Fix UBSan errors (left shift on negative value)

BUG=webrtc:5530

Review-Url: https://codereview.webrtc.org/1999653002
Cr-Commit-Position: refs/heads/master@{#12820}
This commit is contained in:
minyue 2016-05-20 01:23:50 -07:00 committed by Commit bot
parent 03d45b016a
commit 30629957e1

View File

@ -117,7 +117,8 @@ int32_t WebRtcAgc_CalculateGainTable(int32_t *gainTable, // Q16
// limiterLvl = targetLevelDbfs + limiterOffset/compRatio
limiterLvlX = analogTarget - limiterOffset;
limiterIdx =
2 + WebRtcSpl_DivW32W16ResW16((int32_t)limiterLvlX << 13, kLog10_2 / 2);
2 + WebRtcSpl_DivW32W16ResW16((int32_t)limiterLvlX * (1 << 13),
kLog10_2 / 2);
tmp16no1 = WebRtcSpl_DivW32W16ResW16(limiterOffset + (kCompRatio >> 1), kCompRatio);
limiterLvl = targetLevelDbfs + tmp16no1;
@ -143,7 +144,7 @@ int32_t WebRtcAgc_CalculateGainTable(int32_t *gainTable, // Q16
inLevel = WebRtcSpl_DivW32W16(tmp32, kCompRatio); // Q14
// Calculate diffGain-inLevel, to map using the genFuncTable
inLevel = ((int32_t)diffGain << 14) - inLevel; // Q14
inLevel = (int32_t)diffGain * (1 << 14) - inLevel; // Q14
// Make calculations on abs(inLevel) and compensate for the sign afterwards.
absInLevel = (uint32_t)WEBRTC_SPL_ABS_W32(inLevel); // Q14
@ -185,7 +186,7 @@ int32_t WebRtcAgc_CalculateGainTable(int32_t *gainTable, // Q16
logApprox = (tmpU32no1 - tmpU32no2) >> (8 - zerosScale); //Q14
}
}
numFIX = (maxGain * constMaxGain) << 6; // Q14
numFIX = (maxGain * constMaxGain) * (1 << 6); // Q14
numFIX -= (int32_t)logApprox * diffGain; // Q14
// Calculate ratio
@ -198,7 +199,7 @@ int32_t WebRtcAgc_CalculateGainTable(int32_t *gainTable, // Q16
{
zeros = WebRtcSpl_NormW32(den) + 8;
}
numFIX <<= zeros; // Q(14+zeros)
numFIX *= 1 << zeros; // Q(14+zeros)
// Shift den so we end up in Qy1
tmp32no1 = WEBRTC_SPL_SHIFT_W32(den, zeros - 8); // Q(zeros)
@ -213,7 +214,7 @@ int32_t WebRtcAgc_CalculateGainTable(int32_t *gainTable, // Q16
if (limiterEnable && (i < limiterIdx))
{
tmp32 = WEBRTC_SPL_MUL_16_U16(i - 1, kLog10_2); // Q14
tmp32 -= limiterLvl << 14; // Q14
tmp32 -= limiterLvl * (1 << 14); // Q14
y32 = WebRtcSpl_DivW32W16(tmp32 + 10, 20);
}
if (y32 > 39000)
@ -559,8 +560,8 @@ int32_t WebRtcAgc_ProcessDigital(DigitalAgc* stt,
// Apply gain
// handle first sub frame separately
delta = (gains[1] - gains[0]) << (4 - L2);
gain32 = gains[0] << 4;
delta = (gains[1] - gains[0]) * (1 << (4 - L2));
gain32 = gains[0] * (1 << 4);
// iterate over samples
for (n = 0; n < L; n++)
{
@ -587,8 +588,8 @@ int32_t WebRtcAgc_ProcessDigital(DigitalAgc* stt,
// iterate over subframes
for (k = 1; k < 10; k++)
{
delta = (gains[k+1] - gains[k]) << (4 - L2);
gain32 = gains[k] << 4;
delta = (gains[k+1] - gains[k]) * (1 << (4 - L2));
gain32 = gains[k] * (1 << 4);
// iterate over samples
for (n = 0; n < L; n++)
{