Hi all, While hacking a different patch, I've noticed that a couple of %llu did not get the PRIu64 call in the AIO code, and I don't see why we could not switch them. These have been introduced in commits that got into the tree after Peter's 15a79c73111f.
A couple remain even after the attached, which have been left off by Peter for the same reasons as the ones I am guessing here: - launch_backend.c for paramHandle, does not seem worth it. - ecpglib/execute.c, when storing some input, which does not seem worth bothering either. - reconstruct.c, offset handling. - pg_backup_tar.c, ftello() result and some consistency with the surroundings. - pg_verifybackup.c and astreamer_verify.c, for Sizes. That's not necessarily mandatory for v18, for sure, but as this is new code we could as well clean it up before forking the next stable branch. Comments or opinions? -- Michael
From 0c9ccd79203f4709638658cf918d5ebc3d584801 Mon Sep 17 00:00:00 2001
From: Michael Paquier <mich...@paquier.xyz>
Date: Mon, 9 Jun 2025 11:41:57 +0900
Subject: [PATCH] Replace %llu by PRIu64 in AIO io_uring code
---
src/backend/storage/aio/method_io_uring.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/src/backend/storage/aio/method_io_uring.c b/src/backend/storage/aio/method_io_uring.c
index cc312b641ca6..b78048328e11 100644
--- a/src/backend/storage/aio/method_io_uring.c
+++ b/src/backend/storage/aio/method_io_uring.c
@@ -400,9 +400,9 @@ pgaio_uring_wait_one(PgAioHandle *ioh, uint64 ref_generation)
while (true)
{
pgaio_debug_io(DEBUG3, ioh,
- "wait_one io_gen: %llu, ref_gen: %llu, cycle %d",
- (long long unsigned) ioh->generation,
- (long long unsigned) ref_generation,
+ "wait_one io_gen: %" PRIu64 ", ref_gen: %" PRIu64 ", cycle %d",
+ ioh->generation,
+ ref_generation,
waited);
if (pgaio_io_was_recycled(ioh, ref_generation, &state) ||
--
2.49.0
signature.asc
Description: PGP signature
