From: | Abhijit Menon-Sen <ams(at)2ndQuadrant(dot)com> |
---|---|
To: | Andres Freund <andres(at)2ndquadrant(dot)com> |
Cc: | pgsql-hackers(at)postgresql(dot)org, Heikki Linnakangas <hlinnakangas(at)vmware(dot)com>, Robert Haas <robertmhaas(at)gmail(dot)com> |
Subject: | Re: What exactly is our CRC algorithm? |
Date: | 2014-12-29 13:14:18 |
Message-ID: | 20141229131418.GA14274@toroid.org |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
At 2014-12-29 13:22:28 +0100, andres(at)2ndquadrant(dot)com wrote:
>
> How about pg_choose_crc_impl() or something?
Done.
> _sb8? Unless I miss something it's not slice by 8 but rather bytewise?
This is meant to apply on top of the earlier patches I posted to
implement slice-by-8. I'll attach both here.
> Should be marked inline.
Done (and the other one too).
> > +#ifdef __GNUC__
> > + __asm__ ("crc32b %[data], %[crc]\n" : [crc] "+r" (crc) : [data] "rm" (data));
>
> Have you checked which version of gcc introduced named references to
> input/output parameters?
No. The documentation calls them "asmSymbolicName"s, but I can't find
the term (or various likely alternatives, e.g. symbolic_operand may be
related) either in the release notes, the changelog, or the source (on
Github). Does anyone know, or know how to find out?
Meanwhile, I have attached the two patches with the other modifications.
-- Abhijit
Attachment | Content-Type | Size |
---|---|---|
sb8.diff | text/x-diff | 32.6 KB |
crc.diff | text/x-diff | 6.8 KB |
From | Date | Subject | |
---|---|---|---|
Next Message | Kevin Grittner | 2014-12-29 14:03:28 | Re: BUG #12330: ACID is broken for unique constraints |
Previous Message | Andres Freund | 2014-12-29 12:22:28 | Re: What exactly is our CRC algorithm? |