View previous topic :: View next topic |
Author |
Message |
Slashhhh n00b
Joined: 07 Nov 2023 Posts: 27
|
Posted: Fri Apr 19, 2024 12:46 pm Post subject: gcc seems to crash when trying to compile qtgui on riscv |
|
|
I'm on a risc v board, and gcc seems to crash with `unrecognizable insn` when I try to compile qtgui
Build log: https://gist.github.com/XerneraC/1d5cbb3c9a444e89e4bfcac44508755f
note:
Code: |
riscv64-unknown-linux-gnu-g++ -c -mabi=lp64d -march=rv64imafdc_zicsr_zba_zbb -mcpu=sifive-u74 -mtune=sifive-7-series --param l1-cache-size=32 --param l2-cache-size=2048 -O2 -std=c++1z -fvisibility=hidden -fvisibility-inlines-hidden -fno-exceptions -Wall -Wextra -Wvla -Wdate-time -Wshift-overflow=2 -Wduplicated-cond -Wno-stringop-overflow -Wno-format-overflow -D_REENTRANT -fPIC -DQT_NO_LINKED_LIST -DQT_NO_JAVA_STYLE_ITERATORS -DQT_NO_USING_NAMESPACE -DQT_NO_FOREACH -DMD4C_USE_UTF8 -DENABLE_PIXMAN_DRAWHELPERS -DQT_NO_NARROWING_CONVERSIONS_IN_CONNECT -DQT_BUILD_GUI_LIB -DQT_BUILDING_QT -DQT_NO_CAST_TO_ASCII -DQT_ASCII_CAST_WARNINGS -DQT_MOC_COMPAT -DQT_USE_QSTRINGBUILDER -DQT_DEPRECATED_WARNINGS -DQT_DISABLE_DEPRECATED_BEFORE=0x050000 -DQT_DEPRECATED_WARNINGS_SINCE=0x060000 -DQT_NO_EXCEPTIONS -D_LARGEFILE64_SOURCE -D_LARGEFILE_SOURCE -DQT_NO_DEBUG -DQT_CORE_LIB -I/var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13/src/gui -I. -I/var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13/src/3rdparty/md4c -I/var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13/src/3rdparty/VulkanMemoryAllocator -I/var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13/include -I/var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13/include/QtGui -I../../include -I../../include/QtGui -I/var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13/include/QtGui/5.15.13 -I/var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13/include/QtGui/5.15.13/QtGui -I../../include/QtGui/5.15.13 -I../../include/QtGui/5.15.13/QtGui -I.tracegen -I/usr/include/qt5/QtCore/5.15.13 -I/usr/include/qt5/QtCore/5.15.13/QtCore -I/usr/include/qt5 -I/usr/include/qt5/QtCore -I.moc -I/usr/include/libpng16 -I/var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13/mkspecs/linux-g++ -o .obj/qshaderformat.o /var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13/src/gui/util/qshaderformat.cpp
/var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13/src/gui/util/qgridlayoutengine.cpp: In member function ‘void QGridLayoutRowData::calculateGeometries(int, int, qreal, qreal*, qreal*, qreal*, const QGridLayoutBox&, const QGridLayoutRowInfo&, bool)’:
/var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13/src/gui/util/qgridlayoutengine.cpp:447:1: error: unrecognizable insn:
447 | }
| ^
(insn 681 680 682 97 (set (reg:DI 793)
(if_then_else:DI (gt:DF (reg/v:DF 537 [ targetSize ])
(reg:DF 146 [ _16 ]))
(reg:DI 794)
(reg:DI 795))) -1
(nil))
during RTL pass: vregs
/var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13/src/gui/util/qgridlayoutengine.cpp:447:1: internal compiler error: in extract_insn, at recog.cc:2791
0x11c820d internal_error(char const*, ...)
???:0
0x2bffbb fancy_abort(char const*, int, char const*)
???:0
0x2be56d _fatal_insn(char const*, rtx_def const*, char const*, int, char const*)
???:0
0x2be58f _fatal_insn_not_found(rtx_def const*, char const*, int, char const*)
???:0
0x845e3b extract_insn(rtx_insn*)
???:0
Please submit a full bug report, with preprocessed source (by using -freport-bug).
Please include the complete backtrace with any bug report.
See <https://bugs.gentoo.org/> for instructions.
|
A bit higher up I also noticed a bunch of other things that could be errors, similar to this:
Code: |
Checking for Support for Intel Control-flow Enforcement Technology...
