[Qemu-devel] [PATCH v6 1/4] Modify DIRTY_FLAG value and introduce DIRTY_IDX to use as indexes of bit-based dirty bitmap.

[Yoshiaki Tamura]

Replaces byte-based dirty bitmap with four (MASTER, VGA, CODE,
MIGRATION) bit-based dirty bitmap.  On allocation, it sets all bits in
the bitmap.  It uses ffs() to convert DIRTY_FLAG to DIRTY_IDX.

Modifies wrapper functions for byte-based dirty bitmap to bit-based
dirty bitmap.  MASTER works as a buffer, and upon get_diry() or
get_dirty_flags(), it calls cpu_physical_memory_sync_master() to
update VGA and MIGRATION.

Signed-off-by: Yoshiaki Tamura <address@hidden>
---
 cpu-all.h     |  127 ++++++++++++++++++++++++++++++++++++++++++++++++---------
 exec.c        |   14 +++++--
 qemu-common.h |    3 +
 3 files changed, 120 insertions(+), 24 deletions(-)

diff --git a/cpu-all.h b/cpu-all.h
index e31c2de..fcccf6f 100644
--- a/cpu-all.h
+++ b/cpu-all.h
@@ -37,6 +37,9 @@
 
 #include "softfloat.h"
 
+/* to use ffs in flag_to_idx() */
+#include <strings.h>
+
 #if defined(HOST_WORDS_BIGENDIAN) != defined(TARGET_WORDS_BIGENDIAN)
 #define BSWAP_NEEDED
 #endif
@@ -861,6 +864,18 @@ target_phys_addr_t cpu_get_phys_page_debug(CPUState *env, 
target_ulong addr);
 extern int phys_ram_fd;
 extern ram_addr_t ram_size;
 
+/* Use DIRTY_IDX as indexes of bit-based phys_ram_dirty. */
+#define MASTER_DIRTY_IDX    0
+#define VGA_DIRTY_IDX       1
+#define CODE_DIRTY_IDX      2
+#define MIGRATION_DIRTY_IDX 3
+#define NUM_DIRTY_IDX       4
+
+#define MASTER_DIRTY_FLAG    (1 << MASTER_DIRTY_IDX)
+#define VGA_DIRTY_FLAG       (1 << VGA_DIRTY_IDX)
+#define CODE_DIRTY_FLAG      (1 << CODE_DIRTY_IDX)
+#define MIGRATION_DIRTY_FLAG (1 << MIGRATION_DIRTY_IDX)
+
 typedef struct RAMBlock {
     uint8_t *host;
     ram_addr_t offset;
@@ -869,7 +884,7 @@ typedef struct RAMBlock {
 } RAMBlock;
 
 typedef struct RAMList {
-    uint8_t *phys_dirty;
+    unsigned long *phys_dirty[NUM_DIRTY_IDX];
     ram_addr_t last_offset;
     QLIST_HEAD(ram, RAMBlock) blocks;
 } RAMList;
@@ -896,51 +911,123 @@ extern int mem_prealloc;
 /* Set if TLB entry is an IO callback.  */
 #define TLB_MMIO        (1 << 5)
 
-#define VGA_DIRTY_FLAG       0x01
-#define CODE_DIRTY_FLAG      0x02
-#define MIGRATION_DIRTY_FLAG 0x08
+static inline int dirty_flag_to_idx(int flag)
+{
+    return ffs(flag) - 1;
+}
+
+static inline int dirty_idx_to_flag(int idx)
+{
+    return 1 << idx;
+}
 
