Introduced in c152d602.

Thanks to Pedro Becerra for the report.

Introduced in c152d602.

Thanks to Pedro Becerra for the report.

http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0068b/CIHBJEHG.html
 says that "Rd cannot be the same as Rm."
http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0068b/CIHBJEHG.html
 says that "RdLo, RdHi, and Rm must all be different registers."
This means that some of the early clobbers I removed really should
 have been there (to prevent aliasing Rd, RdLo, or RdHi with Rm).
It also means that we should reverse some of the operands in the
 FFT's complex multiplies.
This should only affect the ARMv4 optimizations.

Thanks to Nils Wallménius for the report.

While we're here, audit the commutative pair flags again, since I
 screwed up at least one of them, and eliminate some dead code.

In most cases these will use __builtin_clz().
In a follow-up, we should audit usage of silk_CLZ32() and convert
 the places where its argument must be non-zero to use EC_ILOG()
 directly to avoid the test for zero (which is necessary on x86).

Since the last patch originally had them mangled (presumably by
 mailer, http server, or something else), let's just get rid of
 them.

Original patch by Aurélien Zanelli <aurelien.zanelli@parrot.com>:
 http://lists.xiph.org/pipermail/opus/2013-May/002078.html

Revised version:
- Add autconf detection (ported from libtheora).
- Rename ARM5E to ARMv5E (an ARM5 is not the same thing as ARMv5!).
- Use actual macros so they can still be selectively overridden.
- Split out ARMv4 parts and add a few more ARMv4 macros.
- Label blocks to make them easy to find in generated assembly.
- Fix MULT16_32_Q15() so we can pass make check.
  The MDCT test passes in values larger than 2**30 for b.
  The new version should be just as fast (or faster, since it's
   easier to merge the shift with following instructions), and
   there's no appreciable impact on accuracy (FFT/MDCT SNR actually
   goes up in most cases).
- Fix register constraints.
  We were using early-clobber flags in a bunch of places that
   didn't need them, and commutative-pair flags in a bunch of
   places that weren't actually commutative.
  This was Jean-Marc's fault (the original code came from Speex).
- Simplify silk_CLZ16().
- Port over iFFT C_MULC asm by Andree Buschmann
   <AndreeBuschmann@t-online.de> from Rockbox.
- Speed up the C_MULC asm by using LDRD, allowing more flexible
   addressing, re-ordering instructions to avoid some stalls,
   allowing more flexible register allocation, and getting things
   out of the inline asm block so the compiler can schedule them
   better.
- Add C_MUL and C_MUL4 asm for the FFT to the encoder based, on the
   new C_MULC.

In total, this patch gives a 22.3% speed-up on test_opus_encoder on
 a 600 MHz Cortex A8 using gcc 4.2.1,
When restricted to ARMv4 optimizations, it gives a 9.6% speed-up
 on the same processor/compiler.
On the conformance test vectors:
 Average mono quality is 97.0583 %
 Average stereo quality is 97.775 %

This is a follow-up on ac76b150. When Schur blows up, we set the last RC to
+/- 0.99 instead of eliminating it.

This replaces the OPUS_VERSION removed by ron's auto-updater.

We stop the schur recursion before any reflection coefficient
goes outside of ]-1,1[ and we force reporting a residual energy
of at least 1.
Assertion was:
Fatal (internal) error in ../silk/fixed/noise_shape_analysis_FIX.c, line 290: assertion failed: nrg >= 0
triggered by:
opus_demo voip 16000 1 12500 -bandwidth WB -complexity 10 pl04f087.stp-crash out.pcm

silk_setup_resamples() was using x_bufFIX for two purposes, and I
 only allocated enough space for one of them.
This patch also switches to slightly more descriptive variable
 names than nSamples_temp and computes the resampler input/ouput
 sizes in a way that a little more obviously doesn't have issues
 with fractional samples (and replaces a divide with a variable
 divisor by one with a constant divisor).

This makes all remaining large stack allocations use the vararray
 macros.
This continues the work of 6f2d9f50 to allow compiling with
 NONTHREADSAFE_PSEUDOSTACK to move the memory for large buffers
 off the stack for devices where it is very limited.

It also does this for some additional large buffers used by the
 PLC in the decoder.

1) The memcpy's were using sizeof(opus_int32), but the type of the
    local buffer was opus_int16.
2) Because the size was wrong, this potentially allowed the source
    and destination regions of the memcpy overlap.
   I _believe_ that nSamplesIn is at least fs_in_khZ, which is at
    least 8.
   Since RESAMPLER_ORDER_FIR_12 is only 8, I don't think that's a
    problem once you fix the type size.
3) The size of the buffer used RESAMPLER_MAX_BATCH_SIZE_IN, but the
    data stored in it was actually _twice_ the input batch size
    (nSamplesIn<<1).

