Difference between revisions of "Fall 2022 SPO600 Weekly Schedule"

From CDOT Wiki
Jump to: navigation, search
(Video)
(Week 12 Deliverables)
(31 intermediate revisions by 2 users not shown)
Line 19: Line 19:
 
|2||Sep 12||[[#Week 2 - Class I|Introduction to 6502 Assembly]]||[[#Week 2 - Class II|6502 Math / Jumps, Branches, and Subroutines]]||[[#Week 2 Deliverables|Lab 2]]
 
|2||Sep 12||[[#Week 2 - Class I|Introduction to 6502 Assembly]]||[[#Week 2 - Class II|6502 Math / Jumps, Branches, and Subroutines]]||[[#Week 2 Deliverables|Lab 2]]
 
|-
 
|-
|3||Sep 19||[[#Week 3 - Class I|6502 Strings]]||[[#Week 3 - Class II|Building Code / Make and Makefiles / Autotools and Friends]]||[[#Week 3 Deliverables|Lab 3]]
+
|3||Sep 19||[[#Week 3 - Class I|6502 Strings]]||[[#Week 3 - Class II|6502 String Input / Building Code: Make and Makefiles]]||[[#Week 3 Deliverables|Lab 3]]
 
|-
 
|-
|4||Sep 26||[[#Week 4 - Class I|Compiler Optimizations]]||[[#Week 4 - Class II|ELF Files / Shared Libaries]]||[[#Week 4 Deliverables|Lab 3, September blog posts]]
+
|4||Sep 26||[[#Week 4 - Class I|Compiler Optimizations]]||[[#Week 4 - Class II|Building Code: Compiler Options, GNU Autotools/Automake]]||[[#Week 4 Deliverables|Lab 3, September blog posts]]
 
|-
 
|-
|5||Oct 3||[[#Week 5 - Class I|Introduction to 64-bit Architectures and Assembly Language (x86_64 and AArch64)]]||[[#Week 5 - Class II|Memory on 64-bit Systems]]||[[#Week 5 Deliverables|Lab 3]]
+
|5||Oct 3||[[#Week 5 - Class I|Introduction to 64-bit Architectures and Assembly Language (x86_64 and AArch64)]]||[[#Week 5 - Class II|Memory on 64-bit Systems]]||[[#Week 5 Deliverables|Lab 4]]
 
|-
 
|-
|6||Oct 10||[[#Week 6 - Class I|Single Instruction, Multiple Data (SIMD) / Scalable Vector Extensions (SVE/SVE2)]]||[[#Week 6 - Class II|Indirect Functions (GCC ifunc)]]||[[#Week 6 Deliverables|Lab 4]]
+
|6||Oct 10||[[#Week 6 - Class I|Mid-semester Sync Discussion]]||[[#Week 6 - Class II|Algorithm Selection / In-line Assembler / SIMD]]||[[#Week 6 Deliverables|Lab 5]]
 
|-
 
|-
|7||Oct 17||[[#Week 7 - Class I|Project Introduction]]||[[#Week 7 - Class II|Project Selection]]||[[#Week 7 Deliverables|Lab 5]]
+
|7||Oct 17||[[#Week 7 - Class I|Exploring 64-bit Code]]||[[#Week 7 - Class II|SVE2]]||[[#Week 7 Deliverables|Wrap up lab 5]]
 
|-
 
|-
 
|Reading||Oct 24||style="background: #f0f0ff" colspan="3" align="center"|Reading Week
 
|Reading||Oct 24||style="background: #f0f0ff" colspan="3" align="center"|Reading Week
 
|-
 
|-
|8||Oct 31||[[#Week 8 - Class I|Optimization Trade-Offs / Algorithm Selection]]||[[#Week 8 - Class II|Inline Assembler]]||[[#Week 8 Deliverables|Lab 6, October blog posts]]  
+
|8||Oct 31||[[#Week 8 - Class I|Optimization Trade-Offs / Algorithm Selection / Inline Assembler / SIMD]]||[[#Week 8 - Class II|Scalable Vector Extensions (SVE/SVE2) via Inline Assembler and C Intrinsics]]||[[#Week 8 Deliverables|October blog posts]]  
 
|-
 
|-
|9||Nov 7||[[#Week 9 - Class I|Project Discussion]]||[[#Week 9 - Class II|Demo/discussion of SVE2 Examples]]||[[#Week 9 Deliverables|Blog about project work]]
+
|9||Nov 7||[[#Week 9 - Class I|GNU ifunc & Project Overview]]||[[#Week 9 - Class II|Project Detail]]||[[#Week 9 Deliverables|Blog about ifunc and your project work]]
 
|-
 
|-
|10||Nov 14||[[#Week 10 - Class I|Project Discussion]]||[[#Week 10 - Class II|Memory Barriers]]||[[#Week 10 Deliverables|Blog about project work]]
+
|10||Nov 14||[[#Week 10 - Class I|Project Tips]]||[[#Week 10 - Class II|Advanced Memory]]||[[#Week 10 Deliverables|Blog about project work]]
 
|-
 
|-
|11||Nov 21||[[#Week 11 - Class I|Project Discussion]]||[[#Week 11 - Class II|Advanced Memory Systems]]||[[#Week 11 Deliverables|Blog about project work]]
+
|11||Nov 21||[[#Week 11 - Class I|Project Techniques]]||[[#Week 11 - Class II|Project Demo]]||[[#Week 11 Deliverables|Blog about project work]]
 
|-
 
|-
|12||Nov 28||[[#Week 12 - Class I|Project Discussion]]||[[#Week 12 - Class II|Compiler Technology]]||[[#Week 12 Deliverables|Blog about project work, November blog posts]]
+
|12||Nov 28||[[#Week 12 - Class I|Benchmarking]]||[[#Week 12 - Class II|Step-by-Step Project Minimum Requirements]]||[[#Week 12 Deliverables|Blog about project work, November blog posts]]
 
|-
 
|-
|13||Dec 5||[[#Week 13 - Class I|Project Discussion]]||[[#Week 13 - Class II|Final project instructions]]||[[#Week 13 Deliverables|Blog about project work]]
+
|13||Dec 5||[[#Week 13 - Class I|Enhancing Your Project]]||[[#Week 13 - Class II|Project Discussion]]||[[#Week 13 Deliverables|Blog about project work; Project Stage 2 due Thursday December 8 at Noon]]
 
|-
 
|-
 
|14||Dec 12||[[#Week 14 - Class I|Future Directions in Architecture]]||style="background: #f0f0ff"|(No class)||[[#Week 14 Deliverables|Project Stage 3, December blog posts]]
 
|14||Dec 12||[[#Week 14 - Class I|Future Directions in Architecture]]||style="background: #f0f0ff"|(No class)||[[#Week 14 Deliverables|Project Stage 3, December blog posts]]
Line 80: Line 80:
 
*** [[Endian|Endianness]]
 
*** [[Endian|Endianness]]
 
** Code that takes advantage of platform-specific features
 
** Code that takes advantage of platform-specific features
* Reasons for writing code in Assembly Langauge include:
+
* Reasons for writing code in Assembly Language include:
 
** Performance
 
** Performance
 
** [[Atomic Operation|Atomic Operations]]
 
** [[Atomic Operation|Atomic Operations]]
Line 105: Line 105:
 
===== Build Process =====
 
===== Build Process =====
  
Building software is a complex task that many developers gloss over. The simple act of compiling a program invokes a process with five or more stages, including pre-proccessing, compiling, optimizing, assembling, and linking. However, a complex software system will have hundreds or even thousands of source files, as well as dozens or hundreds of build configuration options, auto configuration scripts (cmake, autotools), build scripts (such as Makefiles) to coordinate the process, test suites, and more.
+
Building software is a complex task that many developers gloss over. The simple act of compiling a program invokes a process with five or more stages, including pre-processing, compiling, optimizing, assembling, and linking. However, a complex software system will have hundreds or even thousands of source files, as well as dozens or hundreds of build configuration options, auto configuration scripts (cmake, autotools), build scripts (such as Makefiles) to coordinate the process, test suites, and more.
  
 
The build process varies significantly between software packages. Most software distribution projects (including Linux distributions such as Ubuntu and Fedora) use a packaging system that further wraps the build process in a standardized script format, so that different software packages can be built using a consistent process.
 
The build process varies significantly between software packages. Most software distribution projects (including Linux distributions such as Ubuntu and Fedora) use a packaging system that further wraps the build process in a standardized script format, so that different software packages can be built using a consistent process.
Line 138: Line 138:
 
*** Logically: false or true.
 
*** Logically: false or true.
 
** Binary numbers are resistant to errors, especially when compared to other systems such as analog voltages.
 
** Binary numbers are resistant to errors, especially when compared to other systems such as analog voltages.
*** To represent the numbers 0-10 as an analog electical value, we could use a voltage from 0 - 10 volts. However, if we use a long cable, there will be signal loss and the voltage will drop: we could apply 10 volts on one end of the cable, but only observe (say) 9.1 volts on the other end of the cable. Alternately, electromagnetic interference from nearby devices could slightly increase the signal.
+
*** To represent the numbers 0-10 as an analog electrical value, we could use a voltage from 0 - 10 volts. However, if we use a long cable, there will be signal loss and the voltage will drop: we could apply 10 volts on one end of the cable, but only observe (say) 9.1 volts on the other end of the cable. Alternately, electromagnetic interference from nearby devices could slightly increase the signal.
 
*** If we instead use the same voltages and cable length to carry a binary signal, where 0 volts == off == "0" and 10 volts == on == "1", a signal that had degraded from 10 volts to 9.1 volts would still be counted as a "1" and a 0 volt signal with some stray electromagnetic interference presenting as (say) 0.4 volts would still be counted as "0". However, we will need to use multiple bits to carry larger numbers -- either in parallel (multiple wires side-by-side), or sequentially (multiple bits presented over the same wire in sequence).
 
*** If we instead use the same voltages and cable length to carry a binary signal, where 0 volts == off == "0" and 10 volts == on == "1", a signal that had degraded from 10 volts to 9.1 volts would still be counted as a "1" and a 0 volt signal with some stray electromagnetic interference presenting as (say) 0.4 volts would still be counted as "0". However, we will need to use multiple bits to carry larger numbers -- either in parallel (multiple wires side-by-side), or sequentially (multiple bits presented over the same wire in sequence).
 
* Integers
 
* Integers
Line 173: Line 173:
 
*** Instead of fixed-length numbers, variable-length numbers are used, with the most common values encoded in the smallest number of bits. This is an effective strategy if the distribution of values in the data set is uneven.
 
*** Instead of fixed-length numbers, variable-length numbers are used, with the most common values encoded in the smallest number of bits. This is an effective strategy if the distribution of values in the data set is uneven.
 
** Repeated sequence encoding (1D, 2D, 3D)
 
** Repeated sequence encoding (1D, 2D, 3D)
*** Run length encoding is an encoding scheme that records the number of repeated values. For example, fax messages are encoded as a series of numbers representing the number of white pixels, then the number of black pixels, then white pixels, then black pixels, alternating to the end of each line. These numbers are then represented with adaptive artithmetic encoding.
+
*** Run length encoding is an encoding scheme that records the number of repeated values. For example, fax messages are encoded as a series of numbers representing the number of white pixels, then the number of black pixels, then white pixels, then black pixels, alternating to the end of each line. These numbers are then represented with adaptive arithmetic encoding.
 
*** Text data can be compressed by building a dictionary of common sequences, which may represent words or complete phrases, where each entry in the dictionary is numbered. The compressed data contains the dictionary plus a sequence of numbers which represent the occurrence of the sequences in the original text. On standard text, this typically enables 10:1 compression.
 
*** Text data can be compressed by building a dictionary of common sequences, which may represent words or complete phrases, where each entry in the dictionary is numbered. The compressed data contains the dictionary plus a sequence of numbers which represent the occurrence of the sequences in the original text. On standard text, this typically enables 10:1 compression.
 
** Decomposition
 
** Decomposition
*** Compound audio wavforms can be decomposed into individual signals, which can then be modelled as repeated sequences. For example, a waveform consisting of two notes being played at different frequencies can be decomposed into those separate notes; since each note consists of a number of repetitions of a particular wave pattern, they can individually be represented in a more compact format by describing the frequency, waveform shape, and amplitude characteristics.
+
*** Compound audio waveforms can be decomposed into individual signals, which can then be modelled as repeated sequences. For example, a waveform consisting of two notes being played at different frequencies can be decomposed into those separate notes; since each note consists of a number of repetitions of a particular wave pattern, they can individually be represented in a more compact format by describing the frequency, waveform shape, and amplitude characteristics.
 
