input: bitmap update for sh_keysc V2
Subsystems (3)
PATCH 1/1 input: bitmap update for sh_keysc V2
From: Magnus Damm <[email protected]>
Update the sh_keysc driver with proper bitmap support V2. This instead of using a fixed 32-bit integer to keep track for the key states. With this change in place the driver supports key pads with more than 32 keys.
---
Changes since V1:
- ditched the wrapping macros
- use __set_bit() and __clear_bit()
drivers/input/keyboard/sh_keysc.c | 71 +++++++++++++++++++++++-------------- 1 file changed, 46 insertions(+), 25 deletions(-)
drivers/input/keyboard/sh_keysc.c +46 -25
@@ -19,6 +19,7 @@#include <linux/platform_device.h> #include <linux/input.h> #include <linux/input/sh_keysc.h> +#include <linux/bitmap.h> #include <linux/clk.h> #include <linux/io.h>
@@ -32,10 +33,14 @@ static const struct { [SH_KEYSC_MODE_5] = { 4, 6, 7 }, }; +struct sh_keysc_map { + DECLARE_BITMAP(b, SH_KEYSC_MAXKEYS); +}; +struct sh_keysc_priv {void __iomem *iomem_base;struct clk *clk; -unsigned long last_keys; +struct sh_keysc_map last_keys;struct input_dev *input;struct sh_keysc_info pdata; };
@@ -71,69 +76,85 @@ static void sh_keysc_level_mode(struct sudelay (pdata->kycr2_delay); } +static void sh_keysc_map_dbg (struct device *dev,struct sh_keysc_map *map, +const char *str) +{ +int k; + +for (k =0 ; k <BITS_TO_LONGS (SH_KEYSC_MAXKEYS ); k++) +dev_dbg (dev,"%s[%d] 0x%lx\n" , str, k, map->b[k]); +} +static irqreturn_t sh_keysc_isr (int irq,void *dev_id) {struct platform_device *pdev = dev_id;struct sh_keysc_priv *priv =platform_get_drvdata (pdev);struct sh_keysc_info *pdata = &priv->pdata; -unsigned long keys, keys1, keys0, mask; +int keyout_nr = sh_keysc_mode[pdata->mode].keyout; +int keyin_nr = sh_keysc_mode[pdata->mode].keyin; +struct sh_keysc_map keys, keys1, keys0;unsigned char keyin_set, tmp; -int i, k; +int i, k, n;dev_dbg (&pdev->dev,"isr!\n" ); - keys1 = ~0 ; - keys0 =0 ; +bitmap_fill (keys1.b,SH_KEYSC_MAXKEYS ); +bitmap_zero (keys0.b,SH_KEYSC_MAXKEYS );do { - keys =0 ; +bitmap_zero (keys.b,SH_KEYSC_MAXKEYS ); keyin_set =0 ;sh_keysc_write (priv,KYCR2 ,KYCR2_IRQ_DISABLED ); -for (i =0 ; i < sh_keysc_mode[pdata->mode].keyout; i++) { +for (i =0 ; i < keyout_nr; i++) { + n = keyin_nr * i; + +/* drive one KEYOUT pin low, read KEYIN pins */ sh_keysc_write (priv,KYOUTDR ,0xfff ^ (3 << (i *2 )));udelay (pdata->delay); tmp =sh_keysc_read (priv,KYINDR ); - keys |= tmp << (sh_keysc_mode[pdata->mode].keyin * i); - tmp ^= (1 << sh_keysc_mode[pdata->mode].keyin) -1 ; - keyin_set |= tmp; +/* set bit if key press has been detected */ +for (k =0 ; k < keyin_nr; k++) { +if (tmp & (1 << k)) +__set_bit (n + k, keys.b); + } + +/* keep track of which KEYIN bits that have been set */ + keyin_set |= tmp ^ ((1 << keyin_nr) -1 ); }sh_keysc_level_mode (priv, keyin_set); - keys ^= ~0 ; - keys &= (1 << (sh_keysc_mode[pdata->mode].keyin * - sh_keysc_mode[pdata->mode].keyout)) -1 ; - keys1 &= keys; - keys0 |= keys; +bitmap_complement (keys.b, keys.b,SH_KEYSC_MAXKEYS ); +bitmap_and (keys1.b, keys1.b, keys.b,SH_KEYSC_MAXKEYS ); +bitmap_or (keys0.b, keys0.b, keys.b,SH_KEYSC_MAXKEYS ); -dev_dbg (&pdev->dev,"keys 0x%08lx\n" , keys); +sh_keysc_map_dbg (&pdev->dev, &keys,"keys" ); }while (sh_keysc_read (priv,KYCR2 ) &0x01 ); -dev_dbg (&pdev->dev,"last_keys 0x%08lx keys0 0x%08lx keys1 0x%08lx\n" , - priv->last_keys, keys0, keys1); +sh_keysc_map_dbg (&pdev->dev, &priv->last_keys,"last_keys" ); +sh_keysc_map_dbg (&pdev->dev, &keys0,"keys0" ); +sh_keysc_map_dbg (&pdev->dev, &keys1,"keys1" );for (i =0 ; i <SH_KEYSC_MAXKEYS ; i++) { k = pdata->keycodes[i];if (!k)continue ; - mask =1 << i; - -if (!