Because this never blew up in testing, I suspect that in practice
 the batch sizes are reasonable enough that none of these things
 was ever a problem, but proving that seems non-obvious.

This patch just converts the whole thing to use CELT's vararrays.
This fixes the buffer size problems (since we allocate a buffer
 with the actual size we use) and gets these large buffers off the
 stack on devices using the pseudo-stack.
It also fixes the memcpy problems by changing the sizeof to
 opus_int16.
It turns out sFIR, which saved state between calls, was being used
 elsewhere as opus_int32, so this converts it to a union to make
 this sharing explicit.

You can't use /* within a comment.

1) In cases where the SILK desired bandwidth went down, then quickly up, we
count get stuck in a mode with the LP variation going the wrong way.
2) Bandwidth detection can no longer force SILK to go below wideband to
avoid switches that require redundancy.

By not reinitializing the stereo state during a switch, the old stereo
memory would create a false impulse (i.e. single sample) in the decoded audio.
This change affects the normative part of the decoder. Fortunately,
the modified decoder is still compliant with the specification because
it still easily passes the testvectors. For example, for the float decoder
at 48 kHz, the opus_compare (arbitrary) "quality score" changes from
from 99.9333% to 99.925%.

Specify the correct debug runtime libraries for the test
and demo executables to prevent mismatch warnings with
the default runtime.

Signed-off-by: Ralph Giles <giles@mozilla.com>

Koen Vos authored 12 years ago and Jean-Marc Valin committed 12 years ago

Normalizes the cost function by (x+y) instead of sqrt(x*y)

Philip Jägenstedt authored 12 years ago and Gregory Maxwell committed 12 years ago

I stumbled upon the typo in README.draft, so took the opportunity to
grep for common misspellings using List_of_common_misspellings.txt for
hunspell.

This fixes a problem where we could end up starving the redundancy
frame, especially for CBR. The solution is to make sure that some
bits are left available -- assuming we use the same rate for redundancy
as for the rest of the frame.

Diego Elio Pettenò authored 12 years ago and Jean-Marc Valin committed 12 years ago

These are only being used outside if the translation unit and only in
a few codepaths; by making them static in the header, the compiler can
optimize them quickly.

This avoids GOT-based access to the external symbol on Unix, and can
allow for constant propagation to optimize the code.

Signed-off-by: Diego Elio Pettenò <flameeyes@flameeyes.eu>

Diego Elio Pettenò authored 12 years ago and Jean-Marc Valin committed 12 years ago

Signed-off-by: Diego Elio Pettenò <flameeyes@flameeyes.eu>

Diego Elio Pettenò authored 12 years ago and Jean-Marc Valin committed 12 years ago

This is similar to the change for functions, but is only used to
perform DCE.

Signed-off-by: Diego Elio Pettenò <flameeyes@flameeyes.eu>

Diego Elio Pettenò authored 12 years ago and Jean-Marc Valin committed 12 years ago

This allows the compiler to perform more optimization on them as it
knows how the functions are being called.

Signed-off-by: Diego Elio Pettenò <flameeyes@flameeyes.eu>

Timothy B. Terriberry authored 12 years ago and Jean-Marc Valin committed 12 years ago

This allows the decoder to be compiled with
 NONTHREADSAFE_PSEUDOSTACK to move the memory for large buffers off
 the stack for devices where it is very limited.
This patch only attempts to do this for the decoder.
The encoder still requires more than 10 kB of stack.

Cyril Lashkevich authored 12 years ago and Gregory Maxwell committed 12 years ago

Signed-off-by: Gregory Maxwell <greg@xiph.org>
silk_get_TOC is unused, unreachable, and not useful without some
packet pre-processing, but until we remove it it should be correct.
This was also reported by Mozilla.

gcp authored 12 years ago and Gregory Maxwell committed 12 years ago

Add more MSVC files to .gitignore.
Make all configurations use default floating point model.

gcp authored 12 years ago and Gregory Maxwell committed 12 years ago

Also make the 64-bit builds use the default output directories.
Fix the type of the default Opus output to be a library.

gcp authored 12 years ago and Gregory Maxwell committed 12 years ago

The (non-default) temporary & output dirs configured in the projects could
cause issues when doing compilation on a multiprocessor machine. Cleaned
this up a little to use the defaults, which makes that work again and
doesn't make the output structure any worse. Also removed debug output for
release binaries & a tiny whitespace fix.