Fix a race condition for multi-thread qemu dma, where vmx linux guests

show warning "dma interrupt lost" and dma becomes very slow.

root cause: In the time between set ide irq and set dma status, if guest
receive the irq and query the status, it will find the status is not
ready and therefore treat it as pseudo interrupt. Change the order of
set irq and set dma status fixes this issue.

Signed-off-by: Ke Yu <ke.yu@intel.com>

diff --git a/tools/ioemu/hw/ide.c b/tools/ioemu/hw/ide.c
index 71ad561154065a5749a3156d87b42eb248bdb8c2..7c012cebcb8bd7d371785102a38553c56d32dc04 100644 (file)
--- a/tools/ioemu/hw/ide.c
+++ b/tools/ioemu/hw/ide.c
@@ -669,6 +669,8 @@ static int ide_read_dma_cb(IDEState *s,
     }
     if (s->io_buffer_index >= s->io_buffer_size && s->nsector == 0) {
         s->status = READY_STAT | SEEK_STAT;
+        s->bmdma->status &= ~BM_STATUS_DMAING;
+        s->bmdma->status |= BM_STATUS_INT;
         ide_set_irq(s);
 #ifdef DEBUG_IDE_ATAPI
         printf("dma status=0x%x\n", s->status);
@@ -736,6 +738,8 @@ static int ide_write_dma_cb(IDEState *s,
             if (n == 0) {
                 /* end of transfer */
                 s->status = READY_STAT | SEEK_STAT;
+                s->bmdma->status &= ~BM_STATUS_DMAING;
+                s->bmdma->status |= BM_STATUS_INT;
                 ide_set_irq(s);
                 return 0;
             }
@@ -983,6 +987,8 @@ static int ide_atapi_cmd_read_dma_cb(IDEState *s,
     if (s->packet_transfer_size <= 0) {
         s->status = READY_STAT;
         s->nsector = (s->nsector & ~7) | ATAPI_INT_REASON_IO | ATAPI_INT_REASON_CD;
+        s->bmdma->status &= ~BM_STATUS_DMAING;
+        s->bmdma->status |= BM_STATUS_INT;
         ide_set_irq(s);
 #ifdef DEBUG_IDE_ATAPI
         printf("dma status=0x%x\n", s->status);
@@ -2065,8 +2071,6 @@ static void ide_dma_loop(BMDMAState *bm)
     }
     /* end of transfer */
  the_end:
-    bm->status &= ~BM_STATUS_DMAING;
-    bm->status |= BM_STATUS_INT;
     bm->dma_cb = NULL;
     bm->ide_if = NULL;
 }