Open Source Cloud Authors: Elizabeth White, Liz McMillan, Yeshim Deniz, Zakia Bouachraoui, Pat Romanski

Related Topics: Open Source Cloud

Open Source Cloud: Article

Using Interrupts For Timing

Using Interrupts For Timing

This is an excerpt from Chapter 11 of Embedded Linux by Craig Hollabaugh, published by Addison-Wesley. ISBN 0672322269. No part of this excerpt may be reproduced, in any form or by any means without the prior written permission of the publisher. (c) 2002 Addison-Wesley.

The first snowfall at Silverjack filled the air with excitement. Soon the village would bustle and the slopes would open. As in past years, many skiers and snowboarders would race against the clock on the mountain's racecourse. This year, however, Project Trailblazer wants to collect, display, log, and post race results. Using embedded Linux for Silverjack race timing will reduce data entry errors, increase efficiency, and enhance the racers' experience.

This article series explores using Linux as an event timer with 1-millisecond resolution. The Project Trailblazer engineers need to develop interrupt handlers for the target boards-the MZ104, the MediaEngine, and the RPX-CLLF. Then they need to use these handlers to measure average interrupt latencies. The engineers need to design a race timer that uses a split interrupt approach with "top half" and "bottom half" handler routines, tasklets, and a self-rescheduling kernel timer.

Linux Timing Sources
Linux provides several mechanisms for event timing, ranging from 1-second down to submicrosecond resolution. This section discusses the following timing sources:

  • date-The resolution of date is 1s. Using the +%s formatter, date returns the number of seconds that have elapsed since 00:00:00, January 1, 1970.
  • jiffies-The resolution of jiffies is 10ms. On all platforms, Linux configures a hardware timer that interrupts the processor periodically-typically every 10ms (this value is defined in the kernel source as HZ). The timer interrupt routine increments the kernel variable jiffies by 1. Therefore, the value of jiffies represents the number of 10ms increments that have occurred since booting. It is possible to increase jiffies resolution by altering the HZ source declaration. However, this could ultimately lower overall system performance due to increased interrupt processing. Other mechanisms offer higher resolution.
  • PSR-The resolution of PSR (which stands for processor-specific registers) is various. Some newer processors contain two 32-bit counters that increment on every system clock cycle. The counters accessed through PSR provide resolutions that are dependent on processor clock speed. Kernel source code provides function calls to access the PSR. For example, the rdtsc function returns the timestamp counter (TSC) on Pentium and newer processors, and the get_tbl function returns the mftb register value on PowerPCs. 486 and ARM processors do not contain system clock counters.
  • get_cycles-The resolution of get_cycles is various. The get_cycles function, which is defined all on platforms, returns a count of system clock cycles that fit into a single CPU register. Typically this is the lower half of the two 32-bit counters mentioned previously. If the processor doesn't contain a clock counter, get_cycles returns 0. get_cycles returns 0 on 486 and ARM processors.
  • do_gettimeofday-The resolution of do_gettimeofday is about 1ms. The do_gettimeofday function fills a timeval data structure with the number of seconds and microseconds elapsed since booting. The x86 and PowerPC kernel source claim near-microsecond resolution for do_gettimeofday.

    The Project Trailblazer race timer needs to have 1ms resolution. This resolution eliminates using date and jiffies. The MZ104's 486 and the MediaEngine's ARM processors don't contain system clock counters, so the Project Trailblazer engineers also can't use PSR. The get_cycles function returns only a 32-bit counter value. On the slowest clocked target board, the RPX-CLLF (at 70MHz), the get_cycles function can only count 61 seconds' worth of microseconds. A typical race will be longer than 61 seconds, so the get_cycles function won't work for Project Trailblazer. This leaves the do_gettimeofday function. With near-microsecond resolution and a 32-bit counter of seconds (232 seconds = 49,710 days), the do_gettimeofday will provide accuracy for even the longest race. The race timer's interrupt routine can timestamp the race's start and finish by using do_gettimeofday. A scheduled task can later compute the race's overall time.

    Interrupt latency is a measure of the time between an event occurring (for example, the race start) and when the processor executes the interrupt handler code. Even with microsecond timing access, race timing won't have millisecond accuracy if the system interrupt latency exceeds a millisecond. The next section examines the Project Trailblazer engineers' development of interrupt handlers within a device driver to measure average interrupt latencies for the MZ104, MediaEngine, and RPX-CLLF target boards.

