Re: [Mesa-dev] [PATCH] r600g: fix op3 abs issue
On Tue, 31 Mar 2015 07:27:50 +0200, Dave Airlie airl...@gmail.com wrote:
From: Dave Airlie airl...@redhat.com
This code to handle absolute values on op3 srcs was a bit too simple,
it really needs a temp reg per src, not one per channel, make it
easier and let sb clean up the mess.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89831
Signed-off-by: Dave Airlie airl...@redhat.com
---
src/gallium/drivers/r600/r600_shader.c | 51
++
1 file changed, 34 insertions(+), 17 deletions(-)
diff --git a/src/gallium/drivers/r600/r600_shader.c
b/src/gallium/drivers/r600/r600_shader.c
index ff2c784..1367599 100644
--- a/src/gallium/drivers/r600/r600_shader.c
+++ b/src/gallium/drivers/r600/r600_shader.c
@@ -4864,10 +4864,9 @@ static int tgsi_helper_copy(struct
r600_shader_ctx *ctx, struct tgsi_full_instru
}
static int tgsi_make_src_for_op3(struct r600_shader_ctx *ctx,
- unsigned temp, int temp_chan,
- struct r600_bytecode_alu_src *bc_src,
- const struct r600_shader_src
*shader_src,
- unsigned chan)
+ unsigned temp, int chan,
+ struct r600_bytecode_alu_src *bc_src,
+ const struct r600_shader_src
*shader_src)
{
struct r600_bytecode_alu alu;
int r;
@@ -4880,7 +4879,7 @@ static int tgsi_make_src_for_op3(struct
r600_shader_ctx *ctx,
memset(alu, 0, sizeof(struct r600_bytecode_alu));
alu.op = ALU_OP1_MOV;
alu.dst.sel = temp;
- alu.dst.chan = temp_chan;
+ alu.dst.chan = chan;
alu.dst.write = 1;
alu.src[0] = *bc_src;
@@ -4891,7 +4890,7 @@ static int tgsi_make_src_for_op3(struct
r600_shader_ctx *ctx,
memset(bc_src, 0, sizeof(*bc_src));
bc_src-sel = temp;
- bc_src-chan = temp_chan;
+ bc_src-chan = chan;
}
return 0;
}
@@ -4902,7 +4901,13 @@ static int tgsi_op3(struct r600_shader_ctx *ctx)
struct r600_bytecode_alu alu;
int i, j, r;
int lasti = tgsi_last_instruction(inst-Dst[0].Register.WriteMask);
+ int temp_regs[4];
+ for (j = 0; j inst-Instruction.NumSrcRegs; j++) {
+ temp_regs[j] = 0;
+ if (ctx-src[j].abs)
+ temp_regs[j] = r600_get_temp(ctx);
+ }
for (i = 0; i lasti + 1; i++) {
if (!(inst-Dst[0].Register.WriteMask (1 i)))
continue;
@@ -4910,7 +4915,7 @@ static int tgsi_op3(struct r600_shader_ctx *ctx)
memset(alu, 0, sizeof(struct r600_bytecode_alu));
alu.op = ctx-inst_info-op;
for (j = 0; j inst-Instruction.NumSrcRegs; j++) {
- r = tgsi_make_src_for_op3(ctx, ctx-temp_reg, j, alu.src[j],
ctx-src[j], i);
+ r = tgsi_make_src_for_op3(ctx, temp_regs[j], i, alu.src[j],
ctx-src[j]);
if (r)
return r;
}
@@ -6003,7 +6008,7 @@ static int tgsi_lrp(struct r600_shader_ctx *ctx)
struct tgsi_full_instruction *inst =
ctx-parse.FullToken.FullInstruction;
struct r600_bytecode_alu alu;
int lasti = tgsi_last_instruction(inst-Dst[0].Register.WriteMask);
- unsigned i, extra_temp;
+ unsigned i, temp_regs[2];
int r;
/* optimize if it's just an equal balance */
@@ -6073,10 +6078,15 @@ static int tgsi_lrp(struct r600_shader_ctx *ctx)
}
/* src0 * src1 + (1 - src0) * src2 */
- if (ctx-src[0].abs || ctx-src[1].abs) /* XXX avoid dupliating
condition */
- extra_temp = r600_get_temp(ctx);
+if (ctx-src[0].abs)
Rest of patch uses tabs for indendation, this line spaces.
