Github messages for voidlinux
 help / color / mirror / Atom feed
From: JohnGebbie <JohnGebbie@users.noreply.github.com>
To: ml@inbox.vuxu.org
Subject: Re: [PR PATCH] [Updated] vosk-api: update to 0.3.45
Date: Sun, 05 Feb 2023 13:59:31 +0100	[thread overview]
Message-ID: <20230205125931.qAeR8tPURjqUToddP3hUHm04mop1dNB2W8tyiqYgkMU@z> (raw)
In-Reply-To: <gh-mailinglist-notifications-41a7ca26-5023-4802-975b-f1789d68868e-void-packages-42091@inbox.vuxu.org>

[-- Attachment #1: Type: text/plain, Size: 1269 bytes --]

There is an updated pull request by JohnGebbie against master on the void-packages repository

https://github.com/JohnGebbie/void-packages vosk-api
https://github.com/void-linux/void-packages/pull/42091

vosk-api: update to 0.3.45
<!-- Uncomment relevant sections and delete options which are not applicable -->

#### Testing the changes
- I tested the changes in this PR: **YES**|**briefly**|**NO**

<!--
#### New package
- This new package conforms to the [package requirements](https://github.com/void-linux/void-packages/blob/master/CONTRIBUTING.md#package-requirements): **YES**|**NO**
-->

<!-- Note: If the build is likely to take more than 2 hours, please add ci skip tag as described in
https://github.com/void-linux/void-packages/blob/master/CONTRIBUTING.md#continuous-integration
and test at least one native build and, if supported, at least one cross build.
Ignore this section if this PR is not skipping CI.
-->
<!--
#### Local build testing
- I built this PR locally for my native architecture, (ARCH-LIBC)
- I built this PR locally for these architectures (if supported. mark crossbuilds):
  - aarch64-musl
  - armv7l
  - armv6l-musl
-->


A patch file from https://github.com/void-linux/void-packages/pull/42091.patch is attached

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: github-pr-vosk-api-42091.patch --]
[-- Type: text/x-diff, Size: 7230 bytes --]

From 1dfd79012b1f7b5c58ba20575948ab902d2c8f1f Mon Sep 17 00:00:00 2001
From: John Gebbie <me@johngebbie.com>
Date: Sun, 5 Feb 2023 12:10:37 +0000
Subject: [PATCH 1/2] vosk-api: update to 0.3.45.

---
 common/shlibs             | 2 +-
 srcpkgs/vosk-api/template | 8 ++++----
 2 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/common/shlibs b/common/shlibs
index 59444164928b..bc77277ce8d1 100644
--- a/common/shlibs
+++ b/common/shlibs
@@ -4213,7 +4213,7 @@ libtracecmd.so.1 libtracecmd-3.1.2_1
 libkshark.so.2 libkshark-2.1.1_1
 libcpp-hocon.so.0.3.0 cpp-hocon-0.3.0_5
 libfst.so.25 openfst-1.8.2_1
-libvosk.so vosk-api-0.3.43_1
+libvosk.so vosk-api-0.3.45_1
 libdspy-1.so.1 d-spy-1.4.0_1
 libpanel-1.so.1 libpanel-1.0.1_1
 libqrtr.so.1 qrtr-ns-1.0_1
diff --git a/srcpkgs/vosk-api/template b/srcpkgs/vosk-api/template
index b86c74452b95..41d53e654042 100644
--- a/srcpkgs/vosk-api/template
+++ b/srcpkgs/vosk-api/template
@@ -1,12 +1,12 @@
 # Template file for 'vosk-api'
 pkgname=vosk-api
-version=0.3.43
+version=0.3.45
 revision=1
 create_wrksrc=yes
 build_wrksrc="vosk-api-${version}/src"
 # Vosk is built with their own vendored version of Kaldi which has no releases
 # of its own, so the commit must be determined by the date of the vosk release.
-_kaldi_commit=ecb4b47159501ef0d67888633e65b8fc0c1f79d1
+_kaldi_commit=93ef0019b847272a239fbb485ef97f29feb1d587
 build_style=gnu-makefile
 make_use_env=yes
 make_build_args="KALDI_ROOT=../../kaldi HAVE_OPENBLAS_NO_F2C=1 HAVE_OPENBLAS_CLAPACK=0"
@@ -18,8 +18,8 @@ homepage="https://alphacephei.com/vosk"
 distfiles="https://github.com/alphacep/vosk-api/archive/v${version}/vosk-api-${version}.tar.gz
  https://github.com/alphacep/kaldi/archive/${_kaldi_commit}.tar.gz
  https://alphacephei.com/vosk/models/vosk-model-small-en-us-0.15.zip"
-checksum="50fa82bfbfe3fc5be868237c84ac180eb55da0b65b1803bf022899673afda024
- 39cadd5a351a5513e3e4e1f71ac4b4f018a592a856a55f1d32c159e22e799f64
+checksum="930fb9cfa6c1b3035d3730feee7d670fb893caa0c71bd2159ee7623102674c26
+ f47a996af546884a8e1cb961323b796c5c238c1bbeabaf657bb300b760e431b2
  30f26242c4eb449f948e42cb302dd7a686cb29a3423a8367f99ff41780942498"
 shlib_provides="libvosk.so"
 patch_args="-Np1 -d ${pkgname}-${version}"

