HNHacker News
TopNewBestAskShowJobs

Joky

850 karma · joined October 17, 2014

submissionscomments
Joky··on A note of reflection after 10 years of “A successful Git branching model”
The original post was describing "Administrator is in charge of merging, maintains a linear history. Releases are just tags on master", you seem to discuss the various branching strategy while it seems to me the explanation was about the fact that we need a branching strategy in the first place. Do I miss something?
Joky··on IBM to transition z/OS, Power and AIX Compilers to being LLVM/Clang-based
The fact that Apple basically bootstrapped the project 10 years ago does not mean it is forever “Apple’s compiler”. The community is much much larger than any single vendor out there.
Joky··on MLIR: A Compiler Infrastructure for the End of Moore's Law
Exactly: the "why not just LLVM" can be greatly explained by the need to target heterogeneous hardware from DSLs.
Joky··on MLIR: A Compiler Infrastructure for the End of Moore's Law
You are making good point, but I think you're missing some aspects and the title hints about it: we are also trying to address today's heterogeneity. By having a flexible infrastructure and (hopefully) an ecosystem with which the interaction cost is lowered, you can re-assemble more easily custom compilers for specific use-cases.

This does not make MLIR the best infrastructure for building an industrial embedded Javascript compiler for instance (just like you wrote B3 to move away from LLVM), but I am not convinced that between these two, MLIR is the niche ;-) Time will tell!

Joky··on MLIR: A Compiler Infrastructure for the End of Moore's Law
I wouldn't say that we "recommend" ANTLR, but yes it is likely "possible" to represent some AST in MLIR, but we just aren't convinced that MLIR is the best tool there so far.
Joky··on MLIR: A Compiler Infrastructure for the End of Moore's Law
As far as I understand, it is entirely different: IREE is a runtime system.
Joky··on MLIR: A Compiler Infrastructure for the End of Moore's Law
Basically 6 man months of some of the most senior Apple engineers? Seems fair. But at this point you "just" have an IR, and it is dedicated to your own use case, which has its advantage but also some inconvenient, especially when it comes to interacting with other tools/framework. So if every time you want to create a DSL for a use-case you need this amount of investment, this is a non-trivial cost.

Now think about use-cases like https://drive.google.com/file/d/19pSpEsi4I9-MKLRodD-po82HFCW... and what the MLIR ecosystem (if it develop can provide them). After 6 months of work, instead of having "just" an IR for their abstraction, they would likely have iterated on the IR, revisited their abstraction, write optimizations, and most importantly mapped their use-cases to heterogeneous HW! They can benefit fairly easily from the work of PlaidML on the affine/polyhedral optimizations for instance.

On the other hand, if a new HW vendor would like to expose their accelerator to a wide range of use-cases (whether it is Fortran OpenACC or the Stencil DSL), plugin lowering/backend to MLIR is a much efficient way than the alternative (which are almost non-existent today by the way).

If it was so easy to write its own compiler end-to-end from scratch, wouldn't LLVM be out of users by now?

Joky··on Clang vs. other open-source compilers
Minor precision: it is Apache 2.0 with LLVM exceptions, the exceptions are intended to address libc++ and other runtime use-cases I believe.
Joky··on Chris Lattner to Lead SiFive Platform Engineering Team
You seem to have the impression that a machine learning model is all you need to drive the car: this isn't true. There is a lot of real-time system programming and overall architecture.

> Andrej Karpathy basically threw everything out once he replaced Latner.

What are you basing this on? I don't believe there is any truth to this. First Chris was not replaced by Andrej but by Jim Keller and then Stuart Bowers. Andrej is amazing in his field and has done impressive things at Tesla, but he is managing the ML part of the org, which was a relatively small portion of the Autopilot Software group overall.

Joky··on Chris Lattner to Lead SiFive Platform Engineering Team
"2020-01-24: Thoughts on Tensor Code Generation in MLIR" here: https://mlir.llvm.org/talks/ for reference.
Joky··on Chris Lattner to Lead SiFive Platform Engineering Team
I wouldn't be worried for MLIR.

MLIR is getting used internally more and more inside TensorFlow, but also by separate team in different projects, like IREE for example (https://github.com/google/iree ). The TensorFlow lite converter has been replaced by the new MLIR-based one, similarly for the Edge TPU.