    Measuring Interrupt Latency
    Interrupt latency is comprised of hardware propagation time, register saving, and software execution. Propagation time and register saving occur extremely fast-on the order of 10s of nanoseconds. Software execution, on the other hand, can be extremely slow. The Linux kernel allows device drivers to disable interrupts by using the cli system call. While interrupts are disabled, other interrupts can occur, but their service routines are not executed until interrupts are re-enabled. Developers disable interrupts to protect critical sections of their code. For example, a video card driver might disable interrupts for 16ms while waiting for video sync. Or a serial card driver might disable interrupts during a byte transmission. Therefore, maximum interrupt latency is a combination of system hardware, peripherals, and the quality of the device driver software. Even with source code availability, this combination and the asynchronous nature of interrupts makes calculating the maximum interrupt latency a difficult, if not impossible, task. On a lightly loaded system with little or no peripherals, it's likely that the average interrupt latency is much smaller than the maximum. Even so, the Project Trailblazer engineers had no idea of the average interrupt latency magnitude. Was it microseconds or milliseconds? They decided to measure it for each target board.

    When you are writing device drivers, you should disable interrupts only when your driver requires absolutely no interruption. While interrupts are disabled, Linux does not update system timers, transfer network packets to and from buffers, or update video information. You should perform your interrupt task as quickly as possible. For lengthy tasks, you should split your handler and use tasklets to perform most of the processing.

    Each target board-the MZ104, the MediaEngine, and the RPX-CLLF-has external connections that are capable of generating processor interrupts. In addition, each board can generate an interrupt signal by using one of its own output ports. In the device driver code we're about to discuss, one function asserts the interrupt signal, which causes the interrupt handler to execute. The handler simply deasserts the interrupt signal and exits. The interrupt signal assertion and deassertion can be viewed on an oscilloscope, to determine the average interrupt latency.

    To measure the average interrupt latency, the Project Trailblazer engineers need to do the following: 1. Write a device driver that contains the following:

  • Configuration code for the board's interrupt controller, if necessary
  • A request for the interrupt from Linux
  • A function that asserts the interrupt signal
  • An interrupt handler that deasserts the interrupt signal
    2. Connect an output port pin to a processor interrupt pin and to the oscilloscope.
    3. Load the device driver.
    4. Assert the interrupt pin.
    5. Watch for the deassertion.
    6. Measure the interrupt latency.
    7. Repeat steps 4, 5, and 6 to determine the average interrupt latency.

    As discussed in the section "Linux Timing Sources," earlier, the do_gettimeofday function provides timing information with microsecond resolution. By using two calls to do_gettimeofday, the device driver itself can calculate its own instantaneous interrupt latency. (Instantaneous in this case refers to the interrupt latency associated with a single interrupt event.) The instantaneous interrupt latency value will not be the same for interrupt events. The average interrupt latency is the numerical average of several instantaneous interrupt latency values.

    The magnitude of the average interrupt latency is what the engineers want to determine for each of their target boards. They need to find out if it is greater than 1ms. A race timer with 1ms accuracy requires a timing source with interrupt latencies of less than 1ms. The engineers need to write an interrupt latency device driver for each of their target boards. They then need to execute this device driver several times, record the instantaneous interrupt latency values, and compute an average interrupt latency value for each board. Using the interrupt latency device driver's output combined with an oscilloscope measurement, the engineers can also verify the microsecond accuracy of the do_gettimeofday function.

    The three target board interrupt latency device drivers contain four functions: init, proc_read, interrupt_latency, and cleanup. The init function creates a /proc directory entry called interrupt_latency and configures the processor's interrupt controller and port settings. Reading from the interrupt_latency file calls the function proc_read, which generates an interrupt electrical signal. The device driver handles the interrupt by calling the interrupt_latency function. It then computes and prints the instantaneous interrupt latency value. The driver also maintains an interrupt counter to aid in debugging.

    Part 2 will cover measuring interrupt latency on the MZ104.

  • More Stories By Maureen O'Gara

    Maureen O'Gara the most read technology reporter for the past 20 years, is the Cloud Computing and Virtualization News Desk editor of SYS-CON Media. She is the publisher of famous "Billygrams" and the editor-in-chief of "Client/Server News" for more than a decade. One of the most respected technology reporters in the business, Maureen can be reached by email at maureen(at)sys-con.com or paperboy(at)g2news.com, and by phone at 516 759-7025. Twitter: @MaureenOGara

    Comments (1) View Comments

    Share your thoughts on this story.