** Palletization
 
** Palletization
 
*** Images often contain repeated colours, and rarely use all of the available colours in the original encoding scheme. For example, a 1920x1080 "full HD" image contains about 2 million pixels, so if every pixel was a different colour, there would be a maximum of 2 million colours. But it's likely that many of the pixels in the image are the same colour, so there might only be (perhaps) 4000 colours in the image. If each pixel is encoded as a 24-bit value, there are potentially 16 million colours available, and there is no possibility that they are all used. Instead, a palette can be provided which specifies each of the 4000 colours used in the picture, and then each pixel can be encoded as a 12-bit number which selects one of the colours from the palette. The total storage requirement for the original 24-bit scheme is 1920*1080*3 bytes per pixel = 5.9 MB. Using a 12-bit pallette, the storage requirement is 3 * 4096 bytes for the palette plus 1920*1080*1.5 bytes for the image, for a total of 3 MB -- a reduction of almost 50%
 
*** Images often contain repeated colours, and rarely use all of the available colours in the original encoding scheme. For example, a 1920x1080 "full HD" image contains about 2 million pixels, so if every pixel was a different colour, there would be a maximum of 2 million colours. But it's likely that many of the pixels in the image are the same colour, so there might only be (perhaps) 4000 colours in the image. If each pixel is encoded as a 24-bit value, there are potentially 16 million colours available, and there is no possibility that they are all used. Instead, a palette can be provided which specifies each of the 4000 colours used in the picture, and then each pixel can be encoded as a 12-bit number which selects one of the colours from the palette. The total storage requirement for the original 24-bit scheme is 1920*1080*3 bytes per pixel = 5.9 MB. Using a 12-bit pallette, the storage requirement is 3 * 4096 bytes for the palette plus 1920*1080*1.5 bytes for the image, for a total of 3 MB -- a reduction of almost 50%
Line 186: Line 186:
 
# Follow the [[SPO600 Communication Tools]] set-up instructions.
 
# Follow the [[SPO600 Communication Tools]] set-up instructions.
 
# Optional (strongly recommended): [[SPO600 Host Setup|Set up a personal Linux system]].
 
# Optional (strongly recommended): [[SPO600 Host Setup|Set up a personal Linux system]].
# Optional: If you have an AArch64 development board (such as a Raspberry Pi 4, Raspberry Pi 400, or [http://96boards.org 96Boards] device, consider installing a 64-bit Linux operating system such as Fedora on it.
+
# Optional: If you have an AArch64 development board (such as a Raspberry Pi 4, Raspberry Pi 400, or [http://96boards.org 96Boards] device), consider installing a 64-bit Linux operating system such as Fedora on it.
 
# Start work on [[SPO600 Code Review Lab|Lab 1]]. Blog your work.
 
# Start work on [[SPO600 Code Review Lab|Lab 1]]. Blog your work.
 
  
 
== Week 2 ==
 
== Week 2 ==
Line 202: Line 201:
 
{{Admon/tip|Follow the Links!|To get the full benefit of the following material, please follow the links embedded within it. For additional detail, see the Category links at the bottom of those pages -- for example, the [[Category:Computer Architecture|Computer Architecture]] category linked from many of the following pages has over 30 pages of content.}}
 
{{Admon/tip|Follow the Links!|To get the full benefit of the following material, please follow the links embedded within it. For additional detail, see the Category links at the bottom of those pages -- for example, the [[Category:Computer Architecture|Computer Architecture]] category linked from many of the following pages has over 30 pages of content.}}
  
* Although we program computers in a variety of languages, they can really only execute one langauge: [[Machine Language]], which is encoded in an architecture-specific binary code, sometimes called object code.
+
* Although we program computers in a variety of languages, they can really only execute one language: [[Machine Language]], which is encoded in an architecture-specific binary code, sometimes called object code.
 
* Machine language is not easy to read. [[Assembly Language]] corresponds very closely to machine language, but is (sort of!) human-readable.
 
* Machine language is not easy to read. [[Assembly Language]] corresponds very closely to machine language, but is (sort of!) human-readable.
 
* Assembly language is converted into machine code by a particular type of compiler called an [[Assembler]] (sometimes the language itself is also referred to as "Assembler").
 
* Assembly language is converted into machine code by a particular type of compiler called an [[Assembler]] (sometimes the language itself is also referred to as "Assembler").
  
 
==== 6502 ====
 
==== 6502 ====
Modern processors are complex - the reference manual for 64-bit ARM processors is over 11000 pages long! - so we're going to look at assembly lanaguage on a much simpler processor to get started. This processor is the 6502, a processor used in many early home and personal computers as well as video game systems, including the Commodore PET, VIC-20, C64; the Apple II; the Atari 400 and 800 computers and 2600 video game systems; and many others.
+
Modern processors are complex - the reference manual for 64-bit ARM processors is over 11000 pages long! - so we're going to look at assembly language on a much simpler processor to get started. This processor is the 6502, a processor used in many early home and personal computers as well as video game systems, including the Commodore PET, VIC-20, C64; the Apple II; the Atari 400 and 800 computers and 2600 video game systems; and many others.
  
 
* Introduction to the [[6502]] (note the Resources links on that page)
 
* Introduction to the [[6502]] (note the Resources links on that page)
Line 213: Line 212:
 
* Introduction to the [[6502 Addressing Modes]]
 
* Introduction to the [[6502 Addressing Modes]]
 
* Information about the [[6502 Emulator]] which we will use in this course, and some [[6502_Emulator_Example_Code|example code]]
 
* Information about the [[6502 Emulator]] which we will use in this course, and some [[6502_Emulator_Example_Code|example code]]
* Link to the actual [http://6502.cdot.system 6502 emulator]
+
* Link to the actual [http://6502.cdot.systems 6502 emulator]
  
 
==== Lab 2 ====
 
==== Lab 2 ====
Line 235: Line 234:
 
# Study the [[6502 Instructions - Introduction|6502 Instructions]] and [[6502 Addressing Modes]] and make sure you understand what each one does.
 
# Study the [[6502 Instructions - Introduction|6502 Instructions]] and [[6502 Addressing Modes]] and make sure you understand what each one does.
 
# Complete [[6502 Assembly Language Lab|Lab 2]] and blog your results.
 
# Complete [[6502 Assembly Language Lab|Lab 2]] and blog your results.
 
<!-- Memory System Design - Paging ; Memory - Cache/Numa ; Memory - Observability, Barriers -->
 
<!--
 
 
#############################################
 
#############################################
 
#############################################
 
#############################################
 
            Winter 2022
 
#############################################
 
#############################################
 
#############################################
 
 
== Week 1 ==
 
 
=== Week 1 - Class I ===
 
 
==== Video ====
 
 
* [https://web.microsoftstream.com/video/3dc687df-94ab-4f32-af43-b1ad9abe2ca0 Recording of Week 1 Class I] (January 11) - with light editing to correct some of the sound and video issues.
 
 
==== Introduction to the Problems ====
 
 
===== Porting and Portability =====
 
* Most software is written in a '''high-level language''' which can be compiled into [[Machine Language|machine code]] for a specific computer architecture. In many cases, this code can be compiled for multiple architectures. However, there is a lot of existing code that contains some architecture-specific code fragments which contains assumptions based on the architecture written in architecture-specific high-level code or in [[Assembly Language]].
 
* Reasons that code is architecture-specific:
 
** System assumptions that don't hold true on other platforms
 
*** Variable or [[Word|word]] size
 
*** [[Endian|Endianness]]
 
** Code that takes advantage of platform-specific features
 
* Reasons for writing code in Assembl
 
== Week 2 ==
 
 
=== Week 2 - Class I ===
 
 
==== Video ====
 
* [https://web.microsoftstream.com/video/cf653b70-1000-4dbe-a45f-184625eb45c1 Summary video recording from class]
 
* '''Reminder:''' The Tuesday classes are live. An edited recording is provided for reference only - it is no substitute for attending class, taking notes, and asking questions!
 
 
==== Machine Language, Assembly Language ====
 
{{Admon/tip|Follow the Links!|To get the full benefit of the following material, please follow the links embedded within it. For additional detail, see the Category links at the bottom of those pages -- for example, the [[Category:Computer Architecture|Computer Architecture]] category linked from many of the following pages has over 30 pages of content.}}
 
 
* Although we program computers in a variety of languages, they can really only execute one langauge: [[Machine Language]], which is encoded in an architecture-specific binary code, sometimes called object code.
 
* Machine language is not easy to read. [[Assembly Language]] corresponds very closely to machine language, but is (sort of!) human-readable.
 
* Assembly language is converted into machine code by a particular type of compiler called an [[Assembler]] (sometimes the language itself is also referred to as "Assembler").
 
 
==== 6502 ====
 
Modern processors are complex - the reference manual for 64-bit ARM processors is over 7000 pages long! - so we 're going to look at assembly lanaguage on a much simpler processor to get started. This processor is the 6502.
 
 
* Introduction to the [[6502]] (note the Resources links on that page)
 
* Introduction to the [[6502 Instructions - Introduction|6502 Instructions]]
 
* Information about the [[6502 Emulator]] which we will use in this course, and some [[6502_Emulator_Example_Code|example code]]
 
* Link to the actual [http://6502.cdot.system 6502 emulator]
 
 
=== Week 2 - Class II ===
 
 
==== Videos ====
 
* [https://web.microsoftstream.com/video/1e70e6a8-0fff-40de-b0e3-1297442572b8 Week 2 Announcements]
 
* [https://web.microsoftstream.com/video/01d6cae5-e490-40fd-85a1-2000025fdb68 6502 Emulator and Debugger]
 
* [https://web.microsoftstream.com/video/ed7aedf1-fe6f-4b72-bbf1-c9b4e6e80af9 Calculating 6502 Program Execution Time]
 
 
==== Lab 2 ====
 
* [[6502 Assembly Language Lab]] - Lab 2
 
 
=== Week 2 Deliverables ===
 
# If not already completed:
 
## Set up your [[SPO600 Communication Tools]]
 
## Complete [[SPO600 Code Review Lab|Lab 1]] and blog your work.
 
# Study the [[6502 Instructions - Introduction|6502 Instructions]] and make sure you understand what each one does
 
# Complete [[6502 Assembly Language Lab|Lab 2]] and blog your results
 
y Langauge include:
 
** Performance
 
** [[Atomic Operation|Atomic Operations]]
 
** Direct access to hardware features, e.g., CPUID registers
 
* Most of the historical reasons for including assembler are no longer valid. Modern compilers can out-perform most hand-optimized assembly code, atomic operations can be handled by libraries or [[Compiler Intrinsics|compiler intrinsics]], and most hardware access should be performed through the operating system or appropriate libraries.
 
* A new architecture has appeared: AArch64, which is part of [http://www.arm.com/products/processors/instruction-set-architectures/armv8-architecture.php ARMv8]. This is the first new [[Computer Architecture|computer architecture]] to appear in several years (at least, the first mainstream computer architecture).
 
* At this point, most key open source software (the software typically present in a Linux distribution such as Ubuntu or Fedora, for example) now runs on AArch64. However, it may not run as well as on older architectures (such as x86_64).
 
 
===== Benchmarking and Profiling =====
 
 
Benchmarking involves testing software performance under controlled conditions so that the performance can be compared to other software, the same software operating on other types of computers, or so that the impact of a change to the software can be gauged.
 
 
Profiling is the process of analyzing software performance on finer scale, determining resource usage per program part (typically per function/method). This can identify software bottlenecks and potential targets for optimization.
 
 
===== Optimization =====
 
Optimization is the process of evaluating different ways that software can be written or built and selecting the option that has the best performance tradeoffs.
 
 
Optimization may involve substituting software algorithms, altering the sequence of operations, using architecture-specific code, or altering the build process. It is important to ensure that the optimized software produces correct results and does not cause an unacceptable performance regression for other use-cases, system configurations, operating systems, or architectures.
 
 
The definition of "performance" varies according to the target system and the operating goals. For example, in some contexts, low memory or storage usage is important; in other cases, fast operation; and in other cases, low CPU utilization or long battery life may be the most important factor. It is often possible to trade off performance in one area for another; using a lookup table, for example, can reduce CPU utilization and improve battery life in some algorithms, in return for increased memory consumption.
 
 
Most advanced compilers perform some level of optimization, and the options selected for compilation can have a significant effect on the trade-offs made by the compiler, affecting memory usage, execution speed, executable size, power consumption, and debuggability.
 
 
===== Build Process =====
 
 
Building software is a complex task that many developers gloss over. The simple act of compiling a program invokes a process with five or more stages, including pre-proccessing, compiling, optimizing, assembling, and linking. However, a complex software system will have hundreds or even thousands of source files, as well as dozens or hundreds of build configuration options, auto configuration scripts (cmake, autotools), build scripts (such as Makefiles) to coordinate the process, test suites, and more.
 
 
The build process varies significantly between software packages. Most software distribution projects (including Linux distributions such as Ubuntu and Fedora) use a packaging system that further wraps the build process in a standardized script format, so that different software packages can be built using a consistent process.
 
 
In order to get consistent and comparable benchmark results, you need to ensure that the software is being built in a consistent way. Altering the build process is one way of optimizing software.
 
 
Note that the build time for a complex package can range up to hours or even days!
 
 
==== General Course Information ====
 
* Course resources are linked from the CDOT wiki, starting at https://wiki.cdot.senecacollege.ca/wiki/SPO600 (Quick find: This page will usually be Google's top result for a search on "SPO600").
 
* Coursework is submitted by blogging.
 
* Quizzes will be short (1 page) and will be held without announcement at the start of any synchronous class. There is no opportunity to re-take a missed quiz, but your lowest three quiz scores will not be counted, so do not worry if you miss one or two.
 
** Students with test accommodations: an alternate monthly quiz can be made available via the Test Centre. See your professor for details.
 
* Course marks (see Weekly Schedule for dates):
 
** 60% - Project Deliverables
 
** 20% - Communication (Blog and Wiki writing)
 
** 20% - Labs and Quizzes (10% labs - completed/not completed; 10% for quizzes - lowest 3 scores not counted)
 
 
==== Classes ====
 
* Tuesday: synchronous (live) classes on Big Blue Button at 11:40 am - login to learn.senecacollege.ca ("Blackboard"), go to SPO600, and select the "Tuesday Classes" option on the left-hand menu.
 
* Friday: these classes will usually be asynchronous (pre-recorded) - see this page for details each week.
 
 
==== Course Setup ====
 
Follow the instructions on the [[SPO600 Communication Tools]] page to set up a blog, create SSH keys, and send your blog URLs and public key to me.
 
