Messages

301

Operating Systems and Porting / Re: kernel config: page size 4k vs 64k

« on: September 16, 2020, 07:02:22 pm »

Maybe that is why Thunderbird is not working at all either.

These were pretty low-level changes, so I don't know if they're what's blocking Thunderbird. But I run and test Firefox all the time. Perhaps, if there's interest, I can provide my binaries (I already provide my .mozconfigs).

302

Operating Systems and Porting / Re: kernel config: page size 4k vs 64k

« on: September 14, 2020, 08:47:48 pm »

Some emulation and virtualization tools have dependencies on page size. We had some page size issues in Firefox initially, though I think we smoked most of those out.

I'd certainly prefer a 4K page size as it's more "typical" but given RHEL's prominence on HPC POWER I can understand why Fedora ppc64le defaults to 64K as well. I certainly wouldn't expect them to do two flavours, especially since they don't even support big-endian ppc64 anymore, which I think would be more significant architecturally than page size.

303

Operating Systems and Porting / Re: testing a new machine, memory test, etc

« on: September 08, 2020, 11:05:57 pm »

Seems like a statically-built testing binary you could indeed run from the Petitboot shell would be feasible.

304

Applications and Porting / Re: Wine patch to support PPC64

« on: August 31, 2020, 10:53:11 pm »

Exactly. You'd need something like Hangover to do that.

305

Blackbird / Re: Fedora 32 Linux kernel 5.7.x crashes

« on: August 28, 2020, 11:14:42 am »

No problem with my WX7100 (Polaris), so it may be specific to Navi cards.

306

Applications and Porting / Re: Raspberry Pie

« on: August 20, 2020, 11:03:00 pm »

Seems like most of the components are open source, though.

307

Firmware / Re: Basic understanding of OPAL

« on: August 18, 2020, 06:37:53 pm »

Skiboot is not a Linux kernel. Skiboot's most important component is OPAL. It is also responsible for setting up the PCIe Host Bridges (PHBs) and certain other kinds of midlevel initialization. Once this is done, it then starts a kernel stored in the BOOTKERNEL firmware partition, which is usually Skiroot/Petitboot.

308

Firmware / Re: Basic understanding of OPAL

« on: August 17, 2020, 12:29:37 pm »

That's not really its purpose; its purpose is to make implementation differences between systems appear unified to an OS running at a higher level. Things like USB and hardware attacks are probably best approached from the OS itself by simply disabling the conduits for those attacks, which certainly could be done through OPAL calls and thus will work on any class of OpenPOWER machine. But the role it serves is far more general than that.

309

Operating Systems and Porting / Re: Linux 5.8 and nx-gzip

« on: August 16, 2020, 04:08:32 pm »

Well, how many resources is it reporting (instrumenting that assertion)?

310

Firmware / Re: Basic understanding of OPAL

« on: August 16, 2020, 04:02:27 pm »

I guess you could think of it as a sort of hypervisor, but what it's actually doing is serving as an abstraction layer (hence the acronym) to give a common entry point for doing necessary low-level services. OPAL is part of Skiboot, so as you update your firmware, OPAL is updated as well.

The full list of OPAL calls by number defines its services. It largely concerns itself with low-level functions like PCI devices and interrupts. Here's one such list maintained in FreeBSD: http://fxr.watson.org/fxr/source/powerpc/powernv/opal.h

311

GPU Compute / Accelerators / Re: Which GPU would *you* choose?

« on: August 11, 2020, 03:16:41 pm »

AMD Polaris, largely because of familiarity, and it seems to have the best track record right now. Doing just fine on the Raptor BTO WX7100, so either that or the RX equivalent should be spiffy.

312

GPU Compute / Accelerators / Re: the microcode question

« on: August 11, 2020, 03:16:03 pm »

Wow, libreSOC seems to have gotten along further than I thought. Worth keeping an eye on.

313

GPU Compute / Accelerators / Re: the microcode question

« on: August 10, 2020, 11:16:59 am »

Maury, that doesn't help him because that still requires an AMD blob.

I'm not aware of a fully free 4K framebuffer ... yet. I'm told some people were putting RISC-V to such a task but I haven't heard much further and it might just be wishful thinking.

314

General CPU Discussion / Re: POWER9 water cooling kit

« on: August 10, 2020, 11:15:46 am »

Outstanding!

315

Operating Systems and Porting / Re: [NEWS] Linux kernel 5.8 is out!

« on: August 03, 2020, 08:37:30 pm »

Does that mean an Axone system in the works, potentially?