2754
Comment: Building on hppa
|
2954
segfault in hppa.
|
Deletions are marked like this. | Additions are marked like this. |
Line 42: | Line 42: |
* hppa (paer) (Building now) | * hppa (paer) |
Line 62: | Line 62: |
now. | now. I got to a segfault: {{{ (sid)marcot@paer:~/ghc6-6.12.1$ inplace/lib/ghc-stage2 --interactive GHCi, version 6.12.1: http://www.haskell.org/ghc/ :? for help Segmentation fault }}} It's like with powerpc. |
Marco Túlio Gontijo e Silva
Email: <marcot AT SPAMFREE debian DOT org>
IRC: marcot, irc.oftc.net, #debian-haskell.
Debian Developer information
My bugs : marcot@debian.org
My QA page : http://qa.debian.org/developer.php?login=LOGIN@debian.org
Current work
It's possible that I left some files in the Debian machines I've used, so I'll check my homedir in them latter.
runhaskell is not available in arches without ghci
It should be, since it can compile the code an run it latter. I'll write a patch to ghc about it. The GHCI macro will be useful.
agda-stdlib
Laney@#debian-haskell@irc.oftc.net asked me to test if agda-stdlib ?from git builds in the arches that failed to build before, so I'm building it now.
No LibFFI in Alpha
I'm building ghc6-6.12.1-6 with a patch to see if it corrects the bug on libffi.
No GHCi in all arches
Where to work on
This is the list of arches without a working GHCi.
- alpha (albeniz) (Building now)
- armel (agricola) (Building now)
- hppa (paer)
- powerpc (pescetti)
- s390 (zelenka)
I could not connect to mips or mipsel to check about it in them.
alpha
I'm checking about libffi first, then I'll investigate on it.
armel
I've made this patch to armel, to check if it builds correctly. I'm building it now.
hppa
I've made a patch to it, but I got a compile error. It seems that it EM_PARISC is not defined in the build. I think I'll try to replace it with 15 just to see what happens. I'm building now. I got to a segfault:
(sid)marcot@paer:~/ghc6-6.12.1$ inplace/lib/ghc-stage2 --interactive GHCi, version 6.12.1: http://www.haskell.org/ghc/ :? for help Segmentation fault
It's like with powerpc.
powerpc
segfault when ghci is started. No idea about it.
s390
It seems that the code for Elf Relocation must be written by a s390 expert.
GHC is not up-to-date in all ports
I could not connect to m68k to check about it.
hurd-i386
It seems that there's a non-deterministic bug happening when ghc 6.8 is executed. Maybe it's a bug in something introduced after 6.6. This needs more investigation.