| To: | technoboy85@gmail.com |
|---|---|
| Subject: | Re: [PATCH][MIPS]: AR7 GPIO |
| From: | Atsushi Nemoto <anemo@mba.ocn.ne.jp> |
| Date: | Sun, 13 Jan 2008 21:32:44 +0900 (JST) |
| Cc: | linux-mips@linux-mips.org, florian@openwrt.org, nbd@openwrt.org, ejka@imfi.kspu.ru, nico@openwrt.org, ralf@linux-mips.org, akpm@linux-foundation.org |
| In-reply-to: | <200801121818.02550.technoboy85@gmail.com> |
| Original-recipient: | rfc822;linux-mips@linux-mips.org |
| References: | <200801121818.02550.technoboy85@gmail.com> |
| Sender: | linux-mips-bounce@linux-mips.org |
On Sat, 12 Jan 2008 18:18:02 +0100, Matteo Croce <technoboy85@gmail.com> wrote:
> +static inline void gpio_set_value(unsigned gpio, int value)
> +{
> + static void __iomem *gpio_out;
> + unsigned tmp;
> +
> + if (!gpio_out)
> + gpio_out = (void __iomem *)
> + KSEG1ADDR(AR7_REGS_GPIO + AR7_GPIO_OUTPUT);
> +
> + tmp = readl(gpio_out) & ~(1 << gpio);
> + if (value)
> + tmp |= 1 << gpio;
> + writel(tmp, gpio_out);
> +}
It seems the compiler can calculate gpio_out value at compile time.
So I think caching it just make the function slower.
---
Atsushi Nemoto
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: [PATCH][MIPS]: AR7 GPIO, Atsushi Nemoto |
|---|---|
| Next by Date: | Re: [SPAM] flush_cache_page, Markus Gothe |
| Previous by Thread: | Re: [PATCH][MIPS]: AR7 GPIO, Atsushi Nemoto |
| Next by Thread: | Re: [UPDATED PATCH] SGISEEQ: use cached memory access to make driver work on IP28, Jeff Garzik |
| Indexes: | [Date] [Thread] [Top] [All Lists] |