[PATCH] sdhci-pci register legacy PCI PM or runtime PM not both

[PATCH] sdhci-pci register legacy PCI PM or runtime PM not both

[David Fries]

From: David Fries <David@Fries.net>

pci_has_legacy_pm_support says to not register both legacy PM
and runtime PM callbacks, modified ifdefs.

Signed-off-by: David Fries <David@Fries.net>
Cc: linux-mmc@vger.kernel.org
Cc: Chris Ball <cjb@laptop.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Yunpeng Gao <yunpeng.gao@intel.com>
---
 drivers/mmc/host/sdhci-pci.c |    7 ++++---
 1 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/mmc/host/sdhci-pci.c b/drivers/mmc/host/sdhci-pci.c
index d833d9c..eca6e84 100644
--- a/drivers/mmc/host/sdhci-pci.c
+++ b/drivers/mmc/host/sdhci-pci.c
@@ -991,7 +991,8 @@ static struct sdhci_ops sdhci_pci_ops = {
  *                                                                           *
 \*****************************************************************************/
 
-#ifdef CONFIG_PM
+/* Register legacy PM or PM runtime, not both. */
+#if defined(CONFIG_PM) && !defined(CONFIG_PM_RUNTIME)
 
 static int sdhci_pci_suspend(struct pci_dev *pdev, pm_message_t state)
 {
@@ -1085,12 +1086,12 @@ static int sdhci_pci_resume(struct pci_dev *pdev)
 	return 0;
 }
 
-#else /* CONFIG_PM */
+#else /* CONFIG_PM && !CONFIG_PM_RUNTIME */
 
 #define sdhci_pci_suspend NULL
 #define sdhci_pci_resume NULL
 
-#endif /* CONFIG_PM */
+#endif /* CONFIG_PM && !CONFIG_PM_RUNTIME */
 
 #ifdef CONFIG_PM_RUNTIME
 
-- 
1.7.7.1

Re: [PATCH] sdhci-pci register legacy PCI PM or runtime PM not both

[Adrian Hunter]

On 20/11/11 23:29, David Fries wrote:
> From: David Fries <David@Fries.net>
> 
> pci_has_legacy_pm_support says to not register both legacy PM
> and runtime PM callbacks, modified ifdefs.

Hasn't this already been done in Chris mmc tree i.e.

commit f353f593fe449178a8cb01cefe559dba125ab8b1
Author: Manuel Lauss <manuel.lauss@googlemail.com>
Date:   Thu Nov 3 11:09:45 2011 +0100

    mmc: sdhci: remove "state" argument from sdhci_suspend_host

    Drop the "state" argument from sdhci_suspend_host.  Its only user is the
    PCI glue;  this allows to move all SDHCI glues to use dev_pm_ops instead.

    Signed-off-by: Manuel Lauss <manuel.lauss@googlemail.com>
    Acked-by: Adrian Hunter <adrian.hunter@intel.com>
    Signed-off-by: Chris Ball <cjb@laptop.org>

 drivers/mmc/host/sdhci-cns3xxx.c   |    5 +----
 drivers/mmc/host/sdhci-dove.c      |    5 +----
 drivers/mmc/host/sdhci-esdhc-imx.c |    5 +----
 drivers/mmc/host/sdhci-of-esdhc.c  |    5 +----
 drivers/mmc/host/sdhci-of-hlwd.c   |    5 +----
 drivers/mmc/host/sdhci-pci.c       |   26 +++++++++++++-------------
 drivers/mmc/host/sdhci-pltfm.c     |   18 +++++++++++-------
 drivers/mmc/host/sdhci-pltfm.h     |    6 ++++--
 drivers/mmc/host/sdhci-pxav2.c     |    5 +----
 drivers/mmc/host/sdhci-pxav3.c     |    5 +----
 drivers/mmc/host/sdhci-s3c.c       |   21 ++++++++++++++-------
 drivers/mmc/host/sdhci-tegra.c     |    5 +----
 drivers/mmc/host/sdhci.c           |    2 +-
 drivers/mmc/host/sdhci.h           |    2 +-
 14 files changed, 52 insertions(+), 63 deletions(-)



> 
> Signed-off-by: David Fries <David@Fries.net>
> Cc: linux-mmc@vger.kernel.org
> Cc: Chris Ball <cjb@laptop.org>
> Cc: Adrian Hunter <adrian.hunter@intel.com>
> Cc: Yunpeng Gao <yunpeng.gao@intel.com>
> ---
>  drivers/mmc/host/sdhci-pci.c |    7 ++++---
>  1 files changed, 4 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/mmc/host/sdhci-pci.c b/drivers/mmc/host/sdhci-pci.c
> index d833d9c..eca6e84 100644
> --- a/drivers/mmc/host/sdhci-pci.c
> +++ b/drivers/mmc/host/sdhci-pci.c
> @@ -991,7 +991,8 @@ static struct sdhci_ops sdhci_pci_ops = {
>   *                                                                           *
>  \*****************************************************************************/
>  
> -#ifdef CONFIG_PM
> +/* Register legacy PM or PM runtime, not both. */
> +#if defined(CONFIG_PM) && !defined(CONFIG_PM_RUNTIME)
>  
>  static int sdhci_pci_suspend(struct pci_dev *pdev, pm_message_t state)
>  {
> @@ -1085,12 +1086,12 @@ static int sdhci_pci_resume(struct pci_dev *pdev)
>  	return 0;
>  }
>  
> -#else /* CONFIG_PM */
> +#else /* CONFIG_PM && !CONFIG_PM_RUNTIME */
>  
>  #define sdhci_pci_suspend NULL
>  #define sdhci_pci_resume NULL
>  
> -#endif /* CONFIG_PM */
> +#endif /* CONFIG_PM && !CONFIG_PM_RUNTIME */
>  
>  #ifdef CONFIG_PM_RUNTIME
>  

Re: [PATCH] sdhci-pci register legacy PCI PM or runtime PM not both

[David Fries]

On Mon, Nov 21, 2011 at 10:48:20AM +0200, Adrian Hunter wrote:
> On 20/11/11 23:29, David Fries wrote:
> > From: David Fries <David@Fries.net>
> > 
> > pci_has_legacy_pm_support says to not register both legacy PM
> > and runtime PM callbacks, modified ifdefs.
> 
> Hasn't this already been done in Chris mmc tree i.e.
> 
> commit f353f593fe449178a8cb01cefe559dba125ab8b1
> Author: Manuel Lauss <manuel.lauss@googlemail.com>
> Date:   Thu Nov 3 11:09:45 2011 +0100
> 
>     mmc: sdhci: remove "state" argument from sdhci_suspend_host

Good call.  It looks like a better solution than what I posted.  I
tried it and it also worked.  I would classify this as a regression in
mainline (suspend didn't used to generate warnings) until a fix is
merged.  I'm good for now, but I doubt I'll be the only one running
into it.

-- 
David Fries <david@fries.net>    PGP pub CB1EE8F0
http://fries.net/~david/