exact contents of PCB and TCB?exact shared flags to clone?only analyze a concurrency problem?hardware?thread advantages same as processes? (lec12 or 13)kernel memory?linux specific code?windows specific anything?threads in kernel vs userturnaround time for different schedulers, how much variation?critical sections simultaneous?amdahl's lawshort answer questions code based/conceptual?coding mostly concurrency?example: multiple queuesschedulabilityutilizationidentify things i tell younothere are hardware questionsthreads give the advantages within the same program.processes launch other instances of the same program.linux: kernel physical mem is mapped to every process's address spaceyou shouldn't need a separate saved kernel stack for every process.(but someone might do it that way idk, not on the exam)coding neither confirm nor deny.linux-specificity will be limitednoscheduler can see threads in kernel, not in userhere are processes, find the schedule. (i will give quantums)only if data is not sharednumbers will be easyboth?probablyin general, can be separate virtual space from processes, implementation-dependentif you need to write code, it will be somethingyou've seen in classif you need to write code, it will be somethingyou've seen in classnamereleaseprocess timeP1045604P2P3quantums: (1, 2, 4)highest priorityqueuesq1: P1 (4), P2 (5)q2:q4:queuesq1: P2 (5)q2: P1 (3)q4:queuesq1:q2: P1 (3), P2 (4)q4:queuesq1: P3 (6)q2: P2 (4)q4: P1 (1)0: P11: P22: P14: P3process go back up in priority if they did not get full quantumq1q1q2U = sum( pk / Tk )U = sum( pk / Tk ) is for Real-Time, single core, CPU-bound1 - p^n equation is for i/o bound, single core, multiprogrammingMidterm Q & A