 
Once this information has been submitted, I will:
 
# Update the [[Current SPO600 Participants]] page with your information, and
 
# Create an account for you on the [[SPO600 Servers]].
 
 
This updating is done in batches once or twice a week -- allow some time!
 
 
==== How open source communities work ====
 
* Do the [[SPO600 Code Review Lab|Code Review Lab (Lab 1)]] as homework.
 
 
=== Week 1 - Class II ===
 
 
==== Video ====
 
* [https://web.microsoftstream.com/video/0bc04f34-5438-4a15-b089-ef35b0a65cd3 SPO600 - Binary Representation of Data]
 
 
==== Binary Representation of Data ====
 
* Binary
 
** Binary is a system which uses "bits" (''binary digits'') to represent values.
 
** Each bit has one of two values, signified by the symbols 0 and 1. These correspond to:
 
*** Electrically: typically off/on, or low/high voltage, or low/high current. Many other electrical representations are possible.
 
*** Logically: false or true.
 
** Binary numbers are resistant to errors, especially when compared to other systems such as analog voltages.
 
*** To represent the numbers 0-5 as an analog electical value, we could use a voltage from 0 - 5 volts. However, if we use a long cable, there will be signal loss and the voltage will drop: we could apply 5 volts on one end of the cable, but only observe (say) 4.1 volts on the other end of the cable. Alternately, electromagnetic interference from nearby devices could slightly increase the signal.
 
*** If we use instead use the same voltages and cable length to carry a binary signal, where 0 volts == off == "0" and 5 volts == on == "1", a signal that had degraded from 5 volts to 4.1 volts would still be counted as a "1" and a 0 volt signal with some stray electromagnetic interference presenting as (say) 0.4 volts would still be counted as "0". However, we will need to use multiple bits to carry larger numbers -- either in parallel (multiple wires side-by-side), or sequentially (multiple bits presented over the same wire in sequence).
 
* Integers
 
** Integers are the basic building block of binary numbering schemes.
 
** In an unsigned integer, the bits are numbered from right to left starting at 0, and the value of each bit is <code>2<sup>bit</sup></code>. The value represented is the sum of each bit multiplied by its corresponding bit value. The range of an unsigned integer is <code>0:2<sup>bits</sup>-1</code> where bits is the number of bits in the unsigned integer.
 
** Signed integers are generally stored in twos-complement format, where the highest bit is used as a sign bit. If that bit is set, the value represented is <code>-(!value)-1</code> where ! is the NOT operation (each bit gets flipped from 0&rarr;1 and 1&rarr;0)
 
* Fixed-point
 
** A fixed-point value is encoded the same as an integer, except that some of the bits are fractional -- they're considered to be to the right of the "binary point" (binary version of "decimal point" - or more generically, the ''radix point''). For example, binary 000001.00 is decimal 1.0, and 000001.11 is decimal 1.75.
 
** An alternative to fixed-point values is integer values in a smaller unit of measurement. For example, some accounting software may use integer values representing cents. For input and display purposes, dollar and cent values are converted to/from cent values.
 
* Floating-point
 
** The most commonly-used floating point formats are defined in the [[IEEE 754]] standard.
 
** IEEE754 floating point numbers have three parts: a ''sign bit'' (0 for positive, 1 for negative), a ''mantissa'' or ''significand'', and an ''exponent''. The significand has an implied 1 and radix point preceeding the stored value. The exponent is stored as an unsigned integer to which a ''bias'' value has been added; the bias value is 2<sup>(number of exponent bits - 1)</sup> - 1. The floating point value is interpreted in normal cases as <code>''sign'' mantissa * 2<sup>(exponent - bias)</sup></code>. Exponent values which are all-zeros or all-ones encode four categories of special cases: zero, infinity, Not a Number (NaN), and subnormal numbers (numbers which are close to zero, where the significand does not have an implied 1 to the left of the radix point); in these special cases, the sign bit and significand values may have special meanings.
 
* Characters
 
** Characters are encoded as integers, where each integer corresponds to one "code point" in a character table (e.g., code 65 in ASCII corresponds to the character "A").
 
** Historically, many different coding schemes have been used, but the two most common ones were the American Standard Code for Information Interchange (ASCII), and Extended Binary Coded Decimal Interchange Code (EBCDIC - primarily used on IBM midrange and mainframe systems).
 
** ASCII characters occupied seven bits (code points 0-127), and contains only characters used in North American English. ASCII characters are usually encoded in bytes, so many vendors of ASCII-based systems used the remaining codes 128-255 for special characters such as
 
graphics, line symbols (horizontal, vertical, connector, and corner line symbols for drawing tables), and accented characters; these were called "extended ASCII".
 
** Several ISO standards exist in an attempt to standardize the "extended ascii" characters, such as ISO8859, which was intended to enable the encoding of European languages by adding currency symbols and accented characters. However, the original version of ISO8859-1 does not include all accented characters and was created before the Euro symbol was standardized, so there are multiple versions of ISO8859, ranging from ISO8859-1 through ISO8859-15.
 
** The Unicode and ISO10646 initiatives were initiated to create a single character code set that would encode all symbols used in human writing, both for current and obsolete languages. These initiatives were merged, and the Unicode and ISO10646 standards define a common character set with 2<sup>32</sup> potential code points. However, Unicode also describes transformation formats for data interchange, rendering and composition/decomposition recommendations, and font symbol recommendations.
 
** The first 127 code points in Unicode correspond to ASCII code points, and the first 255 code points correspond to ISO8869-1 code points. The first 65536 code points form the Basic Multilingual Pane (BMP), which contains most of the characters required to write in all contemporary languages. Therefore, for many applications, it is inefficient to store Unicode as full 32-bit values. To solve this issue, several Unicode Transformation Formats (also known -- technically incorrectly -- as Unicode Transfer Formats) have been defined, including UTF-8, UTF-16, and UTF-32 (32-bit). UTF-8 represents ASCII and some ISO-8859 characters as a single byte, the remainder of the BMP as 2-3 bytes per character, and the remaining characters using 3-4 bytes per character. UTF-16 is similar, encoding much of the BMP in a single 16-bit value, and most other characters as two 16-bit values.
 
* Sound
 
** Sound waves are air pressure vibrations.
 
** Digital sound is most often represented in raw form as a series of time-based measurements of air pressure, called Pulse Coded Modulation (PCM).
 
** PCM takes a lot of storage, so sound is often compressed in either a lossless (perfectly recoverable) or lossy format (higher compression, but the decompressed data doesn't perfectly match the original data). To permit high compression ratios with minimal impact on quality, psychoacoustic compression is used - sound variations that most people can't perceive are removed.
 
* Graphics
 
** The human eye perceives luminance (brightness) as well as hue (colour). Our main hue receptors ("cones") are generally sensitive to three wavelengths: red, green, and blue (RGB). We can stimulate the eye to perceive most colours by presenting a combination of light at these three wavelengths utilizing [https://en.wikipedia.org/wiki/Metamerism_(color) metamerism].
 
** Digital displays emit RGB colours, which are mixed together and perceived by the viewer. This is called ''additive'' colour.
 
** For printing, cyan (C)/yellow (Y)/magenta (M) pigmented inks are used, plus black (K) to reduce the amount of colour ink required to represent dark tones; this is known as CYMK colour. These pigments absorb light at specific frequencies, subtracting energy from white or near-white sunlight or artificial light. This is called ''subtractive'' colour.
 
** Images are broken into picture elements (''pixels'') and each pixel is usually represented by a group of values for RGB or CYMK channels, where each channel is represented by an integer or floating-point value. For example, using an 8-bit-per-pixel integer scheme (also known as 24-bit colour), the brightest blue could be represented as R=0,G=0,B=255; the brightest yellow would be R=255,G=255,B=0; black would be R=0,G=0,B=0; and white would be R=255,G=255,B=255. With this scheme, the number of unique colours available is 256^3 ~= 16 million.
 
** As with sound, the raw storage of sampled data requires a lot of storage space, so various lossy and lossless compression schemes are used. Highest compression is achieved with psychovisual compression (e.g., JPEG).
 
** Moving pictures (video, animations) are stored as sequential images, often compressed by encoding only the differences between frames to save storage space. Motion compensation can further compress the data stream by describing how portions of the previous frame should be moved and positioned in the current frame.
 
* Compression techniques
 
** Huffman encoding / Adaptive arithmetic encoding
 
*** Instead of fixed-length numbers,
 
== Week 2 ==
 
 
=== Week 2 - Class I ===
 
 
==== Video ====
 
* [https://web.microsoftstream.com/video/cf653b70-1000-4dbe-a45f-184625eb45c1 Summary video recording from class]
 
* '''Reminder:''' The Tuesday classes are live. An edited recording is provided for reference only - it is no substitute for attending class, taking notes, and asking questions!
 
 
==== Machine Language, Assembly Language ====
 
{{Admon/tip|Follow the Links!|To get the full benefit of the following material, please follow the links embedded within it. For additional detail, see the Category links at the bottom of those pages -- for example, the [[Category:Computer Architecture|Computer Architecture]] category linked from many of the following pages has over 30 pages of content.}}
 
 
* Although we program computers in a variety of languages, they can really only execute one langauge: [[Machine Language]], which is encoded in an architecture-specific binary code, sometimes called object code.
 
* Machine language is not easy to read. [[Assembly Language]] corresponds very closely to machine language, but is (sort of!) human-readable.
 
* Assembly language is converted into machine code by a particular type of compiler called an [[Assembler]] (sometimes the language itself is also referred to as "Assembler").
 
 
==== 6502 ====
 
Modern processors are complex - the reference manual for 64-bit ARM processors is over 7000 pages long! - so we 're going to look at assembly lanaguage on a much simpler processor to get started. This processor is the 6502.
 
 
* Introduction to the [[6502]] (note the Resources links on that page)
 
* Introduction to the [[6502 Instructions - Introduction|6502 Instructions]]
 
* Information about the [[6502 Emulator]] which we will use in this course, and some [[6502_Emulator_Example_Code|example code]]
 
* Link to the actual [http://6502.cdot.system 6502 emulator]
 
 
=== Week 2 - Class II ===
 
 
==== Videos ====
 
* [https://web.microsoftstream.com/video/1e70e6a8-0fff-40de-b0e3-1297442572b8 Week 2 Announcements]
 
* [https://web.microsoftstream.com/video/01d6cae5-e490-40fd-85a1-2000025fdb68 6502 Emulator and Debugger]
 
* [https://web.microsoftstream.com/video/ed7aedf1-fe6f-4b72-bbf1-c9b4e6e80af9 Calculating 6502 Program Execution Time]
 
 
==== Lab 2 ====
 
* [[6502 Assembly Language Lab]] - Lab 2
 
 
=== Week 2 Deliverables ===
 
# If not already completed:
 
## Set up your [[SPO600 Communication Tools]]
 
## Complete [[SPO600 Code Review Lab|Lab 1]] and blog your work.
 
# Study the [[6502 Instructions - Introduction|6502 Instructions]] and make sure you understand what each one does
 
# Complete [[6502 Assembly Language Lab|Lab 2]] and blog your results
 
variable-length numbers are used, with the most common values encoded in the smallest number of bits. This is an effective strategy if the distribution of values in the data set is uneven.
 
** Repeated sequence encoding (1D, 2D, 3D)
 
*** Run length encoding is an encoding scheme that records the number of repeated values. For example, fax messages are encoded as a series of numbers representing the number of white pixels, then the number of black pixels, then white pixels, then black pixels, alternating to the end of each line. These numbers are then represented with adaptive artithmetic encoding.
 
*** Text data can be compressed by building a dictionary of common sequences, which may represent words or complete phrases, where each entry in the dictionary is numbered. The compressed data contains the dictionary plus a sequence of numbers which represent the occurrence of the sequences in the original text. On standard text, this typically enables 10:1 compression.
 
** Decomposition
 
*** Compound audio wavforms can be decomposed into individual signals, which can then be modelled as repeated sequences. For example, a waveform consisting of two notes being played at different frequencies can be decomposed into those separate notes; since each note consists of a number of repetitions of a particular wave pattern, they can individually be represented in a more compact format by describing the frequency, waveform shape, and amplitude characteristics.
 
** Palletization
 
*** Images often contain repeated colours, and rarely use all of the available colours in the original encoding scheme. For example, a 1920x1080 "full HD" image contains about 2 million pixels, so if every pixel was a different colour, there would be a maximum of 2 million colours. But it's likely that many of the pixels in the image are the same colour, so there might only be (perhaps) 4000 colours in the image. If each pixel is encoded as a 24-bit value, there are potentially 16 million colours available, and there is no possibility that they are all used. Instead, a palette can be provided which specifies each of the 4000 colours used in the picture, and then each pixel can be encoded as a 12-bit number which selects one of the colours from the palette. The total storage requirement for the original 24-bit scheme is 1920*1080*3 bytes per pixel = 5.9 MB. Using a 12-bit pallette, the storage requirement is 3 * 4096 bytes for the palette plus 1920*1080*1.5 bytes for the image, for a total of 3 MB -- a reduction of almost 50%
 
** Psychoacoustic and psychovisual compression
 
*** Much of the data in sound and images cannot be perceived by humans. Psychoacoustic and psychovisual compression remove artifacts which are least likely to be perceived. As a simple example, if two pixels on opposite sides of a large image are almost but not exactly the same, most people won't be able to tell the difference, so these can be encoded as the same colour if that saves space (for example, by reducing the size of the colour palette).
 
 
=== Week 1 Deliverables ===
 
# Follow the [[SPO600 Communication Tools]] set-up instructions.
 
# Optional (strongly recommended): [[SPO600 Host Setup|Set up a personal Linux system]].
 
# Optional: Purchase an AArch64 development board (such as a Raspberry Pi 4, Raspberry Pi 400, or [http://96boards.org 96Boards] device. (Note: install a 64-bit Linux operating system on it, not a 32-bit version).
 
