Re: [HACKERS] Hash joins vs small-integer join values

2007-06-01 Thread Tom Lane
Gregory Stark <[EMAIL PROTECTED]> writes: > hashany of a 4-byte value degenerates to pretty much just a call to mix(). > Perhaps we should just expose a hash12() that takes three integers and calls > mix() on them like hash_any does. I don't see any use for that, but probably there is a use in has

Re: [HACKERS] Hash joins vs small-integer join values

2007-05-31 Thread Tom Lane
Josh Berkus <[EMAIL PROTECTED]> writes: > Tom, >> The problem is that this is not a back-patchable >> answer, because changing the hash functions would corrupt existing hash >> indexes. > Does anyone *use* hash indexes? We get bug reports on 'em, so yes ... regards, tom

Re: [HACKERS] Hash joins vs small-integer join values

2007-05-31 Thread Gregory Stark
"Tom Lane" <[EMAIL PROTECTED]> writes: > The best idea I can come up with for the back branches is > to make ExecHashGetBucketAndBatch do hash_any internally, say hashany of a 4-byte value degenerates to pretty much just a call to mix(). Perhaps we should just expose a hash12() that takes three

Re: [HACKERS] Hash joins vs small-integer join values

2007-05-31 Thread Josh Berkus
Tom, > The problem is that this is not a back-patchable answer, because changing the hash functions would corrupt existing hash indexes. Does anyone *use* hash indexes? Comments, better ideas? I was just talking to Luke today and he said they had a considerable amount of cleanup on hash