((priv->last_keys ^ keys0) & mask)) +if (test_bit (i, keys0.b) =test_bit (i, priv->last_keys.b))continue ; -if ((keys1 | keys0) & mask) { +if (test_bit (i, keys1.b) ||test_bit (i, keys0.b)) {input_event (priv->input,EV_KEY , k,1 ); - priv->last_keys |= mask; +__set_bit (i, priv->last_keys.b); } -if (!(keys1 & mask)) { +if (!test_bit (i, keys1.b)) {input_event (priv->input,EV_KEY , k,0 ); - priv->last_keys &= ~mask; +__clear_bit (i, priv->last_keys.b); } }
Dmitry Torokhov
maintainerQuoted 7 lines · Magnus Damm
From: Magnus Damm <[email protected]>
Update the
sh_keyscdriver with proper bitmap support V2. This instead of using a fixed 32-bit integer to keep track for the key states. With this change in place the driver supports key pads with more than 32 keys.
Quoted 6 lines
Signed-off-by Magnus Damm <[email protected]>---
Changes since V1:
- ditched the wrapping macros
- use__set_bit()and__clear_bit()
Thank you for making the change. It indeed is easier to read than the original since one does not have to go and look up what wrappers are doing.
Signature · 1 line
Dmitry
Quoted 11 lines · Dmitry Torokhov
Quoted 10 lines · Magnus Damm
From: Magnus Damm <[email protected]>
Update the
sh_keyscdriver with proper bitmap support V2. This instead of using a fixed 32-bit integer to keep track for the key states. With this change in place the driver supports key pads with more than 32 keys.Signed-off-by Magnus Damm <[email protected]>---
Quoted 7 lines
Quoted 3 lines
Changes since V1:
- ditched the wrapping macros
- use__set_bit()and__clear_bit()Thank you for making the change. It indeed is easier to read than the original since one does not have to go and look up what wrappers are doing.
Yeah, I initially disliked the SH_KEYSC_MAXKEYS that now are sprinkled all over the place, but I have to admit that not using wrappers made me realize that I should use __set_bit() and __clear_bit() instead of the atomic ones. So all good.
Thanks!
/ magnus
Dmitry Torokhov
maintainerQuoted 17 lines · Magnus Damm
Quoted 15 lines · Dmitry Torokhov
Quoted 13 lines · Magnus Damm
From: Magnus Damm <[email protected]>
Update the
sh_keyscdriver with proper bitmap support V2. This instead of using a fixed 32-bit integer to keep track for the key states. With this change in place the driver supports key pads with more than 32 keys.Signed-off-by Magnus Damm <[email protected]>---
Changes since V1:
- ditched the wrapping macros
- use__set_bit()and__clear_bit()
Quoted 9 lines
Quoted 3 lines
Thank you for making the change. It indeed is easier to read than the original since one does not have to go and look up what wrappers are doing.
Yeah, I initially disliked the
SH_KEYSC_MAXKEYSthat now are sprinkled all over the place, but I have to admit that not using wrappers made me realize that I should use__set_bit()and__clear_bit()instead of the atomic ones. So all good.
BTW, that sh_keysc_map seems to be a holdover from the V1 of the patch and is not needed anymore. Does the patch below still work for you?
Thanks!
