Re: [PATCH v3 02/10] fsdax: Factor helper: dax_fault_actor()

[Ritesh Harjani <riteshh@linux.ibm.com>]

On 3/19/21 7:22 AM, Shiyang Ruan wrote:
> The core logic in the two dax page fault functions is similar. So, move
> the logic into a common helper function. Also, to facilitate the
> addition of new features, such as CoW, switch-case is no longer used to
> handle different iomap types.
> 
> Signed-off-by: Shiyang Ruan <ruansy.fnst@fujitsu.com>
> ---
>   fs/dax.c | 291 +++++++++++++++++++++++++++----------------------------
>   1 file changed, 145 insertions(+), 146 deletions(-)
> 
> diff --git a/fs/dax.c b/fs/dax.c
> index 7031e4302b13..33ddad0f3091 100644
> --- a/fs/dax.c
> +++ b/fs/dax.c
> @@ -1053,6 +1053,66 @@ static vm_fault_t dax_load_hole(struct xa_state *xas,
>   	return ret;
>   }
>   
> +#ifdef CONFIG_FS_DAX_PMD
> +static vm_fault_t dax_pmd_load_hole(struct xa_state *xas, struct vm_fault *vmf,
> +		struct iomap *iomap, void **entry)
> +{
> +	struct address_space *mapping = vmf->vma->vm_file->f_mapping;
> +	unsigned long pmd_addr = vmf->address & PMD_MASK;
> +	struct vm_area_struct *vma = vmf->vma;
> +	struct inode *inode = mapping->host;
> +	pgtable_t pgtable = NULL;
> +	struct page *zero_page;
> +	spinlock_t *ptl;
> +	pmd_t pmd_entry;
> +	pfn_t pfn;
> +
> +	zero_page = mm_get_huge_zero_page(vmf->vma->vm_mm);
> +
> +	if (unlikely(!zero_page))
> +		goto fallback;
> +
> +	pfn = page_to_pfn_t(zero_page);
> +	*entry = dax_insert_entry(xas, mapping, vmf, *entry, pfn,
> +			DAX_PMD | DAX_ZERO_PAGE, false);
> +
> +	if (arch_needs_pgtable_deposit()) {
> +		pgtable = pte_alloc_one(vma->vm_mm);
> +		if (!pgtable)
> +			return VM_FAULT_OOM;
> +	}
> +
> +	ptl = pmd_lock(vmf->vma->vm_mm, vmf->pmd);
> +	if (!pmd_none(*(vmf->pmd))) {
> +		spin_unlock(ptl);
> +		goto fallback;
> +	}
> +
> +	if (pgtable) {
> +		pgtable_trans_huge_deposit(vma->vm_mm, vmf->pmd, pgtable);
> +		mm_inc_nr_ptes(vma->vm_mm);
> +	}
> +	pmd_entry = mk_pmd(zero_page, vmf->vma->vm_page_prot);
> +	pmd_entry = pmd_mkhuge(pmd_entry);
> +	set_pmd_at(vmf->vma->vm_mm, pmd_addr, vmf->pmd, pmd_entry);
> +	spin_unlock(ptl);
> +	trace_dax_pmd_load_hole(inode, vmf, zero_page, *entry);
> +	return VM_FAULT_NOPAGE;
> +
> +fallback:
> +	if (pgtable)
> +		pte_free(vma->vm_mm, pgtable);
> +	trace_dax_pmd_load_hole_fallback(inode, vmf, zero_page, *entry);
> +	return VM_FAULT_FALLBACK;
> +}
> +#else
> +static vm_fault_t dax_pmd_load_hole(struct xa_state *xas, struct vm_fault *vmf,
> +		struct iomap *iomap, void **entry)
> +{
> +	return VM_FAULT_FALLBACK;
> +}
> +#endif /* CONFIG_FS_DAX_PMD */
> +
>   s64 dax_iomap_zero(loff_t pos, u64 length, struct iomap *iomap)
>   {
>   	sector_t sector = iomap_sector(iomap, pos & PAGE_MASK);
> @@ -1289,6 +1349,61 @@ static int dax_fault_cow_page(struct vm_fault *vmf, struct iomap *iomap,
>   	return 0;
>   }
>   
> +/**
> + * dax_fault_actor - Common actor to handle pfn insertion in PTE/PMD fault.
> + * @vmf:	vm fault instance
> + * @pfnp:	pfn to be returned
> + * @xas:	the dax mapping tree of a file
> + * @entry:	an unlocked dax entry to be inserted
> + * @pmd:	distinguish whether it is a pmd fault
> + * @flags:	iomap flags
> + * @iomap:	from iomap_begin()
> + * @srcmap:	from iomap_begin(), not equal to iomap if it is a CoW
> + */
> +static vm_fault_t dax_fault_actor(struct vm_fault *vmf, pfn_t *pfnp,
> +		struct xa_state *xas, void *entry, bool pmd, unsigned int flags,
> +		struct iomap *iomap, struct iomap *srcmap)
> +{
> +	struct address_space *mapping = vmf->vma->vm_file->f_mapping;
> +	size_t size = pmd ? PMD_SIZE : PAGE_SIZE;
> +	loff_t pos = (loff_t)xas->xa_offset << PAGE_SHIFT;

