- 09 Jul, 2015 7 commits
-
-
Morris Jette authored
Some dead assignments, memory leaks, unitialized variables, etc.
-
Morris Jette authored
Changed spaces to tabs at start of lines. Minor changes to some formatting. Added the new files to the RPM (slurm.spec file). Prevent memory leak of "l_name" variable if which_power_layout() function is called more than once. Initialize "cpufreq" variable in powercap_get_cpufreq() function. Array "tmp_max_watts_dvfs" could be NULL and used if "max_watts_dvfs" variable is NULL in powercap_get_node_bitmap_maxwatts_dvfs() Variable "tmp_pcap_cpu_freq" could be used with uninitialized value in function _get_req_features() Variable "tmp_max_watts" could be used with uninitialized value in function _get_req_features() Array "tmp_max_watts_dvfs" could be used with uninitialized value in function _get_req_features() Array "allowed_freqs" could be NULL and used if "node_record_count" variable is zero in powercap_get_job_nodes_numfreq() Overwriting a memory buffer header (especially with different data types) is just asking for something bad to happen. This code from function powercap_get_job_nodes_numfreq(): allowed_freqs = xmalloc(sizeof(int)*((int)num_freq+2)); allowed_freqs[-1] = (int) num_freq; Clean up memory on slurmctld shutdown
-
Yiannis Georgiou authored
-
Yiannis Georgiou authored
[PATCH 4/5] correct bugs in powercapping logic and fix function for optimal frequency
-
Yiannis Georgiou authored
[PATCH 3/5] Adapt the power_cpufreq layout and set the CurrentCoreWatts value based on the selection of the resources and the choice of CPU Frequency. CurrentCoreWatts are set to IdleWatts when one or more jobs occupy other resources of the node and set to 0 when the node is free. The above algorithm is activated only when powercapping is active.
-
Yiannis Georgiou authored
[PATCH 2/5] Add powercapping logic to make use of the power_cpufreq layout. The new logic allows the usage of only one of the existing layouts either power or power_cpufreq to take into account for the power adaptive scheduling. It considers the cpu frequency possibilities of the nodes as populated within the layouts along with the jobs' choices for allowed cpu frequencies for their execution. Based on the available power budget and the cluster's current maximum consumption it calculates if the job can be allocated and if yes with which allowed CPU frequency.
-
Danny Auble authored
-
- 08 Jul, 2015 15 commits
-
-
Danny Auble authored
-
Yiannis Georgiou authored
[PATCH 1/5] powercapping: add new power_cpufreq layout to use as an alternative to the power layout for power adaptive scheduling
-
Morris Jette authored
-
Morris Jette authored
-
Danny Auble authored
-
David Bigagli authored
-
Morris Jette authored
-
Morris Jette authored
It was missing a "\"
-
Morris Jette authored
-
Morris Jette authored
Conflicts: META NEWS
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
-
- 07 Jul, 2015 18 commits
-
-
Nathan Yee authored
-
Nathan Yee authored
-
Danny Auble authored
-
Nathan Yee authored
-
Danny Auble authored
-
Trey Dockendorf authored
-
Trey Dockendorf authored
Add job record qos field and partition record allow_qos field.
-
Trey Dockendorf authored
-
Morris Jette authored
-
Trey Dockendorf authored
This patch moves the QOS update of an existing job to be before the partition update. This ensures a new QOS value is the value used when doing validations against things like a partition's AllowQOS and DenyQOS. Currently if a two partitions have AllowQOS that do not share any QOS, the order of updates prevents a job from being moved from one partition to another using something like the following: scontrol update job=<jobID> partition=<new part> qos=<new qos>
-
Morris Jette authored
-
Morris Jette authored
-
Danny Auble authored
-
Danny Auble authored
-
Morris Jette authored
-
David Bigagli authored
-
Danny Auble authored
-
Danny Auble authored
value for booleans.
-