lp:~bhill/epics-gateway/epics-gateway-preempt

Branch merges

Propose for merging

No branches dependent on this one.

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related bugs

Related blueprints

184. By Bruce Logan Hill <bhill@rhel6-64h> on 2014-08-15

The changes in this checkin were done by Sheng Peng in an effort
to run the gateway using a preemptive channel access callback context.
I have very little specific info on the testing other than to say we ran it
on several gateways for a few weeks.

There were a few anecdotal reports of better throughput, but we also
had it crash 5 times over 9 days and switched to using a gateway version
using non-preemptive ca with file descriptors via USE_FDS.

This needs more work, possibly changing to using the multi-threaded CA
server, as the only multi-threaded portion of this version is the CA client side
which spawns two threads per connection. We probably need some mutex protection
in the callbacks as well to prevent crashes.

183. By Ralph Lange on 2014-08-07

Fix CPU load formula for SMP (Posix) (provided by J. Anderson)

182. By Ralph Lange on 2014-07-31

Merge caputlog branch by J. Priller / M. Konrad (MSU/FRIB)

181. By Ralph Lange on 2014-07-30

Add appropriate .bzrignore (copied from EPICS Base)

180. By Janet B. Anderson on 2011-04-05

Committed freebsd portability changes contributed by Gerrit Kuhn.

179. By Janet B. Anderson on 2009-09-17

Moved lib gdd before Com. libgdd.a(aitConvert.o) uses cvtDoubleToString.

178. By Janet B. Anderson on 2009-09-17

Added heartbeat pv note and gateway version change.

177. By Janet B. Anderson on 2009-09-17

New Release 2.0.4.0

176. By Jeff Hill on 2009-08-31

o Fixed issue where put callback response from IOC with failure status does not result in put notify response being sent to gw's client
o Fixed problems with writes not being processed when the gateway runs in -no_cache mode and a stale variable in the gateway is reactivated
o Improved my patch to the gateway allowing it to determine if a write is from a ca_put or ca_put_callback, and to accordingly dispatch the appropriate form of ca_put to the IOC.

175. By Jeff Hill on 2009-08-27

more fixes for mantis 362