[Qemu-devel] [PATCH 1/9] dma-helpers: Expose the sg mapping logic

[Alex Pyrgiotis <apyrgio@arrikto.com>]

The mapping of scatter-gather lists from physical addresses (as
perceived by the guest kernel) to the virtual address space of the QEMU
process is a vital step for a DMA operation. This step is currently
implemented, amongst other things, in dma_blk_cb(), making it impossible
to be used by anyone else.

In order to pave the way for the DMA support of ioctl commands, expose
the aforementioned logic in a separate function called "dma_map_sg".
Also, expose some other important pieces too, such as the initialization
of the dbs structure.

Signed-off-by: Alex Pyrgiotis <apyrgio@arrikto.com>
Signed-off-by: Dimitris Aragiorgis <dimara@arrikto.com>

diff --git a/dma-helpers.c b/dma-helpers.c
index 4faec5d..c38661e 100644
--- a/dma-helpers.c
+++ b/dma-helpers.c
@@ -79,9 +79,10 @@ typedef struct {
     QEMUIOVector iov;
     QEMUBH *bh;
     DMAIOFunc *io_func;
+    BlockCompletionFunc *dma_cb;
 } DMAAIOCB;
 
-static void dma_blk_cb(void *opaque, int ret);
+static void dma_blk_io_cb(void *opaque, int ret);
 
 static void reschedule_dma(void *opaque)
 {
@@ -89,7 +90,7 @@ static void reschedule_dma(void *opaque)
 
     qemu_bh_delete(dbs->bh);
     dbs->bh = NULL;
-    dma_blk_cb(dbs, 0);
+    dbs->dma_cb(dbs, 0);
 }
 
 static void dma_blk_unmap(DMAAIOCB *dbs)
@@ -120,21 +121,19 @@ static void dma_complete(DMAAIOCB *dbs, int ret)
     qemu_aio_unref(dbs);
 }
 
-static void dma_blk_cb(void *opaque, int ret)
+/*
+ * Create a QEMUIOVector from a scatter-gather list.
+ *
+ * This function does not copy the data of the scatter-gather list. Instead, it
+ * uses the dma_memory_map() function to map physical memory regions of the
+ * virtual device (as interpreted by the guest kernel) into the address space
+ * of the QEMU process, in order to have access to the data.
+ */
+static void dma_map_sg(DMAAIOCB *dbs)
 {
-    DMAAIOCB *dbs = (DMAAIOCB *)opaque;
     dma_addr_t cur_addr, cur_len;
     void *mem;
 
-    trace_dma_blk_cb(dbs, ret);
-
-    dbs->acb = NULL;
-    dbs->sector_num += dbs->iov.size / 512;
-
-    if (dbs->sg_cur_index == dbs->sg->nsg || ret < 0) {
-        dma_complete(dbs, ret);
-        return;
-    }
     dma_blk_unmap(dbs);
 
     while (dbs->sg_cur_index < dbs->sg->nsg) {
@@ -162,9 +161,38 @@ static void dma_blk_cb(void *opaque, int ret)
     if (dbs->iov.size & ~BDRV_SECTOR_MASK) {
         qemu_iovec_discard_back(&dbs->iov, dbs->iov.size & ~BDRV_SECTOR_MASK);
     }
+}
+
+/*
+ * Callback function for DMA read/write operations.
+ *
+ * This function initiates the read/write operation and also acts as a
+ * completion callback. It uses the dma_map_sg() function to map the
+ * scatter-gather list to a QEMUIOVector and then passes this iovec to the
+ * underlying read/write I/O function.
+ *
+ * If the DMA operation cannot take place in one step, e.g. it couldn't map all
+ * the scatter-gather entries, then this function will do a partial I/O
+ * operation and once done, it will be called and will retry the I/O operation.
+ */
+static void dma_blk_io_cb(void *opaque, int ret)
+{
+    DMAAIOCB *dbs = (DMAAIOCB *)opaque;
+
+    trace_dma_blk_io_cb(dbs, ret);
+
+    dbs->acb = NULL;
+    dbs->sector_num += dbs->iov.size / 512;
+
+    if (dbs->sg_cur_index == dbs->sg->nsg || ret < 0) {
+        dma_complete(dbs, ret);
+        return;
+    }
+
+    dma_map_sg(dbs);
 
     dbs->acb = dbs->io_func(dbs->blk, dbs->sector_num, &dbs->iov,
-                            dbs->iov.size / 512, dma_blk_cb, dbs);
+                            dbs->iov.size / 512, dma_blk_io_cb, dbs);
     assert(dbs->acb);
 }
 
@@ -190,6 +218,22 @@ static const AIOCBInfo dma_aiocb_info = {
     .cancel_async       = dma_aio_cancel,
 };
 
+/*
+ * Initialize the dbs structure and the QEMUIOVector to sane defaults.
+ */
+static void dma_init_dbs(DMAAIOCB *dbs, BlockBackend *blk, QEMUSGList *sg,
+                         DMADirection dir)
+{
+    dbs->acb = NULL;
+    dbs->blk = blk;
+    dbs->sg = sg;
+    dbs->sg_cur_index = 0;
+    dbs->sg_cur_byte = 0;
+    dbs->dir = dir;
+    dbs->bh = NULL;
+    qemu_iovec_init(&dbs->iov, sg->nsg);
+}
+
 BlockAIOCB *dma_blk_io(
     BlockBackend *blk, QEMUSGList *sg, uint64_t sector_num,
     DMAIOFunc *io_func, BlockCompletionFunc *cb,
@@ -199,21 +243,14 @@ BlockAIOCB *dma_blk_io(
 
     trace_dma_blk_io(dbs, blk, sector_num, (dir == DMA_DIRECTION_TO_DEVICE));
 
-    dbs->acb = NULL;
-    dbs->blk = blk;
-    dbs->sg = sg;
+    dma_init_dbs(dbs, blk, sg, dir);
     dbs->sector_num = sector_num;
-    dbs->sg_cur_index = 0;
-    dbs->sg_cur_byte = 0;
-    dbs->dir = dir;
     dbs->io_func = io_func;
-    dbs->bh = NULL;
-    qemu_iovec_init(&dbs->iov, sg->nsg);
-    dma_blk_cb(dbs, 0);
+    dbs->dma_cb = dma_blk_io_cb;
+    dbs->dma_cb(dbs, 0);
     return &dbs->common;
 }
 
-
 BlockAIOCB *dma_blk_read(BlockBackend *blk,
                          QEMUSGList *sg, uint64_t sector,
                          void (*cb)(void *opaque, int ret), void *opaque)
diff --git a/trace-events b/trace-events
index 2fce98e..120cdd4 100644
--- a/trace-events
+++ b/trace-events
@@ -1127,7 +1127,7 @@ win_helper_retry(uint32_t tl) "tl=%d"
 dma_blk_io(void *dbs, void *bs, int64_t sector_num, bool to_dev) "dbs=%p bs=%p sector_num=%" PRId64 " to_dev=%d"
 dma_aio_cancel(void *dbs) "dbs=%p"
 dma_complete(void *dbs, int ret, void *cb) "dbs=%p ret=%d cb=%p"
-dma_blk_cb(void *dbs, int ret) "dbs=%p ret=%d"
+dma_blk_io_cb(void *dbs, int ret) "dbs=%p ret=%d"
 dma_map_wait(void *dbs) "dbs=%p"
 
 # ui/console.c
-- 
2.6.2