git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Torsten Bögershausen" <tboegi@web.de>
To: Jonathan Nieder <jrnieder@gmail.com>, Stefan Beller <sbeller@google.com>
Cc: Junio C Hamano <gitster@pobox.com>,
	git@vger.kernel.org, Michael Haggerty <mhagger@alum.mit.edu>,
	Jeff King <peff@peff.net>
Subject: Re: [PATCH 04/14] hold_lock_file_for_append: pass error message back through a strbuf
Date: Wed, 03 Dec 2014 07:09:28 +0100	[thread overview]
Message-ID: <547EA918.30003@web.de> (raw)
In-Reply-To: <20141203051459.GN6527@google.com>

On 12/03/2014 06:14 AM, Jonathan Nieder wrote:
> This way, the code does not need to carefully safeguard errno to allow
> callers to print a reasonable error message when they choose to do
> some cleanup before die()-ing.
>
> Fixes a bug waiting to happen where copy_fd would clobber the errno
> passed back via hold_lock_file_for_append from read() or write() when
> flags did not contain LOCK_DIE_ON_ERROR.  Luckily the only caller uses
> flags == LOCK_DIE_ON_ERROR, avoiding that bug in practice.
>
> Reported-by: Ronnie Sahlberg <sahlberg@google.com>
> Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
> ---
> The advertised bugfix.
>
>   lockfile.c  | 29 ++++++++++-------------------
>   lockfile.h  |  3 ++-
>   sha1_file.c |  7 ++++++-
>   3 files changed, 18 insertions(+), 21 deletions(-)
>
> diff --git a/lockfile.c b/lockfile.c
> index b3020f3..8685c68 100644
> --- a/lockfile.c
> +++ b/lockfile.c
> @@ -179,45 +179,36 @@ int hold_lock_file_for_update(struct lock_file *lk, const char *path, int flags)
>   	return fd;
>   }
>   
> -int hold_lock_file_for_append(struct lock_file *lk, const char *path, int flags)
> +int hold_lock_file_for_append(struct lock_file *lk, const char *path,
> +			      int flags, struct strbuf *err)
>   {
>   	int fd, orig_fd;
> -	struct strbuf err = STRBUF_INIT;
> +
> +	assert(!(flags & LOCK_DIE_ON_ERROR));
> +	assert(err && !err->len);
>   
What do I miss ?
In the old code we die() printing the errno when LOCK_DIE_ON_ERRORis set.
Now we die() in an assert() or two ?
And should'nt we assert in  strbuf_addf() instead ?

And in general, does the commit message deserve an update?

>Luckily the only caller uses flags == LOCK_DIE_ON_ERROR,
The first impression was: Why do we not simplify the code and remove
the flag completely ?
This feels somewhat like maintaining dead code, which may be used later.
(Unless it will be used in later commit, and the we could mention it here)

  reply	other threads:[~2014-12-03  6:11 UTC|newest]

