[Qemu-devel] [PATCH 08/16] nvdimm: init backend memory mapping and config data area

[Xiao Guangrong]

The parameter @file is used as backed memory for NVDIMM which is
divided into two parts:
- first parts is (0, size - 128K], which is used as PMEM (Persistent
  Memory)
- 128K at the end of the file, which is used as Config Data Area, it's
  used to store Label namespace data

The @file supports both regular file and block device, of course we
can assign any these two kinds of files for test and emulation, however,
in the real word for performance reason, we usually used these files as
NVDIMM backed file:
- the regular file in the filesystem with DAX enabled created on NVDIMM
  device on host
- the raw PMEM device on host, e,g /dev/pmem0

Signed-off-by: Xiao Guangrong <address@hidden>
---
 hw/mem/pc-nvdimm.c         | 102 ++++++++++++++++++++++++++++++++++++++++++++-
 include/hw/mem/pc-nvdimm.h |   5 +++
 2 files changed, 106 insertions(+), 1 deletion(-)

diff --git a/hw/mem/pc-nvdimm.c b/hw/mem/pc-nvdimm.c
index b40d4e7..9531935 100644
--- a/hw/mem/pc-nvdimm.c
+++ b/hw/mem/pc-nvdimm.c
@@ -22,12 +22,20 @@
  * License along with this library; if not, see <http://www.gnu.org/licenses/>
  */
 
+#include <sys/mman.h>
+#include <sys/ioctl.h>
+#include <linux/fs.h>
+
+#include "exec/address-spaces.h"
 #include "hw/mem/pc-nvdimm.h"
 
-#define PAGE_SIZE      (1UL << 12)
+#define PAGE_SIZE               (1UL << 12)
+
+#define MIN_CONFIG_DATA_SIZE    (128 << 10)
 
 static struct nvdimms_info {
     ram_addr_t current_addr;
+    int device_index;
 } nvdimms_info;
 
 /* the address range [offset, ~0ULL) is reserved for NVDIMM. */
@@ -37,6 +45,26 @@ void pc_nvdimm_reserve_range(ram_addr_t offset)
     nvdimms_info.current_addr = offset;
 }
 
+static ram_addr_t reserved_range_push(uint64_t size)
+{
+    uint64_t current;
+
+    current = ROUND_UP(nvdimms_info.current_addr, PAGE_SIZE);
+
+    /* do not have enough space? */
+    if (current + size < current) {
+        return 0;
+    }
+
+    nvdimms_info.current_addr = current + size;
+    return current;
+}
+
+static uint32_t new_device_index(void)
+{
+    return nvdimms_info.device_index++;
+}
+
 static char *get_file(Object *obj, Error **errp)
 {
     PCNVDIMMDevice *nvdimm = PC_NVDIMM(obj);
@@ -48,6 +76,11 @@ static void set_file(Object *obj, const char *str, Error 
**errp)
 {
     PCNVDIMMDevice *nvdimm = PC_NVDIMM(obj);
 
+    if (memory_region_size(&nvdimm->mr)) {
+        error_setg(errp, "cannot change property value");
+        return;
+    }
+
     if (nvdimm->file) {
         g_free(nvdimm->file);
     }
@@ -60,13 +93,80 @@ static void pc_nvdimm_init(Object *obj)
     object_property_add_str(obj, "file", get_file, set_file, NULL);
 }
 
+static uint64_t get_file_size(int fd)
+{
+    struct stat stat_buf;
+    uint64_t size;
+
+    if (fstat(fd, &stat_buf) < 0) {
+        return 0;
+    }
+
+    if (S_ISREG(stat_buf.st_mode)) {
+        return stat_buf.st_size;
+    }
+
+    if (S_ISBLK(stat_buf.st_mode) && !ioctl(fd, BLKGETSIZE64, &size)) {
+        return size;
+    }
+
+    return 0;
+}
+
 static void pc_nvdimm_realize(DeviceState *dev, Error **errp)
 {
     PCNVDIMMDevice *nvdimm = PC_NVDIMM(dev);
+    char name[512];
+    void *buf;
+    ram_addr_t addr;
+    uint64_t size;
+    int fd;
 
     if (!nvdimm->file) {
         error_setg(errp, "file property is not set");
     }
+
+    fd = open(nvdimm->file, O_RDWR);
+    if (fd < 0) {
+        error_setg(errp, "can not open %s", nvdimm->file);
+        return;
+    }
+
+    /* reserve MIN_CONFIGDATA_AREA_SIZE for configue data */
+    size = get_file_size(fd) - MIN_CONFIG_DATA_SIZE;
+    if ((int64_t)size <= 0) {
+        error_setg(errp, "file size is too small to store NVDIMM"
+                         " configure data");
+        goto do_close;
+    }
+
+    buf = mmap(NULL, size + MIN_CONFIG_DATA_SIZE, PROT_READ | PROT_WRITE,
+               MAP_SHARED, fd, 0);
+    if (buf == MAP_FAILED) {
+        error_setg(errp, "can not do mmap on %s", nvdimm->file);
+        goto do_close;
+    }
+
+    addr = reserved_range_push(size);
+    if (!addr) {
+        error_setg(errp, "do not have enough space for size %#lx.\n", size);
+        goto do_unmap;
+    }
+
+    nvdimm->device_index = new_device_index();
+    sprintf(name, "NVDIMM-%d", nvdimm->device_index);
+    memory_region_init_ram_ptr(&nvdimm->mr, OBJECT(dev), name, size, buf);
+    vmstate_register_ram(&nvdimm->mr, DEVICE(dev));
+    memory_region_add_subregion(get_system_memory(), addr, &nvdimm->mr);
+
+    nvdimm->config_data_addr = buf + size;
+    nvdimm->config_data_size = MIN_CONFIG_DATA_SIZE;
+
+    return;
+do_unmap:
+    munmap(buf, size);
+do_close:
+    close(fd);
 }
 
 static void pc_nvdimm_class_init(ObjectClass *oc, void *data)
diff --git a/include/hw/mem/pc-nvdimm.h b/include/hw/mem/pc-nvdimm.h
index 2081e7c..e743ed1 100644
--- a/include/hw/mem/pc-nvdimm.h
+++ b/include/hw/mem/pc-nvdimm.h
@@ -21,6 +21,11 @@ typedef struct PCNVDIMMDevice {
     DeviceState parent_obj;
 
     char *file;
+    void *config_data_addr;
+    uint64_t config_data_size;
+
+    int device_index;
+    MemoryRegion mr;
 } PCNVDIMMDevice;
 
 #define TYPE_PC_NVDIMM "pc-nvdimm"
-- 
2.1.0