this patch has resolved the hang problem when aborting a tool prepare or tool 
load 

I tried to be 'minimally invasive' to force-queue the abort to iocontrol only 
when the previous message was a tool prepare or tool load.

the more I think about it (and tried to abort another blocking operation needed 
for iocontrol-v2), I think aborts should always be force-queued regardless of 
the previous command. The reasons are twofold: first, an EMC-side abort is 
really an out-of-band notification in nature and should always travel 
first-class, not cattle-class. Second, the current patch introduces a 
dependency - abort handling is privileged for tool-prepare and tool-load, but 
not for others, which means newly added messages might again be exposed to the 
behaviour already fixed for prepare and load until explicitely added to 
emctaskmain.cc (which I had to do).

I've been running emc with unconditional force-queueing of aborts to iocontrol 
and see no negative consequences so far.

Unless somebody comes up with a good reason why not to do this, I'd like to 
generalize this as above and provide a patch to do so (which is actually a 
simplification).

-Michael


http://git.linuxcnc.org/gitweb?p=emc2.git;a=commit;h=5ebf364ad2bc051d04c779bb635ebc89ab4ed09b
------------------------------------------------------------------------------
Learn how Oracle Real Application Clusters (RAC) One Node allows customers
to consolidate database storage, standardize their database environment, and, 
should the need arise, upgrade to a full multi-node Oracle RAC database 
without downtime or disruption
http://p.sf.net/sfu/oracle-sfdevnl
_______________________________________________
Emc-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-developers

Reply via email to