lab 30 - wikifs

By

NAME

lab 30 - wikifs for inferno

DESCRIPTION

I ported plan 9's wikifs to inferno. There were various aspects of the application I wanted to learn about--such as, the cache, the transforms from wiki syntax to html, the method for using html templates, and the details of creating the namespace--that it seemed worthwhile to write it in limbo to absorb it all. I also had a hunch I'd be borrowing a lot of the code for my next file system.

The differences between the original and the port are small. The one significant difference is the approach to locking data structures.

The original uses locks for the Map and the Cache memory as well as file locks for accessing disk.

For the port I attempted to avoid the memory locks altogether. One reason being that when the Flush message is received I attempt to kill the proc processing the request, but that might leave stray locks. The file locks get freed when the file references are garbage collected.

To eliminate locks, I take advantage of limbo's garbage collection and try to use immutable types, or pass-by-value semantics.

For example, the Map is stored in a global variable. When the Map is read in from a file a new structure is created locally then assigned to that variable. The map is not changed (except for timestamps) after that and always remains self consistent.

The find routine for the map makes a local copy of the reference to the map and uses that safe knowing that the reference will stay unchanged for the duration of it's use. All memory will be collected if in the mean time a new map has been loaded.

With the cache I use a last-in-wins policy. This saves us the trouble of worrying about locks at the cost of extra reads of the disk file because of data added to the cache but lost because of overwriting the global cache references. Once data is cached is is unchanged (except for timestamps) so no locking is required once a function has a local reference to the cache data.

Here's an example of code from the cache handling. I create a new Wcache object, c, and I want to add it to the cache. The cache is an 

       array of list of ref Wcache

so when I add an item I create a new list and assign it to an array location, overwriting whatever list reference may have been there. It will be garbage collected if no other code references it. Also, the cache needs to be bounded, so I set a max length for every list. I remove items from the local copy of the list until there's room to add the new entry. Here tab is the global array. 

      h := n%Nhash;
      ol := tab[h];
      c := ref Wcache(0, 0, 0, 0, nil, nil, nil, nil);
      # ... c is properly initialized before added to the list
      while(len ol >= Mcache){
          evict := -1;
          t := Sys->Maxint;
          for(l := ol; l != nil; l = tl l){
              if((hd l).use < t){
                  t = (hd l).use;
                  evict = (hd l).n;
              }
          }
          l = nil;
          for(;ol != nil; ol = tl ol)
              if((hd ol).n != evict)
                  l = hd ol :: l;
          ol = l;
      }
      # last in wins!
      tab[h] = c :: ol;

Because limbo lists are read only we don't need to worry that the references in the list change (although the referenced objects might). We must guard against changing the referenced objects, except only trivially for the timestamps, and treat them as read only.

Not having to worry about locks does simplify the code. Enough that I'd look for opportunities to eliminate locks like this style of programming in the future.

FILES

lookup.b testwrite.b wiki.b wiki.m wiki2html.b wiki2text.b wikifs.b wikipost.b

Comments

fgergo said…
how do you get current httpd to serve pages by wikifs?
3:39 PM
Post a Comment

Popular posts from this blog

lab 110 - inferno archive edition

By
I've been occupied recently with archiving my digital media. I've been copying home videos on DV tapes to hard-disk, ripping audio CD's to WAV files, gathering photo collections, and trying to copy documents from Iomega disks, floppies, and my dusty old Acorn RiscPC. The plan is to have a copy of this data to give to each of my children. My Dad recently scanned and sent me all his photographs of me and my siblings growing up; he also included pictures of himself and my Mother when they met in Africa. With technology today each generation can build a digital library of family history to hand on to the next generation. In the past a family album may have been passed on to only one person. The accumulation of digital data still presents problems. It requires discipline to store files that are open and not locked into devices or proprietary formats. With digital preservation in mind I've tried to use file formats recommended for long term archiving. WAV files for audio, D
Read more

lab 107 - midiplay

By
NAME lab 107 - midiplay NOTES Midiplay plays back MIDI files. It uses the synthesizer I described in lab 62 and the MIDI module from lab 73. The command takes only one argument, the path to the midi file. I've included one in the lab to demonstrate. Bind the audio device before using it. % bind -a '#A' /dev % midiplay invent15.mid The synthesizer has 16 note polyphony. It uses three oscillators, one at the pitch, one at half pitch, one at double pitch. There is also a filter, two delays and a vibrato. The sample rate is 8000hz and there is one mono channel (MIDI channel events are ignored). It performs well enough to work on my laptop without JIT turned on. All the synthesizer parameters can only be tweaked from inside the code at the moment. FILES lab - 107
Read more

The Cartesian Theater of AI

By
Image
The Cartesian Theater was an intuition pump proposed by philosopher Daniel Dennett to describe a false theory of consciousness. Inside the mind is a stage watched by a humunculus that observes what passes on the stage, hears the sounds, and controls the rest of the mind by responding to the action of what the smaller version of the self sees. Of course, what makes the smaller self conscious? It is an infinite regression. Our understanding of mind is that there is no such theater. There is no one place in the mind that is the locality of consciousness. I want to offer another thought experiment based on the concept of a Cartesian Theater. Imagine a theater containing a screen with a camera pointing at it, and a speaker with a microphone directed toward the speaker. The camera is pointing at the screen. The mic is pointing at the speaker. However, instead of there being an infinite loop due to a direct connection between the camera and screen, insert an AI in the loop. The imag
Read more