[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20210208105530.3072869-7-ruansy.fnst@cn.fujitsu.com>
Date: Mon, 8 Feb 2021 18:55:25 +0800
From: Shiyang Ruan <ruansy.fnst@...fujitsu.com>
To: <linux-kernel@...r.kernel.org>, <linux-xfs@...r.kernel.org>,
<linux-nvdimm@...ts.01.org>, <linux-mm@...ck.org>,
<linux-fsdevel@...r.kernel.org>, <dm-devel@...hat.com>
CC: <darrick.wong@...cle.com>, <dan.j.williams@...el.com>,
<david@...morbit.com>, <hch@....de>, <agk@...hat.com>,
<snitzer@...hat.com>, <rgoldwyn@...e.de>, <qi.fuli@...itsu.com>,
<y-goto@...itsu.com>
Subject: [PATCH v3 06/11] mm, pmem: Implement ->memory_failure() in pmem driver
Call the ->memory_failure() which is implemented by pmem driver, in
order to finally notify filesystem to handle the corrupted data. The
handler which collects and kills processes are moved into
mf_dax_mapping_kill_procs(), which will be called by filesystem.
Keep the old handler in order to roll back if driver/device/filesystem
does not support ->memory_failure()/->corrupted_range().
Signed-off-by: Shiyang Ruan <ruansy.fnst@...fujitsu.com>
---
drivers/nvdimm/pmem.c | 26 +++++++++++
mm/memory-failure.c | 102 +++++++++++++++++++++++++-----------------
2 files changed, 87 insertions(+), 41 deletions(-)
diff --git a/drivers/nvdimm/pmem.c b/drivers/nvdimm/pmem.c
index 875076b0ea6c..c77c80e3d155 100644
--- a/drivers/nvdimm/pmem.c
+++ b/drivers/nvdimm/pmem.c
@@ -363,9 +363,35 @@ static void pmem_release_disk(void *__pmem)
put_disk(pmem->disk);
}
+static int pmem_pagemap_memory_failure(struct dev_pagemap *pgmap,
+ unsigned long pfn, int flags)
+{
+ struct pmem_device *pdev;
+ struct gendisk *disk;
+ loff_t disk_offset;
+ int rc = 0;
+ unsigned long size = page_size(pfn_to_page(pfn));
+
+ pdev = container_of(pgmap, struct pmem_device, pgmap);
+ disk = pdev->disk;
+ if (!disk)
+ return -ENXIO;
+
+ disk_offset = PFN_PHYS(pfn) - pdev->phys_addr - pdev->data_offset;
+ if (disk->fops->corrupted_range) {
+ rc = disk->fops->corrupted_range(disk, NULL, disk_offset,
+ size, &flags);
+ if (rc == -ENODEV)
+ rc = -ENXIO;
+ } else
+ rc = -EOPNOTSUPP;
+ return rc;
+}
+
static const struct dev_pagemap_ops fsdax_pagemap_ops = {
.kill = pmem_pagemap_kill,
.cleanup = pmem_pagemap_cleanup,
+ .memory_failure = pmem_pagemap_memory_failure,
};
static int pmem_attach_disk(struct device *dev,
diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index 158fe0c8e602..670e29cd263e 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -1219,6 +1219,54 @@ static int try_to_split_thp_page(struct page *page, const char *msg)
return 0;
}
+int mf_generic_kill_procs(unsigned long long pfn, int flags)
+{
+ struct page *page = pfn_to_page(pfn);
+ const bool unmap_success = true;
+ unsigned long size = 0;
+ struct to_kill *tk;
+ LIST_HEAD(to_kill);
+ loff_t start;
+ dax_entry_t cookie;
+
+ /*
+ * Prevent the inode from being freed while we are interrogating
+ * the address_space, typically this would be handled by
+ * lock_page(), but dax pages do not use the page lock. This
+ * also prevents changes to the mapping of this pfn until
+ * poison signaling is complete.
