Tarantool development patches archive
 help / color / mirror / Atom feed
From: Vladislav Shpilevoy <v.shpilevoy@tarantool.org>
To: Alexander Turenko <alexander.turenko@tarantool.org>
Cc: tarantool-patches@freelists.org
Subject: [tarantool-patches] Re: [PATCH v2 3/4] tuple: use global msgpack serializer in Lua tuple
Date: Sat, 14 Sep 2019 00:32:39 +0200	[thread overview]
Message-ID: <e089763d-f4a0-c092-2665-0a593e8beff1@tarantool.org> (raw)
In-Reply-To: <20190912232758.qvb3kloft3ngv6fj@tkn_work_nb>

Thanks for the review!

On 13/09/2019 01:27, Alexander Turenko wrote:
> LGTM.
> 
> Several questions are below.
> 
> I agree that the variant with triggers looks more natural.
> 
> Should not we add box_lua_tuple_free() and call trigger_destroy()
> inside? I know, we now don't call tarantool_lua_free(), but I hope it
> will be fixed someday.
> 
> If you'll going to add it, please, add also the call to luaT_tuple_new.c
> for the symmetry with box_lua_tuple_init(). It also will help to keep
> this test clean from ASAN / Valgrind point of view: I did verify it
> aganst one of those tools at the time of writing the test (don't
> remember against which of them).

I tried to add box_lua_tuple_free, but as you mentioned, we don't
have box_lua_free() so that function is unused. I don't think it
should be added now.

Also I stumbled into a problem, that luaT_tuple_new.c uses
luaopen_msgpack, but we don't have luaclose_msgpack. So there
are 2 holes. Moreover, as I understand, no one of our Lua C
modules really drops its bindings from Lua namespace. It is
not related to this issue, but I remember that someday we wanted
to be able to 'uncfg' box, like it was not called.

> 
> Hm. We don't have box_lua_free() at all. I'm doubtful now.
> 
> WBR, Alexander Turenko.
> 
>> +static inline void
>> +tuple_serializer_fill(void)
>> +{
>> +	luaL_serializer_copy_options(&tuple_serializer, luaL_msgpack_default);
>> +	tuple_serializer.encode_sparse_ratio = 0;
>> +}
> 
> Is not this name quite common? Maybe tuple_serializer_update_options()?
> 

Honestly, I don't really care that much. The function with all its invocations
fits in one screen. Both names are ok for me:

diff --git a/src/box/lua/tuple.c b/src/box/lua/tuple.c
index fc22f5572..8b59466b9 100644
--- a/src/box/lua/tuple.c
+++ b/src/box/lua/tuple.c
@@ -554,7 +554,7 @@ static const struct luaL_Reg lbox_tuple_iterator_meta[] = {
 /* }}} */
 
 static inline void
-tuple_serializer_fill(void)
+tuple_serializer_update_options(void)
 {
 	luaL_serializer_copy_options(&tuple_serializer, luaL_msgpack_default);
 	tuple_serializer.encode_sparse_ratio = 0;
@@ -565,7 +565,7 @@ on_msgpack_serializer_update(struct trigger *trigger, void *event)
 {
 	(void) trigger;
 	(void) event;
-	tuple_serializer_fill();
+	tuple_serializer_update_options();
 }
 
 void
@@ -584,7 +584,7 @@ box_lua_tuple_init(struct lua_State *L)
 
 	luamp_set_encode_extension(luamp_encode_extension_box);
 
-	tuple_serializer_fill();
+	tuple_serializer_update_options();
 	trigger_create(&tuple_serializer.update_trigger,
 		       on_msgpack_serializer_update, NULL, NULL);
 	trigger_add(&luaL_msgpack_default->on_update,

  reply	other threads:[~2019-09-13 22:28 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-09-10 20:24 [tarantool-patches] [PATCH v2 0/4] Serializer bugs Vladislav Shpilevoy
2019-09-09 19:00 ` [tarantool-patches] [PATCH v2 1/4] app: serializers update now is reflected in Lua Vladislav Shpilevoy
2019-09-12 23:22   ` [tarantool-patches] " Alexander Turenko
2019-09-13 22:32     ` Vladislav Shpilevoy
2019-09-09 19:00 ` [tarantool-patches] [PATCH v2 2/4] msgpack: make msgpackffi use encode_max_depth option Vladislav Shpilevoy
2019-09-12 23:24   ` [tarantool-patches] " Alexander Turenko
2019-09-13 22:32     ` Vladislav Shpilevoy
2019-09-09 19:00 ` [tarantool-patches] [PATCH v2 3/4] tuple: use global msgpack serializer in Lua tuple Vladislav Shpilevoy
2019-09-12 23:27   ` [tarantool-patches] " Alexander Turenko
2019-09-13 22:32     ` Vladislav Shpilevoy [this message]
2019-09-09 19:00 ` [tarantool-patches] [PATCH v2 4/4] app: allow to raise an error on too nested tables Vladislav Shpilevoy
2019-09-12 23:32   ` [tarantool-patches] " Alexander Turenko
2019-09-13 22:32     ` Vladislav Shpilevoy
2019-09-10 20:25 ` [tarantool-patches] Re: [PATCH v2 0/4] Serializer bugs Vladislav Shpilevoy
2019-09-12 23:44 ` Alexander Turenko
2019-09-13 22:32   ` Vladislav Shpilevoy

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=e089763d-f4a0-c092-2665-0a593e8beff1@tarantool.org \
    --to=v.shpilevoy@tarantool.org \
    --cc=alexander.turenko@tarantool.org \
    --cc=tarantool-patches@freelists.org \
    --subject='[tarantool-patches] Re: [PATCH v2 3/4] tuple: use global msgpack serializer in Lua tuple' \
    /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