# Start work on [[SPO600 Code Review Lab|Lab 1]]. Blog your work.
 
 
== Week 2 ==
 
 
=== Week 2 - Class I ===
 
 
==== Video ====
 
* [https://web.microsoftstream.com/video/cf653b70-1000-4dbe-a45f-184625eb45c1 Summary video recording from class]
 
* '''Reminder:''' The Tuesday classes are live. An edited recording is provided for reference only - it is no substitute for attending class, taking notes, and asking questions!
 
 
==== Machine Language, Assembly Language ====
 
{{Admon/tip|Follow the Links!|To get the full benefit of the following material, please follow the links embedded within it. For additional detail, see the Category links at the bottom of those pages -- for example, the [[Category:Computer Architecture|Computer Architecture]] category linked from many of the following pages has over 30 pages of content.}}
 
 
* Although we program computers in a variety of languages, they can really only execute one langauge: [[Machine Language]], which is encoded in an architecture-specific binary code, sometimes called object code.
 
* Machine language is not easy to read. [[Assembly Language]] corresponds very closely to machine language, but is (sort of!) human-readable.
 
* Assembly language is converted into machine code by a particular type of compiler called an [[Assembler]] (sometimes the language itself is also referred to as "Assembler").
 
 
==== 6502 ====
 
Modern processors are complex - the reference manual for 64-bit ARM processors is over 7000 pages long! - so we 're going to look at assembly lanaguage on a much simpler processor to get started. This processor is the 6502.
 
 
* Introduction to the [[6502]] (note the Resources links on that page)
 
* Introduction to the [[6502 Instructions - Introduction|6502 Instructions]]
 
* Information about the [[6502 Emulator]] which we will use in this course, and some [[6502_Emulator_Example_Code|example code]]
 
* Link to the actual [http://6502.cdot.system 6502 emulator]
 
 
=== Week 2 - Class II ===
 
 
==== Videos ====
 
* [https://web.microsoftstream.com/video/1e70e6a8-0fff-40de-b0e3-1297442572b8 Week 2 Announcements]
 
* [https://web.microsoftstream.com/video/01d6cae5-e490-40fd-85a1-2000025fdb68 6502 Emulator and Debugger]
 
* [https://web.microsoftstream.com/video/ed7aedf1-fe6f-4b72-bbf1-c9b4e6e80af9 Calculating 6502 Program Execution Time]
 
 
==== Lab 2 ====
 
* [[6502 Assembly Language Lab]] - Lab 2
 
 
=== Week 2 Deliverables ===
 
# If not already completed:
 
## Set up your [[SPO600 Communication Tools]]
 
## Complete [[SPO600 Code Review Lab|Lab 1]] and blog your work.
 
# Study the [[6502 Instructions - Introduction|6502 Instructions]] and make sure you understand what each one does
 
# Complete [[6502 Assembly Language Lab|Lab 2]] and blog your results
 
  
 
== Week 3 ==
 
== Week 3 ==
Line 511: Line 240:
  
 
==== Video ====
 
==== Video ====
* No video is available due to issues with the audio.
+
* [https://web.microsoftstream.com/video/a4707dd6-b0df-409b-8168-58ec21a06c1b Summary video from class on 6502 Strings]
* The links below contain the same information.
+
 +
==== Lab ====
 +
* [[6502 Math and Strings Lab]] (Lab 3)
  
==== More 6502 Assembly ====
 
* [[6502 Math]]
 
  
 
=== Week 3 - Class II ===
 
=== Week 3 - Class II ===
  
==== Videos ====
+
==== Video ====
* [https://web.microsoftstream.com/video/792a1653-36a0-4f15-b3fc-36461dbbb969 6502 Jumps, Branches, and Procedures]
+
* 6502 Assembly Language
* [https://web.microsoftstream.com/video/a60f8b98-6057-4c83-aecb-b40eb7f5bd8f 6502 Characters & Strings]
+
** [https://web.microsoftstream.com/video/9caa5e8d-0f15-4b8b-9293-0151c82f77b1 6502 String Input]
 +
** [https://web.microsoftstream.com/video/6a645edd-3537-4910-843c-6d32f6678e79 A 6502 Assembly Hack]
 +
** [https://web.microsoftstream.com/video/1775931c-b9eb-4b2a-a7bd-598d7d725853 6502 Assembly Sample Code]
 +
* 6502 - Additional Resources
 +
** [https://web.microsoftstream.com/video/6a645edd-3537-4910-843c-6d32f6678e79 An old video on the basics of using the 6502 Emulator]
 +
** [https://web.microsoftstream.com/video/f22220d6-9c87-4d23-aaf8-95f681756c41 6502 Assembler Directives] - using "define" and "dcb"
 +
* Building code: make
 +
** [https://web.microsoftstream.com/video/6b83e243-2b82-4afe-848e-e8c26881199a make and Makefiles]
  
==== More 6502 Assembly ====
+
==== Resources ====
* [[6502 Jumps, Branches, and Procedures]]
+
* [[Make and Makefiles]]
 +
* 6502 Example Code
 +
** [[6502 Emulator Example Code]] page on this wiki
 +
** Chris Tyler's [https://github.com/ctyler/6502js-code/ 6502js-code] repository on GitHub (includes Wordle-like example)
 +
** [http://6502asm.com/ 6502asm.com] - a site with an early version of the 6502 Emulator - see the "Examples" pull-down menu (these examples will run in [[http://6502.cdot.systems|our emulator]]
  
==== Lab 3 ====
+
=== Week 3 Deliverables ===
* [[6502 Math and Strings Lab]] (Lab 3)
+
* [[6502 Math and Strings Lab|Lab 3]]
 +
* Note that September blog posts are due at the end of next week, so don't get behind in your blogging
  
=== Week 3 Deliverables ===
 
* Perform [[6502 Math and Strings Lab|Lab 3]]
 
* Continue to blog
 
* Make sure you've submitted the form with your blog URL and public key (see [[#Week 1 Deliverables|Week 1 Deliverables]])
 
  
 
== Week 4 ==
 
== Week 4 ==
Line 539: Line 276:
  
 
==== Video ====
 
==== Video ====
* [https://web.microsoftstream.com/video/05311ea3-8536-4b78-ace3-438902d78d67 Edited summary video]
+
* [https://web.microsoftstream.com/video/30fa002e-9e3d-41f6-95db-36832a8a509c Edited Class Summary Video]
* '''Reminder:''' the videos are a summary/recap only - they're no substitute for attending, taking notes, and asking questions in class!
 
  
==== Compiler Optimizations ====
+
==== Reading Resources ====
 
* [[Compiler Optimizations]]
 
* [[Compiler Optimizations]]
 +
* Connecting to course servers
 +
** [[SPO600 Servers]]
 +
** [[SSH]]
 +
** [[Screen Tutorial|Screen utility]] - allows disconnection/reconnection to remote host
  
=== Week 5 - Class II ===
+
=== Week 4 - Class II ===
* Video content will be delayed due to a storage error
+
 
* Please continue work on [[6502 Math and Strings Lab|Lab 3]]
+
==== Video ====
 +
* [https://web.microsoftstream.com/video/48f2d7a8-67d3-4e49-b02e-a29e7d9b656c Building Code: Compiler Options]
 +
* [https://web.microsoftstream.com/video/38b050c6-6aad-4e64-b564-95ceb53adc7c Building Code: Automake/Autotools (configure scripts)]
 +
 
 +
==== Resouces ====
 +
* [https://www.gnu.org/software/automake/manual/html_node/index.html GNU Autotools/Automake]
 +
* [https://gcc.gnu.org/onlinedocs/ GCC Manual]
  
 
=== Week 4 Deliverables ===
 
=== Week 4 Deliverables ===
* '''By the end of the day on Tuesday, February 1:''' Blogs for January are due (including any labs you're submitting for January). Make sure you've submitted the [[SPO600 Communication Tools|web form]] with your blog URL so I can find it!
+
* September blogs are due this weekend (Sunday, October 2 at 11:59 pm)
* Finish [[6502 Math and Strings Lab|Lab 3]]
 
* Continue to blog
 
  
 
== Week 5 ==
 
== Week 5 ==
Line 559: Line 303:
  
 
==== Video ====
 
==== Video ====
* No recording of the live session is available (bad audio)
+
* [https://web.microsoftstream.com/video/fe744d30-f947-433d-b9f3-f5284e6fb2ad Class Summary Video]
* Pre-recorded lecture/demo: [https://web.microsoftstream.com/video/466bc3a1-6729-434d-b51f-33d4fbb145c5 Make and Makefiles]
 
  
==== Notes ====
+
==== Resources ====
* [[Make and Makefiles]]
+
* [[Assembly Language]]
 
+
* [[ELF]] file format
===== Compiler Options =====
+
* [[X86_64 Register and Instruction Quick Start]]
* (Modern compilers are similar in options, for the sake of this discussion I'm focusing on the GNU C Compiler (gcc), part of the GNU Compiler Collection)
+
* [[Aarch64 Register and Instruction Quick Start]]
* There are hundreds of compiler features available, many of which are optimization options.
+
* ARM 64-bit CPU Instruction Set and Software Developer Manuals
* These features can be controlled from the compiler command line:
+
* ARM Aarch64 documentation
** To enable a feature, specify <code>-f</code> and the option name: <code>-f''builtin''</code>
+
** [http://developer.arm.com/ ARM Developer Information Centre]
** To disable a feature, specify <code>-f</code> and then <code>no-</code> and the option name: <code>-f''no-builtin''</code>
+
*** [https://developer.arm.com/docs/den0024/latest ARM Cortex-A Series Programmer’s Guide for ARMv8-A]
* Example:
+
*** The ''short'' guide to the ARMv8 instruction set: [https://www.element14.com/community/servlet/JiveServlet/previewBody/41836-102-1-229511/ARM.Reference_Manual.pdf ARMv8 Instruction Set Overview] ("ARM ISA Overview")
gcc -fbuiltin -falign-functions -no-caller-saves ''foo''.c -o foo
+
*** The ''long'' guide to the ARMv8 instruction set: [https://developer.arm.com/docs/ddi0487/latest/arm-architecture-reference-manual-armv8-for-armv8-a-architecture-profile ARM Architecture Reference Manual ARMv8, for ARMv8-A architecture profile] ("ARM ARM")
* To see the available optimization features and what each does, view the gcc manpage and/or gcc manual
+
** [https://developer.arm.com/docs/ihi0055/latest/procedure-call-standard-for-the-arm-64-bit-architecture Procedure Call Standard for the ARM 64-bit Architecture (AArch64)]
* It's a pain to specify hundreds of <code>-f</code> options on the command line, so these are grouped into commonly-used sets. The sets can be specified with the <code>-O</code> compiler option (note that that is a capital letter "O", not a lowercase "o" nor a zero "0"), followed by an optimization level:
+
* x86_64 Documentation
** -O0 : almost no optimization
+
** [https://developer.amd.com/resources/developer-guides-manuals/ AMD Developer Guide and Manuals](see the AMD64 Architecture section, particularly the ''AMD64 Architecture Programmer’s Manual Volume 3: General Purpose and System Instructions'')
** -O1 : optimizations that can be quickly performed
+
** [http://www.intel.com/content/www/us/en/processors/architectures-software-developer-manuals.html Intel Software Developers Manuals]
** -O2 : all of the normal optimizations that can be safely applied to all programs (this is the usual default optimization level)
+
* GAS Manual - Using as, The GNU Assembler: https://sourceware.org/binutils/docs/as/
** -O3 : all normal optimization, including some that may in rare cases cause changes to the operation of the program (for example, counting +0 and -0 as the same number -- which is fine in the vast majority of cases, but might interfere with the correct operation of some scientific calculations)
 
** -Os : optimize for smallest size (of both the executable and the memory usage while executing)
 
** -Ofast : optimize for highest speed, even at the cost of more memory usage
 
** -Og : optimize for debugging -- avoid optimizations that will excessively convolute the code, making it harder to see the correlation between the source code and the object code
 
* Note that the set of optimizations considered "safe" may vary over time - for example, vector optimization were previously considered unsafe (<code>-O3</code>) in the gcc compiler, but with improvements and testing are not considered safe and are therefore included in the <code>-O2</code> level in newer versions of gcc.
 
* You can specify a group of options with <code>-O</code> and override the use of individual options with <code>-f</code> by placing the <code>-O</code> group first:
 
gcc -O2 -fno-builtin foo.c -o foo
 
* To see the optimizations that will be applied by a given set of command-line options, use <code>-Q --help=optimizers</code> to query the optimization list that the compiler will use:
 
gcc -O1 -Q --help=optimizers | less
 
  
 
=== Week 5 - Class II ===
 
=== Week 5 - Class II ===
  
 
==== Video ====
 
==== Video ====
* [https://web.microsoftstream.com/video/91554778-ac0f-4928-8a8b-f4636b96a427 x86_64 & AArch64 Introduction - Registers]
+
* [https://web.microsoftstream.com/video/1bcab47b-514a-4f23-bdd4-f73662a0673f Paged Memory Systems]
 
* [https://web.microsoftstream.com/video/880fb0f8-1084-457a-92e0-80f04ad62463 Memory Alignment and Performance]
 
* [https://web.microsoftstream.com/video/880fb0f8-1084-457a-92e0-80f04ad62463 Memory Alignment and Performance]
* [https://web.microsoftstream.com/video/1bcab47b-514a-4f23-bdd4-f73662a0673f Paged Memory Systems]
 
  
=== Resources ===
+
==== Lab 4 ====
* [[X86 64 Register and Instruction Quick Start]]
+
* [[SPO600 64-bit Assembly Language Lab]] (Lab 4)
* [[AArch64 Register and Instruction Quick Start]]
 
* [[Computer Architecture]]
 
  
 
=== Week 5 Deliverables ===
 
=== Week 5 Deliverables ===
* Finish [[6502 Math and Strings Lab|Lab 3]]
+
* [[SPO600 64-bit Assembly Language Lab|Lab 4]]
* Continue to blog
+
 
  
 
== Week 6 ==
 
== Week 6 ==
Line 608: Line 339:
 
=== Week 6 - Class I ===
 
=== Week 6 - Class I ===
  
==== Video ====
+
We used this class for introductions, a discussion of how things are going, and feedback on the course.
* [https://web.microsoftstream.com/video/f8ca9820-2222-4aad-a5f7-17fdc117ec3b Week 6 Class I Summary Video]
 
 
 
==== Class Servers ====
 
* Student accounts on the [[SPO600 Servers]] have been set up
 
* Please test that you can login to both of these machines as soon as possible. Contact me if you have any issues logging in.
 
