[slime-devel] Problems with SLIME and SBCL

Hello, since quite some time I am not able to run SLIME together with SBCL anymore. I did not care too much, since CMUCL and Allegro worked. However, I now am looking again at using OS threads on Linux/AMD64, and SBCL is my only option here. So I am interested to get it working again. Thus, I downloaded sbcl-0.9.9-x86-64-linux, CVS slime from today (Emacs on this machine is version 21.4.1), and did the usual installation as (add-to-list 'load-path PATH-TO-SLIME) (require 'slime) (slime-setup) Unfortunately, M-x slime does not succeed. The inferior-lisp buffer shows: WARNING: These Swank interfaces are unimplemented: (ACTIVATE-STEPPING CALLS-WHO DISASSEMBLE-FRAME INSPECT-FOR-EMACS SLDB-BREAK-AT-START SLDB-BREAK-ON-RETURN WHO-BINDS WHO-CALLS WHO-MACROEXPANDS WHO-REFERENCES WHO-SETS WHO-SPECIALIZES) T * ;; Swank started at port: 34981. NIL * Process inferior-lisp hangup Can anyone help me? Thank you very much, Nicolas.

<Nicolas.Neuss@iwr.uni-heidelberg.de> wrote:
Process inferior-lisp hangup
Are you using a Linux kernel version earlier than 2.6.11? There's some signal handling oddity there, which causes a SIGHUP to be delivered to SBCL when started from an emacs comint buffer. If upgrading the kernel is not an option for you, there's a workaround of setting inferior-lisp-program to "nohup sbcl" instead of "sbcl". -- Juho Snellman

Juho Snellman <jsnell@iki.fi> writes:
<Nicolas.Neuss@iwr.uni-heidelberg.de> wrote:
Process inferior-lisp hangup
Are you using a Linux kernel version earlier than 2.6.11?
Yes, it is 2.6.9.
signal handling oddity there, which causes a SIGHUP to be delivered to SBCL when started from an emacs comint buffer.
If upgrading the kernel is not an option for you, there's a workaround of setting inferior-lisp-program to "nohup sbcl" instead of "sbcl".
Thank you! Indeed I cannot upgrade the kernel, but "nohup sbcl" seems to work fine. Maybe this workaround could be mentioned in slime/PROBLEMS? Yours, Nicolas.
participants (2)
-
Juho Snellman
-
Nicolas Neuss