From 14f0b7ea98d5cb2f805fc22796596ef878ae24cb Mon Sep 17 00:00:00 2001 From: Roland Scheidegger Date: Tue, 13 Mar 2007 13:44:23 +0100 Subject: use movdqu instead of movdqa for unaligned load avoiding a segfault (bug 10265) --- src/mesa/x86/read_rgba_span_x86.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'src') diff --git a/src/mesa/x86/read_rgba_span_x86.S b/src/mesa/x86/read_rgba_span_x86.S index a690283816..3cbcd71996 100644 --- a/src/mesa/x86/read_rgba_span_x86.S +++ b/src/mesa/x86/read_rgba_span_x86.S @@ -369,7 +369,7 @@ _generic_read_RGBA_span_BGRA8888_REV_SSE2: movdqa mask, %xmm1 movdqa mask+16, %xmm2 */ - LOAD_MASK(movdqa,%xmm1,%xmm2) + LOAD_MASK(movdqu,%xmm1,%xmm2) movl 12(%esp), %ebx /* source pointer */ movl 20(%esp), %edx /* number of pixels to copy */ -- cgit v1.2.3 From 4c18d9056b3793a8441ecbfc9c5ea5f2d43555b2 Mon Sep 17 00:00:00 2001 From: Oliver McFadden Date: Tue, 13 Mar 2007 14:46:45 +0000 Subject: r300: Renamed the CACHE_CTLSTAT values to include UNKNOWN in the name; not enough information is known about them to be sure as to what the values mean. --- src/mesa/drivers/dri/r300/r300_ioctl.c | 4 ++-- src/mesa/drivers/dri/r300/r300_reg.h | 8 ++++---- src/mesa/drivers/dri/r300/r300_render.c | 8 ++++---- src/mesa/drivers/dri/r300/radeon_mm.c | 2 +- 4 files changed, 11 insertions(+), 11 deletions(-) (limited to 'src') diff --git a/src/mesa/drivers/dri/r300/r300_ioctl.c b/src/mesa/drivers/dri/r300/r300_ioctl.c index f0741f978d..f674fea7c9 100644 --- a/src/mesa/drivers/dri/r300/r300_ioctl.c +++ b/src/mesa/drivers/dri/r300/r300_ioctl.c @@ -162,11 +162,11 @@ static void r300ClearBuffer(r300ContextPtr r300, int flags, int buffer) cmd2[8].u = r300PackFloat32(ctx->Color.ClearColor[3]); reg_start(R300_RB3D_DSTCACHE_CTLSTAT,0); - e32(R300_RB3D_DSTCACHE_0A); + e32(R300_RB3D_DSTCACHE_UNKNOWN_0A); reg_start(R300_RB3D_ZCACHE_CTLSTAT,0); - e32(R300_RB3D_ZCACHE_CTLSTAT_03); + e32(R300_RB3D_ZCACHE_UNKNOWN_03); cp_wait(rmesa, R300_WAIT_3D | R300_WAIT_3D_CLEAN); } diff --git a/src/mesa/drivers/dri/r300/r300_reg.h b/src/mesa/drivers/dri/r300/r300_reg.h index f9accadf61..3de15752b1 100644 --- a/src/mesa/drivers/dri/r300/r300_reg.h +++ b/src/mesa/drivers/dri/r300/r300_reg.h @@ -1324,8 +1324,8 @@ USE OR OTHER DEALINGS IN THE SOFTWARE. * Set to 0A before 3D operations, set to 02 afterwards. */ #define R300_RB3D_DSTCACHE_CTLSTAT 0x4E4C -# define R300_RB3D_DSTCACHE_02 0x00000002 -# define R300_RB3D_DSTCACHE_0A 0x0000000A +# define R300_RB3D_DSTCACHE_UNKNOWN_02 0x00000002 +# define R300_RB3D_DSTCACHE_UNKNOWN_0A 0x0000000A /* gap */ /* There seems to be no "write only" setting, so use Z-test = ALWAYS @@ -1397,8 +1397,8 @@ USE OR OTHER DEALINGS IN THE SOFTWARE. /* gap */ #define R300_RB3D_ZCACHE_CTLSTAT 0x4F18 /* GUESS */ -# define R300_RB3D_ZCACHE_CTLSTAT_01 0x1 -# define R300_RB3D_ZCACHE_CTLSTAT_03 0x3 +# define R300_RB3D_ZCACHE_UNKNOWN_01 