Re: [PATCH 08/12] Allow xen platform pci device to be compiled asa module

[Stefano Stabellini]

On Tue, 18 May 2010, Jeremy Fitzhardinge wrote:
> > diff --git a/drivers/xen/grant-table.c b/drivers/xen/grant-table.c
> > index 6f5f3ba..f936d30 100644
> > --- a/drivers/xen/grant-table.c
> > +++ b/drivers/xen/grant-table.c
> > @@ -56,6 +56,9 @@
> >  #define GNTTAB_LIST_END 0xffffffff
> >  #define GREFS_PER_GRANT_FRAME (PAGE_SIZE / sizeof(struct grant_entry))
> >  
> > +unsigned long (*alloc_xen_mmio_hook)(unsigned long len);
> > +EXPORT_SYMBOL_GPL(alloc_xen_mmio_hook);
> > +
> >  static grant_ref_t **gnttab_list;
> >  static unsigned int nr_grant_frames;
> >  static unsigned int boot_max_nr_grant_frames;
> > @@ -514,7 +517,7 @@ int gnttab_resume(void)
> >  		return gnttab_map(0, nr_grant_frames - 1);
> >  
> >  	if (!hvm_pv_resume_frames) {
> > -		hvm_pv_resume_frames = alloc_xen_mmio(PAGE_SIZE * max_nr_gframes);
> > +		hvm_pv_resume_frames = alloc_xen_mmio_hook(PAGE_SIZE * max_nr_gframes);
> >   
> 
> This looks like it should be restructured so the pci device driver
> itself is doing this mapping, then calling into the grant subsystem to
> tell it where the mapping is.
> 

OK


> >  		shared = ioremap(hvm_pv_resume_frames, PAGE_SIZE * max_nr_gframes);
> >  		if (shared == NULL) {
> >  			printk(KERN_WARNING
> > @@ -600,6 +603,7 @@ int gnttab_init(void)
> >  	kfree(gnttab_list);
> >  	return -ENOMEM;
> >  }
> > +EXPORT_SYMBOL_GPL(gnttab_init);
> >  
> >  static int __devinit __gnttab_init(void)
> >  {
> > diff --git a/drivers/xen/manage.c b/drivers/xen/manage.c
> > index a73edd8..49ee52d 100644
> > --- a/drivers/xen/manage.c
> > +++ b/drivers/xen/manage.c
> > @@ -34,6 +34,13 @@ enum shutdown_state {
> >  	 SHUTDOWN_HALT = 4,
> >  };
> >  
> > +void (*platform_pci_resume_hook)(void);
> > +EXPORT_SYMBOL_GPL(platform_pci_resume_hook);
> > +void (*platform_pci_disable_irq_hook)(void);
> > +EXPORT_SYMBOL_GPL(platform_pci_disable_irq_hook);
> > +void (*platform_pci_enable_irq_hook)(void);
> > +EXPORT_SYMBOL_GPL(platform_pci_enable_irq_hook);
> >   
> 
> If all this _hook stuff is here to support a modular xen platform pci
> device, then something has gone wrong.  The device should be doing this
> via its own suspend/resume handlers.
> 

I'll give it another try.


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