summaryrefslogtreecommitdiff
path: root/.gitignore
AgeCommit message (Expand)Author
2007-08-15Add a set of tests for DRM locking, exposing issues on BSD.Eric Anholt
2007-08-15Add a regression test for the setversion interface.Eric Anholt
2007-08-15Add simple regression test for getstats (does it not crash the kernel?).Eric Anholt
2007-08-13Add a regression test for authentication.Eric Anholt
2007-07-20BSD: Replace symlink building with symlinks in git.Eric Anholt
2007-07-19Add a test for drawable add, remove, and update.Eric Anholt
2007-07-19Add some trivial regression tests, one of which fails.Eric Anholt
2007-07-19Add current BSD stuff to .gitignore.Eric Anholt
2007-02-02Make git ignore generated config.h.in.Michel Dänzer
2007-01-02Make git ignore Emacs style backup files and cscope files.Michel Dänzer
2006-12-01Track linux-core symlinks in git.Michel Dänzer
2006-07-19.cvsignore -> .gitignoreMichel Dänzer
lass="hl kwb">struct drm_device *dev) { struct drm_nouveau_private *dev_priv = dev->dev_private; uint32_t low; /* From kmmio dumps on nv28 this looks like how the blob does this. * It reads the high dword twice, before and after. * The only explanation seems to be that the 64-bit timer counter * advances between high and low dword reads and may corrupt the * result. Not confirmed. */ uint32_t high2 = NV_READ(NV04_PTIMER_TIME_1); uint32_t high1; do { high1 = high2; low = NV_READ(NV04_PTIMER_TIME_0); high2 = NV_READ(NV04_PTIMER_TIME_1); } while(high1 != high2); return (((uint64_t)high2) << 32) | (uint64_t)low; } void nv04_timer_takedown(struct drm_device *dev) { }