From: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
To: Tony Lindgren <tony@atomide.com>
Cc: linux-omap@vger.kernel.org, linux-arm-kernel@lists.infradead.org,
linux-kernel@vger.kernel.org,
Grant Likely <grant.likely@secretlab.ca>
Subject: Re: [PATCH v2 2/7] ARM: OMAP1: ams-delta: convert latches to basic_mmio_gpio
Date: Tue, 20 Dec 2011 01:59:36 +0100 [thread overview]
Message-ID: <201112200159.36558.jkrzyszt@tis.icnet.pl> (raw)
In-Reply-To: <20111220000600.GY6464@atomide.com>
On Tuesday 20 of December 2011 at 01:06:00, Tony Lindgren wrote:
> * Janusz Krzysztofik <jkrzyszt@tis.icnet.pl> [111219 14:41]:
> > Once ready, ams-delta specific device drivers currently calling custom
> > ams_delta_latch[12]_write() functions can be updated to call generic
> > gpio_set_value() instead, which will make them less platform dependent.
> > Even more, some custom ams-delta only drivers can perhaps be dropped
> > from the tree after converting selected ams-delta platform devices to
> > follow generic GPIO based device models.
> >
> > Depends on patch 1/7, "ARM: OMAP1: ams-delta: register latch dependent
> > devices later".
>
> Hmm looking at this maybe you can move the all the latch stuff into
> a device driver?
The latch stuff is just platform data for the existing gpio-generic aka
basic_mmio_gpio driver. I'm not sure if creating a new custom driver by
just squashing an existig driver code with a board specific platform
data is a good idea.
> Then you can have the other drivers register with it and let the
> module dependencies take care of the init order?
It it was more complicated than providing just platform data, not even a
single custom callback, to an existing gpio-generic driver, creating a
custom driver might help indeed. However, I think I have all issues with
initialization order already sorted out without inventing a new driver.
> You should only register the platform_device entries in your board-*.c
> file, I don't think you actually need to do anything there to power
> up things in the board-*.c file execept for the 8250.c driver?
Exactly, but not in a single patch. With this patch, I keep the old API
for all drivers to still work, that's why I have to handle GPIO pins on
behalf of them until updated. Those are updated step by step throghout
the following patches of the series.
> > +static struct gpio latch_gpios[] __initconst = {
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_LED_CAMERA,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "led_camera",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_LED_ADVERT,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "led_advert",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_LED_EMAIL,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "led_email",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_LED_HANDSFREE,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "led_handsfree",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_LED_VOICEMAIL,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "led_voicemail",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_LED_VOICE,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "led_voice",
> > + },
> > + {
> > + .gpio = AMS_DELTA_LATCH1_GPIO_BASE + 6,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "dockit1",
> > + },
> > + {
> > + .gpio = AMS_DELTA_LATCH1_GPIO_BASE + 7,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "dockit2",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_LCD_VBLEN,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "lcd_vblen",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_LCD_NDISP,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "lcd_ndisp",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_NAND_NCE,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "nand_nce",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_NAND_NRE,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "nand_nre",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_NAND_NWP,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "nand_nwp",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_NAND_NWE,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "nand_nwe",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_NAND_ALE,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "nand_ale",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_NAND_CLE,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "nand_cle",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_KEYBRD_PWR,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "keybrd_pwr",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_KEYBRD_DATAOUT,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "keybrd_dataout",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_SCARD_RSTIN,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "scard_rstin",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_SCARD_CMDVCC,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "scard_cmdvcc",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_MODEM_NRESET,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "modem_nreset",
> > + },
> > + {
> > + .gpio = AMS_DELTA_GPIO_PIN_MODEM_CODEC,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "modem_codec",
> > + },
> > + {
> > + .gpio = AMS_DELTA_LATCH2_GPIO_BASE + 14,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "hookflash1",
> > + },
> > + {
> > + .gpio = AMS_DELTA_LATCH2_GPIO_BASE + 15,
> > + .flags = GPIOF_OUT_INIT_LOW,
> > + .label = "hookflash2",
> > + },
> > +};
> > +
> > +void ams_delta_latch_write(int base, int ngpio, u16 mask, u16 value)
> > +{
> > + int bit = 0;
> > + u16 bitpos = 1 << bit;
> > +
> > + for (; bit < ngpio; bit++, bitpos = bitpos << 1) {
> > + if (!(mask & bitpos))
> > + continue;
> > + gpio_set_value(base + bit, (value & bitpos) != 0);
> > + }
> > +}
> > +EXPORT_SYMBOL(ams_delta_latch_write);
>
> This part especially looks like it really should be just a regular
> device driver under drivers/ somewhere.
