Content-Type: multipart/alternative;
 boundary="------------vC4nwfYQCCgAkNe7Spcn0JjI"
Message-ID: <09677462-ed4c-2b2c-7918-8a3961dc772a@gmail.com>
Date: Wed, 27 Dec 2023 10:00:00 +0300
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.4.2
Subject: Re: BUG #18259: Assertion in ExtendBufferedRelLocal() fails after
 no-space-left condition
Content-Language: en-US
To: tender wang <tndrwang@gmail.com>, pgsql-bugs@lists.postgresql.org,
 Andres Freund <andres@anarazel.de>
References: <18259-6e256429825dd435@postgresql.org>
 <CAHewXNm4ktpPy=9g7NAYV=BoW-Ao3taZREPTcXnJDZ6P5esiag@mail.gmail.com>
 <CAHewXNnayN3NM1HfaOCejk=sGfSva6ZDArWxKiTxL7PdDHRtMw@mail.gmail.com>
From: Alexander Lakhin <exclusion@gmail.com>
In-Reply-To: 
 <CAHewXNnayN3NM1HfaOCejk=sGfSva6ZDArWxKiTxL7PdDHRtMw@mail.gmail.com>
Archived-At: 
 <https://www.postgresql.org/message-id/09677462-ed4c-2b2c-7918-8a3961dc772a%40gmail.com>
Precedence: bulk

This is a multi-part message in MIME format.
--------------vC4nwfYQCCgAkNe7Spcn0JjI
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit

Hello tender wang,

26.12.2023 19:55, tender wang write:
> I tried to analyze the issue, and I found that it might be caused by this commit:
> commit dad50f677c42de207168a3f08982ba23c9fc6720
>        bufmgr: Acquire and clean victim buffer separately
>

Thanks for looking into it!

>     ...
>
>         With debug logging added in this code within ExtendBufferedRelLocal():
>                 if (found)
>                 {
>                     BufferDesc *existing_hdr =
>         GetLocalBufferDescriptor(hresult->id);
>                     uint32      buf_state;
>
>         UnpinLocalBuffer(BufferDescriptorGetBuffer(victim_buf_hdr));
>
>                     existing_hdr = GetLocalBufferDescriptor(hresult->id);
>                     PinLocalBuffer(existing_hdr, false);
>                     buffers[i] = BufferDescriptorGetBuffer(existing_hdr);
>
>                     buf_state = pg_atomic_read_u32(&existing_hdr->state);
>                     Assert(buf_state & BM_TAG_VALID);
>                     Assert(!(buf_state & BM_DIRTY));
>                     buf_state &= BM_VALID;
>         pg_atomic_unlocked_write_u32(&existing_hdr->state, buf_state);
>         ...
>         I see that it reached for the second INSERT (and NOSPC error) with
>         existing_hdr->state == 0x2040000, but for the third INSERT I observe
>         state == 0x0.
>

I wonder, if "buf_state &= BM_VALID" is a typo here, maybe it supposed to be
"buf_state &= ~BM_VALID" as in ExtendBufferedRelShared()...

Best regards,
Alexander
--------------vC4nwfYQCCgAkNe7Spcn0JjI
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: 8bit

<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <div class="moz-cite-prefix">Hello tender wang,<br>
      <br>
      26.12.2023 19:55, tender wang write:<br>
    </div>
    <blockquote type="cite"
cite="mid:CAHewXNnayN3NM1HfaOCejk=sGfSva6ZDArWxKiTxL7PdDHRtMw@mail.gmail.com">
      <meta http-equiv="content-type" content="text/html; charset=UTF-8">
      <div dir="ltr">
        <div dir="ltr">I tried to analyze the issue, and I found that it
          might be caused by this commit:<br>
          <div>commit dad50f677c42de207168a3f08982ba23c9fc6720<br>
          </div>
          <div>       bufmgr: Acquire and clean victim buffer separately</div>
          <div><br>
          </div>
        </div>
      </div>
    </blockquote>
    <br>
    Thanks for looking into it!<br>
    <br>
    <blockquote type="cite"
cite="mid:CAHewXNnayN3NM1HfaOCejk=sGfSva6ZDArWxKiTxL7PdDHRtMw@mail.gmail.com">
      <div dir="ltr">
        <div class="gmail_quote">
          <blockquote class="gmail_quote" style="margin:0px 0px 0px
            0.8ex;border-left:1px solid
            rgb(204,204,204);padding-left:1ex">...<br>
            <div class="gmail_quote"><br>
              <blockquote class="gmail_quote" style="margin:0px 0px 0px
                0.8ex;border-left:1px solid
                rgb(204,204,204);padding-left:1ex">
                With debug logging added in this code within
                ExtendBufferedRelLocal():<br>
                        if (found)<br>
                        {<br>
                            BufferDesc *existing_hdr =<br>
                GetLocalBufferDescriptor(hresult-&gt;id);<br>
                            uint32      buf_state;<br>
                <br>
                           
                UnpinLocalBuffer(BufferDescriptorGetBuffer(victim_buf_hdr));<br>
                <br>
                            existing_hdr =
                GetLocalBufferDescriptor(hresult-&gt;id);<br>
                            PinLocalBuffer(existing_hdr, false);<br>
                            buffers[i] =
                BufferDescriptorGetBuffer(existing_hdr);<br>
                <br>
                            buf_state =
                pg_atomic_read_u32(&amp;existing_hdr-&gt;state);<br>
                            Assert(buf_state &amp; BM_TAG_VALID);<br>
                            Assert(!(buf_state &amp; BM_DIRTY));<br>
                            buf_state &amp;= BM_VALID;<br>
                           
                pg_atomic_unlocked_write_u32(&amp;existing_hdr-&gt;state,
                buf_state);<br>
                ...<br>
                I see that it reached for the second INSERT (and NOSPC
                error) with<br>
                existing_hdr-&gt;state == 0x2040000, but for the third
                INSERT I observe<br>
                state == 0x0.<br>
                <br>
              </blockquote>
            </div>
          </blockquote>
        </div>
      </div>
    </blockquote>
    <br>
    I wonder, if "buf_state &amp;= BM_VALID" is a typo here, maybe it
    supposed to be<br>
    "buf_state &amp;= ~BM_VALID" as in ExtendBufferedRelShared()...<br>
    <br>
    Best regards,<br>
    Alexander<br>
  </body>
</html>

--------------vC4nwfYQCCgAkNe7Spcn0JjI--