Another new one to test for Linux, Mac and Windows

REMINDER TO ALL ALPHA TESTERS:
It's far easier for us to fix problems if you send message logs with the appropriate flags set. The main flags are:

<cpu_sched_debug>: problems involving the choice of applications to run.
<work_fetch_debug>: problems involving work fetch (which projects are asked for work, and how much).
<rr_simulation>: problems involving jobs being run in high-priority mode.

Use these flags from the cc_config.xml file.

Report any problems you get with it to the Alpha email list. This list needs registration.

This release contains several work fetch and CPU/GPU scheduling issues as well as two new features. The new features are as follows:

1) Suspend computation if CPU usage exceeds a certain value. Basically if the overall CPU usage of non-BOINC related processes exceed the configured value BOINC suspends computation. I think the default value right now is 25%. Right now the value can be tweaked via web based preferences.

2) The long awaited for backup project mechanism has now been implemented. Projects with a resource share of 0 are considered backup projects.

6.10.30:

- Mac Sandbox: SetupAuthorizationExecuteWithPrivileges() forks to run chmod and chown, leaving zombie processes which can exceed the system-imposed limit of processes MAXUPRC for a user. Call waitpid() to clear these zombies.

- Mac Sandbox: fix comment in Mac_SA_Secure.sh script.

- Mac Installer: update, fix bugs in script for building GridRepublic Installer.

- client: tiny code shuffle

- API: fix crashing bug. Don't memset(0) APP_INIT_DATA; it contains HOST_INFO, which now contains COPROCS, which has a vector. Define a clear() for APP_INIT_DATA.

- client: initial checkin for "don't compute if CPU usage above X"

- client: finish the above. Should add the Adv Prefs dialog as well.

- lib: Don't allow the diagnostics threads to be suspended, if a crash happens within the remaining thread the process will deadlock.

- client simulator: further fiddling around. Not done.

- client simulator: done for now

- client: brute-force attempt at eliminating domino-effect preemption:

if job A is unstarted and EDF, and there's a job B that is later in the list, is started, has the same app version, and has the same arrival time, move A after B.

- client: remove the "temp_dcf" mechanism, which had the same goal but didn't work.

- client: in computing overall debt for a project, subtract a term that reflects pending work. This should reduce repeated fetches from the same project.

- client simulator: tweaks

- client bug fixes

- client: when launching app, delete init_data.xml using switcher prior to trying to write it, in case it's owned by another user (e.g., boinc_project)

- client: fix small bug that could interfere with work fetch on hosts with both NVIDIA and ATI GPU

- client: fix work fetch bug that prevented getting work from an overworked project, even if idle instance or major shortfall.

- client: remove debug msgs

- client: if a project has zero resource share, treat it as a "backup project": fetch work from it only if there is an idle instance and no other projects have work.

- client: fix my last checkin

- client: allow zero resource share



6.10.31

- client: newer ATI drivers crash if struct_size isn't filled out before the structure is used.

6.10.33 released for the Mac only.

This is to correct an issue with the Mac installer requiring admin rights. Mac users should use this version instead.
____________
BOINC blog