Luke Dashjr wrote:
> If you need a source for some reason, feel free to quote me. You won't
> find anyone else more* qualified in the Bitcoin area.
Without this hint, I would not have guessed it:
Luke Dashjr apparently is "the co-founder of Ocean mining pool" [1].
Which means that he has persona
Luke Dashjr wrote:
> Bitcoin is in fact good for the environment, despite whatever FUD you
> dig up on Wikipedia
You cannot make points in a discussion, by repeating wrong claims that
already have been shown to be misinformation. Wikipedia lists sources.
You don't.
> TEPCO is working on a circul
Luke Dashjr wrote:
> Bitcoin is in fact very _good_ for the environment.
Nope. Read
https://en.wikipedia.org/wiki/Environmental_impact_of_bitcoin
Bruno
___
cfarm-users mailing list
cfarm-users@lists.tetaneutral.net
https://lists.tetaneutral.net/listi
Hi,
Zach van Rijn wrote:
> It's been rebooted and should be available again.
Thanks!
Could you please also mount /Volumes/STORAGE again? It's the big disk
that we're recommended to use because / has only limited capacity.
Bruno
___
cfarm-users mail
Paul H. Hargrove wrote:
> It has been a long time since I've attempted to support my software on AIX.
> However, my old notes mention OBJECT_MODE and COMPILE_BITS as environment
> variables which influence the IBM toolchain.
> IIRC, their use could/should eliminate the need for `-q64` and `-X 64`
>
Hi,
> We've been poking around, getting to grips with how all this stuff
> works, doing some test builds of easy stuff on things that aren't linux,
> as "that's not Linux" is very much our focus. Tried to do a test build
> of libarchive on AIX on cfarm111, and it seemed to build fine, then we g
Jing Luo wrote:
> This would be my proposal, IF there are enough people who want such *bsd
> vm systems.
+1 from me.
It is possible to create such VMs on stock x86_64 hardware [1][2][3],
but it's definitely easier to use for everyone if it's based on aarch64
hardware and one can log in without h
Peter Gutmann wrote:
> >/opt/csw/bin
>
> I'd looked in there but couldn't find anything (make, ar, strip, etc).
Look at the PATH column in
https://gitlab.com/ghwiki/gnow-how/-/wikis/Platforms/Configuration.
> There's
> g present, but I was hoping to test with Sun tools to verify the
> ability to
Peter Gutmann wrote:
> >cfarm119 has two compilers installed:
> > - A GCC that works, in /opt/freeware/bin.
> > - An IBM derivate of clang, named ibm-clang and ibm-clang++,
> >in /opt/IBM/openxlC/17.1.1/bin. But it is broken (miscompiles various
> >things here and there).
>
> Ah, OK. Ac
Peter Gutmann wrote:
> Playing with my own build-on-everything script I've found some issues with
> several systems
Special compiler options or environment variables are needed on many systems.
See https://gitlab.com/ghwiki/gnow-how/-/wikis/Platforms/Configuration .
> cfarm119 has no dev tools th
Jonathan Wakely wrote:
> > It would not be straightforward to track all SSH access on the farm, both
> > for privacy reasons and technical reasons (the farm has very diverse
> > systems, and some people run jobs via cron).
>
> What are the privacy reasons?
>
> It's a free, public service offered
> We are pleased to announce the availability of two new machines, cfarm215
> and cfarm216. They run the latest Solaris 11.4 SRU (Support Repository
> Update). The first one is based on a recent Intel x86 CPU, while the
> second one is based on a SPARC T8-1 CPU.
Thanks a lot for these machines!
>
Gregor Riepl wrote:
> OpenShift is a full-blown container
> orchestrator with many baseline services and is not comparable to a
> local Podman runtime and a handful of containers launched directly by users.
Thanks for explaining, and for blowing away my prejudice :)
__
Gregor Riepl wrote:
> Podman[1] on all boxes where it is supported.
Use of containers generally tends to consume more memory (both RAM and disk),
compared to traditional deployment/installation methodology, right?
At least that's the impression I get from the minimum system requirements
of OpenShi
Peter Gutmann wrote:
> the 'make touch' workaround, which conveniently also gets around wrong-system-
> time problems, time-zone-set-wrong problems, ...
