[FFmpeg-devel] [PATCH] mips/float_dsp: fix a bug in vector_fmul_window_mips
Nedeljko Babic
nedeljko.babic at imgtec.com
Thu Mar 5 12:23:59 CET 2015
Loop was unrolled eight times although in heder there is assumption
that len is multiple of 4.
This is fixed, and assembly code is rewritten to be more optimal and
to simplify clobber list.
Signed-off-by: Nedeljko Babic <nedeljko.babic at imgtec.com>
---
libavutil/mips/float_dsp_mips.c | 180 +++++++++++++++++-----------------------
1 file changed, 74 insertions(+), 106 deletions(-)
diff --git a/libavutil/mips/float_dsp_mips.c b/libavutil/mips/float_dsp_mips.c
index 06d52dc..a54f959 100644
--- a/libavutil/mips/float_dsp_mips.c
+++ b/libavutil/mips/float_dsp_mips.c
@@ -144,114 +144,82 @@ static void vector_fmul_scalar_mips(float *dst, const float *src, float mul,
}
static void vector_fmul_window_mips(float *dst, const float *src0,
- const float *src1, const float *win, int len)
+ const float *src1, const float *win, int len)
{
- int i, j;
- /*
- * variables used in inline assembler
- */
- float * dst_i, * dst_j, * dst_i2, * dst_j2;
- float temp, temp1, temp2, temp3, temp4, temp5, temp6, temp7;
+ float * dst_j, *win_j, *src0_i, *src1_j, *dst_i, *win_i;
+ float temp, temp1, temp2, temp3;
+ float s0, s01, s1, s11;
+ float wi, wi1, wi2, wi3;
+ float wj, wj1, wj2, wj3;
+ const float * lp_end = win + len;
+
+ win_i = (float *)win;
+ win_j = (float *)(win + 2 * len -1);
+ src1_j = (float *)(src1 + len - 1);
+ src0_i = (float *)src0;
+ dst_i = (float *)dst;
+ dst_j = (float *)(dst + 2 * len -1);
- dst += len;
- win += len;
- src0 += len;
-
- for (i = -len, j = len - 1; i < 0; i += 8, j -= 8) {
-
- dst_i = dst + i;
- dst_j = dst + j;
-
- dst_i2 = dst + i + 4;
- dst_j2 = dst + j - 4;
-
- __asm__ volatile (
- "mul.s %[temp], %[s1], %[wi] \n\t"
- "mul.s %[temp1], %[s1], %[wj] \n\t"
- "mul.s %[temp2], %[s11], %[wi1] \n\t"
- "mul.s %[temp3], %[s11], %[wj1] \n\t"
-
- "msub.s %[temp], %[temp], %[s0], %[wj] \n\t"
- "madd.s %[temp1], %[temp1], %[s0], %[wi] \n\t"
- "msub.s %[temp2], %[temp2], %[s01], %[wj1] \n\t"
- "madd.s %[temp3], %[temp3], %[s01], %[wi1] \n\t"
-
- "swc1 %[temp], 0(%[dst_i]) \n\t" /* dst[i] = s0*wj - s1*wi; */
- "swc1 %[temp1], 0(%[dst_j]) \n\t" /* dst[j] = s0*wi + s1*wj; */
- "swc1 %[temp2], 4(%[dst_i]) \n\t" /* dst[i+1] = s01*wj1 - s11*wi1; */
- "swc1 %[temp3], -4(%[dst_j]) \n\t" /* dst[j-1] = s01*wi1 + s11*wj1; */
-
- "mul.s %[temp4], %[s12], %[wi2] \n\t"
- "mul.s %[temp5], %[s12], %[wj2] \n\t"
- "mul.s %[temp6], %[s13], %[wi3] \n\t"
- "mul.s %[temp7], %[s13], %[wj3] \n\t"
-
- "msub.s %[temp4], %[temp4], %[s02], %[wj2] \n\t"
- "madd.s %[temp5], %[temp5], %[s02], %[wi2] \n\t"
- "msub.s %[temp6], %[temp6], %[s03], %[wj3] \n\t"
- "madd.s %[temp7], %[temp7], %[s03], %[wi3] \n\t"