Signature · 17 lines
Dmitry Input: sh_keysc - switch to using bitmaps From: Magnus Damm <[email protected]> Use bitmaps instead of using 32-bit integers to keep track of the key states. With this change in place the driver supports key pads with more than 32 keys. Signed-off-by: Magnus Damm <[email protected]> Signed-off-by: Dmitry Torokhov <[email protected]> --- drivers/input/keyboard/sh_keysc.c | 69 ++++++++++++++++++++++++------------- 1 files changed, 44 insertions(+), 25 deletions(-)
drivers/input/keyboard/sh_keysc.c +44 -25
@@ -19,6 +19,7 @@#include <linux/platform_device.h> #include <linux/input.h> #include <linux/input/sh_keysc.h> +#include <linux/bitmap.h> #include <linux/clk.h> #include <linux/io.h>
@@ -35,7 +36,7 @@ static const struct {struct sh_keysc_priv {void __iomem *iomem_base;struct clk *clk; -unsigned long last_keys; + DECLARE_BITMAP(last_keys, SH_KEYSC_MAXKEYS);struct input_dev *input;struct sh_keysc_info pdata; };
@@ -71,69 +72,87 @@ static void sh_keysc_level_mode(struct sh_keysc_priv *p,udelay (pdata->kycr2_delay); } +static void sh_keysc_map_dbg (struct device *dev,unsigned long *map, +const char *str) +{ +int k; + +for (k =0 ; k <BITS_TO_LONGS (SH_KEYSC_MAXKEYS ); k++) +dev_dbg (dev,"%s[%d] 0x%lx\n" , str, k, map[k]); +} +static irqreturn_t sh_keysc_isr (int irq,void *dev_id) {struct platform_device *pdev = dev_id;struct sh_keysc_priv *priv =platform_get_drvdata (pdev);struct sh_keysc_info *pdata = &priv->pdata; -unsigned long keys, keys1, keys0, mask; +int keyout_nr = sh_keysc_mode[pdata->mode].keyout; +int keyin_nr = sh_keysc_mode[pdata->mode].keyin; +DECLARE_BITMAP (keys,SH_KEYSC_MAXKEYS ); +DECLARE_BITMAP (keys0,SH_KEYSC_MAXKEYS ); +DECLARE_BITMAP (keys1,SH_KEYSC_MAXKEYS );unsigned char keyin_set, tmp; -int i, k; +int i, k, n;dev_dbg (&pdev->dev,"isr!\n" ); - keys1 = ~0 ; - keys0 =0 ; +bitmap_fill (keys1,SH_KEYSC_MAXKEYS ); +bitmap_zero (keys0,SH_KEYSC_MAXKEYS );do { - keys =0 ; +bitmap_zero (keys,SH_KEYSC_MAXKEYS ); keyin_set =0 ;sh_keysc_write (priv,KYCR2 ,KYCR2_IRQ_DISABLED ); -for (i =0 ; i < sh_keysc_mode[pdata->mode].keyout; i++) { +for (i =0 ; i < keyout_nr; i++) { + n = keyin_nr * i; + +/* drive one KEYOUT pin low, read KEYIN pins */ sh_keysc_write (priv,KYOUTDR ,0xfff ^ (3 << (i *2 )));udelay (pdata->delay); tmp =sh_keysc_read (priv,KYINDR ); - keys |= tmp << (sh_keysc_mode[pdata->mode].keyin * i); - tmp ^= (1 << sh_keysc_mode[pdata->mode].keyin) -1 ; - keyin_set |= tmp; +/* set bit if key press has been detected */ +for (k =0 ; k < keyin_nr; k++) { +if (tmp & (1 << k)) +__set_bit (n + k, keys); + } + +/* keep track of which KEYIN bits that have been set */ + keyin_set |= tmp ^ ((1 << keyin_nr) -1 ); }sh_keysc_level_mode (priv, keyin_set); - keys ^= ~0 ; - keys &= (1 << (sh_keysc_mode[pdata->mode].keyin * - sh_keysc_mode[pdata->mode].keyout)) -1 ; - keys1 &= keys; - keys0 |= keys; +bitmap_complement (keys, keys,SH_KEYSC_MAXKEYS ); +bitmap_and (keys1, keys1, keys,SH_KEYSC_MAXKEYS ); +bitmap_or (keys0, keys0, keys,SH_KEYSC_MAXKEYS ); -dev_dbg (&pdev->dev,"keys 0x%08lx\n" , keys); +sh_keysc_map_dbg (&pdev->dev, keys,"keys" ); }while (sh_keysc_read (priv,KYCR2 ) &0x01 ); -dev_dbg (&pdev->dev,"last_keys 0x%08lx keys0 0x%08lx keys1 0x%08lx\n" , - priv->last_keys, keys0, keys1); +sh_keysc_map_dbg (&pdev->dev, priv->last_keys,"last_keys" ); +sh_keysc_map_dbg (&pdev->dev, keys0,"keys0" ); +sh_keysc_map_dbg (&pdev->dev, keys1,"keys1" );for (i =0 ; i <SH_KEYSC_MAXKEYS ; i++) { k = pdata->keycodes[i];if (!k)continue ; - mask =1 << i; - -if (!