linux-kernel - Re: [Bug #13306] hibernate slow on _second

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 for Android: free password hash cracker in your pocket

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

Message-Id: <1244643882.18481.62.camel@johannes.local>
Date:	Wed, 10 Jun 2009 16:24:42 +0200
From:	Johannes Berg <johannes@...solutions.net>
To:	"Rafael J. Wysocki" <rjw@...k.pl>
Cc:	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
	Kernel Testers List <kernel-testers@...r.kernel.org>,
	Steven Rostedt <srostedt@...hat.com>
Subject: Re: [Bug #13306] hibernate slow on _second_ run

On Mon, 2009-06-08 at 22:54 +0200, Rafael J. Wysocki wrote:
> On Thursday 04 June 2009, Johannes Berg wrote:
> > Alright, I have managed to trace this.
> > 
> > http://johannes.sipsolutions.net/files/hibernate-delay-dmesg.txt
> > 
> > has the dmesg, you can clearly see the delay:
> > [50265.143330] ACPI: Waking up from system sleep state S4
> > [50265.679141] HDA Intel 0000:00:08.0: restoring config space at offset 0x1 (was 0xb00006, writing 0xb00002)
> > [50265.689525] forcedeth 0000:00:0a.0: restoring config space at offset 0x1 (was 0xb00003, writing 0xb00007)
> > [50265.699775] ahci 0000:00:0b.0: restoring config space at offset 0x1 (was 0xb00403, writing 0xb00407)
> > [50286.294910] pci 0000:00:03.2: PME# disabled
> 
> Quite frankly, it looks like resume_device_irqs() itself causes the delay to happen.
> 
> The next step would be to put a couple of printk()s in there and see which part
> of it triggers the delay.

--- a/kernel/irq/pm.c
+++ b/kernel/irq/pm.c
@@ -56,8 +56,11 @@ void resume_device_irqs(void)
                if (!(desc->status & IRQ_SUSPENDED))
                        continue;
 
+               printk(KERN_DEBUG "lock irq %d\n", irq);
                spin_lock_irqsave(&desc->lock, flags);
+               printk(KERN_DEBUG "locked\n");
                __enable_irq(desc, irq, true);
+               printk(KERN_DEBUG "enabled\n");
                spin_unlock_irqrestore(&desc->lock, flags);
        }
 }


and I get:
[63702.285944] ACPI: Waking up from system sleep state S4
[63702.811457] HDA Intel 0000:00:08.0: restoring config space at offset 0x1 (was 0xb00006, writing 0xb00002)
[63702.819369] pci 0000:00:09.0: restoring config space at offset 0x7 (was 0x228000f0, writing 0x428000f0)
[63702.827574] forcedeth 0000:00:0a.0: restoring config space at offset 0x1 (was 0xb00003, writing 0xb00007)
[63702.836052] ahci 0000:00:0b.0: restoring config space at offset 0x1 (was 0xb00403, writing 0xb00407)
[63702.845288] lock irq 9
[63702.853902] locked
[63702.862433] enabled
[63702.870967] lock irq 17
[63702.879377] locked
[63702.880964] enabled
[63702.895811] lock irq 18
[63702.903919] locked
[63702.911879] enabled
[63702.919781] lock irq 19
[63702.927595] locked
[63702.935260] enabled
[63702.942989] lock irq 20
[63702.950764] locked
[63702.958425] enabled
[63702.966061] lock irq 25
[63702.973641] locked
[63702.981102] enabled
[63722.620744] pci 0000:00:03.2: PME# disabled
[63722.628069] ohci_hcd 0000:00:04.0: PME# disabled

so it doesn't look like the delay is there.

johannes

Download attachment "signature.asc" of type "application/pgp-signature" (802 bytes)