dax: enable dax fault handler to report VM_FAULT_HWPOISON

When multiple processes mmap() a dax file, then at some point,
a process issues a 'load' and consumes a hwpoison, the process
receives a SIGBUS with si_code = BUS_MCEERR_AR and with si_lsb
set for the poison scope. Soon after, any other process issues
a 'load' to the poisoned page (that is unmapped from the kernel
side by memory_failure), it receives a SIGBUS with
si_code = BUS_ADRERR and without valid si_lsb.

This is confusing to user, and is different from page fault due
to poison in RAM memory, also some helpful information is lost.

Channel dax backend driver's poison detection to the filesystem
such that instead of reporting VM_FAULT_SIGBUS, it could report
VM_FAULT_HWPOISON.

If user level block IO syscalls fail due to poison, the errno will
be converted to EIO to maintain block API consistency.

Signed-off-by: Jane Chu <jane.chu@oracle.com>
Link: https://lore.kernel.org/r/20230615181325.1327259-2-jane.chu@oracle.com
Reviewed-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>

diff --git a/drivers/dax/super.c b/drivers/dax/super.c
index c4c4728a36e43380e0fc122139fcee2566bcf382..0da9232ea1754b862455fe6d2d1c0e57398cfde1 100644 (file)
--- a/drivers/dax/super.c
+++ b/drivers/dax/super.c
@@ -203,6 +203,8 @@ size_t dax_copy_to_iter(struct dax_device *dax_dev, pgoff_t pgoff, void *addr,
 int dax_zero_page_range(struct dax_device *dax_dev, pgoff_t pgoff,
                        size_t nr_pages)
 {
+       int ret;
+
        if (!dax_alive(dax_dev))
                return -ENXIO;
        /*
@@ -213,7 +215,8 @@ int dax_zero_page_range(struct dax_device *dax_dev, pgoff_t pgoff,
        if (nr_pages != 1)
                return -EIO;
 
-       return dax_dev->ops->zero_page_range(dax_dev, pgoff, nr_pages);
+       ret = dax_dev->ops->zero_page_range(dax_dev, pgoff, nr_pages);
+       return dax_mem2blk_err(ret);
 }
 EXPORT_SYMBOL_GPL(dax_zero_page_range);
 

diff --git a/drivers/nvdimm/pmem.c b/drivers/nvdimm/pmem.c
index ceea55f621cc7fdc5d4fbccebe8203bd8384be40..46e094e56159f651ca95dc09c95e4d9d5c88b51e 100644 (file)
--- a/drivers/nvdimm/pmem.c
+++ b/drivers/nvdimm/pmem.c
@@ -260,7 +260,7 @@ __weak long __pmem_direct_access(struct pmem_device *pmem, pgoff_t pgoff,
                long actual_nr;
 
                if (mode != DAX_RECOVERY_WRITE)
-                       return -EIO;
+                       return -EHWPOISON;
 
                /*
                 * Set the recovery stride is set to kernel page size because

diff --git a/drivers/s390/block/dcssblk.c b/drivers/s390/block/dcssblk.c
index c09f2e053bf86358f303927637fc2c6fffb8bbf6..ee47ac520cd45bb64c5824026848bcf70310b265 100644 (file)
--- a/drivers/s390/block/dcssblk.c
+++ b/drivers/s390/block/dcssblk.c
@@ -54,7 +54,8 @@ static int dcssblk_dax_zero_page_range(struct dax_device *dax_dev,
        rc = dax_direct_access(dax_dev, pgoff, nr_pages, DAX_ACCESS,
                        &kaddr, NULL);
        if (rc < 0)
-               return rc;
+               return dax_mem2blk_err(rc);
+
        memset(kaddr, 0, nr_pages << PAGE_SHIFT);
        dax_flush(dax_dev, kaddr, nr_pages << PAGE_SHIFT);
        return 0;

diff --git a/fs/dax.c b/fs/dax.c
index cb36c6746fc4d39dd58d6eee5de015d78d47661b..906ecbd541a3de074f784a77cdeeef73f49ace0a 100644 (file)
--- a/fs/dax.c
+++ b/fs/dax.c
@@ -1148,7 +1148,7 @@ static int dax_iomap_copy_around(loff_t pos, uint64_t length, size_t align_size,
        if (!zero_edge) {
                ret = dax_iomap_direct_access(srcmap, pos, size, &saddr, NULL);
                if (ret)
-                       return ret;
+                       return dax_mem2blk_err(ret);
        }
 
