From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <v.shpilevoy@tarantool.org>
Received: from smtpng3.m.smailru.net (smtpng3.m.smailru.net [94.100.177.149])
 (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256
 bits)) (No client certificate requested)
 by dev.tarantool.org (Postfix) with ESMTPS id 33C534696C3
 for <tarantool-patches@dev.tarantool.org>;
 Fri,  3 Apr 2020 02:41:16 +0300 (MSK)
References: <cover.1585304087.git.imun@tarantool.org>
 <50fe58f83ebb4e4971528641e74f99fe2f9fd8f2.1585304087.git.imun@tarantool.org>
From: Vladislav Shpilevoy <v.shpilevoy@tarantool.org>
Message-ID: <a8c17710-dbfc-5d5c-ebc4-6f6e1d3f302d@tarantool.org>
Date: Fri, 3 Apr 2020 01:41:14 +0200
MIME-Version: 1.0
In-Reply-To: <50fe58f83ebb4e4971528641e74f99fe2f9fd8f2.1585304087.git.imun@tarantool.org>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Subject: Re: [Tarantool-patches] [PATCH luajit 1/2] jit: abort trace
 recording and execution for C API
List-Id: Tarantool development patches <tarantool-patches.dev.tarantool.org>
List-Unsubscribe: <https://lists.tarantool.org/mailman/options/tarantool-patches>, 
 <mailto:tarantool-patches-request@dev.tarantool.org?subject=unsubscribe>
List-Archive: <https://lists.tarantool.org/pipermail/tarantool-patches/>
List-Post: <mailto:tarantool-patches@dev.tarantool.org>
List-Help: <mailto:tarantool-patches-request@dev.tarantool.org?subject=help>
List-Subscribe: <https://lists.tarantool.org/mailman/listinfo/tarantool-patches>, 
 <mailto:tarantool-patches-request@dev.tarantool.org?subject=subscribe>
To: Igor Munkin <imun@tarantool.org>, Sergey Ostanevich <sergos@tarantool.org>, Kirill Yukhin <kyukhin@tarantool.org>
Cc: tarantool-patches@dev.tarantool.org

Hi! Thanks for the patch!

> diff --git a/src/lj_api.c b/src/lj_api.c
> index a5e2465..c1f53e0 100644
> --- a/src/lj_api.c
> +++ b/src/lj_api.c
> @@ -300,7 +311,7 @@ LUA_API int lua_equal(lua_State *L, int idx1, int idx2)
>        return (int)(uintptr_t)base;
>      } else {
>        L->top = base+2;
> -      lj_vm_call(L, base, 1+1);
> +      call(L, base, 1+1);

These changes look dangerous to perf. So if I had operators ==
and < overloaded for an FFI struct, they are not jitted anymore,
right? Why? Does someone really yield in them? Could you measure
how does it affect perf of these calls? From what I see, these
changes basically kill FFI making it not better than Lua C. It no
longer helps JITting anything, does it?

The same for concat, and other things. Looks like an overkill. We
would need sandwich protection only for a few places. Is it possible
to avoid such drastic changes?