 /* read dirty bit (return 0 or 1) */
 static inline int cpu_physical_memory_is_dirty(ram_addr_t addr)
 {
-    return ram_list.phys_dirty[addr >> TARGET_PAGE_BITS] == 0xff;
+    unsigned long mask;
+    ram_addr_t index = (addr >> TARGET_PAGE_BITS) / HOST_LONG_BITS;
+    int offset = (addr >> TARGET_PAGE_BITS) & (HOST_LONG_BITS - 1);
+ 
+    mask = 1UL << offset;
+    return (ram_list.phys_dirty[MASTER_DIRTY_IDX][index] & mask) == mask;
+}
+
+static inline void cpu_physical_memory_sync_master(ram_addr_t index)
+{
+    if (ram_list.phys_dirty[MASTER_DIRTY_IDX][index]) {
+        ram_list.phys_dirty[VGA_DIRTY_IDX][index]
+            |=  ram_list.phys_dirty[MASTER_DIRTY_IDX][index];
+        ram_list.phys_dirty[MIGRATION_DIRTY_IDX][index]
+            |=  ram_list.phys_dirty[MASTER_DIRTY_IDX][index];
+        ram_list.phys_dirty[MASTER_DIRTY_IDX][index] = 0UL;
+    }
 }
 
 static inline int cpu_physical_memory_get_dirty_flags(ram_addr_t addr)
 {
-    return ram_list.phys_dirty[addr >> TARGET_PAGE_BITS];
+     unsigned long mask;
+     ram_addr_t index = (addr >> TARGET_PAGE_BITS) / HOST_LONG_BITS;
+     int offset = (addr >> TARGET_PAGE_BITS) & (HOST_LONG_BITS - 1);
+     int ret = 0, i;
+ 
+     mask = 1UL << offset;
+     cpu_physical_memory_sync_master(index);
+
+     for (i = VGA_DIRTY_IDX; i <= MIGRATION_DIRTY_IDX; i++) {
+         if (ram_list.phys_dirty[i][index] & mask) {
+             ret |= dirty_idx_to_flag(i);
+         }
+     }
+ 
+     return ret;
+}
+
+static inline int cpu_physical_memory_get_dirty_idx(ram_addr_t addr,
+                                                    int dirty_idx)
+{
+    unsigned long mask;
+    ram_addr_t index = (addr >> TARGET_PAGE_BITS) / HOST_LONG_BITS;
+    int offset = (addr >> TARGET_PAGE_BITS) & (HOST_LONG_BITS - 1);
+
+    mask = 1UL << offset;
+    cpu_physical_memory_sync_master(index);
+    return (ram_list.phys_dirty[dirty_idx][index] & mask) == mask;
 }
 
 static inline int cpu_physical_memory_get_dirty(ram_addr_t addr,
                                                 int dirty_flags)
 {
-    return ram_list.phys_dirty[addr >> TARGET_PAGE_BITS] & dirty_flags;
+    return cpu_physical_memory_get_dirty_idx(addr,
+                                             dirty_flag_to_idx(dirty_flags));
 }
 
 static inline void cpu_physical_memory_set_dirty(ram_addr_t addr)
 {
-    ram_list.phys_dirty[addr >> TARGET_PAGE_BITS] = 0xff;
+    unsigned long mask;
+    ram_addr_t index = (addr >> TARGET_PAGE_BITS) / HOST_LONG_BITS;
+    int offset = (addr >> TARGET_PAGE_BITS) & (HOST_LONG_BITS - 1);
+
+    mask = 1UL << offset;
+    ram_list.phys_dirty[MASTER_DIRTY_IDX][index] |= mask;
 }
 
-static inline int cpu_physical_memory_set_dirty_flags(ram_addr_t addr,
-                                                      int dirty_flags)
+static inline void cpu_physical_memory_set_dirty_range(ram_addr_t addr,
+                                                       unsigned long mask)
 {
-    return ram_list.phys_dirty[addr >> TARGET_PAGE_BITS] |= dirty_flags;
+    ram_addr_t index = (addr >> TARGET_PAGE_BITS) / HOST_LONG_BITS;
+
+    ram_list.phys_dirty[MASTER_DIRTY_IDX][index] |= mask;
+}
+
+static inline void cpu_physical_memory_set_dirty_flags(ram_addr_t addr,
+                                                       int dirty_flags)
+{
+    unsigned long mask;
+    ram_addr_t index = (addr >> TARGET_PAGE_BITS) / HOST_LONG_BITS;
+    int offset = (addr >> TARGET_PAGE_BITS) & (HOST_LONG_BITS - 1);
+
+    mask = 1UL << offset;
+    ram_list.phys_dirty[MASTER_DIRTY_IDX][index] |= mask;
+
+    if (dirty_flags & CODE_DIRTY_FLAG) {
+        ram_list.phys_dirty[CODE_DIRTY_IDX][index] |= mask;
+    }
 }
 