        if (copy_all) {
@@ -1310,7 +1310,7 @@ static s64 dax_unshare_iter(struct iomap_iter *iter)
 
 out_unlock:
        dax_read_unlock(id);
-       return ret;
+       return dax_mem2blk_err(ret);
 }
 
 int dax_file_unshare(struct inode *inode, loff_t pos, loff_t len,
@@ -1342,7 +1342,8 @@ static int dax_memzero(struct iomap_iter *iter, loff_t pos, size_t size)
        ret = dax_direct_access(iomap->dax_dev, pgoff, 1, DAX_ACCESS, &kaddr,
                                NULL);
        if (ret < 0)
-               return ret;
+               return dax_mem2blk_err(ret);
+
        memset(kaddr + offset, 0, size);
        if (iomap->flags & IOMAP_F_SHARED)
                ret = dax_iomap_copy_around(pos, size, PAGE_SIZE, srcmap,
@@ -1498,7 +1499,7 @@ static loff_t dax_iomap_iter(const struct iomap_iter *iomi,
 
                map_len = dax_direct_access(dax_dev, pgoff, PHYS_PFN(size),
                                DAX_ACCESS, &kaddr, NULL);
-               if (map_len == -EIO && iov_iter_rw(iter) == WRITE) {
+               if (map_len == -EHWPOISON && iov_iter_rw(iter) == WRITE) {
                        map_len = dax_direct_access(dax_dev, pgoff,
                                        PHYS_PFN(size), DAX_RECOVERY_WRITE,
                                        &kaddr, NULL);
@@ -1506,7 +1507,7 @@ static loff_t dax_iomap_iter(const struct iomap_iter *iomi,
                                recovery = true;
                }
                if (map_len < 0) {
-                       ret = map_len;
+                       ret = dax_mem2blk_err(map_len);
                        break;
                }
 

diff --git a/fs/fuse/virtio_fs.c b/fs/fuse/virtio_fs.c
index 4d8d4f16c727b9d607dc884312d7663e6cfd4b78..5f1be1da92ce942cc1c8828bed888296a87c4d4e 100644 (file)
--- a/fs/fuse/virtio_fs.c
+++ b/fs/fuse/virtio_fs.c
@@ -775,7 +775,8 @@ static int virtio_fs_zero_page_range(struct dax_device *dax_dev,
        rc = dax_direct_access(dax_dev, pgoff, nr_pages, DAX_ACCESS, &kaddr,
                               NULL);
        if (rc < 0)
-               return rc;
+               return dax_mem2blk_err(rc);
+
        memset(kaddr, 0, nr_pages << PAGE_SHIFT);
        dax_flush(dax_dev, kaddr, nr_pages << PAGE_SHIFT);
        return 0;

diff --git a/include/linux/dax.h b/include/linux/dax.h
index bf6258472e495857d33aed314ba8fa6b6f15ac2b..261944ec0887ca8e6ce2599b7bcf7db171e42b7b 100644 (file)
--- a/include/linux/dax.h
+++ b/include/linux/dax.h
@@ -261,6 +261,19 @@ static inline bool dax_mapping(struct address_space *mapping)
        return mapping->host && IS_DAX(mapping->host);
 }
 
+/*
+ * Due to dax's memory and block duo personalities, hwpoison reporting
+ * takes into consideration which personality is presently visible.
+ * When dax acts like a block device, such as in block IO, an encounter of
+ * dax hwpoison is reported as -EIO.
+ * When dax acts like memory, such as in page fault, a detection of hwpoison
+ * is reported as -EHWPOISON which leads to VM_FAULT_HWPOISON.
+ */
+static inline int dax_mem2blk_err(int err)
+{
+       return (err == -EHWPOISON) ? -EIO : err;
+}
+
 #ifdef CONFIG_DEV_DAX_HMEM_DEVICES
 void hmem_register_resource(int target_nid, struct resource *r);
 #else

diff --git a/include/linux/mm.h b/include/linux/mm.h
index 27ce77080c79c7a026e641e491246fcf6f7e26c0..052ac9317365e7d871227b822fd4ca4a64162620 100644 (file)
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -3342,6 +3342,8 @@ static inline vm_fault_t vmf_error(int err)
 {
        if (err == -ENOMEM)
                return VM_FAULT_OOM;
+       else if (err == -EHWPOISON)
+               return VM_FAULT_HWPOISON;
        return VM_FAULT_SIGBUS;
 }