search

IGCIT / Intel-GPU-Community-Issue-Tracker-IGCIT

IGCIT is a Community-driven issue tracker for Intel GPUs.
GNU General Public License v3.0
116 stars 4 forks source link

Bug check 0x116 on wake from sleep AND unlock #313

Closed cutterjohn closed 1 year ago

cutterjohn commented 1 year ago

Checklist [README]

Application [Required]

windows 10 22H2 pro

Processor / Processor Number [Required]

7900X

Graphic Card [Required]

Intel ARC a770 LE

GPU Driver Version [Required]

Rendering API [Required]

Windows Build Number [Required]

Other Windows build number

19045.2728

Intel System Support Utility report

igcit_ssu.txt

Description and steps to reproduce [Required]

Upon wake from sleep or attempting to unlock results in a bugcheck 0x116

Device / Platform

ASROCK x670e PG Lightning

Crash dumps [Required, if applicable]

MEMORY.zip

Application / Windows logs

No response

cutterjohn commented 1 year ago

I need to add I believe since looking at the intel util output that I have TWO monitors. I do NOT see the second listed, unsurprisingly since Intel sw also thinks that rebar is NOT enabled meanwhile GPUZ hwinfo64 etc detect everything fine... (WTF does intel hire for sw? wah HuC is too teh hard on linux?! GTFO!)

I digress.

Anyways primary monitor is DP 1440p as in the intel utility. Secondary monitor is HDMI 1080p that Im using w/an Amazon basics DP->HDMI converter, which was a debugging effort as same monitor on a770 HDMI port would continuously connect/disconnect cycle on screen blank and lock. Using the adapter mitigated that in that on blank HDMI monitor now sleeps.

Additionally ran these monitors on previous eVGA 1070 FTW, which a770 replaced, and only problem was sleep w/HDMI that monitor would not go to standby, workaround power off.

Additionally ran this on 1.07 ASROCK x670e PG lightning BIOS AND currently 1.024AS02(in logss above, was motivated to finally update BIOS because of the recent overvolt situation... otherwise even 1.07 practically launch BIOS was completely stable for my setup... well hardwarewise... more or less...).

Also have run since Thursday w/4x16GB g.skill flare x5 6000 RAM. Previously ran under 1.07 BIOS with 2x16GB g.skill flare x5 at JEDEC DEFAULT 4800. Same problem in ALL cases.

It is VERY strange that the problem is limited to wake from screen, and unlock...

PSU is eVGA 1kW, and using two separate PCIe power cables.

Can't think of any extras ATM, but fire away.

cutterjohn commented 1 year ago

AH forgot one item... may or may not have DISABLED AFAICT iGPU in BIOS, but I don't recall and haven't checked today if the last set of changes, since I enabled EXPO when I updated to 1.24AS02 BIOS. Left it since it was still stable, certainly a better experience than X370...

So I will check BIOS settings and see if I did disable the iGPU on the IO chiplet...

No problems with games, applications, etc. other than as expected... so far the actual usage has been decent, although Ive not utilized video encoding, and only decoding would be by browser primarily w/youtube in firefox, and a few in chrome. Plus I guess whatever ads use in Edge as I use that for my addon free bog standard browser for sites that do NOT work well w/extensions/addons...

AV is Avast free

Lastly I did use the 4314 WHQL stable driver for a while, then decided to try the beta 4335 both behave the same for wrt to the problem report. Deeper tech differences may be existing but Ive not really done any deep dive, just stuck to checking out power connections, DP->HDMI adapter, and beta driver...

Cooling is Corsair h110i GT plus forget how many case fans, 3 or 4 IIRC w/o checking, corsair carbide 400D inverse ATX...

