intel: use pwrite for batch

It's faster. Not only is the memcpy more efficiently performed in the
kernel (making up for the system call overhead), but by not using mmap
we remove the greater overhead of tracking the vma of every batch.

And it means we can read back from the batch buffer without incurring
the cost of a uncached read through the GTT.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

1 files changed, 5 insertions, 6 deletions

diff --git a/src/mesa/drivers/dri/intel/intel_context.c b/src/mesa/drivers/dri/intel/intel_context.c
index 65c4148e73..c470febfb3 100644
--- a/src/mesa/drivers/dri/intel/intel_context.c
+++ b/src/mesa/drivers/dri/intel/intel_context.c
@@ -554,8 +554,8 @@ intel_flush(struct gl_context *ctx)
    if (intel->gen < 4)
       INTEL_FIREVERTICES(intel);
 
-   if (intel->batch->map != intel->batch->ptr)
-      intel_batchbuffer_flush(intel->batch);
+   if (intel->batch.used)
+      intel_batchbuffer_flush(intel);
 }
 
 static void
@@ -751,7 +751,7 @@ intelInitContext(struct intel_context *intel,
    if (intelScreen->deviceID == PCI_CHIP_I865_G)
       intel->maxBatchSize = 4096;
    else
-      intel->maxBatchSize = BATCH_SZ;
+      intel->maxBatchSize = sizeof(intel->batch.map);
 
    intel->bufmgr = intelScreen->bufmgr;
 
@@ -863,7 +863,7 @@ intelInitContext(struct intel_context *intel,
    if (INTEL_DEBUG & DEBUG_BUFMGR)
       dri_bufmgr_set_debug(intel->bufmgr, GL_TRUE);
 
-   intel->batch = intel_batchbuffer_alloc(intel);
+   intel_batchbuffer_reset(intel);
 
    intel_fbo_init(intel);
 
@@ -920,8 +920,7 @@ intelDestroyContext(__DRIcontext * driContextPriv)
       _swrast_DestroyContext(&intel->ctx);
       intel->Fallback = 0x0;      /* don't call _swrast_Flush later */
 
-      intel_batchbuffer_free(intel->batch);
-      intel->batch = NULL;
+      intel_batchbuffer_free(intel);
 
       free(intel->prim.vb);
       intel->prim.vb = NULL;