Re: Memory buffer alignment
| От | Tom Lane |
|---|---|
| Тема | Re: Memory buffer alignment |
| Дата | |
| Msg-id | 3149.1063921594@sss.pgh.pa.us обсуждение исходный текст |
| Ответ на | Memory buffer alignment (Manfred Spraul <manfred@colorfullife.com>) |
| Ответы |
Re: Memory buffer alignment
|
| Список | pgsql-hackers |
Manfred Spraul <manfred@colorfullife.com> writes:
> Unfortunately it's the slower one: Intel cpus have a microcode fastpath
> for memcopies that are 8-byte aligned. This fastpath is around 50%
> faster than the manual copy that is used for "misaligned" (i.e. only
> 4-byte aligned) pointers.
Maybe it'd be worth setting MAXIMUM_ALIGNOF to 8 on such CPUs? Or at
least hacking ShmemAlloc and friends to use 8-byte alignment. I assume
the major issue here is that the shared buffers don't get 8-byte-aligned
within the shared memory segment.
Are there any machines where it'd be worth forcing an even larger
alignment for the buffers?
regards, tom lane
В списке pgsql-hackers по дате отправления: