git@vger.kernel.org list mirror (unofficial, one of many)
 help / color / mirror / code / Atom feed
From: Jeff King <peff@peff.net>
To: Junio C Hamano <gitster@pobox.com>
Cc: "Nguyễn Thái Ngọc Duy" <pclouds@gmail.com>,
	git@vger.kernel.org, "Stefan Beller" <sbeller@google.com>,
	"Brandon Williams" <bmwill@google.com>,
	"Johannes Sixt" <j6t@kdbg.org>
Subject: Re: [PATCH v3 4/4] trace.c: be smart about what env to print in trace_run_command()
Date: Fri, 12 Jan 2018 13:45:46 -0500	[thread overview]
Message-ID: <20180112184546.GA15514@sigill.intra.peff.net> (raw)
In-Reply-To: <xmqqpo6fvstf.fsf@gitster.mtv.corp.google.com>

On Fri, Jan 12, 2018 at 10:24:28AM -0800, Junio C Hamano wrote:

> Jeff King <peff@peff.net> writes:
> 
> >> +		/*
> >> +		 * Do we have a sequence of "unset GIT_DIR; GIT_DIR=foo"?
> >> +		 * Then don't bother with the unset thing.
> >> +		 */
> >> +		if (i + 1 < envs.nr &&
> >> +		    !strcmp(env, envs.items[i + 1].string))
> >>  			continue;
> >
> > Are we guaranteed that "FOO" and "FOO=bar" appear next to each other in the
> > sorted list? I think "FOO123=bar" could come between.
> 
> At this point, envs is a string list whose key is FOO for both "FOO"
> (unset) and "FOO=bar" (set); "FOO123=bar" would sort after these two.

Ah, right, I didn't notice that we took just the key name.

> But I did not see anything that attempts to guarantee that "FOO"
> sorts before "FOO=bar" with string_list_sort().  If the sort used in
> the function is stable, and if the case we care about is unset
> followed by set, then the above would catch the case, but even if
> that were the case, it is unclear what we want to do when a set of
> FOO is followed by an unset of FOO.

The actual prep_childenv() code looks like it would always do
last-one-wins, so we should treat FOO as unset in that final case. But
that only kicks in on non-Windows.

On Windows we feed cmd->env straight to mingw_spawnvpe().  That ends up
in make_environment_block(), which looks like it does something similar.

It's too bad the prep code is not shared, since then we could probably
just ask _it_ which deltas it applied. I suspect it would be possible to
share it, since mingw_spawnvpe is our own compat thing (so we could
change its interface to take the whole prepared environment block). But
I won't blame Duy if he doesn't want to refactor all of the
cross-platform exec code. ;)

> And if the sort is not stable, then the above code is just simply
> broken.

Agreed.

> 
> > I also think this is a special case of a more general problem. FOO could
> > appear any number of times in the "env" array, as a deletion or with
> > multiple values. Our prep_childenv() would treat that as "last one
> > wins", I think. Could we just do the same here?
> 
> Yeah, if the last one is to set FOO=bar, then it feels sufficient to
> just check if FOO is set to bar in the original and deciding to show
> or hide; similarly if the last one is to unset FOO, it does not matter
> if the caller sets it to some other value before unsetting, so it
> feels sufficient to just check if FOO is set to anything in the
> original environment.

Yep, agreed again.

-Peff

  reply	other threads:[~2018-01-12 18:45 UTC|newest]

