↑ ↓ ← → "Carlos Santander B." <csantander619 gmail.com>
writes:
Since I don't know what gdc version is the most recent on cygwin, I
decided to build it.
I downloaded the gcc sources (gcc, gcc-core, gcc-g++, gcc-java) and then
gdc 0.9. Then (from /usr/src/gcc-3.4.1), this happens:
$ patch -p1 < d/patch-gcc-3.4.x
can't find file to patch at input line 4
Perhaps you used the wrong -p or --strip option?
The text leading up to this was:
--------------------------
|diff -c gcc-gcc-3.4.0-orig/real.h gcc-gcc-3.4.0/real.h
|*** gcc-gcc-3.4.0-orig/real.h Fri Oct 10 20:33:05 2003
|--- gcc-gcc-3.4.0/real.h Mon Sep 27 14:06:40 2004
--------------------------
File to patch:
There's no gcc-gcc-3.4.0 or gcc-gcc-3.4.0-orig, but real.h and all the
other files mentioned in the patch are in the gcc directory. If I change
all the references, the output doesn't change.
Notice I've never compiled gcc and I have no experience with patch or
diff or any other like that.
_______________________
Carlos Santander Bernal
↑ ↓ ← → =?ISO-8859-1?Q?Anders_F_Bj=F6rklund?= <afb algonet.se>
writes:
Carlos Santander B. wrote:
There's no gcc-gcc-3.4.0 or gcc-gcc-3.4.0-orig, but real.h and all the
other files mentioned in the patch are in the gcc directory. If I change
all the references, the output doesn't change.
Notice I've never compiled gcc and I have no experience with patch or
diff or any other like that.
The patch should be run from the gcc-3.4.?/gcc directory.
-p1 makes it ignore the directory name, i.e. "gcc-gcc"
You can use my Makefile to build all of this if you like:
http://www.algonet.se/~afb/d/Makefile
make prep
make gdc
sudo make gdc-install
make phobos
sudo make phobos-install
make clean
It hasn't been tested on Cygwin yet, but works on Darwin/Linux ?
--anders
↑ ↓ ← → "Carlos Santander B." <csantander619 gmail.com>
writes:
Anders F Björklund wrote:
Carlos Santander B. wrote:
There's no gcc-gcc-3.4.0 or gcc-gcc-3.4.0-orig, but real.h and all the
other files mentioned in the patch are in the gcc directory. If I
change all the references, the output doesn't change.
Notice I've never compiled gcc and I have no experience with patch or
diff or any other like that.
The patch should be run from the gcc-3.4.?/gcc directory.
-p1 makes it ignore the directory name, i.e. "gcc-gcc"
That seems to work. make is running right now without problems.
You can use my Makefile to build all of this if you like:
http://www.algonet.se/~afb/d/Makefile
make prep
make gdc
sudo make gdc-install
make phobos
sudo make phobos-install
make clean
It hasn't been tested on Cygwin yet, but works on Darwin/Linux ?
--anders
I'll try it if I have more problems.
Thanks!
_______________________
Carlos Santander Bernal
↑ ↓
← → Dale Dude <dale daledude.com>
writes:
Anders F Björklund wrote:
Carlos Santander B. wrote:
There's no gcc-gcc-3.4.0 or gcc-gcc-3.4.0-orig, but real.h and all the
other files mentioned in the patch are in the gcc directory. If I
change all the references, the output doesn't change.
Notice I've never compiled gcc and I have no experience with patch or
diff or any other like that.
The patch should be run from the gcc-3.4.?/gcc directory.
-p1 makes it ignore the directory name, i.e. "gcc-gcc"
You can use my Makefile to build all of this if you like:
http://www.algonet.se/~afb/d/Makefile
make prep
make gdc
sudo make gdc-install
make phobos
sudo make phobos-install
make clean
It hasn't been tested on Cygwin yet, but works on Darwin/Linux ?
--anders
I tried your Makefile and get the following error using gcc 3.4.1...
cc1plus: warning: command line option "-Wstrict-prototypes" is valid for
Ada/C/ObjC but not for C++
cc1plus: warning: command line option "-Wmissing-prototypes" is valid
for Ada/C/ObjC but not for C++
In file included from ./d/dmd/stringtable.h:19,
from ./d/dmd/mtype.h:24,
from d/d-lang.cc:24:
./d/dmd/lstring.h:23: warning: ignoring #pragma warning
In file included from d/d-gcc-includes.h:44,
from d/d-lang.cc:30:
./real.h:45: error: expected class-name before numeric constant
./real.h:45: error: expected `{' before numeric constant
./real.h:45: error: expected unqualified-id before numeric constant
In file included from d/d-gcc-includes.h:52,
from d/d-lang.cc:30:
./rtl.h:124: error: declaration of `rtx_def*rtunion_def::rtx'
./coretypes.h:41: error: changes meaning of `rtx' from `typedef struct
rtx_def*rtx'
./rtl.h:125: error: declaration of `rtvec_def*rtunion_def::rtvec'
./coretypes.h:43: error: changes meaning of `rtvec' from `typedef struct
rtvec_def*rtvec'
d/d-lang.cc: In function `bool d_init()':
d/d-lang.cc:176: warning: deprecated conversion from string constant to
`char*'
d/d-lang.cc:178: warning: deprecated conversion from string constant to
`char*'
d/d-lang.cc:181: warning: deprecated conversion from string constant to
`char*'
d/d-lang.cc:190: warning: deprecated conversion from string constant to
`char*'
d/d-lang.cc:194: warning: deprecated conversion from string constant to
`char*'
d/d-lang.cc:199: warning: deprecated conversion from string constant to
`char*'
d/d-lang.cc:204: warning: deprecated conversion from string constant to
`char*'
d/d-lang.cc:210: warning: deprecated conversion from string constant to
`char*'
d/d-lang.cc:239: warning: deprecated conversion from string constant to
`char*'
make[2]: *** [d/d-lang.glue.o] Error 1
make[2]: Leaving directory `/home/root/t/gcc-3.4.3/gcc'
make[1]: *** [all-gcc] Error 2
make[1]: Leaving directory `/home/root/t/gcc-3.4.3'
make: *** [gdc] Error 2
↑ ↓ ← → =?ISO-8859-1?Q?Anders_F_Bj=F6rklund?= <afb algonet.se>
writes:
Dale Dude wrote:
It hasn't been tested on Cygwin yet, but works on Darwin/Linux ?
I tried your Makefile and get the following error using gcc 3.4.1...
In file included from d/d-gcc-includes.h:52,
from d/d-lang.cc:30:
./rtl.h:124: error: declaration of `rtx_def*rtunion_def::rtx'
./coretypes.h:41: error: changes meaning of `rtx' from `typedef struct
rtx_def*rtx'
./rtl.h:125: error: declaration of `rtvec_def*rtunion_def::rtvec'
./coretypes.h:43: error: changes meaning of `rtvec' from `typedef struct
rtvec_def*rtvec'
Seems like the Makefile failed to apply the GCC patch,
probably because you are missing the "patch" program ?
I thought Make should catch this, but maybe I need to
add an explicit "patch || exit" to the prep script...
Anyway, you need to install GNU patch (and diff first.
http://www.gnu.org/software/diffutils/diffutils.html
--anders
↑ ↓ ← → Dale Dude <dale daledude.com>
writes:
Anders F Björklund wrote:
Dale Dude wrote:
It hasn't been tested on Cygwin yet, but works on Darwin/Linux ?
I tried your Makefile and get the following error using gcc 3.4.1...
In file included from d/d-gcc-includes.h:52,
from d/d-lang.cc:30:
./rtl.h:124: error: declaration of `rtx_def*rtunion_def::rtx'
./coretypes.h:41: error: changes meaning of `rtx' from `typedef struct
rtx_def*rtx'
./rtl.h:125: error: declaration of `rtvec_def*rtunion_def::rtvec'
./coretypes.h:43: error: changes meaning of `rtvec' from `typedef
struct rtvec_def*rtvec'
Seems like the Makefile failed to apply the GCC patch,
probably because you are missing the "patch" program ?
I thought Make should catch this, but maybe I need to
add an explicit "patch || exit" to the prep script...
Anyway, you need to install GNU patch (and diff first.
http://www.gnu.org/software/diffutils/diffutils.html
--anders
You are right. The install of patch didn't seem to take the first time.
Thanks for your useful Makefile!
Regards,
Dale
↑ ↓
← → Brian Chapman <nospam-for-brian see-post-for-address.net>
writes:
Speaking of Cygwin. I was trying to build phobos and I got this error
right away:
gen_unix.c: In function `c_signal':
gen_unix.c:226: error: `SA_SIGINFO' undeclared (first use in this function)
gen_unix.c:226: error: (Each undeclared identifier is reported only once
gen_unix.c:226: error: for each function it appears in.)
gen_unix.c:248: error: `siginfo_t' undeclared (first use in this function)
gen_unix.c:248: error: parse error before "rec"
gen_unix.c:250: error: `rec' undeclared (first use in this function)
gen_unix.c:261: error: structure has no member named `sa_sigaction'
gen_unix.c:261: error: structure has no member named `sa_sigaction'
make: *** [config/gen_unix.o] Error 1
I'm not sure why this is happening. I looked into <signal.h> and
<sys/signal.h> and it is most definitely defined. (?)