0x1 +# define R300_RB3D_ZCACHE_UNKNOWN_03 0x3 /* gap */ diff --git a/src/mesa/drivers/dri/r300/r300_render.c b/src/mesa/drivers/dri/r300/r300_render.c index 659ec3ff54..0864558e8d 100644 --- a/src/mesa/drivers/dri/r300/r300_render.c +++ b/src/mesa/drivers/dri/r300/r300_render.c @@ -344,10 +344,10 @@ GLboolean r300_run_vb_render(GLcontext *ctx, r300UpdateShaderStates(rmesa); reg_start(R300_RB3D_DSTCACHE_CTLSTAT,0); - e32(R300_RB3D_DSTCACHE_0A); + e32(R300_RB3D_DSTCACHE_UNKNOWN_0A); reg_start(R300_RB3D_ZCACHE_CTLSTAT,0); - e32(R300_RB3D_ZCACHE_CTLSTAT_03); + e32(R300_RB3D_ZCACHE_UNKNOWN_03); r300EmitState(rmesa); @@ -360,10 +360,10 @@ GLboolean r300_run_vb_render(GLcontext *ctx, } reg_start(R300_RB3D_DSTCACHE_CTLSTAT,0); - e32(R300_RB3D_DSTCACHE_0A /*R300_RB3D_DSTCACHE_02*/); + e32(R300_RB3D_DSTCACHE_UNKNOWN_0A /*R300_RB3D_DSTCACHE_UNKNOWN_02*/); reg_start(R300_RB3D_ZCACHE_CTLSTAT,0); - e32(R300_RB3D_ZCACHE_CTLSTAT_03 /*R300_RB3D_ZCACHE_CTLSTAT_01*/); + e32(R300_RB3D_ZCACHE_UNKNOWN_03 /*R300_RB3D_ZCACHE_UNKNOWN_01*/); #ifdef USER_BUFFERS r300UseArrays(ctx); diff --git a/src/mesa/drivers/dri/r300/radeon_mm.c b/src/mesa/drivers/dri/r300/radeon_mm.c index 065f288e4c..1502dac8fc 100644 --- a/src/mesa/drivers/dri/r300/radeon_mm.c +++ b/src/mesa/drivers/dri/r300/radeon_mm.c @@ -284,7 +284,7 @@ static void emit_lin_cp(r300ContextPtr rmesa, unsigned long dst, unsigned long s } reg_start(R300_RB3D_DSTCACHE_CTLSTAT,0); - e32(R300_RB3D_DSTCACHE_0A); + e32(R300_RB3D_DSTCACHE_UNKNOWN_0A); reg_start(0x342c,0); e32(0x00000005); -- cgit v1.2.3 From 37e6f760fde506728077c6e8b48bb8c34d53dfd3 Mon Sep 17 00:00:00 2001 From: Roland Scheidegger Date: Tue, 13 Mar 2007 17:43:21 +0100 Subject: enable ARB_vertex_buffer_object for more dri drivers ARB_vertex_buffer_object looks like a useful extension even for old chips. The drivers should not need any code to be able to use this extension since they just use mesa's vbo code anyway. Newly enabled for i810, mach64, mga, r128, radeon, savage, sis and unichrome. --- src/mesa/drivers/dri/i810/i810context.c | 2 ++ src/mesa/drivers/dri/mach64/mach64_context.c | 2 ++ src/mesa/drivers/dri/mga/mga_xmesa.c | 2 ++ src/mesa/drivers/dri/r128/r128_context.c | 2 ++ src/mesa/drivers/dri/radeon/radeon_context.c | 2 ++ src/mesa/drivers/dri/savage/savage_xmesa.c | 2 ++ src/mesa/drivers/dri/sis/sis_context.c | 2 ++ src/mesa/drivers/dri/unichrome/via_context.c | 2 ++ 8 files changed, 16 insertions(+) (limited to 'src') diff --git a/src/mesa/drivers/dri/i810/i810context.c b/src/mesa/drivers/dri/i810/i810context.c index db8f7a19a2..f0332d9ae8 100644 --- a/src/mesa/drivers/dri/i810/i810context.c +++ b/src/mesa/drivers/dri/i810/i810context.c @@ -65,6 +65,7 @@ SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. #define need_GL_ARB_multisample #define need_GL_ARB_texture_compression +#define need_GL_ARB_vertex_buffer_object #include "extension_helper.h" #ifndef I810_DEBUG @@ -129,6 +130,7 @@ const struct dri_extension card_extensions[] = { "GL_ARB_texture_env_combine", NULL }, { "GL_ARB_texture_env_crossbar", NULL }, { "GL_ARB_texture_mirrored_repeat", NULL }, + { "GL_ARB_vertex_buffer_object", GL_ARB_vertex_buffer_object_functions }, { "GL_EXT_stencil_wrap", NULL }, { "GL_EXT_texture_edge_clamp", NULL }, { "GL_EXT_texture_env_combine", NULL }, diff --git a/src/mesa/drivers/dri/mach64/mach64_context.c b/src/mesa/drivers/dri/mach64/mach64_context.c index 5a6c301da2..ad661e198c 100644 --- a/src/mesa/drivers/dri/mach64/mach64_context.c +++ b/src/mesa/drivers/dri/mach64/mach64_context.c @@ -58,6 +58,7 @@ #include "vblank.h" #define need_GL_ARB_multisample +#define need_GL_ARB_vertex_buffer_object #include "extension_helper.h" #ifndef MACH64_DEBUG @@ -83,6 +84,7 @@ const struct dri_extension card_extensions[] = { { "GL_ARB_multisample", GL_ARB_multisample_functions }, { "GL_ARB_multitexture", NULL }, + { "GL_ARB_vertex_buffer_object", GL_ARB_vertex_buffer_object_functions }, { "GL_EXT_texture_edge_clamp", NULL }, { "GL_MESA_ycbcr_texture", NULL }, { "GL_SGIS_generate_mipmap", NULL }, diff --git a/src/mesa/drivers/dri/mga/mga_xmesa.c b/src/mesa/drivers/dri/mga/mga_xmesa.c index 67a6f8bdf0..ca2c8faa18 100644 --- a/src/mesa/drivers/dri/mga/mga_xmesa.c +++ b/src/mesa/drivers/dri/mga/mga_xmesa.c @@ -72,6 +72,7 @@ #define need_GL_ARB_multisample #define need_GL_ARB_texture_compression +#define need_GL_ARB_vertex_buffer_object #define need_GL_ARB_vertex_program #define need_GL_EXT_fog_coord #define need_GL_EXT_multi_draw_arrays @@ -401,6 +402,7 @@ static const struct dri_extension card_extensions[] = { "GL_ARB_multisample", GL_ARB_multisample_functions }, { "GL_ARB_texture_compression", GL_ARB_texture_compression_functions }, { "GL_ARB_texture_rectangle", NULL }, + { "GL_ARB_vertex_buffer_object", GL_ARB_vertex_buffer_object_functions }, { "GL_EXT_blend_logic_op", NULL }, { "GL_EXT_fog_coord", GL_EXT_fog_coord_functions }, { "GL_EXT_multi_draw_arrays", GL_EXT_multi_draw_arrays_functions }, diff --git a/src/mesa/drivers/dri/r128/r128_context.c b/src/mesa/drivers/dri/r128/r128_context.c index 89ddafa02a..95e54a6af5 100644 --- a/src/mesa/drivers/dri/r128/r128_context.c +++ b/src/mesa/drivers/dri/r128/r128_context.c @@ -68,6 +68,7 @@ int R128_DEBUG = 0; #define need_GL_ARB_multisample #define need_GL_ARB_texture_compression +#define need_GL_ARB_vertex_buffer_object #define need_GL_EXT_blend_minmax #define need_GL_EXT_fog_coord #define need_GL_EXT_secondary_color @@ -80,6 +81,7 @@ const struct dri_extension card_extensions[] = { "GL_ARB_texture_compression", GL_ARB_texture_compression_functions }, { "GL_ARB_texture_env_add", NULL }, { "GL_ARB_texture_mirrored_repeat", NULL }, + { "GL_ARB_vertex_buffer_object", GL_ARB_vertex_buffer_object_functions }, { "GL_EXT_blend_subtract", GL_EXT_blend_minmax_functions }, { "GL_EXT_fog_coord", GL_EXT_fog_coord_functions }, { "GL_EXT_texture_edge_clamp", NULL }, diff --git a/src/mesa/drivers/dri/radeon/radeon_context.c b/src/mesa/drivers/dri/radeon/radeon_context.c index 0d25951b64..d7c2d1407d 100644 --- a/src/mesa/drivers/dri/radeon/radeon_context.c +++ b/src/mesa/drivers/dri/radeon/radeon_context.c @@ -64,6 +64,7 @@ WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. #define need_GL_ARB_multisample #define need_GL_ARB_texture_compression +#define need_GL_ARB_vertex_buffer_object #define need_GL_EXT_blend_minmax #define need_GL_EXT_fog_coord #define need_GL_EXT_secondary_color @@ -122,6 +123,7 @@ const struct dri_extension card_extensions[] = { "GL_ARB_texture_env_crossbar", NULL }, { "GL_ARB_texture_env_dot3", NULL }, { "GL_ARB_texture_mirrored_repeat", NULL }, + { "GL_ARB_vertex_buffer_object", GL_ARB_vertex_buffer_object_functions }, { "GL_EXT_blend_logic_op", NULL }, { "GL_EXT_blend_subtract", GL_EXT_blend_minmax_functions }, { "GL_EXT_fog_coord", GL_EXT_fog_coord_functions }, diff --git a/src/mesa/drivers/dri/savage/savage_xmesa.c b/src/mesa/drivers/dri/savage/savage_xmesa.c index f859217069..43422db9a8 100644 --- a/src/mesa/drivers/dri/savage/savage_xmesa.c +++ b/src/mesa/drivers/dri/savage/savage_xmesa.c @@ -61,6 +61,7 @@ #define need_GL_ARB_multisample #define need_GL_ARB_texture_compression +#define need_GL_ARB_vertex_buffer_object #define need_GL_EXT_secondary_color #include "extension_helper.h" @@ -135,6 +136,7 @@ static const struct dri_extension card_extensions[] = { "GL_ARB_multisample", GL_ARB_multisample_functions }, { "GL_ARB_multitexture", NULL }, { "GL_ARB_texture_compression", GL_ARB_texture_compression_functions }, + { "GL_ARB_vertex_buffer_object", GL_ARB_vertex_buffer_object_functions }, { "GL_EXT_stencil_wrap", NULL }, { "GL_EXT_texture_lod_bias", NULL }, { "GL_EXT_secondary_color", GL_EXT_secondary_color_functions }, diff --git a/src/mesa/drivers/dri/sis/sis_context.c b/src/mesa/drivers/dri/sis/sis_context.c index 89b81da347..b21df0a61e 100644 --- a/src/mesa/drivers/dri/sis/sis_context.c +++ b/src/mesa/drivers/dri/sis/sis_context.c @@ -60,6 +60,7 @@ USE OR OTHER DEALINGS IN THE SOFTWARE. #define need_GL_ARB_multisample #define need_GL_ARB_texture_compression +#define need_GL_ARB_vertex_buffer_object #define need_GL_EXT_fog_coord #define need_GL_EXT_secondary_color #include "extension_helper.h" @@ -79,6 +80,7 @@ struct dri_extension card_extensions[] = { "GL_ARB_texture_border_clamp", NULL }, { "GL_ARB_texture_compression", GL_ARB_texture_compression_functions }, { "GL_ARB_texture_mirrored_repeat", NULL }, + { "GL_ARB_vertex_buffer_object", GL_ARB_vertex_buffer_object_functions }, /*{ "GL_EXT_fog_coord", GL_EXT_fog_coord_functions },*/ { "GL_EXT_texture_lod_bias", NULL }, { "GL_EXT_secondary_color", GL_EXT_secondary_color_functions }, diff --git a/src/mesa/drivers/dri/unichrome/via_context.c b/src/mesa/drivers/dri/unichrome/via_context.c index bc5a414df6..89533421b4 100644 --- a/src/mesa/drivers/dri/unichrome/via_context.c +++ b/src/mesa/drivers/dri/unichrome/via_context.c @@ -64,6 +64,7 @@ #define need_GL_ARB_multisample #define need_GL_ARB_point_parameters +#define need_GL_ARB_vertex_buffer_object #define need_GL_EXT_fog_coord #define need_GL_EXT_secondary_color #include "extension_helper.h" @@ -372,6 +373,7 @@ const struct dri_extension card_extensions[] = { "GL_ARB_texture_env_combine", NULL }, /* { "GL_ARB_texture_env_dot3", NULL }, */ { "GL_ARB_texture_mirrored_repeat", NULL }, + { "GL_ARB_vertex_buffer_object", GL_ARB_vertex_buffer_object_functions }, { "GL_EXT_fog_coord", GL_EXT_fog_coord_functions }, { "GL_EXT_secondary_color", GL_EXT_secondary_color_functions }, { "GL_EXT_stencil_wrap", NULL }, -- cgit v1.2.3 From a6cc9ab493a2efa9a0ea91cddba0e85c8c8c83f1 Mon Sep 17 00:00:00 2001 From: Roland Scheidegger Date: Tue, 13 Mar 2007 19:04:28 +0100 Subject: sync up t_vp_build.c brw_vs_tnl.c a bit Bring over the optimizations for fog and normalized spot dir from t_vp_build.c to brw_vs_tnl.c. Likewise, port a fix for point size calc from brw_vs_tnl.c to t_vp_build.c (use ABS(eyez) instead of -eyez). Leave the now differing point size calcs alone though, not sure what's better (it's basically MOV, ABS, MUL, DP3 vs. ABS, MAD, MAD). --- src/mesa/drivers/dri/i965/brw_vs_tnl.c | 27 ++++++++++++--------------- src/mesa/tnl/t_vp_build.c | 12 +++++++----- 2 files changed, 19 insertions(+), 20 deletions(-) (limited to 'src') diff --git a/src/mesa/drivers/dri/i965/brw_vs_tnl.c b/src/mesa/drivers/dri/i965/brw_vs_tnl.c index c05a9b5ea1..e22a26b291 100644 --- a/src/mesa/drivers/dri/i965/brw_vs_tnl.c +++ b/src/mesa/drivers/dri/i965/brw_vs_tnl.c @@ -850,14 +850,13 @@ static struct ureg calculate_light_attenuation( struct tnl_program *p, /* Calculate spot attenuation: */ if (!p->state->unit[i].light_spotcutoff_is_180) { - struct ureg spot_dir = register_param3(p, STATE_LIGHT, i, - STATE_SPOT_DIRECTION); + struct ureg spot_dir_norm = register_param3(p, STATE_INTERNAL, + STATE_SPOT_DIR_NORMALIZED, i); struct ureg spot = get_temp(p); struct ureg slt = get_temp(p); - - emit_normalize_vec3( p, spot, spot_dir ); /* XXX: precompute! */ - emit_op2(p, OPCODE_DP3, spot, 0, ureg_negate(VPpli), spot); - emit_op2(p, OPCODE_SLT, slt, 0, swizzle1(spot_dir,W), spot); + + emit_op2(p, OPCODE_DP3, spot, 0, ureg_negate(VPpli), spot_dir_norm); + emit_op2(p, OPCODE_SLT, slt, 0, swizzle1(spot_dir_norm,W), spot); emit_op2(p, OPCODE_POW, spot, 0, spot, swizzle1(attenuation, W)); emit_op2(p, OPCODE_MUL, att, 0, slt, spot); @@ -1166,7 +1165,8 @@ static void build_fog( struct tnl_program *p ) if (p->state->fog_option && p->state->tnl_do_vertex_fog) { - struct ureg params = register_param1(p, STATE_FOG_PARAMS); + struct