Searching All Intel Documentation
65672 Matching Results
Not finding what you're looking for? Search legacy documents.
Adding Video fromIf you want to embed a video in an IDZ page, and the video is not already hosted on a supported site, you must send it to the WebOps team in a YouTrack ticket. WebOps will upload the video and create an IDZ video page that can be embedded in your page. Required Title (60 characters max, including spaces) Short description (100 characters max, including spaces) Overview/Long description (500 characters max, including spaces) Please do not use the same text as Short description; this should be a more detailed explanation of the video content. Primary Author The name of the primary... Last updated on 10/11/2017 - 11:28
Yocto* Image fromThe SDK runs on a custom Yocto image configured for real-time tasks. The configuration parameters assume the best-effort core is 0 and the target is a four-core system. The best-effort core is the one that Linux is directed to for non-critical operations. Specifically, the parameter isolcpus=1-3 means we prohibit the Linux kernel from scheduling processes on cores 1, 2, and 3. Cores 1, 2, and 3 are available for real-time workloads. The following parameters are configured: processor.max_cstate=0 intel.max_cstate=0 processor_idle.max_cstate=0 intel_idle.max_cstate=0 clocksource=tsc tsc=... Last updated on 04/19/2018 - 14:08
Code Structure fromWhen incorporating the APIs into your application, use the following basic structure: Start with tcc_init(). This function initializes the libraries. There are no required parameters, so you can use the function like this: tcc_init(); Or you can add the optional parameter .cpuid to specify the core on which your application should run. If passed, tcc_init() will set the CPU affinity and use cpuid for subsequent tcc_alloc() calls. When you call tcc_init() with .cpuid = 3, for example, the API moves the application to Core 3, where it will run. static int cpuid = 3; tcc_init(.cpuid=cpuid);... Last updated on 04/19/2018 - 14:08
Real-Time APIs fromThese APIs implement Intel® TCC features or measure the effects of those features. Cache Lock API Measurement API Last updated on 04/19/2018 - 14:08
Cache Lock API fromThe Cache Lock API enables you to control latency by reserving a region of L2 cache for time-sensitive code and data. When a CPU requests data, it first tries to find the data in the cache memory. If the data is in cache memory, a cache hit results. If the data is not in the cache memory, a cache miss results, and the CPU fetches the data from the RAM. Fetching data from cache memory is faster than fetching data from RAM. Internally, the API uses a processor feature called cache pseudo-locking (provided by L2 Cache Allocation Technology or CAT). The API makes it easy to use this feature in... Last updated on 04/19/2018 - 14:08
Measurement API fromThe Measurement API provides the ability to accurately measure the average, maximum, and minimum latencies for user-defined periods. The API is a helper based on the CPU's time stamp counter (TSC) to collect and output rudimentary execution timing for workloads. The API defines and makes use of a tcc_measurement structure to hold start and end times for one or more phases of a workload's cycle time actuals and reports the resulting processing statistics upon completion. API Flow and Implementation When using the API, keep in mind the following steps: Define the tcc_measurement structure.... Last updated on 04/19/2018 - 14:08
For more complete information about compiler optimizations, see our Optimization Notice.