-
- "swc1 %[temp4], 8(%[dst_i]) \n\t" /* dst[i+2] = s02*wj2 - s12*wi2; */
- "swc1 %[temp5], -8(%[dst_j]) \n\t" /* dst[j-2] = s02*wi2 + s12*wj2; */
- "swc1 %[temp6], 12(%[dst_i]) \n\t" /* dst[i+2] = s03*wj3 - s13*wi3; */
- "swc1 %[temp7], -12(%[dst_j]) \n\t" /* dst[j-3] = s03*wi3 + s13*wj3; */
- : [temp]"=&f"(temp), [temp1]"=&f"(temp1), [temp2]"=&f"(temp2),
- [temp3]"=&f"(temp3), [temp4]"=&f"(temp4), [temp5]"=&f"(temp5),
- [temp6]"=&f"(temp6), [temp7]"=&f"(temp7)
- : [dst_j]"r"(dst_j), [dst_i]"r" (dst_i),
- [s0] "f"(src0[i]), [wj] "f"(win[j]), [s1] "f"(src1[j]),
- [wi] "f"(win[i]), [s01]"f"(src0[i + 1]),[wj1]"f"(win[j - 1]),
- [s11]"f"(src1[j - 1]), [wi1]"f"(win[i + 1]), [s02]"f"(src0[i + 2]),
- [wj2]"f"(win[j - 2]), [s12]"f"(src1[j - 2]),[wi2]"f"(win[i + 2]),
- [s03]"f"(src0[i + 3]), [wj3]"f"(win[j - 3]), [s13]"f"(src1[j - 3]),
- [wi3]"f"(win[i + 3])
- : "memory"
- );
-
- __asm__ volatile (
- "mul.s %[temp], %[s1], %[wi] \n\t"
- "mul.s %[temp1], %[s1], %[wj] \n\t"
- "mul.s %[temp2], %[s11], %[wi1] \n\t"
- "mul.s %[temp3], %[s11], %[wj1] \n\t"
-
- "msub.s %[temp], %[temp], %[s0], %[wj] \n\t"
- "madd.s %[temp1], %[temp1], %[s0], %[wi] \n\t"
- "msub.s %[temp2], %[temp2], %[s01], %[wj1] \n\t"
- "madd.s %[temp3], %[temp3], %[s01], %[wi1] \n\t"
-
- "swc1 %[temp], 0(%[dst_i2]) \n\t" /* dst[i] = s0*wj - s1*wi; */
- "swc1 %[temp1], 0(%[dst_j2]) \n\t" /* dst[j] = s0*wi + s1*wj; */
- "swc1 %[temp2], 4(%[dst_i2]) \n\t" /* dst[i+1] = s01*wj1 - s11*wi1; */
- "swc1 %[temp3], -4(%[dst_j2]) \n\t" /* dst[j-1] = s01*wi1 + s11*wj1; */
-
- "mul.s %[temp4], %[s12], %[wi2] \n\t"
- "mul.s %[temp5], %[s12], %[wj2] \n\t"
- "mul.s %[temp6], %[s13], %[wi3] \n\t"
- "mul.s %[temp7], %[s13], %[wj3] \n\t"
-
- "msub.s %[temp4], %[temp4], %[s02], %[wj2] \n\t"
- "madd.s %[temp5], %[temp5], %[s02], %[wi2] \n\t"
- "msub.s %[temp6], %[temp6], %[s03], %[wj3] \n\t"
- "madd.s %[temp7], %[temp7], %[s03], %[wi3] \n\t"
-
- "swc1 %[temp4], 8(%[dst_i2]) \n\t" /* dst[i+2] = s02*wj2 - s12*wi2; */
- "swc1 %[temp5], -8(%[dst_j2]) \n\t" /* dst[j-2] = s02*wi2 + s12*wj2; */
- "swc1 %[temp6], 12(%[dst_i2]) \n\t" /* dst[i+2] = s03*wj3 - s13*wi3; */
- "swc1 %[temp7], -12(%[dst_j2]) \n\t" /* dst[j-3] = s03*wi3 + s13*wj3; */
- : [temp]"=&f"(temp),
- [temp1]"=&f"(temp1), [temp2]"=&f"(temp2), [temp3]"=&f"(temp3),
- [temp4]"=&f"(temp4), [temp5]"=&f"(temp5), [temp6]"=&f"(temp6),
- [temp7] "=&f" (temp7)
- : [dst_j2]"r"(dst_j2), [dst_i2]"r"(dst_i2),
- [s0] "f"(src0[i + 4]), [wj] "f"(win[j - 4]), [s1] "f"(src1[j - 4]),
- [wi] "f"(win[i + 4]), [s01]"f"(src0[i + 5]),[wj1]"f"(win[j - 5]),
- [s11]"f"(src1[j - 5]), [wi1]"f"(win[i + 5]), [s02]"f"(src0[i + 6]),
- [wj2]"f"(win[j - 6]), [s12]"f"(src1[j - 6]),[wi2]"f"(win[i + 6]),
- [s03]"f"(src0[i + 7]), [wj3]"f"(win[j - 7]), [s13]"f"(src1[j - 7]),
