[U-Boot] [PATCH 1/3 v8] arm926ejs: add NXP LPC32x0 cpu series support

[Vladimir Zapolskiy]

Hi Albert,

On 10.02.2012 20:00, Albert ARIBAUD wrote:
> Hi Vladimir,
>
> Le 13/01/2012 01:42, Vladimir Zapolskiy a écrit :
>
>> diff --git a/arch/arm/cpu/arm926ejs/lpc32xx/cpu.c
>> b/arch/arm/cpu/arm926ejs/lpc32xx/cpu.c
>> new file mode 100644
>> index 0000000..e29e130
>> --- /dev/null
>> +++ b/arch/arm/cpu/arm926ejs/lpc32xx/cpu.c
>> @@ -0,0 +1,70 @@
>> +/*
>> + * Copyright (C) 2011 by Vladimir Zapolskiy<vz at mleia.com>
>> + *
>> + * This program is free software; you can redistribute it and/or
>> + * modify it under the terms of the GNU General Public License
>> + * as published by the Free Software Foundation; either version 2
>> + * of the License, or (at your option) any later version.
>> + *
>> + * This program is distributed in the hope that it will be useful,
>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
>> + * GNU General Public License for more details.
>> + *
>> + * You should have received a copy of the GNU General Public License
>> + * along with this program; if not, write to the Free Software
>> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
>> + * MA 02110-1301, USA.
>> + */
>> +
>> +#include<common.h>
>> +#include<asm/arch/cpu.h>
>> +#include<asm/arch/clk.h>
>> +#include<asm/arch/wdt.h>
>> +#include<asm/io.h>
>> +
>> +static struct clk_pm_regs *clk = (struct clk_pm_regs *)CLK_PM_BASE;
>> +static struct wdt_regs *wdt = (struct wdt_regs *)WDT_BASE;
>> +
>> +void reset_cpu(ulong addr)
>> +{
>> + /* Enable watchdog clock */
>> + setbits_le32(&clk->timclk_ctrl, CLK_TIMCLK_WATCHDOG);
>> +
>> + /* Reset pulse length is 13005 peripheral clock frames */
>> + writel(13000,&wdt->pulse);
>
> I can't help but ask, if the comment says 13005, then why does the code
> say 13000?
>
the provided comment is done according to the descriptions of 
WDTIM_PULSE register from User Manual, for register value X it gives the 
RESET pulse length of (X + 5) PERIPH_CLKs.

>> + /* Force WDOG_RESET2 and RESOUT_N signal active */
>> + writel(WDTIM_MCTRL_RESFRC2 | WDTIM_MCTRL_RESFRC1 | WDTIM_MCTRL_M_RES2,
>> + &wdt->mctrl);
>> +
>> + while (1)
>> + /* NOP */;
>> +}
>> +
>> +#if defined(CONFIG_ARCH_CPU_INIT)
>> +int arch_cpu_init(void)
>> +{
>> + /*
>> + * It might be necessary to flush data cache, if U-boot is loaded
>> + * from kickstart bootloader, e.g. from S1L loader
>> + */
>> + flush_dcache_all();
>> +
>> + return 0;
>> +}
>> +#else
>> +#error "You have to select CONFIG_ARCH_CPU_INIT"
>> +#endif
>
> No other code ever has this kind of #error on CONFIG_ARCH_CPU_INIT not
> being defined. What's the point of it? People won't suddenly decide to
> undefine CONFIG_ARCH_CPU_INIT for fun, I think.
>
Generally for all future boards it's a good idea to clean up caches 
after a bootstrap/kickstart bootloader's work, which is expected to be 
present mostly on all LPC32XX boards.

The macro CONFIG_ARCH_CPU_INIT is defined by default in header 
include/asm/arch0lpc32xx/config.h, which is included in board 
configuration files for shared definitions, but of course technically 
for some of future boards it can be undefined after inclusion of the header.

Originally I didn't have such an #else case, however Marek convinced me 
to add the clause as a clear enough reminder about its importance for 
board developers.

If you think this #else case is redundant, I won't argue and it could be 
just removed.

>> diff --git a/arch/arm/include/asm/arch-lpc32xx/sys_proto.h
>> b/arch/arm/include/asm/arch-lpc32xx/sys_proto.h
>> new file mode 100644
>> index 0000000..b2cbfcd
>> --- /dev/null
>> +++ b/arch/arm/include/asm/arch-lpc32xx/sys_proto.h
>> @@ -0,0 +1,25 @@
>> +/*
>> + * Copyright (C) 2011 Vladimir Zapolskiy<vz at mleia.com>
>> + *
>> + * This program is free software; you can redistribute it and/or
>> + * modify it under the terms of the GNU General Public License
>> + * as published by the Free Software Foundation; either version 2
>> + * of the License, or (at your option) any later version.
>> + *
>> + * This program is distributed in the hope that it will be useful,
>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
>> + * GNU General Public License for more details.
>> + *
>> + * You should have received a copy of the GNU General Public License
>> + * along with this program; if not, write to the Free Software
>> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
>> + * MA 02110-1301, USA.
>> + */
>> +
>> +#ifndef _LPC32XX_SYS_PROTO_H
>> +#define _LPC32XX_SYS_PROTO_H
>> +
>> +void lpc32xx_uart_init(unsigned int uart_id);
>> +
>> +#endif /* _LPC32XX_SYS_PROTO_H */
>
> Do we need this .h for a single function when we already have uart.h?
>
I think we need it, I found that's quite a common practice to have such 
a header with SoC specific declarations for boards, and these headers 
have the same name. I expect that later on this header will contain SoC 
EMAC specific declarations and may be something else depending on U-boot 
support of other LPC32XX peripherals.

Thank you for review.

--
With best wishes,
Vladimir