From: Junio C Hamano <gitster@pobox.com>
To: Michael Haggerty <mhagger@alum.mit.edu>
Cc: "Jeff King" <peff@peff.net>,
"David Turner" <dturner@twopensource.com>,
git@vger.kernel.org, "Nguyễn Thái Ngọc Duy" <pclouds@gmail.com>,
"Ramsay Jones" <ramsay@ramsayjones.plus.com>
Subject: Re: [PATCH 19/29] refs: don't dereference on rename
Date: Fri, 29 Apr 2016 08:29:10 -0700 [thread overview]
Message-ID: <xmqqmvoc1lix.fsf@gitster.mtv.corp.google.com> (raw)
In-Reply-To: <572367B4.4050207@alum.mit.edu> (Michael Haggerty's message of "Fri, 29 Apr 2016 15:55:00 +0200")
Michael Haggerty <mhagger@alum.mit.edu> writes:
> On 04/29/2016 02:12 PM, Jeff King wrote:
>> On Fri, Apr 29, 2016 at 12:57:29PM +0200, Michael Haggerty wrote:
>>
>>> Remember, we're talking about rename_ref() only, not reference deletion
>>> in general. rename_ref() is not very robust anyway--it doesn't happen in
>>> a single transaction, and it is vulnerable to being defeated by
>>> simultaneous reference updates by other processes. It *does* wrap the
>>> deletion of newrefname in a transaction; the only question is whether an
>>> old_sha1 is supplied to that transaction.
>>>
>>> So, suppose that newrefname starts at value A, and there are two updates
>>> running simultaneously:
>>>
>>> 1. An update of reference newrefname from A -> B
>>>
>>> 2. A rename of reference oldrefname to newrefname, which includes
>>> a. read_ref_full("newrefname") and
>>> b. delete_ref("newrefname").
>>
>> I wondered at first if you meant "oldrefname" in 2b. That is, a rename
>> would involve writing to "newrefname" and then deleting "oldrefname",
>> and not doing the old_sha1 and normal locking on the deletion of
>> oldrefname would be bad (in case somebody else updated it while we were
>> operating).
>>
>> But reading the patch again, that's not what's going on. You're talking
>> just about the case where we force-overwrite an existing newrefname, and
>> delete it first to do so. But what I don't understand is:
>
> It's beyond the ambition of this patch to fix this old rename_ref()
> code, but...
Thanks for a clear explanation. The NULL might want to be explained
in in-code comment (e.g. "here we do not care about verifying old
value because..."), then.
next prev parent reply other threads:[~2016-04-29 15:29 UTC|newest]
Thread overview: 70+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-04-27 16:57 [PATCH 00/29] Yet more preparation for reference backends Michael Haggerty
2016-04-27 16:57 ` [PATCH 01/29] safe_create_leading_directories(): improve docstring Michael Haggerty
2016-04-27 16:57 ` [PATCH 02/29] remove_dir_recursively(): add docstring Michael Haggerty
2016-04-27 16:57 ` [PATCH 03/29] refname_is_safe(): use skip_prefix() Michael Haggerty
2016-04-27 16:57 ` [PATCH 04/29] refname_is_safe(): don't allow the empty string Michael Haggerty
2016-04-27 16:57 ` [PATCH 05/29] refname_is_safe(): insist that the refname already be normalized Michael Haggerty
2016-04-27 17:59 ` Junio C Hamano
2016-04-27 20:10 ` David Turner
2016-04-27 20:15 ` Jeff King
2016-04-27 20:34 ` David Turner
2016-04-27 20:37 ` Jeff King
2016-04-27 22:19 ` Jeff King
2016-04-28 17:44 ` David Turner
2016-04-27 16:57 ` [PATCH 06/29] commit_ref_update(): write error message to *err, not stderr Michael Haggerty
2016-04-27 16:57 ` [PATCH 07/29] rename_ref(): remove unneeded local variable Michael Haggerty
2016-04-27 16:57 ` [PATCH 08/29] ref_transaction_commit(): remove local variable n Michael Haggerty
2016-04-27 18:16 ` Junio C Hamano
