Disqus - Latest Comments for tmm1

Re: Creating a Java VM from Android Native Code

tmm1 — Fri, 12 May 2017 15:05:57 -0000

Great writeup! Unfortunately it appears that as of Android M, the WithFramework class has been removed: https://android-review.goog...

I assume you were running this on an older Android device?

There's some similar discussions on http://stackoverflow.com/qu...

Re: an obscure kernel feature to get more info about dying processes

tmm1 — Wed, 02 May 2012 20:40:26 -0000

Note that this technique often results in a truncated coredump, due to a known upstream kernel bug. See https://bugzilla.redhat.com...

Re: Garbage Collection Slides from LA Ruby Conference

tmm1 — Tue, 23 Feb 2010 23:46:56 -0000

The patch is for 1.9, written by @nari3 and is available at http://www.narihiro.info/in...

Re: Garbage Collection Slides from LA Ruby Conference

tmm1 — Tue, 23 Feb 2010 23:45:22 -0000

our first hack at a rails3 patch is at http://gist.github.com/308003, but it needs some work before it can be committed back to rails3

Re: Garbage Collection Slides from LA Ruby Conference

tmm1 — Tue, 23 Feb 2010 23:44:10 -0000

PDF: http://dl.dropbox.com/u/635...

Re: Ruby statement modifers behave differently than conditional statements - Ryan Angilly's Blog

tmm1 — Mon, 28 Dec 2009 13:57:36 -0000

This has nothing to do with statement modifiers vs conditionals. Consider the following:

>> if false
>> a = 1
>> end

>> defined? a
=> "local-variable"

>> a
=> nil

The ruby parser sees 'a=1' and defines 'a' as a local variable, even though it is impossible for that code to be run.

Re: memprof: A Ruby level memory profiler

tmm1 — Wed, 16 Dec 2009 15:55:07 -0000

The Memprof.stats API is designed to replace bleak_house. Additional features are being developed- for instance, the Memprof.dump API described at http://timetobleed.com/what... will print out all objects in detail, including the number of bytes of character data associated with all strings.

Re: What is a ruby object? (introducing Memprof.dump)

tmm1 — Tue, 15 Dec 2009 16:15:34 -0000

I plan on adding a Memprof.dump_all which will produce a full heap dump which can be analyzed to calculate the size of any given object. It's tricky because the size of an Object has to take into account all the instance variables and singleton methods attached to it, etc.

Re: What is a ruby object? (introducing Memprof.dump)

tmm1 — Mon, 14 Dec 2009 20:50:42 -0000

Not really sure what CREFs are yet =)

Re: Extending ltrace to make your Ruby/Python/Perl/PHP apps faster

tmm1 — Mon, 12 Oct 2009 14:58:12 -0000

Also ltrace can trace system calls, so you can use it instead of strace.

Re: Extending ltrace to make your Ruby/Python/Perl/PHP apps faster

tmm1 — Mon, 12 Oct 2009 14:57:50 -0000

Using strace to trace mysql traffic is painful. You have to know which file descriptors to trace (check via lsof first), then trace all network traffic on those fds in verbose mode to see the queries themselves (doesn't work if the mysql protocol has compression enabled), and then try to figure out how long each of the queries took to execute by matching queries and responses.

ltraceing mysql_send_query() gives you all this information in one line.

Re: Extending ltrace to make your Ruby/Python/Perl/PHP apps faster

tmm1 — Sun, 11 Oct 2009 03:33:09 -0000

strace traces system calls. ltrace traces library calls.

Re: nirvdrum's weblog :: Lessons Learned in Large Computations with Ruby

tmm1 — Thu, 17 Sep 2009 19:12:18 -0000

FYI I uploaded a new perftools.rb 0.2.5 gem yesterday that fixes a lot of the segfault issues. If you get a chance, give it a try and let me know if you still have problems.

Re: Beware Of EventMachine Periodic TImers

tmm1 — Fri, 21 Aug 2009 18:49:46 -0000

Ah. Ideally you would use something like em-http-request to make those calls, so they're scheduled back into the reactor and there's no need for EM.defer or threads.

Re: Beware Of EventMachine Periodic TImers

tmm1 — Fri, 21 Aug 2009 16:06:53 -0000

What was your longer running task in this case?

Re: Superfeedr Blog : Performance monitoring with collectd

tmm1 — Sat, 08 Aug 2009 07:20:18 -0000

This is awsome!

Re: git ready » smartly save stashes

tmm1 — Thu, 26 Mar 2009 15:23:05 -0000

You can also git add -p the changes you don't want into your index, and then git stash save --keep-index to ignore those.