| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| ... | |
| |
|
|
|
| |
lj_arch.h:405:2: error: #error "No support for PowerPC 64 bit mode (yet)"
#error "No support for PowerPC 64 bit mode (yet)"
|
| |
|
|
| |
status
|
| |
|
|
|
|
|
| |
Stripping manually to avoid strip non-CARCH binaries, as reported:
>>> xf86-video-rendition*: Stripping binaries
strip: Unable to recognise the format of the input file `./usr/lib/xorg/modules/v20002d.uc'
|
| |
|
|
|
|
|
| |
Simplifying build script for xf86-video-rendition, since prepare() is
not required on this package, since there is no patches, and, if it
was required, I understand that default_prepare() should do the same
job as current prepare(), thus removing prepare() section.
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
https://kripken.github.io/emscripten-site/
The Clang + LLVM backend for Emscripten
|
| |
|
|
|
|
|
| |
Nagios-plugins is not building check-game plugin anymore, thus,
abuild is failing to build this package because it tries to remove
(it is not important for Alpine) a package that is not being built
anymore.
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
strace build requires <linux/ptrace.h> and it includes <asm/ptrace.h>
that, by its turn, defines the pt_regs struct. However the same struct
is also define in <bits/user.h> from musl-dev, creating therefore a
conflict. A solution is to add the __ASSEMBLY__ guard so pt_regs struct
from <asm/ptrace.h> is not include twice, avoiding the collision in
question.
There is no other way to worki around that issue on Musl since it does
not include the linux headers and hence has to define again the pt_regs
struct in user.h and the name clashes. In glibc, for instance, user.h
includes <asm/ptrace.h> from kernel and does not redefine it.
Also Musl community has no better solution at the moment for this issue
on ppc, so basically nothing change since [1].
[1] http://www.openwall.com/lists/musl/2016/12/30/5
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
| |
During the build script, there is a command that tries to copy all
the files from inside a directory, which happens to have a sub-directory,
thus failing the build.
This patch just execute a recursive copy, copying the sub-directory, and
not failing the build.
|
| | |
|
| |
|
|
|
|
|
| |
Fail2ban added ipv6 support in version 0.10.0 and the default ban
actions (such as iptables-common.conf) use ip6tables as well as
iptables. Not having the ip6tables installed results in errors
during jail start.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
py-curl does not build on a clean environment due to lack of
libressl-dev package, that is not listed as a make-dependency.
This is the failure:
In file included from src/docstrings.c:4:0:
src/pycurl.h:170:31: fatal error: openssl/crypto.h: No such file or directory
# include <openssl/crypto.h>
^
compilation terminated.
error: command 'gcc' failed with exit status 1
>>> ERROR: py-curl: all failed
|
| |
|
|
| |
The package is outdated and unmaintained.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
Fix the path when creating cmake folder. It had a typo
in $pkgdir variable.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
| |
smokeping fails to build on a clean build machine due to the lack
of perl-dev package, which provides pod2man. Adding this package
as a make-dependency.
Adding a dependency for perl-try-tiny also, otherwise it fails to
build on a clean environment with the following error:
checking checking for perl module 'LWP'... Can't locate Try/Tiny.pm in @INC (you may need to install the Try::Tiny module)
|
| | |
|
| |
|
|
| |
update config guess before build
|
| |
|
|
|
|
| |
change pushd command, that is a bash built-in command and is not
working, for cd command. As a popd is not being used, it can be
changed
|
| |
|
|
|
| |
Currently silc-client FTBFS on ppc64le due to unrecognized platform.
Updating config.guess solves this problem and make it buildable.
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Review notes by @jirutka:
Upstream has republished the lpeg-1.0.1 tarball with two minor
modifications in file test.lua:
--- old/lpeg-1.0.1/test.lua
+++ new/lpeg-1.0.1/test.lua
@@ -1,6 +1,6 @@
#!/usr/bin/env lua
--- $Id: test.lua,v 1.111 2016/09/20 17:51:30 roberto Exp $
+-- $Id: test.lua,v 1.112 2017/01/14 18:55:22 roberto Exp $
-- require"strict" -- just to be pedantic
@@ -1132,7 +1132,7 @@
local function manyCmt (n)
return m.Cmt("a", function ()
local a = {}; for i = 1, n do a[i] = n - i end
- return true, table.unpack(a)
+ return true, unpack(a)
end)
end
@@ -1147,7 +1147,7 @@
-- bug in 1.0: problems with math-times returning too many captures
do
- local lim = 2^15 - 10
+ local lim = 2^11 - 10
local res = {m.match(manyCmt(lim), "a")}
assert(#res == lim and res[1] == lim - 1 and res[lim] == 0)
checkerr("too many", m.match, manyCmt(2^15), "a")
|
| |
|
|
| |
py() is completely wrong, but I'll fix it later.
|
| |
|
|
| |
Adjust path to a valid one when changing directory
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
is in archive
|
| |
|
|
| |
in archive
|
