gstreamer/gst/volume/gstvolumeorc.orc


.init gst_volume_orc_init

.function orc_scalarmultiply_f64_ns
.dest 8 d1 double
.doubleparam 8 p1

muld d1, d1, p1

.function orc_scalarmultiply_f32_ns
.dest 4 d1 float
.floatparam 4 p1

mulf d1, d1, p1

.function orc_process_int32
.dest 4 d1 gint32
.param 4 p1
.temp 8 t1

mulslq t1, d1, p1
shrsq t1, t1, 27
convql d1, t1

.function orc_process_int32_clamp
.dest 4 d1 gint32
.param 4 p1
.temp 8 t1

mulslq t1, d1, p1
shrsq t1, t1, 27
convsssql d1, t1

.function orc_process_int16
.dest 2 d1 gint16
.param 2 p1
.temp 4 t1

mulswl t1, d1, p1
shrsl t1, t1, 11
convlw d1, t1


.function orc_process_int16_clamp
.dest 2 d1 gint16
.param 2 p1
.temp 4 t1

mulswl t1, d1, p1
shrsl t1, t1, 11
convssslw d1, t1

.function orc_process_int8
.dest 1 d1 gint8
.param 1 p1
.temp 2 t1

mulsbw t1, d1, p1
shrsw t1, t1, 3
convwb d1, t1


.function orc_process_int8_clamp
.dest 1 d1 gint8
.param 1 p1
.temp 2 t1

mulsbw t1, d1, p1
shrsw t1, t1, 3
convssswb d1, t1

.function orc_memset_f64
.dest 8 d1 gdouble
.doubleparam 8 p1

copyq d1, p1

.function orc_prepare_volumes
.dest 8 d1 gdouble
.source 4 s1 gboolean
.temp 8 t1

convld t1, s1
subd t1, 0x3FF0000000000000L, t1
muld d1, d1, t1

.function orc_process_controlled_f64_1ch
.dest 8 d1 gdouble
.source 8 s1 gdouble

muld d1, d1, s1

.function orc_process_controlled_f32_1ch
.dest 4 d1 gfloat
.source 8 s1 gdouble
.temp 4 t1

convdf t1, s1
mulf d1, d1, t1

.function orc_process_controlled_f32_2ch
.dest 8 d1 gfloat
.source 8 s1 gdouble
.temp 4 t1
.temp 8 t2

convdf t1, s1
mergelq t2, t1, t1
x2 mulf d1, d1, t2

.function orc_process_controlled_int32_1ch
.dest 4 d1 gint32
.source 8 s1 gdouble
.temp 8 t1

convld t1, d1
muld t1, t1, s1
convdl d1, t1

.function orc_process_controlled_int16_1ch
.dest 2 d1 gint16
.source 8 s1 gdouble
.temp 4 t1
.temp 4 t2

convswl t1, d1
convlf t1, t1
convdf t2, s1
mulf t1, t1, t2
convfl t1, t1
convssslw d1, t1

.function orc_process_controlled_int16_2ch
.dest 4 d1 gint16
.source 8 s1 gdouble
.temp 8 t1
.temp 4 t2
.temp 8 t3

x2 convswl t1, d1
x2 convlf t1, t1
convdf t2, s1
mergelq t3, t2, t2
x2 mulf t3, t3, t1
x2 convfl t3, t3
x2 convssslw d1, t3

.function orc_process_controlled_int8_1ch
.dest 1 d1 gint8
.source 8 s1 gdouble
.temp 2 t1
.temp 4 t2
.temp 4 t3

convsbw t1, d1
convswl t2, t1
convlf t2, t2
convdf t3, s1
mulf t2, t2, t3
convfl t2, t2
convlw t1, t2
convssswb d1, t1

