From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <alexander.turenko@tarantool.org>
Received: from smtp17.mail.ru (smtp17.mail.ru [94.100.176.154])
 (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 C28644765E0
 for <tarantool-patches@dev.tarantool.org>;
 Tue, 29 Dec 2020 23:10:47 +0300 (MSK)
Date: Tue, 29 Dec 2020 23:11:02 +0300
From: Alexander Turenko <alexander.turenko@tarantool.org>
Message-ID: <20201229201102.cmrgcrzvlenyn7wf@tkn_work_nb>
References: <cover.1609259010.git.artemreyt@tarantool.org>
 <72c07be7cadc1e78c9394bbc8f1bb64b1a4a4a0c.1609259010.git.artemreyt@tarantool.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <72c07be7cadc1e78c9394bbc8f1bb64b1a4a4a0c.1609259010.git.artemreyt@tarantool.org>
Subject: Re: [Tarantool-patches] [PATCH 1/3] test: tarantool -e test add
 more details and avoid busy loop
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: Artem Starshov <artemreyt@tarantool.org>
Cc: tarantool-patches@dev.tarantool.org

On Tue, Dec 29, 2020 at 07:25:33PM +0300, Artem Starshov wrote:
> From: Alexander Turenko <alexander.turenko@tarantool.org>

I would prefer if you would send patches from youself.

> 
> Background: the test fails on waiting the process termination, but
> surprisingly only on GitHub hosted runners.
> 
> My guess is that the process becomes zombie, but we don't collect it.

I wrote this as brief notes for my experiments, then we observed that it
is not so. The process is not a direct child of our one.

> 
> A busy loop is not good in general. It is better to pass execution to
> other fibers if there is nothing to do. This way the waiting will not
> eat 100% CPU time.
> 
> Added collect more details in tarantool -e 'assert(false)' test.
> ---
>  .../gh-4983-tnt-e-assert-false-hangs.test.lua       | 13 ++++++++++++-
>  1 file changed, 12 insertions(+), 1 deletion(-)
> 
> diff --git a/test/app-tap/gh-4983-tnt-e-assert-false-hangs.test.lua b/test/app-tap/gh-4983-tnt-e-assert-false-hangs.test.lua
> index 656fe212b..ff3281e0f 100755
> --- a/test/app-tap/gh-4983-tnt-e-assert-false-hangs.test.lua
> +++ b/test/app-tap/gh-4983-tnt-e-assert-false-hangs.test.lua
> @@ -31,6 +31,7 @@ local function wait_process_completion(pid, timeout)
>              process_completed = true
>              break
>          end
> +        fiber.sleep(0.01)
>      end
>      return process_completed
>  end
> @@ -74,8 +75,18 @@ local res = tap.test('gh-4983-tnt-e-assert-false-hangs', function(test)
>      local process_completed = wait_process_completion(pid,
>              process_waiting_timeout)
>  
> +    local details
> +    pcall(function()
> +        details = {
> +            cmdline = fio.open(('/proc/%d/cmdline'):format(pid),
> +                               {'O_RDONLY'}):read(1000000),
> +            status = fio.open(('/proc/%d/status'):format(pid),
> +                              {'O_RDONLY'}):read(1000000),
> +        }
> +    end)

It is Linux specific, so the change is dubious.

>      test:ok(process_completed,
> -            ('tarantool process with pid = %d completed'):format(pid))
> +            ('tarantool process with pid = %d completed'):format(pid),
> +            details)
>  
>      -- Kill process if hangs.
>      if not process_completed then ffi.C.kill(pid, 9) end
> -- 
> 2.28.0
>