similar to: [patch] some buildsystem fixes for crosscompiling

http://bugzilla.mindrot.org/show_bug.cgi?id=321 Summary: configure does not work when cross compiling Product: Portable OpenSSH Version: -current Platform: MIPS OS/Version: Linux Status: NEW Severity: normal Priority: P2 Component: Build system AssignedTo: openssh-unix-dev at mindrot.org

../openssh-3.1p1/configure --host=mips-linux --build=i686-linux --with-pam does not work. It selects the correct toolchain prefix, but the configure script bails on cross-compilation. Attached is a patch that *might* make the right paranoid assumptions, but I am not positive. -- bryan --- configure.ac.orig Tue Feb 26 22:12:35 2002 +++ configure.ac Wed Mar 27 14:28:02 2002 @@ -437,20 +437,6

http://bugzilla.mindrot.org/show_bug.cgi?id=336 ------- Additional Comments From gaston at ips.edu.ar 2002-09-03 12:27 ------- configure:13893: checking for msg_accrights field in struct msghdr configure:13922: gcc -o conftest -g -O2 -Wall -Wpointer-arith -Wno- uninitialized -I/usr/local/ssl /include -L/usr/local/ssl/lib conftest.c -lbsd -lz -lcrypto >&5 configure: In function

Hi, below is a patch which simply removes a Cygwin-specific piece of code from OpenSSH. Input from a system admin educated me how useful this option could be on Cygwin as well. However, way back when the call to read_etc_default_login got disabled on Cygwin hardcoded in session.c, rather than just setting --disable-etc-default-login when building the Cygwin version of OpenSSH. So, here's

Hi, I was trying to crosscompile dovecot, but I got some error: --with-notify=inotify don't work, as it try to run a test program to verify if inotify is supported in the building platform. Solved with --with-notify=dnotify. Ihmo, is more a hack than a solution. | checking whether posix_fallocate() works... configure: error: cannot run test program while cross compiling I can't

Since I use the Dante SOCKS library (instead of the NEC libraries), I decided to hack support for them into OpenSSH. Here is the results. Thanks, David $NetBSD$ --- configure.in.orig Wed Dec 29 08:37:01 1999 +++ configure.in Wed Dec 29 08:37:25 1999 @@ -334,6 +341,20 @@ AC_MSG_WARN([*** Disabling lastlog support *** ]) AC_DEFINE(DISABLE_LASTLOG) fi + +dnl Compile with dante SOCKS library

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Anton Korobeynikov schreef: > Hello, Koen. > >> So my question is: How should I build llvm and llvm-gcc4 to have it cross-compile from x86 >> to ARM/EABI? > Well, just supply appropriate --host/--target options to configure. To llvm or llvm-gcc4? > I haven't tried ARM, but this definitely works for crosscompiling from

http://bugzilla.mindrot.org/show_bug.cgi?id=184 Summary: 3.1p1 openssh fails to build a working sshd on Trusted HP-UX 10.26 Product: Portable OpenSSH Version: -current Platform: HPPA OS/Version: HP-UX Status: NEW Severity: major Priority: P2 Component: sshd AssignedTo:

I am trying to build openssh-4.3p2 portable source under Cygwin running under Windows XP SP2. I have installed Cygwin 1.5.19-4 along with these significant packages (installed using the cygwin setup program): gcc 3.4.4-1 minires-devel 1.00-1 openssl 0.9.8a-1 openssl-devel 0.9.8a-1 openssl097 0.9.7i-1 tcp_wrappers 7.6-1 zlib 1.2.3-1 I used the instructions in openssh-4.3p2/contrib./cygwin/README

Hello, Koen. > So my question is: How should I build llvm and llvm-gcc4 to have it cross-compile from x86 > to ARM/EABI? Well, just supply appropriate --host/--target options to configure. I haven't tried ARM, but this definitely works for crosscompiling from linux to mingw32. Please also note, that you'll need native (=arm) tools (binutils, libcheaders, etc) during gcc build

Hi, I am trying to use LLVM on x86 with some code intended for another architecture (IBM Cell). The plan is: 1. foreach $c_files llvm-gcc -emit-llvm -c -o $bc_file $c_file 2. link $bc_files 3. run my passes with 'opt' 4. llc -march=c 5. cell_cc The original code is for the Cell, and uses altivec intrinsics. However, llvm-gcc complains about not recognizing them. I read that

On Apr 5, 2008, at 3:31 PM, Sunjae Park wrote: > Hi, I am trying to use LLVM on x86 with some code intended for another > architecture (IBM Cell). The plan is: > > 1. foreach $c_files > llvm-gcc -emit-llvm -c -o $bc_file $c_file > 2. link $bc_files > 3. run my passes with 'opt' > 4. llc -march=c > 5. cell_cc > > The original code is for the Cell, and

On Apr 6, 2008, at 11:24 AM, Sunjae Park wrote: > On Sunday 06 April 2008 06:53:06 am Bill Wendling wrote: >> (snip) >> Hi Sunjae, >> >> You'll need to compile llvm-gcc as a cross-compiler to get it to >> accept altivec intrinsics and PPC-specific flags. >> >> -bw > > Thanks, Bill. I was afraid of that. > > Any pointers on how to do

> If you want to use llvm-gcc as a cross compiler, you must configure it as > a cross compiler, and compile the output with the appropriate cross > toolchain. LLVM does not implicitly make your code portable, see: > http://llvm.org/docs/tutorial/LangImpl8.html#targetindep > > for example. > > -Chris Hm.. so the C backend works differently if I configure LLVM to be a cross

Hi, I'm still working on this problem by first trying code without the Altivec intrinsics. I'm trying to use the cross compiler on the c code outputted by "llc -march c" and have run into this linker problem (spewed by the cross compiler): Command line : error: L0039: reference to undefined symbol .fopen64 in file /tmp/cc4tgLu5.o Command line : error: L0039: reference to

Hi Koen, The process to build llvm-gcc as a crosscompiler is basically the same as build a normal gcc as crosscompiler. The differences: you need a LLVM installed (compiled for the host). You must configure llvm-gcc with --enable-llvm. Problems: - Today LLVM supports ARM/EABI, but llvm-gcc don't. You must apply the patch:

On Sunday 06 April 2008 06:53:06 am Bill Wendling wrote: > (snip) > Hi Sunjae, > > You'll need to compile llvm-gcc as a cross-compiler to get it to > accept altivec intrinsics and PPC-specific flags. > > -bw Thanks, Bill. I was afraid of that. Any pointers on how to do this? I'm able to build llvm-gcc-4.2 x86 natively, but not the powerpc version. Using

I imagine this "where are your header files" problem comes up in other packages, so I'm asking this as a general R question. How should configure scripts be re-written so they look in more places? Briefly, the problem is that Fedora-Extras installs the header files in a subdirectory /usr/include/netcdf-3 rather than /usr/include: # rpm -ql netcdf-devel /usr/include/netcdf-3

Greetings all, I recently tried to install RNetCDF from within R (install.packages) on Fedora Core 8 (with netcdf 3.6.2 and netcdf-devel 3.6.2 already installed). This resulted in an error because the netcdf header files are installed in /usr/include/netcdf-3 rather than /usr/include which is where RNetCDF looks for them. This problem was described in a 2006 post (see below), but no

This might be a croosmail, but necessary. I did us openssl-SNAP-20160104 minorss erros and installed. I tried openssh-SNAP-20160105 and the OpenSSL libraries cannot be found In the openssh configuration file I see # Determine OpenSSL library version { $as_echo "$as_me:${as_lineno-$LINENO}: checking OpenSSL library versi on" >&5 $as_echo_n "checking