qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Eric Blake <eblake@redhat.com>
To: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>,
	"qemu-block@nongnu.org" <qemu-block@nongnu.org>
Cc: "kwolf@redhat.com" <kwolf@redhat.com>,
	"fam@euphon.net" <fam@euphon.net>,
	Denis Lunev <den@virtuozzo.com>,
	"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>,
	"mreitz@redhat.com" <mreitz@redhat.com>,
	"jsnow@redhat.com" <jsnow@redhat.com>
Subject: Re: [PATCH 08/10] nbd/server: introduce NBDExtentArray
Date: Fri, 18 Oct 2019 11:34:48 -0500	[thread overview]
Message-ID: <39449490-8cea-0387-6840-96bc06af55b2@redhat.com> (raw)
In-Reply-To: <ffee3ef7-d931-bcac-be82-2ae3533aa981@virtuozzo.com>

On 10/18/19 11:07 AM, Vladimir Sementsov-Ogievskiy wrote:

>>>    static int nbd_co_send_extents(NBDClient *client, uint64_t handle,
>>> -                               NBDExtent *extents, unsigned int nb_extents,
>>> -                               uint64_t length, bool last,
>>> -                               uint32_t context_id, Error **errp)
>>> +                               NBDExtentArray *ea,
>>> +                               bool last, uint32_t context_id, Error **errp)
>>>    {
>>>        NBDStructuredMeta chunk;
>>> -
>>> +    size_t len = ea->count * sizeof(ea->extents[0]);
>>> +    g_autofree NBDExtent *extents = g_memdup(ea->extents, len);
>>
>> Why do we need memdup here?  What's wrong with modifying ea->extents in place?...
> 
> To not make ea to be IN-OUT parameter.. I don't like functions with side effects.
> It will break the code if at some point we call nbd_co_send_extents twice on same
> ea object.
> 
> What is the true way? To not memdup, nbd_co_send_extents should consume the whole
> ea object..
> 
> Seems, g_autoptr attribute can't be used for function parameter, gcc complains:
> nbd/server.c:1983:32: error: ‘cleanup’ attribute ignored [-Werror=attributes]
>    1983 |                                g_autoptr(NBDExtentArray) ea,
>         |                                ^~~~~~~~~
> 
> so, is it better
> to call nbd_co_send_external(... g_steal_pointer(&ea) ...)
> 
> and than in nbd_co_send_external do
> 
> g_autoptr(NBDExtentArray) local_ea = ea;
> NBDExtent *extents = local_ea->extents;
> 
> ?
> 

No, that makes it worse.  It's that much more confusing to track who is 
allocating what and where it gets cleaned up.

I personally don't see the need to avoid jumping through hoops to avoid 
an in-out parameter (if we're going to rework code later, we'll notice 
that we documented how things are supposed to be used), but if in-out 
parameters bother you, then the approach you used, even with an extra 
memdup(), is the simplest way to maintain, even if it is not the most 
efficient.

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org


  reply	other threads:[~2019-10-18 16:36 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-09-30 15:14 [PATCH 00/10] Further bitmaps improvements Vladimir Sementsov-Ogievskiy
2019-09-30 15:14 ` [PATCH 01/10] hbitmap: introduce HBITMAP_MAX_ORIG_SIZE Vladimir Sementsov-Ogievskiy
2019-10-09 15:34   ` Eric Blake
2019-10-09 16:04     ` Vladimir Sementsov-Ogievskiy
2019-09-30 15:14 ` [PATCH 02/10] hbitmap: move hbitmap_iter_next_word to hbitmap.c Vladimir Sementsov-Ogievskiy
2019-09-30 15:14 ` [PATCH 03/10] hbitmap: unpublish hbitmap_iter_skip_words Vladimir Sementsov-Ogievskiy
2019-09-30 15:14 ` [PATCH 04/10] hbitmap: drop meta bitmaps as they are unused Vladimir Sementsov-Ogievskiy
2019-09-30 15:14 ` [PATCH 05/10] block/dirty-bitmap: switch _next_dirty_area and _next_zero to int64_t Vladimir Sementsov-Ogievskiy
2019-09-30 15:14 ` [PATCH 06/10] block/dirty-bitmap: add _next_dirty API Vladimir Sementsov-Ogievskiy
2019-09-30 15:14 ` [PATCH 07/10] block/dirty-bitmap: improve _next_dirty_area API Vladimir Sementsov-Ogievskiy
2019-09-30 15:15 ` [PATCH 08/10] nbd/server: introduce NBDExtentArray Vladimir Sementsov-Ogievskiy
2019-10-09 17:02   ` Eric Blake
2019-10-18 16:07     ` Vladimir Sementsov-Ogievskiy
2019-10-18 16:34       ` Eric Blake [this message]
2019-09-30 15:15 ` [PATCH 09/10] nbd/server: use bdrv_dirty_bitmap_next_dirty_area Vladimir Sementsov-Ogievskiy
2019-10-09 18:26   ` Eric Blake
2019-09-30 15:15 ` [PATCH 10/10] block/qcow2-bitmap: use bdrv_dirty_bitmap_next_dirty Vladimir Sementsov-Ogievskiy

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=39449490-8cea-0387-6840-96bc06af55b2@redhat.com \
    --to=eblake@redhat.com \
    --cc=den@virtuozzo.com \
    --cc=fam@euphon.net \
    --cc=jsnow@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=mreitz@redhat.com \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=vsementsov@virtuozzo.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).