Horton

Thread

Commits

  1. Remove duplicate check added by commit b2a5545bd6.

  2. Don't archive bogus recycled or preallocated files after timeline switch.

  1. Remove a redundant condition check

    Ádám Balogh <adam.balogh@ericsson.com> — 2020-06-25T15:27:14Z

    Hello,

A one line change to remove a duplicate check. This duplicate check was detected during testing my contribution to a static code analysis tool. There is no functional change, no new tests needed.

Regards,

Ádám Balogh
CodeChecker Team
Ericsson Hungary

  2. Re: Remove a redundant condition check

    Amit Kapila <amit.kapila16@gmail.com> — 2020-06-26T09:09:22Z

    On Thu, Jun 25, 2020 at 11:23 PM Ádám Balogh <adam.balogh@ericsson.com> wrote:
>
>
> A one line change to remove a duplicate check. This duplicate check was detected during testing my contribution to a static code analysis tool. There is no functional change, no new tests needed.
>
>

Yeah, this duplicate check is added as part of commit b2a5545bd6.  See
below part of change.

- /*
- * If this record was a timeline switch, wake up any
- * walsenders to notice that we are on a new timeline.
- */
- if (switchedTLI && AllowCascadeReplication())
- WalSndWakeup();
+ /* Is this a timeline switch? */
+ if (switchedTLI)
+ {
+ /*
+ * Before we continue on the new timeline, clean up any
+ * (possibly bogus) future WAL segments on the old timeline.
+ */
+ RemoveNonParentXlogFiles(EndRecPtr, ThisTimeLineID);
+
+ /*
+ * Wake up any walsenders to notice that we are on a new
+ * timeline.
+ */
+ if (switchedTLI && AllowCascadeReplication())
+ WalSndWakeup();
+ }

It seems we forgot to remove the additional check for switchedTLI
while adding a new check.  I think we can remove this duplicate check
in the HEAD code.  I am not sure if it is worth to backpatch such a
change.

-- 
With Regards,
Amit Kapila.
EnterpriseDB: http://www.enterprisedb.com

  3. Re: Remove a redundant condition check

    Michael Paquier <michael@paquier.xyz> — 2020-06-26T10:02:39Z

    On Fri, Jun 26, 2020 at 02:39:22PM +0530, Amit Kapila wrote:
> It seems we forgot to remove the additional check for switchedTLI
> while adding a new check.  I think we can remove this duplicate check
> in the HEAD code.  I am not sure if it is worth to backpatch such a
> change.

Yes, there is no point to keep this check so let's clean up this
code.  I also see no need to do a backpatch here, this is purely
cosmetic.
--
Michael

  4. RE: Remove a redundant condition check

    Ádám Balogh <adam.balogh@ericsson.com> — 2020-06-26T10:43:48Z

    Hello,

-----Original Message-----
From: Amit Kapila <amit.kapila16@gmail.com> 
Sent: 2020. június 26., péntek 11:09
To: Ádám Balogh <adam.balogh@ericsson.com>
Cc: PostgreSQL Hackers <pgsql-hackers@postgresql.org>
Subject: Re: Remove a redundant condition check

>On Thu, Jun 25, 2020 at 11:23 PM Ádám Balogh <adam.balogh@ericsson.com> wrote:
>>
>>
>> A one line change to remove a duplicate check. This duplicate check was detected during testing my contribution to a static code analysis tool. There is no functional change, no new tests needed.
>
> Yeah, this duplicate check is added as part of commit b2a5545bd6.  See below part of change.
>
> - /*
> - * If this record was a timeline switch, wake up any
> - * walsenders to notice that we are on a new timeline.
> - */
> - if (switchedTLI && AllowCascadeReplication())
> - WalSndWakeup();
> + /* Is this a timeline switch? */
> + if (switchedTLI)
> + {
> + /*
> + * Before we continue on the new timeline, clean up any
> + * (possibly bogus) future WAL segments on the old timeline.
> + */
> + RemoveNonParentXlogFiles(EndRecPtr, ThisTimeLineID);
> +
> + /*
> + * Wake up any walsenders to notice that we are on a new
> + * timeline.
> + */
> + if (switchedTLI && AllowCascadeReplication()) WalSndWakeup(); }
>
> It seems we forgot to remove the additional check for switchedTLI while adding a new check.  I think we can remove this duplicate > > check in the HEAD code.  I am not sure if it is worth to backpatch such a change.

Thank you for confirming it. I do not think it is worth to backpatch, it is just a readability issue. 
Regards,

Ádám

  5. Re: Remove a redundant condition check

    Ranier Vilela <ranier.vf@gmail.com> — 2020-06-26T11:14:55Z

    Em sex., 26 de jun. de 2020 às 06:09, Amit Kapila <amit.kapila16@gmail.com>
escreveu:

> On Thu, Jun 25, 2020 at 11:23 PM Ádám Balogh <adam.balogh@ericsson.com>
> wrote:
> >
> >
> > A one line change to remove a duplicate check. This duplicate check was
> detected during testing my contribution to a static code analysis tool.
> There is no functional change, no new tests needed.
> >
> >
>
> Yeah, this duplicate check is added as part of commit b2a5545bd6.  See
> below part of change.
>
> - /*
> - * If this record was a timeline switch, wake up any
> - * walsenders to notice that we are on a new timeline.
> - */
> - if (switchedTLI && AllowCascadeReplication())
> - WalSndWakeup();
> + /* Is this a timeline switch? */
> + if (switchedTLI)
> + {
> + /*
> + * Before we continue on the new timeline, clean up any
> + * (possibly bogus) future WAL segments on the old timeline.
> + */
> + RemoveNonParentXlogFiles(EndRecPtr, ThisTimeLineID);
> +
> + /*
> + * Wake up any walsenders to notice that we are on a new
> + * timeline.
> + */
> + if (switchedTLI && AllowCascadeReplication())
> + WalSndWakeup();
> + }
>
> It seems we forgot to remove the additional check for switchedTLI
> while adding a new check.  I think we can remove this duplicate check
> in the HEAD code.  I am not sure if it is worth to backpatch such a
> change.
>
+1
Great to know, that this is finally going to be fixed. (1)

regards,
Ranier Vilela
1.
https://www.postgresql.org/message-id/CAEudQAocMqfqt0t64HNo39Z73jMey60WmeryB%2BWFDg3BZpCf%3Dg%40mail.gmail.com

  6. Re: Remove a redundant condition check

    Amit Kapila <amit.kapila16@gmail.com> — 2020-06-27T05:22:49Z

    On Fri, Jun 26, 2020 at 3:32 PM Michael Paquier <michael@paquier.xyz> wrote:
>
> On Fri, Jun 26, 2020 at 02:39:22PM +0530, Amit Kapila wrote:
> > It seems we forgot to remove the additional check for switchedTLI
> > while adding a new check.  I think we can remove this duplicate check
> > in the HEAD code.  I am not sure if it is worth to backpatch such a
> > change.
>
> Yes, there is no point to keep this check so let's clean up this
> code.  I also see no need to do a backpatch here, this is purely
> cosmetic.
>

Thanks for the confirmation, pushed!

-- 
With Regards,
Amit Kapila.
EnterpriseDB: http://www.enterprisedb.com