Closed GoogleCodeExporter closed 9 years ago
i couldn't reproduce the problem(at Cygwin32 at Win7 64 mosh/nmosh consumes
about 36MB).
Please provide following information :
- CPU
- OS(32bit or 64bit kernel?)
- Physical memory size
- Any GC warnings observed (unset GC_NO_BLACKLIST_WARNING or
GC_LARGE_ALLOC_WARN_INTERVAL)
and, please try :
- run with --verbose. nmosh falls into infinite-loop at some situation(see
Issue 115)
- sync with git master, then regenerate nmosh runtime by ./gen-git-build.sh
Original comment by oku...@gmail.com
on 20 Jul 2010 at 10:54
s/Issue 115/Issue 155/
Original comment by oku...@gmail.com
on 20 Jul 2010 at 10:54
Hi.
Linux 2.6.32-23-generic #37-Ubuntu SMP Fri Jun 11 08:03:28 UTC 2010 x86_64
GNU/Linux
Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz
Memory 1GB
No GC warnings.
Original comment by hige...@gmail.com
on 21 Jul 2010 at 8:43
the problem was not reproduced on the same kernel, same CPU MA(Core 2 Duo),
same amount of memory..
if your problem is highly reproductive, your nmosh cache and/or nmosh image may
be corrupted.
remove ~/.nmosh-cache directory and regenerate it.
while nmosh expander is not weak-GC robust, data-centric workload should:
- execute cached script
-- cache your library before executing your program ("nmosh xxxxx.sls" can do
this. it will expand, compile, cache the library and exit.)
-- syntax object's usage environment and debug symbols may not freed if they
are pointed by false-pointers.
- avoid cons-ing many times
-- not to use format or display. use FASL instead and display or process it in
separate programs.
Debugging information will consume 1.5x - 2x memory compared with psyntax-mosh.
Currently, there is no way to disable it.
Original comment by oku...@gmail.com
on 21 Jul 2010 at 12:17
status: won't fix at 0.2.6. I couldn't reproduce the problem.
Original comment by oku...@gmail.com
on 27 Sep 2010 at 3:33
Original comment by hige...@gmail.com
on 24 Oct 2010 at 12:37
Original issue reported on code.google.com by
hige...@gmail.com
on 20 Jul 2010 at 5:44Attachments: