linux-kernel - 2.6.26.6-rt11 boot failure on PREEMPT

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-ID: <48F932B1.6020405@ct.jp.nec.com>
Date:	Fri, 17 Oct 2008 17:49:53 -0700
From:	Hiroshi Shimamoto <h-shimamoto@...jp.nec.com>
To:	Peter Zijlstra <a.p.zijlstra@...llo.nl>,
	Steven Rostedt <rostedt@...dmis.org>,
	Ingo Molnar <mingo@...e.hu>,
	Thomas Gleixner <tglx@...utronix.de>
Cc:	linux-kernel@...r.kernel.org, linux-rt-users@...r.kernel.org
Subject: 2.6.26.6-rt11 boot failure on PREEMPT_DESKTOP

Hi,

I got an issue, 2.6.26.6-rt11 PREEMPT_DESKTOP doesn't boot on my box.
Attached my .config.

The boot stopped the following point.
[    1.285988] Real Time Clock Driver v1.12ac
[    1.290816] intel_rng: FWH not detected
[    1.295992] Linux agpgart interface v0.103
[    1.300385] Serial: 8250/16550 driver $Revision: 1.90 $ 4 ports, IRQ sharing disabled
[    1.308734] serial8250: ttyS0 at I/O 0x3f8 (irq = 4) is a 16550A
[    1.316818] serial8250: ttyS1 at I/O 0x2f8 (irq = 3) is a 16550A
[    1.325397] 00:0a: ttyS1 at I/O 0x2f8 (irq = 3) is a 16550A
[    1.329632] 00:0b: ttyS0 at I/O 0x3f8 (irq = 4) is a 16550A
[    1.336132] Floppy drive(s): fd0 is 1.44M
[    4.356923] floppy0: no floppy controllers found

I've bisected, the patch
sched-properly-account-irq-and-rt-load.patch
causes this issue.

I was looking for what's wrong, but couldn't find the root cause.
But tried to change load.weight to 1 instead of 0 for RT process,
the kernel booted fine.
This zero load weight causes rq load weight zero, I guess it might
make some badness. But I'm not sure about this.
I don't think this is correct fix.

Could you please check it?

OK log with below patch.
[    1.304459] Real Time Clock Driver v1.12ac
[    1.309223] intel_rng: FWH not detected
[    1.313788] Linux agpgart interface v0.103
[    1.322107] Serial: 8250/16550 driver $Revision: 1.90 $ 4 ports, IRQ sharing disabled
[    1.330237] serial8250: ttyS0 at I/O 0x3f8 (irq = 4) is a 16550A
[    1.334635] serial8250: ttyS1 at I/O 0x2f8 (irq = 3) is a 16550A
[    1.342883] 00:0a: ttyS1 at I/O 0x2f8 (irq = 3) is a 16550A
[    1.348902] 00:0b: ttyS0 at I/O 0x3f8 (irq = 4) is a 16550A
[    1.355337] Floppy drive(s): fd0 is 1.44M
[    1.377551] FDC 0 is a National Semiconductor PC87306
[    1.386192] brd: module loaded
[    1.390410] loop: module loaded

---
 kernel/sched.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/kernel/sched.c b/kernel/sched.c
index 0d68e79..8ac0f25 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -1737,8 +1737,8 @@ static void set_load_weight(struct task_struct *p)
 		 * Real-time tasks do not contribute to SCHED_OTHER load
 		 * this is compensated by sched_scale_load() usage.
 		 */
-		p->se.load.weight = 0;
-		p->se.load.inv_weight = 0;
+		p->se.load.weight = 1;
+		p->se.load.inv_weight = 1;
 		return;
 	}
 
-- 
1.5.6


View attachment "rt11-desktop.config" of type "text/plain" (31553 bytes)