| From: | Heikki Linnakangas <heikki(dot)linnakangas(at)enterprisedb(dot)com> |
|---|---|
| To: | Jan Urbański <wulczer(at)wulczer(dot)org> |
| Cc: | Stephen Frost <sfrost(at)snowman(dot)net>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, David Fetter <david(at)fetter(dot)org>, PG Hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Re: [COMMITTERS] pgsql: Fix plpgsql to release SPI plans when a function or DO block is |
| Date: | 2011-03-31 09:43:26 |
| Message-ID: | 4D944CBE.1050307@enterprisedb.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-committers pgsql-hackers |
On 31.03.2011 12:30, Jan Urbański wrote:
> On 31/03/11 07:35, Heikki Linnakangas wrote:
>> On 30.03.2011 21:21, Jan Urbański wrote:
>>> Valgrind showed me the way. PFA a trivial patch to avoid leaking a
>>> PLyProcedure struct in inline blocks.
>>
>> Hmm, any reason the PLyProcedure struct needs to be allocated in
>> TopMemoryContext in the first place? Could you palloc0 it in a
>> shorter-lived context, or even better, just allocate it in stack?
>
> Yeah, you're right, you can keep it on the stack.
>
>> PS. I don't think the volatile qualifier in 'proc' is in necessary. The
>> variable is not changed in PG_TRY-block.
>
> That always confuses me, but I guess you're right, the variable does not
> change, only the memory it points to.
Ok then, committed.
--
Heikki Linnakangas
EnterpriseDB http://www.enterprisedb.com
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Heikki Linnakangas | 2011-03-31 10:41:35 | pgsql: Increase SHMEM_INDEX_SIZE from 32 to 64. We're currently at 40 e |
| Previous Message | Heikki Linnakangas | 2011-03-31 09:43:01 | pgsql: Don't leak the temporary PLyProcedure struct we create for inlin |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Noah Misch | 2011-03-31 10:06:49 | Re: BUG #5856: pg_attribute.attinhcount is not correct. |
| Previous Message | Jan Urbański | 2011-03-31 09:30:29 | Re: Re: [COMMITTERS] pgsql: Fix plpgsql to release SPI plans when a function or DO block is |