GEN11_FEATURES(1, 1, subslices(1), 6),
};
+static void
+gen_device_info_set_eu_mask(struct gen_device_info *devinfo,
+ unsigned slice,
+ unsigned subslice,
+ unsigned eu_mask)
+{
+ unsigned subslice_offset = slice * devinfo->eu_slice_stride +
+ subslice * devinfo->eu_subslice_stride;
+
+ for (unsigned b_eu = 0; b_eu < devinfo->eu_subslice_stride; b_eu++) {
+ devinfo->eu_masks[subslice_offset + b_eu] =
+ (((1U << devinfo->num_eu_per_subslice) - 1) >> (b_eu * 8)) & 0xff;
+ }
+}
+
+/* Generate slice/subslice/eu masks from number of
+ * slices/subslices/eu_per_subslices in the per generation/gt gen_device_info
+ * structure.
+ *
+ * These can be overridden with values reported by the kernel either from
+ * getparam SLICE_MASK/SUBSLICE_MASK values or from the kernel version 4.17+
+ * through the i915 query uapi.
+ */
+static void
+fill_masks(struct gen_device_info *devinfo)
+{
+ devinfo->slice_masks = (1U << devinfo->num_slices) - 1;
+
+ /* Subslice masks */
+ unsigned max_subslices = 0;
+ for (int s = 0; s < devinfo->num_slices; s++)
+ max_subslices = MAX2(devinfo->num_subslices[s], max_subslices);
+ devinfo->subslice_slice_stride = DIV_ROUND_UP(max_subslices, 8);
+
+ for (int s = 0; s < devinfo->num_slices; s++) {
+ devinfo->subslice_masks[s * devinfo->subslice_slice_stride] =
+ (1U << devinfo->num_subslices[s]) - 1;
+ }
+
+ /* EU masks */
+ devinfo->eu_subslice_stride = DIV_ROUND_UP(devinfo->num_eu_per_subslice, 8);
+ devinfo->eu_slice_stride = max_subslices * devinfo->eu_subslice_stride;
+
+ for (int s = 0; s < devinfo->num_slices; s++) {
+ for (int ss = 0; ss < devinfo->num_subslices[s]; ss++) {
+ gen_device_info_set_eu_mask(devinfo, s, ss,
+ (1U << devinfo->num_eu_per_subslice) - 1);
+ }
+ }
+}
+
bool
gen_get_device_info(int devid, struct gen_device_info *devinfo)
{
return false;
}
+ fill_masks(devinfo);
+
/* From the Skylake PRM, 3DSTATE_PS::Scratch Space Base Pointer:
*
* "Scratch Space per slice is computed based on 4 sub-slices. SW must
#include <stdbool.h>
#include <stdint.h>
+#include "util/macros.h"
+
#ifdef __cplusplus
extern "C" {
#endif
+#define GEN_DEVICE_MAX_SLICES (6) /* Maximum on gen10 */
+#define GEN_DEVICE_MAX_SUBSLICES (8) /* Maximum on gen11 */
+#define GEN_DEVICE_MAX_EUS_PER_SUBSLICE (10) /* Maximum on Haswell */
+
/**
* Intel hardware information and quirks
*/
/**
* Number of subslices for each slice (used to be uniform until CNL).
*/
- unsigned num_subslices[3];
+ unsigned num_subslices[GEN_DEVICE_MAX_SUBSLICES];
/**
* Number of EU per subslice.
*/
unsigned num_thread_per_eu;
+ /**
+ * A bit mask of the slices available.
+ */
+ uint8_t slice_masks;
+
+ /**
+ * An array of bit mask of the subslices available, use subslice_slice_stride
+ * to access this array.
+ */
+ uint8_t subslice_masks[GEN_DEVICE_MAX_SLICES *
+ DIV_ROUND_UP(GEN_DEVICE_MAX_SUBSLICES, 8)];
+
+ /**
+ * An array of bit mask of EUs available, use eu_slice_stride &
+ * eu_subslice_stride to access this array.
+ */
+ uint8_t eu_masks[GEN_DEVICE_MAX_SLICES *
+ GEN_DEVICE_MAX_SUBSLICES *
+ DIV_ROUND_UP(GEN_DEVICE_MAX_EUS_PER_SUBSLICE, 8)];
+
+ /**
+ * Stride to access subslice_masks[].
+ */
+ uint16_t subslice_slice_stride;
+
+ /**
+ * Strides to access eu_masks[].
+ */
+ uint16_t eu_slice_stride;
+ uint16_t eu_subslice_stride;
+
unsigned l3_banks;
unsigned max_vs_threads; /**< Maximum Vertex Shader threads */
unsigned max_tcs_threads; /**< Maximum Hull Shader threads */