OSDN Git Service

drm/amd/display: Enable regamma 25 segments and use double buffer.
authorVitaly Prosyak <vitaly.prosyak@amd.com>
Tue, 7 Feb 2017 16:41:37 +0000 (10:41 -0600)
committerAlex Deucher <alexander.deucher@amd.com>
Tue, 26 Sep 2017 21:14:18 +0000 (17:14 -0400)
Moved custom floating point calculation to the shared place
between dce's.

Signed-off-by: Vitaly Prosyak <vitaly.prosyak@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drivers/gpu/drm/amd/display/dc/calcs/Makefile
drivers/gpu/drm/amd/display/dc/calcs/custom_float.c [new file with mode: 0644]
drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h
drivers/gpu/drm/amd/display/dc/inc/custom_float.h [new file with mode: 0644]

index 4bb08ae..2f4c8e7 100644 (file)
@@ -3,7 +3,7 @@
 # It calculates Bandwidth and Watermarks values for HW programming
 #
 
-BW_CALCS = bandwidth_calcs.o bw_fixed.o
+BW_CALCS = bandwidth_calcs.o bw_fixed.o custom_float.o
 
 AMD_DAL_BW_CALCS = $(addprefix $(AMDDALPATH)/dc/calcs/,$(BW_CALCS))
 
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/custom_float.c b/drivers/gpu/drm/amd/display/dc/calcs/custom_float.c
new file mode 100644 (file)
index 0000000..7243c37
--- /dev/null
@@ -0,0 +1,197 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+#include "dm_services.h"
+#include "custom_float.h"
+
+
+static bool build_custom_float(
+       struct fixed31_32 value,
+       const struct custom_float_format *format,
+       bool *negative,
+       uint32_t *mantissa,
+       uint32_t *exponenta)
+{
+       uint32_t exp_offset = (1 << (format->exponenta_bits - 1)) - 1;
+
+       const struct fixed31_32 mantissa_constant_plus_max_fraction =
+               dal_fixed31_32_from_fraction(
+                       (1LL << (format->mantissa_bits + 1)) - 1,
+                       1LL << format->mantissa_bits);
+
+       struct fixed31_32 mantiss;
+
+       if (dal_fixed31_32_eq(
+               value,
+               dal_fixed31_32_zero)) {
+               *negative = false;
+               *mantissa = 0;
+               *exponenta = 0;
+               return true;
+       }
+
+       if (dal_fixed31_32_lt(
+               value,
+               dal_fixed31_32_zero)) {
+               *negative = format->sign;
+               value = dal_fixed31_32_neg(value);
+       } else {
+               *negative = false;
+       }
+
+       if (dal_fixed31_32_lt(
+               value,
+               dal_fixed31_32_one)) {
+               uint32_t i = 1;
+
+               do {
+                       value = dal_fixed31_32_shl(value, 1);
+                       ++i;
+               } while (dal_fixed31_32_lt(
+                       value,
+                       dal_fixed31_32_one));
+
+               --i;
+
+               if (exp_offset <= i) {
+                       *mantissa = 0;
+                       *exponenta = 0;
+                       return true;
+               }
+
+               *exponenta = exp_offset - i;
+       } else if (dal_fixed31_32_le(
+               mantissa_constant_plus_max_fraction,