+ cd /var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13_build/config.tests/intelcet && /var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13_build/bin/qmake "CONFIG -= qt debug_and_release app_bundle lib_bundle" "CONFIG += shared warn_off console single_arch" QMAKE_AR=riscv64-unknown-linux-gnu-ar QMAKE_CC=riscv64-unknown-linux-gnu-gcc QMAKE_LINK_C=riscv64-unknown-linux-gnu-gcc QMAKE_LINK_C_SHLIB=riscv64-unknown-linux-gnu-gcc QMAKE_CXX=riscv64-unknown-linux-gnu-g++ QMAKE_LINK=riscv64-unknown-linux-gnu-g++ QMAKE_LINK_SHLIB=riscv64-unknown-linux-gnu-g++ QMAKE_OBJCOPY=riscv64-unknown-linux-gnu-objcopy QMAKE_RANLIB= QMAKE_STRIP=riscv64-unknown-linux-gnu-strip 'QMAKE_CFLAGS=-mabi=lp64d -march=rv64imafdc_zicsr_zba_zbb -mcpu=sifive-u74 -mtune=sifive-7-series --param l1-cache-size=32 --param l2-cache-size=2048 -O2' QMAKE_CFLAGS_RELEASE= QMAKE_CFLAGS_DEBUG= 'QMAKE_CXXFLAGS=-mabi=lp64d -march=rv64imafdc_zicsr_zba_zbb -mcpu=sifive-u74 -mtune=sifive-7-series --param l1-cache-size=32 --param l2-cache-size=2048 -O2' QMAKE_CXXFLAGS_RELEASE= QMAKE_CXXFLAGS_DEBUG= 'QMAKE_LFLAGS=-Wl,-O1 -Wl,--as-needed -Wl,--undefined-version' QMAKE_LFLAGS_RELEASE= QMAKE_LFLAGS_DEBUG= /var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13_build/config.tests/intelcet
+ cd /var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13_build/config.tests/intelcet && MAKEFLAGS= /usr/bin/gmake
> riscv64-unknown-linux-gnu-g++ -c -mabi=lp64d -march=rv64imafdc_zicsr_zba_zbb -mcpu=sifive-u74 -mtune=sifive-7-series --param l1-cache-size=32 --param l2-cache-size=2048 -O2 -w -fPIC -I. -I/var/tmp/portage/dev-qt/qtgui-5.15.13/work/qtbase-everywhere-src-5.15.13/mkspecs/linux-g++ -o main.o main.cpp
> main.cpp:7:8: error: #error Intel CET not available
> 7 | # error Intel CET not available
> | ^~~~~
> gmake: *** [Makefile:184: main.o] Error 1
test config.qtbase.tests.intelcet FAILED
|
But as the compile went on afterwards, I assume this is not an issue?
Last edited by Slashhhh on Wed Apr 24, 2024 2:20 pm; edited 1 time in total |
|
Back to top |
|
|
Hu Administrator
Joined: 06 Mar 2007 Posts: 22658
|
Posted: Fri Apr 19, 2024 2:43 pm Post subject: Re: gcc seems to crash when trying to compile qtgui on riscv |
|
|
Slashhhh wrote: | Code: | Please submit a full bug report, with preprocessed source (by using -freport-bug).
Please include the complete backtrace with any bug report.
See <https://bugs.gentoo.org/> for instructions. |
| The output states this is a reportable bug. We may be able to help you prepare a better bug report, but this will need a bug report. Can you trim down the input to produce a minimal test case for this crash? Slashhhh wrote: | A bit higher up I also noticed a bunch of other things that could be errors, similar to this: Code: | Checking for Support for Intel Control-flow Enforcement Technology...
> main.cpp:7:8: error: #error Intel CET not available
> 7 | # error Intel CET not available
> | ^~~~~
> gmake: *** [Makefile:184: main.o] Error 1
| But as the compile went on afterwards, I assume this is not an issue? | You are using make -j4, so fatal errors may not be immediately fatal. It looks wrong to me that a test for an Intel feature was even attempted on a clearly non-Intel-compatible platform. However, I am not familiar with this package, so it might be able to recover from that. |
|
Back to top |
|
|
Slashhhh n00b
Joined: 07 Nov 2023 Posts: 27
|
Posted: Wed Apr 24, 2024 2:19 pm Post subject: |
|
|
Upgraded from gcc13 to the experimental gcc14 and the issue is gone. |
|
Back to top |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|