From: "Rafael J. Wysocki" <rjw@sisk.pl>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Frans Pop <elendil@planet.nl>, Greg KH <greg@kroah.com>,
Ingo Molnar <mingo@elte.hu>,
jbarnes@virtuousgeek.org, lenb@kernel.org,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
tiwai@suse.de, Andrew Morton <akpm@linux-foundation.org>
Subject: Re: Regression from 2.6.26: Hibernation (possibly suspend) broken on Toshiba R500 (bisected)
Date: Sat, 6 Dec 2008 01:04:49 +0100 [thread overview]
Message-ID: <200812060104.50050.rjw@sisk.pl> (raw)
In-Reply-To: <alpine.LFD.2.00.0812051411400.3386@nehalem.linux-foundation.org>
On Friday, 5 of December 2008, Linus Torvalds wrote:
>
> On Fri, 5 Dec 2008, Rafael J. Wysocki wrote:
> >
> > Well, actually I think we should go further and save the standard config
> > registers of _all_ PCI devices in the _late() callbacks (ie. with interrupts
> > disabled) and restore them in the _early() callbacks.
>
> That would certainly simplify the code.
>
> > I don't really understand why pci_restore_state() is not called by the core
> > and every single driver calls it by itself.
>
> The idea was to allow PCI drivers to override it if they wanted to.
>
> That said, the ones that do their own state restore generally do it wrong
> (eg the USB host controllers doing things in the wrong order and enabling
> the device before having actually written back the BAR values), so it's
> arguably broken to let drivers override it.
>
> > IOW, I would split the resume of PCI devices into two parts, the first of
> > which will call pci_restore_state() with interrupts disabled and the second
> > will do the remaining stuff.
>
> I would definitely not disagree with that - leave the regular
> "suspend/resume" callbacks for purely higher-level actions. It would
> interesting to hear what it does for you.
I tested the appended patch with suspend-to-RAM and it just hangs during
resume.
Next, I'll try to do that only for devices the drivers of which don't define their own
suspend-resume callbacks at all.
Thanks,
Rafael
---
drivers/pci/pci-driver.c | 50 ++++++++++++++++++++++++++++++++++++++++++++---
1 file changed, 47 insertions(+), 3 deletions(-)
Index: linux-2.6/drivers/pci/pci-driver.c
===================================================================
--- linux-2.6.orig/drivers/pci/pci-driver.c
+++ linux-2.6/drivers/pci/pci-driver.c
@@ -300,6 +300,46 @@ static void pci_device_shutdown(struct d
#ifdef CONFIG_PM_SLEEP
+static void pci_default_suspend_noirq(struct pci_dev *pci_dev)
+{
+ dev_info(&pci_dev->dev, "saving standard PCI config registers\n");
+
+ /* save the PCI config space */
+ pci_save_state(pci_dev);
+ /*
+ * mark its power state as "unknown", since we don't know if
+ * e.g. the BIOS will change its device state when we suspend.
+ */
+ if (pci_dev->current_state == PCI_D0)
+ pci_dev->current_state = PCI_UNKNOWN;
+}
+
+static void pci_default_resume_noirq(struct pci_dev *pci_dev)
+{
+ dev_info(&pci_dev->dev, "restoring standard PCI config registers\n");
+
+ /* restore the PCI config space */
+ pci_restore_state(pci_dev);
+}
+
+static int pci_default_resume(struct pci_dev *pci_dev)
+{
+ int retval = 0;
+
+ dev_info(&pci_dev->dev, "trying to reenable device\n");
+
+ /* if the device was enabled before suspend, reenable */
+ retval = pci_reenable_device(pci_dev);
+ /*
+ * if the device was busmaster before the suspend, make it busmaster
+ * again
+ */
+ if (pci_dev->is_busmaster)
+ pci_set_master(pci_dev);
+
+ return retval;
+}
+
/*
* Default "suspend" method for devices that have no driver provided suspend,
* or not even a driver at all.