I really don't understand what kind of a driver you might mean here.
The latch_gpios[] table is initially filled with all latch1 and latch2
GPIO pins in order to register and initialize them from the board file
until they are handled by respective existing device drivers (leds,
nand, lcd, serio, serial8250, asoc) instead of those drivers accessing
the latches with those old ams_delta_latch[12]_write() functions. That
table will get almost empty after the transision process is completed,
holding only pins not used by any drivers / connected to unsued devices,
in order to initialize them from the board file for power saving
purposes. A separate driver for the purpose of initializing a few GPIO
pins seems an overkill.
The new ams_delta_latch_write() function is a unified replacement for
those removed ams_delta_latch[12]_write(), and serves as a temporary
wrapper over gpio_set_value(), providing the old API for those not yet
updated device drivers, and will be removed after all drivers are
converted.
Perhaps I was not clear enough with my intention of a smooth step by
step transition to the GPIO API without breaking any signle driver with
any single patch.
> That might simplify things quite a bit for you..
Will be simplified, step by step, while moving GPIO handling from the
board file to all those existing device drivers.
I hope this clarifies things enough.
Thanks,
Janusz
next prev parent reply other threads:[~2011-12-20 1:00 UTC|newest]
Thread overview: 63+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-12-11 20:11 [PATCH 00/10] ARM: OMAP1: ams-delta: replace custom I/O with GPIO Janusz Krzysztofik
2011-12-11 20:11 ` [PATCH 01/10] GPIO: gpio-generic: Move initialization up to postcore Janusz Krzysztofik
2011-12-12 18:04 ` Tony Lindgren
2011-12-12 23:05 ` Janusz Krzysztofik
2011-12-12 23:15 ` Tony Lindgren
2011-12-12 23:44 ` Janusz Krzysztofik
2011-12-12 23:55 ` Tony Lindgren
2011-12-13 0:15 ` Janusz Krzysztofik
2011-12-14 13:10 ` Janusz Krzysztofik
2011-12-14 18:21 ` Tony Lindgren
2011-12-14 20:07 ` Janusz Krzysztofik
2011-12-14 22:10 ` Tony Lindgren
2011-12-14 11:13 ` Janusz Krzysztofik
2011-12-11 20:12 ` [PATCH 02/10] ARM: OMAP1: ams-delta: Convert latches to basic_mmio_gpio Janusz Krzysztofik
2011-12-11 20:12 ` [PATCH 03/10] ARM: OMAP1: ams-delta: Supersede custom led device by leds-gpio Janusz Krzysztofik
2011-12-11 20:12 ` [PATCH 04/10] LED: Drop leds-ams-delta driver Janusz Krzysztofik
2011-12-11 20:12 ` [PATCH 05/10] MTD: NAND: ams-delta: Use GPIO instead of custom I/O Janusz Krzysztofik
2011-12-11 20:12 ` [PATCH 06/10] ARM: OMAP1: ams-delta: Use GPIO API in modem setup Janusz Krzysztofik
2011-12-11 20:12 ` [PATCH 07/10] ASoC: OMAP: ams-delta: Drive modem/codec pins over GPIO API Janusz Krzysztofik
2011-12-12 4:29 ` Mark Brown
