I would appreciate if everyone stepped back on this.
1. repeatedly reading constant data and shoving it into the kernel
software RNG is safe. Such values do not saturate or decrease the
quality.
2. The kernel software RNG does not require a hardware RNG to be
available.
3. The people
I wanted to see how this would behave on my APUs. One gave 000,
another consistently. But then I had this diff in my tree
and I rebuilt for my workstation. That has
cpu0: AMD Ryzen 9 7950X 16-Core Processor, 4500.00 MHz, 19-61-02
...
ccp0 at pci21 dev 0 function 2 "AMD 17h/90h
On Apr 23 23:17:10, h...@stare.cz wrote:
> On Apr 23 21:00:35, na...@mips.inka.de wrote:
> > > That was in 2022. Lots of people will have machines without new BIOS.
> > I have the latest firmware and the ccp(4) RNG returns nothing but 0.
>
> With your diff, my APU2d's and APU2e's report all
On Apr 23 21:00:35, na...@mips.inka.de wrote:
> > That was in 2022. Lots of people will have machines without new BIOS.
> I have the latest firmware and the ccp(4) RNG returns nothing but 0.
With your diff, my APU2d's and APU2e's report all zeroes
with each of bios v4.11.0.5, v4.17.0.1,
Theo de Raadt:
> That was in 2022. Lots of people will have machines without new BIOS.
I have the latest firmware and the ccp(4) RNG returns nothing but 0.
> I wonder if our kernel should have similar code to enable the registers.
I tried that yesterday to no effect... but I'm not certain
That was in 2022. Lots of people will have machines without new BIOS.
I wonder if our kernel should have similar code to enable the registers.
Jan Stary wrote:
> On Apr 21 17:27:37, dera...@openbsd.org wrote:
> > Christian Weisgerber wrote:
> >
> > > Christian Weisgerber:
> > >
> > > > I
On Apr 21 17:27:37, dera...@openbsd.org wrote:
> Christian Weisgerber wrote:
>
> > Christian Weisgerber:
> >
> > > I built a kernel with an instrumented driver. Unfortunately, no
> > > entropy is provided:
> >
> > FWIW, it appears to work on the SoftIron OverDrive 1000:
> >
> > ccp: rng
Christian Weisgerber wrote:
> Christian Weisgerber:
>
> > I built a kernel with an instrumented driver. Unfortunately, no
> > entropy is provided:
>
> FWIW, it appears to work on the SoftIron OverDrive 1000:
>
> ccp: rng 058f9dad
> ccp: rng f0a495ba
> ccp: rng a757bdf7
> ccp: rng 31b21d19
>
Christian Weisgerber:
> I built a kernel with an instrumented driver. Unfortunately, no
> entropy is provided:
FWIW, it appears to work on the SoftIron OverDrive 1000:
ccp: rng 058f9dad
ccp: rng f0a495ba
ccp: rng a757bdf7
ccp: rng 31b21d19
ccp: rng d1ce1c78
ccp: rng 863c9199
--
Christian
"Theo de Raadt" writes:
> Maybe the driver is broken.
>
> Maybe it fails to initialize it. Maybe in other cases, the BIOS initializes
> it.
>
> So maybe on this machine, it is broken, but on other machines it is
> not broken.
Indeed, it works e.g. on this machine.
ccp0 at pci9 dev 0 function
Maybe the driver is broken.
Maybe it fails to initialize it. Maybe in other cases, the BIOS initializes it.
So maybe on this machine, it is broken, but on other machines it is not broken.
Pushing 0's to the random subsystem doesn't make the random state worse.
It just fails to make it better.
Christian Weisgerber:
> ccp(4) attaches, so presumably it is used as a source of entropy.
> Whether the hardware actually provides random output, I don't know.
I built a kernel with an instrumented driver. Unfortunately, no
entropy is provided:
ccp: rng
ccp: rng
ccp: rng
Jan Stary:
> Does OpenBSD use any hardware RNG on the PC Engines APUs?
ccp0 at pci0 dev 8 function 0 "AMD 16h Crypto" rev 0x00
ccp(4) attaches, so presumably it is used as a source of entropy.
Whether the hardware actually provides random output, I don't know.
--
Christian "naddy" Weisgerber
Reading random(4),
System activity (such as disk, network, and clock device interrupts),
and hardware random generator output is collected, ...
Does OpenBSD use any hardware RNG on the PC Engines APUs?
https://github.com/pcengines/apu2-documentation/issues/112
discusses the firmware support
14 matches
Mail list logo