+               value)) {
+               uint32_t i = 1;
+
+               do {
+                       value = dal_fixed31_32_shr(value, 1);
+                       ++i;
+               } while (dal_fixed31_32_lt(
+                       mantissa_constant_plus_max_fraction,
+                       value));
+
+               *exponenta = exp_offset + i - 1;
+       } else {
+               *exponenta = exp_offset;
+       }
+
+       mantiss = dal_fixed31_32_sub(
+               value,
+               dal_fixed31_32_one);
+
+       if (dal_fixed31_32_lt(
+                       mantiss,
+                       dal_fixed31_32_zero) ||
+               dal_fixed31_32_lt(
+                       dal_fixed31_32_one,
+                       mantiss))
+               mantiss = dal_fixed31_32_zero;
+       else
+               mantiss = dal_fixed31_32_shl(
+                       mantiss,
+                       format->mantissa_bits);
+
+       *mantissa = dal_fixed31_32_floor(mantiss);
+
+       return true;
+}
+
+static bool setup_custom_float(
+       const struct custom_float_format *format,
+       bool negative,
+       uint32_t mantissa,
+       uint32_t exponenta,
+       uint32_t *result)
+{
+       uint32_t i = 0;
+       uint32_t j = 0;
+
+       uint32_t value = 0;
+
+       /* verification code:
+        * once calculation is ok we can remove it
+        */
+
+       const uint32_t mantissa_mask =
+               (1 << (format->mantissa_bits + 1)) - 1;
+
+       const uint32_t exponenta_mask =
+               (1 << (format->exponenta_bits + 1)) - 1;
+
+       if (mantissa & ~mantissa_mask) {
+               BREAK_TO_DEBUGGER();
+               mantissa = mantissa_mask;
+       }
+
+       if (exponenta & ~exponenta_mask) {
+               BREAK_TO_DEBUGGER();
+               exponenta = exponenta_mask;
+       }
+
+       /* end of verification code */
+
+       while (i < format->mantissa_bits) {
+               uint32_t mask = 1 << i;
+
+               if (mantissa & mask)
+                       value |= mask;
+
+               ++i;
+       }
+
+       while (j < format->exponenta_bits) {
+               uint32_t mask = 1 << j;
+
+               if (exponenta & mask)
+                       value |= mask << i;
+
+               ++j;
+       }
+
+       if (negative && format->sign)
+               value |= 1 << (i + j);
+
+       *result = value;
+
+       return true;
+}
+
+bool convert_to_custom_float_format(
+       struct fixed31_32 value,
+       const struct custom_float_format *format,
+       uint32_t *result)
+{
+       uint32_t mantissa;
+       uint32_t exponenta;
+       bool negative;
+
+       return build_custom_float(
+               value, format, &negative, &mantissa, &exponenta) &&
+       setup_custom_float(
+               format, negative, mantissa, exponenta, result);
+}
+
+
index 6a5cec0..88a2fa9 100644 (file)
@@ -49,6 +49,7 @@
 /* include DCE11 register header files */
 #include "dce/dce_11_0_d.h"
 #include "dce/dce_11_0_sh_mask.h"
+#include "custom_float.h"
 
