Updates! Pedigree-Apps and POSIX stuff.

Over the past couple of months I've been working less on Pedigree as a whole and working on less intensive projects, such as playing video games, or writing small userspace applications. There is nothing that sorts out early burnout symptoms faster than taking a short break - and that means getting away from low-level programming completely! So progress is definitely slowing down, but that doesn't mean things aren't getting done.

As for the "Pedigree Team." Eduard (eddyb) has now started high school and so has a lot less time to contribute to Pedigree. He's still got a fair few ideas flying through his mind though and that's always a good thing. Can't deny it - since he first showed up in our IRC channel his behaviour (not to mention his English) has improved significantly. Just have to hope that doesn't go to his head ;-).

I still consider Pedigree a unique project: proof of what can happen when a group of mostly like-minded individuals come together and use their combined skillset to make something really significant. It's a shame people like James and Joerg have left us, but this is just the way things work. I don't know about Joerg, but James has certainly used Pedigree as a springboard into some pretty amazing work with his Horizon project.

Enough talk about logistics and teamwork though! Let's talk about code.

The first and foremost thing I really want to talk about is the "Pedigree-Apps" repository. This was set up a while back as a place for all of our ported packages to sit, and as a staging point for a master package repository for whichever package manager we ended up using.

History lesson time! We initially started using "pacman", Arch Linux's package manager. It's a nice piece of software, and does the job very well. The dependencies it asks for are not quite as nice (especially libdownload, which is essentially impossible to find) however. It also doesn't lend itself well necessarily to cross-compiling packages.

So recently I put together a new piece of software called "pup" - "Package Updater for Pedigree." It takes a path to a directory containing a Pedigree-style tree of directories that it tarballs and compresses and adds to a package repository somewhere. The databases used on both the server (for synchronising local "available package" lists) and on the client side ("installed packages") are nice and simple sqlite databases that Python has modules to read out of the box. Oh yes, that's right - pup is completely written in Python. Why write pup in Python? The answer is simple: Python is already a Pedigree build dependency. You have to have Python to build Pedigree.

In the future the main Pedigree repository will ship with a copy of pup ready to run, and instead of distributing hundreds of megabytes worth of binary files for a hard disk image, anyone who builds Pedigree will be allowed to choose which packages to add to their hard disk image. This will hopefully improve the size of source snapshots and shallow clones, and also make it easier to customise a hard disk image to a specific testcase.

As for Pedigree-Apps itself, it now has a shiny build system that caters directly to both cross-compiling and automated builds. Package maintainers just provide patches and a set of scripts (for each phase of the build - prebuild, configuration, build, install, and for packages which provide libraries that future package builds might depend on, a script to set up library links for the cross-compiler) and the rest falls into place. Compared to the old system, this is much easier for package maintainers and also significantly reduces the total size of the repository, which is fantastic for those with slower internet connections.

Now that all that is aside, I have a quick few things to add about a few changes that have gone through the POSIX subsystem. And by quick, I mean bullet points! Yay!

• Brand new pthread locks
• I got fairly sick of having every pthread lock function call into the kernel, and it seriously hammered performance. So I spent a few hours rewriting both the spinlock and mutex functions to use atomic operations all the way out in userspace, only calling into the kernel to put a thread to sleep or wake a thread up. The resulting performance boost is significant: a few mutex tests that I ran as a benchmark averaged a completion time very close to that of Linux. Obviously it's hard to truly benchmark such an OS-specific thing, but I think that if a VM is hitting speeds close to that of Linux for a similar construct, I've done pretty well.
• Rework of statfs/statvfs
• Only one of these functions is actually standardised - statvfs. The other is not. So I've removed statfs completely and stuck with statvfs, and so far nothing has complained. I'm yet to actually implement either of these functions, but I finally have a method of enumerating mounted filesystems in the VFS making an implementation of these functions fairly simple.
• Condition variables
• I'm hesitant to put this here because I really don't like the implementation I wrote, but I'll do it anyway. Condition variables of the pthread variety are now implemented. Badly. There is very little in the way of atomicity, and they end up being two mutexes tied together that are locked and unlocked unatomically. It's pretty nasty all up. Eventually I'll just rewrite these functions to use their own specific implementation without building upon the existing mutex implementation, but that will probably only really come when this implementation causes a bug.

All up there's a lot still going on with Pedigree. Most tasks are now becoming longer subprojects - this is just a part of building such a massive piece of software and is to be expected.