Gotta say that this x670e PG lightning has been MUCH better to work with than the X370 Fatal1ty Professional Gaming mobo that I had before... (had to take what I could get at Ryzen launch, ASUS X370 was self bricking and microcenter had no Taichis although Fatal1ty is Taichi with crappy aesthetics and 5Gbps ether)

(couldn't believe how easily and stably I could run RAM at rated 'speeds' and even more so when adding two stick to completely fill RAM slots... X370 was yeah best I could do w/3200 was 2933 stably... probably couldve spent time tweaking it to 3200 but after time spent I was nah good @ 2933)

cutterjohn commented 1 year ago

In case I left something out, I started a thread on IntelArc on reddit, link: https://www.reddit.com/r/IntelArc/comments/139vwo3/after_nearly_5d_of_use/

The one photo where the guy shows what looks like noise at the bottom of the screen, since I switched HDMI monitor to DP->HDMI adapter which results in my 1440p DP monitor being primary, its ENTIRE screen shows that what appears to 'noise'

EDIT: next up I need to add back second antenna for my NGFF AX210 wifi card, so I will(or may have to) reseat a770 again although Im really doubting thats the problem as in normal operation even under stress its fine, problem only shows up in sleep so Imma gonna guess BIOS/Windows/a770 firmware power management is the root cause of failure, and Im thinking with a770 flakiness connecting to monitors that yeah... the firmware probably needs some work and lets hope that its not its teh too hard HuC for linux driver for Alchemist... GTFO... too hard?! Grow up!

jpovixwm commented 1 year ago

Similar to #309 minus the cold boot part.

cutterjohn commented 1 year ago

yes second photo from https://github.com/IGCIT/Intel-GPU-Community-Issue-Tracker-IGCIT/issues/309 is what I get on main monitor via DP from wake from sleep before it bugchecks 0x116

So yes it APPEARS to be similar, and not entirely uncommon as I was able to find similar problems on reddit posts.

EDIT: when I was using HDMI the screens remained black, seems to me IF I GUESSED that it MAY be related to Intel firmware HDMI handling..

I had more commentary but decided to delete it since intel sw... yeah... not good...

Perhaps Ill try running with a single monitor and see if that changes anything...

cutterjohn commented 1 year ago

WTH are there so many game bug reports? Hell I ran X2 The Threat bench for fun and it rated 200+ FPS, looked as smooth as I ever remembered it on any GPU and I did not notice any obvious artifacts although oh man it looks dated nowadays even moreso than back at release... I BELIEVE that it's dx9, but I could be incorrect... Anyways yeah just ffelt the need for commentary on all the game bug reports, they're unimportant in the big picture...

Need stable hardware/firmware before we go yapping about games or other software, and this is nowhere clode to stable hardware/firmware yet as I fully expect sleep problem, its not power o,w, card would fail in general use, etc. and dont tell me too teh hardz... get to fscking work on the linux drivers, no too teh hardz?! those guys are weak, otherwise theyd welcome the challenge...

Karen-Intel commented 1 year ago

Hey @cutterjohn sorry for the late reply

We have a new driver release v.4639. I have done some tests (leaving my PC in idle for periods of 45 minutes > 1 hour and a half > 13 hours) and it has resumed from sleep without any issues, using this driver version.

Could you please give it a go? We also recommend to use the latest DDU version for the upgrade

Let me know if you see any improvement, if not, I'll be glad to report them otherwise.

Karen

cutterjohn commented 1 year ago

4369, yeah saw that this morning opted out of beta and let arc control do an express install, just hadn't gotten around to testing until about half an hour ago...

Just tried twice once for c. 8m and a second time for a minute or two. Same results as before, what looks to be noise on primary display... going to try again as I changed PCIe power from L1.1 to disabled(default value) and will try screen lock this time.

If I get another bugcheck I'll zip and attach mem dump. Yes in both crashes Im still getting bugcheck 0x116 from waking from sleep. give me a few minutes, then Ive got to move on to other things and any further testing will have to wait likely until tomorrow or maybe weekend.

EDIT: well screen lock for 5m or so long enough for monitors to go into standy(both) allowed me to log in.

Let me let sit in screen lock for a lengthier period, then I'll see if I can try sleep again.

