perform
> +	 * a path lookup and use the resolved path as the command to exec. If
> +	 * there are dir separator characters, we have exec attempt to invoke
> +	 * the command directly.
>  	 */
> -	if (!strchr(out->argv[1], '/')) {
> +	if (find_last_dir_sep(out->argv[1]) == NULL) {
>  		char *program = locate_in_PATH(out->argv[1]);
>  		if (program) {
>  			free((char *)out->argv[1]);
>
> base-commit: 274b9cc25322d9ee79aa8e6d4e86f0ffe5ced925

  reply index

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-03-23 21:13 András Kucsma via GitGitGadget
2020-03-24 20:51 ` Junio C Hamano [this message]
2020-03-25 13:45 ` [PATCH v2] " András Kucsma via GitGitGadget
2020-03-25 16:35   ` Torsten Bögershausen
2020-03-25 17:09     ` András Kucsma
2020-03-26 21:16     ` Junio C Hamano
2020-03-26 21:14   ` Junio C Hamano
2020-03-27  0:21     ` András Kucsma
2020-03-27  0:36   ` [PATCH v3] run-command: trigger PATH lookup properly on Cygwin András Kucsma via GitGitGadget
2020-03-27 18:04     ` Junio C Hamano
2020-03-27 18:10       ` András Kucsma
2020-03-27 18:41       ` Andreas Schwab
2020-03-27 21:27         ` Junio C Hamano

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

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

  git send-email \
    --in-reply-to=xmqqlfnp1np6.fsf@gitster.c.googlers.com \
    --to=gitster@pobox.com \
    --cc=git@vger.kernel.org \
    --cc=gitgitgadget@gmail.com \
    --cc=r0maikx02b@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

Git Mailing List Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/git/0 git/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 git git/ https://lore.kernel.org/git \
		git@vger.kernel.org
	public-inbox-index git

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.git


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git