Hi folks!
When I was trying to fix a halting bug in GNU Artanis' new runner module
based on futures for async requests. I found it doesn't work for my cloud
server which has only one CPU.
The reason is because the%worker-count will be zero when there's only one
cpu. So it will do nothing for the existing futures in the queue. The
proper fix should be always keep %worker-count when there's just one cpu.
It was tested successfully on my server machine.
Attached the patch.
Thanks!
From e675b77a1c7de7ca03566a1f71ac75ea3ed24cb1 Mon Sep 17 00:00:00 2001
From: Nala Ginrut <mu...@gnu.org>
Date: Thu, 14 Aug 2025 21:37:42 +0900
Subject: [PATCH] Fix futures, %worker-count shouldn't be zero to prevent
halting
---
module/ice-9/futures.scm | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/module/ice-9/futures.scm b/module/ice-9/futures.scm
index 4a462839a..a830858b0 100644
--- a/module/ice-9/futures.scm
+++ b/module/ice-9/futures.scm
@@ -264,7 +264,9 @@ adding it to the waiter queue."
(define %worker-count
(if (provided? 'threads)
- (- (current-processor-count) 1)
+ (if (= 1 (current-processor-count))
+ 1
+ (- (current-processor-count) 1))
0))
;; A dock of workers that stay here forever.
--
2.43.0