EDIT2: nope, screen lock rebooted the machine again before I had to a chance to let it sit for half an hour... bugcheck 0x116

Attaching mem dump... nm just realized 7zio crapped out on first, empty archive and second found out max file size 25MB, so no dump, its about 800MB... igcit-dumps_40b03d48.zip apparently extension can be changed and key still works...

EDIT3

alright trying the helper util... see if I have space for memory dump on my gdrive... oh how nice .7z is not accepted lemme check the wiki again and well that was helpful do not rename, yet 7z cant be uploaded... wonderful.

EDIT4

yeah helper util file got added wrong place, here's a google drive link for memory dump from mem dump after bugcheck 0x116 after screen lock

https://drive.google.com/file/d/1MQbW6UDWH-rV0ZMugKyMo0R3t78mCmP3/view?usp=sharing

I'll try to add another from sleep tonight, lemme know if you cannot access it as I generally don't share files but I set it to anyone w/link....

https://drive.google.com/file/d/1MQbW6UDWH-rV0ZMugKyMo0R3t78mCmP3/view?usp=sharing

Karen-Intel commented 1 year ago

Hey @cutterjohn it is very weird you're seeing all these crashes. Thank you for the dumps, analysis in progress

Question: I saw you did use the express install but how about DDU? Have you used it? We have seen it helps in some cases where we start seeing odd behaviors.

Also, did this behavior start with driver 4314? If you use and older version do you see any difference? If you do, what is the method you follow to upgrade your driver? Is it DDU too or just the express install?

It is very important for us to isolate if this behavior started happening with a public release, specially as we have not been able to see such behavior lately in configs with 2 or even 3 monitors.

Please let me know, thanks

Karen

cutterjohn commented 1 year ago

no go, same results as expected after running DDU...

Im thinking that I need to try one monitor at a time, and Im beginning to think HDMI monitors are a problem for ARC, but that will have to wait. I'll send my $450/h bill to Intel once I isolate this problem along with billable hours.

BTW Dundas around? Someone I knew that was once on Itanic and then gfx.... go blue

I do realize that his card more or less works in operation but sleep functionality and screen lock functionality or lack thereof is yeah not good.

Here's the memory dump link of bugcheck 0x116 after sleep after DDU just like I expected... color me surprised...

https://drive.google.com/file/d/1xx8aA23B6Uod9_aeTwWBFD5RXnnS3dru/view?usp=sharing

EDIT:

a;lso let JD know that he ruint Star control for me, and that I still love the shotgun in Doom for the unwary

EDIT2:

A;lso the DP->HDMI amazon basics HDMI monitor yeah not showing its max 75Hz refresh rate although I dont care about that but it DOES INDEED point to a problem with HDMI monitors I personally think. I do NOT have enough hardware to test this with, and quite frankly the HDMI support causing crashing wouldn't surprise me in the slightest, although that could be an entire red herring...

Karen-Intel commented 1 year ago

@cutterjohn Alright. I'm already checking this behavior with the development team. In the meantime, I'll get back to the lab to do some more testing. Apologies for all the trouble, I'm sure we will find a solution soon and I will keep you posted as soon as I get any update

Karen

cutterjohn commented 1 year ago

two things, my description of lock Im thinking may have been wrong since it just bugchecked and rebooted while I was away yesterday, and I'll try running single DP and HDMI(via builtin DP-HDMI adapter which quite frankly works better given that it elimated the connect/disconnect that screen lock USED to give me although Im not sure that bugcheck/reboot is any better my apps all default to primary monitor again... BUT hey BIOS shows on my primary which is great since secondary is portrait which was a MAJOR PITA)

cutterjohn commented 1 year ago

single monitor, the 1440p monitor connected to first DP port, IIRC its DP1.4

Same problem wakes to what looks like noise(static, and by that I mean not moving not like old analog TV noise...)