From 3714c762f47b91a303def8e9c8608f99c935f9aa Mon Sep 17 00:00:00 2001
From: John Gebbie <me@johngebbie.com>
Date: Sun, 5 Feb 2023 12:59:28 +0000
Subject: [PATCH 2/2] openfst: unrestrict archs

---
 srcpkgs/openfst/patches/32bit.patch         | 71 +++++++++++++++++++++
 srcpkgs/openfst/patches/cross-compile.patch | 23 -------
 srcpkgs/openfst/template                    |  5 +-
 3 files changed, 72 insertions(+), 27 deletions(-)
 create mode 100644 srcpkgs/openfst/patches/32bit.patch
 delete mode 100644 srcpkgs/openfst/patches/cross-compile.patch

diff --git a/srcpkgs/openfst/patches/32bit.patch b/srcpkgs/openfst/patches/32bit.patch
new file mode 100644
index 000000000000..78953f3afb21
--- /dev/null
+++ b/srcpkgs/openfst/patches/32bit.patch
@@ -0,0 +1,71 @@
+This removes checks which stops at least i686 natively compiling:
+https://github.com/kaldi-asr/kaldi/issues/4808
+
+diff --git a/configure.ac b/configure.ac
+index a2328ff..ddacaae 100644
+--- a/configure.ac
++++ b/configure.ac
+@@ -149,39 +149,6 @@ AC_ARG_WITH([libfstdir],
+ 
+ AC_SUBST([libfstdir], $with_libfstdir)
+ 
+-# Flags may be changed after configuring, so this is checked again by
+-# weight_test.cc.  The check here is to save time in the common case,
+-# or when someone does not run `make check`.
+-AC_RUN_IFELSE([AC_LANG_PROGRAM([
+-                 #include <cstdio>
+-
+-                 template <typename T>
+-                 bool FloatEqIsReflexive(T m) {
+-                   volatile T x = 1.111;
+-                   x *= m;
+-
+-                   T y = 1.111;
+-                   y *= m;
+-
+-                   return x == y;
+-                 }
+-               ], [
+-                 volatile double test_value = 1.1;
+-                 if (!FloatEqIsReflexive(static_cast<float>(test_value))) {
+-                   printf("float FAIL\n");
+-                   return 1;
+-                 }
+-                 if (!FloatEqIsReflexive(test_value)) {
+-                   printf("double FAIL\n");
+-                   return 1;
+-                 }
+-               ])],
+-              [echo "Float equality is good"],
+-              [AC_MSG_FAILURE(m4_normalize([
+-                   Test float equality failed!
+-                   Compile with -msse -mfpmath=sse if using g++.
+-              ]))])
+-
+ AC_CHECK_LIB([dl], dlopen, [DL_LIBS=-ldl])
+ AC_SUBST([DL_LIBS])
+ 
+diff --git a/src/include/fst/float-weight.h b/src/include/fst/float-weight.h
+index 5cf4386..38b7d48 100644
+--- a/src/include/fst/float-weight.h
++++ b/src/include/fst/float-weight.h
+@@ -109,6 +109,19 @@ class FloatWeightTpl {
+ // Single-precision float weight.
+ using FloatWeight = FloatWeightTpl<float>;
+ 
++template <class T>
++constexpr bool operator==(const FloatWeightTpl<T> &w1,
++                          const FloatWeightTpl<T> &w2) {
++#if (defined(__i386__) || defined(__x86_64__)) && !defined(__SSE2_MATH__)
++// With i387 instructions, excess precision on a weight in an 80-bit
