Accept negative -par values to leave N CPUs free.

[dserrano5]

Giving negative values to the -par parameter causes bitcoin to run with 1 verification thread. It can be useful to tell bitcoin to use all existing CPUs but leave e.g. 1 of them free for another tasks; this patch allows that by specifying -par=-1.

[BitcoinPullTester]

Automatic sanity-testing: PASSED, see http://jenkins.bluematt.me/pull-tester/817abe0c4b89d743dab617fb1912a5317dc75b13 for binaries and test log.

[gavinandresen]

ACK

[sipa]

ACK; needs a rebase, though.

[Diapolo]

@sipa What about -rpcthreads, does this make sense there too? Or for -genproclimit (which seems missing from our help message in init.cpp?)?

[sipa]

@Diapolo As long as we don't have good lock isolation, RPC processing is effectively single-threaded anyway. I don't even think there's any good reason for more than 2 RPC threads right now.

[Diapolo]

@sipa AFAIK we currently use 4 RPC threads as default ;). I just was asking if these changes could be valuable for other thread-related command-line parameters also.

[dserrano5]

Re "needs a rebase", am I supposed to do that? (my knowledge of git/github is pretty basic still)

[dserrano5]

Ok, did the rebase, with help from @rdponticelli.

[sipa]

Thanks. If you can squash the two commits together also, I'll merge.

[dserrano5]

'twas easier to reset both diffs and commit a new one again ("easier" == "I've done it in the past" :)).

[BitcoinPullTester]

Automatic sanity-testing: PASSED, see http://jenkins.bluematt.me/pull-tester/22192ca744892568dc8d68fcbcec5d6d0aaa8d2e for binaries and test log.
This test script verifies pulls every time they are updated. It, however, dies sometimes and fails to test properly. If you are waiting on a test, please check timestamps to verify that the test.log is moving at http://jenkins.bluematt.me/pull-tester/current/
Contact BlueMatt on freenode if something looks broken.