2011-12-21 18:33 ` Liam Girdwood
2011-12-21 22:14 ` Janusz Krzysztofik
2011-12-22 9:25 ` Girdwood, Liam
2011-12-11 20:12 ` [PATCH 08/10] omapfb: lcd_ams_delta: Drive control lines over GPIO Janusz Krzysztofik
2011-12-12 11:24 ` Tomi Valkeinen
2011-12-11 20:12 ` [PATCH 09/10] input: serio: ams-delta: Toggle keyboard power " Janusz Krzysztofik
2011-12-12 8:18 ` Dmitry Torokhov
2011-12-11 20:12 ` [PATCH 10/10] ARM: OMAP1: ams-delta: Drop custom I/O functions Janusz Krzysztofik
2011-12-12 5:00 ` [PATCH 00/10] ARM: OMAP1: ams-delta: replace custom I/O with GPIO Jonathan McDowell
2011-12-19 23:08 ` [PATCH v2 0/7] " Janusz Krzysztofik
2011-12-19 23:08 ` [PATCH v2 2/7] ARM: OMAP1: ams-delta: convert latches to basic_mmio_gpio Janusz Krzysztofik
2011-12-20 0:06 ` Tony Lindgren
2011-12-20 0:59 ` Janusz Krzysztofik [this message]
2011-12-20 1:04 ` Tony Lindgren
2011-12-20 1:18 ` Janusz Krzysztofik
2011-12-20 2:13 ` Tony Lindgren
2011-12-20 2:24 ` [PATCH v2 2/7 v2] " Janusz Krzysztofik
2011-12-19 23:08 ` [PATCH v2 3/7] ARM: OMAP1: ams-delta: supersede custom led device by leds-gpio Janusz Krzysztofik
2011-12-19 23:08 ` [PATCH v2 4/7] LED: drop leds-ams-delta driver Janusz Krzysztofik
2011-12-19 23:08 ` [PATCH v2 5/7] MTD: NAND: ams-delta: use GPIO instead of custom I/O Janusz Krzysztofik
2011-12-21 19:12 ` Tony Lindgren
2011-12-21 19:45 ` Janusz Krzysztofik
2011-12-21 20:56 ` Artem Bityutskiy
2011-12-19 23:08 ` [PATCH v2 6/7] omapfb: lcd_ams_delta: drive control lines over GPIO Janusz Krzysztofik
2011-12-19 23:08 ` [PATCH v2 7/7] input: serio: ams-delta: toggle keyboard power " Janusz Krzysztofik
2011-12-20 22:10 ` [PATCH] " Janusz Krzysztofik
2011-12-21 19:09 ` Tony Lindgren
2011-12-21 19:55 ` Janusz Krzysztofik
2011-12-22 17:45 ` Dmitry Torokhov
2011-12-22 18:06 ` Tony Lindgren
[not found] ` <83e934adfc691b347534edb7788a67ab2e6bd7e1.1324331816.git.jkrzyszt@tis.icnet.pl>
2011-12-19 23:28 ` [PATCH v2 1/7][RESEND] ARM: OMAP1: ams-delta: register latch dependent devices later Janusz Krzysztofik
2011-12-20 18:06 ` Tony Lindgren
2011-12-20 20:34 ` Janusz Krzysztofik
2011-12-20 20:57 ` Tony Lindgren
2011-12-20 20:40 ` Russell King - ARM Linux
2011-12-20 20:51 ` [alsa-devel] " Janusz Krzysztofik
2011-12-20 21:54 ` [PATCH v2 1/7 v2] " Janusz Krzysztofik
2011-12-21 19:08 ` Tony Lindgren
2011-12-21 19:51 ` Janusz Krzysztofik
2011-12-21 20:07 ` Tony Lindgren
2011-12-22 10:39 ` Janusz Krzysztofik
2011-12-22 11:08 ` Jarkko Nikula
2011-12-22 11:10 ` Mark Brown
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=201112200159.36558.jkrzyszt@tis.icnet.pl \
--to=jkrzyszt@tis.icnet.pl \
--cc=grant.likely@secretlab.ca \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-omap@vger.kernel.org \
--cc=tony@atomide.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®