Closed verement closed 11 years ago
The trouble appears to stem from:
7 vm_backtrace_push (arg=0xf779fed8, file=272247168, line_no=136, name=0)
at vm.c:766
where vm.c:766 tries to reference RSTRING_PTR(name)
and name
is NULL.
At this point however I'm not sure whether this is a fault with my implementation or a fault with Ruby.
Upgrading Ruby to version 1.9.2.180-5 of Debian's ruby1.9.1 package seems to have solved this, so I suspect it was in fact a bug in Ruby.
I get segfault and I do have /spu mounted and I use ruby 1.9.2p180 with YDL.
Any idea? BTW running cellminer with -d switch doesn't provide any extra information.
That looks to be a very different problem; it appears to be segfaulting within spe_program_load()
. Could you open a different tracking issue for that, and include a build log as well as the segfault trace?
Thanks.
When
spe_context_create()
fails, for example if spufs is not mounted on /spu,rb_raise()
is called within theBitcoin::SPUMiner
allocator function. However, during the subsequent process of creating a stack trace, Ruby will segfault.