 struct dce110_hw_seq_reg_offsets {
        uint32_t crtc;
@@ -286,174 +287,6 @@ static bool dce110_set_input_transfer_func(
        return result;
 }
 
-static bool build_custom_float(
-       struct fixed31_32 value,
-       const struct custom_float_format *format,
-       bool *negative,
-       uint32_t *mantissa,
-       uint32_t *exponenta)
-{
-       uint32_t exp_offset = (1 << (format->exponenta_bits - 1)) - 1;
-
-       const struct fixed31_32 mantissa_constant_plus_max_fraction =
-               dal_fixed31_32_from_fraction(
-                       (1LL << (format->mantissa_bits + 1)) - 1,
-                       1LL << format->mantissa_bits);
-
-       struct fixed31_32 mantiss;
-
-       if (dal_fixed31_32_eq(
-               value,
-               dal_fixed31_32_zero)) {
-               *negative = false;
-               *mantissa = 0;
-               *exponenta = 0;
-               return true;
-       }
-
-       if (dal_fixed31_32_lt(
-               value,
-               dal_fixed31_32_zero)) {
-               *negative = format->sign;
-               value = dal_fixed31_32_neg(value);
-       } else {
-               *negative = false;
-       }
-
-       if (dal_fixed31_32_lt(
-               value,
-               dal_fixed31_32_one)) {
-               uint32_t i = 1;
-
-               do {
-                       value = dal_fixed31_32_shl(value, 1);
-                       ++i;
-               } while (dal_fixed31_32_lt(
-                       value,
-                       dal_fixed31_32_one));
-
-               --i;
-
-               if (exp_offset <= i) {
-                       *mantissa = 0;
-                       *exponenta = 0;
-                       return true;
-               }
-
-               *exponenta = exp_offset - i;
-       } else if (dal_fixed31_32_le(
-               mantissa_constant_plus_max_fraction,
-               value)) {
-               uint32_t i = 1;
-
-               do {
-                       value = dal_fixed31_32_shr(value, 1);
-                       ++i;
-               } while (dal_fixed31_32_lt(
-                       mantissa_constant_plus_max_fraction,
-                       value));
-
-               *exponenta = exp_offset + i - 1;
-       } else {
-               *exponenta = exp_offset;
-       }
-
-       mantiss = dal_fixed31_32_sub(
-               value,
-               dal_fixed31_32_one);
-
-       if (dal_fixed31_32_lt(
-                       mantiss,
-                       dal_fixed31_32_zero) ||
-               dal_fixed31_32_lt(
-                       dal_fixed31_32_one,
-                       mantiss))
-               mantiss = dal_fixed31_32_zero;
-       else
-               mantiss = dal_fixed31_32_shl(
-                       mantiss,
-                       format->mantissa_bits);
-
-       *mantissa = dal_fixed31_32_floor(mantiss);
-
-       return true;
-}
-
-static bool setup_custom_float(
-       const struct custom_float_format *format,
-       bool negative,
-       uint32_t mantissa,
-       uint32_t exponenta,
-       uint32_t *result)
-{
-       uint32_t i = 0;
-       uint32_t j = 0;
-
-       uint32_t value = 0;
-
-       /* verification code:
-        * once calculation is ok we can remove it
-        */
-
-       const uint32_t mantissa_mask =
-               (1 << (format->mantissa_bits + 1)) - 1;
-
-       const uint32_t exponenta_mask =
-               (1 << (format->exponenta_bits + 1)) - 1;
-
-       if (mantissa & ~mantissa_mask) {
-               BREAK_TO_DEBUGGER();
-               mantissa = mantissa_mask;
-       }
-
-       if (exponenta & ~exponenta_mask) {
-               BREAK_TO_DEBUGGER();
-               exponenta = exponenta_mask;
-       }
-
-       /* end of verification code */
-
-       while (i < format->mantissa_bits) {
-               uint32_t mask = 1 << i;
-
-               if (mantissa & mask)
-                       value |= mask;
-
-               ++i;
-       }
-
-       while (j < format->exponenta_bits) {
-               uint32_t mask = 1 << j;
-
-               if (exponenta & mask)
-                       value |= mask << i;
-
-               ++j;
-       }
-
-       if (negative && format->sign)
-               value |= 1 << (i + j);
-
-       *result = value;
-
-       return true;
-}
-
-static bool convert_to_custom_float_format(
-       struct fixed31_32 value,
-       const struct custom_float_format *format,
-       uint32_t *result)
-{
-       uint32_t mantissa;
-       uint32_t exponenta;
-       bool negative;
-
-       return build_custom_float(
-               value, format, &negative, &mantissa, &exponenta) &&
-       setup_custom_float(
-               format, negative, mantissa, exponenta, result);
-}
-
 static bool convert_to_custom_float(
                struct pwl_result_data *rgb_resulted,
                struct curve_points *arr_points,
@@ -579,7 +412,7 @@ static bool convert_to_custom_float(
        return true;
 }
 
-bool dce110_translate_regamma_to_hw_format(const struct dc_transfer_func
+static bool dce110_translate_regamma_to_hw_format(const struct dc_transfer_func
                *output_tf, struct pwl_params *regamma_params)
 {
        struct curve_points *arr_points;
index 68632dd..a6b4d0d 100644 (file)
@@ -58,8 +58,5 @@ void dce110_power_down(struct core_dc *dc);
 
 void dce110_update_pending_status(struct pipe_ctx *pipe_ctx);
 
-bool dce110_translate_regamma_to_hw_format(const struct dc_transfer_func
-               *output_tf, struct pwl_params *regamma_params);
-
 #endif /* __DC_HWSS_DCE110_H__ */
 
diff --git a/drivers/gpu/drm/amd/display/dc/inc/custom_float.h b/drivers/gpu/drm/amd/display/dc/inc/custom_float.h
new file mode 100644 (file)
index 0000000..f572396
--- /dev/null
@@ -0,0 +1,40 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef CUSTOM_FLOAT_H_
+#define CUSTOM_FLOAT_H_
+
+#include "bw_fixed.h"
+#include "hw_shared.h"
+#include "opp.h"
+
+
+bool convert_to_custom_float_format(
+       struct fixed31_32 value,
+       const struct custom_float_format *format,
+       uint32_t *result);
+
+
+#endif //CUSTOM_FLOAT_H_