Fix segfault in ExportWaveAsCode (#3769)
`char *txtData = (char *)RL_CALLOC(waveDataSize * 6 + 2000, sizeof(char));` assumes every chunk being added to txtData is 6 bytes. This is not always true, sometimes a newline is involved and the data becomes 12 bytes instead, and this can cause a random segfault. This commit changes `6` to `12`, and explains why in the comment.
This commit is contained in:
parent
c0b081fac0
commit
7ec43022c1
@ -1081,8 +1081,9 @@ bool ExportWaveAsCode(Wave wave, const char *fileName)
|
||||
int waveDataSize = wave.frameCount*wave.channels*wave.sampleSize/8;
|
||||
|
||||
// NOTE: Text data buffer size is estimated considering wave data size in bytes
|
||||
// and requiring 6 char bytes for every byte: "0x00, "
|
||||
char *txtData = (char *)RL_CALLOC(waveDataSize*6 + 2000, sizeof(char));
|
||||
// and requiring 12 char bytes for every byte; the actual size varies, but
|
||||
// the longest possible char being appended is "%.4ff,\n ", which is 12 bytes.
|
||||
char *txtData = (char *)RL_CALLOC(waveDataSize*12 + 2000, sizeof(char));
|
||||
|
||||
int byteCount = 0;
|
||||
byteCount += sprintf(txtData + byteCount, "\n//////////////////////////////////////////////////////////////////////////////////\n");
|
||||
|
Loading…
Reference in New Issue
Block a user