.function orc_process_controlled_int8_2ch
.dest 2 d1 gint8
.source 8 s1 gdouble
.temp 4 t1
.temp 8 t2
.temp 8 t3

x2 convsbw t1, d1
x2 convswl t2, t1
x2 convlf t2, t2
convdf t1, s1
mergelq t3, t1, t1
x2 mulf t2, t2, t3
x2 convfl t2, t2
x2 convlw t1, t2
x2 convssswb d1, t1
volume: convert from liboil to orc 2010-06-07 06:48:15 +00:00
Revert "Revert "Use init functions for Orc code"" This reverts commit 93aa13639d74449dc68296427e5dbcfe8aca5f51. Everything should work now after regenerating the disted source files. 2010-09-09 08:57:41 +00:00			`.init gst_volume_orc_init`

volume: Implement f64 scaling with orc This requires orc 0.4.10 2010-10-01 08:44:37 +00:00			`.function orc_scalarmultiply_f64_ns`
			`.dest 8 d1 double`
volume: Update for orc changes double parameters are declared with .doubleparam now. 2010-10-07 21:54:25 +00:00			`.doubleparam 8 p1`
volume: Implement f64 scaling with orc This requires orc 0.4.10 2010-10-01 08:44:37 +00:00
			`muld d1, d1, p1`

volume: enable ORC for float in volume 2010-08-25 16:01:57 +00:00			`.function orc_scalarmultiply_f32_ns`
			`.dest 4 d1 float`
			`.floatparam 4 p1`
volume: convert from liboil to orc 2010-06-07 06:48:15 +00:00
volume: Improve f32 scaling by using only a single array Passing the same array as dest and src is invalid anyway because they're maked with the restrict qualifier. 2010-10-01 08:42:27 +00:00			`mulf d1, d1, p1`
volume: convert from liboil to orc 2010-06-07 06:48:15 +00:00
volume: Implement int32 processing with orc 2010-10-03 09:20:37 +00:00			`.function orc_process_int32`
			`.dest 4 d1 gint32`
			`.param 4 p1`
			`.temp 8 t1`

			`mulslq t1, d1, p1`
			`shrsq t1, t1, 27`
			`convql d1, t1`

			`.function orc_process_int32_clamp`
			`.dest 4 d1 gint32`
			`.param 4 p1`
			`.temp 8 t1`

			`mulslq t1, d1, p1`
			`shrsq t1, t1, 27`
			`convsssql d1, t1`
volume: convert from liboil to orc 2010-06-07 06:48:15 +00:00
			`.function orc_process_int16`
			`.dest 2 d1 gint16`
			`.param 2 p1`
			`.temp 4 t1`

			`mulswl t1, d1, p1`
volume: Fix handling of volume>=4.0 for 8 and 16 bit integer formats Also add a unit test for this. Previously volumes bigger than 4.0 would have resulted in overflows in the fixed point processing. Fixes bug #649642. 2011-05-31 09:05:03 +00:00			`shrsl t1, t1, 11`
volume: convert from liboil to orc 2010-06-07 06:48:15 +00:00			`convlw d1, t1`


			`.function orc_process_int16_clamp`
			`.dest 2 d1 gint16`
			`.param 2 p1`
			`.temp 4 t1`

			`mulswl t1, d1, p1`
volume: Fix handling of volume>=4.0 for 8 and 16 bit integer formats Also add a unit test for this. Previously volumes bigger than 4.0 would have resulted in overflows in the fixed point processing. Fixes bug #649642. 2011-05-31 09:05:03 +00:00			`shrsl t1, t1, 11`
volume: convert from liboil to orc 2010-06-07 06:48:15 +00:00			`convssslw d1, t1`

			`.function orc_process_int8`
			`.dest 1 d1 gint8`
			`.param 1 p1`
			`.temp 2 t1`

			`mulsbw t1, d1, p1`
volume: Fix handling of volume>=4.0 for 8 and 16 bit integer formats Also add a unit test for this. Previously volumes bigger than 4.0 would have resulted in overflows in the fixed point processing. Fixes bug #649642. 2011-05-31 09:05:03 +00:00			`shrsw t1, t1, 3`
volume: convert from liboil to orc 2010-06-07 06:48:15 +00:00			`convwb d1, t1`