as an aside this is the amazon basics dp->hdmi dongle that Im using https://www.amazon.com/dp/B08989D7RZ?ref=ppx_yo2ov_dt_b_product_details&th=1

DP cable: https://www.amazon.com/dp/B08989D7RZ?ref=ppx_yo2ov_dt_b_product_details&th=1

HDMI cable https://www.amazon.com/dp/B014I8T0YQ?ref=ppx_yo2ov_dt_b_product_details&th=1

anyways, here's the memory dump for single DP monitor wake from sleep crash (1440p)\ https://drive.google.com/file/d/1xx8aA23B6Uod9_aeTwWBFD5RXnnS3dru/view?usp=sharing

lemme know if I did not set it up correctly. Debating trying DP->HDMI monitor connection only next or HDMI->HDMI monitor next

all other serttings as noted above I have left unchanged. Personally I suspect monitor detection and/or something else perhaps in the card firmware, but hey WTF do I know... i.e. just my $0.02

EDIT:

https://www.newegg.com/evga-supernova-1000-g2-120-g2-1000-xr-1000w/p/N82E16817438010?Item=N82E16817438010

EDIT2:

was readomg what was happening in https://github.com/IGCIT/Intel-GPU-Community-Issue-Tracker-IGCIT/issues/309, so I'll add here that fast startup has been disabled since probably shortly after I originally installed windows. IIRC turning it off correct some other problems, and well, it wasnt really 'fast' anyways...

cutterjohn commented 1 year ago

screen blank(aka turn monitor(s) off after XX minutes) also does trigger bugcheck 0x116... probably happen after monitors go into standby but not immediately as I have gotten back with monitors having just blanked and already in standby but was able to resume normally. This time about 30m passed since the monitors should have blanked...

driver 4369

The computer has rebooted from a bugcheck. The bugcheck was: 0x00000116 (0xffffd585082be010, 0xfffff8007a2d6bb0, 0x0000000000000000, 0x000000000000000d). A dump was saved in: C:\WINDOWS\MEMORY.DMP. Report Id: 8bd6a74f-0011-4e06-937d-f814a19178a5.

If you'd like further memory dumps let me know othgerwise Ill just leave the two that I have up there...

cutterjohn commented 1 year ago

Hell. I compressed the memory dump from the screen lock crash. Apparently from event view happened ~4m after screen lock from my best estimate...

https://drive.google.com/file/d/1YDyz0_H4QYXSx97sgE54wUrIcUDMc5kF/view?usp=sharing

just to reiterate here are the others: https://drive.google.com/file/d/1MQbW6UDWH-rV0ZMugKyMo0R3t78mCmP3/view?usp=sharing https://drive.google.com/file/d/1WfYgKlpsR0urZz9pAEtUJCMuXORCRotl/view?usp=sharing https://drive.google.com/file/d/1xx8aA23B6Uod9_aeTwWBFD5RXnnS3dru/view?usp=sharing

not expecting much as Ive never done any debugging work with a bunch of generic hardware/software, and in yep fixed hardware/software memory dumps arent really useful, but here ya go for chaos... probably shouldve named them better and added in explicity file name dates but ah well...

EDIT:

Interesting that both screen lock and screen blank/turn off screen trigger the bug check fairly quickly(well if minutes are 'quickly'), while sleep it only happen upon wake... apparently something about changing power states would be my guess, but WTF do I know...

I would say ALL AMD, but I cannot. I would have said all AM5 but DEFINITELY I cannot as 309 report is earlier ZEN which also means that I cannot directly blame BIOS UNLESS there are overlaps in low level firmware between Zen-Zen4 which there may be. eh. I give up, good luckm best wishes, etc. et. al. long day, cold, well chilly day for late May...

cutterjohn commented 1 year ago

and here's a memory dump from this morning after resumig from sleep which I believe is not included in any of the above

https://drive.google.com/file/d/1IoZZZ3-kRxcBGqElqm7XTzdu6n3AZmw6/view?usp=sharing

lemme know when these are downloaded or no longer needed as I would prefer to cut off the share sooner than later... and/or if they're just generally not helpful...