  
 
=== Week 6 - Class II ===
 
=== Week 6 - Class II ===
  
==== Videos ====
+
==== Video ====
* [https://web.microsoftstream.com/video/8c3c1353-5729-4217-b1ba-371410f14ad4 64-Bit Assembly Language - Part II]
+
* [https://web.microsoftstream.com/video/d208a737-7777-4b5a-b276-1b19dc78145c Inline Assembly Language] - Inserting assembly language code into programs written in other languages (in this case, C)
 +
* [https://web.microsoftstream.com/video/f60b92c6-9db3-4f57-b0b9-7c35ea0c054f Single Instruction, Multiple Data (SIMD)]
 +
* [https://web.microsoftstream.com/video/2a82da88-bf5b-4112-953a-7408fbab30c1 Algorithm Selection and Benchmarking]
  
==== Reading ====
+
==== Lab 5 ====
* [[Assembly Language]]
+
* [https://wiki.cdot.senecacollege.ca/wiki/SPO600_Algorithm_Selection_Lab Algorithm Selection Lab] (Lab 5)
* [[Assembler Basics]]
 
  
==== Lab ====
 
* [[SPO600 64-bit Assembly Language Lab]] (Lab 4)
 
  
 
=== Week 6 Deliverables ===
 
=== Week 6 Deliverables ===
* [[SPO600 64-bit Assembly Language Lab|Lab 4]]
+
* [https://wiki.cdot.senecacollege.ca/wiki/SPO600_Algorithm_Selection_Lab Lab 5]
* Continue to Blog
+
 
  
 
== Week 7 ==
 
== Week 7 ==
Line 636: Line 361:
  
 
==== Video ====
 
==== Video ====
* A summary video will be posted after editing
+
* Video summary will be posted after editing
 
 
  
 
=== Week 7 - Class II ===
 
=== Week 7 - Class II ===
  
==== Videos ====
+
'''Please catch up on course material to this point. If you are fully caught up, you can start to take a look at SVE2:'''
* [https://web.microsoftstream.com/video/f60b92c6-9db3-4f57-b0b9-7c35ea0c054f Single Instruction, Multiple Data (SIMD)]
 
* [https://web.microsoftstream.com/video/d208a737-7777-4b5a-b276-1b19dc78145c Inline Assembly Language]
 
* [https://web.microsoftstream.com/video/2a82da88-bf5b-4112-953a-7408fbab30c1 Algorithm Selection]
 
* [https://web.microsoftstream.com/video/d56ec6ff-2c2c-40d6-8967-52d829e413cc Linux Tips] (This is an older video -- the systems mentioned, such as xerxes, were previous versions of the [[SPO600 Servers|class servers]].)
 
  
 
==== Reading ====
 
==== Reading ====
* [[Inline Assembly Language]]
+
* [[SVE2]]
  
==== Lab ====
+
==== SVE2 Demonstration ====
* [[SPO600 Algorithm Selection Lab]] (Lab 5)
+
* Code available here: https://github.com/ctyler/sve2-test
 +
* This is an implementation of a very simple program which takes an image file, adjusts the red/green/blue channels of that file, and then writes an output file. Each channel is adjusted by a factor in the range 0.0 to 2.0 (with saturation).
 +
* The image adjustment is performed in the function <code>adjust_channels()</code> in the file <code>adjust_channels.c</code>. There are three implementations:
 +
*# A basic (naive) implementation in C. Although this is a very basic implementation, it is potentially subject to autovectorization.
 +
*# An implementation using inline assembler for SVE2 with strucure loads.
 +
*# An implementation using inline assembler for SVE2 with an interleaved factor table.
 +
*# An implementation using ACLE compile intrinsics.
 +
* The implementation built is dependent on the value of the ADJUST_CHANNEL_IMPLEMENTATION macro.
 +
* The provided Makefile will build four versions of the binary -- one using each of the four implementations -- and it will run through 3 tests with each binary. The tests use the input image file <code>tests/input/bree.jpg</code> (a picture of a cat) and place the output in the files <code>tests/output/bree[1234][abc].jpg</code>. The output files are processed with adjustment factors of 0.5/0.5/0.5, 1.0/1.0/1.0, and 2.0/2.0/2.0.
 +
* '''Please examine, build, and test the code, compare the implementations, and note how it works - there are extensive comments in the code, especially for implementation 2.'''
 +
* Your observations about the code might make a good blog post!
  
  
 
=== Week 7 Deliverables ===
 
=== Week 7 Deliverables ===
* [[SPO600 Algorithm Selection Lab|Lab 5]]
+
* Complete [[SPO600 64-bit Assembly Language Lab|Lab 4]] and [https://wiki.cdot.senecacollege.ca/wiki/SPO600_Algorithm_Selection_Lab Lab 5]
* '''Note:''' Blog for February are due at 11:59 pm on February 28 (Tuesday). I'll be looking for an average of 1-2 blog posts per week, or 4-8 blog posts for February. Please review your posts for accuracy and completeness.
+
* Remember that October blogs are due soon.
  
 
== Week 8 ==
 
== Week 8 ==
Line 663: Line 393:
  
 
==== Video ====
 
==== Video ====
* [https://web.microsoftstream.com/video/333ff7df-7c82-48c7-8051-39a04592c849 Edited summary video]
+
* [https://web.microsoftstream.com/video/f67c0185-fc67-43fb-ac39-57cae26792a8 SIMD - Edited Summary Video]
 
 
==== Reading ====
 
* [[SVE2]]
 
 
 
==== Lab ====
 
* [[SPO600 SVE2 Lab]] (Lab 6)
 
  
 
=== Week 8 - Class II ===
 
=== Week 8 - Class II ===
 
==== Project ====
 
* [[Winter 2022 SPO600 Project]] - We will discuss this in the next class
 
  
 
==== Video ====
 
==== Video ====
* [https://web.microsoftstream.com/video/ad72567e-047d-4dda-9f24-0c525429c7d1 Searching a Codebase]
+
* [https://web.microsoftstream.com/video/a6b892e4-b408-4bc7-9fc1-d78e4efb8e0e SVE & SVE2 - Edited Summary Video]
 
 
=== Week 8 Deliverables ===
 
* Blog about your [[SPO600 SVE2 Lab|Lab 6]]
 
 
 
== Week 9 ==
 
 
 
=== Week 9 - Class I ===
 
 
 
==== Video ====
 
* Summary video: [https://web.microsoftstream.com/video/b9ccfc5f-f816-45af-828f-98f5fad5c8c8 Project Discussion 1]
 
 
 
=== Week 9 - Class II ===
 
* No content posted.
 
 
 
== Week 10 ==
 
 
 
=== Week 10 - Class I ===
 
 
 
==== Video ====
 
* Summary video: [https://web.microsoftstream.com/video/058bf7e4-ba17-47a0-9db9-9221a24da548 Project Discussion 2]
 
 
 
=== Week 10 - Class II ===
 
 
 
==== Videos ====
 
* [https://web.microsoftstream.com/video/b62a6499-012b-4203-b320-126b978e6ae3 GCC and SVE2] - A discussion of compiler flags, macros, pre-processor directives, and disassembly analysis that may be useful in project stage 2.
 
* [https://web.microsoftstream.com/video/52c0d800-9618-46ac-8a5e-1a5477b5e4f0 Bitwise Operations] - this video covers AND, OR, XOR/EOR, and NOT operations. It should be review, but I've seen these operations misused a few times lately so it may be useful to you, especially if you are not familiar with the use of masks with these operations.
 
 
 
=== Week 10 - Deliverables ===
 
* Blog about your [[Winter 2022 SPO600 Project|project work]].
 
* '''Note:''' Your [[Winter 2022 SPO600 Project|Project Stage 1]] is due on Monday, March 28 by midnight.
 
 
 
== Week 11 ==
 
 
 
=== Week 11 - Class I ===
 
 
 
==== Videos ====
 
* [https://web.microsoftstream.com/video/5f633a25-c478-4aab-af90-c447d581a631 Project Discussion 3] - An edited recording of the March 28, 2022 SPO600 class.
 
* [https://web.microsoftstream.com/video/b197e528-3385-41ef-b648-7041d054c0d2 Benchmarking and Profiling] - An optional video that may be useful to some projects in Stage 2. This video discusses benchmarking (overall program performance analysis) and profiling (per-function/per-method performance analysis) principles and techniques. (This is an edited version of a previous-semester video. There are a couple of small audio and video glitches in the recording).
 
  
 
==== Reading ====
 
==== Reading ====
* [https://locklessinc.com/articles/vectorize/ Auto-vectorization with GCC 4.7] - Although based on an earlier version of GCC (and a number of new features have been added to the GCC autovectorizer since this article was written), it discusses some of the techniques and code adjustments that may be required to get the GCC compiler to vectorize code. Note that the <code>-fopt-info-vec-all</code> or <code>-fopt-info-vec-missed</code> options are useful in conjunction with this technique, as they will cause the compiler to emit information about the vectorization decisions that it is making.
+
* [[SVE2]]
* [https://www.phoronix.com/scan.php?page=news_item&px=GCC-12-Auto-Vec-O2 GCC 12 Enables Auto-Vectorization for -O2 Optimization Level] - a short news article from October 2021 regarding the [https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=2b8453c401b699ed93c085d0413ab4b5030bcdb8 commit] that added autovectorization to the <code>-O2</code> optimization level, which is the default for many projects. GCC12 is expected to ship in April 2022, according to a recent [https://gcc.gnu.org/pipermail/gcc/2022-January/238136.html status update].
 
 
 
=== Week 11 - Class II ===
 
  
 
==== SVE2 Demonstration ====
 
==== SVE2 Demonstration ====
 
* Code available here: https://github.com/ctyler/sve2-test
 
* Code available here: https://github.com/ctyler/sve2-test
 +
** You can clone this to israel.cdot.systems with: <code>git clone https://github.com/ctyler/sve2-test.git</code>
 
* This is an implementation of a very simple program which takes an image file, adjusts the red/green/blue channels of that file, and then writes an output file. Each channel is adjusted by a factor in the range 0.0 to 2.0 (with saturation).
 
* This is an implementation of a very simple program which takes an image file, adjusts the red/green/blue channels of that file, and then writes an output file. Each channel is adjusted by a factor in the range 0.0 to 2.0 (with saturation).
 
* The image adjustment is performed in the function <code>adjust_channels()</code> in the file <code>adjust_channels.c</code>. There are three implementations:
 
* The image adjustment is performed in the function <code>adjust_channels()</code> in the file <code>adjust_channels.c</code>. There are three implementations:
 
*# A basic (naive) implementation in C. Although this is a very basic implementation, it is potentially subject to autovectorization.
 
*# A basic (naive) implementation in C. Although this is a very basic implementation, it is potentially subject to autovectorization.
*# An implementation using inline assembler for SVE2.
+
*# An implementation using inline assembler for SVE2 with strucure loads.
*# (Future) An implementation using ACLE compile intrinsics.
+
*# An implementation using inline assembler for SVE2 with an interleaved factor table.
 +
*# An implementation using ACLE compile intrinsics.
 
* The implementation built is dependent on the value of the ADJUST_CHANNEL_IMPLEMENTATION macro.
 
* The implementation built is dependent on the value of the ADJUST_CHANNEL_IMPLEMENTATION macro.
* The provided Makefile will build two versions of the binary, one using implementation 1 (named <code>image_adjust1</code>) and one using implementation 2 (named <code>image_adjust2</code>), and it will run through 3 tests with each binary. The tests use the input image file <code>tests/input/bree.jpg</code> (a picture of a cat) and place the output in the files <code>tests/output/bree[12][abc].jpg</code>. The output files are processed with adjustment factors of 0.5/0.5/0.5, 1.0/1.0/1.0, and 2.0/2.0/2.0.
+
* The provided Makefile will build four versions of the binary -- one using each of the four implementations -- and it will run through 3 tests with each binary. The tests use the input image file <code>tests/input/bree.jpg</code> (a picture of a cat) and place the output in the files <code>tests/output/bree[1234][abc].jpg</code>. The output files are processed with adjustment factors of 0.5/0.5/0.5, 1.0/1.0/1.0, and 2.0/2.0/2.0.
 
* '''Please examine, build, and test the code, compare the implementations, and note how it works - there are extensive comments in the code, especially for implementation 2.'''
 
* '''Please examine, build, and test the code, compare the implementations, and note how it works - there are extensive comments in the code, especially for implementation 2.'''
 
* Your observations about the code might make a good blog post!
 
* Your observations about the code might make a good blog post!
  
