diff options
| author | Christoph Bumiller <[email protected]> | 2011-04-16 15:24:07 +0200 |
|---|---|---|
| committer | Christoph Bumiller <[email protected]> | 2011-04-17 00:17:14 +0200 |
| commit | 25a7f662145f56ad15dbdf3da203a8a85594276f (patch) | |
| tree | 556cbdf4474b9c86adf37b43e12ce65d8e986c80 | |
| parent | f4ad25fefb33ad4930b09c685688ba71b4a21f32 (diff) | |
nvc0: work around what looks like a code prefetch bug
Sounds very unlikely, but I don't have a better explanation at the
moment.
The GPU throws page faults at the first page after the code buffer
quite frequently on startup, and traces don't show us overflowing.
| -rw-r--r-- | src/gallium/drivers/nvc0/nvc0_screen.c | 5 |
1 files changed, 4 insertions, 1 deletions
diff --git a/src/gallium/drivers/nvc0/nvc0_screen.c b/src/gallium/drivers/nvc0/nvc0_screen.c index c4cdface502..de62fd68b28 100644 --- a/src/gallium/drivers/nvc0/nvc0_screen.c +++ b/src/gallium/drivers/nvc0/nvc0_screen.c @@ -447,7 +447,10 @@ nvc0_screen_create(struct pipe_winsys *ws, struct nouveau_device *dev) if (ret) goto fail; - nouveau_resource_init(&screen->text_heap, 0, 1 << 20); + /* XXX: getting a page fault at the end of the code buffer every few + * launches, don't use the last 256 bytes to work around them - prefetch ? + */ + nouveau_resource_init(&screen->text_heap, 0, (1 << 20) - 0x100); ret = nouveau_bo_new(dev, NOUVEAU_BO_VRAM, 1 << 12, 6 << 16, &screen->uniforms); |