Also fix an effective-type violation. Exposed by https://fate.ffmpeg.org/report.cgi?time=20240312011016&slot=ppc-linux-gcc-13.2-ubsan-altivec-qemu
Showing the changes to libavcodec (1 of 1 file). Show all
Unified · Split
| @@ -401,30 +401,29 @@ static inline void write16x4(uint8_t *dst, int dst_stride, · 400 unmodified lines | |||
| 401 | register vec_u8 r0, register vec_u8 r1, | 401 | register vec_u8 r0, register vec_u8 r1, |
| 402 | register vec_u8 r2, register vec_u8 r3) { | 402 | register vec_u8 r2, register vec_u8 r3) { |
| 403 | DECLARE_ALIGNED(16, unsigned char, result)[64]; | 403 | DECLARE_ALIGNED(16, unsigned char, result)[64]; |
| 404 | uint32_t *src_int = (uint32_t *)result, *dst_int = (uint32_t *)dst; | 404 | uint32_t *src_int = (uint32_t *)result; |
| 405 | int int_dst_stride = dst_stride/4; | ||
| 406 | 405 | ||
| 407 | vec_st(r0, 0, result); | 406 | vec_st(r0, 0, result); |
| 408 | vec_st(r1, 16, result); | 407 | vec_st(r1, 16, result); |
| 409 | vec_st(r2, 32, result); | 408 | vec_st(r2, 32, result); |
| 410 | vec_st(r3, 48, result); | 409 | vec_st(r3, 48, result); |
| 411 | /* FIXME: there has to be a better way!!!! */ | 410 | /* FIXME: there has to be a better way!!!! */ |
| 412 | *dst_int = *src_int; | 411 | AV_WN32(dst, AV_RN32A(src_int + 0)); |
| 413 | *(dst_int+ int_dst_stride) = *(src_int + 1); | 412 | AV_WN32(dst + dst_stride, AV_RN32A(src_int + 1)); |
| 414 | *(dst_int+ 2*int_dst_stride) = *(src_int + 2); | 413 | AV_WN32(dst + 2 * dst_stride, AV_RN32A(src_int + 2)); |
| 415 | *(dst_int+ 3*int_dst_stride) = *(src_int + 3); | 414 | AV_WN32(dst + 3 * dst_stride, AV_RN32A(src_int + 3)); |
| 416 | *(dst_int+ 4*int_dst_stride) = *(src_int + 4); | 415 | AV_WN32(dst + 4 * dst_stride, AV_RN32A(src_int + 4)); |
| 417 | *(dst_int+ 5*int_dst_stride) = *(src_int + 5); | 416 | AV_WN32(dst + 5 * dst_stride, AV_RN32A(src_int + 5)); |
| 418 | *(dst_int+ 6*int_dst_stride) = *(src_int + 6); | 417 | AV_WN32(dst + 6 * dst_stride, AV_RN32A(src_int + 6)); |
| 419 | *(dst_int+ 7*int_dst_stride) = *(src_int + 7); | 418 | AV_WN32(dst + 7 * dst_stride, AV_RN32A(src_int + 7)); |
| 420 | *(dst_int+ 8*int_dst_stride) = *(src_int + 8); | 419 | AV_WN32(dst + 8 * dst_stride, AV_RN32A(src_int + 8)); |
| 421 | *(dst_int+ 9*int_dst_stride) = *(src_int + 9); | 420 | AV_WN32(dst + 9 * dst_stride, AV_RN32A(src_int + 9)); |
| 422 | *(dst_int+10*int_dst_stride) = *(src_int + 10); | 421 | AV_WN32(dst + 10 * dst_stride, AV_RN32A(src_int + 10)); |
| 423 | *(dst_int+11*int_dst_stride) = *(src_int + 11); | 422 | AV_WN32(dst + 11 * dst_stride, AV_RN32A(src_int + 11)); |
| 424 | *(dst_int+12*int_dst_stride) = *(src_int + 12); | 423 | AV_WN32(dst + 12 * dst_stride, AV_RN32A(src_int + 12)); |
| 425 | *(dst_int+13*int_dst_stride) = *(src_int + 13); | 424 | AV_WN32(dst + 13 * dst_stride, AV_RN32A(src_int + 13)); |
| 426 | *(dst_int+14*int_dst_stride) = *(src_int + 14); | 425 | AV_WN32(dst + 14 * dst_stride, AV_RN32A(src_int + 14)); |
| 427 | *(dst_int+15*int_dst_stride) = *(src_int + 15); | 426 | AV_WN32(dst + 15 * dst_stride, AV_RN32A(src_int + 15)); |
| 428 | } | 427 | } |
| 429 | 428 | ||
| 430 | /** @brief performs a 6x16 transpose of data in src, and stores it to dst | 429 | /** @brief performs a 6x16 transpose of data in src, and stores it to dst |