=== Week 11 - Deliverables ===
+
=== Week 8 Deliverables ===
* Blog about your project work.
+
* Continue your blogging
 +
* Include blogging on SVE/SVE
 +
* The second group of blog posts is due on or before this Sunday (November 6, 11:59 pm)
  
 +
== Week 9 ==
  
== Week 12 ==
+
=== Week 9 - Class I ===
 
 
=== Week 12 - Class I ===
 
  
 
==== Video ====
 
==== Video ====
* [https://web.microsoftstream.com/video/00172f3b-f0cb-486f-bf15-42c73e8916b4 SVE2 Examples] - Summary video of the SPO600 class on Tuesday, April 5.
+
* [https://web.microsoftstream.com/video/28a4f8e7-c96c-4662-9adc-8d67aa409868 Edited summary]
  
==== SVE2 Demonstration ====
+
==== iFunc ====
* The SVE2 [https://github.com/ctyler/sve2-test example code] has been extended with an additional inline assembley implementation, plus an ACLE implementation.
 
  
=== Week 12 - Class II ===
+
GNU iFunc is a facility for handling indirect functions. The basic premise is that you prototype the function to be called, add the <code>ifunc</code> attribute to that prototype, and provide the name of a resolver function. The resolver function is called at program initialization, and returns a pointer to the function to be executed when the function referenced in the prototype is called. The resolver typically picks one of several implementations based on the capabilities of the machine on which the code is running; for example, it could return a pointer to a non-SVE, SVE, or SVE2 implementation of a function based on cpu capabilities (on an Aarch64 system) or it could return a pointer to an SSE, SSE2, AVX, or AVX512 implementation (on an x86_64 system).
  
==== Video ====
+
There is a [https://github.com/ctyler/ifunc-aarch64-demo GitHub repository] available with example iFunc code -- please clone this to [[SPO600 Servers#AArch64:_israel.cdot.systems|israel.cdot.systems]] and build and test the code there. You should see different results if you run the output executable directly (<code>./ifunc-test</code>) and run it through the qemu-aarch64 tool, which will emultate SVE2 capabilities (<code>qemu-aarch64 ./ifunc-test</code>). Make sure you understand how the code works.
* [https://web.microsoftstream.com/video/11def15f-20df-41b5-84f0-6fd5bd96bc2a SVE2 Examples - Part 2] - Part 2 of a look at the [https://github.com/ctyler/sve2-test example code] - A discussion of the bug that existed in the ACLE/intrinsics code discussed in the last class, plus an examination of the disassembly of the naive/autovectorized version of the code (implementation #1).
 
  
=== Week 12 - Deliverables ===
+
==== Reading/Resources ====
* Continue to blog about your project work
 
* Project Stage 2 will be due on '''Wednesday, April 13''' (11:59 pm EDT).
 
  
== Week 13 ==
+
* [https://gcc.gnu.org/onlinedocs/gcc-12.2.0/gcc/Common-Function-Attributes.html#index-ifunc-function-attribute GNU iFunc attribute in GCC manual]
 +
* [https://sourceware.org/glibc/wiki/GNU_IFUNC iFunc on the glibc wiki]
  
=== Week 13 - Class I ===
+
=== Week 9 - Class II ===
  
 
==== Video ====
 
==== Video ====
* [https://web.microsoftstream.com/video/274ee2d2-a19c-4d12-8c1b-78141cfb4566 Memory Systems] summary video
+
* [https://web.microsoftstream.com/video/edc09b0a-1a7f-45d1-a27e-7f4901bba03d Edited summary video] - '''Important!''' This video contains a detailed discussion of the requirements for the course project.
 +
** Project discussion starts at beginning of video
 +
** Demo of what the project needs to do (manually performing the same steps) starts at 0:27:47
 +
** Recap/summary of the demo starts around 1:02:05
  
=== Week 13 - Class II ===
+
==== Project ====
* Good Friday - no class
+
* [[Fall 2022 SPO600 Project]]
  
=== Week 13 Deliverables ===
+
=== Week 9 Deliverables ===
* Continue to post about your project.
+
* Investigate the iFunc example code
 +
* Blog about your investigation
 +
* Start blogging about your project
  
== Week 14 ==
+
== Week 10 ==
 +
=== Week 10 - Class I ===
 +
==== Video ====
 +
* [https://web.microsoftstream.com/video/76aae456-492d-4edd-a735-5009caf59bf6 Summary Video] - Tips and techniques for working productively in a remote aarch64 Linux system.
 +
=== Week 10 - Class II ===
 +
* A discussion of of Advanced Memory Systems
 +
* No video available at this time
 +
* Not directly related to the course project
 +
=== Week 10 Deliverables ===
 +
* Blog about your project work.
  
=== Week 14 - Class I ===
+
== Week 11 ==
 +
=== Week 11 - Class I ===
 +
==== Video ====
 +
* [https://web.microsoftstream.com/video/29183138-9105-4ca6-b085-7ac84f19f7a3 Summary Video] - Techniques you can use in your project.
 +
=== Week 11 - Class II ===
 +
==== Video ====
 +
* [https://web.microsoftstream.com/video/4a29385e-e924-4ef1-a5eb-b526e3aa6390 Summary Video] - A demonstration of one solution to the course project.
 +
=== Week 11 Deliverables ===
 +
* Blog about your project work.
  
 +
== Week 12 ==
 +
=== Week 12 - Class I ===
 +
* A discussion of benchmarking
 +
* No video available at this time
 +
* Not directly related to the course project
 +
=== Week 12 - Class II ===
 
==== Video ====
 
==== Video ====
 +
* [https://web.microsoftstream.com/video/fcc73002-b59c-4bc7-b0e6-c73706b492e8 Summary Video] - Step-by-Step discussion of the minimum requirements for a successful project
 +
==== Minimum Requirements for a Successful Project ====
 +
* Remember: This project is about building a ''tool'' that processes C source code. The tool itself can be in any suitable computer language. It is recommended that you use a language that can manipulate text easily. The tool will receive an input file containing a C function. The tool will output a modified file(s) with ifunc capability for asimd/sve/sve2 computers.
 +
* Requirements for a minimum implementation:
 +
# Find the protype and function name by analyzing the C function.
 +
#* Example name: <code>adjust_channels</code>
 +
#* Example prototype: <code>void adjust_channels(unsigned char *image,int x_size,int y_size,float red_factor,float green_factor,float blue_factor);</code>
 +
#* Hint/suggestion: compiling the function file with the <code>-S</code> option will produce an assembly-language output file. You can parse this file for lines like <code>.type adjust_channels, %function</code> to find the function name(s).
 +
#* Hint/suggestion #2: the <code>makeheaders</code> utility can produce a header file from an C source code input file. It is easier to find the function prototypes in the header file than in the C source file.
 +
# Place the prototype into the output file, adding the ifunc atttribute and resolver name. This produces the indirect function.
 +
#* Example indirect function: <code>__attribute__ (( ifunc("''name_of_resolver_function''") )) void adjust_channels(unsigned char *image,int x_size,int y_size,float red_factor,float green_factor,float blue_factor);</code>
 +
# Repeat these next two steps three times, once for each target (see the table below)
 +
## Place a GCC pragma directive into the output file to select one of the three targets: <code>#pragma GCC target arch="''put the target architecture string here''"</code>
 +
## Paste in the function, changing the function name so that the various versions don't conflict. For example, you could append a suffix identifying the architecture target, changing the function name <code>adjust_channels</code> to something like <code>adjust_channels_sve</code>.
 +
##* Tip: After the last function, add a pragma to switch the compiler back to targeting basic <code>armv8-a</code>
 +
# Add the resolver function, being sure to include the header file <code><nowiki><sys/auxv.h></nowiki></code> so that we can access the auxilliary vector (and therefore the description of available hardware capabilities).
 +
static void (*''name_of_resolver_function''(void)) {
 +
        long hwcaps  = getauxval(AT_HWCAP);
 +
        long hwcaps2 = getauxval(AT_HWCAP2);
 +
 +
        printf("\n### Resolver function - selecting the implementation to use for  foo()\n");
 +
        if (hwcaps2 & HWCAP2_SVE2) {
 +
                return adjust_channels_sve2;
 +
        } else if (hwcaps & HWCAP_SVE) {
 +
                return adjust_channels_sve;
 +
        } else {
 +
                return adjust_channels_asimd;
 +
        }
 +
};
 +
<ol><li value="5"> Build the software using the output file in place of the input file.
 +
<li> Test it thoroughly! Use your tool and compile the output. Look at the final binary to check for asimd/sve/sve2 code, and make sure the program runs in both contexts (run the binary directly to test it in asimd mode, and use the qemu-aarch64 tool to run the binary in sve2 mode).
 +
<li> Write it up in one or more blog posts.
 +
* The mark assigned will depend on:
 +
** The quality of the implementation - For example, does it accept various input files and process them correctly? Are the output files reliably usable? Is the tool user-friendly?
 +
** The amount and quality of the testing - For example, have you demonstrated that the tool works correctly? Can you show the asimd, sve, and sve2 code in the compiled binary?
 +
** The quality of the writing - For example, have you fully explained the code? Have you documented how to use your tool with clear, step-by-step instructions? Is the code provided in a form that makes it easy to test?
 +
** If the basic operation of the program is good, bonus marks may be assigned for additional features and for overcoming the (permitted) limitations. See the project page for details.
 +
</ol>
 +
==== Build Targets ====
 +
These are the build targets:
 +
{|cellspacing="0" width="100%" cellpadding="5" border="1"
 +
|-
 +
!Name!!Abbreviation!!Typical GCC Target String!!Description
 +
|-
 +
|Advanced SIMD||asimd||armv8-a||Basic "Advanced SIMD" implementation present in all aarch64 systems. Fixed-length 128-bit SIMD implementation.
 +
|-
 +
|Scalable Vector Extensions||sve||armv8-a+sve||Original version of the Scalable Vector Extensions. Variable-length 128-to-2048 bit SIMD implementation with predicate registers and first-fault load/store operations.
 +
|-
 +
|Scalable Vector Extensions version 2||sve2||armv8-a+sve2 (also armv9-a)||New, expanded version of the Scalable Vector Extensions used in Arm architecture version 9 systems. Has the same implementation details as sve, but with many additional instructions, such as operations useful for digital signal processing (dsp).
 +
|}
  
* (Not available)
+
GCC target strings may be used:
 +
* On the GCC command line, using the <code>-march=...</code> option
 +
* In function attributes: <code>__attribute__(( target("...") ))</code>
 +
* In pragmas: <code>#pragma GCC target "..."</code>
  
=== Week 14 - Class II ===
+
=== Week 12 Deliverables ===
 +
* Blog about your Project
 +
* November blog posts are due Sunday, December 4, at 11:59 pm.
 +
* Project Stage II is due next Thursday, December 8, at 12 noon.
  
=== Week 14 Deliverables ===
 
* [[Winter_2022_SPO600_Project|Project Stage 2]] due Monday April 18 (by 11:59 pm)
 
* [[Winter_2022_SPO600_Project|Project Stage 3]] due Friday April 22 (by 11:59 pm)
 
* April blog posts due Friday April 22 (by 11:59 pm)
 
  
###############################################
 
###############################################
 
###############################################
 
###############################################
 
###############################################
 
###############################################
 
###############################################
 
-->
 
<br/>
 
<br/>
 
<br/>
 
<br/>
 
<br/>
 
<br/>
 
<br/>
 
<br/>
 
<br/>
 
<br/>
 
<br/>
 
<br/>
 
<br/>
 
<br/>
 
<br/>
 
 
<br/>
 
<br/>
 
<br/>
 
<br/>

Revision as of 00:46, 7 December 2022

This is the schedule and main index page for the SPO600 Software Portability and Optimization course for Fall 2022.

Important.png
It's Alive!
This SPO600 weekly schedule will be updated as the course proceeds - dates and content are subject to change. The cells in the summary table will be linked to relevant resources and labs as the course progresses.


Schedule Summary Table

Please follow the links in each cell for additional detail which will be added below as the course proceeds -- especially for the Deliverables column.

Week Week of... Class I
Wednesday 11:40-1:25
Class II
Friday 11:40-1:25
Deliverables
(Summary - click for details)
1 Sep 5 Introduction to the Course / Introduction to the Problem / Computer Architecture Basics Binary Representation of Data Set up for the course / Lab 1
2 Sep 12 Introduction to 6502 Assembly 6502 Math / Jumps, Branches, and Subroutines Lab 2
3 Sep 19 6502 Strings 6502 String Input / Building Code: Make and Makefiles Lab 3
4 Sep 26 Compiler Optimizations Building Code: Compiler Options, GNU Autotools/Automake Lab 3, September blog posts
5 Oct 3 Introduction to 64-bit Architectures and Assembly Language (x86_64 and AArch64) Memory on 64-bit Systems Lab 4
6 Oct 10 Mid-semester Sync Discussion Algorithm Selection / In-line Assembler / SIMD Lab 5
7 Oct 17 Exploring 64-bit Code SVE2 Wrap up lab 5
Reading Oct 24 Reading Week
8 Oct 31 Optimization Trade-Offs / Algorithm Selection / Inline Assembler / SIMD Scalable Vector Extensions (SVE/SVE2) via Inline Assembler and C Intrinsics October blog posts
9 Nov 7 GNU ifunc & Project Overview Project Detail Blog about ifunc and your project work
10 Nov 14 Project Tips Advanced Memory Blog about project work
11 Nov 21 Project Techniques Project Demo Blog about project work
12 Nov 28 Benchmarking Step-by-Step Project Minimum Requirements Blog about project work, November blog posts
13 Dec 5 Enhancing Your Project Project Discussion Blog about project work; Project Stage 2 due Thursday December 8 at Noon
14 Dec 12 Future Directions in Architecture (No class) Project Stage 3, December blog posts


Contents

Week 1

Week 1 - Class I

Video

General Course Information

  • Course resources are linked from the CDOT wiki, starting at https://wiki.cdot.senecacollege.ca/wiki/SPO600 (Quick find: This page will usually be Google's top result for a search on "SPO600"), arranged by week and class. There will be lots of hyperlinks -- be sure to follow these links.
  • Coursework is submitted by blogging. The only exception to this is quizzes.
  • Quizzes will be short (~1 page) and will be held without announcement at the start of any synchronous class. There is no opportunity to re-take a missed quiz, but your lowest three quiz scores will not be counted, so do not worry if you miss one or two.
    • Students with test accommodations: an alternate monthly quiz can be made available via the Test Centre. Communicate with your professor for details.
  • Course marks (see Weekly Schedule for dates):
    • 60% - Project Deliverables in three phases (15%, 20%, 25%)
    • 20% - Communication (Blog writing, in four phases roughly a month long each, 5% each)
    • 20% - Labs and Quizzes (10% labs; 10% for quizzes - lowest 3 quiz scores not counted)
About SPO600 Classes
  • Wednesday: synchronous (live) classes at 11:40 am - login to learn.senecacollege.ca ("Blackboard"), go to SPO600, and select the "Wednesday Classes" option on the left-hand menu.
  • Friday: these classes will usually be asynchronous (pre-recorded) - see this page for details each week.
  • There may be occasional exceptions to this pattern.