We're looking at possibly doing a "milestone" release in a month or two just to get our current code out into the wild (which does well for finding both bugs and feature requests). Stay tuned!

New Core Features

It's been a long time since I posted anything here. Moving to a new city has slowed my development down significantly, but things are getting back to normal again.

Recently there has been a fair bit of activity, and I'll summarise what's been going on in this post. There's a couple new features, some changes to Pedigree itself, and some bug fixes. In-depth discussion of each, where necessary, follows the list.

• ZombieQueue (commit d5314bf7)
  
• Using > 4 GB of RAM now works, with the x86_64 port (commit bb6b954b)
• MemoryPool (commit 64454975)
• UdpLogger (versatile log callbacks) (commit ff5fddba)

ZombieQueue

ZombieQueue solves a problem that has been lying around for quite some time. Things such as signals require that a Process object is deleted as part of it being killed, so it is no longer schedulable. However, this was dealt with originally by a "delete this;" line. Any form of "delete this" can be dangerous when the next line is not a return statement. Even when the next line is a return statement, a reschedule out of the current context and into another, which allocates memory, may overwrite the existing object... making "this" invalid. That said, it did suffice for quite some time.

The idea of ZombieQueue is that you queue these deletions, allowing the thread to be rescheduled (or the call that needs deletion to return) and the object to be deleted at a later time, when it's safe to do so. Eventually ZombieQueue might also have a "delay" parameter, which ensures that it never deletes an object until after a specific time period has passed. This could easily be used for things such as TIME_WAIT handling in the TCP stack.

ZombieQueue has already been implemented into Process, Pipe, and Socket, all of which needed to delete themselves somehow. Now they call into ZombieQueue, which is able to safely delete the object.

x86_64 port using > 4 GB of RAM

A TODO all through the physical memory manager for quite some time has been to handle regions above 4 GB in the x86_64 port. Until this change was made, no ranges of memory above 4 GB were actually accessible. In fact, using the x86_64 port on a system with over 4 GB of memory actually caused anything from a panic to a triple fault and reboot.

This has been rectified now. I'm still not sure it'll handle a full 64-bit physical address space yet, but it will handle a massive amount of memory nonetheless.

MemoryPool

This one is something I'm surprised hasn't been implemented yet. The idea of MemoryPool is to take buffer allocation out of the kernel heap, where it consumes far too much memory - and sometimes even consumed the full size of the heap.

MemoryPool takes a block of memory, outside of the heap, and splits it up into buffers of a specified size. These buffers are distributed to modules and applications where necessary, and returned to the pool when they are no longer needed. Buffers in, for example, the network stack, are around 2048 bytes (rounded up to the next power of 2 from 1500). Rather than have every incoming packet cause an allocation on the kernel heap of, on average, 800 bytes, this moves the processing into a pool of memory dedicated to the job.

A key feature to MemoryPool is its blocking nature. If an allocation cannot be satisfied by the current pool of memory (ie, all buffers are allocated out), it will block until a buffer becomes available. Compared to the heap, this means a constant memory usage is enforced. Effectively, the pool provides a strict contract on its memory consumption: a pool will never resize to fit a new allocation.

This does bring up a minor problem: some code can't block - an interrupt handler for example. For this situation, a non-blocking allocation method is available, which merely returns NULL if no buffers are available.

UdpLogger

It has long been possible to install custom Log callbacks which take log entries and write them to a destination. Callbacks that already exist include the Serial logger, and a callback is used for the loading screen log dump. For testing on real machines, however, neither of these suffice. None of my test machines have a serial port, for example. They are all networked via Ethernet, however.

UdpLogger is able to be installed after the route table is installed, so it's only really useful for debugging if you can actually boot to a shell on the test machine. It dumps all log entries to a given IP address and port. A test machine can then dump all its debugging output to a development machine running netcat.

Here it is, in action:

Log::installCallback has now also been updated to dump all the old log entries to the new callback, so every logging destination should have the same data once they are all installed.

These are just a couple of exciting new features and fixes that have happened in the past couple of weeks. At this stage we have not planned a second release - we're still working through the bugs from Foster and finally getting around to fixing all those TODOs scattered in the code.

At the moment git master is being kept "stable" - that is, it builds and runs to the bash shell (apart from x86_64) - so feel free to grab the latest code and have a play!