Loongson3's extention instructions (prefixed with gs) are widely used
in our MMI codebase. However, these instructions are not avilable on
Loongson-2E/F while MMI code should work on these processors.
Previously we introduced mmiutils marcos to provide backward compactbility
but newly commited code didn't follow that. In this patch I revised the
codebase and converted all these instructions into MMI marcos to get
Loongson2 supproted again.
Signed-off-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
Reviewed-by: Shiyou Yin <yinshiyou-hf@loongson.cn>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Clang is more strict on the type of asm operands, float or double
type variable should use constraint 'f', integer variable should
use constraint 'r'.
Signed-off-by: Jin Bo <jinbo@loongson.cn>
Reviewed-by: yinshiyou-hf@loongson.cn
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
1.'xor,or,and' to 'pxor,por,pand'. In the case of operating FPR,
gcc supports both of them, clang only supports the second type.
2.'dsrl,srl' to 'ssrld,ssrlw'. In the case of operating FPR, gcc
supports both of them, clang only supports the second type.
Signed-off-by: Jin Bo <jinbo@loongson.cn>
Reviewed-by: yinshiyou-hf@loongson.cn
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
GCC complains:
warning: listing the stack pointer register ‘$29’ in a clobber
list is deprecated [-Wdeprecated]
Actually stack pointer was restored at the end of the inline assembly
so there is no reason to add it to the clobber list.
Also use $sp insted of $29 to make our intention much more clear.
Signed-off-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
Reviewed-by: Shiyou Yin <yinshiyou-hf@loongson.cn>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
In function ff_h264_add_pixels4_8_mmi, there is no need to reset '%[ftmp0]'
to 0, because it's value has never changed since the start of the asm block.
This patch remove the redundant 'xor' and set src to zero once it was loaded.
In function ff_h264_idct_add_8_mmi, 'block' is seted to zero twice.
This patch removed the first setting zero operation and move the second one
after the load operation of block.
In function ff_h264_idct8_add_8_mmi, 'block' is seted to zero twice too.
This patch just removed the second setting zero operation.
This patch mainly simplifies the implementation of functions above,
the effect on the performance of whole h264 decoding process is not obvious.
According to the perf data, proportion of ff_h264_idct_add_8_mmi decreased from
0.29% to 0.26% and ff_h264_idct8_add_8_mmi decreased from 0.62% to 0.59% when decoding
H264 format on loongson 3A3000(For reference only , not very stable.).
Reviewed-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Optimized memset with mmi in following functions:
1. ff_h264_add_pixels4_8_mmi.
2. ff_h264_idct_add_8_mmi.
3. ff_h264_idct8_add_8_mmi.
This optimization improved h264 decoding performance about 1.3%(tested on loongson 3A3000).
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
1.MMI_ load/store macros are defined in libavutil/mips/mmiutils.h
2.Replace some unnecessary unaligned access with aligned operator
3.The MMI_ load/store is compatible with cpu loongson2e/2f which not support instructions start with gs
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
1. no longer use the register names directly and optimized code format
2. to be compatible with O32, specify type of address variable with mips_reg and handle the address variable with PTR_ operator
3. optimize some unaligned loads and stores
4. use uld and mtc1 to workaround cpu 3A2000 gslwlc1 bug (gslwlc1 instruction extension bug in O32 ABI)
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
