[PATCH v3 2/3] ARM: static_vm: introduce an infrastructure for static mapped area

[Joonsoo Kim]

Hello, Nicolas.

On Tue, Jan 29, 2013 at 07:05:32PM -0500, Nicolas Pitre wrote:
> On Thu, 24 Jan 2013, Joonsoo Kim wrote:
> 
> > From: Joonsoo Kim <js1304 at gmail.com>
> > 
> > In current implementation, we used ARM-specific flag, that is,
> > VM_ARM_STATIC_MAPPING, for distinguishing ARM specific static mapped area.
> > The purpose of static mapped area is to re-use static mapped area when
> > entire physical address range of the ioremap request can be covered
> > by this area.
> > 
> > This implementation causes needless overhead for some cases.
> > For example, assume that there is only one static mapped area and
> > vmlist has 300 areas. Every time we call ioremap, we check 300 areas for
> > deciding whether it is matched or not. Moreover, even if there is
> > no static mapped area and vmlist has 300 areas, every time we call
> > ioremap, we check 300 areas in now.
> > 
> > If we construct a extra list for static mapped area, we can eliminate
> > above mentioned overhead.
> > With a extra list, if there is one static mapped area,
> > we just check only one area and proceed next operation quickly.
> > 
> > In fact, it is not a critical problem, because ioremap is not frequently
> > used. But reducing overhead is better idea.
> > 
> > Another reason for doing this work is for removing architecture dependency
> > on vmalloc layer. I think that vmlist and vmlist_lock is internal data
> > structure for vmalloc layer. Some codes for debugging and stat inevitably
> > use vmlist and vmlist_lock. But it is preferable that they are used
> > as least as possible in outside of vmalloc.c
> > 
> > Now, I introduce an ARM-specific infrastructure for static mapped area. In
> > the following patch, we will use this and resolve above mentioned problem.
> > 
> > Signed-off-by: Joonsoo Kim <js1304 at gmail.com>
> > Signed-off-by: Joonsoo Kim <iamjoonsoo.kim at lge.com>
> 
> First of all, I don't think you really need a new file with a global 
> scope header file.  Given that this code is meant to be used only for 
> ioremap optimization  on ARM, it is probably a better idea to simply put 
> it all into arch/arm/mm/ioremap.c instead.  The only function that needs 
> to be exported out of ioremap.c is insert_static_vm(), and only for the 
> benefit of arch/arm/mm/mmu.c, therefore this function prototype may as 
> well just be added to arch/arm/mm/mm.h.

I agree with your all opinions.
I will re-work and will re-send v4 as soon as possible.
Thanks for review.

> More comments below.
> 
> > diff --git a/arch/arm/include/asm/mach/static_vm.h b/arch/arm/include/asm/mach/static_vm.h
> > new file mode 100644
> > index 0000000..72c8339
> > --- /dev/null
> > +++ b/arch/arm/include/asm/mach/static_vm.h
> > @@ -0,0 +1,45 @@
> > +/*
> > + * arch/arm/include/asm/mach/static_vm.h
> > + *
> > + * Copyright (C) 2012 LG Electronics, Joonsoo Kim <iamjoonsoo.kim at lge.com>
> > + *
> > + * 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.
> > + *
> > + * 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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
> > + */
> > +
> > +#ifndef _ASM_MACH_STATIC_VM_H
> > +#define _ASM_MACH_STATIC_VM_H
> > +
> > +#include <linux/types.h>
> > +#include <linux/vmalloc.h>
> > +
> > +struct static_vm {
> > +	struct static_vm	*next;
> > +	void			*vaddr;
> > +	unsigned long		size;
> > +	unsigned long		flags;
> > +	phys_addr_t		paddr;
> > +	const void		*caller;
> > +};
> 
> Here you're duplicating most of the vm_struct content for no obvious 
> reasons.  Patch #3 even allocates both a vm_struct and a static_vm 
> instance in parallel for each mapping.  Instead, you should consider 
> something like this:
> 
> struct static_vm {
> 	struct static_vm *next;
> 	struct vm_struct vm;
> };
> 
> This way, you only need to allocate one structure:
> 
> 	struct static_vm *svm = early_alloc(...);
> 	...
> 	svm->vm.addr = addr;
> 	...
> 	vm_area_add_early(&svm->vm);
> 	insert_static_vm(svm);

Yes!
It's good idea.

> And then, it would make sense for the insert_static_vm() to do the 
> vm_area_add_early() call itself as well.

Okay.

> Maybe rename insert_static_vm() to static_vm_area_add_early() to better 
> identify its purpose as well.  It shouldn't be used for any other 
> purpose anyway.

Okay.

> > +
> > +extern struct static_vm *static_vmlist;
> > +extern spinlock_t static_vmlist_lock;
> 
> Your patch is providing the proper accessors to manipulate those.  They 
> therefore should not be exported globally.

Okay.

> > +
> > +extern struct static_vm *find_static_vm_paddr(phys_addr_t paddr,
> > +			size_t size, unsigned long flags);
> > +extern struct static_vm *find_static_vm_vaddr(void *vaddr);
> > +extern void init_static_vm(struct static_vm *static_vm,
> > +			struct vm_struct *vm, unsigned long flags);
> 
> Since those are only used in ioremap.c, and because I suggested their 
> implementation be moved there as well, you shouldn't need prototype 
> declarations anymore.  And init_static_vm() would be useless with my 
> previous suggestions.

Yes.

> 
> Nicolas
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/