mirror of https://github.com/mstorsjo/fdk-aac.git
bef1b4dd96
An overflow in the fft was observed for certain configurations. The root cause was in the scaling of the fft input data in dct_IV and dst_IV. The following patch revises the scaling in dct and removes the unused arm implementation. Bug: 112661057 Bug: 109936783 Bug: 112022208 Test: atest DecoderTestXheAac ; atest DecoderTestAacDrc Change-Id: Idc9c44a9755b8151d5c7d8107696f43ac592413b |
||
|---|---|---|
| .. | ||
| arm | ||
| mips | ||
| FDK_bitbuffer.cpp | ||
| FDK_core.cpp | ||
| FDK_crc.cpp | ||
| FDK_decorrelate.cpp | ||
| FDK_hybrid.cpp | ||
| FDK_lpc.cpp | ||
| FDK_matrixCalloc.cpp | ||
| FDK_qmf_domain.cpp | ||
| FDK_tools_rom.cpp | ||
| FDK_trigFcts.cpp | ||
| autocorr2nd.cpp | ||
| dct.cpp | ||
| fft.cpp | ||
| fft_rad2.cpp | ||
| fixpoint_math.cpp | ||
| huff_nodes.cpp | ||
| mdct.cpp | ||
| nlc_dec.cpp | ||
| qmf.cpp | ||
| scale.cpp | ||