JIT works only partially with meson build?

[pg-hub-mirror]

Read-only mirror. Reply and review on pgsql-hackers; activity here is not sent upstream.

• Original author: shihao zhong <zhong950419(at)gmail(dot)com>
• Mailing list: pgsql-hackers
• Message-ID: CAGRkXqQd26aCrGX03fXiV9MWZmAtYsFDd9N5HzHOxsiSfoKsXA@mail.gmail.com
• Original email

Patch files:

• v3_fix_doc_jit_meson.patch

---

Hi Yugo,
cfbot fails because v2 is a plain diff, and git am cannot apply it.
Attached v3 is the same change made with git format-patch, plus a
commit message. It applies to master and REL_16_STABLE through
REL_19_STABLE.
One suggestion: EXPLAIN still shows "Inlining: true" in this case,
since it only checks PGJIT_INLINE. Maybe the note should say so.
Regards,
Shihao

[pg-hub-mirror]

Yugo Nagata <nagata(at)sraoss(dot)co(dot)jp> via pgsql-hackers · original email

Hi Shihao,
On Thu, 17 Sep 2026 22:13:02 -0400
shihao zhong <zhong950419(at)gmail(dot)com> wrote:

Hi Yugo,

cfbot fails because v2 is a plain diff, and git am cannot apply it.
Attached v3 is the same change made with git format-patch, plus a
commit message. It applies to master and REL_16_STABLE through
REL_19_STABLE.

One suggestion: EXPLAIN still shows "Inlining: true" in this case,
since it only checks PGJIT_INLINE. Maybe the note should say so.
Thak you for updating the patch.
I've attached an updated path incorporating your suggestion.
Regards,]
Yugo Nagata
--
Yugo Nagata <nagata(at)sraoss(dot)co(dot)jp>

New patch set: v4-0001-doc-Mention-that-JIT-inlining-does-not-work-with-.patch

[pg-hub-mirror]

Yugo Nagata <nagata(at)sraoss(dot)co(dot)jp> via pgsql-hackers · original email

On Fri, 18 Sep 2026 21:00:20 +0900
Yugo Nagata <nagata(at)sraoss(dot)co(dot)jp> wrote:

Hi Shihao,

On Thu, 17 Sep 2026 22:13:02 -0400
shihao zhong <zhong950419(at)gmail(dot)com> wrote:

Hi Yugo,

cfbot fails because v2 is a plain diff, and git am cannot apply it.
Attached v3 is the same change made with git format-patch, plus a
commit message. It applies to master and REL_16_STABLE through
REL_19_STABLE.

One suggestion: EXPLAIN still shows "Inlining: true" in this case,
since it only checks PGJIT_INLINE. Maybe the note should say so.

Thak you for updating the patch.
I've attached an updated path incorporating your suggestion.
It may be unnecessary once the LLVM bitcode emission patches [1] are
committed, though.
[1] https://www.postgresql.org/message-id/flat/CAN55FZ26ohMftt20NUtVAxU8Feqjhz-jEfUvFKATYEaRR%2Bxmqg%40mail.gmail.com#a7b90ab42f2c38fcddca74f063741c9d
Regards,
Yugo Nagata
--
Yugo Nagata <nagata(at)sraoss(dot)co(dot)jp>