shouldn't we use xa_index here for pos ?
(loff_t)xas->xa_index << PAGE_SHIFT;

> +	bool write = vmf->flags & FAULT_FLAG_WRITE;
> +	bool sync = dax_fault_is_synchronous(flags, vmf->vma, iomap);
> +	int err = 0;
> +	pfn_t pfn;
> +
> +	/* if we are reading UNWRITTEN and HOLE, return a hole. */
> +	if (!write &&
> +	    (iomap->type == IOMAP_UNWRITTEN || iomap->type == IOMAP_HOLE)) {
> +		if (!pmd)
> +			return dax_load_hole(xas, mapping, &entry, vmf);
> +		else
> +			return dax_pmd_load_hole(xas, vmf, iomap, &entry);
> +	}
> +
> +	if (iomap->type != IOMAP_MAPPED) {
> +		WARN_ON_ONCE(1);
> +		return VM_FAULT_SIGBUS;
> +	}

So now in case if mapping is not mapped, we always cause
VM_FAULT_SIGBUG. But earlier we were only doing WARN_ON_ONCE(1).
Can you pls help answer why the change in behavior?




> +
> +	err = dax_iomap_pfn(iomap, pos, size, &pfn);
> +	if (err)
> +		return dax_fault_return(err);

Same case here as well. This could return SIGBUS while earlier I am not 
sure why were we only returning FALLBACK?


> +
> +	entry = dax_insert_entry(xas, mapping, vmf, entry, pfn, 0,
> +				 write && !sync);

In dax_insert_entry() we are passing 0 as flags.
We should be passing DAX_PMD/DAX_PTE no?


> +
> +	if (sync)
> +		return dax_fault_synchronous_pfnp(pfnp, pfn);
> +


/* handle PMD case here */
> +	if (pmd)
> +		return vmf_insert_pfn_pmd(vmf, pfn, write);

/* handle PTE case here */
> +	if (write)
> +		return vmf_insert_mixed_mkwrite(vmf->vma, vmf->address, pfn);
> +	else
> +		return vmf_insert_mixed(vmf->vma, vmf->address, pfn);
> +}

It is easy to miss the return from if(pmd) case while reading.
A comment like above could be helpful for code review.


