Re: SQL/JSON features for v15

[Jonathan S. Katz]

On 8/19/22 10:11 AM, Jonathan S. Katz wrote:

Hi,

On 8/17/22 11:45 PM, Nikita Glukhov wrote:

Hi,

On 17.08.2022 04:45, Jonathan S. Katz wrote:

On 8/15/22 10:14 PM, Andres Freund wrote:

I pushed a few cleanups to https://github.com/anarazel/postgres/commits/json while I was hacking on this (ignore that it's based on the meson tree, that's just faster for me). Some of them might not be applicable anymore, but it

might still make sense for you to look at.

1. Nikita: Did you have a chance to review Andres's changes as well?

With RMT hat on, this appears to be making progress. A few questions / comments for the group:

Yes, I have reviewed Andres's changes, they all are ok.

Thank you!

Then I started to do on the top of it other fixes that help to avoid
subtransactions when they are not needed. And it ended in the new
refactoring of coercion code.  Also I moved here from v6-0003 fix of
ExecEvalJsonNeedSubtransaction() which considers more cases.

Great.

Andres, Robert: Do these changes address your concerns about the use of substransactions and reduce the risk of xid wraparound?

On 16.08.2022 05:14, Andres Freund wrote:

But for JIT I still had to construct additional ExprState with a
function compiled from subexpression steps.

Why did you have to do this?

I simply did not dare to implement compilation of recursively-callable
function with additional parameter stepno.  In the v8 patch I did it
by adding a switch with all possible jump addresses of EEOP_SUBTRANS
steps in the beginning of the function.  And it really seems to work
faster, but needs more exploration.  See patch 0003, where both
variants preserved using #ifdef.


The desciprion of the v7 patches:

0001 Simplify JsonExpr execution
  Andres's changes + mine:
   - Added JsonCoercionType enum, fields like via_io replaced with it
   - Emit only context item steps in JSON_TABLE_OP case
   - Skip coercion of NULLs to non-domain types (is it correct?)

0002 Fix returning of json[b] domains in JSON_VALUE:
   simply rebase of v6 onto 0001

0003 Add EEOP_SUBTRANS executor step
   v6 + new recursive JIT

0004 Split JsonExpr execution into steps
   simply rebase of v6 + used LLMBuildSwitch() in EEOP_JSONEXPR

What do folks think of these patches?

Thanks,

Jonathan

Andres, Andrew, Amit, Robert -- as you have either worked on this or expressed opinions -- any thoughts on this current patch set?

OpenPGP_signature
Description: OpenPGP digital signature