Thread overview: 83+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-11-17 22:14 [PATCH] copy.c: make copy_fd preserve meaningful errno Stefan Beller
2014-11-17 23:08 ` Junio C Hamano
2014-11-17 23:13   ` Stefan Beller
2014-11-17 23:35 ` Jonathan Nieder
2014-11-18  0:18   ` Stefan Beller
2014-11-18  0:48     ` Jonathan Nieder
2014-11-18  1:01       ` Stefan Beller
2014-12-03  5:02         ` [PATCH 0/14] " Jonathan Nieder
2014-12-03  5:10           ` [PATCH 01/14] strbuf: introduce strbuf_prefixf() Jonathan Nieder
2014-12-03 20:10             ` Eric Sunshine
2014-12-03 20:14               ` Jonathan Nieder
2014-12-03 21:45             ` Junio C Hamano
2014-12-03 21:59               ` Jonathan Nieder
2014-12-03 22:09                 ` Jonathan Nieder
2014-12-03 22:40                 ` Junio C Hamano
2014-12-03  5:12           ` [PATCH 02/14] add_to_alternates_file: respect GIT_OBJECT_DIRECTORY Jonathan Nieder
2014-12-03 18:53             ` Junio C Hamano
2014-12-03  5:13           ` [PATCH 03/14] copy_fd: pass error message back through a strbuf Jonathan Nieder
2014-12-03 19:01             ` Junio C Hamano
2014-12-03 20:28               ` Jonathan Nieder
2014-12-03 21:00               ` Jeff King
2014-12-03 21:38                 ` Jonathan Nieder
2014-12-04  7:59                   ` Jeff King
2014-12-04  8:36                     ` Stefan Beller
2014-12-04  9:04                       ` Jeff King
2014-12-10 17:02                     ` Michael Haggerty
2014-12-10 19:00                       ` Junio C Hamano
2014-12-10 19:14                         ` Jeff King
2014-12-04  3:01               ` [PATCH/RFC] doc: document error handling functions and conventions (Re: [PATCH 03/14] copy_fd: pass error message back through a strbuf) Jonathan Nieder
2014-12-04 23:27                 ` Junio C Hamano
2014-12-04 23:37                   ` Jonathan Nieder
2014-12-04 23:41                   ` Jonathan Nieder
2014-12-04 23:44                     ` Jeff King
2014-12-04 23:52                       ` Junio C Hamano
2014-12-05  0:01                         ` Jeff King
2014-12-05 18:00                           ` Junio C Hamano
2014-12-07 10:07                             ` Jeff King
2014-12-09 18:43                               ` Junio C Hamano
2014-12-09 18:49                                 ` Jeff King
2015-02-12 23:08                                 ` Junio C Hamano
2015-02-17 15:50                                   ` Michael Haggerty
2015-02-17 16:03                                     ` Junio C Hamano
2015-02-17 16:05                                       ` Jeff King
2015-02-17 22:46                                         ` Junio C Hamano
2014-12-03 20:02             ` [PATCH 03/14] copy_fd: pass error message back through a strbuf Junio C Hamano
2014-12-03 20:18               ` Jonathan Nieder
2014-12-03 21:43                 ` Junio C Hamano
2014-12-03 21:51                   ` Jonathan Nieder
2014-12-03 20:20               ` Stefan Beller
2014-12-03  5:14           ` [PATCH 04/14] hold_lock_file_for_append: " Jonathan Nieder
2014-12-03  6:09             ` Torsten Bögershausen [this message]
2014-12-03  7:04               ` Jonathan Nieder
2014-12-03  5:16           ` [PATCH 05/14] lock_packed_refs: " Jonathan Nieder
2014-12-03  5:19           ` [PATCH 06/14] lockfile: introduce flag for locks outside .git Jonathan Nieder
2014-12-03 23:13             ` Junio C Hamano
2014-12-03 23:24               ` Jonathan Nieder
2014-12-03 23:25               ` Junio C Hamano
2014-12-03 23:29                 ` Jonathan Nieder
2014-12-03 23:38                   ` Junio C Hamano
2014-12-03 23:41                     ` Jonathan Nieder
2014-12-03 23:43                     ` Junio C Hamano
2014-12-03 23:57                   ` Jeff King
2014-12-04  5:51                   ` Junio C Hamano
2014-12-04 17:56                     ` Jonathan Nieder
2014-12-03  5:19           ` [PATCH 07/14] fast-import: use message from lockfile API when writing marks fails Jonathan Nieder
2014-12-03  5:20           ` [PATCH 08/14] credentials: use message from lockfile API when locking ~/.git-credentials fails Jonathan Nieder
2014-12-03  5:21           ` [PATCH 09/14] config: use message from lockfile API when locking config file fails Jonathan Nieder
2014-12-03 19:59             ` Junio C Hamano
2014-12-03 20:16               ` Jonathan Nieder
2014-12-03  5:22           ` [PATCH 10/14] rerere: error out on autoupdate failure Jonathan Nieder
2014-12-03  5:25           ` [PATCH 11/14] hold_locked_index: pass error message back through a strbuf Jonathan Nieder
2014-12-03  5:26           ` [PATCH 12/14] hold_lock_file_for_update: " Jonathan Nieder
2014-12-03 18:53             ` Jonathan Nieder
2014-12-03  5:26           ` [PATCH 13/14] lockfile: remove unused function 'unable_to_lock_die' Jonathan Nieder
2014-12-03  5:27           ` [PATCH 14/14] lockfile: make 'unable_to_lock_message' private Jonathan Nieder
2014-12-03 20:42             ` Stefan Beller
2014-11-18 16:32   ` [PATCH] copy.c: make copy_fd preserve meaningful errno Junio C Hamano
2014-11-18 17:08     ` Junio C Hamano
2014-11-21  9:14 ` Michael Haggerty
2014-11-21  9:17   ` Michael Haggerty
2014-11-21 17:48     ` Junio C Hamano
2014-11-21 17:54       ` Jeff King
2014-11-21 18:31         ` 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=547EA918.30003@web.de \
    --to=tboegi@web.de \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=jrnieder@gmail.com \
    --cc=mhagger@alum.mit.edu \
    --cc=peff@peff.net \
    --cc=sbeller@google.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.
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 NNTP newsgroup(s).