			`.function orc_process_int8_clamp`
			`.dest 1 d1 gint8`
			`.param 1 p1`
			`.temp 2 t1`

			`mulsbw t1, d1, p1`
volume: Fix handling of volume>=4.0 for 8 and 16 bit integer formats Also add a unit test for this. Previously volumes bigger than 4.0 would have resulted in overflows in the fixed point processing. Fixes bug #649642. 2011-05-31 09:05:03 +00:00			`shrsw t1, t1, 3`
volume: convert from liboil to orc 2010-06-07 06:48:15 +00:00			`convssswb d1, t1`

volume: Convert parts of the controlled processing to orc 2010-10-01 09:00:54 +00:00			`.function orc_memset_f64`
			`.dest 8 d1 gdouble`
volume: Update for orc changes double parameters are declared with .doubleparam now. 2010-10-07 21:54:25 +00:00			`.doubleparam 8 p1`
volume: Convert parts of the controlled processing to orc 2010-10-01 09:00:54 +00:00
			`copyq d1, p1`

			`.function orc_prepare_volumes`
			`.dest 8 d1 gdouble`
			`.source 4 s1 gboolean`
			`.temp 8 t1`

			`convld t1, s1`
			`subd t1, 0x3FF0000000000000L, t1`
			`muld d1, d1, t1`
volume: convert from liboil to orc 2010-06-07 06:48:15 +00:00
volume: Implement controlled processing for f64/1ch and f32/1-2ch in orc 2010-10-01 09:13:01 +00:00			`.function orc_process_controlled_f64_1ch`
			`.dest 8 d1 gdouble`
			`.source 8 s1 gdouble`

			`muld d1, d1, s1`

			`.function orc_process_controlled_f32_1ch`
			`.dest 4 d1 gfloat`
			`.source 8 s1 gdouble`
			`.temp 4 t1`

			`convdf t1, s1`
			`mulf d1, d1, t1`

			`.function orc_process_controlled_f32_2ch`
			`.dest 8 d1 gfloat`
			`.source 8 s1 gdouble`
			`.temp 4 t1`
			`.temp 8 t2`

			`convdf t1, s1`
			`mergelq t2, t1, t1`
			`x2 mulf d1, d1, t2`
volume: Implement controlled processing for int16/1-2ch and int8/1,2,4ch with orc 2010-10-01 10:21:52 +00:00
volume: Implement int32 processing with orc 2010-10-03 09:20:37 +00:00			`.function orc_process_controlled_int32_1ch`
			`.dest 4 d1 gint32`
			`.source 8 s1 gdouble`
			`.temp 8 t1`

volume: Fix unit test failure for the controlled processing functions Going over integer arithmetic will lead to minimal rounding errors, leading to +/-1 changes for volume==1.0. Implement the controlled processing with floating point arithmetic, which was already done for the C versions anyway. 2010-10-08 09:49:09 +00:00			`convld t1, d1`
			`muld t1, t1, s1`
			`convdl d1, t1`
volume: Implement int32 processing with orc 2010-10-03 09:20:37 +00:00
volume: Implement controlled processing for int16/1-2ch and int8/1,2,4ch with orc 2010-10-01 10:21:52 +00:00			`.function orc_process_controlled_int16_1ch`
			`.dest 2 d1 gint16`
			`.source 8 s1 gdouble`
volume: Fix unit test failure for the controlled processing functions Going over integer arithmetic will lead to minimal rounding errors, leading to +/-1 changes for volume==1.0. Implement the controlled processing with floating point arithmetic, which was already done for the C versions anyway. 2010-10-08 09:49:09 +00:00			`.temp 4 t1`
volume: Implement controlled processing for int16/1-2ch and int8/1,2,4ch with orc 2010-10-01 10:21:52 +00:00			`.temp 4 t2`

