similar to: Int overflow fix not present in tree.

Felicia Lim wrote: > Hi Yves, > > At that time, the issue wasn't reproducible in opus master, but there > had been some significant changes since the last tagged release that > made it hard to pinpoint a single patch to cherry pick to Chromium, > hence this local fix. I'm not convinced that using a saturating add here is the correct thing to do, either (as opposed

Hi Yves, At that time, the issue wasn't reproducible in opus master, but there had been some significant changes since the last tagged release that made it hard to pinpoint a single patch to cherry pick to Chromium, hence this local fix. Cheers, Felicia On Tue, Sep 10, 2019 at 7:56 AM Yves Gerey <yvesg at google.com> wrote: > Hi! I'm Yves, bit fiddler, hoping to roll a fresh

Hi, I'm still Yves. This other fix <https://chromium-review.googlesource.com/c/chromium/src/+/1061499/> never made it to opus repository. It seems safe and sane enough to be applied to opus master branch. What do you think? -- Warm regards, Yves -------------- next part -------------- An HTML attachment was scrubbed... URL:

NSQ_LPC_BUF_LENGTH is independent of DECISION_DELAY. --- silk/define.h | 4 ---- 1 file changed, 4 deletions(-) diff --git a/silk/define.h b/silk/define.h index 781cfdc..1286048 100644 --- a/silk/define.h +++ b/silk/define.h @@ -173,11 +173,7 @@ extern "C" #define MAX_MATRIX_SIZE MAX_LPC_ORDER /* Max of LPC Order and LTP order */ -#if( MAX_LPC_ORDER >

On 06/01/2016 02:06 PM, Felicia Lim wrote: > That was my intention with refactoring out the subframe encoding and > repacketizing bit. Or do you mean I should merge the explicit check for > 120 ms frame and the existing checks for 40/60 ms wideband? What I mean is that this line in opus_encoder.c: if (frame_size > st->Fs/50 && (st->mode == MODE_CELT_ONLY ||

Thanks for your comments, Ralph and Jean-Marc. Please find attached the amended patch: - decodes a sequence of input packets rather than just one (I'm planning on using the Opus test vectors as the seed corpus) - decides on decoder setup and FEC independently of the packet data - uses Opus functions to parse ToC Cheers, Felicia On Sun, Jan 29, 2017 at 9:48 PM Jean-Marc Valin <jmvalin at

Hi, I wondered if are there any further thoughts on these patches? Thanks, Felicia On Thu, Jun 2, 2016 at 2:13 PM Felicia Lim <flim at google.com> wrote: > OK, I've amended the second patch and also added 80 and 100 ms. > > Thanks, > Felicia > > > On Thu, Jun 2, 2016 at 7:20 AM Jean-Marc Valin <jmvalin at jmvalin.ca> wrote: > >> On 06/01/2016 02:06

Hi, Attached is a patch with arm neon optimizations for silk_warped_autocorrelation_FIX(). Please review. Thanks, Felicia -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.xiph.org/pipermail/opus/attachments/20170131/9a912bb4/attachment-0001.html> -------------- next part -------------- A non-text attachment was scrubbed... Name:

Attached is the amended second patch. It now extends the multistream API as well to 80/100/120 ms and incorporates changes based on Mark's comments. Thanks, Felicia On Mon, Jun 13, 2016 at 4:21 PM Felicia Lim <flim at google.com> wrote: > Hi Mark, Jean-Marc, > > Thanks for your comments. > > On Sun, Jun 12, 2016 at 6:34 AM Mark Harris <mark.hsj at gmail.com>

Hi all, We (WebRTC/Google) would like to extend Opus to natively support 120 ms encoding instead of relying on repacketization as a post processing step. This is to ensure that a valid 120 ms packet is always available. I've attached a couple of patches to add this to opus_encoder(), based on the internal repacketization process carried out by 60 ms CELT. We intend to extend this later for

Hi Felicia, A few comments: > - /* CELT can only support up to 20 ms */ > subframe_size = st->Fs/50; > - nb_subframes = frame_size > st->Fs/25 ? 3 : 2; > + nb_subframes = frame_size/subframe_size; This will use six 20ms frames to make a 120ms packet, even for SILK-only mode where frames can be up to 60ms. For SILK, two 60ms frames would be a more

Hi all, I'm working on fuzzing Opus with OSS-Fuzz and have started with the decoder. Attached is a patch to add the corresponding fuzz target. Please let me know if there are any concerns? Thanks, Felicia -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.xiph.org/pipermail/opus/attachments/20170127/f69951aa/attachment-0001.html>

Hi Ulrich, thanks for the suggestion. My concern is that one of the valid inputs is "2.5", which would require conversion to an int, e.g. x10, but doing something like this would start to affect the code readability. On Mon, Jun 27, 2016 at 3:02 PM Ulrich Windl < Ulrich.Windl at rz.uni-regensburg.de> wrote: > Hi! > > A note on style: Looking at this chunk of the patch

Hi, Please find attached a patch to fix alignment exceptions. Without this change, we were seeing occasional alignment faults when using this with clang. Thanks, Felicia -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.xiph.org/pipermail/opus/attachments/20170818/e0d6bb06/attachment.html> -------------- next part -------------- A non-text

We see the MOVQ instruction but this patch deliberately uses it rather than MOVQDA (load 128-bits aligned). We were seeing that with the trace below, the final invocation is not 128-bit aligned but MOVQDA insists on it (the calling function was pitch_sse4_1.c:90, in the 4-way N - i >= 4 loop). 07-31 11:00:13.469 210 2540 <(469)%20210-2540> D opus_sse1: RBE celt_inner_prod_sse4_1: x

Hi all, The attached patch adds/fixes a few null argument checks in the multistream API. Do these changes make sense? Thanks, Felicia -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.xiph.org/pipermail/opus/attachments/20170324/1afee1e1/attachment.html> -------------- next part -------------- A non-text attachment was scrubbed... Name:

|frame_size_enum| in tests/test_opus_encode.cl:117 is flagged as potentially uninitialized but get_frame_size_enum() will fail anyway if a valid value is not found. --- tests/test_opus_common.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tests/test_opus_common.h b/tests/test_opus_common.h index ff7f0142..8b878607 100644 --- a/tests/test_opus_common.h +++ b/tests/test_opus_common.h @@

Hi Felicia, I still don't quite understand why you need to make 120 ms a special case, rather than extend the code that already handles 40 ms and 60 ms. Cheers, Jean-Marc On 06/01/2016 12:58 PM, Felicia Lim wrote: > Hi all, > > I've just realized that there's a better and simpler way of doing this > which ensures that analysis and selection of the mode/bandwidth etc

Hi Jean-Marc, Thanks a lot for reviewing this huge assembly function! silk_warped_autocorrelation_FIX_c()'s kernel part is for( n = 0; n < length; n++ ) { tmp1_QS = silk_LSHIFT32( (opus_int32)input[ n ], QS ); /* Loop over allpass sections */ for( i = 0; i < order; i++ ) { /* Output of allpass section */ tmp2_QS = silk_SMLAWB(

Jonathan, Here's the code difference we see with the recent change -- what amounts to reverting your change from a couple years back. It doesn't look like we're getting superfluous instructions from clang now. the bad behavior for us was the alignment exception on the movdqa instructions when the input data wasn't 128-bit aligned. We had to change something because the code as-is