Re: [PATCH v8 10/22] Replace the XIP page fault handler with the DAX page fault handler

[Matthew Wilcox <willy@linux.intel.com>]

On Wed, Jul 23, 2014 at 03:10:25PM +0300, Kirill A. Shutemov wrote:
> > +int dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf,
> > +			get_block_t get_block)
> > +{
> > +	int result;
> > +	struct super_block *sb = file_inode(vma->vm_file)->i_sb;
> > +
> > +	if (vmf->flags & FAULT_FLAG_WRITE) {
> 
> Nobody seems calls sb_start_pagefault() in fault handler.
> Do you mean FAULT_FLAG_MKWRITE?

We need to call sb_start_pagefault() if we're going to make a modification
to the filesystem.  Admittedly, we don't know if we're going to make a
modification at this point, but if we're taking a write fault on a hole,
we will be.  We can skip the call to sb_start_pagefault() if we're taking
a read fault.

> > +int dax_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf,
> > +			get_block_t get_block)
> > +{
> > +	return dax_fault(vma, vmf, get_block);
> > +}
> > +EXPORT_SYMBOL_GPL(dax_mkwrite);
> 
> I don't think we want to introduce new exported symbol just for dummy
> wrapper. Just use ".page_mkwrite = foo_fault,". perf and selinux do
> this.
> Or add #define into header file if you want better readability.

They were different at one time ... agreed, I can just make them an alias
for now.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>