Your 'make touch' workaround does not work for packages that contain some
*.m4 files, an aclocal.m4 file, a configure file, and a config.h.in fil
Peter Gutmann wrote:
> If you bundle something up on a system that's (say) 12 hours ahead of the
> target system and unpack it there, many of the files will have dates up to 12
> hours in the future, messing up the build process. So time zones do matter.
If time zones do matter, either your packi
Michael Felt wrote:
> Migration complete.
Thanks!
The system clock of this machine is 42 minutes backwards:
$ LC_ALL=C TZ=UTC date ; ssh gcc119.fsffrance.org TZ=UTC date
Fri Aug 18 11:27:41 UTC 2023
Fri Aug 18 10:45:40 UTC 2023
Could this be fixed, please? Because in the current state, it means
Jaroslav Fojtik wrote:
> >In particular, on Solaris 10, you need to have /usr/ccs/bin in PATH.
> This should be a responsibility of anybody who attempts to compile my package.
Yes, modifying PATH should not be done in the configure script. It is the
responsibility of the person whole compiles the
Jaroslav Fojtik wrote:
> ***
>
> ar rcs libatoms.a struct.o dbllist.o sets.o std_str.o struct.o lists.o
> strings.o stack.o
> utf8.o mtx_impl.o
> gmake[2]: ar: No such file or directory
> gmake[2]: *** [makefi
Jacob Bachmeyer wrote:
> Speculation: (based on a bug I encountered long ago in the ext3 driver)
There are also other causes of "BUG: soft lockup - CPU#n stuck". One of them
is bad RAM.
You can diagnose bad RAM more reliably through 'memtester' [1]. That did it
in my case (on an x86_64 machine)
The gcc104 filesystem is full. Please move what you can to the 'STORAGE' disk.
$ df -m .
Filesystem 1M-blocks Used Available Capacity iused ifree %iused Mounted on
/dev/disk3s5233752 21769319 100% 1284579 204760 86%
/System/Volumes/Data
$ df -m /Volumes/STORAGE
Filesystem
Jonathan Wakely wrote:
> May be that this is a bug of autoconf.
AC_PROG_MAKE_SET, as documented [1], uses $MAKE if set or else ‘make’.
I think that's good enough, because ultimately it is the user's responsibility
to make sure that $PATH and other environment variables are set appropriately.
If
Jaroslav Fojtik wrote:
> make is not installed on system.
> There is "gmake".
Indeed:
* On gcc210, there is /usr/xpg4/bin/make = /usr/ccs/bin/make and
/opt/csw/bin/gmake.
* On gcc211, there is only /opt/csw/bin/gmake.
$ pkg search /usr/bin/make
and
$ pkg search /usr/ccs/bin/m
David Edelsohn wrote:
> Someone created a huge file in /var/tmp. I have removed it, so there now
> is space.
Thank you!
> > 2) A symbolic link for /usr/lib/libncurses.a exists, but linking with
> > this library produces a non-functional program:
> >
> > $ cat conftest.c
> > extern int tgetent (c
Kaz Kylheku wrote:
> gcc -E - -D_XOPEN_SOURCE=500
Defining _XOPEN_SOURCE — or worse, _POSIX_SOURCE — on arbitrary platforms
is a recipe for producing compilation errors.
On HP-UX/hppa, -D_XOPEN_SOURCE=500 is useful.
On HP-UX/ia64, -D_XOPEN_SOURCE=600 is useful, but only in combination with
-D_HPU
Kaz Kylheku wrote:
> Simple code like this doesn't compile:
>
> #include
>
> // ...
>
> void foo(struct in6_addr *addr)
> {
> return bar(addr->s6_addr, 16);
> }
After fixing the gcc warning "‘return’ with a value, in function returning
void", this code compiles perfectly fine, bo
On the gcc119 (AIX 7.2) machine, there are two problems:
1) The /var partition is full:
$ df -m /tmp /var/tmp
FilesystemMB blocks Free %UsedIused %Iused Mounted on
/dev/hd3 16384.00 12377.05 25% 132106 5% /tmp
/dev/hd9var 4096.00 0.00 100% 244033% /va
Hi Jeffrey,
> Would it be possible to have libtool installed, please?
