qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Qemu-devel] Re: [PATCH 1/6] Make fw_cfg interface 32-bit aware


From: Alexander Graf
Subject: [Qemu-devel] Re: [PATCH 1/6] Make fw_cfg interface 32-bit aware
Date: Fri, 13 Nov 2009 07:15:16 +0100


On 13.11.2009, at 01:48, Glauber Costa wrote:

On Thu, Nov 12, 2009 at 09:53:10PM +0100, Alexander Graf wrote:
The fw_cfg interface can only handle up to 16 bits of data for its streams. While that isn't too much of a problem when handling integers, we would
like to stream full kernel images over that interface!

So let's extend it to 32 bit length variables.

Signed-off-by: Alexander Graf <address@hidden>

---

v1 -> v2:

 - add savevm compat code (untested!)
---
hw/fw_cfg.c |   30 ++++++++++++++++++++++++------
hw/fw_cfg.h |    2 +-
2 files changed, 25 insertions(+), 7 deletions(-)

diff --git a/hw/fw_cfg.c b/hw/fw_cfg.c
index a6d811b..0cd6f68 100644
--- a/hw/fw_cfg.c
+++ b/hw/fw_cfg.c
@@ -39,7 +39,7 @@
#define FW_CFG_SIZE 2

typedef struct _FWCfgEntry {
-    uint16_t len;
+    uint32_t len;
    uint8_t *data;
    void *callback_opaque;
    FWCfgCallback callback;
@@ -48,7 +48,7 @@ typedef struct _FWCfgEntry {
typedef struct _FWCfgState {
    FWCfgEntry entries[2][FW_CFG_MAX_ENTRY];
    uint16_t cur_entry;
-    uint16_t cur_offset;
+    uint32_t cur_offset;
} FWCfgState;

static void fw_cfg_write(FWCfgState *s, uint8_t value)
@@ -164,19 +164,37 @@ static void fw_cfg_reset(void *opaque)
    fw_cfg_select(s, 0);
}

+static int fw_cfg_load_old(QEMUFile *f, void *opaque, int version_id)
+{
+    FWCfgState *s = opaque;
+    uint16_t cur_offset;
+
+    if (version_id != 1)
+        return -EINVAL;
+
+    qemu_get_be16s(f, &s->cur_entry);
+
+    /* Convert old 16 bit value to new 32 bit width */
+    qemu_get_be16s(f, &cur_offset);
+    s->cur_offset = cur_offset;
+
+    return 0;
+}
+
static const VMStateDescription vmstate_fw_cfg = {
    .name = "fw_cfg",
-    .version_id = 1,
-    .minimum_version_id = 1,
+    .version_id = 2,
+    .minimum_version_id = 2,
    .minimum_version_id_old = 1,
+    .load_state_old = fw_cfg_load_old,
    .fields      = (VMStateField []) {
        VMSTATE_UINT16(cur_entry, FWCfgState),
-        VMSTATE_UINT16(cur_offset, FWCfgState),
+        VMSTATE_UINT32(cur_offset, FWCfgState),
        VMSTATE_END_OF_LIST()
    }
};

Why don't we just add another field for the upper bits, and add it through
VMSTATE_UINT16_V ?

Because that would mean I'd have to deal with it in the code later on and I don't see the point of writing code that's not in the load/save cycle because of limitations there.

Alex




reply via email to

[Prev in Thread] Current Thread [Next in Thread]