@@ -346,8 +386,6 @@ static int pci_legacy_suspend(struct dev
if (drv && drv->suspend) {
i = drv->suspend(pci_dev, state);
suspend_report_result(drv->suspend, i);
- } else {
- pci_default_pm_suspend(pci_dev);
}
return i;
}
@@ -362,6 +400,9 @@ static int pci_legacy_suspend_late(struc
i = drv->suspend_late(pci_dev, state);
suspend_report_result(drv->suspend_late, i);
}
+
+ pci_default_suspend_noirq(pci_dev);
+
return i;
}
@@ -374,7 +415,7 @@ static int pci_legacy_resume(struct devi
if (drv && drv->resume)
error = drv->resume(pci_dev);
else
- error = pci_default_pm_resume(pci_dev);
+ error = pci_default_resume(pci_dev);
return error;
}
@@ -384,8 +425,11 @@ static int pci_legacy_resume_early(struc
struct pci_dev * pci_dev = to_pci_dev(dev);
struct pci_driver * drv = pci_dev->driver;
+ pci_default_resume_noirq(pci_dev);
+
if (drv && drv->resume_early)
error = drv->resume_early(pci_dev);
+
return error;
}
next prev parent reply other threads:[~2008-12-06 0:06 UTC|newest]
Thread overview: 137+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-12-02 2:20 Rafael J. Wysocki
2008-12-02 3:32 ` Linus Torvalds
2008-12-02 3:42 ` Linus Torvalds
2008-12-02 4:31 ` Frans Pop
2008-12-02 4:46 ` Linus Torvalds
2008-12-02 5:29 ` Frans Pop
2008-12-02 5:56 ` Frans Pop
2008-12-02 15:46 ` Linus Torvalds
2008-12-02 17:46 ` Frans Pop
2008-12-02 18:17 ` Linus Torvalds
2008-12-05 8:53 ` MSI changes in .28 Frans Pop
2008-12-05 9:09 ` Yinghai Lu
2008-12-05 12:20 ` Ingo Molnar
2008-12-05 13:04 ` Eric Dumazet
2008-12-05 17:49 ` H. Peter Anvin
2008-12-02 4:13 ` Regression from 2.6.26: Hibernation (possibly suspend) broken on Toshiba R500 (bisected) Frans Pop
2008-12-02 4:36 ` Linus Torvalds
2008-12-02 22:38 ` Rafael J. Wysocki
2008-12-02 23:37 ` Linus Torvalds
2008-12-03 0:00 ` Rafael J. Wysocki
2008-12-03 0:05 ` Rafael J. Wysocki
2008-12-03 0:31 ` Rafael J. Wysocki
2008-12-03 0:41 ` Linus Torvalds
2008-12-03 1:22 ` Rafael J. Wysocki
2008-12-03 2:02 ` Linus Torvalds
2008-12-03 7:40 ` Rafael J. Wysocki
2008-12-03 7:52 ` Rafael J. Wysocki
2008-12-03 11:20 ` Rafael J. Wysocki
2008-12-03 15:53 ` Linus Torvalds
2008-12-04 1:23 ` Rafael J. Wysocki
2008-12-04 4:40 ` Linus Torvalds
2008-12-04 8:21 ` Frans Pop
2008-12-04 22:01 ` Rafael J. Wysocki
2008-12-04 11:29 ` Frans Pop
2008-12-04 16:17 ` Linus Torvalds
2008-12-04 18:00 ` Frans Pop
2008-12-04 20:03 ` Linus Torvalds
2008-12-05 21:26 ` Linus Torvalds
2008-12-05 22:01 ` Rafael J. Wysocki
2008-12-05 22:14 ` Linus Torvalds
2008-12-06 0:04 ` Rafael J. Wysocki [this message]
2008-12-06 0:50 ` Linus Torvalds
2008-12-06 1:18 ` Rafael J. Wysocki
2008-12-06 1:55 ` Linus Torvalds
2008-12-06 2:18 ` Rafael J. Wysocki
2008-12-06 13:53 ` Rafael J. Wysocki
2008-12-06 2:45 ` Greg KH
2009-01-28 12:00 ` Frans Pop
2009-01-29 14:11 ` Ingo Molnar
2009-01-29 14:48 ` Rafael J. Wysocki
2009-01-29 16:44 ` Alexey Starikovskiy
2009-01-30 4:35 ` Frans Pop
2008-12-06 9:20 ` [patch,rfc] usb: restore config before enabling device on resume Frans Pop
2008-12-06 13:48 ` Rafael J. Wysocki
2008-12-06 15:02 ` Frans Pop
2008-12-10 14:06 ` "APIC error on CPU1: 00(40)" during resume (was: Regression from 2.6.26: Hibernation (possibly suspend) broken on Toshiba R500) Frans Pop
2008-12-10 15:51 ` Linus Torvalds
2008-12-10 16:05 ` Frans Pop
2008-12-10 16:26 ` Linus Torvalds
2008-12-10 16:52 ` Matthew Garrett
2008-12-10 17:13 ` Linus Torvalds
2008-12-10 17:33 ` Ingo Molnar
2008-12-10 18:41 ` Maxim Levitsky
2008-12-20 21:31 ` "APIC error on CPU1: 00(40)" during resume Frans Pop
