00:24:33
sech1:
Add a kernel call (possibly two kernel calls - one for aesenc, one for aesdec) to that latency. It can't be efficient enough for the main loop AES.
01:35:27
DataHoarder:
I wonder if they do that AES for single rounds or do they have specific n round calls?
01:44:37
DataHoarder:
on the coprocessor
07:26:46
sech1:
They must have N round calls, or even some sort of programming language for this accelerator
07:26:59
sech1:
Something like OpenCL, but for crypto primitives
07:27:13
sech1:
Someone needs to get their hands on SDK :)
07:30:27
DataHoarder:
https://github.com/sophgo/linux-riscv/tree/sg2044-dev-6.12
07:30:43
DataHoarder:
they have a couple of active branches here
07:32:44
DataHoarder:
https://github.com/sophgo/linux-riscv/tree/d74a073f36ecc302399bb0e5bf5b8b807400b798/drivers/soc/sophgo
07:41:14
DataHoarder:
forked from above too https://github.com/revyos/sg2044-vendor-kernel
07:41:56
DataHoarder:
https://github.com/revyos/sg2044-vendor-kernel/blob/35403068220226f17c364aaa45565e428a38c0c2/arch/riscv/boot/dts/sophgo/sg2044-cpus.dtsi
07:42:12
DataHoarder:
this has isa extension list
07:45:08
DataHoarder:
/say // pka device, change size from 0x1000 to 0x6000
07:45:08
DataHoarder:
pka: pka@7040040000 {
07:45:08
DataHoarder:
compatible = "snps,designware-pka";
07:46:04
DataHoarder:
actual list upstream on their repo https://github.com/sophgo/linux-riscv/blob/sg2044-dev-6.12/arch/riscv/boot/dts/sophgo/sg2044-cpus.dtsi
07:46:09
DataHoarder:
that folder has more definitions
07:51:43
DataHoarder:
seems that just accelerates ECC/EdSA/RSA
07:53:57
DataHoarder:
it's probably another desingware IP if I had to guess
07:55:07
DataHoarder:
enjoy digging
08:05:30
sech1:
I found all kinds of aes, even implementations for different architectures, but not their SPACC
08:08:41
sech1:
#define SG2044_RST_SPACC 188
08:08:48
sech1:
the only mention I found
08:27:00
DataHoarder:
indeed
20:01:52
gingeropolous:
well i got a benchmarking script working for windows in powershell, except for the msr mod. requires manual stuff. maybe we'll get more #s if windows users can bench
20:09:16
sech1:
They can bench with xmrig (dev branch)
20:09:48
sech1:
But it's hard to automate because xmrig doesn't auto-exit after benchmark finishes