- [wi3]"f"(win[i + 7])
- : "memory"
- );
- }
+ /* loop unrolled 4 times */
+ __asm__ volatile (
+ "1:"
+ "lwc1 %[s1], 0(%[src1_j]) \n\t"
+ "lwc1 %[wi], 0(%[win_i]) \n\t"
+ "lwc1 %[wj], 0(%[win_j]) \n\t"
+ "lwc1 %[s11], -4(%[src1_j]) \n\t"
+ "lwc1 %[wi1], 4(%[win_i]) \n\t"
+ "lwc1 %[wj1], -4(%[win_j]) \n\t"
+ "lwc1 %[s0], 0(%[src0_i]) \n\t"
+ "lwc1 %[s01], 4(%[src0_i]) \n\t"
+ "mul.s %[temp], %[s1], %[wi] \n\t"
+ "mul.s %[temp1], %[s1], %[wj] \n\t"
+ "mul.s %[temp2], %[s11], %[wi1] \n\t"
+ "mul.s %[temp3], %[s11], %[wj1] \n\t"
+ "lwc1 %[s1], -8(%[src1_j]) \n\t"
+ "lwc1 %[wi2], 8(%[win_i]) \n\t"
+ "lwc1 %[wj2], -8(%[win_j]) \n\t"
+ "lwc1 %[s11], -12(%[src1_j]) \n\t"
+ "msub.s %[temp], %[temp], %[s0], %[wj] \n\t"
+ "madd.s %[temp1], %[temp1], %[s0], %[wi] \n\t"
+ "msub.s %[temp2], %[temp2], %[s01], %[wj1] \n\t"
+ "madd.s %[temp3], %[temp3], %[s01], %[wi1] \n\t"
+ "lwc1 %[wi3], 12(%[win_i]) \n\t"
+ "lwc1 %[wj3], -12(%[win_j]) \n\t"
+ "lwc1 %[s0], 8(%[src0_i]) \n\t"
+ "lwc1 %[s01], 12(%[src0_i]) \n\t"
+ "addiu %[src1_j],-16 \n\t"
+ "addiu %[win_i], 16 \n\t"
+ "addiu %[win_j], -16 \n\t"
+ "addiu %[src0_i], 16 \n\t"
+ "swc1 %[temp], 0(%[dst_i]) \n\t" /* dst[i] = s0*wj - s1*wi; */
+ "swc1 %[temp1], 0(%[dst_j]) \n\t" /* dst[j] = s0*wi + s1*wj; */
+ "swc1 %[temp2], 4(%[dst_i]) \n\t" /* dst[i+1] = s01*wj1 - s11*wi1; */
+ "swc1 %[temp3], -4(%[dst_j]) \n\t" /* dst[j-1] = s01*wi1 + s11*wj1; */
+ "mul.s %[temp], %[s1], %[wi2] \n\t"
+ "mul.s %[temp1], %[s1], %[wj2] \n\t"
+ "mul.s %[temp2], %[s11], %[wi3] \n\t"
+ "mul.s %[temp3], %[s11], %[wj3] \n\t"
+ "msub.s %[temp], %[temp], %[s0], %[wj2] \n\t"
+ "madd.s %[temp1], %[temp1], %[s0], %[wi2] \n\t"
+ "msub.s %[temp2], %[temp2], %[s01], %[wj3] \n\t"
+ "madd.s %[temp3], %[temp3], %[s01], %[wi3] \n\t"
+ "swc1 %[temp], 8(%[dst_i]) \n\t" /* dst[i+2] = s0*wj2 - s1*wi2; */
+ "swc1 %[temp1], -8(%[dst_j]) \n\t" /* dst[j-2] = s0*wi2 + s1*wj2; */
+ "swc1 %[temp2], 12(%[dst_i]) \n\t" /* dst[i+2] = s01*wj3 - s11*wi3; */
+ "swc1 %[temp3], -12(%[dst_j]) \n\t" /* dst[j-3] = s01*wi3 + s11*wj3; */
+ "addiu %[dst_i], 16 \n\t"
+ "addiu %[dst_j], -16 \n\t"
+ "bne %[win_i], %[lp_end], 1b \n\t"
+ : [temp]"=&f"(temp), [temp1]"=&f"(temp1), [temp2]"=&f"(temp2),
+ [temp3]"=&f"(temp3), [src0_i]"+r"(src0_i), [win_i]"+r"(win_i),
+ [src1_j]"+r"(src1_j), [win_j]"+r"(win_j), [dst_i]"+r"(dst_i),
+ [dst_j]"+r"(dst_j), [s0] "=&f"(s0), [s01]"=&f"(s01), [s1] "=&f"(s1),
+ [s11]"=&f"(s11), [wi] "=&f"(wi), [wj] "=&f"(wj), [wi2]"=&f"(wi2),
+ [wj2]"=&f"(wj2), [wi3]"=&f"(wi3), [wj3]"=&f"(wj3), [wi1]"=&f"(wi1),
+ [wj1]"=&f"(wj1)
+ : [lp_end]"r"(lp_end)
+ : "memory"
+ );
}
static void butterflies_float_mips(float *av_restrict v1, float *av_restrict v2,
--
1.8.2.1
More information about the ffmpeg-devel
mailing list