garryt wrote:
> Hello,
>
> Please find attached an attempt of mach-imx board support. I have been able
> to
> test it with xenomai and could make other tests if needed. Does someone else
> use
> this kind of board ? Could any of the mailing list gurus have a quick look
> and
> tell if there is something missing /nok ?
Comments follow within the patch.
>
> This have been made with the support of the "Adapting ARM I-pipe patch to a
> new
> board." document on the xenomai site, and the already available pxa support.
> BTW the __ipipe_mach_get_tscinfo is missing in this doc.
Yes, I know, I will add this.
>
> kernel used 2.6.20-12, adeos patch 1.7-03
> diff -aburN ./linux-2.6.20.12-imx1/arch/arm/mach-imx/imx_irq.h
> ./linux-2.6.20.12-ipipe/arch/arm/mach-imx/imx_irq.h
> --- ./linux-2.6.20.12-imx1/arch/arm/mach-imx/imx_irq.h 1970-01-01
> 01:00:00.000000000 +0100
> +++ ./linux-2.6.20.12-ipipe/arch/arm/mach-imx/imx_irq.h 2007-06-13
> 12:00:44.000000000 +0200
> @@ -0,0 +1,80 @@
> +/*
> + * linux/arch/arm/mach-imx/imx_irq.h
> + *
> + * Author: [EMAIL PROTECTED]
> + * Still exprimental
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + */
> +#include <asm/arch/irqs.h>
> +
> +void __init imx_init_priority(void);
> +
> +static unsigned int imx_default_irq_priority[IMX_IRQS] __initdata = {
> (snip)
> + };
> diff -aburN ./linux-2.6.20.12-imx1/arch/arm/mach-imx/irq.c
> ./linux-2.6.20.12-ipipe/arch/arm/mach-imx/irq.c
> --- ./linux-2.6.20.12-imx1/arch/arm/mach-imx/irq.c 2007-06-22
> 09:54:30.000000000 +0200
> +++ ./linux-2.6.20.12-ipipe/arch/arm/mach-imx/irq.c 2007-06-22
> 09:51:38.000000000 +0200
> @@ -26,12 +26,19 @@
> #include <linux/init.h>
> #include <linux/list.h>
> #include <linux/timer.h>
> +#include <linux/ipipe.h>
>
> #include <asm/hardware.h>
> #include <asm/irq.h>
> #include <asm/io.h>
>
> #include <asm/mach/irq.h>
> +#include <asm/arch/imx-regs.h>
> +
> +/* Used for IMX INTERRUPT priority: Still Experimental */
> +#ifdef CONFIG_IPIPE
> +#include "imx_irq.h"
> +#endif
Here, I would put imx_default_irq_priority definition directly in irq.c,
since it is used only here, there is no reason to put it in a
header. Besides, imx_irq.h is not a good name, there is already an
include/asm-arm/mach-imx/irqs.h, what is the difference ?
> (...)
> - .set_type = imx_irq_type,
> + .set_type = imx_internal_irq_type,
What is the use of this renaming ?
> };
>
> static struct irq_chip imx_gpio_chip = {
> @@ -222,6 +229,25 @@
> .set_type = imx_gpio_irq_type,
> };
>
> +#ifdef CONFIG_IPIPE
> +void __init imx_init_priority(void)
> +{
> + unsigned int irq;
> + IMX_PRIO0=0;
> + IMX_PRIO1=0;
> + IMX_PRIO2=0;
> + IMX_PRIO3=0;
> + IMX_PRIO4=0;
> + IMX_PRIO5=0;
> + IMX_PRIO6=0;
> + IMX_PRIO7=0;
> + printk(KERN_INFO "Initializing imx interrupt priorities\n");
> + for (irq = 0; irq < IMX_IRQS; irq++) {
> + IMX_PRIO(irq)|=((imx_default_irq_priority[irq]&15)<<((irq%8)*4));
> + }
Coding style here. You do not need braces, and the indentation is
wrong.
> (...)
> +
> +#ifdef CONFIG_IPIPE
> + imx_timer_initialized = 1;
> + tsc = (union tsc_reg *) __ipipe_tsc_area;
> + barrier();
The aim of the barrier is to ensure that the "tsc" variable is
initialized before imx_timer_initialized is set to 1, so that
ipipe_mach_get_tsc will not dereference an uninitialized pointer.
Besides, tsc emulation in user-space can not work in the CONFIG_SMP
case, so this should be:
#ifdef CONFIG_SMP
tsc = (union tsc_reg *) __ipipe_tsc_area;
barrier();
#endif
imx_timer_initialized = 1;
> (...)
> +void __ipipe_mach_get_tscinfo(struct __ipipe_tscinfo *info)
> +{
> + info->type = IPIPE_TSC_TYPE_FREERUNNING;
> + info->u.fr.counter = (unsigned *) (0x10+IMX_TIM1_BASE);
> + info->u.fr.mask = 0xffffffff;
> + info->u.fr.tsc = &tsc->full;
> +}
There should be some #ifdef CONFIG_SMP here too.
> (...)
> +#ifdef CONFIG_IPIPE
> +#define __ipipe_mach_irq_mux_p(irq) (((irq) == GPIO_INT_PORTA ) || ((irq) ==
> GPIO_INT_PORTB ) || ((irq) == GPIO_INT_PORTC ) || ((irq) == GPIO_INT_PORTD)
> )
> +#endif /* CONFIG_IPIPE */
This test should be as fast as possible, so you should test a bit in a
mask, the way s3c2410 does it. Something like:
#define __ipipe_irqbit(irq) (1ULL << (irq))
#define __ipipe_muxed_irqmask (__ipipe_irqbit(GPIO_INT_PORTA) | \
__ipipe_irqbit(GPIO_INT_PORTB) | \
__ipipe_irqbit(GPIO_INT_PORTC) | \
__ipipe_irqbit(GPIO_INT_PORTD))
#define __ipipe_mach_irq_mux_p(irq) (__ipipe_irqbit(irq) \
& __ipipe_muxed_irqmask)
--
Gilles Chanteperdrix.
_______________________________________________
Adeos-main mailing list
[email protected]
https://mail.gna.org/listinfo/adeos-main
