[PATCH linux v1 1/3] drivers: misc: Platform driver to update display on the POST card
Rick Altherr
raltherr at google.com
Thu Oct 13 12:09:20 AEDT 2016
On Wed, Oct 12, 2016 at 5:41 PM, Jaghathiswari Rankappagounder Natarajan <
jaghu at google.com> wrote:
> Platform driver to periodically update the display on the POST card.
> POST card hardware assumed is 74HC164 wired to two 7-segment displays.
>
> Signed-off-by: Jaghathiswari Rankappagounder Natarajan <jaghu at google.com>
> ---
> drivers/misc/Kconfig | 6 ++
> drivers/misc/Makefile | 1 +
> drivers/misc/update-postcard.c | 219 ++++++++++++++++++++++++++++++
> +++++++++++
> drivers/misc/update-postcard.h | 1 +
> 4 files changed, 227 insertions(+)
> create mode 100644 drivers/misc/update-postcard.c
> create mode 100644 drivers/misc/update-postcard.h
>
> diff --git a/drivers/misc/Kconfig b/drivers/misc/Kconfig
> index 4617ddc..d65b83f 100644
> --- a/drivers/misc/Kconfig
> +++ b/drivers/misc/Kconfig
> @@ -804,6 +804,12 @@ config PANEL_BOOT_MESSAGE
> An empty message will only clear the display at driver init
> time. Any other
> printf()-formatted message is valid with newline and escape
> codes.
>
> +config ASPEED_UPDATE_POSTCARD
> + tristate "Platform driver to update POST card"
> + help
> + Support to periodically update display on the POST card. The POST
> card
> + hardware assumed is 74HC164 wired to two 7-segment displays.
> +
>
Nothing about this driver is associated with Aspeed. UPDATE_POSTCARD reads
to me as describing an action, not a device. Maybe SEVEN_SEGMENT_DISPLAY?
> config ASPEED_BT_IPMI_HOST
> tristate "BT IPMI host driver"
> help
> diff --git a/drivers/misc/Makefile b/drivers/misc/Makefile
> index 724861b..730c6c2 100644
> --- a/drivers/misc/Makefile
> +++ b/drivers/misc/Makefile
> @@ -57,4 +57,5 @@ obj-$(CONFIG_ECHO) += echo/
> obj-$(CONFIG_VEXPRESS_SYSCFG) += vexpress-syscfg.o
> obj-$(CONFIG_CXL_BASE) += cxl/
> obj-$(CONFIG_PANEL) += panel.o
> +obj-$(CONFIG_ASPEED_UPDATE_POSTCARD) += update-postcard.o
> obj-$(CONFIG_ASPEED_BT_IPMI_HOST) += bt-host.o
> diff --git a/drivers/misc/update-postcard.c b/drivers/misc/update-
> postcard.c
> new file mode 100644
> index 0000000..3516c24
> --- /dev/null
> +++ b/drivers/misc/update-postcard.c
> @@ -0,0 +1,219 @@
> +/* Platform driver to periodically update the display on the POST card.
> + * POST card hardware assumed is 74HC164 wired to two 7-segment displays.
> + */
> +
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/timer.h>
> +#include <linux/jiffies.h>
> +#include <linux/sizes.h>
> +#include <linux/map_to_7segment.h>
> +#include <linux/io.h>
> +#include <linux/delay.h>
> +#include <linux/uaccess.h>
> +#include <linux/mutex.h>
> +#include <linux/of_platform.h>
> +#include <linux/gpio/consumer.h>
> +
> +#define POST_CODE_UPDATE_INTVL 1000
> +
> +#define MAX_PC_SIZE 3
> +
> +#define LED_DOT 0x01
> +
> +#define CLOCK_GPIO_NAME "clk"
> +#define DATA_GPIO_NAME "data"
> +#define CLEAR_GPIO_NAME "clr"
> +
> +static struct mutex mutex;
> +
> +static struct timer_list update_timer;
> +
> +static char curr_postcode[3];
> +static u8 pc_valid;
> +
>
+static struct gpio_desc *clk_gpio;
> +static struct gpio_desc *data_gpio;
> +static struct gpio_desc *clr_gpio;
>
I realize there should really only ever be one of these in a system but
static data members in modules bug me. I'd much rather see all this
configuration and these buffers in a struct that is saved with
platform_set_drvdata().
