Daniel Ostrow

Ok, so jumping on the trend started by Simon and Diego here is the 'What did ppc and ppc64 do in 2005?' status update.

• The first thing to mention is 2 very successful releases each adding futher support for the machines using the powerpc processor. 2005.0 and 2005.1 were both successful. Additionally 2005.1-r1 fixed some minor issues on PPC64. Thanks got to Pylon, wolf31o2, jforman and the entire PPC, PPC64, RelEng and Infra teams.
• Along with 2005.1 we merged the ppc and ppc64 profiles into one common parent to better match the efforts of sparc and mips which both support similar structures. The merge was mostly stylistic but a further blending will be coming with 2006.0
• Support for the PPC970 processor found in the G5 transitioned over to the ppc64 team. Needing an easy way to transition users we created pure 32-bit, pure 64-bit and multilib userland profiles (the latter thanks in large part to the AMD64 team whos work made it possible). Gentoo is the only distro out there that fully supports all three types of installs. That is huge.
• 2005 saw the first support for Gnome in a pure 64-bit environment on PPC64 as we finally got mozilla to compile. Mozilla and Firefox still don't work but efforts are continuing to make these browsers ppc64 64-bit friendly.
• Hardened support has improved on ppc and ppc64 saw the first hardened profile. Neither one is really ready for prime time, but thanks to the great work of our Hardened team things are getting there.
• We started using ATs on the ppc team, they have been a great help, thanks all of you and thanks to the AMD64 team for coming up with the idea.
• Due to all these improvements, and the continued improvement of the PPC Faq and the Handbook we were able to close just shy of 950 bugs between the two groups.
• Because we have made such a name for ourselves as a strong reliable distro for both ppc and ppc64 we saw huge hardware sponsorship from Genesi and IBM. This relationship I'm sure will continue to grow. Thanks go out to both companies for their continued support.
• While not technically a Gentoo accomplishment work continues on the bcm43xx driver for the AirportExtreme (among others). I would personally like to thank JoseJX and Kugelfang for their contributions in bringing this project to where it is today. There is still work to be done for sure, but hell, I have wireless on my iBook now so I can't complain. I'd also like to thank all those who work on the driver that are not directly part of the Gentoo community, good job guys!
• I'd also like to thank all those that work on the ppc32 and ppc64 kernels as 2005 saw support for quite a bit of new hardware and without them it would not be possible.
• Finally I'll leave off quoting Simon, as it truely is the most important aspect of all: "We had lots of fun".

All told I'd say that's one hell of a year, here is to another great year for Gentoo, the PPC architecture and OpenSource as a whole.

Yeah, even though I'm on vacation I just had to jump on the band wagon. Damn peer preasure........

Just a heads up, I'm working to bring the Gentoo hardened profile to a ppc64 near you. A big thanks to solar for putting in the time to help me with this. I now return you to your regularly scheduled programing.
Some preliminary PaXtest data (no toolchain or noexec/pageexec yet):

Mode: blackhat
Linux Strife64 2.6.11-hardened-r1 #4 SMP Wed Mar 16 21:08:23 EST 2005 ppc64 PPC970, altivec supported PowerMac7,2 GNU/Linux

Executable anonymous mapping : Killed
Executable bss : Killed
Executable data : Killed
Executable heap : Killed
Executable stack : Killed
Executable anonymous mapping (mprotect) : Killed
Executable bss (mprotect) : Killed
Executable data (mprotect) : Killed
Executable heap (mprotect) : Killed
Executable stack (mprotect) : Killed
Executable shared library bss (mprotect) : Killed
Executable shared library data (mprotect): Killed
Writable text segments : Vulnerable
Anonymous mapping randomisation test : 24 bits (guessed)
Heap randomisation test (ET_EXEC) : 14 bits (guessed)
Heap randomisation test (ET_DYN) : 32 bits (guessed)
Main executable randomisation (ET_EXEC) : 20 bits (guessed)
Main executable randomisation (ET_DYN) : No randomisation
Shared library randomisation test : 24 bits (guessed)
Stack randomisation test (SEGMEXEC) : 32 bits (guessed)
Stack randomisation test (PAGEEXEC) : 32 bits (guessed)
Return to function (strcpy) : paxtest: bad luck, try different compiler options.
Return to function (memcpy) : Killed
Return to function (strcpy, RANDEXEC) : paxtest: bad luck, try different compiler options.
Return to function (memcpy, RANDEXEC) : Killed
Executable shared library bss : Killed
Executable shared library data : Killed

Looks like dual core G5s aren't that far off, if you take the update to MONster to be any indication. If you all remember last year the 970FX definition showed up all of 3 months before the machines hit the shelves. Apple has a tendency to only do major product releases three times a year, Mac World Expo in San Francisco, WWDC and Mac World Expo Paris. If the past is any indication of future results it looks like they are trying to push for production machines by WWDC in June. With the recent updates to the ppc64 kernel, and new fun stuff like AGP and iMac-G5 patches coming down the pike it looks like ppc64 is going to grow fast from here on out. Now if I could only get multilib working...

10 PRINT Hello_World
20 BEEP
30 GOTO 10
Ah gotta love Apple Basic.

A little story for introduction:

At the edge of the Architecture map the intrepid programmer found the words "Here there be PowerPCs". Having no fear of these mysterious processors he set his sails to catch the wind and found that indeed the world was not flat. What he found over the horizon was a land where code was no longer bound by the tyranny of x86, a veritable paradise. The programmer set up shop and hung a sign outside his door; "PowerPC to the People" it read. As people slowly realized there was another way they broke free from their shackles and came to the new land. Welcome the programmer said, stay a while.