Re: Uh, this is *not* a 64-bit CRC ...
От | Vadim Mikheev |
---|---|
Тема | Re: Uh, this is *not* a 64-bit CRC ... |
Дата | |
Msg-id | 046c01c0a6d9$d9470c40$4879583f@sectorbase.com обсуждение исходный текст |
Ответ на | Uh, this is *not* a 64-bit CRC ... (Tom Lane <tgl@sss.pgh.pa.us>) |
Ответы |
Re: Re: Uh, this is *not* a 64-bit CRC ...
|
Список | pgsql-hackers |
> This isn't a 64-bit CRC. It's two independent 32-bit CRCs, one done > on just the odd-numbered bytes and one on just the even-numbered bytes > of the datastream. That's hardly any stronger than a single 32-bit CRC; I believe that the longer data the more chance to get same CRC/hash for different data sets (if data length > CRC/hash length). Or am I wrong? Having no crc64 implementation (see below) I decided to use 2 CRC32 instead of one - it looked better, without any additional cost (record header is 8 byte aligned anyway, on, mmm, most platform). > it's certainly not what I thought we had agreed to implement. I've asked if anyone can send crc64 impl to me and got only one from Nathan Myers. Unfortunately, SWISS-PROT impl assumes that long long is 8 bytes - is it portable? Vadim
В списке pgsql-hackers по дате отправления: