Jul 19, 2018 1:08:06 GMT -8

Reply Quote | Like

	daedalus Newbie	Right now I'm Offline I have made 2 posts I joined July 2018		I currently have no messenger contact information filled out.	I currently have no social network information filled out.

Jul 18, 2018 14:48:23 GMT -8 tcrider84 said:

Managed to get it resolved! Sure enough kernel needed to be recompiled with CONFIG_CRYPTO_DEV_SP_PSP=n and I also changed these:

CONFIG_KVM_AMD=m to CONFIG_KVM_AMD=y just in case. Probably overkill but it's for sure working again.

Good that you got it working

Jul 17 23:36:16 zenith kernel: [  131.619989] sysrq: SysRq : Show Blocked State
Jul 17 23:36:16 zenith kernel: [  131.620016]   task                        PC stack   pid father
Jul 17 23:36:16 zenith kernel: [  131.620086] systemd-udevd   D12624  7505   6593 0x80000120
Jul 17 23:36:16 zenith kernel: [  131.620121] Call Trace:
Jul 17 23:36:16 zenith kernel: [  131.620143]  ? __schedule+0x35b/0x730
Jul 17 23:36:16 zenith kernel: [  131.620169]  schedule+0x37/0x90
Jul 17 23:36:16 zenith kernel: [  131.620195]  __sev_do_cmd_locked+0xb3/0x130 [ccp]
Jul 17 23:36:16 zenith kernel: [  131.620233]  ? wait_woken+0x80/0x80
Jul 17 23:36:16 zenith kernel: [  131.620260]  __sev_platform_init_locked+0x26/0x40 [ccp]
Jul 17 23:36:16 zenith kernel: [  131.620296]  sev_platform_init+0x18/0x30 [ccp]
Jul 17 23:36:16 zenith kernel: [  131.620328]  psp_pci_init+0x2b/0xb0 [ccp]
Jul 17 23:36:16 zenith kernel: [  131.620356]  ? driver_register+0x7b/0xc0
Jul 17 23:36:16 zenith kernel: [  131.620387]  sp_mod_init+0x11/0x1000 [ccp]
Jul 17 23:36:16 zenith kernel: [  131.620413]  ? 0xffffffffc09b6000
Jul 17 23:36:16 zenith kernel: [  131.620435]  do_one_initcall+0x35/0x1c0
Jul 17 23:36:16 zenith kernel: [  131.620461]  ? free_unref_page_commit.isra.90+0x5f/0xc0
Jul 17 23:36:16 zenith kernel: [  131.620494]  ? kmem_cache_alloc_trace+0x167/0x1d0
Jul 17 23:36:16 zenith kernel: [  131.620526]  do_init_module+0x56/0x1ef
Jul 17 23:36:16 zenith kernel: [  131.620551]  load_module+0x20be/0x2970
Jul 17 23:36:16 zenith kernel: [  131.620578]  ? kernel_read_file+0x16d/0x190
Jul 17 23:36:16 zenith kernel: [  131.620607]  ? __do_sys_finit_module+0xa2/0xb0
Jul 17 23:36:16 zenith kernel: [  131.620634]  __do_sys_finit_module+0xa2/0xb0
Jul 17 23:36:16 zenith kernel: [  131.620663]  do_syscall_64+0x43/0xf0

Disabling PSP prevents those SEV calls being made to PSP, which lead to lockup.

So I guess the real BUG is in the PSP firmware which now does something unexpected when SEV is being queried. I'd imagine that with previous AGESA versions PSP returned something like "error, not implemented" (TR does not support SEV) but now returns something unexpected or never returns anything.

Jul 19, 2018 3:08:06 GMT -8 | via mobile

Reply Quote | Like

	ValkyrieStar Newbie	Right now I'm Offline I have made 12 posts I joined June 2018 I've received 1 like		I currently have no messenger contact information filled out.	I currently have no social network information filled out.

Just a note:

My modded bios (F3g with 1.0.0.6) - 13265
Gigabytes latest bios (F10 1.1.0.0) - 12851

Proof: www.3dmark.com/compare/fs/15951497/fs/15935627

All parts of the test resulted in quite significantly higher scores on my modded bios. Both tests were with ram timings set manually as otherwise i cannot run 3200 c14.

Jul 25, 2018 4:18:32 GMT -8

Reply Quote | Like

pcpcpcp
Guest

Can confirm first and last core are stuck at max clock. I have the X399 Designare EX.

Jul 17, 2018 4:52:17 GMT -8 ValkyrieStar said:

My findings,

Goods;
-The stutter of F3j felt gone
-Latest agesa revision for TR1 is now implemented (same as my modded bios)
-New microcode for TR1+
-PWM/Voltage modes are back and working (yay).

Bads;
-CBS options are half gone
-memory controller options, are all over the place and nearly unfindable
-First and Last cores seem to already run at max clock, preventing the chip from idling properly, resulting in 10c higher idle temps and high idle power. Changing the Powersupply idle mode did nothing to change that. Nor did any combination of c states/coolnquiet etc

Note the agesa “1.1.0.0” is for next gen cpus (zen1+), its still only “1.0.0.6” for zen1. Ive been learning alot about AMD microcode implementation and CPUID stuff too.

Aug 2, 2018 1:26:52 GMT -8

Reply Quote | Like

	kebene Newbie	Right now I'm Offline I have made 5 posts I joined July 2018	My gender is Male	I currently have no messenger contact information filled out.	I currently have no social network information filled out.

Has anyone ever turned to support and indicated the errors?

Aug 6, 2018 10:41:07 GMT -8

Reply Quote | Like

ruffy91
Guest

Yes, they always open a RMA so you can send your board to them

Did this 2 times when primary bios was corrupted.

I then began fixing it myself using a recovery stick until i discovered that disabling ErP/EuP stops the BIOS from corrupting itself.

I sent the support my findings and they offered to open a RMA so I can send them my board...

Aug 8, 2018 7:11:35 GMT -8

Reply Quote | Like

	riphur Newbie	Right now I'm Offline I have made 2 posts I joined August 2018		I currently have no messenger contact information filled out.	I currently have no social network information filled out.

So, the F10 bios will be updated again for Threadripper 2 support?

Aug 9, 2018 8:25:38 GMT -8

Reply Quote | Like

	riphur Newbie	Right now I'm Offline I have made 2 posts I joined August 2018		I currently have no messenger contact information filled out.	I currently have no social network information filled out.

The esupport of Gigabyte reply me.
The Aorus 7 Gaming support 2990WX with F10 bios

Edited Aug 9, 2018 8:26:05 GMT -8 by riphur

Aug 9, 2018 17:41:57 GMT -8

Reply Quote | Like

	srb Pro	Right now I'm Offline I have made 254 posts I joined April 2017 I've received 11 likes	My gender is Male	I currently have no messenger contact information filled out.	I currently have no social network information filled out.

Well, for what it's worth, F10 worked ok, but the two cores stuck at 3700 had my water pump

jumping high all of the time. Got tired of the noise and went back to F3j.

In another note, I've been having fits getting 3200 memory stable. F3g would fail to boot intermittently,

and f3j refused to run higher that 2933. I saw someone's advice to try the ryzen memory calculator.

When a new version came out, I used tahiphoon burner to export my current memory data, ran it, and

manually set every selection per what it said. Low and behold, first reboot was perfect with no issues for

the first time with f3j. Have only rebooted a few times, but no issues so far. Must admit, I was skeptical,

but so far so good. This is with Corsair 4x16gb 3200 c16 dual rank samsung b die.

Aug 13, 2018 18:56:31 GMT -8

Reply Quote | Like

	teknoman117 Newbie	Right now I'm Offline I have made 1 post I joined August 2018		I currently have no messenger contact information filled out.	I currently have no social network information filled out.

I'd also like to echo the comment on the last core not downclocking.

On my Gentoo system the cpufreq driver is not registering on the last core only (the other 15 are fine). Oddly enough, its registering on the first thread of the 16th core, but not the second thread.

VFIO is also working fine on my box. Both my Linux and Windows VMs work fine.

Aug 14, 2018 12:55:33 GMT -8

Reply Quote | Like

	echto Elite	Right now I'm Offline I have made 121 posts I joined September 2017 I've received 5 likes	My gender is Male	I currently have no messenger contact information filled out.	I currently have no social network information filled out.

Aug 13, 2018 18:56:31 GMT -8 teknoman117 said:

Nice core speeds with F3j. I'm not going to touch F10.

~ # cpupower frequency-set -g performance && inxi -C

CPU: 12 core AMD Ryzen Threadripper 1920X (-MT-MCP-) cache: 6144 KB
clock speeds: max: 3500 MHz 1: 3613 MHz 2: 3781 MHz 3: 3693 MHz 4: 3708 MHz 5: 3643 MHz
6: 3660 MHz 7: 3681 MHz 8: 3691 MHz 9: 3684 MHz 10: 3691 MHz 11: 3684 MHz 12: 3628 MHz
13: 3851 MHz 14: 3835 MHz 15: 3669 MHz 16: 3677 MHz 17: 3691 MHz 18: 3722 MHz 19: 3662 MHz
20: 3692 MHz 21: 3685 MHz 22: 3692 MHz 23: 3656 MHz 24: 3666 MHz

~ # cpupower frequency-set -g powersave && inxi -C

CPU: 12 core AMD Ryzen Threadripper 1920X (-MT-MCP-) cache: 6144 KB
clock speeds: max: 3500 MHz 1: 2195 MHz 2: 2195 MHz 3: 2195 MHz 4: 2195 MHz 5: 2195 MHz
6: 2195 MHz 7: 2195 MHz 8: 2195 MHz 9: 2195 MHz 10: 2195 MHz 11: 2195 MHz 12: 2195 MHz
13: 2195 MHz 14: 2195 MHz 15: 2195 MHz 16: 2195 MHz 17: 2195 MHz 18: 2195 MHz 19: 2195 MHz
20: 2195 MHz 21: 2195 MHz 22: 2195 MHz 23: 2195 MHz 24: 2195 MHz

Edited Aug 14, 2018 12:56:35 GMT -8 by echto

Aug 31, 2018 0:55:34 GMT -8

Reply Quote | Like

SBS
Guest

Since updating to F10, the system appears to have a (quasi) cold-boot issues. The system starts, but the monitors stay blank. Only after a power down with the power button and a second start (with the power button) does the system start up normally. So far, this has occurred every other day, so I would classify this as random (as it start up normally on the first try, sometime).

No overclock anywhere all default clocks on CPU and RAM.

System:

- AMD TR 1900X
- AMD Radeon Pro WX 7100
- Gigabyte X399 Designare EX
- Kingston Value RAM 32GB ECC (4x8GB) at 2400 MHz, CL 17
- Samsung 960 Pro 1TB
- Noctua NH-U14S TR4-SP3
- Win 10 Pro 64-bit

gabeomatic wrote: Mar 3, 2021 18:28:57 GMT -8
If anyone can chime in and give me some ram OC tips in my thread I'd really appreciate it! forum.gigabyte.us/thread/11741/z390m-gaming-hynix-3600mhz-tuning

greggg wrote: Mar 3, 2021 11:03:15 GMT -8
I can't get gigabyte warranty service to respond to an RMA. Keep getting an error message saying 'server error. Please try again later. I have been trying since last November. Anyone else have this problem?

reddorki wrote: Mar 1, 2021 4:06:13 GMT -8
ricks22: Seems all AMD based motherboards of Gigabyte suffer from the same issue. Has 3 of them, AB350 Gaming 3, X470 Aorus Gaming Ultra and a X570 Aorus Ultra. None of them is hitting the rated XMP values of the memory. *