You don't need libtool to be preinstalled in /usr or /usr/local, since
you can easily install it on any machine within a few seconds:
$ wget https://ftp.gnu.org/gnu/libtool/libtool-2.4.6.tar.gz
$ tar xfz libtool-2.4.6.tar.
> on these machines, terminals are messed up.
>
> TERM is set to xterm-256colors (standard for KDE).
Here's a local fix:
1) Save the attached file (taken from an Ubuntu machine).
2) $ mkdir $HOME/.terminfo
$ tic -o $HOME/.terminfo xterm-256color.ubuntu
Bruno
# Reconstructed via infocmp
Hi,
Thank you for providing these machines!
Where to find the GCC sources? I want to build this GCC as a cross-compiler.
I tried https://github.com/loongson/gcc, but
- with it, "gcc --version" reports "12.0.0 20210528 (experimental)",
not "12.0.0 20211202 (experimental)",
- my cross-compi
Jeffrey Walton wrote:
> > > A quick
> > > workaround if you're using zipped archives is to add -DD to the command
> > > line.
> >
> > What do you mean? Neither 'tar' not 'make' takes a '-DD' option.
>
> I believe that would be the -D and -DD option to unzip (1).
This workaround would
1) requir
Peter Gutmann wrote:
> A quick
> workaround if you're using zipped archives is to add -DD to the command line.
What do you mean? Neither 'tar' not 'make' takes a '-DD' option.
Bruno
___
cfarm-users mailing list
cfarm-users@lists.tetaneutral.net
https
On both machines, the clock is 8 hours behind reality. When unpacking
freshly created tarballs, this leads to warning messages such as
tar: foobar-20220107: time stamp 2022-01-08 00:36:16 is 28740.13808862 s in the
future
and subsequent build failures.
Would it be possible to correct this? I th
Hi,
/var and /var/tmp on gcc111.fsffrance.org are full. 40% of this partition
are occupied by a single file: /var/spool/mail/atkins3 .
$ ls -l /var/spool/mail/atkins3
-rw-rw1 atkins3 mail 1729518131 Apr 12 05:45
/var/spool/mail/atkins3
Can someone please shrink this huge file?
Br
Some machines of the compile farm require a non-default port [1].
But it's annoying to remember in every command that while ssh takes a
'-p' option, scp wants a '-P' option.
Here's a trick: You can omit the -p / -P options if you add this to
your $HOME/.ssh/config file:
=-=-=-=-=-=-=-=-=-=-=-=-=
Hi Kaz,
> On Mac OS, there are, by default, bizarre restrictions on IP datagram
> sizes
> in the protocol stack.
> ...
> I have a test case involving UDP that works on numerous platforms, the
> oddball being MacOS
So, the test machine was useful to you: it has told you that a test case
(and pres
Pierre Muller asked:
>Are executables randomized in memory?
Yes, they are. I have a program which prints its memory map. When I run
it twice in a row, it prints different addresses:
10272-102724000 r-x
102724000-102728000 r--
102728000-10272c000 rw-
10272c000-10273 r--
10273-10273
> 6. There is no way to do interactive debugging, AFAICS:
Post-mortem debugging, through core dumps, does not work either [1][2].
Bruno
[1]
https://stackoverflow.com/questions/65278351/no-core-dump-file-generated-after-segmentation-fault-in-macos-big-sur
[2]
https://stackoverflow.com/questions
Christian Jullien wrote:
> TinyCC is not an interpreter, it compiles in memory then executes the code
> so fast that its looks to act as an interpreter.
But the essential difference is that it does so in the same process, without
using ptrace.
> AFAIK, TinyCC has no specific debugger.
A debugger
Christian Jullien wrote:
> Again, we'll have to probably adapt the code to Silicon which no longer
> allows to make pages writable and executable. Few extra operations are
> required.
There are two approaches for doing that. One is to create a writable
mapping and an executable mapping of the same
Hi Christian,
> To port TinyCC on Silicon ...
It's good that you mention TinyCC. AFAIU, it contains a C interpreter.
Would it be possible to extend it, to contain a C interpreter and debugger?
This would make it possible to do debugging on platforms where 'ptrace'
is not available to normal users
Christian,
> To port TinyCC on Silicon we facing a major issue: libc.dylib (link to
> libSystem.dylib) no longer exists on the filesystem, it is cached and
> hidden somewhere in the system.
Yes. [1][2]
> You can still dlopen libSystem.dylib and then dlsym any symbols you like
> but you can't fop
> I second Jeff''s thanks to those who've made this system available.
+1
> .zprofile (which should be made executable)
No, ~/.zprofile does not need to be made executable.
> 5. Finally a note from my earlier efforts to port to this target:
> If using config.{sub,guess} (such as w/ autoconf) one
Jeffrey Walton wrote:
> It looks like is missing
No, it is not missing.
$ clang --version | head -n 1
OpenBSD clang version 10.0.1
$ ls -l /usr/lib/clang/10.0.1/include/immintrin.h
-r--r--r-- 1 root bin 16225 Oct 5 01:47
/usr/lib/clang/10.0.1/include/immintrin.h
$ ec
Hi,
The gcc119 machine (AIX 7.2) does not currently accept ssh logins,
although it is pingable.
The status page says "Unreach. for a day".
Does it need a reboot, maybe?
Bruno
___
cfarm-users mailing list
cfarm-users@lists.tetaneutral.net
https://list
Jeffrey,
> b2sum-src/b2sum.c:266:26: error: array type has incomplete element
> type 'struct option'
> static struct option long_options[] = {
> ^~~~
When you use 'struct option', you need to
1) use Gnulib module 'getopt-gnu' - see the documentation
Segher Boessenkool wrote:
> $ sh config.sub powerpc64-elfv2-linux
> powerpc64-elfv2-linux-gnu
>
> No, that is not the same thing. It is BE, and it uses ELFv1!
>
>
> $ :|powerpc64-elfv2-linux-gcc -dM -E -|grep BYTE_ORDER
> #define __BYTE_ORDER__ __ORDER_BIG_ENDIAN__
>
> $ :|powerpc64-elfv2-linu
Kaz Kylheku wrote:
> What I'm now wondering is whether my context saving/restoring assembly
> code for PPC has any ABI "gotcha" in it. I've only tested in on the
> above gcc110 machine.
The compilefarm covers all variations of PowerPC ABIs (except for
Mac OS X <= 10.5, which no one uses any more)
Segher Boessenkool wrote:
> > not on powerpc64-elfv2-linux!).
>
> I think you mean powerpc64le-linux?
Yes, this is the same thing. The ABI indicator for this ABI is the
preprocessor define _CALL_ELF=2; it is used on the distros named
powerpc64le-linux.
Bruno
Paulo César Pereira de Andrade wrote:
> When testing on gcc110.fsffrance.org I got confused as there were no
> function descriptors, and the float abi to call varargs functions did
> not match.
Indeed, function descriptors are a difference between the ABIs as well:
On powerpc-aix and powerpc64-
Hi Paulo,
> Just testing for hopefully soon new release of GNU Lightning, I noticed
> now when building with -m32 it appears to have a different abi that I
> could not find any description searching on google; probably I did not
> search enough..
Generally you find this info in the GCC sources,
Hi,
Has there been a recent change to the ssh configuration on the machines
gcc22, gcc23, gcc24 ?
Up until recently I could ssh into these machines. Now it asks for the
password, which indicates a problem with the ~/.ssh/ directory or the
files therein. But https://cfarm.tetaneutral.net/sshkey/li
Martin Jambor wrote:
> Unfortunately it seems that openSUSE Leap 42.1 repositories (as
> opposed to 42.2 and the current 42.3 repositories) are no longer
> online
A copy is still available at
https://ftp5.gwdg.de/pub/opensuse/discontinued/distribution/leap/42.1/repo/oss/
It contains perf-4.1.10-36
> No matter how the cake is cut, our AIX machines (and probably AIX in
> general) need to symlink the default system compiler to `cc'.
No, please do NOT put such a symlink in /usr/bin or /usr/local/bin.
The compilefarm machines are used for portability testing. They should
be as similar as the ma
Ville Voutilainen writes:
> I needed to add
> PubkeyAcceptedKeyTypes ssh-dss
> to my client-side config a long while ago.
If you are using OpenSSH >= 7.0, I'd recommend the syntax
PubkeyAcceptedKeyTypes +ssh-dss
in order to allow ssh-rsa, ssh-ed25519, etc.
See
http://unix.stackexchange.com/que
55 matches
Mail list logo