++// register may cause it to compare unequal to that same weight when
++// stored to memory. This breaks =='s reflexivity, in turn breaking
++// NaturalLess.
++#error "Please compile with -msse -mfpmath=sse, or equivalent."
++#endif
++  return w1.Value() == w2.Value();
++}
++
+ // These seemingly unnecessary overloads are actually needed to make
+ // comparisons like FloatWeightTpl<float> == float compile. If only the
+ // templated version exists, the FloatWeightTpl<float>(float) conversion
diff --git a/srcpkgs/openfst/patches/cross-compile.patch b/srcpkgs/openfst/patches/cross-compile.patch
deleted file mode 100644
index 4014f51dba38..000000000000
--- a/srcpkgs/openfst/patches/cross-compile.patch
+++ /dev/null
@@ -1,23 +0,0 @@
-From 256f83e52112a5cd37e37a34beff2c4f0eae4660 Mon Sep 17 00:00:00 2001
-From: Nickolay Shmyrev <nshmyrev@gmail.com>
-Date: Tue, 16 Jun 2020 15:22:19 +0200
-Subject: [PATCH] Don't fail on cross-compile
-
----
- configure.ac | 3 ++-
- 1 file changed, 2 insertions(+), 1 deletion(-)
-
-diff --git a/configure.ac b/configure.ac
-index d52ed02..0f53b4d 100644
---- a/configure.ac
-+++ b/configure.ac
-@@ -180,7 +180,8 @@ AC_RUN_IFELSE([AC_LANG_PROGRAM([
-               [AC_MSG_FAILURE(m4_normalize([
-                    Test float equality failed!
-                    Compile with -msse -mfpmath=sse if using g++.
--              ]))])
-+              ]))],
-+              [echo "Ignoring test for cross-compilation"])
- 
- AC_CHECK_LIB([dl], dlopen, [DL_LIBS=-ldl])
- AC_SUBST([DL_LIBS])
diff --git a/srcpkgs/openfst/template b/srcpkgs/openfst/template
index 1e62d249f83c..aa5385d2187f 100644
--- a/srcpkgs/openfst/template
+++ b/srcpkgs/openfst/template
@@ -1,10 +1,7 @@
 # Template file for 'openfst'
 pkgname=openfst
 version=1.8.2
-revision=1
-# Limiting archs until resolved why a test fails when natively compiling i686:
-# https://github.com/kaldi-asr/kaldi/issues/4808
-archs="x86_64* aarch64*"
+revision=2
 build_style=gnu-configure
 configure_args="
 	--disable-bin

  parent reply	other threads:[~2023-02-05 12:59 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-02-05 12:18 [PR PATCH] " JohnGebbie
2023-02-05 12:28 ` [PR PATCH] [Updated] " JohnGebbie
2023-02-05 12:45 ` JohnGebbie
2023-02-05 12:59 ` JohnGebbie [this message]
2023-02-05 13:08 ` JohnGebbie
2023-02-05 13:20 ` JohnGebbie
2023-02-05 13:26 ` JohnGebbie
2023-02-05 14:30 ` JohnGebbie
2023-02-05 16:59 ` JohnGebbie
2023-02-05 19:42 ` JohnGebbie
2023-02-09 14:39 ` JohnGebbie
2023-02-10 14:13 ` [PR PATCH] [Updated] " JohnGebbie
2023-02-10 14:44 ` JohnGebbie
2023-02-11 11:26 ` JohnGebbie
2023-02-11 13:51 ` JohnGebbie
2023-02-11 15:02 ` JohnGebbie
2023-02-11 16:05 ` JohnGebbie
2023-02-11 16:06 ` JohnGebbie
2023-02-14 20:29 ` [PR PATCH] [Merged]: " paper42

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20230205125931.qAeR8tPURjqUToddP3hUHm04mop1dNB2W8tyiqYgkMU@z \
    --to=johngebbie@users.noreply.github.com \
    --cc=ml@inbox.vuxu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).