But more than that, what makes me confident is the traction we are getting outside Google. First, we landed MLIR in LLVM last month: https://github.com/llvm/llvm-project/commit/0f0d0ed1c78f1a80...

The LLVM Fortran frontend (f18/flang) which will merge soon in the LLVM monorepo is using MLIR for their own IR. It'll be exiting to develop a non-ML MLIR-based frontend within LLVM! In particular Flang is opening an HPC perspective that could be leveraged by other DSLs later. They are adding an OpenMP dialect to MLIR right now: https://llvm.discourse.group/t/rfc-openmp-dialect-in-mlir/39...

Intel has been actively porting their nGraph/PlaidML framework to be based on MLIR (search for "The Stripe dialect" and "nGraph Dialect" here: https://mlir.llvm.org/talks/ ).

I'm less familiar with S4TF, but I know they recently got some nice new hires, including https://twitter.com/DaveAbrahams/status/1207690883782467584

Joky··on Apple’s Great GPL Purge (2012)
Well: clang exists only because of the decision to get rid of GCC in the first place :)
Joky··on Andrej Karpathy talks about how Tesla's NNs are structured and trained [video]
Pytorch is used to train models on servers/cloud, not to drive the car later. The trained model is converted to something native to the embedded environment of the car.
Joky··on Andrej Karpathy talks about how Tesla's NNs are structured and trained [video]
You seem to use "regulated" as in "forbidden" right now, but maybe I am misreading so can you elaborate?

Today thousands are dying every year on the road, are we just forbidding cars entirely?

> And no, failover control is not acceptable given the past incidents and deaths.

How many incidents/deaths per miles driven? How does it compare to all the other transportation systems?

Joky··on Parallel GCC: a research project aiming to parallelize a real-world compiler
Sure: for instance in LLVM global values (variables, functions) have use-lists, so do constants. That means you can find easily every uses of a function: they are chained in a doubly linked list. In MLIR, it isn't the case. We use symbol name to refer to other global object, we lose the ability to easily find the uses of these global objects. On the other hand functions are well isolated and you can run Function passes in a multi-threaded fashion safely.
Joky··on Parallel GCC: a research project aiming to parallelize a real-world compiler
In LLVM, even with LLVM_ENABLE_THREADS, the compilation of a module is still sequential as far as I know.

The LLVM Context is the unit of isolation between threads. For instance the multi-threaded ThinLTO optimizer/codegen will use one LLVM Context per-thread, and so each thread is processing a single Module / TU.

The issue for concurrency is fairly deep in LLVM (use-lists, constant stored uniquely in context, etc.) that would make it really difficult to address intra-module parallelism.

MLIR for example is designed with this in mind and the pass-manager is already multi-threaded at every level of nesting (function passes would runs on two functions of the same Module in parallel). This is causing other complication/inefficiency in the infrastructure, I'm still not sure how much it is a good tradeoff, we'll see...

Joky··on Speeding Up Our Build Pipelines
The main difficulty with C/C++ is making the build steps "hermetics". The build system is frequently unaware of what files the preprocessor will touch/lookup before hands.

Build systems like Bazel are providing some way to ensure some level of isolation that provides correctness guarantees. It also provide an easy way to statically build a superset of the files needed for each step of the build, allowing to integrate with a distributed service. Some related documentation/discussions can be found starting from https://docs.bazel.build/versions/master/remote-execution.ht... Another good source of information is online videos from BazelCon: https://conf.bazel.build

Joky··on $1B for 20,000 Bay Area homes
I leave in MV for a few years: in my opinion there has been many development of dense complex in the recent years (not to say that it is enough): especially alongside El Camino and also in the San Antonio area.
Joky··on Arm announces its new premium CPU and GPU designs
Tensorflow has the XLA compiler for targeting accelerators. In practice the integration isn't necessarily amazing right now, but the plan is to make it much better and able to generalize more easily.

Disclaimer: I work on the MLIR/XLA team.

Joky··on Arm announces its new premium CPU and GPU designs
I'm not super familiar with the distinction you're making ("array processor" vs "vector processor"), do you have any pointer I could look at?
Joky··on Go to Statement Considered Harmful (1968) [pdf]
Quite the opposite IMO: it gives a name to the computation and simplifies the control flow by making the computation more functional.
Joky··on Google Is Turning Off the Works-with-Nest API
Isn't "Works with Google Assistant" supposed to become the replacement API? It seems like they don't offer direct access to the data or the device but plan to allow "some" level of control? (I don't know the space, apologize if my questions are naive...)