+ temp_regs[0] = r600_get_temp(ctx);
+ else
+ temp_regs[0] = 0;
+ if (ctx-src[1].abs)
+ temp_regs[1] = r600_get_temp(ctx);
else
- extra_temp = 0;
+ temp_regs[1] = 0;
+
for (i = 0; i lasti + 1; i++) {
if (!(inst-Dst[0].Register.WriteMask (1 i)))
continue;
@@ -6084,10 +6094,10 @@ static int tgsi_lrp(struct r600_shader_ctx *ctx)
memset(alu, 0, sizeof(struct r600_bytecode_alu));
alu.op = ALU_OP3_MULADD;
alu.is_op3 = 1;
- r = tgsi_make_src_for_op3(ctx, extra_temp, 0, alu.src[0],
ctx-src[0], i);
+ r = tgsi_make_src_for_op3(ctx, temp_regs[0], i, alu.src[0],
ctx-src[0]);
if (r)
return r;
- r = tgsi_make_src_for_op3(ctx, extra_temp, 1, alu.src[1],
ctx-src[1], i);
+ r = tgsi_make_src_for_op3(ctx, temp_regs[1], i, alu.src[1],
ctx-src[1]);
if (r)
return r;
[Mesa-dev] [PATCH] r600g: fix op3 abs issue
From: Dave Airlie airl...@redhat.com
This code to handle absolute values on op3 srcs was a bit too simple,
it really needs a temp reg per src, not one per channel, make it
easier and let sb clean up the mess.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89831
Signed-off-by: Dave Airlie airl...@redhat.com
---
src/gallium/drivers/r600/r600_shader.c | 51 ++
1 file changed, 34 insertions(+), 17 deletions(-)
diff --git a/src/gallium/drivers/r600/r600_shader.c
b/src/gallium/drivers/r600/r600_shader.c
index ff2c784..1367599 100644
--- a/src/gallium/drivers/r600/r600_shader.c
+++ b/src/gallium/drivers/r600/r600_shader.c
@@ -4864,10 +4864,9 @@ static int tgsi_helper_copy(struct r600_shader_ctx *ctx,
struct tgsi_full_instru
}
static int tgsi_make_src_for_op3(struct r600_shader_ctx *ctx,
- unsigned temp, int temp_chan,
- struct r600_bytecode_alu_src *bc_src,
- const struct r600_shader_src *shader_src,
- unsigned chan)
+ unsigned temp, int chan,
+ struct r600_bytecode_alu_src *bc_src,
+ const struct r600_shader_src *shader_src)
{
struct r600_bytecode_alu alu;
int r;
@@ -4880,7 +4879,7 @@ static int tgsi_make_src_for_op3(struct r600_shader_ctx
*ctx,
memset(alu, 0, sizeof(struct r600_bytecode_alu));
alu.op = ALU_OP1_MOV;
alu.dst.sel = temp;
- alu.dst.chan = temp_chan;
+ alu.dst.chan = chan;
alu.dst.write = 1;
alu.src[0] = *bc_src;
@@ -4891,7 +4890,7 @@ static int tgsi_make_src_for_op3(struct r600_shader_ctx
*ctx,
memset(bc_src, 0, sizeof(*bc_src));
bc_src-sel = temp;
- bc_src-chan = temp_chan;
+ bc_src-chan = chan;
}
return 0;
}
@@ -4902,7 +4901,13 @@ static int tgsi_op3(struct r600_shader_ctx *ctx)
struct r600_bytecode_alu alu;
int i, j, r;
int lasti = tgsi_last_instruction(inst-Dst[0].Register.WriteMask);
+ int temp_regs[4];
+ for (j = 0; j inst-Instruction.NumSrcRegs; j++) {
+ temp_regs[j] = 0;
+ if (ctx-src[j].abs)
+ temp_regs[j] = r600_get_temp(ctx);
+ }
for (i = 0; i lasti + 1; i++) {
if (!(inst-Dst[0].Register.WriteMask (1 i)))
continue;
@@ -4910,7 +4915,7 @@ static int tgsi_op3(struct r600_shader_ctx *ctx)
memset(alu, 0, sizeof(struct r600_bytecode_alu));
alu.op = ctx-inst_info-op;
for (j = 0; j inst-Instruction.NumSrcRegs; j++) {
- r = tgsi_make_src_for_op3(ctx, ctx-temp_reg, j,
alu.src[j], ctx-src[j], i);
+ r = tgsi_make_src_for_op3(ctx, temp_regs[j], i,
alu.src[j], ctx-src[j]);
if (r)
return r;
}
@@ -6003,7 +6008,7 @@ static int tgsi_lrp(struct r600_shader_ctx *ctx)
struct tgsi_full_instruction *inst =
ctx-parse.FullToken.FullInstruction;
struct r600_bytecode_alu alu;
int lasti = tgsi_last_instruction(inst-Dst[0].Register.WriteMask);
- unsigned i, extra_temp;
+ unsigned i, temp_regs[2];
int r;
/* optimize if it's just an equal balance */
@@ -6073,10 +6078,15 @@ static int tgsi_lrp(struct r600_shader_ctx *ctx)
}
/* src0 * src1 + (1 - src0) * src2 */
- if (ctx-src[0].abs || ctx-src[1].abs) /* XXX avoid dupliating
condition */
- extra_temp = r600_get_temp(ctx);
+if (ctx-src[0].abs)
+ temp_regs[0] = r600_get_temp(ctx);
+ else
+ temp_regs[0] = 0;
+ if (ctx-src[1].abs)
+ temp_regs[1] = r600_get_temp(ctx);
else
- extra_temp = 0;
+ temp_regs[1] = 0;
+
for (i = 0; i lasti + 1; i++) {
if (!(inst-Dst[0].Register.WriteMask (1 i)))
continue;
@@ -6084,10 +6094,10 @@ static int tgsi_lrp(struct r600_shader_ctx *ctx)
memset(alu, 0, sizeof(struct r600_bytecode_alu));
alu.op = ALU_OP3_MULADD;
alu.is_op3 = 1;
- r = tgsi_make_src_for_op3(ctx, extra_temp, 0, alu.src[0],
ctx-src[0], i);
+ r = tgsi_make_src_for_op3(ctx, temp_regs[0], i, alu.src[0],
ctx-src[0]);
if (r)
return r;
- r = tgsi_make_src_for_op3(ctx, extra_temp, 1, alu.src[1],
ctx-src[1], i);
+ r = tgsi_make_src_for_op3(ctx, temp_regs[1], i, alu.src[1],
ctx-src[1]);
if (r)
return r;
alu.src[2].sel = ctx-temp_reg;
@@ -6109,8 +6119,15 @@ static int