EDIT:

Also are these problems ALL on AMD platforms? Might want to send them something if so, and/or get info from them, some undocumented/ill defined power behavior in their platforms. I know that the one guy was on zen3/5XXX but I imagine zen4/7XXX aren't all that different in some aspects, just an idea... especially given that intel software INSISTS that rebar is NOT enabled meanwhile CPUZ GPUZ hwinfo all are yeah rebar enabled and CAN detect VRAM(Steam sysinfo also has a problem with this reporting 0 VRAM), either way just tossing out bones to see if any turn up golden...

Also did not notice any way to explicitly disable IO die iGPU in ASROCK BIOS. I thought that I had seen it before, but it's tiresome digging through every undocumented BIOS section with broken and/or questionable descriptiongs... engrish FTW

I probably should check their taichi am5 bord dox to see if that one at least has some BIOS dox... my old x370 Fatal1ty Gaming Professional had a separate printed manual IIRC, but these 'cheap'(and I use the term cheap loosely as they're $100 range boards in a past life) have little documentation and none for the BIOS...

EDIT2:

I mean to say that I expect the vast majority of functionality is provided by the AMI BIOS itself and the various board mfgs just customize it to their particular configurations, which means that mostly if I can find a BIOS reference and translate it to my board in particular(engrish), as Im really suspecting some sort of power state thing here... drivers or BIOS, or both....somewhere in the chain something clearly isn't jiving and either nVidia has a workaround that Intel does not, AMD drivers are 'fixing' nvidia or Intel is doing things wrong, which wouldn't surprise me as ax210 6G drivers pickup 6G until a very certain driver version on windows 10(and apparently windows 11 has 6G driver problems as well, can I start a thread here about that?)

EDIT 3(I think, yes it is):

Can I also start a bug about HuC NOT being supported for alchemist under linux? Windows drivers guys got it to work, so why can't linux drivers guys? If answer is politics, close, if it's too teh hard, GTFO

EDIT4:

You're also blocking me with this problem in installing an nVME to 5 port SATA adapter as I do NOT want to add in additional hardware that may or may not be flakey... bottomline is that I need more SATA ports screw nVME...

Karen-Intel commented 1 year ago

@cutterjohn thank you for all the information We're currently working on the issue isolation on systems similar to yours+ providing data to the dev team thanks to all the dumps you sent. FYI the fix may take from 3 to 6 months to be released so please leave this post open in case I have updated info If I need any additional information I will get back to you

Thank you for your help! Karen

cutterjohn commented 1 year ago

@karen am I able to remove the memory dump links? i.e. have you or team downloaded them or do not need them etc? as I am NOT comfortable leaving them up as well there's no really good way to encrypt or password them...

If I do not see a response by Wednesday I am removing the files linked, although thats probably already too late... probably too late a picosecond after posting them...

EDIT:

Additionally I AM using the ASROCK supplied chipset drivers for AM5 which are older than amd generic chipset drivers for AM5, so Im going to try those as I almost always prefer ODM drivers over whatever the OEM provides given that I really do not think that they modify drivers in any meaningful way if at all.

EDIT2:

I am disabling linked memory dumps, either anyone interested has them, or can ask for them again in the fitire...

Karen-Intel commented 1 year ago

Hey @cutterjohn I have downloaded them, no worries You can remove the links if you want

Karen

cutterjohn commented 1 year ago

Karen, they're disabled as of about an hour ago

Files are in tact if needed for now.

EDIT:

Best of luck as other than this problem, these cards seem to work fairly well although I have NOT exercised compute yet, and I am REALLY REALLY HOPING that linux support will be COMPLETE. I've gotten dxvk intro by virtue of purchasing a steam deck, unfortunately its parked for a while, but it's a nice not quite so little device for portable gaming, plenty of input, but that also is it's own problem remembering buttons etc for each game... plus some games, yeah, just not working well... sometimes games that should work well but coming from other sources than steam just not working as well or at all... anyways I digress...

