Blogs

Haiku Down Under 2012 Report

Blog post by Sikosis on Sun, 2012-08-26 06:41

If anything could go wrong -- it did go wrong at the Fifth Annual Haiku Down Under Conference for Users and Developers, held at The University of Queensland, Brisbane, Australia on 19th August, 2012. At least, that's how it initially played out.

GP South Building at The University of Queensland

A brief summary of HAIKU's package management.

Blog post by mmadia on Mon, 2012-08-20 21:26

With the announcement of Ingo and Oliver's contracts for package management, it is worthwhile to revisit how package management will function. When reading, keep in mind that this explanation will be condensed, simplified, and partially incomplete. Nonetheless, it will provide a general overview on how things will work.

NFSv4 client: three quarter term report

Blog post by Paweł Dziepak on Mon, 2012-08-06 22:09

I've recently been working on caching in NFSv4 client. It was essential in order to allow the client to be comfortably used. I can gladly say that the traffic generated by NFS client has been greatly reduced, thanks to metadata, directory, lookup and file caching. I've also implemented support for open delegations which, though not always available, allow the client to perform virtually all file operations without immediate server participation.

BFS Partition Resizer: Three-quarter-term Report

Blog post by ahenriksson on Mon, 2012-08-06 20:31

For this period, I have been working on getting resizing to work from within Haiku, rather than just in bfs_shell. In its current state, the code works, sometimes, if you don't stress it too much and write data to the partition while resizing. On the bright side, recovery from various errors is working well :). In terms of functionality, the only thing missing is the ability to grow full or almost full file system. The problem with this is that we need to grow the bitmap that tracks allocated blocks before we can actually make use of the new blocks. This can be overcome with a little slyness, but it's a bit of work, and adds some complexity.

I think it would be better to focus on getting the current code into a finished state. So my plan for the remaining time is to try to exterminate the remaining bugs, and polish the code. I'd also like to rebase my work into a neat patch set. The chances of finding time to add resize support to DriveSetup seem fairly slim at the moment. But since the task is independent of my other work, it might be something to start on if I run out of things to do in the last few days.

On a different note, I will be travelling away for two weeks at the end of the coding period (on the 18th), returning the September 2nd. Last time I went to an internet café my email account was hijacked, so I'll probably be accessing the internet sparingly. I hope this won't cause any trouble.

OpenJDK port: three quarter term report

Blog post by hamish on Sun, 2012-08-05 22:03

Since midterm I have been working on the jsound port, which provides audio, MIDI input/output and the ability to control mixer volume and other parameters.

After getting my head around some of the media kit concepts the implementation has gone smoothly. I implemented audio output support first, as I guessed this would be the most used component. It works well. Then I implemented MIDI input and output support. This is untested so far because I don't have any MIDI hardware. In the end I will probably end up constructing some dummy MIDI endpoints in another app for rudimentary testing. Audio input support is awaiting the availability of the SoundConsumer class, which might be included as a private API in libmedia, as the file cannot be included in OpenJDK because of licensing restrictions. Once this is in place I'll get working on the audio input support.

After that I'll implement support for "ports", which gives Java apps the ability to configure parameters on the system audio mixer. I should be finished in time for the suggested pencils-down date, the 13th of August. I didn't have time to get a new build ready for this blog post, but it's not very difficult to build from source. Build instructions are here: http://pastebin.com/4FhDAnyg

cpuidle: three quarter term report

Blog post by yongcong on Sun, 2012-08-05 11:34

I began to implement the acpi cpuidle driver so that the power saving feature can benefit all x86 platforms(In theory although). The acpi is more complicated than I thought. The main time is spent on implementing "_CST" evaluation and decoding.

First of all, to evaluate any acpi object/method needs acpi handle. Since haiku doesn't export AcpiWalkSpace method of acpica, so after system booting, I can get the acpi processor handle. the only solution is using the device manager so that the acpi cpuidle driver can be loaded during boot. This requires cpuidle modification so that generic cpuidle module is loaded by low level idle driver. The modification is not done because it's simple and I want to get "_CST" evaluation done firstly.

To evaluate "_CST", we need to evaluate "_PDC" or "_OSC" firstly. While "_OSC" is preferred than "_PDC" from ACPI 3.0. So my implementation evaluate "_OSC", fallback to "_PDC" if failed.

The "_CST" decoding is simpler than "_CST" evaluation. Some code is ready during gsoc bonding period. I just make it finished.

Then I came across one big problem which block me for one week--under haiku, the acpi processor is enumerated in "cpu8=>cpu7...=>cpu1" rather than "cpu1=>cpu2=>...>cpu8". The later order is taken by Linux, FreeBSD, NetBSD, etc. To evaluate "_CST" of cpu8 needs "_CST" of CPU1, so haiku's enumeration make the acpica reports AE_NOT_FOUND. But I dunno such requirement before, I even dig into acpica code but found nothing. After frustrated for about one week, I realized that the problem may exists in the evaluation order so I dumped the according processor id every acpi processor and found the reason. In no more than ten minutes, I implemented one workaround and it works!! Here is the CSTATES information dumped from my laptop

KERN: evaluate _CST @0x821e7698
KERN: cpuidle found 2 cstates
KERN: c1
KERN: Latency: 1
KERN: power: 1000
KERN: bufer length 17
KERN: cpu_reg size: 15
KERN: FFH method
KERN: c3
KERN: Latency: 104
KERN: power: 350
KERN: bufer length 17
KERN: cpu_reg size: 15
KERN: IO method

NOTE: Here, Cx is acpi reported Cx rather than OS's cx. For example, acpi's C2 is missing if AC is plugged. But Linux take ACPI C3 as the OS C2. I would like the take similar solution to make the generic cpuidle module happy

x86_64 port: three quarter term report

Blog post by xyzzy on Sat, 2012-08-04 10:50

I have continued to make good progress since my midterm report. All the kernel functionality except for user debugging is implemented, and I have ported a basic set of drivers, including PCI, disk drivers, BFS and PS/2 input. For most drivers, porting is just a matter of fixing compiler warnings. For some, there are 64-bit issues which make porting more difficult. For example, the USB stack will require a bit more work as it currently assumes that addr_t is 32-bit everywhere.

I have also made some progress in porting userland to x86_64. I currently have libroot, libbe, bash, and most of the command line utilities ported. I have got an interactive bash shell running on top of consoled (which is usually used to run gdb on if app_server crashes).

Syndicate content