[Bug 265087] Changes to pipe/socket handling causes configure test program to loop endless
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 265087] Changes to pipe/socket handling causes configure test program to loop endless"
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 265087] Changes to pipe/socket handling causes configure test program to loop endless"
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 265087] Changes to pipe/socket handling causes configure test program to loop endless"
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 265087] Changes to pipe/socket handling causes configure test program to loop endless"
- Reply: bugzilla-noreply_a_freebsd.org: "[Bug 265087] Changes to pipe/socket handling causes configure test program to loop endless"
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Fri, 08 Jul 2022 09:19:33 UTC
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=265087
Bug ID: 265087
Summary: Changes to pipe/socket handling causes configure test
program to loop endless
Product: Base System
Version: CURRENT
Hardware: amd64
OS: Any
Status: New
Severity: Affects Only Me
Priority: ---
Component: kern
Assignee: bugs@FreeBSD.org
Reporter: rz-rpi03@h-ka.de
Attachment #235121 text/plain
mime type:
Created attachment 235121
--> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=235121&action=edit
Extraced "checking read/send-side pipe system..." C program
Hello,
recent changes in CURRENT causes the Dante socks client to fail.
Usually this could be fixed simply by recompiling and reinstalling net/dante.
This time it failed because the configure script stops at
"checking read/send-side pipe system..."
Investigation shows that the C program executed loops forever.
Trying to bisect CURRENT stops for me at
716fd348e01c5f2ba125f878a634a753436c2994
because the boot zpool was upgraded meanwhile and an older CURRENT version
can not boot from such a pool.
716fd348e01c5f2ba125f878a634a753436c2994 does expose the new (loop) behavior
as well, so the cause for the endless loop must be earlier.
Adding some additional fprintf's showed that the C program does not exit
the wile(1)-loop in sendtest() at the first call of sendtest().
Ralf
--
You are receiving this mail because:
You are the assignee for the bug.