Recent increases in hard fault rates in modern chip multi-processors have led to a variety of approaches to try and save manufacturing yield. Among these are: fine-grain fault tolerance (such as error correction coding, redundant cache lines, and redundant functional units), and large-grain fault tolerance (such as disabling of faulty cores, adding extra cores, and core salvaging techniques). This paper considers the case of core salvaging techniques and the heterogeneous per- formance introduced when these techniques have some salvaged and some non-faulty cores. It proposes a hypervisor-based hardware thread scheduler, triggered by detection of spin locks and thread imbalance, that mitigates the loss of throughput resulting from this het- erogeneity. Specifically, a new algorithm, called Most ProgressMade algorithm, reduces the number of synchronization locks held on a salvaged core and balances the time each thread in an application spends running on that core. For some benchmarks, the results show as much as a 2.68x increase in performance over a salvaged chip multi-processor without this technique.

Les travaux présentés portent sur l'intégration de la programmation basée entités des machines-outils à commande numérique (MOCN). Cette programmation émergente s'appuie sur les concepts véhiculés par le standard STEP-NC et la norme ISO 14649. Quatre axes de recherche sont proposés. Le premier concerne l'évolution de la programmation actuelle vers la programmation basée entités et l'enrichissement du standard STEP-NC. Pour ceci, nous proposons deux contributions principales au standard STEP-NC dans le cadre de notre participation au comité normatif de l'ISO 14649 : l'intégration de stratégies d'usinage à motifs avec pour application l'usinage trochoïdal et le tréflage, et l'élaboration d'un modèle fonctionnel STEP-NC des MOCN. Le deuxième axe se focalise sur l'évolution du système de fabrication pour tirer profit de la programmation basée entités. Nous proposons un concept de système étendu de fabrication intégrée XMIS (eXtended Manufacturing Integrated System) ainsi qu'une déclinaison concrète à travers la CAO étendue et la CN étendue. Au cours du troisième axe, nous étendons l'étude à l''intégration de plusieurs procédés au sein d'une même gamme de fabrication STEP-NC et étudions l'organisation des outils de simulation dans ce contexte multi-process avec la description d'espaces de simulation. Le quatrième axe porte sur la validation et la mise en œuvre des concepts proposés avec le développement de la plate-forme SPAIM (STEP-NC Platform for Advanced and Intelligent Manufacturing) qui intégre la programmation basée entités du modèle à la pièce usinée sur des moyens industriels et montre l'impact multidirectionnel des modifications tout au long de la chaîne numérique.

Objective: Dog owners have been shown to walk more per week compared to non-dog owners; however, 60% of dog owners are still not walking their dogs at intensities sufficient to reap optimal health benefits. The aim of this study was to evaluate the feasibility, acceptability, and efficacy of a 9-week feasibility randomized controlled trial involving a program of six weekly scheduled instructor-led group dog walks supplemented with theory-based strategies to encourage increased dog walking among dog owners in Greater Victoria, BC. Methods: This study was based on the multi-process action control (M-PAC) framework and utilized an open parallel randomized controlled trial design involving experimental and waitlist-control group participants. Quantitative data was collected using pedometers and self-report measures. A program evaluation survey was administered upon the completion of the study. Primary outcomes examined the feasibility and acceptability of the program; secondary outcomes analyzed pedometry and self-report moderate-to-vigorous physical activity (MVPA) data; and tertiary outcomes observed changes in participants’ perceptions of M-PAC constructs. Percentage calculations were used to obtain primary outcomes, and analysis of covariance (ANCOVA; controlling for baseline) was performed to examine secondary and tertiary outcomes to explore the direction of effects and obtain a first estimate of expected effect sizes. Eligibility: Male and female adults aged 18+ living in Greater Victoria, BC, who owned at least one healthy and friendly dog aged six months and above, who were not meeting recommended guidelines of 150 minutes of MVPA per week, and who were medically cleared to participate. Results: Feasibility outcomes included 74 interested responses, 23% recruitment rate (n = 17), 94% retention rate (n = 16), and 94% adherence rate (n = 15). Program participants were overall (very) satisfied with the program – worksheets (62.5%), program instructor (100%), various program/group dog walks logistics (75% to 100%). Total weekly step counts and average daily step counts, MVPA dog walking, and MVPA with dog increased at the end of the program and at follow-up, resulting in large effect sizes when compared to the waitlist-control group. MVPA dog walking and total MVPA (with and without dog) exceeded recommended guidelines at follow-up. Positive changes across time were observed for dog responsibility and M-PAC constructs of affective judgments, opportunity, planning, identity, and habit, resulting in medium and large effect sizes when compared to the waitlist-control group. Conclusions: This six-week group dog walking program is overall feasible, acceptable, and efficacious in encouraging increased dog walking and MVPA among dog owners. Attendance at weekly scheduled instructor-led group dog walks and exposure to the M-PAC construct worksheets resulted in program participants’ adoption and maintenance of positive behavioral changes at the end of the program and at follow- up. Program participants reported enjoying the program and being (very) satisfied with it. It is recommended for future studies to refine/modify initial recruitment strategies and eligibility criteria, reimburse medical/veterinarian clearance costs to reduce cost-related barriers to participation, offer a variety of options for program delivery (e.g., different locations/schedules/seasons, online programs, multi-site study) to accommodate more participants, and apply the M-PAC framework to a larger sample. / Graduate

Tato práce rozšiřuje nástroj ANaConDA pro dynamickou analýzu vícevláknových programů o možnost analyzovat také programy víceprocesové. Část práce se soustředí na popis nástroje ANaConDA a mechanismů, které pro monitorování využívá, a na jejich nutné úpravy vzhledem k rozdílům procesů a vláken. Tyto zahrnují nutnost složitějších mechanismů pro meziprocesovou komunikaci, nutnost překládat logické adresy na jiný jednoznačný identifikátor a monitorování obecných semaforů. Rozšíření pro monitorování procesů tyto problémy řeší za vývojáře analyzátorů, čímž velmi zjednodušuje jejich vývoj. Užitečnost rozšíření je ukázána na implementaci dvou analyzátorů pro detekci souběhu (AtomRace a FastTrack), které bylo dosud možné využít pouze na vícevláknové programy. Implementace algoritmu FastTrack využívá happens-before relaci pro obecné semafory, která byla také definována jako součást této práce. Experimenty s analyzátory na studentských projektech ukázaly, že nástroj ANaConDA je nyní schopen detekovat paralelní chyby i ve víceprocesových programech a může tak pomoci při vývoji další skupiny paralelních programů.