Re: [RFT][PATCH 3/3] swsusp: improve freeing of memory

["Rafael J. Wysocki" <rjw@sisk.pl>]

Hi,

On Sunday, 13 of November 2005 22:14, Pavel Machek wrote:
> Hi!
> 
> > This patch makes swsusp free only as much memory as needed and not as much
> > as possible.
> 
> Looks okay to me. ACK, modulo few small things.
> 
> > -
> >  /* References to section boundaries */
> >  extern const void __nosave_begin, __nosave_end;
> >  
> >  extern unsigned int nr_copy_pages;
> > -extern suspend_pagedir_t *pagedir_nosave;
> > -extern suspend_pagedir_t *pagedir_save;
> > +extern struct pbe *pagedir_nosave;
> > +
> > +/*
> > + * This compilation switch determines the way in which memory will be freed
> > + * during suspend.  If defined, only as much memory will be freed as needed
> > + * to complete the suspend.  Otherwise, the largest possible amount of memory
> > + * will be freed.
> > + */
> > +#define OPPORTUNISTIC_SHRINKING		1
> 
> Can you use little less tabelators? Also shorter name for this one
> might be "FREE_ALL". 

OK

> > +/*
> > + * During suspend, on each attempt to free some more memory SHRINK_BITE
> > + * is used as the number of pages to free
> > + */
> > +#define SHRINK_BITE	10000
> 
> Does this really need this kind of visibility? There's nothing user
> should tweak here.

By setting this to a smaller value you can make swsusp free more memory
sometimes, but of course it need not be visible.  I'll move it to swsusp.c

> >  /**
> > + *	On resume it is necessary to trace and eventually free the unsafe
> > + *	pages that have been allocated, because they are needed for I/O
> > + *	(on x86-64 we likely will "eat" these pages once again while
> > + *	creating the temporary page translation tables)
> > + */
> > +
> > +struct eaten_page {
> > +	struct eaten_page	*next;
> > +	char			padding[PAGE_SIZE - sizeof(void *)];
> > +};
> 
> Less tabelators here, please...

OK

Greetings,
Rafael