> As a Works with Nest developer and partner, you will not be able to access or control Nest devices once the Works with Nest APIs are turned off on August 31, 2019. Moving forward, our team will focus on making Works with Google Assistant the most helpful and intelligent ecosystem for the home, enabling all of the products in your users’ homes to work together. We encourage all smart home developers to visit the Actions on Google Smart Home developer site to learn how to integrate your devices or services with the Google Assistant.

Joky··on Google Is Turning Off the Works-with-Nest API
Are the physical devices not working after this change, or rather some of the applications developed for these devices need to switch to another API?
Joky··on Tesla Live Stream – Autonomy Day [video]
> My point was that it appears that Tesla doesn't have a large and varied dataset. It has a small and pre-selected data set,

I don't thinks it is a fair characterization: first the notion of "large" is fairly subjective. But more importantly the fact that they collect data on pre-determined triggers is just a guarantee that the dataset is not over-fitted (let say to the 280 and 101 in the Bay Area and to Elon's commute in LA) and instead has good coverage of the world.

Their capabilities of triggering on situation allows them to grow the dataset quickly in a supervised way. It is all about the granularity of these triggers, imagine that you can express "collect situation in tunnels with jerk higher than X m/s3" or "collect all lane change abort in snow condition", ... In the next 24h you get data from all over the world and this data is automatically tagged and classified by the neural net.

Joky··on Tesla Live Stream – Autonomy Day [video]
I think the key thing is that you assume a single dot-product per RAM load. In general you get higher compute performance by doing more than this. You load weights for a layer in the SRAM, then stream from RAM tiles of data that you can compute multiple operations on.

In the same way, to reach peak FLOPs on a GPU you better use the local/shared memory as much as possible.

Joky··on Commit messages guide
> When I talk about a particular point in _history_, I use past tense. :)

When you read a novel written in the present tense, you don't assume it tells a story that is currently happening right? ;)

Joky··on Commit messages guide
Past-tense is about the what the programmer did to get to the patch, present (imperative...) tense to me is describing in the absolute what is a diff is doing (it describes the act of `git am` this patch).

While the past tense makes sense from the author point of view describing what just happened until they typed `git commit`, I'm in the "present" camp since I find it more relevant to me when reading `git show <rev>`: it tells me what the commit is doing at this particular point in history. I don't really care about the programmer/author centric point of view at the time the message was written.

Joky··on F18 Fortran compiler and runtime project is accepted as part of LLVM project
This is ignoring that LLVM is first and before all designed as reusable libraries.

Bumping requirements is causing other problems than just being able to build with the just built compiler: namely deployment and integration in other products. For example, you may be forced to isolate components in dynamic library that talks with each others through C APIs and each statically link the C++ standard library. It can become fairly complicated when this does not fit your setup and you get forced into this because LLVM decided to bump their standard.

Joky··on F18 Fortran compiler and runtime project is accepted as part of LLVM project
LLVM supports old platforms: some CI bots are targeting stable Debian, or Ubuntu 14.04.

LLVM is currently moving to C++14, which is a pain for some users like TensorFlow for instance which supports Ubuntu 14.04: the c++ standard library is too old. There are technical solutions for everything but for a component like LLVM, not pressuring its users into convoluted workflow is important.

Of course the LLVM developpers would love to just use the latest version of the standard all the time. For the library part of the standard, it is worked around by having its own: https://github.com/llvm/llvm-project/blob/master/llvm/includ... (and others in the same directory)

Joky··on Think Like a Martian About Money and Universal Basic Income
Actually I'm not against UBI, but I don't think that the martians argument about the money system was a good one: the current system is just much civilized than anything that existed before as far as I can tell.

UBI may be part of the next evolutionary step of society, we'll see, but I'm slightly bothered that it relies on the comfort and the wealthiness of the "first world" built to the detriment of the "third world". I personally feel that "people should be allowed to live" would be more convincing to me if it was about worldwide inequality rather than UBI in westerner society.

← PreviousPage 2 of 14Next →