Just got my new system up and running. And it all very nice.
I have put together a machine in a micro ATX box using the GA-A75M-UD2H motherboard with A8-3850 APU (CPU and ATI HD6550D GPU) and 8 Gb memory and all for around 450£.
I have installed a 64 bit system using ~AMD64 and Gnome 3.2.1. Thought I'd change my application policy and keep the unstable AMD64 branch. Instead I am adding a few masks to stop a few applications being upgraded without me actually choosing it (GCC, python, kernel).
Not sure about GCC. I set the cflags to march=native. I gues it shouldn't be too bad. Far as i know this CPU is more or less a phenom quad core with the added graphics on chip.
One of the things I might want to have a look I guess. Trying to get the onboard sound to work now.
im not sure but you can compile one or two programs with march=amdfam10 which is the march for the phenom, "IF" native works correctly it might pick that up, if not you might be running 386 or something, don't quote me im not exactly sure how or if its possible for march=native to work correctly, though it does seem like a sorta special processor, but as long as you know 100% sure that march setting will work its no different than a properly working march native. but you can run i donno something that you compile like mplayer or whatever just to see if the binary would work before you commit to emerging system wide changes.
i use a healthy dose of redundancy incase the -march or something similar gets filtered by an ebuild or programmer or something because one of the flags in -march=amdfam10 broke something not all the optimizations are lost by filtering the flag
“Music is a moral law. It gives a soul to the Universe, wings to the mind, flight to the imagination, a charm to sadness, gaiety and life to everything. It is the essence of order, and leads to all that is good and just and beautiful.”
processor : 3
vendor_id : AuthenticAMD
cpu family : 18
model : 1
model name : AMD A8-3850 APU with Radeon(tm) HD Graphics
stepping : 0
cpu MHz : 2899.824
cache size : 1024 KB
dj, I notice you have the sse etc things in your CFLAGS. I have just put them in the use flags, wouldn't that be the same? Or is your method making sure its used better?
I'll give your cflags a try now.
EDIT: yes that works too. I guess amdfamily10 is the latest for now?
djdunn wrote:i use a healthy dose of redundancy incase the -march or something similar gets filtered by an ebuild or programmer or something because one of the flags in -march=amdfam10 broke something not all the optimizations are lost by filtering the flag
Yeah, and then we developers need to filter out even more.
Congrats! I too swam back to Gentoo; setup was quite a bit different from what I remembered it, but I am surprised how smoothly, quickly, and pain free it was this go.
Thought I'd put today's discovery up here in case other people have come across it.
I have a brand new LG blueray burner with my machine. this one has a SATA connection so I put it in sata port 1 and the hd on port 0. The bios asked to use AHCI mode for the hd, which I said yes to. During installation the blueray worked fine and booted the rescucd (actually a dvd) just fine.
The problem came a few days ago when I wanted to test the burner functionality. Brasero couldn't even find it. So I installed gnomebaker, which did. But it froze.
I found some posts that mentioned problems with burners in AHCI mode, and looking at my motherborard manual, I find that sata port 4 can be configured independently from the others to run IDE mode instead of AHCI. So I moved the drive to port 4 and changed the setting in bios for IDE on this port. Had to rebuild the kernel with the IDE support (ATAIIPX module in this case) too.
I haven't found any good answer to why burners don't like AHCI mode. is it just my drive, or a limitation in the linux kernel?
