Skip to content

fix: WHPG compatibility #444

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
usamoi merged 3 commits into from
Mar 24, 2026
+11 −10
Merged

fix: WHPG compatibility #444

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
2f50fcb
Do not hardcode 8k but use BLCKSZ to accomodate whpg's 32k blocks
Ngalstyan4 Feb 21, 2026
0c20146
Reduce default index creation verbosity by reducing log level of a fe…
Ngalstyan4 Feb 18, 2026
19d5b3e
Fix sphere() functions to use explicit cast for WHPG compatibility
Ngalstyan4 Feb 24, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 4 additions & 3 deletions src/index/vchordrq/am/am_build.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -517,9 +517,10 @@ mod vchordrq_cached {
}
let pages_s = buffer.len();
buffer.extend(pages.iter().flat_map(|x| unsafe {
std::mem::transmute::<&PostgresPage<vchordrq::Opaque>, &[u8; 8192]>(
x.as_ref(),
)
std::mem::transmute::<
&PostgresPage<vchordrq::Opaque>,
&[u8; pgrx::pg_sys::BLCKSZ as usize],
>(x.as_ref())
}));
let pages_e = buffer.len();
while buffer.len() % ALIGN != 0 {
Expand Down
6 changes: 3 additions & 3 deletions src/index/vchordrq/dispatch.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -179,15 +179,15 @@ pub fn maintain<R>(
check,
),
};
pgrx::info!(
pgrx::debug1!(
Copy link
Copy Markdown
Contributor

@usamoi usamoi Mar 18, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why do this?

Copy link
Copy Markdown
Contributor Author

@Ngalstyan4 Ngalstyan4 Mar 18, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Everywhere else in postgres or common extensions, CREATE EXTENSION does not produce any logs in the common success case. This change brings vchord to par with this expectation.

It is annoying when I am directly running commands in psql and unnecessary output logs move more important commands and outputs out of my terminal view.

Copy link
Copy Markdown
Contributor

@usamoi usamoi Mar 19, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Everywhere else in postgres or common extensions, CREATE EXTENSION does not produce any logs in the common success case.

Do you mean CREATE INDEX? VectorChord should not print logs in CREATE EXTENSION.

This change brings vchord to par with this expectation.

We consider this a feature, so it would be inappropriate to hide it directly.

It is annoying when I am directly running commands in psql and unnecessary output logs move more important commands and outputs out of my terminal view.

You can change all pgrx::info in the code to pgrx::notice. Then, use SET client_min_messages = warning to hide these messages.

Copy link
Copy Markdown
Contributor

@usamoi usamoi Mar 19, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I found that you're not hiding the logs of clustering. I think this change is reasonable, although I don't understand why you ignore the 10+ lines of clustering-related logs above and only focus on the 3 lines at the bottom.

Copy link
Copy Markdown
Contributor Author

@Ngalstyan4 Ngalstyan4 Mar 19, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ah, those had not triggered in my test run. now I see what you mean. happy to turn those into debug1 as well if that works for you!

Copy link
Copy Markdown
Contributor

@usamoi usamoi Mar 20, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

happy to turn those into debug1 as well if that works for you!

Keep it as it is.

"maintain: number_of_formerly_allocated_pages = {}",
maintain.number_of_formerly_allocated_pages
);
pgrx::info!(
pgrx::debug1!(
"maintain: number_of_freshly_allocated_pages = {}",
maintain.number_of_freshly_allocated_pages
);
pgrx::info!(
pgrx::debug1!(
"maintain: number_of_freed_pages = {}",
maintain.number_of_freed_pages
);
Expand Down
8 changes: 4 additions & 4 deletions src/sql/finalize.sql
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -189,16 +189,16 @@ CREATE OPERATOR @# (
-- List of functions

CREATE FUNCTION sphere(vector, real) RETURNS sphere_vector
IMMUTABLE PARALLEL SAFE LANGUAGE sql AS 'SELECT ROW($1, $2)';
IMMUTABLE PARALLEL SAFE LANGUAGE sql AS 'SELECT ROW($1, $2)::sphere_vector';

CREATE FUNCTION sphere(halfvec, real) RETURNS sphere_halfvec
IMMUTABLE PARALLEL SAFE LANGUAGE sql AS 'SELECT ROW($1, $2)';
IMMUTABLE PARALLEL SAFE LANGUAGE sql AS 'SELECT ROW($1, $2)::sphere_halfvec';

CREATE FUNCTION sphere(rabitq8, real) RETURNS sphere_rabitq8
IMMUTABLE PARALLEL SAFE LANGUAGE sql AS 'SELECT ROW($1, $2)';
IMMUTABLE PARALLEL SAFE LANGUAGE sql AS 'SELECT ROW($1, $2)::sphere_rabitq8';

CREATE FUNCTION sphere(rabitq4, real) RETURNS sphere_rabitq4
IMMUTABLE PARALLEL SAFE LANGUAGE sql AS 'SELECT ROW($1, $2)';
IMMUTABLE PARALLEL SAFE LANGUAGE sql AS 'SELECT ROW($1, $2)::sphere_rabitq4';

CREATE FUNCTION quantize_to_rabitq8(vector) RETURNS rabitq8
IMMUTABLE STRICT PARALLEL SAFE LANGUAGE c AS 'MODULE_PATHNAME', '_vchord_vector_quantize_to_rabitq8_wrapper';
Expand Down
Loading