Re: Interrupts vs signals

Heikki Linnakangas <hlinnaka@iki.fi>

From: Heikki Linnakangas <hlinnaka@iki.fi>

To: Michael Paquier <michael@paquier.xyz>, Robert Haas <robertmhaas@gmail.com>

Cc: Thomas Munro <thomas.munro@gmail.com>, Fujii Masao <masao.fujii@gmail.com>, pgsql-hackers <pgsql-hackers@postgresql.org>, Andres Freund <andres@anarazel.de>

Date: 2024-11-15T12:39:15Z

Lists: pgsql-hackers

Attachments

v4-0001-Replace-Latches-with-Interrupts.patch (text/x-patch) patch v4-0001
v4-0002-Fix-lost-wakeup-issue-in-logical-replication-laun.patch (text/x-patch) patch v4-0002

On 04/11/2024 21:41, Heikki Linnakangas wrote:
> I pushed the preliminary cleanup patches from this patch set earlier, 
> only the main patches remain. Attached is a new version of those, with 
> mostly comment cleanups.

Another rebased version attached, no other changes. I think this is 
pretty much ready to be committed, but I'd love to get some review.

After this, some work remains to publish the compatibility header on the 
wiki, along with some instructions on how to upgrade your extension code.

-- 
Heikki Linnakangas
Neon (https://neon.tech)

Commits

Ignore SIGINT in walwriter and walsummarizer
- a92b809f9da1 19 (unreleased) landed
Split WaitEventSet functions to separate source file
- 393e0d231405 18.0 landed
Use ModifyWaitEvent to update exit_on_postmaster_death
- 84e5b2f07a5e 18.0 landed
Remove unused ShutdownLatchSupport() function
- a98e4dee63ce 18.0 landed
Rename two functions that wake up other processes
- 368d8270c838 18.0 landed
Use ProcNumbers instead of direct Latch pointers to address other procs
- a9c546a5a378 18.0 landed
Clean up WaitLatch calls that passed latch without WL_LATCH_SET
- f9ecb57a506a 18.0 landed
Remove unneeded #include
- 094ae071605d 18.0 landed
Remove unused latch
- 6c0c49f7d37d 18.0 landed
Remove support for background workers without BGWORKER_SHMEM_ACCESS.
- 80a8f95b3bca 15.0 cited