ureg params = register_param2(p, STATE_INTERNAL, + STATE_FOG_PARAMS_OPTIMIZED); struct ureg tmp = get_temp(p); struct ureg id = get_identity_param(p); @@ -1174,8 +1174,7 @@ static void build_fog( struct tnl_program *p ) switch (p->state->fog_option) { case FOG_LINEAR: { - emit_op1(p, OPCODE_ABS, tmp, 0, input); - emit_op2(p, OPCODE_SUB, tmp, 0, swizzle1(params,Z), tmp); + emit_op3(p, OPCODE_MAD, tmp, 0, input, swizzle1(params,X), swizzle1(params,Y)); emit_op2(p, OPCODE_MUL, tmp, 0, tmp, swizzle1(params,W)); emit_op2(p, OPCODE_MAX, tmp, 0, tmp, swizzle1(id,X)); /* saturate */ emit_op2(p, OPCODE_MIN, fog, WRITEMASK_X, tmp, swizzle1(id,W)); @@ -1183,15 +1182,13 @@ static void build_fog( struct tnl_program *p ) } case FOG_EXP: emit_op1(p, OPCODE_ABS, tmp, 0, input); - emit_op2(p, OPCODE_MUL, tmp, 0, tmp, swizzle1(params,X)); - emit_op2(p, OPCODE_POW, fog, WRITEMASK_X, - register_const1f(p, M_E), ureg_negate(tmp)); + emit_op2(p, OPCODE_MUL, tmp, 0, tmp, swizzle1(params,Z)); + emit_op1(p, OPCODE_EX2, fog, WRITEMASK_X, ureg_negate(tmp)); break; case FOG_EXP2: - emit_op2(p, OPCODE_MUL, tmp, 0, input, swizzle1(params,X)); + emit_op2(p, OPCODE_MUL, tmp, 0, input, swizzle1(params,W)); emit_op2(p, OPCODE_MUL, tmp, 0, tmp, tmp); - emit_op2(p, OPCODE_POW, fog, WRITEMASK_X, - register_const1f(p, M_E), ureg_negate(tmp)); + emit_op1(p, OPCODE_EX2, fog, WRITEMASK_X, ureg_negate(tmp)); break; } diff --git a/src/mesa/tnl/t_vp_build.c b/src/mesa/tnl/t_vp_build.c index 2a97a1c8c7..2663d99304 100644 --- a/src/mesa/tnl/t_vp_build.c +++ b/src/mesa/tnl/t_vp_build.c @@ -1329,14 +1329,16 @@ static void build_pointsize( struct tnl_program *p ) struct ureg out = register_output(p, VERT_RESULT_PSIZ); struct ureg ut = get_temp(p); + /* dist = |eyez| */ + emit_op1(p, OPCODE_ABS, ut, WRITEMASK_Y, swizzle1(eye, Z)); /* p1 + dist * (p2 + dist * p3); */ - emit_op3(p, OPCODE_MAD, ut, 0, negate(swizzle1(eye, Z)), + emit_op3(p, OPCODE_MAD, ut, WRITEMASK_X, swizzle1(ut, Y), swizzle1(state_attenuation, Z), swizzle1(state_attenuation, Y)); - emit_op3(p, OPCODE_MAD, ut, 0, negate(swizzle1(eye, Z)), + emit_op3(p, OPCODE_MAD, ut, WRITEMASK_X, swizzle1(ut, Y), ut, swizzle1(state_attenuation, X)); /* 1 / sqrt(factor) */ - emit_op1(p, OPCODE_RSQ, ut, 0, ut ); + emit_op1(p, OPCODE_RSQ, ut, WRITEMASK_X, ut ); #if 1 /* out = pointSize / sqrt(factor) */ @@ -1344,8 +1346,8 @@ static void build_pointsize( struct tnl_program *p ) #else /* not sure, might make sense to do clamping here, but it's not done in t_vb_points neither */ - emit_op2(p, OPCODE_MUL, ut, 0, ut, state_size); - emit_op2(p, OPCODE_MAX, ut, 0, ut, swizzle1(state_size, Y)); + emit_op2(p, OPCODE_MUL, ut, WRITEMASK_X, ut, state_size); + emit_op2(p, OPCODE_MAX, ut, WRITEMASK_X, ut, swizzle1(state_size, Y)); emit_op2(p, OPCODE_MIN, out, WRITEMASK_X, ut, swizzle1(state_size, Z)); #endif -- cgit v1.2.3