Thanks for the help, and and yeah I'll be tracking driver updates as lack of sleep/screen lock/display sleep is a problem.

Karen-Intel commented 1 year ago

Hey @cutterjohn yeah I had downloaded them already :D And we really really appreciate all the evidence you shared with us, be sure it will help many other users and it is helping us already to have a better product, however it may take a while for the fix to be implemented. Also, if you have spotted the OEM's recommended driver version that works well, could you share it with me? If we have that info to identify the regression it would be great! Thanks!

Talk to you soon!

Karen

cutterjohn commented 1 year ago

Karen, Ive noticed that the ASROCK chipset drivers seem to be behind AMD's latests generic drivers, although I truly suspoect that there are zero differences, but will try them. Hell even on video cards I generally use nVidia/ATI(AMD)/(and now) Intel(even though my card is Intel drivers.

Ive never noticed any difference betwee AIB/OEM drivers other than the graohics of the installer, and that they lag behind ODM drivers, sometimes by quite a bit,

That said Im not expecting much but Ill try AMD latest chipset etc. drivers versus ASROCK and explore any BIOS options as time allows as I kind of expect that it may be a BIOS option is this problem is restricted to AMD platform.

BTW also noticed today dxdiag correctly detects VRAM, leaving only steam and Intel to not correctly detect/display VRAM so Im also left wondering about that given thatr cpuz, gpuz, hwinfo all detect rebar, etc. enabled, and also detect VRAM meanwhile only steam and Intel's own software fails at that, so Im not entirely certain that its a BIOS setting, but maybe even perhaps intel driver not properly detecting card and features, but Im just guessing at this point, although intel sw not detecting rebar and VRAM is concerning to me...

Karen, thank you for kind assistance in this matter and I look forward to this problem being resolved via driver updates than hardware failure, either of which are still in flight...

cutterjohn commented 1 year ago

Karen? If I sacrifice my older desktop, core i7-3930k to the silicon gods, does that help move things along mor quickly? (was my primary machine until I went back to AMD with ryzen 1800X)... and it's still ready to run(sort of) as a backup... but I can stoke the fires and burn the sacrifice to the spikey gods...

Karen-Intel commented 1 year ago

Hey @cutterjohn there's no need to do so. Our driver cadence release is kinda fixed already so we're going to have to wait a while for it to be released, like I mentioned before. I think (so far) we have everything we need to follow this up. FYI here's a temporary WA for this but once the fix has been released to the public I will let you know in this thread

@cutterjohn thank you for all the information We're currently working on the issue isolation on systems similar to yours+ providing data to the dev team thanks to all the dumps you sent. FYI the fix may take from 3 to 6 months to be released so please leave this post open in case I have updated info If I need any additional information I will get back to you

Thank you for your help! Karen

Karen

cutterjohn commented 1 year ago

Karen, indeterminate result here. I decided to 'upgrade' to windows 11(ridicuously) after a couple pops. Install seemed to go fine, so I decided eh lemme try sleep.

Left it for minimum 6m, machine indicated sleep and both monitors went into standy before leaving it, came back, punched a keybord key to wake it up et voila login screen...

...and then I was introduced to the 'new' macroshaft, sorry your credntials phail(mind you I purposefully reset macroshaft login crap and was using a LOCAL account), so that was as far as I got, and then proceeded to waste half an hour re-imaging the larger SSD from backed up SSD again as I did about a month ago...

Addenda which perhaps may be important, I DID UNINSTALL virtualbox while doing the windows 11 'upgrade'(total crap UI BTW reminds me of windows 8 guess those proponents are still around), so yeah take this with a grain of salt as the re-imaged drive yes still does have virtualbox on it, but Im really suspecting microsoft reverting to their old ways, as windows 11 nagged every which way to sunday to log into their account...