2008-12-21 8:29 ` Ingo Molnar
2008-12-23 4:28 ` Len Brown
2008-12-04 22:46 ` Regression from 2.6.26: Hibernation (possibly suspend) broken on Toshiba R500 (bisected) Rafael J. Wysocki
2008-12-04 22:40 ` Rafael J. Wysocki
2008-12-04 23:22 ` Linus Torvalds
2008-12-04 23:45 ` Rafael J. Wysocki
2008-12-05 0:07 ` Linus Torvalds
2008-12-05 0:20 ` Rafael J. Wysocki
2008-12-05 6:55 ` Frans Pop
2008-12-04 22:09 ` Rafael J. Wysocki
2008-12-04 22:20 ` Linus Torvalds
2008-12-04 23:31 ` Rafael J. Wysocki
2008-12-05 0:03 ` Linus Torvalds
2008-12-05 0:45 ` Linus Torvalds
2008-12-05 1:08 ` Rafael J. Wysocki
2008-12-05 1:45 ` Linus Torvalds
2008-12-05 2:55 ` Linus Torvalds
2008-12-05 3:25 ` Linus Torvalds
2008-12-05 6:44 ` Frans Pop
2008-12-05 8:27 ` Frans Pop
2008-12-05 12:00 ` Rafael J. Wysocki
2008-12-05 15:57 ` Linus Torvalds
2008-12-05 21:32 ` Rafael J. Wysocki
2008-12-05 17:25 ` Jesse Barnes
2008-12-02 15:49 ` Rafael J. Wysocki
2008-12-06 14:05 ` [PATCH 0/3] Fix hibernation regression on Toshiba Portege R500 Rafael J. Wysocki
2008-12-06 14:07 ` [PATCH 1/3] PCI: Rework default handling of suspend and resume Rafael J. Wysocki
2008-12-06 17:07 ` Linus Torvalds
2008-12-06 17:22 ` Rafael J. Wysocki
2008-12-06 17:33 ` Linus Torvalds
2008-12-06 17:43 ` Rafael J. Wysocki
2008-12-06 18:00 ` Linus Torvalds
2008-12-06 21:24 ` Rafael J. Wysocki
2008-12-07 4:44 ` Jesse Barnes
2008-12-07 5:41 ` Greg KH
2008-12-07 12:47 ` Rafael J. Wysocki
2008-12-07 16:44 ` Linus Torvalds
2008-12-07 21:02 ` Rafael J. Wysocki
2008-12-07 17:26 ` Greg KH
2008-12-07 23:34 ` [PATCH 1/3] PCI: Rework default handling of suspend and resume (rebased) Rafael J. Wysocki
2008-12-06 18:30 ` [linux-pm] [PATCH 1/3] PCI: Rework default handling of suspend and resume Alan Stern
2008-12-06 21:36 ` Rafael J. Wysocki
2008-12-06 22:24 ` Linus Torvalds
2008-12-06 23:25 ` Arjan van de Ven
2008-12-06 23:35 ` Alan Cox
2008-12-07 6:00 ` Linus Torvalds
2008-12-07 6:03 ` Linus Torvalds
2008-12-07 13:39 ` Rafael J. Wysocki
2008-12-07 16:34 ` Linus Torvalds
2008-12-14 9:28 ` Pavel Machek
2008-12-07 17:18 ` Arjan van de Ven
2008-12-07 9:44 ` Takashi Iwai
2008-12-07 0:02 ` Alan Stern
2008-12-07 13:14 ` Rafael J. Wysocki
2008-12-06 21:09 ` Alan Cox
2008-12-06 21:50 ` Rafael J. Wysocki
2008-12-06 14:07 ` [PATCH 2/3] PCI: Suspend and resume PCI Express ports with interrupts disabled Rafael J. Wysocki
2008-12-06 17:15 ` Linus Torvalds
2008-12-06 17:25 ` Rafael J. Wysocki
2008-12-06 17:38 ` Linus Torvalds
2008-12-06 17:46 ` Rafael J. Wysocki
2008-12-07 2:18 ` Jesse Barnes
2008-12-07 12:53 ` Rafael J. Wysocki
2008-12-06 14:09 ` [PATCH 3/3] Sound (HDA Intel): Restore PCI configuration space with interrupts off Rafael J. Wysocki
2008-12-07 4:45 ` Jesse Barnes
2008-12-07 9:47 ` Takashi Iwai
2008-12-11 7:07 ` Takashi Iwai
2008-12-11 20:03 ` Rafael J. Wysocki
2008-12-11 20:27 ` Takashi Iwai
2008-12-11 20:38 ` Rafael J. Wysocki
2008-12-12 6:32 ` Takashi Iwai
2008-12-06 19:30 ` [PATCH 0/3] Fix hibernation regression on Toshiba Portege R500 Frans Pop
2008-12-02 7:53 Regression from 2.6.26: Hibernation (possibly suspend) broken on Toshiba R500 (bisected) Frans Pop
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=200812060104.50050.rjw@sisk.pl \
--to=rjw@sisk.pl \
--cc=akpm@linux-foundation.org \
--cc=elendil@planet.nl \
--cc=greg@kroah.com \
--cc=jbarnes@virtuousgeek.org \
--cc=lenb@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=tiwai@suse.de \
--cc=torvalds@linux-foundation.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Powered by JetHome