| From: | Heikki Linnakangas <hlinnakangas(at)vmware(dot)com> |
|---|---|
| To: | Kevin Grittner <kgrittn(at)ymail(dot)com> |
| Cc: | Robert Haas <robertmhaas(at)gmail(dot)com>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Race condition in b-tree page deletion |
| Date: | 2014-01-28 11:13:43 |
| Message-ID: | 52E790E7.60400@vmware.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On 01/27/2014 04:34 PM, Heikki Linnakangas wrote:
> While fixing the above issues, and reviewing this again in general, I
> noticed that btree_xlog_unlink_page() was a few bricks shy of a load.
> When unlinking an internal page, it didn't update the pointer to the new
> topmost page in the to-be-deleted chain. Fixed that.
>
> Attached is a new version. I also re-worded the README changes; I hope
> it reads better now.
I dusted off the scripts I posted yesterday, to visualize b-tree
structure using graphviz, and tried it on this patch. I found a number
of bugs. Attached is a new version with those bugs fixed.
- Heikki
| Attachment | Content-Type | Size |
|---|---|---|
| fix-btree-page-deletion-4.patch | text/x-diff | 64.5 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andres Freund | 2014-01-28 11:28:35 | Re: Observed Compilation warning in WIN32 build |
| Previous Message | Kyotaro HORIGUCHI | 2014-01-28 11:12:37 | Re: Retain dynamic shared memory segments for postmaster lifetime |