Introduction to the Problems

Porting and Portability
  • Most software is written in a high-level language which can be compiled into machine code for a specific computer architecture. In many cases, this code can be compiled or interpreted for execution on multiple computer architectures - this is called 'portable' code. However, there is a lot of existing code that contains some architecture-specific code fragments which contains assumptions about the architecture, resulting in architecture-specific high-level or Assembly Language code.
  • Reasons that code is architecture-specific:
    • System assumptions that don't hold true on other platforms
    • Code that takes advantage of platform-specific features
  • Reasons for writing code in Assembly Language include:
    • Performance
    • Atomic Operations
    • Direct access to hardware features, e.g., CPUID registers
  • Most of the historical reasons for using assembler are no longer valid. Modern compilers can out-perform most hand-optimized assembly code, atomic operations can be handled by libraries or compiler intrinsics, and most hardware access should be performed through the operating system or appropriate libraries.
  • A new architecture has appeared: AArch64, which is part of ARMv8. This is the first new computer architecture to appear in several years (at least, the first mainstream computer architecture).
  • At this point, most key open source software (the software typically present in a Linux distribution such as Ubuntu or Fedora, for example) now runs on AArch64. However, it may not yet be as extensively optimized as on older architectures (such as x86_64).
Benchmarking and Profiling

Benchmarking involves testing software performance under controlled conditions so that the performance can be compared to other software, the same software operating on other types of computers, or so that the impact of a change to the software can be gauged.

Profiling is the process of analyzing software performance on finer scale, determining resource usage per program part (typically per function/method). This can identify software bottlenecks and potential targets for optimization. The resource utilization studies may include memory, CPU cycles/time, or power.

Optimization

Optimization is the process of evaluating different ways that software can be written or built and selecting the option that has the best performance tradeoffs.

Optimization may involve substituting software algorithms, altering the sequence of operations, using architecture-specific code, or altering the build process. It is important to ensure that the optimized software produces correct results and does not cause an unacceptable performance regression for other use-cases, system configurations, operating systems, or architectures.

The definition of "performance" varies according to the target system and the operating goals. For example, in some contexts, low memory or storage usage is important; in other cases, fast operation; and in other cases, low CPU utilization or long battery life may be the most important factor. It is often possible to trade off performance in one area for another; using a lookup table, for example, can reduce CPU utilization and improve battery life in some algorithms, in return for increased memory consumption.

Most advanced compilers perform some level of optimization, and the options selected for compilation can have a significant effect on the trade-offs made by the compiler, affecting memory usage, execution speed, executable size, power consumption, and debuggability.

Build Process

Building software is a complex task that many developers gloss over. The simple act of compiling a program invokes a process with five or more stages, including pre-processing, compiling, optimizing, assembling, and linking. However, a complex software system will have hundreds or even thousands of source files, as well as dozens or hundreds of build configuration options, auto configuration scripts (cmake, autotools), build scripts (such as Makefiles) to coordinate the process, test suites, and more.

The build process varies significantly between software packages. Most software distribution projects (including Linux distributions such as Ubuntu and Fedora) use a packaging system that further wraps the build process in a standardized script format, so that different software packages can be built using a consistent process.

In order to get consistent and comparable benchmark results, you need to ensure that the software is being built in a consistent way. Altering the build process is one way of optimizing software.

Note that the build time for a complex package can range up to hours or even days!

Course Setup

Follow the instructions on the SPO600 Communication Tools page to set up a blog, create SSH keys, and send your blog URLs and public key to me.

I will use this information to:

  1. Update the Current SPO600 Participants page with your information, and
  2. Create an account for you on the SPO600 Servers.

This updating is done in batches once or twice a week -- allow some time!

How open source communities work

Week 1 - Class II

Video

Binary Representation of Data

  • Binary
    • Binary is a system which uses "bits" (binary digits) to represent values.
    • Each bit has one of two values, signified by the symbols 0 and 1. These correspond to:
      • Electrically: typically off/on, or low/high voltage, or low/high current. Many other electrical representations are possible.
      • Logically: false or true.
    • Binary numbers are resistant to errors, especially when compared to other systems such as analog voltages.
      • To represent the numbers 0-10 as an analog electrical value, we could use a voltage from 0 - 10 volts. However, if we use a long cable, there will be signal loss and the voltage will drop: we could apply 10 volts on one end of the cable, but only observe (say) 9.1 volts on the other end of the cable. Alternately, electromagnetic interference from nearby devices could slightly increase the signal.
      • If we instead use the same voltages and cable length to carry a binary signal, where 0 volts == off == "0" and 10 volts == on == "1", a signal that had degraded from 10 volts to 9.1 volts would still be counted as a "1" and a 0 volt signal with some stray electromagnetic interference presenting as (say) 0.4 volts would still be counted as "0". However, we will need to use multiple bits to carry larger numbers -- either in parallel (multiple wires side-by-side), or sequentially (multiple bits presented over the same wire in sequence).
  • Integers
    • Integers are the basic building block of binary numbering schemes.
    • In an unsigned integer, the bits are numbered from right to left starting at 0, and the value of each bit is 2bit. The value represented is the sum of each bit multiplied by its corresponding bit value. The range of an unsigned integer is 0:2bits-1 where bits is the number of bits in the unsigned integer -- for example, an 8-bit unsigned integer has a range of 0 through 28-1 = 255.
    • Signed integers are generally stored in twos-complement format, where the highest bit is used as a sign bit. If that bit is set, the value represented is -(!value)-1 where ! is the NOT operation (each bit gets flipped from 0→1 and 1→0)
  • Fixed-point
    • A fixed-point value is encoded the same as an integer, except that some of the bits are fractional -- they're considered to be to the right of the "binary point" (binary version of "decimal point" - or more generically, the radix point). For example, binary 000001.00 is decimal 1.0, and 000001.11 is decimal 1.75.
    • An alternative to fixed-point fractional values is integer values in a smaller unit of measurement. For example, some accounting software may use integer values representing cents. For input and display purposes, dollar and cent values are converted to/from cent values. Similarly, a program that stores measurements could use milimetres instead of fractional meters.
  • Floating-point
    • The most commonly-used floating point formats are defined in the IEEE 754 standard.
    • IEEE754 floating point numbers have three parts: a sign bit (0 for positive, 1 for negative), a mantissa or significand, and an exponent. The significand has an implied 1 and radix point preceeding the stored value. The exponent is stored as an unsigned integer to which a bias value has been added; the bias value is 2(number of exponent bits - 1) - 1. The floating point value is interpreted in normal cases as sign mantissa * 2(exponent - bias). Exponent values which are all-zeros or all-ones encode four categories of special cases: zero, infinity, Not a Number (NaN), and subnormal numbers (numbers which are close to zero, where the significand does not have an implied 1 to the left of the radix point); in these special cases, the sign bit and significand values may have special meanings.
    • There are some new floating-point formats appearing, such as Brain Float 16, a 16-bit format with the same dynamic range as 32-bit IEEE 754 floating point but with less accuracy, intended for use in machine learning applications.
  • Characters
    • Characters are encoded as integers, where each integer corresponds to one code point in a character table (e.g., code 65 in ASCII corresponds to the character "A").
    • Historically, many different coding schemes have been used, but the two most common ones were the American Standard Code for Information Interchange (ASCII), and Extended Binary Coded Decimal Interchange Code (EBCDIC - primarily used on IBM midrange and mainframe systems).
    • ASCII characters occupied seven bits (code points 0-127), and contains only characters used in North American English. ASCII characters are usually encoded in bytes, so many vendors of ASCII-based systems used the remaining codes 128-255 for special characters such as graphics, line symbols (horizontal, vertical, connector, and corner line symbols for drawing tables), and accented characters; these were called "extended ASCII".
    • Several ISO standards exist in an attempt to standardize the "extended ascii" characters, such as ISO8859, which was intended to enable the encoding of European languages by adding currency symbols and accented characters. However, the original version of ISO8859 (called ISO8859-1) does not include all accented characters and was created before the Euro symbol was standardized, so there are multiple versions of ISO8859, ranging from ISO8859-1 through ISO8859-15.
    • The Unicode and ISO10646 initiatives were initiated to create a single character code set that would encode all symbols used in human writing, both for current and obsolete languages. These initiatives were merged, and the Unicode and ISO10646 standards define a common character set with 232 potential code points. However, Unicode also describes transformation formats for data interchange, rendering and composition/decomposition recommendations, and font symbol recommendations.
    • The first 127 code points in Unicode correspond to ASCII code points, and the first 255 code points correspond to ISO8869-1 code points. The first 65536 code points form the Basic Multilingual Pane (BMP), which contains most of the characters required to write in all contemporary languages. Therefore, for many applications, it is inefficient to store Unicode as full 32-bit values. To solve this issue, several Unicode Transformation Formats (also known -- technically incorrectly -- as Unicode Transfer Formats) have been defined, including UTF-8, UTF-16, and UTF-32 (32-bit). UTF-8 represents ASCII and some ISO-8859 characters as a single byte, the remainder of the BMP as 2-3 bytes per character, and the remaining characters using 3-4 bytes per character. UTF-16 is similar, encoding much of the BMP in a single 16-bit value, and most other characters as two 16-bit values.
  • Sound
    • Sound waves are air pressure vibrations.
    • Sound is most often represented in raw digital form as a series of time-based measurements of air pressure, called Pulse Coded Modulation (PCM).
    • PCM takes a lot of storage, so sound is often compressed in either a lossless (perfectly recoverable) or lossy format (higher compression, but the decompressed data doesn't perfectly match the original data). To permit high compression ratios with minimal impact on quality, psychoacoustic compression is used - sound variations that most people can't perceive are removed.
  • Graphics
    • The human eye perceives luminance (brightness) as well as hue (colour). Our main hue receptors ("cones") are generally sensitive to three wavelengths: red, green, and blue (RGB). We can stimulate the eye to perceive most colours by presenting a combination of light at these three wavelengths utilizing metamerism.
    • Digital displays emit RGB colours, which are mixed together and perceived by the viewer. This is called additive colour.
    • For printing, cyan (C)/yellow (Y)/magenta (M) pigmented inks are used, plus black (K) to reduce the amount of colour ink required to represent dark tones; this is known as CYMK colour. These pigments absorb light at specific frequencies, subtracting energy from white or near-white sunlight or artificial light. This is called subtractive colour.
    • Images are broken into picture elements (pixels) and each pixel is usually represented by a group of values for RGB or CYMK channels, where each channel is represented by an integer or floating-point value. For example, using an 8-bit-per-channel integer scheme (also known as 24-bit colour), the brightest blue could be represented as R=0,G=0,B=255; the brightest yellow would be R=255,G=255,B=0; black would be R=0,G=0,B=0; and white would be R=255,G=255,B=255. With this 8-bit-per-channel (24 bit total) scheme, the number of unique colours available is 256^3 ~= 16 million.
    • As with sound, the raw storage of sampled data requires a lot of storage space, so various lossy and lossless compression schemes are used. Highest compression is achieved with psychovisual compression (e.g., JPEG).
    • Moving pictures (video, animations) are stored as sequential images, often compressed by encoding only the differences between frames to save storage space. Motion compensation can further compress the data stream by describing how portions of the previous frame should be moved and positioned in the current frame.
  • Compression techniques
    • Huffman encoding / Adaptive arithmetic encoding
      • Instead of fixed-length numbers, variable-length numbers are used, with the most common values encoded in the smallest number of bits. This is an effective strategy if the distribution of values in the data set is uneven.
    • Repeated sequence encoding (1D, 2D, 3D)
      • Run length encoding is an encoding scheme that records the number of repeated values. For example, fax messages are encoded as a series of numbers representing the number of white pixels, then the number of black pixels, then white pixels, then black pixels, alternating to the end of each line. These numbers are then represented with adaptive arithmetic encoding.
      • Text data can be compressed by building a dictionary of common sequences, which may represent words or complete phrases, where each entry in the dictionary is numbered. The compressed data contains the dictionary plus a sequence of numbers which represent the occurrence of the sequences in the original text. On standard text, this typically enables 10:1 compression.
    • Decomposition
      • Compound audio waveforms can be decomposed into individual signals, which can then be modelled as repeated sequences. For example, a waveform consisting of two notes being played at different frequencies can be decomposed into those separate notes; since each note consists of a number of repetitions of a particular wave pattern, they can individually be represented in a more compact format by describing the frequency, waveform shape, and amplitude characteristics.
    • Palletization
      • Images often contain repeated colours, and rarely use all of the available colours in the original encoding scheme. For example, a 1920x1080 "full HD" image contains about 2 million pixels, so if every pixel was a different colour, there would be a maximum of 2 million colours. But it's likely that many of the pixels in the image are the same colour, so there might only be (perhaps) 4000 colours in the image. If each pixel is encoded as a 24-bit value, there are potentially 16 million colours available, and there is no possibility that they are all used. Instead, a palette can be provided which specifies each of the 4000 colours used in the picture, and then each pixel can be encoded as a 12-bit number which selects one of the colours from the palette. The total storage requirement for the original 24-bit scheme is 1920*1080*3 bytes per pixel = 5.9 MB. Using a 12-bit pallette, the storage requirement is 3 * 4096 bytes for the palette plus 1920*1080*1.5 bytes for the image, for a total of 3 MB -- a reduction of almost 50%
    • Psychoacoustic and psychovisual compression
      • Much of the data in sound and images cannot be perceived by humans. Psychoacoustic and psychovisual compression remove artifacts which are least likely to be perceived. As a simple example, if two pixels on opposite sides of a large image are almost but not exactly the same, most people won't be able to tell the difference, so these can be encoded as the same colour if that saves space (for example, by reducing the size of the colour palette).


