tpm/tpm_tis: Add missing ifdef CONFIG_ACPI for pnp_acpi_device
diff mbox

Message ID 1421872815-16644-1-git-send-email-peterhuewe@gmx.de
State New, archived
Headers show

Commit Message

Peter Huewe Jan. 21, 2015, 8:40 p.m. UTC
This fixes a build failure if CONFIG_PNP is set but CONFIG_ACPI is not:
drivers/char/tpm/tpm_tis.c: In function ‘tpm_tis_pnp_init’:
drivers/char/tpm/tpm_tis.c:912:45: error: invalid type argument of
‘->’ (have ‘int’)
   acpi_dev_handle = pnp_acpi_device(pnp_dev)->handle;

If CONFIG_PNPACPI is not set pnp_acpi_device is defined as 0 and thus
accesing the handle is not possible.

Fixes: 0dc553652102 ("tpm: fix raciness of PPI interface lookup")
Reported-by: Jim Davis <jim.epost@gmail.com>
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
---
 drivers/char/tpm/tpm_tis.c | 2 ++
 1 file changed, 2 insertions(+)

Comments

Stephen Rothwell Jan. 21, 2015, 9:23 p.m. UTC | #1
Hi Peter,

On Wed, 21 Jan 2015 21:40:15 +0100 Peter Huewe <peterhuewe@gmx.de> wrote:
>
> This fixes a build failure if CONFIG_PNP is set but CONFIG_ACPI is not:
> drivers/char/tpm/tpm_tis.c: In function ‘tpm_tis_pnp_init’:
> drivers/char/tpm/tpm_tis.c:912:45: error: invalid type argument of
> ‘->’ (have ‘int’)
>    acpi_dev_handle = pnp_acpi_device(pnp_dev)->handle;
> 
> If CONFIG_PNPACPI is not set pnp_acpi_device is defined as 0 and thus
> accesing the handle is not possible.
> 
> Fixes: 0dc553652102 ("tpm: fix raciness of PPI interface lookup")
> Reported-by: Jim Davis <jim.epost@gmail.com>
> Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
> ---
>  drivers/char/tpm/tpm_tis.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/drivers/char/tpm/tpm_tis.c b/drivers/char/tpm/tpm_tis.c
> index 20a61bc98db8..6725bef7cb96 100644
> --- a/drivers/char/tpm/tpm_tis.c
> +++ b/drivers/char/tpm/tpm_tis.c
> @@ -908,8 +908,10 @@ static int tpm_tis_pnp_init(struct pnp_dev *pnp_dev,
>  	if (is_itpm(pnp_dev))
>  		itpm = true;
>  
> +#ifdef CONFIG_ACPI
>  	if (pnp_acpi_device(pnp_dev))
>  		acpi_dev_handle = pnp_acpi_device(pnp_dev)->handle;
> +#endif
>  
>  	return tpm_tis_init(&pnp_dev->dev, acpi_dev_handle, start, len, irq);
>  }

To save on the ifdef, why not

	struct acpi_device *acpi;

	.
	.

	acpi = pnp_acpi_device(dev);
	if (acpi)
		acpi_dev_handle = acpi->handle;

As an aside, the dummy pnp_acpi_device() should be returning NULL not
0, and could be a static inline function (as could several other things
in that header.
Stephen Rothwell Jan. 21, 2015, 9:34 p.m. UTC | #2
Hi Peter,

On Thu, 22 Jan 2015 08:23:50 +1100 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> To save on the ifdef, why not
> 
> 	struct acpi_device *acpi;
> 
> 	.
> 	.
> 
> 	acpi = pnp_acpi_device(dev);
> 	if (acpi)
> 		acpi_dev_handle = acpi->handle;

I guess struct acpi_device may not be visible when CONFIG_ACPI is not
set? That would be a pain :-(
Peter Huewe Jan. 21, 2015, 10:40 p.m. UTC | #3
Am Mittwoch, 21. Januar 2015, 22:34:32 schrieb Stephen Rothwell:
> Hi Peter,
> 
> On Thu, 22 Jan 2015 08:23:50 +1100 Stephen Rothwell <sfr@canb.auug.org.au> 
wrote:
> > To save on the ifdef, why not
> > 
> > 	struct acpi_device *acpi;
> > 	
> > 	.
> > 	.
> > 	
> > 	acpi = pnp_acpi_device(dev);
> > 	if (acpi)
> > 	
> > 		acpi_dev_handle = acpi->handle;
> 
> I guess struct acpi_device may not be visible when CONFIG_ACPI is not
> set? That would be a pain :-(

Yeah seems like it :(
I also thought about this way but then decided against it for the obvious 
reason.

Stephen, do you pull this patch in directly or shall it go via my tpmdd-tree -
> james -> -next ?

> As an aside, the dummy pnp_acpi_device() should be returning NULL not
> 0, and could be a static inline function (as could several other things
> in that header.
Yeah, probably.
(although this function is only in exactly two places within the kernel... so 
not much is gained here:)
Maybe I'll post a patch or maybe we can get rid of this whole thing 
altogether.
Let me think about it.



Peter

------------------------------------------------------------------------------
New Year. New Location. New Benefits. New Data Center in Ashburn, VA.
GigeNET is offering a free month of service with a new server in Ashburn.
Choose from 2 high performing configs, both with 100TB of bandwidth.
Higher redundancy.Lower latency.Increased capacity.Completely compliant.
http://p.sf.net/sfu/gigenet
Stephen Rothwell Jan. 21, 2015, 11:05 p.m. UTC | #4
Hi Peter,

On Wed, 21 Jan 2015 23:40:04 +0100 Peter Hüwe <PeterHuewe@gmx.de> wrote:
>
> Stephen, do you pull this patch in directly or shall it go via my tpmdd-tree -
> > james -> -next ?

Normally these go via the tree maintainer (unless they affect me
directly - which this doesn't).

Patch
diff mbox

diff --git a/drivers/char/tpm/tpm_tis.c b/drivers/char/tpm/tpm_tis.c
index 20a61bc98db8..6725bef7cb96 100644
--- a/drivers/char/tpm/tpm_tis.c
+++ b/drivers/char/tpm/tpm_tis.c
@@ -908,8 +908,10 @@  static int tpm_tis_pnp_init(struct pnp_dev *pnp_dev,
 	if (is_itpm(pnp_dev))
 		itpm = true;
 
+#ifdef CONFIG_ACPI
 	if (pnp_acpi_device(pnp_dev))
 		acpi_dev_handle = pnp_acpi_device(pnp_dev)->handle;
+#endif
 
 	return tpm_tis_init(&pnp_dev->dev, acpi_dev_handle, start, len, irq);
 }