[PATCH 1/1] Input: tegra-kbc - remove pre Tegra20 definitions

Add support for Tegra30 by correcting definitions. This is necessary to
make driver useful in Tegra30.

Signed-off-by: Rakesh Iyer <redacted>
---
 arch/arm/mach-tegra/include/mach/kbc.h |    5 -----
 1 files changed, 0 insertions(+), 5 deletions(-)

diff --git a/arch/arm/mach-tegra/include/mach/kbc.h b/arch/arm/mach-tegra/include/mach/kbc.h
index 4f3572a..3610568 100644
--- a/arch/arm/mach-tegra/include/mach/kbc.h
+++ b/arch/arm/mach-tegra/include/mach/kbc.h

@@ -24,13 +24,8 @@
 #include <linux/types.h>
 #include <linux/input/matrix_keypad.h>
 
-#ifdef CONFIG_ARCH_TEGRA_2x_SOC
 #define KBC_MAX_GPIO	24
 #define KBC_MAX_KPENT	8
-#else
-#define KBC_MAX_GPIO	20
-#define KBC_MAX_KPENT	7
-#endif
 
 #define KBC_MAX_ROW	16
 #define KBC_MAX_COL	8

-- 
1.7.4.1

On Tue, Jan 17, 2012 at 03:26:50PM -0800, Rakesh Iyer wrote:

Add support for Tegra30 by correcting definitions. This is necessary to
make driver useful in Tegra30.

So what about older versions? There still could be users...

-- 
Dmitry

Dmitry Torokhov wrote at Tuesday, January 17, 2012 5:30 PM:

On Tue, Jan 17, 2012 at 03:26:50PM -0800, Rakesh Iyer wrote:

quoted

Add support for Tegra30 by correcting definitions. This is necessary to
make driver useful in Tegra30.

So what about older versions? There still could be users...

I wonder if there will be problems with this patch; the loop in
tegra_kbc_report_keys() is going to process 8 bits more of the last
register and the loop in tegra_kbc_config_pins() will now touch one
more row and one more column register. I think these loops should be
limited at run-time based on the SoC actually being executed upon?

-- 
nvpublic

Stephen Warren wrote at Wednesday, January 18, 2012 9:36 AM:

Dmitry Torokhov wrote at Tuesday, January 17, 2012 5:30 PM:

quoted

On Tue, Jan 17, 2012 at 03:26:50PM -0800, Rakesh Iyer wrote:

quoted

Add support for Tegra30 by correcting definitions. This is necessary to
make driver useful in Tegra30.

So what about older versions? There still could be users...

I wonder if there will be problems with this patch; the loop in
tegra_kbc_report_keys() is going to process 8 bits more of the last
register and the loop in tegra_kbc_config_pins() will now touch one
more row and one more column register. I think these loops should be
limited at run-time based on the SoC actually being executed upon?

Sorry, I read the ifdef the wrong way around. The new values are for
Tegra20 and should have been in use all along. There are no old
versions to consider here, since chips before Tegra20 aren't supported
by mainline (or any??) Linux.

So,

Acked-by: Stephen Warren: <redacted>

Sorry for the confusion.

-- 
nvpublic