From Fedora Project Wiki
< Architectures | AArch64
Yselkowitz (talk | contribs) (posted msp430-gcc patch) |
Yselkowitz (talk | contribs) (posted webkitgtk3 patch to fix seed) |
||
Line 60: | Line 60: | ||
** backport of aarch64 '''host''' configuration | ** backport of aarch64 '''host''' configuration | ||
** https://bugzilla.redhat.com/show_bug.cgi?id=926177 | ** https://bugzilla.redhat.com/show_bug.cgi?id=926177 | ||
* seed | |||
** JavaScriptCore (in webkitgtk3 at least, probably affects other copies) builds but does not run properly due to an improper assumption with 64k pages (attempts to mprotect 16k not 64k-aligned) | |||
** https://bugzilla.redhat.com/show_bug.cgi?id=1074093 | |||
** https://bugzilla.redhat.com/show_bug.cgi?id=1113345 | |||
== Affecting other architectures == | == Affecting other architectures == | ||
Line 104: | Line 109: | ||
* community-mysql | * community-mysql | ||
** ICE: http://arm.koji.fedoraproject.org/koji/taskinfo?taskID=2450897 | ** ICE: http://arm.koji.fedoraproject.org/koji/taskinfo?taskID=2450897 | ||
= Useful hints = | = Useful hints = |
Revision as of 02:45, 27 June 2014
This page lists packages which failed to build for AArch64 (so called FTBFS which means 'Failed To Build From Source').
Failures
High priority
- hdf:
- python3-cairo
- issue is rather in cairo itself as it fails in testsuite
- openbabel
- 3 tests fail
- same on ppc64: https://bugzilla.redhat.com/show_bug.cgi?id=1094513
- reported upstream: https://sourceforge.net/p/openbabel/bugs/927/
- 3 tests fail
Have patches
- ice
- lack of aarch64 support fixed with nice patch from Ubuntu
- still fails on all rawhide architectures due to openjdk 1.8
- trafficserver
- xdaliclock
- old friend: config.{guess,sub}
- https://bugzilla.redhat.com/show_bug.cgi?id=926760
- sockperf
- synthv1
- configure thinks /usr/lib64 is only for x86_64, also affects ppc64
- patch: https://bugzilla.redhat.com/show_bug.cgi?id=1110892
- magic
- 64-bit patch also needed on aarch64
- patch: https://bugzilla.redhat.com/show_bug.cgi?id=1037190#c2
- pl
- unable to find libjvm
- https://bugzilla.redhat.com/show_bug.cgi?id=1112012
- fixed in rawhide
- msp430-binutils
- config.{guess,sub}, also affects ppc64le
- https://bugzilla.redhat.com/show_bug.cgi?id=926176
- msp430-gcc
- backport of aarch64 host configuration
- https://bugzilla.redhat.com/show_bug.cgi?id=926177
- seed
- JavaScriptCore (in webkitgtk3 at least, probably affects other copies) builds but does not run properly due to an improper assumption with 64k pages (attempts to mprotect 16k not 64k-aligned)
- https://bugzilla.redhat.com/show_bug.cgi?id=1074093
- https://bugzilla.redhat.com/show_bug.cgi?id=1113345
Affecting other architectures
- ice
- lack of aarch64 support fixed with nice patch from Ubuntu
- still fails anyway but on other thing
- package fails on all rawhide architectures due to openjdk 1.8
Need porting
- gccxml
- fails to build for aarch64
- code is based on gcc 4.2.1 ;(
- reported upstream: http://www.gccxml.org/Bug/view.php?id=14919
- no need to worry about it for now
- blocks: CableSwig, InsightToolkit
- gcc aarch64 config does not cleanly backport to 4.2; upstream will likely need to forward-port to 4.8+ (yselkowitz has a diff of the gccxml-specific changes on top of 4.2.1)
- luajit
- port needed
Unsorted
- eclipse
- coq
- documentation generation fails
DEBUG: sh: line 1: 6265 Segmentation fault (core dumped) '/builddir/build/BUILD/coq-8.4pl4/bin/coqtop' -boot < /tmp/coq_texbce448.v > /tmp/coq_tex297040.coq_output 2>&1
- xosview
- DONE
- need to upstream
- lttng-ust
- koji-shadow weirdness
- flint
- koji-shadow weirdness, but did expose that perl-Tk was quietly disabling Xft support due to freetype-2.5 header rearrangement (filed patch in bug 1110872)
- community-mysql
Useful hints
Useful GIT alias (paste into ~/.gitconfig):
[alias] fpaste = !sh -c 'git format-patch -1 -C -M1 --patience --stdout $1 | fpaste -l diff' -