[LinuxBIOS] r418 - in LinuxBIOSv3: arch/x86 device include include/arch/x86 include/arch/x86/arch include/device lib util/dtc

[Uwe Hermann]

On Fri, Jun 29, 2007 at 06:57:23PM +0200, svn at openbios.org wrote:
> * start using arch/foo.h again instead of archfoo.h (trivial)
> * make constructor an initializer.

I don't understand. What's an initializer in this context?

 
> +	if(c && c->ops && c->ops->constructor)
> +		c->ops->constructor(dev, c);
> +	else
> +		printk(BIOS_INFO, "No constructor called for %s.\n", 
> +			dev_id_string(&c->id));

BIOS_DEBUG?


> Copied: LinuxBIOSv3/include/arch/x86/arch/elf.h (from rev 417, LinuxBIOSv3/include/arch/x86/archelf.h)
> ===================================================================
> --- LinuxBIOSv3/include/arch/x86/arch/elf.h	                        (rev 0)
> +++ LinuxBIOSv3/include/arch/x86/arch/elf.h	2007-06-29 16:57:23 UTC (rev 418)

Missing license header. Trivial file, so feel free to add yourself as
copyright owner.


> @@ -0,0 +1,8 @@
> +#ifndef ARCH_ELF_H
> +#define ARCH_ELF_H

Following our current guidelines this should be ARCH_X86_ARCH_ELF_H,
which looks a bit stupid, but I think it's necessary.

ARCH_ELF_H doesn't work as we'd have another ARCH_ELF_H if we add
another architecture...


> +
> +#define ELF_CLASS	ELFCLASS32
> +#define ELF_DATA	ELFDATA2LSB
> +#define ELF_ARCH	EM_386
> +
> +#endif /* ARCH_ELF_H */
> 
> Added: LinuxBIOSv3/include/arch/x86/arch/spinlock.h
> ===================================================================
> --- LinuxBIOSv3/include/arch/x86/arch/spinlock.h	                        (rev 0)
> +++ LinuxBIOSv3/include/arch/x86/arch/spinlock.h	2007-06-29 16:57:23 UTC (rev 418)
> @@ -0,0 +1,83 @@
> +/*
> + * This file is part of the LinuxBIOS project.
> + *
> + * Copyright (C) 2001 Linux Networx
> + *
> + * 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 should be GPLv2, I think. We know from email conversation with
Eric Biederman / Linux Networx that their code is GPLv2 only.

But this file is based on Linux code anyway, thus definately GPLv2,
I'll update the header and HACKING accordingly (later).


> + * 
> + * 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 St, Fifth Floor, Boston, MA 02110-1301 USA
> + */
> +
> +#ifndef ARCH_SPINLOCK_H
> +#define ARCH_SPINLOCK_H

See above.

> +
> +/*
> + * Your basic SMP spinlocks, allowing only a single CPU anywhere
> + */
> +
> +typedef struct {
> +	volatile unsigned int lock;

Why volatile? I think the Linux file doesn't have that.


> +} spinlock_t;

Drop spinlock_t and use 'struct spinlock' as we do for other structs?


> +
> +
> +#define SPIN_LOCK_UNLOCKED (spinlock_t) { 1 }
> +
> +/*
> + * Simple spin lock operations.  There are two variants, one clears IRQ's
> + * on the local processor, one does not.
> + *
> + * We make no fairness assumptions. They have a cost.
> + */
> +#define barrier() __asm__ __volatile__("": : :"memory")

#define barrier() (__asm__ __volatile__("": : :"memory"))

(better safe than sorry)


> +#define spin_is_locked(x)	(*(volatile char *)(&(x)->lock) <= 0)
> +#define spin_unlock_wait(x)	do { barrier(); } while(spin_is_locked(x))

#define spin_unlock_wait(x) (do { barrier(); } while(spin_is_locked(x)))


> +
> +#define spin_lock_string \
> +	"\n1:\t" \
> +	"lock ; decb %0\n\t" \
> +	"js 2f\n" \
> +	".section .text.lock,\"ax\"\n" \
> +	"2:\t" \
> +	"cmpb $0,%0\n\t" \
> +	"rep;nop\n\t" \
> +	"jle 2b\n\t" \
> +	"jmp 1b\n" \
> +	".previous"
> +
> +/*
> + * This works. Despite all the confusion.
> + */
> +#define spin_unlock_string \
> +	"movb $1,%0"

Shall we drop these #defines? They're only used in one place and a small
comment which says what the asm does should be enough.


> +static inline __attribute__((always_inline)) void spin_lock(spinlock_t *lock)
> +{
> +	__asm__ __volatile__(
> +		spin_lock_string
> +		:"=m" (lock->lock) : : "memory");
> +}
> +
> +static inline __attribute__((always_inline)) void spin_unlock(spinlock_t *lock)
> +{
> +	__asm__ __volatile__(
> +		spin_unlock_string
> +		:"=m" (lock->lock) : : "memory");
> +}
 

> Modified: LinuxBIOSv3/util/dtc/dtc-parser.y
> ===================================================================
> --- LinuxBIOSv3/util/dtc/dtc-parser.y	2007-06-29 15:32:19 UTC (rev 417)
> +++ LinuxBIOSv3/util/dtc/dtc-parser.y	2007-06-29 16:57:23 UTC (rev 418)
> @@ -142,7 +142,7 @@
>  		}
>  		')' ';' { 
>  				/* convention: first property is labeled with path */
> -				$6->label = strdup($3.val);
> +				$6->label = strdup((char *)$3.val);

Is this the correct fix for the warning? I didn't notice this warning
before so it must have been introduced by some recent change?


Uwe.
-- 
http://www.hermann-uwe.de  | http://www.holsham-traders.de
http://www.crazy-hacks.org | http://www.unmaintained-free-software.org
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://www.coreboot.org/pipermail/coreboot/attachments/20070630/a6b4ec10/attachment.sig>