> +
> +/*
> + * 0 1 2 3 4 5 6 7 8 9 A B C D E F
> + * _ _ _ _ _ _ _ _ _ _ _ _
> + * | | | _| _| |_| |_ |_ | |_| |_| |_| |_ | _| |_ |_
> + * |_| | |_ _| | _| |_| | |_| | | | |_| |_ |_| |_ |
> + *
> + * data[7:1] = led[a:g]
> + * lookup table used for both bytes of lpc post code and lower byte of
> + * bmc post code
> + */
> +const u8 seven_seg_bits[] = {
> + 0xFC, 0x60, 0xDA, 0xF2, 0x66, 0xB6, 0xBE, 0xE0,
> + 0xFE, 0xF6, 0xEE, 0x3E, 0x9C, 0x7A, 0x9E, 0x8E
> + };
> +
> +/*
> + * 0 1 2 3 4 5 6 7 8 9 A B C D E F
> + * _ _ _ _ _
> + * | |_ |_| |_ _ _ _ _ _ _ _ |_ _| _| | |
> + * |_ |_ | | _| |_| |_| | |
> + *
> + * data[7:1] = led[a:g]
> + * lookup table used for higher byte of bmc post code
> + */
> +const u8 special_seven_seg_bits[] = {
> + 0x00, 0x9C, 0x1E, 0xCE, 0x8E, 0x02, 0x02, 0x02,
> + 0x02, 0x02, 0x02, 0x02, 0xB6, 0x7A, 0x7A, 0xEC
> + };
> +
> +void update_postcode(char *buf)
> +{
> + mutex_lock(&mutex);
> + strncpy(curr_postcode, buf, sizeof(curr_postcode));
> + mutex_unlock(&mutex);
> + pc_valid = 1;
> +}
> +EXPORT_SYMBOL(update_postcode);
> +
> +/* convert postcode to led pattern
> + * 7-bits used for each 7-seg display and
> + * 1 bit used for the 'dot' on both digits.
> + *
> + * @param N/A
> + * @return N/A
> + */
> +static u16 postcode_to_led_signal(void)
> +{
> + u8 low_display;
> + u8 high_display;
> + u16 led_value;
> + char buf[3];
> +
> + mutex_lock(&mutex);
> + strncpy(buf, curr_postcode, sizeof(buf));
> + mutex_unlock(&mutex);
> +
> + low_display = seven_seg_bits[hex_to_bin(buf[2])];
> +
> + high_display = (buf[0] == '1') ?
> + special_seven_seg_bits[hex_to_bin(buf[1])] :
> + seven_seg_bits[hex_to_bin(buf[1])];
> +
> + led_value = low_display | (high_display << 8);
> + if (buf[0] == '1') {
> + led_value |= LED_DOT | (LED_DOT << 8);
> + }
> +
> + return led_value;
> +}
> +
> +static void update_sgpio(void)
> +{
> + u16 led_signal;
> + int i;
> +
> + /* Convert two bytes of post code to 16 bit led pattern */
> + led_signal = postcode_to_led_signal();
>
Why do you wait until the display update to convert the string to the LED
bit pattern? For that matter, why does update_postcode() take a string at
all? If you can only display at most a 16-bit number, take a u16 and have
the caller deal with whatever processing they need.