    Add your comment
    You must be signed in to add a comment. Sign-in | Register

    In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.

    Most Recent Comments
    Michael Choo 12/15/03 02:15:00 AM EST

    Dear sir/madam,

    hi! If I write interrupt handler make it as long interrupt latency. I mean inside isr function, call many functions. after then, iret exit isr function. What will happen?

    my experience is:
    1. beginning, system will service interrupts few times. after then stop response even has inteerupt event in.

    2. some unpredicatable happen.

    What is your result? Do you have any document mention this kind of problem?

    Thank you!

    Michael Choo

    IoT & Smart Cities Stories
    Early Bird Registration Discount Expires on August 31, 2018 Conference Registration Link ▸ HERE. Pick from all 200 sessions in all 10 tracks, plus 22 Keynotes & General Sessions! Lunch is served two days. EXPIRES AUGUST 31, 2018. Ticket prices: ($1,295-Aug 31) ($1,495-Oct 31) ($1,995-Nov 12) ($2,500-Walk-in)
    Andrew Keys is Co-Founder of ConsenSys Enterprise. He comes to ConsenSys Enterprise with capital markets, technology and entrepreneurial experience. Previously, he worked for UBS investment bank in equities analysis. Later, he was responsible for the creation and distribution of life settlement products to hedge funds and investment banks. After, he co-founded a revenue cycle management company where he learned about Bitcoin and eventually Ethereal. Andrew's role at ConsenSys Enterprise is a mul...
    Nicolas Fierro is CEO of MIMIR Blockchain Solutions. He is a programmer, technologist, and operations dev who has worked with Ethereum and blockchain since 2014. His knowledge in blockchain dates to when he performed dev ops services to the Ethereum Foundation as one the privileged few developers to work with the original core team in Switzerland.
    René Bostic is the Technical VP of the IBM Cloud Unit in North America. Enjoying her career with IBM during the modern millennial technological era, she is an expert in cloud computing, DevOps and emerging cloud technologies such as Blockchain. Her strengths and core competencies include a proven record of accomplishments in consensus building at all levels to assess, plan, and implement enterprise and cloud computing solutions. René is a member of the Society of Women Engineers (SWE) and a m...
    Digital Transformation and Disruption, Amazon Style - What You Can Learn. Chris Kocher is a co-founder of Grey Heron, a management and strategic marketing consulting firm. He has 25+ years in both strategic and hands-on operating experience helping executives and investors build revenues and shareholder value. He has consulted with over 130 companies on innovating with new business models, product strategies and monetization. Chris has held management positions at HP and Symantec in addition to ...
    The challenges of aggregating data from consumer-oriented devices, such as wearable technologies and smart thermostats, are fairly well-understood. However, there are a new set of challenges for IoT devices that generate megabytes or gigabytes of data per second. Certainly, the infrastructure will have to change, as those volumes of data will likely overwhelm the available bandwidth for aggregating the data into a central repository. Ochandarena discusses a whole new way to think about your next...
    CloudEXPO | DevOpsSUMMIT | DXWorldEXPO are the world's most influential, independent events where Cloud Computing was coined and where technology buyers and vendors meet to experience and discuss the big picture of Digital Transformation and all of the strategies, tactics, and tools they need to realize their goals. Sponsors of DXWorldEXPO | CloudEXPO benefit from unmatched branding, profile building and lead generation opportunities.
    Dynatrace is an application performance management software company with products for the information technology departments and digital business owners of medium and large businesses. Building the Future of Monitoring with Artificial Intelligence. Today we can collect lots and lots of performance data. We build beautiful dashboards and even have fancy query languages to access and transform the data. Still performance data is a secret language only a couple of people understand. The more busine...
    All in Mobile is a place where we continually maximize their impact by fostering understanding, empathy, insights, creativity and joy. They believe that a truly useful and desirable mobile app doesn't need the brightest idea or the most advanced technology. A great product begins with understanding people. It's easy to think that customers will love your app, but can you justify it? They make sure your final app is something that users truly want and need. The only way to do this is by ...
    DXWorldEXPO LLC announced today that Big Data Federation to Exhibit at the 22nd International CloudEXPO, colocated with DevOpsSUMMIT and DXWorldEXPO, November 12-13, 2018 in New York City. Big Data Federation, Inc. develops and applies artificial intelligence to predict financial and economic events that matter. The company uncovers patterns and precise drivers of performance and outcomes with the aid of machine-learning algorithms, big data, and fundamental analysis. Their products are deployed...