volume: Fix unit test failure for the controlled processing functions Going over integer arithmetic will lead to minimal rounding errors, leading to +/-1 changes for volume==1.0. Implement the controlled processing with floating point arithmetic, which was already done for the C versions anyway. 2010-10-08 09:49:09 +00:00			`convswl t1, d1`
			`convlf t1, t1`
			`convdf t2, s1`
			`mulf t1, t1, t2`
			`convfl t1, t1`
			`convssslw d1, t1`
volume: Implement controlled processing for int16/1-2ch and int8/1,2,4ch with orc 2010-10-01 10:21:52 +00:00
			`.function orc_process_controlled_int16_2ch`
			`.dest 4 d1 gint16`
			`.source 8 s1 gdouble`
			`.temp 8 t1`
			`.temp 4 t2`
volume: Fix unit test failure for the controlled processing functions Going over integer arithmetic will lead to minimal rounding errors, leading to +/-1 changes for volume==1.0. Implement the controlled processing with floating point arithmetic, which was already done for the C versions anyway. 2010-10-08 09:49:09 +00:00			`.temp 8 t3`
volume: Implement controlled processing for int16/1-2ch and int8/1,2,4ch with orc 2010-10-01 10:21:52 +00:00
volume: Fix unit test failure for the controlled processing functions Going over integer arithmetic will lead to minimal rounding errors, leading to +/-1 changes for volume==1.0. Implement the controlled processing with floating point arithmetic, which was already done for the C versions anyway. 2010-10-08 09:49:09 +00:00			`x2 convswl t1, d1`
			`x2 convlf t1, t1`
			`convdf t2, s1`
			`mergelq t3, t2, t2`
			`x2 mulf t3, t3, t1`
			`x2 convfl t3, t3`
			`x2 convssslw d1, t3`
volume: Implement controlled processing for int16/1-2ch and int8/1,2,4ch with orc 2010-10-01 10:21:52 +00:00
			`.function orc_process_controlled_int8_1ch`
			`.dest 1 d1 gint8`
			`.source 8 s1 gdouble`
volume: Fix unit test failure for the controlled processing functions Going over integer arithmetic will lead to minimal rounding errors, leading to +/-1 changes for volume==1.0. Implement the controlled processing with floating point arithmetic, which was already done for the C versions anyway. 2010-10-08 09:49:09 +00:00			`.temp 2 t1`
volume: Implement controlled processing for int16/1-2ch and int8/1,2,4ch with orc 2010-10-01 10:21:52 +00:00			`.temp 4 t2`
volume: Fix unit test failure for the controlled processing functions Going over integer arithmetic will lead to minimal rounding errors, leading to +/-1 changes for volume==1.0. Implement the controlled processing with floating point arithmetic, which was already done for the C versions anyway. 2010-10-08 09:49:09 +00:00			`.temp 4 t3`

			`convsbw t1, d1`
			`convswl t2, t1`
			`convlf t2, t2`
			`convdf t3, s1`
			`mulf t2, t2, t3`
			`convfl t2, t2`
			`convlw t1, t2`
			`convssswb d1, t1`
volume: Implement controlled processing for int16/1-2ch and int8/1,2,4ch with orc 2010-10-01 10:21:52 +00:00
			`.function orc_process_controlled_int8_2ch`
			`.dest 2 d1 gint8`
			`.source 8 s1 gdouble`
volume: Fix unit test failure for the controlled processing functions Going over integer arithmetic will lead to minimal rounding errors, leading to +/-1 changes for volume==1.0. Implement the controlled processing with floating point arithmetic, which was already done for the C versions anyway. 2010-10-08 09:49:09 +00:00			`.temp 4 t1`
			`.temp 8 t2`
			`.temp 8 t3`

			`x2 convsbw t1, d1`
			`x2 convswl t2, t1`
			`x2 convlf t2, t2`
			`convdf t1, s1`
			`mergelq t3, t1, t1`
			`x2 mulf t2, t2, t3`
			`x2 convfl t2, t2`
			`x2 convlw t1, t2`
			`x2 convssswb d1, t1`
volume: Implement controlled processing for int16/1-2ch and int8/1,2,4ch with orc 2010-10-01 10:21:52 +00:00