> +
> + /* Set the clear signal to low */
> + gpiod_set_value(clr_gpio, 0);
> + udelay(1);
> + /* Set the clear clear to high */
> + gpiod_set_value(clr_gpio, 1);
> + udelay(1);
> +
> + /* Bitbang the 16 bit led pattern
> + * 7-bits used for each 7-seg display
> + * 1 bit used for the 'dot' on both digits
> + */
> + for (i = 0; i < 16; i++) {
> + if (led_signal & 0x01) {
> + /* Set the data signal to high */
> + gpiod_set_value(data_gpio, 1);
> + } else {
> + /* Set the data signal to low */
> + gpiod_set_value(data_gpio, 0);
> + }
> + udelay(1);
> +
> + /* Set the clock signal to low */
> + gpiod_set_value(clk_gpio, 0);
> + udelay(1);
> + /* Set the clock signal to high */
> + gpiod_set_value(clk_gpio, 1);
> + udelay(1);
> +
> + led_signal >>= 1;
> + }
> +}
> +
> +static void update_timer_handler(unsigned long data)
> +{
> + if (pc_valid == 1) {
>
Is this to avoid the time after the driver is register until
update_postcode() is called the first time? Is there an API for clearing
the display?
> + update_sgpio();
> + }
> + mod_timer(&update_timer,
> + jiffies + msecs_to_jiffies(POST_CODE_UPDATE_INTVL));
> +}
> +
> +static const struct of_device_id of_postcard_match[] = {
> + { .compatible = "postcard-display" },
> + {},
> +};
> +
> +MODULE_DEVICE_TABLE(of, of_postcard_match);
> +
> +static int postcard_probe(struct platform_device *pdev)
> +{
> + int result;
> + struct device *dev = &pdev->dev;
> +
> + /* Requesting the clock gpio */
> + clk_gpio = devm_gpiod_get(dev, CLOCK_GPIO_NAME,
> + GPIOD_OUT_HIGH);
> + if (IS_ERR(clk_gpio))
> + return PTR_ERR(data_gpio);
> +
> + /* Requesting the data gpio */
> + data_gpio = devm_gpiod_get(dev, DATA_GPIO_NAME,
> + GPIOD_OUT_HIGH);
> + if (IS_ERR(data_gpio))
> + return PTR_ERR(data_gpio);
> +
> + /* Requesting the clear gpio */
> + clr_gpio = devm_gpiod_get(dev, CLEAR_GPIO_NAME,
> + GPIOD_OUT_HIGH);
> + if (IS_ERR(clr_gpio))
> + return PTR_ERR(clr_gpio);
> +
> + /* Start timer to update post code every second */
> + setup_timer(&update_timer, update_timer_handler, 0);
> + result = mod_timer(&update_timer,
> + jiffies + msecs_to_jiffies(POST_CODE_UPDATE_INTVL));
> +
> + if (result)
> + return result;
> +
> + mutex_init(&mutex);
> +
> + return 0;
> +}
> +
> +static void postcard_remove(struct platform_device *pdev)
> +{
> + del_timer(&update_timer);
> +}
> +
> +static struct platform_driver postcard_display_driver = {
> + .probe = postcard_probe,
> + .shutdown = postcard_remove,
> + .driver = {
> + .name = "postcard-display",
> + .of_match_table = of_postcard_match,
> + },
> +};
> +
> +module_platform_driver(postcard_display_driver);
+
> +MODULE_LICENSE("GPL");
> +MODULE_AUTHOR("Jaghathiswari Rankappagounder Natarajan <jaghu at google.com
> >");
> +MODULE_DESCRIPTION("Post card display update driver");
> diff --git a/drivers/misc/update-postcard.h b/drivers/misc/update-
> postcard.h
> new file mode 100644
> index 0000000..ef82e9091d
> --- /dev/null
> +++ b/drivers/misc/update-postcard.h
> @@ -0,0 +1 @@
> +void update_postcode(char *buf);
> --
> 2.8.0.rc3.226.g39d4020
>
> _______________________________________________
> openbmc mailing list
> openbmc at lists.ozlabs.org
> https://lists.ozlabs.org/listinfo/openbmc
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ozlabs.org/pipermail/openbmc/attachments/20161012/47d11c74/attachment-0001.html>
More information about the openbmc
mailing list