+ */
+ cookie = dax_lock_page(page);
+ if (!cookie)
+ return -EBUSY;
+ /*
+ * Unlike System-RAM there is no possibility to swap in a
+ * different physical page at a given virtual address, so all
+ * userspace consumption of ZONE_DEVICE memory necessitates
+ * SIGBUS (i.e. MF_MUST_KILL)
+ */
+ flags |= MF_ACTION_REQUIRED | MF_MUST_KILL;
+ collect_procs(page, &to_kill, flags & MF_ACTION_REQUIRED);
+
+ list_for_each_entry(tk, &to_kill, nd)
+ if (tk->size_shift)
+ size = max(size, 1UL << tk->size_shift);
+ if (size) {
+ /*
+ * Unmap the largest mapping to avoid breaking up
+ * device-dax mappings which are constant size. The
+ * actual size of the mapping being torn down is
+ * communicated in siginfo, see kill_proc()
+ */
+ start = (page->index << PAGE_SHIFT) & ~(size - 1);
+ unmap_mapping_range(page->mapping, start, start + size, 0);
+ }
+ kill_procs(&to_kill, flags & MF_MUST_KILL, !unmap_success, pfn, flags);
+
+ dax_unlock_page(page, cookie);
+ return 0;
+}
+
int mf_dax_mapping_kill_procs(struct address_space *mapping, pgoff_t index, int flags)
{
const bool unmap_success = true;
@@ -1343,13 +1391,7 @@ static int memory_failure_dev_pagemap(unsigned long pfn, int flags,
struct dev_pagemap *pgmap)
{
struct page *page = pfn_to_page(pfn);
- const bool unmap_success = true;
- unsigned long size = 0;
- struct to_kill *tk;
- LIST_HEAD(to_kill);
int rc = -EBUSY;
- loff_t start;
- dax_entry_t cookie;
if (flags & MF_COUNT_INCREASED)
/*
@@ -1357,20 +1399,9 @@ static int memory_failure_dev_pagemap(unsigned long pfn, int flags,
*/
put_page(page);
- /*
- * Prevent the inode from being freed while we are interrogating
- * the address_space, typically this would be handled by
- * lock_page(), but dax pages do not use the page lock. This
- * also prevents changes to the mapping of this pfn until
- * poison signaling is complete.
- */
- cookie = dax_lock_page(page);
- if (!cookie)
- goto out;
-
if (hwpoison_filter(page)) {
rc = 0;
- goto unlock;
+ goto out;
}
if (pgmap->type == MEMORY_DEVICE_PRIVATE) {
@@ -1378,7 +1409,7 @@ static int memory_failure_dev_pagemap(unsigned long pfn, int flags,
* TODO: Handle HMM pages which may need coordination
* with device-side memory.
*/
- goto unlock;
+ goto out;
}
/*
@@ -1388,32 +1419,21 @@ static int memory_failure_dev_pagemap(unsigned long pfn, int flags,
SetPageHWPoison(page);
/*
- * Unlike System-RAM there is no possibility to swap in a
- * different physical page at a given virtual address, so all
- * userspace consumption of ZONE_DEVICE memory necessitates
- * SIGBUS (i.e. MF_MUST_KILL)
+ * Call driver's implementation to handle the memory failure,
+ * otherwise roll back to generic handler.
*/
- flags |= MF_ACTION_REQUIRED | MF_MUST_KILL;
- collect_procs_file(page, page->mapping, page->index, &to_kill,
- flags & MF_ACTION_REQUIRED);
-
- list_for_each_entry(tk, &to_kill, nd)
- if (tk->size_shift)
- size = max(size, 1UL << tk->size_shift);
- if (size) {
+ if (pgmap->ops->memory_failure) {
+ rc = pgmap->ops->memory_failure(pgmap, pfn, flags);
/*
- * Unmap the largest mapping to avoid breaking up
- * device-dax mappings which are constant size. The
- * actual size of the mapping being torn down is
- * communicated in siginfo, see kill_proc()
+ * Roll back to generic handler too if operation is not
+ * supported inside the driver/device/filesystem.
*/
- start = (page->index << PAGE_SHIFT) & ~(size - 1);
- unmap_mapping_range(page->mapping, start, start + size, 0);
+ if (rc != EOPNOTSUPP)
+ goto out;
}
- kill_procs(&to_kill, flags & MF_MUST_KILL, !unmap_success, pfn, flags);
- rc = 0;
-unlock:
- dax_unlock_page(page, cookie);
+
+ rc = mf_generic_kill_procs(pfn, flags);
+
out:
/* drop pgmap ref acquired in caller */
put_dev_pagemap(pgmap);
--
2.30.0
Powered by blists - more mailing lists