The patch titled
Subject: mm-dax-make-pmd_fault-and-friends-to-be-the-same-as-fault-v7
has been removed from the -mm tree. Its filename was
mm-dax-make-pmd_fault-and-friends-to-be-the-same-as-fault-v7.patch
This patch was dropped because it was folded into mm-dax-make-pmd_fault-and-friends-to-be-the-same-as-fault.patch
------------------------------------------------------
From: Dave Jiang <dave.jiang@xxxxxxxxx>
Subject: mm-dax-make-pmd_fault-and-friends-to-be-the-same-as-fault-v7
v7:
Xiong found issue with xfs_tests stall when DAX option is turned out. It
seems the issue has to do with dax_iomap_pmd_fault(). The pgoff needs
to be linear_page_index(vma, vmf->address & PMD_MASK) in this funciton.
However, the passed in address is linear_page_index(vma, vmf->address)
and based on the PTE and not the PMD. The vmf->pgoff needs to be what was
passed in when returned. Therefore we go back to having a local variable
of pgoff in the dax_iomap_pmd_fault that is based on address & PMD_MASK.
Link: http://lkml.kernel.org/r/148469861071.195597.3619476895250028518.stgit@xxxxxxxxxxxxxxxxxxxxxxxxxx
Signed-off-by: Dave Jiang <dave.jiang@xxxxxxxxx>
Reviewed-by: Ross Zwisler <ross.zwisler@xxxxxxxxxxxxxxx>
Reviewed-by: Jan Kara <jack@xxxxxxx>
Cc: Dave Chinner <david@xxxxxxxxxxxxx>
Cc: Matthew Wilcox <mawilcox@xxxxxxxxxxxxx>
Cc: Steven Rostedt <rostedt@xxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
fs/dax.c | 14 +++++++-------
1 file changed, 7 insertions(+), 7 deletions(-)
diff -puN fs/dax.c~mm-dax-make-pmd_fault-and-friends-to-be-the-same-as-fault-v7 fs/dax.c
--- a/fs/dax.c~mm-dax-make-pmd_fault-and-friends-to-be-the-same-as-fault-v7
+++ a/fs/dax.c
@@ -1350,7 +1350,7 @@ int dax_iomap_pmd_fault(struct vm_area_s
struct inode *inode = mapping->host;
int result = VM_FAULT_FALLBACK;
struct iomap iomap = { 0 };
- pgoff_t max_pgoff;
+ pgoff_t max_pgoff, pgoff;
void *entry;
loff_t pos;
int error;
@@ -1360,7 +1360,7 @@ int dax_iomap_pmd_fault(struct vm_area_s
* supposed to hold locks serializing us with truncate / punch hole so
* this is a reliable test.
*/
- vmf->pgoff = linear_page_index(vma, pmd_addr);
+ pgoff = linear_page_index(vma, pmd_addr);
max_pgoff = (i_size_read(inode) - 1) >> PAGE_SHIFT;
trace_dax_pmd_fault(inode, vma, vmf, max_pgoff, 0);
@@ -1375,13 +1375,13 @@ int dax_iomap_pmd_fault(struct vm_area_s
if ((pmd_addr + PMD_SIZE) > vma->vm_end)
goto fallback;
- if (vmf->pgoff > max_pgoff) {
+ if (pgoff > max_pgoff) {
result = VM_FAULT_SIGBUS;
goto out;
}
/* If the PMD would extend beyond the file size */
- if ((vmf->pgoff | PG_PMD_COLOUR) > max_pgoff)
+ if ((pgoff | PG_PMD_COLOUR) > max_pgoff)
goto fallback;
/*
@@ -1389,7 +1389,7 @@ int dax_iomap_pmd_fault(struct vm_area_s
* setting up a mapping, so really we're using iomap_begin() as a way
* to look up our filesystem block.
*/
- pos = (loff_t)vmf->pgoff << PAGE_SHIFT;
+ pos = (loff_t)pgoff << PAGE_SHIFT;
error = ops->iomap_begin(inode, pos, PMD_SIZE, iomap_flags, &iomap);
if (error)
goto fallback;
@@ -1403,7 +1403,7 @@ int dax_iomap_pmd_fault(struct vm_area_s
* the tree, for instance), it will return -EEXIST and we just fall
* back to 4k entries.
*/
- entry = grab_mapping_entry(mapping, vmf->pgoff, RADIX_DAX_PMD);
+ entry = grab_mapping_entry(mapping, pgoff, RADIX_DAX_PMD);
if (IS_ERR(entry))
goto finish_iomap;
@@ -1425,7 +1425,7 @@ int dax_iomap_pmd_fault(struct vm_area_s
}
unlock_entry:
- put_locked_mapping_entry(mapping, vmf->pgoff, entry);
+ put_locked_mapping_entry(mapping, pgoff, entry);
finish_iomap:
if (ops->iomap_end) {
int copied = PMD_SIZE;
_
Patches currently in -mm which might be from dave.jiang@xxxxxxxxx are
mm-dax-make-pmd_fault-and-friends-to-be-the-same-as-fault.patch
mm-dax-move-pmd_fault-to-take-only-vmf-parameter.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
