Category Archives: Computer Science

Things I’ve Learned This Week (March 30 – April 3, 2015)

This is my second post in a weekly series, where I attempt to distill my week down into some lessons or facts that I’ve picked up. Let’s get to it!

ES6 – what’s safe to use in browser development?

As of March 27, 2015, ES6 classes are still not yet safe for use in production browser code. There’s code to support them in Firefox, but they’re Nightly-only behind a build-time pref.

Array.prototype.includes and ArrayBuffer.transfer are also Nightly only at this time.

However, any of the rest of the ES6 Harmony work currently implemented by Nightly is fair-game for use, according to jorendorff. The JS team is also working on a Wiki page to tell us Firefox developers what ES6 stuff is safe for use and what is not.

Getting a profile from a hung process

Update [January 23rd, 2017]: Markus Stange contacted me to let me know that the function mozilla_sampler_save_profile_to_file has been changed to profiler_save_profile_to_file. This happened in this bug. I’ve updated the post to reflect this.

According to mstange, it is possible to get profiles from hung Firefox processes using lldb1.

  1. After the process has hung, attach lldb.
  2. Type in2, :
    p (void)profiler_save_profile_to_file("somepath/profile.txt")
  3. Clone mstange’s handy profile analysis repository.
  4. Run:
    python symbolicate_profile.py somepath/profile.txt

    To graft symbols into the profile. mstange’s scripts do some fairly clever things to get those symbols – if your Firefox was built by Mozilla, then it will retrieve the symbols from the Mozilla symbol server. If you built Firefox yourself, it will attempt to use some cleverness3 to grab the symbols from your binary.

    Your profile will now, hopefully, be updated with symbols.

    Then, load up Cleopatra, and upload the profile.

    I haven’t yet had the opportunity to try this, but I hope to next week. I’d be eager to hear people’s experience giving this a go – it might be a great tool in determining what’s going on in Firefox when it’s hung4!

Parameter vs. Argument

I noticed that when I talked about “things that I passed to functions5”, I would use “arguments” and “parameters” interchangeably. I recently learned that there is more to those terms than I had originally thought.

According to this MSDN article, an argument is what is passed in to a function by a caller. To the function, it has received parameters. It’s like two sides of a coin. Or, as the article puts it, like cars and parking spaces:

You can think of the parameter as a parking space and the argument as an automobile. Just as different automobiles can park in a parking space at different times, the calling code can pass a different argument to the same parameter every time that it calls the procedure.6

Not that it really makes much difference, but I like knowing the details.


  1. Unfortunately, this technique will not work for Windows. 🙁  

  2. Assuming you’re running a build after this revision landed. 

  3. A binary called dump_syms_mac in mstange’s toolkit, and nm on Linux 

  4. I’m particularly interested in knowing if we can get Javascript stacks via this technique – I can see that being particularly useful with hung content processes. 

  5. Or methods. 

  6. Source 

The Joy of Coding (Ep. 8): View Source Hacking

In this episode, I again started with some code review. I reviewed this patch for this bug by fellow Firefox hacker Gijs, and refreshed my memory on var hoisting. I’ve been using let for so long that it was really, really weird to see how var worked.

After that, I quickly gave an update on my plugin crash UI bug I had been working on the last episode – the patches are up, and are currently undergoing review, so there wasn’t much to do there.

Next, I started on a brand new bug1, explained the bug2, and then laid out my plan for attacking it.

Specifically, I’m going to try an experiment: I will only be working on that bug during Joy of Coding sessions. That way, there is continuity from video to video, and you won’t miss any of the development that goes on between episodes.

We sliced off a chunk to get done, and hit some minor roadblocks (as expected). The View Source code is old and crufty, and I have to do my best to make sure I don’t break any of the other applications that depend on it (like Thunderbird and SeaMonkey).

So that was the name of the game – looking to see how other applications use View Source, and trying to come up with a plan for making sure we don’t break them, while at the same time refactoring View Source to be easier to code against (and work with a frame script and messages).

It was a long slog3, but we got to a good point by the end. Let’s see how far we get next week!

Episode Agenda

References

Bug 1148807 – Method moveToAlertPosition in dialog.xml should check if opener is not null

Bug 1110887 – With e10s, plugin crash submit UI is brokenNotes

Bug 1025146 – [e10s] Never load the source off of the network when viewing sourceNotes


  1. I say brand new, except that, as I explain in the video, I had already attacked this bug early on in my e10s work, and had only recently come back to it. 

  2. The View Source tool sometimes re-retrieves the source off of the network when opened from an e10s-browser 

  3. My longest episode ever, clocking in at over 2.5 hours. 

The Joy of Coding (Episode 7): Code review, and a Regression

In this episode, I started with some code review. I was reviewing a patch to make the Findbar (particularly, the Find As You Type feature) e10s-friendly.

With that review out of the way, I had to swap a bunch of information about the plugin crash UI for e10s in my head – and in particular, some non-determinism that we have to handle. I explained that stuff (and hopefully didn’t spend too much time on it).

Then, I showed how far I’d gotten with the plugin crash UI for e10s. I was able to submit a crash report, but I found I wasn’t able to type into the comment text area.

After a while, I noticed that I couldn’t type into the comment text area on Nightly, even without my patch. And then I reproduced it in Aurora. And then in Beta. Luckily, I couldn’t reproduce it in Release – but with Beta transitioning to Release in only a few days, I didn’t have a lot of time to get a bug on file to shine some light on it.

Luckily, our brilliant Steven Michaud was on the case, and has just landed a patch to fix this. Talk about fast work!

Episode Agenda

References:
Bug 1133981 – [e10s] Stop sending unsafe CPOWs after the findbar has been closed in a remote browser

Bug 1110887 – With e10s, plugin crash submit UI is brokenNotes

Bug 1147521 – Cannot type into comment area of plugin crash UI

The Joy of Coding (Episode 6): Plugins!

In this episode, I took the feedback of my audience, and did a bit of code review, but also a little bit of work on a bug. Specifically, I was figuring out the relationship between NPAPI plugins and Gecko Media Plugins, and how to crash the latter type (which is necessary for me in order to work on the crash report submission UI).

A minor goof – for the first few minutes, I forgot to switch my camera to my desktop, so you get prolonged exposure to my mug as I figure out how I’m going to review a patch. I eventually figured it out though. Phew!

Episode Agenda

References:
Bug 1134222 – [e10s] “Save Link As…”/”Bookmark This Link” in remote browser causes unsafe CPOW usage warning

Bug 1110887 – With e10s, plugin crash submit UI is brokenNotes

The Joy of Coding (Episode 5): Much Code Review

In this fifth episode, I didn’t work on any bugs. Instead, I did a bunch of code review. Ever wanted to know what a Firefox engineer does to review a patch? If so, then this episode is for you!

Episode Agenda

References:
Bug 1140898 – [e10s] “View” > “Switch Page Direction” doesn’t work in e10s

Bug 1140878 – [e10s] “Switch Page Direction” in remote browser causes unsafe CPOW usage warnings

Bug 1066531 – [e10s] Switching tabs can result in old content being displayed for a split second after the tab bar is updated