[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <4AC6638E.3020803@sbcglobal.net>
Date: Fri, 02 Oct 2009 15:33:18 -0500
From: Jayson King <jaysonk@...global.net>
To: kernel@...ivas.org, LKML <linux-kernel@...r.kernel.org>
Subject: backporting BFS to 2.6.27
Hello, Con & LKML
I have started a backport of sched-bfs 300 to kernel 2.6.27.35. I
originally thought of two different ways to do it. Either pull in
mainline patches from .28 ~ .31 until bfs applies cleanly, or change bfs
enough to apply to .27 without needing many mainline patches. The latter
approach seemed to be cleaner. I started by comparing the .31 sched.c to
.27.35 sched.c and applied those backward changes to sched_bfs.c.
I did also apply 3 mainline patches to .27.35, because I couldn't decide
what to do about reverting the account_* functions which bfs changes.
These patches are 457533a7, 79741dd3, and f5f293a4.
The kernel compiles, but it panics:
Kernel alive
Kernel really alive
PANIC: early exception 0e rip 10:ffffffff810ad10b error 0 cr2 18.
There's no stack trace but that address is in __kmalloc.
Can you or others give a hint of what might be going wrong or what I can do?
I've attached the .27'ized patches mentioned above as well as the .27
sched-bfs patch. These patches apply to 2.6.27.35 in order:
linux-2.6.27-457533a7.patch
linux-2.6.27-79741dd3.patch
linux-2.6.27-f5f293a4.patch
2.6.27-sched-bfs-300.patch
Attached also is the .config file.
Thanks.
(if replying from LKML please CC replies to me)
Jayson
View attachment "linux-2.6.27-457533a7.patch" of type "text/plain" (11639 bytes)
View attachment "linux-2.6.27-79741dd3.patch" of type "text/plain" (14623 bytes)
View attachment "linux-2.6.27-f5f293a4.patch" of type "text/plain" (2009 bytes)
View attachment "2.6.27-sched-bfs-300.patch" of type "text/plain" (181995 bytes)
View attachment ".config" of type "text/plain" (84233 bytes)
Powered by blists - more mailing lists