 static inline void cpu_physical_memory_mask_dirty_range(ram_addr_t start,
-                                                        int length,
+                                                        unsigned long length,
                                                         int dirty_flags)
 {
-    int i, mask, len;
-    uint8_t *p;
+    ram_addr_t addr = start, index;
+    unsigned long mask;
+    int offset, i;
 
-    len = length >> TARGET_PAGE_BITS;
-    mask = ~dirty_flags;
-    p = ram_list.phys_dirty + (start >> TARGET_PAGE_BITS);
-    for (i = 0; i < len; i++) {
-        p[i] &= mask;
-    }
+    for (i = 0;  i < length; i += TARGET_PAGE_SIZE) {
+        index = ((addr + i) >> TARGET_PAGE_BITS) / HOST_LONG_BITS;
+        offset = ((addr + i) >> TARGET_PAGE_BITS) & (HOST_LONG_BITS - 1);
+        mask = ~(1UL << offset);
+        ram_list.phys_dirty[dirty_flag_to_idx(dirty_flags)][index] &= mask;
+     }
 }
 
 void cpu_physical_memory_reset_dirty(ram_addr_t start, ram_addr_t end,
diff --git a/exec.c b/exec.c
index 7f64384..24f0f67 100644
--- a/exec.c
+++ b/exec.c
@@ -2768,6 +2768,7 @@ static void *file_ram_alloc(ram_addr_t memory, const char 
*path)
 ram_addr_t qemu_ram_alloc(ram_addr_t size)
 {
     RAMBlock *new_block;
+    int i;
 
     size = TARGET_PAGE_ALIGN(size);
     new_block = qemu_malloc(sizeof(*new_block));
@@ -2803,10 +2804,15 @@ ram_addr_t qemu_ram_alloc(ram_addr_t size)
 
     QLIST_INSERT_HEAD(&ram_list.blocks, new_block, next);
 
-    ram_list.phys_dirty = qemu_realloc(ram_list.phys_dirty,
-        (ram_list.last_offset + size) >> TARGET_PAGE_BITS);
-    memset(ram_list.phys_dirty + (ram_list.last_offset >> TARGET_PAGE_BITS),
-           0xff, size >> TARGET_PAGE_BITS);
+    for (i = MASTER_DIRTY_IDX; i < NUM_DIRTY_IDX; i++) {
+        ram_list.phys_dirty[i] 
+            = qemu_realloc(ram_list.phys_dirty[i],
+                           BITMAP_SIZE(ram_list.last_offset + size));
+        memset((uint8_t *)ram_list.phys_dirty[i] +
+               BITMAP_SIZE(ram_list.last_offset),
+               0xff, BITMAP_SIZE(ram_list.last_offset + size) -
+               BITMAP_SIZE(ram_list.last_offset));
+    }
 
     ram_list.last_offset += size;
 
diff --git a/qemu-common.h b/qemu-common.h
index d133f35..912d5e9 100644
--- a/qemu-common.h
+++ b/qemu-common.h
@@ -294,6 +294,9 @@ static inline uint8_t from_bcd(uint8_t val)
     return ((val >> 4) * 10) + (val & 0x0f);
 }
 
+#define ALIGN(x, y)  (((x)+(y)-1) & ~((y)-1))
+#define BITMAP_SIZE(m) (ALIGN(((m)>>TARGET_PAGE_BITS), HOST_LONG_BITS) / 8)
+
 #include "module.h"
 
 #endif /* dyngen-exec.h hack */
-- 
1.7.0.31.g1df487