EDIT:

...and I found that my secondary HDMI(OR VGA but Im using that amazon basics DP->HDMI since I had other problems otherswise) SHOWS ALL of it's display rates on Windows 11, thank M$ for being yeah...

I did the 'test' 'upgrade' but running the separate 'application' as the 'integrated' update never offered it to me other than advertising that my machine was compatible... also on a side note I VAGUELY recall reading something about AMD TPM maybe being buggy in windows 11, so I'm going to do more research before I try w11x64pro again, if I ever do as I really did not like the UI although doing some quick research while waiting for drive to be re-imaged I see there are several ways to mitigate account login nagging, which is a non-starter for me.

cutterjohn commented 1 year ago

@karen I have installed a 1TB nVME card and discovered several things, confirming some from the previous post.

1 DO NOT login to a macroshaft account BEFORE running the windows 11 update as that is what apparently caused all my problems logging back in. Why? Not sure, ask macroshaft.

2 yes indeed, screen lock, and wake from sleep do INDEED WORK under windows 11. Why? Apparently again ask macroshaft, along with WTF do ancient beta drivers for the AXE210 show 6g nets meanwhile recent drivers? Nah, there arte no 6g nets...

3 someone high up from Intel needs to have a chat with Macroshaft and set them straight.

4 Windows 11 UI is still as ass as my first experience, and they've gotten even more obnoxious with their data snooping online account login requests, which is incredibly annoying(see #3 where an adult can have a word with macroshaft)

I still have two backup copies of w10x64pro where this was failing and since mirroring to the nvme take ~15m I'd be willing to try that out again later, but yes this is macroshaft fscking with people needlessly, obviously.

Last item, Arc Control now recognizes rebar is enabled on my x670e PG Lighting/7900X machine. Since I think it was 4499 beta. Have not updated to WHQL driver which was SURPRISINGLY offered to me as back in May I had to opt out of beta before being offered the WHQL version which was 4411? Not going to bother to check.

Have a great summer, and let me know if you have a new driver to test on w10x64pro, just don't expect snappy results as I id NOT test w10x64pro behavior on the nvme but went straight to trying w11 'upgrade' again.. I LOATHE it so much so debating registry hacking v. yeah buying a shareware product to fix it just like vista/8/8.1 all over again. You'd think that they'd have learnt their lesson, but nope...

EDIT: and yes I intentionally call them 'macro' which is greek for what? and shaft? which is common for what again? combine them and yeah, there's your answer... adult supervision required somewhere, sooner the better.

Karen-Intel commented 1 year ago

Hahaha @cutterjohn your post made my day.

No worries, I think we can test Win10 on our own, actually, like you, we have found out many issues get fixed with the newest OS so that's probably the best to do now (or for the time being)

Let me know if you are ok with closing the thread for now and reopen if we see similar issues in Win11

Have a wonderful summer too

Karen

cutterjohn commented 1 year ago

@Karen-Intel Last point, I left it in sleep mode overnight and was able to successfully wake up w/no bugcheck.

Additionally I left it in screen lock for over an hour today, which under w10 used to also produce the bugcheck. w11pro no problems, gave me the login screen and away I went.

ax210 driver bug has got to be MS's doing as all the registry entries are present on latest drivers BUT nada 6g nets on w10pro. Install the old BETA (SIGNED) driver and hey LOOK! 6g nets! BUT it IS a flakey fix as it tends to require re-installs occasionally, but I guess to be expected.

The w10x64pro backup copies are really for my own failsafe fallback to re-image the nvme drive in case things go south as Im still stability testing w11pro, the hard way(while actually using it). Nice thing is SATA SSD to nvme takes 15m to re-image, while what I was doing before was replace SATA SSD, connect target SSD to a USB<->SATA adapter which took around 45m to re-image...

Anuways yes I'd be happy to close it, as I believe that the other open bug is probably related... Have a good summer yourself.

Karen-Intel commented 1 year ago

@IGCIT can you help us close this one?

Issue went away after upgrading to Win11.

Thank you!

Karen