Week 1 Deliverables

  1. Follow the SPO600 Communication Tools set-up instructions.
  2. Optional (strongly recommended): Set up a personal Linux system.
  3. Optional: If you have an AArch64 development board (such as a Raspberry Pi 4, Raspberry Pi 400, or 96Boards device), consider installing a 64-bit Linux operating system such as Fedora on it.
  4. Start work on Lab 1. Blog your work.

Week 2

Week 2 - Class I

Video

Machine Language, Assembly Language

Idea.png
Follow the Links!
To get the full benefit of the following material, please follow the links embedded within it. For additional detail, see the Category links at the bottom of those pages -- for example, the category linked from many of the following pages has over 30 pages of content.
  • Although we program computers in a variety of languages, they can really only execute one language: Machine Language, which is encoded in an architecture-specific binary code, sometimes called object code.
  • Machine language is not easy to read. Assembly Language corresponds very closely to machine language, but is (sort of!) human-readable.
  • Assembly language is converted into machine code by a particular type of compiler called an Assembler (sometimes the language itself is also referred to as "Assembler").

6502

Modern processors are complex - the reference manual for 64-bit ARM processors is over 11000 pages long! - so we're going to look at assembly language on a much simpler processor to get started. This processor is the 6502, a processor used in many early home and personal computers as well as video game systems, including the Commodore PET, VIC-20, C64; the Apple II; the Atari 400 and 800 computers and 2600 video game systems; and many others.

Lab 2

Week 2 - Class II

Videos

Reading

Week 2 Deliverables

  1. If not already completed last week:
    1. Set up your SPO600 Communication Tools
    2. Complete Lab 1 and blog your work.
  2. Study the 6502 Instructions and 6502 Addressing Modes and make sure you understand what each one does.
  3. Complete Lab 2 and blog your results.

Week 3

Week 3 - Class I

Video

Lab


Week 3 - Class II

Video

Resources

Week 3 Deliverables

  • Lab 3
  • Note that September blog posts are due at the end of next week, so don't get behind in your blogging


Week 4

Week 4 - Class I

Video

Reading Resources

Week 4 - Class II

Video

Resouces

Week 4 Deliverables

  • September blogs are due this weekend (Sunday, October 2 at 11:59 pm)

Week 5

Week 5 - Class I

Video

Resources

Week 5 - Class II

Video

Lab 4

Week 5 Deliverables


Week 6

Week 6 - Class I

We used this class for introductions, a discussion of how things are going, and feedback on the course.

Week 6 - Class II

Video

Lab 5


Week 6 Deliverables


Week 7

Week 7 - Class I

Video

  • Video summary will be posted after editing

Week 7 - Class II

Please catch up on course material to this point. If you are fully caught up, you can start to take a look at SVE2:

Reading

SVE2 Demonstration

  • Code available here: https://github.com/ctyler/sve2-test
  • This is an implementation of a very simple program which takes an image file, adjusts the red/green/blue channels of that file, and then writes an output file. Each channel is adjusted by a factor in the range 0.0 to 2.0 (with saturation).
  • The image adjustment is performed in the function adjust_channels() in the file adjust_channels.c. There are three implementations:
    1. A basic (naive) implementation in C. Although this is a very basic implementation, it is potentially subject to autovectorization.
    2. An implementation using inline assembler for SVE2 with strucure loads.
    3. An implementation using inline assembler for SVE2 with an interleaved factor table.
    4. An implementation using ACLE compile intrinsics.
  • The implementation built is dependent on the value of the ADJUST_CHANNEL_IMPLEMENTATION macro.
  • The provided Makefile will build four versions of the binary -- one using each of the four implementations -- and it will run through 3 tests with each binary. The tests use the input image file tests/input/bree.jpg (a picture of a cat) and place the output in the files tests/output/bree[1234][abc].jpg. The output files are processed with adjustment factors of 0.5/0.5/0.5, 1.0/1.0/1.0, and 2.0/2.0/2.0.
  • Please examine, build, and test the code, compare the implementations, and note how it works - there are extensive comments in the code, especially for implementation 2.
  • Your observations about the code might make a good blog post!


Week 7 Deliverables

  • Complete Lab 4 and Lab 5
  • Remember that October blogs are due soon.

Week 8

Week 8 - Class I

Video

Week 8 - Class II

Video

Reading

SVE2 Demonstration

  • Code available here: https://github.com/ctyler/sve2-test
  • This is an implementation of a very simple program which takes an image file, adjusts the red/green/blue channels of that file, and then writes an output file. Each channel is adjusted by a factor in the range 0.0 to 2.0 (with saturation).
  • The image adjustment is performed in the function adjust_channels() in the file adjust_channels.c. There are three implementations:
    1. A basic (naive) implementation in C. Although this is a very basic implementation, it is potentially subject to autovectorization.
    2. An implementation using inline assembler for SVE2 with strucure loads.
    3. An implementation using inline assembler for SVE2 with an interleaved factor table.
    4. An implementation using ACLE compile intrinsics.
  • The implementation built is dependent on the value of the ADJUST_CHANNEL_IMPLEMENTATION macro.
  • The provided Makefile will build four versions of the binary -- one using each of the four implementations -- and it will run through 3 tests with each binary. The tests use the input image file tests/input/bree.jpg (a picture of a cat) and place the output in the files tests/output/bree[1234][abc].jpg. The output files are processed with adjustment factors of 0.5/0.5/0.5, 1.0/1.0/1.0, and 2.0/2.0/2.0.
  • Please examine, build, and test the code, compare the implementations, and note how it works - there are extensive comments in the code, especially for implementation 2.
  • Your observations about the code might make a good blog post!

Week 8 Deliverables

  • Continue your blogging
  • Include blogging on SVE/SVE
  • The second group of blog posts is due on or before this Sunday (November 6, 11:59 pm)

Week 9

Week 9 - Class I

Video

iFunc

GNU iFunc is a facility for handling indirect functions. The basic premise is that you prototype the function to be called, add the ifunc attribute to that prototype, and provide the name of a resolver function. The resolver function is called at program initialization, and returns a pointer to the function to be executed when the function referenced in the prototype is called. The resolver typically picks one of several implementations based on the capabilities of the machine on which the code is running; for example, it could return a pointer to a non-SVE, SVE, or SVE2 implementation of a function based on cpu capabilities (on an Aarch64 system) or it could return a pointer to an SSE, SSE2, AVX, or AVX512 implementation (on an x86_64 system).

There is a GitHub repository available with example iFunc code -- please clone this to israel.cdot.systems and build and test the code there. You should see different results if you run the output executable directly (./ifunc-test) and run it through the qemu-aarch64 tool, which will emultate SVE2 capabilities (qemu-aarch64 ./ifunc-test). Make sure you understand how the code works.

Reading/Resources

Week 9 - Class II

Video

  • Edited summary video - Important! This video contains a detailed discussion of the requirements for the course project.
    • Project discussion starts at beginning of video
    • Demo of what the project needs to do (manually performing the same steps) starts at 0:27:47
    • Recap/summary of the demo starts around 1:02:05

Project

Week 9 Deliverables

  • Investigate the iFunc example code
  • Blog about your investigation
  • Start blogging about your project

Week 10

Week 10 - Class I

Video

  • Summary Video - Tips and techniques for working productively in a remote aarch64 Linux system.

Week 10 - Class II

  • A discussion of of Advanced Memory Systems
  • No video available at this time
  • Not directly related to the course project

Week 10 Deliverables

  • Blog about your project work.

Week 11

Week 11 - Class I

Video

Week 11 - Class II

Video

  • Summary Video - A demonstration of one solution to the course project.

Week 11 Deliverables

  • Blog about your project work.

Week 12

Week 12 - Class I

  • A discussion of benchmarking
  • No video available at this time
  • Not directly related to the course project

Week 12 - Class II

Video

  • Summary Video - Step-by-Step discussion of the minimum requirements for a successful project

Minimum Requirements for a Successful Project

  • Remember: This project is about building a tool that processes C source code. The tool itself can be in any suitable computer language. It is recommended that you use a language that can manipulate text easily. The tool will receive an input file containing a C function. The tool will output a modified file(s) with ifunc capability for asimd/sve/sve2 computers.
  • Requirements for a minimum implementation:
  1. Find the protype and function name by analyzing the C function.
    • Example name: adjust_channels
    • Example prototype: void adjust_channels(unsigned char *image,int x_size,int y_size,float red_factor,float green_factor,float blue_factor);
    • Hint/suggestion: compiling the function file with the -S option will produce an assembly-language output file. You can parse this file for lines like .type adjust_channels, %function to find the function name(s).
    • Hint/suggestion #2: the makeheaders utility can produce a header file from an C source code input file. It is easier to find the function prototypes in the header file than in the C source file.
  2. Place the prototype into the output file, adding the ifunc atttribute and resolver name. This produces the indirect function.
    • Example indirect function: __attribute__ (( ifunc("name_of_resolver_function") )) void adjust_channels(unsigned char *image,int x_size,int y_size,float red_factor,float green_factor,float blue_factor);
  3. Repeat these next two steps three times, once for each target (see the table below)
    1. Place a GCC pragma directive into the output file to select one of the three targets: #pragma GCC target arch="put the target architecture string here"
    2. Paste in the function, changing the function name so that the various versions don't conflict. For example, you could append a suffix identifying the architecture target, changing the function name adjust_channels to something like adjust_channels_sve.
      • Tip: After the last function, add a pragma to switch the compiler back to targeting basic armv8-a
  4. Add the resolver function, being sure to include the header file <sys/auxv.h> so that we can access the auxilliary vector (and therefore the description of available hardware capabilities).
static void (*name_of_resolver_function(void)) {
        long hwcaps  = getauxval(AT_HWCAP);
        long hwcaps2 = getauxval(AT_HWCAP2);

        printf("\n### Resolver function - selecting the implementation to use for  foo()\n");
        if (hwcaps2 & HWCAP2_SVE2) {
                return adjust_channels_sve2;
        } else if (hwcaps & HWCAP_SVE) {
                return adjust_channels_sve;
        } else {
                return adjust_channels_asimd;
        }
};
  1. Build the software using the output file in place of the input file.
  2. Test it thoroughly! Use your tool and compile the output. Look at the final binary to check for asimd/sve/sve2 code, and make sure the program runs in both contexts (run the binary directly to test it in asimd mode, and use the qemu-aarch64 tool to run the binary in sve2 mode).
  3. Write it up in one or more blog posts.
    • The mark assigned will depend on:
      • The quality of the implementation - For example, does it accept various input files and process them correctly? Are the output files reliably usable? Is the tool user-friendly?
      • The amount and quality of the testing - For example, have you demonstrated that the tool works correctly? Can you show the asimd, sve, and sve2 code in the compiled binary?
      • The quality of the writing - For example, have you fully explained the code? Have you documented how to use your tool with clear, step-by-step instructions? Is the code provided in a form that makes it easy to test?
      • If the basic operation of the program is good, bonus marks may be assigned for additional features and for overcoming the (permitted) limitations. See the project page for details.

Build Targets

These are the build targets:

Name Abbreviation Typical GCC Target String Description
Advanced SIMD asimd armv8-a Basic "Advanced SIMD" implementation present in all aarch64 systems. Fixed-length 128-bit SIMD implementation.
Scalable Vector Extensions sve armv8-a+sve Original version of the Scalable Vector Extensions. Variable-length 128-to-2048 bit SIMD implementation with predicate registers and first-fault load/store operations.
Scalable Vector Extensions version 2 sve2 armv8-a+sve2 (also armv9-a) New, expanded version of the Scalable Vector Extensions used in Arm architecture version 9 systems. Has the same implementation details as sve, but with many additional instructions, such as operations useful for digital signal processing (dsp).

GCC target strings may be used:

  • On the GCC command line, using the -march=... option
  • In function attributes: __attribute__(( target("...") ))
  • In pragmas: #pragma GCC target "..."

Week 12 Deliverables

  • Blog about your Project
  • November blog posts are due Sunday, December 4, at 11:59 pm.
  • Project Stage II is due next Thursday, December 8, at 12 noon.