> +
>   static vm_fault_t dax_iomap_pte_fault(struct vm_fault *vmf, pfn_t *pfnp,
>   			       int *iomap_errp, const struct iomap_ops *ops)
>   {
> @@ -1296,17 +1411,14 @@ static vm_fault_t dax_iomap_pte_fault(struct vm_fault *vmf, pfn_t *pfnp,
>   	struct address_space *mapping = vma->vm_file->f_mapping;
>   	XA_STATE(xas, &mapping->i_pages, vmf->pgoff);
>   	struct inode *inode = mapping->host;
> -	unsigned long vaddr = vmf->address;
>   	loff_t pos = (loff_t)vmf->pgoff << PAGE_SHIFT;
>   	struct iomap iomap = { .type = IOMAP_HOLE };
>   	struct iomap srcmap = { .type = IOMAP_HOLE };
>   	unsigned flags = IOMAP_FAULT;
>   	int error, major = 0;
>   	bool write = vmf->flags & FAULT_FLAG_WRITE;
> -	bool sync;
>   	vm_fault_t ret = 0;
>   	void *entry;
> -	pfn_t pfn;
>   
>   	trace_dax_pte_fault(inode, vmf, ret);
>   	/*
> @@ -1352,8 +1464,8 @@ static vm_fault_t dax_iomap_pte_fault(struct vm_fault *vmf, pfn_t *pfnp,
>   		goto unlock_entry;
>   	}
>   	if (WARN_ON_ONCE(iomap.offset + iomap.length < pos + PAGE_SIZE)) {
> -		error = -EIO;	/* fs corruption? */
> -		goto error_finish_iomap;
> +		ret = VM_FAULT_SIGBUS;	/* fs corruption? */
> +		goto finish_iomap;
>   	}
>   
>   	if (vmf->cow_page) {
> @@ -1363,49 +1475,19 @@ static vm_fault_t dax_iomap_pte_fault(struct vm_fault *vmf, pfn_t *pfnp,
>   		goto finish_iomap;
>   	}
>   
> -	sync = dax_fault_is_synchronous(flags, vma, &iomap);
> -
> -	switch (iomap.type) {
> -	case IOMAP_MAPPED:
> -		if (iomap.flags & IOMAP_F_NEW) {
> -			count_vm_event(PGMAJFAULT);
> -			count_memcg_event_mm(vma->vm_mm, PGMAJFAULT);
> -			major = VM_FAULT_MAJOR;
> -		}
> -		error = dax_iomap_pfn(&iomap, pos, PAGE_SIZE, &pfn);
> -		if (error < 0)
> -			goto error_finish_iomap;
> -
> -		entry = dax_insert_entry(&xas, mapping, vmf, entry, pfn,
> -						 0, write && !sync);
> -
> -		if (sync) {
> -			ret = dax_fault_synchronous_pfnp(pfnp, pfn);
> -			goto finish_iomap;
> -		}
> -		trace_dax_insert_mapping(inode, vmf, entry);
> -		if (write)
> -			ret = vmf_insert_mixed_mkwrite(vma, vaddr, pfn);
> -		else
> -			ret = vmf_insert_mixed(vma, vaddr, pfn);
> -
> +	ret = dax_fault_actor(vmf, pfnp, &xas, entry, false, flags,
> +			      &iomap, &srcmap);
> +	if (ret == VM_FAULT_SIGBUS)
>   		goto finish_iomap;
> -	case IOMAP_UNWRITTEN:
> -	case IOMAP_HOLE:
> -		if (!write) {
> -			ret = dax_load_hole(&xas, mapping, &entry, vmf);
> -			goto finish_iomap;
> -		}
> -		fallthrough;
> -	default:
> -		WARN_ON_ONCE(1);
> -		error = -EIO;
> -		break;
> +
> +	/* read/write MAPPED, CoW UNWRITTEN */
> +	if (iomap.flags & IOMAP_F_NEW) {
> +		count_vm_event(PGMAJFAULT);
> +		count_memcg_event_mm(vma->vm_mm, PGMAJFAULT);
> +		major = VM_FAULT_MAJOR;
>   	}

It is much better if above accounting is also done in dax_fault_actor()
function itself. Then at the end of this function we need to just do
"return ret"  instead of "return ret | major"


>   
> - error_finish_iomap:
> -	ret = dax_fault_return(error);
> - finish_iomap:
> +finish_iomap:
>   	if (ops->iomap_end) {
>   		int copied = PAGE_SIZE;
>   
> @@ -1419,66 +1501,14 @@ static vm_fault_t dax_iomap_pte_fault(struct vm_fault *vmf, pfn_t *pfnp,
>   		 */
>   		ops->iomap_end(inode, pos, PAGE_SIZE, copied, flags, &iomap);
>   	}
> - unlock_entry:
> +unlock_entry:
>   	dax_unlock_entry(&xas, entry);
> - out:
> +out:
>   	trace_dax_pte_fault_done(inode, vmf, ret);
>   	return ret | major;
>   }


-ritesh