Thread overview: 74+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-01-10 10:48 [PATCH] run-command.c: print env vars when GIT_TRACE is set Nguyễn Thái Ngọc Duy
2018-01-10 18:09 ` Brandon Williams
2018-01-10 19:14   ` Stefan Beller
2018-01-10 19:26     ` Brandon Williams
2018-01-10 19:35       ` Stefan Beller
2018-01-11  9:47 ` [PATCH v2] " Nguyễn Thái Ngọc Duy
2018-01-11 11:25   ` Duy Nguyen
2018-01-11 17:53   ` Brandon Williams
2018-01-11 18:20     ` Stefan Beller
2018-01-11 19:27   ` Junio C Hamano
2018-01-12  9:23     ` Duy Nguyen
2018-01-12  9:56   ` [PATCH v3 0/4] " Nguyễn Thái Ngọc Duy
2018-01-12  9:56     ` [PATCH v3 1/4] trace.c: introduce trace_run_command() Nguyễn Thái Ngọc Duy
2018-01-12 13:05       ` Jeff King
2018-01-12 13:11         ` Jeff King
2018-01-12  9:56     ` [PATCH v3 2/4] trace.c: print program 'git' when child_process.git_cmd is set Nguyễn Thái Ngọc Duy
2018-01-12 13:05       ` Jeff King
2018-01-12  9:56     ` [PATCH v3 3/4] trace.c: print env vars in trace_run_command() Nguyễn Thái Ngọc Duy
2018-01-12 13:13       ` Jeff King
2018-01-12  9:56     ` [PATCH v3 4/4] trace.c: be smart about what env to print " Nguyễn Thái Ngọc Duy
2018-01-12 13:33       ` Jeff King
2018-01-12 18:24         ` Junio C Hamano
2018-01-12 18:45           ` Jeff King [this message]
2018-01-12 19:19             ` Junio C Hamano
2018-01-12 19:23               ` Jeff King
2018-01-12 20:28                 ` Brandon Williams
2018-01-12 22:54         ` Junio C Hamano
2018-01-13  4:54           ` Duy Nguyen
2018-01-13  6:47             ` Duy Nguyen
2018-01-12 13:36     ` [PATCH v3 0/4] run-command.c: print env vars when GIT_TRACE is set Jeff King
2018-01-12 13:38       ` [PATCH 5/4] sq_quote_argv: drop maxlen parameter Jeff King
2018-01-12 23:20         ` Junio C Hamano
2018-01-12 13:39       ` [PATCH 6/4] trace: avoid unnecessary quoting Jeff King
2018-01-12 18:06       ` [PATCH v3 0/4] run-command.c: print env vars when GIT_TRACE is set Stefan Beller
2018-01-12 17:19     ` Stefan Beller
2018-01-13  6:54       ` Duy Nguyen
2018-01-13  6:49     ` [PATCH v4 0/7] Trace env variables in run_command() Nguyễn Thái Ngọc Duy
2018-01-13  6:49       ` [PATCH v4 1/7] sq_quote_argv: drop maxlen parameter Nguyễn Thái Ngọc Duy
2018-01-13  6:49       ` [PATCH v4 2/7] trace: avoid unnecessary quoting Nguyễn Thái Ngọc Duy
2018-01-13  6:49       ` [PATCH v4 3/7] trace.c: move strbuf_release() out of print_trace_line() Nguyễn Thái Ngọc Duy
2018-01-13  7:14         ` Jeff King
2018-01-13  6:49       ` [PATCH v4 4/7] trace.c: introduce trace_run_command() Nguyễn Thái Ngọc Duy
2018-01-13  6:49       ` [PATCH v4 5/7] trace.c: print program 'git' when child_process.git_cmd is set Nguyễn Thái Ngọc Duy
2018-01-13  6:49       ` [PATCH v4 6/7] trace.c: print env vars in trace_run_command() Nguyễn Thái Ngọc Duy
2018-01-13  7:25         ` Jeff King
2018-01-16 22:13         ` Junio C Hamano
2018-01-16 22:20           ` Stefan Beller
2018-01-17  1:32             ` Junio C Hamano
2018-01-13  6:49       ` [PATCH v4 7/7] trace.c: print new cwd " Nguyễn Thái Ngọc Duy
2018-01-13  7:26       ` [PATCH v4 0/7] Trace env variables in run_command() Jeff King
2018-01-15 10:59       ` [PATCH v5 " Nguyễn Thái Ngọc Duy
2018-01-15 10:59         ` [PATCH v5 1/7] sq_quote_argv: drop maxlen parameter Nguyễn Thái Ngọc Duy
2018-01-15 10:59         ` [PATCH v5 2/7] trace: avoid unnecessary quoting Nguyễn Thái Ngọc Duy
2018-01-15 10:59         ` [PATCH v5 3/7] trace.c: move strbuf_release() out of print_trace_line() Nguyễn Thái Ngọc Duy
2018-01-16 18:24           ` Brandon Williams
2018-01-15 10:59         ` [PATCH v5 4/7] trace.c: introduce trace_run_command() Nguyễn Thái Ngọc Duy
2018-01-17 22:21           ` Jeff King
2018-01-17 22:41             ` Junio C Hamano
2018-01-15 10:59         ` [PATCH v5 5/7] trace.c: print program 'git' when child_process.git_cmd is set Nguyễn Thái Ngọc Duy
2018-01-15 10:59         ` [PATCH v5 6/7] trace.c: print env vars in trace_run_command() Nguyễn Thái Ngọc Duy
2018-01-16 18:32           ` Brandon Williams
2018-01-15 10:59         ` [PATCH v5 7/7] trace.c: print new cwd " Nguyễn Thái Ngọc Duy
2018-01-18  9:45         ` [PATCH v6 0/7] Trace env variables in run_command() Nguyễn Thái Ngọc Duy
2018-01-18  9:45           ` [PATCH v6 1/7] sq_quote_argv: drop maxlen parameter Nguyễn Thái Ngọc Duy
2018-01-18  9:45           ` [PATCH v6 2/7] trace: avoid unnecessary quoting Nguyễn Thái Ngọc Duy
2018-01-18  9:45           ` [PATCH v6 3/7] trace.c: move strbuf_release() out of print_trace_line() Nguyễn Thái Ngọc Duy
2018-01-18  9:45           ` [PATCH v6 4/7] run-command.c: introduce trace_run_command() Nguyễn Thái Ngọc Duy
2018-01-18  9:45           ` [PATCH v6 5/7] run-command.c: print program 'git' when tracing git_cmd mode Nguyễn Thái Ngọc Duy
2018-01-18  9:45           ` [PATCH v6 6/7] run-command.c: print env vars in trace_run_command() Nguyễn Thái Ngọc Duy
2018-01-18  9:45           ` [PATCH v6 7/7] run-command.c: print new cwd " Nguyễn Thái Ngọc Duy
2018-01-19 21:39           ` [PATCH v6 0/7] Trace env variables in run_command() Jeff King
2018-01-11 10:07 ` [PATCH] run-command.c: print env vars when GIT_TRACE is set Jeff King
2018-01-11 11:13   ` Duy Nguyen
2018-01-11 18:21   ` Johannes Sixt

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=20180112184546.GA15514@sigill.intra.peff.net \
    --to=peff@peff.net \
    --cc=bmwill@google.com \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=j6t@kdbg.org \
    --cc=pclouds@gmail.com \
    --cc=sbeller@google.com \
    --subject='Re: [PATCH v3 4/4] trace.c: be smart about what env to print in trace_run_command()' \
    /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

Code repositories for project(s) associated with this 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).