On Tue, 20 Aug 2019 04:01:47 +0200, Daniele Ceraolo Spurio
<daniele.ceraolospurio@xxxxxxxxx> wrote:
diff --git a/drivers/gpu/drm/i915/intel_reg_types.h
b/drivers/gpu/drm/i915/intel_reg_types.h
new file mode 100644
index 000000000000..87bce80dd5ed
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_reg_types.h
+
+typedef struct {
+ u32 reg;
+} i915_reg_t;
+
+#define _MMIO(r) ((const i915_reg_t){ .reg = (r) })
+
+#define INVALID_MMIO_REG _MMIO(0)
+
+static inline u32 i915_mmio_reg_offset(i915_reg_t reg)
+{
+ return reg.reg;
+}
+
+static inline bool i915_mmio_reg_equal(i915_reg_t a, i915_reg_t b)
+{
+ return i915_mmio_reg_offset(a) == i915_mmio_reg_offset(b);
+}
+
+static inline bool i915_mmio_reg_valid(i915_reg_t reg)
+{
+ return !i915_mmio_reg_equal(reg, INVALID_MMIO_REG);
+}
+
hmm, there is now disconnection between prefixes in:
'intel'_reg_types.h
and
'i915'_reg_t
'i915'_mmio_reg_xxx()
that is why I was suggesting to keep:
'i915'_reg.h (or at your preference 'i915'_reg_types.h)
with
'i915'_reg_t
'i915'_mmio_reg_xxx()
and use intel_reg* files for actual hw definitions.
if we don't plan to rename i915_reg_t into intel_mmio_reg_t
then maybe better to stay with i915_reg_types.h ?
Michal
ps. i915/intel prefix rules are killing me too ;)
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
