[PATCH] ttusb-budget: make it depend on PCI

[PATCH] ttusb-budget: make it depend on PCI

[Mike Frysinger]

Since dvb-ttusb-budget.c relies on pci_alloc_consistent and
pci_free_consistent, make it depend on PCI in Kconfig.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
---
 drivers/media/dvb/ttusb-budget/Kconfig |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/media/dvb/ttusb-budget/Kconfig b/drivers/media/dvb/ttusb-budget/Kconfig
index f546bcc..2663ae3 100644
--- a/drivers/media/dvb/ttusb-budget/Kconfig
+++ b/drivers/media/dvb/ttusb-budget/Kconfig
@@ -1,6 +1,6 @@
 config DVB_TTUSB_BUDGET
 	tristate "Technotrend/Hauppauge Nova-USB devices"
-	depends on DVB_CORE && USB && I2C
+	depends on DVB_CORE && USB && I2C && PCI
 	select DVB_CX22700 if !DVB_FE_CUSTOMISE
 	select DVB_TDA1004X if !DVB_FE_CUSTOMISE
 	select DVB_VES1820 if !DVB_FE_CUSTOMISE
-- 
1.6.0.3

[PATCH] ttusb-dec: make it depend on PCI

[Mike Frysinger]

Since ttusb_dec.c relies on pci_alloc_consistent and
pci_free_consistent, make it depend on PCI in Kconfig.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
---
 drivers/media/dvb/ttusb-dec/Kconfig |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/media/dvb/ttusb-dec/Kconfig b/drivers/media/dvb/ttusb-dec/Kconfig
index d5f48a3..290254a 100644
--- a/drivers/media/dvb/ttusb-dec/Kconfig
+++ b/drivers/media/dvb/ttusb-dec/Kconfig
@@ -1,6 +1,6 @@
 config DVB_TTUSB_DEC
 	tristate "Technotrend/Hauppauge USB DEC devices"
-	depends on DVB_CORE && USB && INPUT
+	depends on DVB_CORE && USB && INPUT && PCI
 	select CRC32
 	help
 	  Support for external USB adapters designed by Technotrend and
-- 
1.6.0.3

Re: [PATCH] ttusb-budget: make it depend on PCI

[Arnd Bergmann]

On Wednesday 12 November 2008, Mike Frysinger wrote:
> From: 
> Mike Frysinger <vapier@gentoo.org>
>   To: 
> v4l-dvb-maintainer@linuxtv.org, linux-dvb@linuxtv.org, Holger Waechtler <holger@convergence.de>
>   CC: 
> Mauro Carvalho Chehab <mchehab@infradead.org>, linux-kernel@vger.kernel.org
>   Date: 
> Today 13:00:18
>    
> Since dvb-ttusb-budget.c relies on pci_alloc_consistent and
> pci_free_consistent, make it depend on PCI in Kconfig.
> 
> Signed-off-by: Mike Frysinger <vapier@gentoo.org>

Shouldn't you just convert it (trivially) to dma_alloc_coherent and
dma_alloc_coherent?

	Arnd <><

Re: [PATCH] ttusb-budget: make it depend on PCI

[Andi Kleen]

Arnd Bergmann <arnd@arndb.de> writes:

> On Wednesday 12 November 2008, Mike Frysinger wrote:
>> From: 
>> Mike Frysinger <vapier@gentoo.org>
>>   To: 
>> v4l-dvb-maintainer@linuxtv.org, linux-dvb@linuxtv.org, Holger Waechtler <holger@convergence.de>
>>   CC: 
>> Mauro Carvalho Chehab <mchehab@infradead.org>, linux-kernel@vger.kernel.org
>>   Date: 
>> Today 13:00:18
>>    
>> Since dvb-ttusb-budget.c relies on pci_alloc_consistent and
>> pci_free_consistent, make it depend on PCI in Kconfig.
>> 
>> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
>
> Shouldn't you just convert it (trivially) to dma_alloc_coherent and
> dma_alloc_coherent?

Unless it really needs to use the 16MB DMA zone (which I doubt
especially for USB devices) then it should not call either with a NULL
device argument.

-Andi

-- 
ak@linux.intel.com

Re: [PATCH] ttusb-budget: make it depend on PCI

[Mike Frysinger]

On Wed, Nov 12, 2008 at 07:35, Arnd Bergmann wrote:
> On Wednesday 12 November 2008, Mike Frysinger wrote:
>> Since dvb-ttusb-budget.c relies on pci_alloc_consistent and
>> pci_free_consistent, make it depend on PCI in Kconfig.
>>
>> Signed-off-by: Mike Frysinger <vapier@gentoo.org>
>
> Shouldn't you just convert it (trivially) to dma_alloc_coherent and
> dma_alloc_coherent?

i dont have any DVB hardware let alone this device, so there's no way
for me to test things

if the maintainer wishes to go that route, i imagine it'd make other
people happy
-mike

Re: [PATCH] ttusb-budget: make it depend on PCI

[Arnd Bergmann]

On Wednesday 12 November 2008, Andi Kleen wrote:
> Unless it really needs to use the 16MB DMA zone (which I doubt
> especially for USB devices) then it should not call either with a NULL
> device argument.

Yes, of course. It will need to pass &ttusb->dev.dev to dma_alloc_coherent.

	Arnd <><

---

Note to self: read code before commenting on it...