Tarantool development patches archive
 help / color / mirror / Atom feed
From: Nikita Pettik <korablev@tarantool.org>
To: Aleksandr Lyapunov <alyapunov@tarantool.org>
Cc: tarantool-patches@dev.tarantool.org, v.shpilevoy@tarantool.org
Subject: Re: [Tarantool-patches] [PATCH v3 3/3] vinyl: clean-up write iterator if vy_task_write_run() fails
Date: Thu, 7 May 2020 22:16:36 +0000	[thread overview]
Message-ID: <20200507221636.GA13970@tarantool.org> (raw)
In-Reply-To: <7b07eb55-340b-b2ca-231c-9feeffe32a3f@tarantool.org>

On 07 May 10:53, Aleksandr Lyapunov wrote:
> On 5/7/20 3:36 AM, Nikita Pettik wrote:
> > On 06 May 13:37, Aleksandr Lyapunov wrote:
> > > It's very strange, as I see start() method already makes its own cleanup
> > > in case of failure and does not require 'stop' call in this case.
> > > If it does not do it correctly - it must be fixed (or removed completely,
> > > but it seems to be very serious change).
> > Turns out start() doesn't provide proper clean-up: in case
> > vy_write_iterator_add_src() fails, only last source is cleaned up,
> > meanwhile the rest will be destroyed in vy_write_iterator_stop().
> > I've moved clean-up part from stop() method right to start(). Here's diff:
> 
> Could you see commit 2f17c929? I see it in master.
> It does exactly the same and some other good looking related changes.
>

I explored that patch. It is not backported to 1.10 (obviously) since
other changes except for additional clean-up in vy_write_iterator_start()
look like refactoring (we push to 1.10 only bug fixes). It can't be
cherry-picked now since code is bit different and 1.10 lacks some
features (like heap helpers to determine whether heap node is linked or
not). So I suggest to leave my patch as is for 1.10 branch and backport
test case to master branch.
 

  reply	other threads:[~2020-05-07 22:16 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-04-27  0:52 [Tarantool-patches] [PATCH v3 0/3] vinyl: fix uninitialized memory accesses Nikita Pettik
2020-04-27  0:52 ` [Tarantool-patches] [PATCH v3 1/3] vinyl: init all vars before cleanup in vy_lsm_split_range() Nikita Pettik
2020-05-06  9:04   ` Aleksandr Lyapunov
2020-05-06 13:12     ` Nikita Pettik
2020-05-06 17:52       ` Aleksandr Lyapunov
2020-05-07  1:09         ` Nikita Pettik
2020-04-27  0:52 ` [Tarantool-patches] [PATCH v3 2/3] vinyl: clean-up unprocessed read views in *_build_read_views() Nikita Pettik
2020-05-06  9:56   ` Aleksandr Lyapunov
2020-05-07  0:29     ` Nikita Pettik
2020-05-07  8:44       ` Aleksandr Lyapunov
2020-05-07 12:28         ` Nikita Pettik
2020-04-27  0:52 ` [Tarantool-patches] [PATCH v3 3/3] vinyl: clean-up write iterator if vy_task_write_run() fails Nikita Pettik
2020-05-01  0:55   ` Vladislav Shpilevoy
2020-05-03  9:22     ` Konstantin Osipov
2020-05-07  0:38     ` Nikita Pettik
2020-05-06 10:37   ` Aleksandr Lyapunov
2020-05-07  0:36     ` Nikita Pettik
2020-05-07  7:53       ` Aleksandr Lyapunov
2020-05-07 22:16         ` Nikita Pettik [this message]
2020-05-08 16:29           ` Aleksandr Lyapunov

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=20200507221636.GA13970@tarantool.org \
    --to=korablev@tarantool.org \
    --cc=alyapunov@tarantool.org \
    --cc=tarantool-patches@dev.tarantool.org \
    --cc=v.shpilevoy@tarantool.org \
    --subject='Re: [Tarantool-patches] [PATCH v3 3/3] vinyl: clean-up write iterator if vy_task_write_run() fails' \
    /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

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox