RE: Fail to compile 5.04.3 on Solaris

OK,

I managed to get it to work it seems...

I started by upgrading to 5.04.2 as this seemed to be a good idea. The
error was still there. But restarting the make process again worked fine
when I did 'make boot' in hslibs as well.

So I got my compiler after all. But the error is a bit annoying.

Thanks for the help.

        /Josef

On Mon, 14 Apr 2003, Simon Peyton-Jones wrote:

> Strange.
>
> The only things I can think of are
>
> a) you haven't done 'make clean' first and you still have leftover .hi
> files from an earlier version of GHC. With GHC's new binary .hi-file
> format, version changes can give rise to strange behaviour.  (A very
> unsavoury fact.)
>
> b) you haven't done a 'make boot' in hslibs.
>
> Apart from that, I can't explain what you are seeing.  Try with -v to
> see what's going on?
>
> SImon
>
> | -----Original Message-----
> | From: glasgow-haskell-bugs-admin@xxxxxxxxxxx
> [mailto:glasgow-haskell-bugs-admin@xxxxxxxxxxx] On
> | Behalf Of Josef Svenningsson
> | Sent: 14 April 2003 11:16
> | To: glasgow-haskell-bugs@xxxxxxxxxxx
> |
> | Hi!
> |
> | I've tried to compile 5.04.3 on Solaris some time now but I can't make
> it
> | work. Compiling the compiler works fine. The problem is when trying to
> | compile the libraries. Compiling GHC/Handle always fails with either
> Bus
> | error or Segmentation fault. Like this:
> |
> | rm -f GHC/Handle.o; if [ ! -d GHC/Handle_split ]; then mkdir
> | GHC/Handle_split; else /usr/bin/find GHC/Handle_split -name '*.o' |
> xargs
> | rm -f __rm_food; fi;
> | ../../ghc/compiler/ghc-inplace -ldl -fglasgow-exts -cpp -Iinclude
> | -funbox-strict-fields -package-name base -O -Rghc-timing  -split-objs
> | -c GHC/Handle.hs -o GHC/Handle.o
> | Bus Error
> | <<ghc: 486170608 bytes, 1662 GCs, 5862901/9509368 avg/max bytes
> residency
> | (13 samples), 26M in use, 0.01 INIT (0.02 elapsed), 11.98 MUT (30.24
> | elapsed), 8.40 GC (8.76 elapsed) :ghc>>
> |
> | The funny thing is though that if I cd into the libraries directory
> and do
> | `make` afterwards it works fine. Problems show up again when I do
> `make`
> | in hslibs. Then I get:
> | rm -f ByteArray.o; if [ ! -d ByteArray_split ]; then mkdir
> | ByteArray_split; else /usr/bin/find ByteArray_split -name '*.o' |
> xargs rm
> | -f __rm_food; fi;
> | ../../ghc/compiler/ghc-inplace -ldl -cpp -fglasgow-exts -Icbits
> -imonads
> | -package-name lang -O -Rghc-timing  -split-objs    -c ByteArray.lhs -o
> | ByteArray.o
> |
> | ByteArray.lhs:34:
> |     failed to load interface for `PrelByteArr':
> |         Could not find interface file for `PrelByteArr'
> | <<ghc: 10332428 bytes, 4 GCs, 9760/9760 avg/max bytes residency (1
> | samples), 7M in use, 0.02 INIT (0.01 elapsed), 0.26 MUT (0.52
> elapsed),
> | 0.09 GC (0.15 elapsed) :ghc>>
> |
> | I guess that the make process isn't quite made for restarting in the
> | middle of everything...
> | Any hints are welcome.
> |
> | I'm using ghc 5.04 to compile the compiler.
> |
> | > uname -a
> | SunOS 5.7 Generic_106541-12 sun4u sparc SUNW,Ultra-80
> | > gcc --version
> | 2.95.3
> |
> | The linker is Sun's own linker.
> |
> |
> | Cheers,
> |
> |     /Josef
> |
> | _______________________________________________
> | Glasgow-haskell-bugs mailing list
> | Glasgow-haskell-bugs@xxxxxxxxxxx
> | http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
>
>
> _______________________________________________
> Glasgow-haskell-bugs mailing list
> Glasgow-haskell-bugs@xxxxxxxxxxx
> http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
>