Github messages for voidlinux
 help / color / mirror / Atom feed
From: ericonr <ericonr@users.noreply.github.com>
To: ml@inbox.vuxu.org
Subject: Re: [WIP] Qt pyside : update to 5.15.2
Date: Mon, 18 Jan 2021 01:27:54 +0100	[thread overview]
Message-ID: <20210118002754.pBSYD4ZmzOrI7uLCg-59mRjU28d5jvPAHVFf5EBR22Y@z> (raw)
In-Reply-To: <gh-mailinglist-notifications-41a7ca26-5023-4802-975b-f1789d68868e-void-packages-26962@inbox.vuxu.org>

[-- Attachment #1: Type: text/plain, Size: 2802 bytes --]

New comment by ericonr on void-packages repository

https://github.com/void-linux/void-packages/pull/26962#issuecomment-761908822

Comment:
I've tried working around this by adding this patch to shiboken:

```
diff --git ApiExtractor/clangparser/compilersupport.cpp ApiExtractor/clangparser/compilersupport.cpp
index dac5110..72d6ce2 100644
--- sources/shiboken2/ApiExtractor/clangparser/compilersupport.cpp
+++ sources/shiboken2/ApiExtractor/clangparser/compilersupport.cpp
@@ -305,6 +305,12 @@ static QString compilerFromCMake(const QString &defaultCompiler)
 QByteArrayList emulatedCompilerOptions()
 {
     QByteArrayList result;
+
+    QByteArray clang_target_triplet = qgetenv("CLANG_TARGET_TRIPLET");
+    if (!clang_target_triplet.isEmpty()) result.append(clang_target_triplet);
+    QByteArray clang_target_sysroot = qgetenv("CLANG_TARGET_SYSROOT");
+    if (!clang_target_sysroot.isEmpty()) result.append(clang_target_sysroot);
+
 #if defined(Q_CC_MSVC)
     HeaderPaths headerPaths;
     result.append(QByteArrayLiteral("-fms-compatibility-version=19"));
```

and this to the pyside template:

```
diff --git a/srcpkgs/python3-pyside2/template b/srcpkgs/python3-pyside2/template
index c7ea90d0a7..d6de74c181 100644
--- a/srcpkgs/python3-pyside2/template
+++ b/srcpkgs/python3-pyside2/template
@@ -34,21 +34,10 @@ subpackages="libpyside2-python3-devel libpyside2-python3"
 # needed by shiboken2 runtime
 export CLANG_INSTALL_DIR=${XBPS_CROSS_BASE}/usr
 
-if [ "$XBPS_TARGET_NO_ATOMIC8" ]; then
-	# see https://github.com/void-linux/void-packages/pull/26962
-	broken="Requires 8-byte atomics"
-fi
-
 if [ ${CROSS_BUILD} ]; then
 	hostmakedepends+=" qt5-qmake qt5-host-tools qt5-tools qt5-remoteobjects qt5-scxml"
-
-	# shiboken2 calls clang without defining target's platform, so it fails for some platform.
-	# related upsteam's bug https://bugreports.qt.io/browse/PYSIDE-1477
-	if [ "$XBPS_TARGET_LIBC" = "glibc" ]; then
-		case "$XBPS_TARGET_MACHINE" in
-			armv7*) broken="see https://bugreports.qt.io/browse/PYSIDE-1477" ;;
-		esac
-	fi
+	export CLANG_TARGET_TRIPLET="--target=${XBPS_CROSS_TRIPLET}"
+	export CLANG_TARGET_SYSROOT="--sysroot=/usr/${XBPS_CROSS_TRIPLET}"
 fi
 
 if [ "$XBPS_TARGET_ENDIAN" = "le" ]; then
```

But now I get a different error, and it still seems to be using the host headers:

```
/../lib/gcc/arm-linux-musleabihf/9.3.0/../../../../include/c++/9.3.0/type_traits:38:10: fatal: 'bits/c++config.h' file not found
```

@q66 suggested making it `nocross` in general, and in the current circumstance I'd agree, since the aarch64 build probably just works out due to luck, not really because it's doing the right thing. This is quite unfortunate, of course, since there could be FreeCAD users on aarch64.

  parent reply	other threads:[~2021-01-18  0:27 UTC|newest]

Thread overview: 48+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-12-05 13:27 [PR PATCH] [WIP] Qt pyside 5.12.2 yopito
2020-12-05 22:49 ` ericonr
2020-12-05 23:18 ` ericonr
2020-12-05 23:19 ` ericonr
2020-12-06  9:06 ` yopito
2020-12-19 15:48 ` [PR PATCH] [Updated] [WIP] Qt pyside 5.15.2 yopito
2020-12-19 16:18 ` yopito
2020-12-19 17:53 ` yopito
2020-12-19 17:55 ` yopito
2020-12-20 10:43 ` [PR PATCH] [Updated] " yopito
2020-12-20 12:33 ` yopito
2020-12-20 12:35 ` yopito
2020-12-26  4:58 ` sgn
2020-12-26  5:10 ` ericonr
2020-12-26 11:13 ` yopito
2020-12-26 19:57 ` ericonr
2021-01-16 20:06 ` [PR PATCH] [Updated] " yopito
2021-01-16 20:14 ` yopito
2021-01-17  0:33 ` Qt pyside : update to 5.15.2 yopito
2021-01-17  1:27 ` ericonr
2021-01-17  3:11 ` q66
2021-01-17 10:27 ` yopito
2021-01-17 19:06 ` [WIP] " ericonr
2021-01-17 19:25 ` ericonr
2021-01-17 19:26 ` ericonr
2021-01-17 19:50 ` yopito
2021-01-17 20:01 ` ericonr
2021-01-17 20:04 ` ericonr
2021-01-18  0:27 ` ericonr [this message]
2021-01-18 19:29 ` yopito
2021-01-18 19:33 ` ericonr
2021-01-22 18:40 ` yopito
2021-01-22 18:51 ` yopito
2021-01-23 18:59 ` [PR PATCH] [Updated] " yopito
2021-01-23 22:35 ` yopito
2021-01-24  8:19 ` yopito
2021-01-26 21:22 ` [PR REVIEW] " ericonr
2021-01-26 21:38 ` q66
2021-01-26 22:46 ` yopito
2021-01-26 23:16 ` ericonr
2021-01-26 23:43 ` sgn
2021-01-27 19:10 ` [PR PATCH] [Updated] " yopito
2021-01-27 19:14 ` [PR REVIEW] " yopito
2021-01-27 22:15 ` ericonr
2021-01-27 22:55 ` [PR PATCH] [Updated] " yopito
2021-01-27 22:55 ` [PR REVIEW] " yopito
2021-01-28  0:15 ` q66
2021-01-28  3:33 ` [PR PATCH] [Closed]: " ericonr

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=20210118002754.pBSYD4ZmzOrI7uLCg-59mRjU28d5jvPAHVFf5EBR22Y@z \
    --to=ericonr@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).