This is an interesting idea.
Although some catalog tables are not in catcaches,
such as pg_depend, when scanning them, if there is any
SharedInvalidationMessage, the CatalogSnapshot
will be invalidated and recreated ("RelationInvalidatesSnapshotsOnly"
in syscache.c)
Maybe during the system_scan,
On Fri, Jan 12, 2024 at 03:47:13PM -0500, Tom Lane wrote:
> I wrote:
> > This is uncomfortably much in bed with the tuple table slot code,
> > perhaps, but I don't see a way to do it more cleanly unless we want
> > to add some new provisions to that API. Andres, do you have any
> > thoughts about
I wrote:
> Xiaoran Wang writes:
>> Hmm, how about first checking if any invalidated shared messages have been
>> accepted, then rechecking the tuple's visibility?
>> If there is no invalidated shared message accepted during
>> 'toast_flatten_tuple',
>> there is no need to do then visibility
Xiaoran Wang writes:
> Hmm, how about first checking if any invalidated shared messages have been
> accepted, then rechecking the tuple's visibility?
> If there is no invalidated shared message accepted during
> 'toast_flatten_tuple',
> there is no need to do then visibility check, then it can
Hmm, how about first checking if any invalidated shared messages have been
accepted, then rechecking the tuple's visibility?
If there is no invalidated shared message accepted during
'toast_flatten_tuple',
there is no need to do then visibility check, then it can save several
CPU cycles.
Great! That's what exactly we need.
The patch LGTM, +1
Tom Lane 于2024年1月13日周六 04:47写道：
> I wrote:
> > This is uncomfortably much in bed with the tuple table slot code,
> > perhaps, but I don't see a way to do it more cleanly unless we want
> > to add some new provisions to that API. Andres,
I wrote:
> This is uncomfortably much in bed with the tuple table slot code,
> perhaps, but I don't see a way to do it more cleanly unless we want
> to add some new provisions to that API. Andres, do you have any
> thoughts about that?
Oh! After nosing around a bit more I remembered
Xiaoran Wang writes:
>> Maybe, but that undocumented hack in SetHintBits seems completely
>> unacceptable. Isn't there a cleaner way to make this check?
> Maybe we don't need to call 'HeapTupleSatisfiesVisibility' to check if the
> tuple has been deleted.
> As the tuple's xmin must been
> Also, I'm pretty dubious that GetNonHistoricCatalogSnapshot rather
> than GetCatalogSnapshot is the right thing, because the catcaches
> use the latter.
Yes, you are right, should use GetCatalogSnapshot here.
> Maybe, but that undocumented hack in SetHintBits seems completely
> unacceptable.
Xiaoran Wang writes:
>>> The detection of "get an invalidation" could be refined: what I did
>>> here is to check for any advance of SharedInvalidMessageCounter,
>>> which clearly will have a significant number of false positives.
> I have reviewed your patch, and it looks good. But instead of
Hi,
>> BTW, while nosing around I found what seems like a very nasty related
>> bug. Suppose that a catalog tuple being loaded into syscache contains
>> some toasted fields. CatalogCacheCreateEntry will flatten the tuple,
>> involving fetches from toast tables that will certainly cause
>>
I wrote:
> In bug #18163 [1], Alexander proved the misgivings I had in [2]
> about catcache detoasting being possibly unsafe:
> ...
> Attached is a POC patch for fixing this.
The cfbot pointed out that this needed a rebase. No substantive
changes.
regards, tom lane
diff
In bug #18163 [1], Alexander proved the misgivings I had in [2]
about catcache detoasting being possibly unsafe:
>> BTW, while nosing around I found what seems like a very nasty related
>> bug. Suppose that a catalog tuple being loaded into syscache contains
>> some toasted fields.
13 matches
Mail list logo