((priv->last_keys ^ keys0) & mask)) +if (test_bit (i, keys0) =test_bit (i, priv->last_keys))continue ; -if ((keys1 | keys0) & mask) { +if (test_bit (i, keys1) ||test_bit (i, keys0)) {input_event (priv->input,EV_KEY , k,1 ); - priv->last_keys |= mask; +__set_bit (i, priv->last_keys); } -if (!(keys1 & mask)) { +if (!test_bit (i, keys1)) {input_event (priv->input,EV_KEY , k,0 ); - priv->last_keys &= ~mask; +__clear_bit (i, priv->last_keys); } }
Quoted 22 lines · Dmitry Torokhov
Quoted 21 lines · Magnus Damm
Quoted 18 lines · Dmitry Torokhov
Quoted 7 lines · Magnus Damm
From: Magnus Damm <[email protected]>
Update the
sh_keyscdriver with proper bitmap support V2. This instead of using a fixed 32-bit integer to keep track for the key states. With this change in place the driver supports key pads with more than 32 keys.Quoted 6 lines
Signed-off-by Magnus Damm <[email protected]>---
Changes since V1:
- ditched the wrapping macros
- use__set_bit()and__clear_bit()Thank you for making the change. It indeed is easier to read than the original since one does not have to go and look up what wrappers are doing.
Quoted 8 lines
Quoted 5 lines
Yeah, I initially disliked the
SH_KEYSC_MAXKEYSthat now are sprinkled all over the place, but I have to admit that not using wrappers made me realize that I should use__set_bit()and__clear_bit()instead of the atomic ones. So all good.BTW, that
sh_keysc_mapseems to be a holdover from the V1 of the patch and is not needed anymore. Does the patch below still work for you?
Thanks for updating the patch. It still works as expected. Please mark the new patch as V3 if possible so people can
tell the difference between the patches.
Cheers,
/ magnus
Dmitry Torokhov
maintainerQuoted 31 lines · Magnus Damm
On Wed, Feb 10, 2010 at 6:26 PM, Dmitry Torokhov
Quoted 28 lines · <[email protected]> wrote:
Quoted 26 lines · Magnus Damm
On Tue, Feb 9, 2010 at 6:39 PM, Dmitry Torokhov
Quoted 11 lines · <[email protected]> wrote:
Quoted 10 lines · Magnus Damm
From: Magnus Damm <[email protected]>
Update the
sh_keyscdriver with proper bitmap support V2. This instead of using a fixed 32-bit integer to keep track for the key states. With this change in place the driver supports key pads with more than 32 keys.Signed-off-by Magnus Damm <[email protected]>---
Quoted 7 lines
Quoted 3 lines
Changes since V1:
- ditched the wrapping macros
- use__set_bit()and__clear_bit()Thank you for making the change. It indeed is easier to read than the original since one does not have to go and look up what wrappers are doing.
Yeah, I initially disliked the
SH_KEYSC_MAXKEYSthat now are sprinkled all over the place, but I have to admit that not using wrappers made me realize that I should use__set_bit()and__clear_bit()instead of the atomic ones. So all good.
Quoted 4 lines
Quoted 2 lines
BTW, that
sh_keysc_mapseems to be a holdover from the V1 of the patch and is not needed anymore. Does the patch below still work for you?Thanks for updating the patch. It still works as expected.
Great, thank you for testing it.
Quoted 4 lines
Please mark
the new patch as V3 if possible so people can
tell the difference between the patches.
Umm, I usually remove all versioning from the changelogs when I apply the patches to my tree.
Signature · 1 line
Dmitry