2016-04-27 20:45 ` Junio C Hamano
2016-04-27 16:57 ` [PATCH 09/29] read_raw_ref(): rename flags argument to type Michael Haggerty
2016-04-27 16:57 ` [PATCH 10/29] read_raw_ref(): clear *type at start of function Michael Haggerty
2016-04-27 16:57 ` [PATCH 11/29] read_raw_ref(): rename symref argument to referent Michael Haggerty
2016-04-27 16:57 ` [PATCH 12/29] read_raw_ref(): improve docstring Michael Haggerty
2016-04-27 18:31 ` Junio C Hamano
2016-04-27 16:57 ` [PATCH 13/29] lock_ref_sha1_basic(): remove unneeded local variable Michael Haggerty
2016-04-27 16:57 ` [PATCH 14/29] refs: make error messages more consistent Michael Haggerty
2016-04-27 16:57 ` [PATCH 15/29] ref_transaction_create(): disallow recursive pruning Michael Haggerty
2016-04-27 18:47 ` Junio C Hamano
2016-04-27 20:23 ` David Turner
2016-04-27 20:45 ` Junio C Hamano
2016-04-27 21:15 ` Junio C Hamano
2016-04-28 17:48 ` David Turner
2016-04-28 20:15 ` Junio C Hamano
2016-04-29 6:56 ` Michael Haggerty
2016-04-29 8:19 ` Junio C Hamano
2016-04-29 8:41 ` Junio C Hamano
2016-04-29 14:29 ` Michael Haggerty
2016-04-27 16:57 ` [PATCH 16/29] ref_transaction_commit(): correctly report close_ref() failure Michael Haggerty
2016-04-27 16:57 ` [PATCH 17/29] delete_branches(): use resolve_refdup() Michael Haggerty
2016-04-27 16:57 ` [PATCH 18/29] refs: allow log-only updates Michael Haggerty
2016-04-27 16:57 ` [PATCH 19/29] refs: don't dereference on rename Michael Haggerty
2016-04-27 18:55 ` Junio C Hamano
2016-04-29 7:38 ` Michael Haggerty
2016-04-29 8:53 ` Junio C Hamano
2016-04-29 10:57 ` Michael Haggerty
2016-04-29 12:12 ` Jeff King
2016-04-29 13:55 ` Michael Haggerty
2016-04-29 14:08 ` Jeff King
2016-04-29 15:29 ` Junio C Hamano [this message]
2016-04-29 23:21 ` David Turner
2016-04-30 3:48 ` Michael Haggerty
2016-05-02 17:55 ` Junio C Hamano
2016-04-27 16:57 ` [PATCH 20/29] verify_refname_available(): adjust constness in declaration Michael Haggerty
2016-04-27 16:57 ` [PATCH 21/29] add_update(): initialize the whole ref_update Michael Haggerty
2016-04-27 16:57 ` [PATCH 22/29] lock_ref_for_update(): new function Michael Haggerty
2016-04-27 16:57 ` [PATCH 23/29] unlock_ref(): move definition higher in the file Michael Haggerty
2016-04-27 16:57 ` [PATCH 24/29] ref_transaction_update(): check refname_is_safe() at a minimum Michael Haggerty
2016-04-27 20:14 ` Junio C Hamano
2016-04-29 7:42 ` Michael Haggerty
2016-04-29 8:53 ` Junio C Hamano
2016-04-27 16:57 ` [PATCH 25/29] refs: resolve symbolic refs first Michael Haggerty
2016-04-28 23:40 ` David Turner
2016-04-29 9:51 ` Michael Haggerty
2016-04-29 23:14 ` David Turner
2016-05-02 18:06 ` Junio C Hamano
2016-04-27 16:57 ` [PATCH 26/29] lock_ref_for_update(): don't re-read non-symbolic references Michael Haggerty
2016-04-27 16:57 ` [PATCH 27/29] lock_ref_for_update(): don't resolve symrefs Michael Haggerty
2016-04-27 16:57 ` [PATCH 28/29] commit_ref_update(): remove the flags parameter Michael Haggerty
2016-04-27 16:57 ` [PATCH 29/29] lock_ref_sha1_basic(): only handle REF_NODEREF mode Michael Haggerty
2016-04-29 15:43 ` Junio C Hamano
2016-04-29 1:14 ` [PATCH 00/29] Yet more preparation for reference backends David Turner
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=xmqqmvoc1lix.fsf@gitster.mtv.corp.google.com \
--to=gitster@pobox.com \
--cc=dturner@twopensource.com \
--cc=git@vger.kernel.org \
--cc=mhagger@alum.mit.edu \
--cc=pclouds@gmail.com \
--cc=peff@peff.net \
--cc=ramsay@ramsayjones.plus.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).